National Academies Press: OpenBook

Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work (2005)

Chapter: Part II - Review of References Related to Public Comments

« Previous: Part I - General Literature Review
Page 115
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 115
Page 116
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 116
Page 117
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 117
Page 118
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 118
Page 119
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 119
Page 120
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 120
Page 121
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 121
Page 122
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 122
Page 123
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 123
Page 124
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 124
Page 125
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 125
Page 126
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 126
Page 127
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 127
Page 128
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 128
Page 129
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 129
Page 130
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 130
Page 131
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 131
Page 132
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 132
Page 133
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 133
Page 134
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 134
Page 135
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 135
Page 136
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 136
Page 137
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 137
Page 138
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 138
Page 139
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 139
Page 140
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 140
Page 141
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 141
Page 142
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 142
Page 143
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 143
Page 144
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 144
Page 145
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 145
Page 146
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 146
Page 147
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 147
Page 148
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 148
Page 149
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 149
Page 150
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 150
Page 151
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 151
Page 152
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 152
Page 153
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 153
Page 154
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 154
Page 155
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 155
Page 156
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 156
Page 157
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 157
Page 158
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 158
Page 159
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 159
Page 160
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 160
Page 161
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 161
Page 162
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 162
Page 163
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 163
Page 164
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 164
Page 165
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 165
Page 166
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 166
Page 167
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 167
Page 168
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 168
Page 169
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 169
Page 170
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 170
Page 171
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 171
Page 172
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 172
Page 173
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 173
Page 174
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 174
Page 175
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 175
Page 176
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 176
Page 177
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 177
Page 178
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 178
Page 179
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 179
Page 180
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 180
Page 181
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 181
Page 182
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 182
Page 183
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 183
Page 184
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 184
Page 185
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 185
Page 186
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 186
Page 187
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 187
Page 188
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 188
Page 189
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 189
Page 190
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 190
Page 191
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 191
Page 192
Suggested Citation:"Part II - Review of References Related to Public Comments." National Academies of Sciences, Engineering, and Medicine. 2005. Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work. Washington, DC: The National Academies Press. doi: 10.17226/13839.
×
Page 192

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Part II Review of References Related to Public Comments

Since 1995, the U.S. Department of Transportation’s Federal Motor Carrier Safety Administration (FMCSA) [formerly FHWA’s Office of Motor Carriers] actively con- ducted a concentrated program of research, study, and industry outreach education on commercial driver alertness, fatigue, health and wellness. There was much open pub- lic discussion, deliberation, and negotiation over the public rulemaking process from 1996 to 2003. In May 2003, FMCSA issued new Hours of Service (HOS) rules for com- mercial truck drivers with a planned implementation date in January 2004. After sub- stantial amounts of training and preparation by government and the trucking industry, those new HOS rules went into effect January 4, 2004. The January 2004 revised HOS rules extended allowable driving time to 11 hours and reduced overall driver work hours to 14 before requiring a 10-hour break. The old HOS rules limited commercial truck driving to 10 hours and allowed drivers to work 15 hours before taking a mandatory 8-hour break. Public Citizen challenged those HOS rules in a lawsuit, alleging that the new HOS did not properly account for commercial driver health concerns. Responding to that lawsuit, in July 2004, the U.S. Court of Appeals for the Washington District of Colum- bia Circuit ruled that DOT’s FMCSA did not consider truck drivers’ health in the revised HOS rules. FMCSA requested the federal court to stay its order and to keep the current, revised HOS rules in effect until FMCSA could present its case again or could prepare a new set of HOS rules. As one part of its efforts to reply to the Court of Appeals ruling on HOS, the FMCSA requested independent technical assistance from a third-party research team to sum- marize the scientific and technical literature on commercial vehicle operator health, wellness, fatigue, and performance, as they might be related to the hours a person works or to the structure of the work schedule (e.g., on-duty/off-duty cycles, sleep time, etc.). SUMMARY LITERATURE REVIEW ON HEALTH AND FATIGUE ISSUES ASSOCIATED WITH COMMERCIAL MOTOR VEHICLE DRIVER HOURS OF SERVICE Part II: Review of References Related to Public Comments

MaineWay Services was assigned the task of literature review by the Transportation Research Board (TRB). This synthesis was completed and submitted to TRB for review and publication. OBJECTIVE OF SUPPLEMENTAL REVIEW On January 24, 2005, FMCSA published in the Federal Register (70 FR 3339) a Notice of Proposed Rulemaking (NPRM) regarding HOS of commercial motor vehi- cle (CMV) drivers. In that NPRM, FMCSA announced its intention to review and reconsider the regulations on HOS of drivers published on April 28, 2003, and amended on September 30, 2003. In the docket to this January 24, 2005, NPRM, FMCSA re- filed the same Regulatory Impact Analysis (RIA), or comprehensive analysis of eco- nomic benefits and costs of the proposed rule, as was filed in the docket for the April 2003 final rule. Within the responses to this request for comments on this announcement were some 266 references to studies, articles, and literature relating to the health and fatigue effects of the HOS regulations. To assess the significance and relevance of these references, the MaineWay Services research team was asked to review the studies, articles, and lit- erature and provide analyses of those references it deemed relevant to the health and fatigue effects of the existing regulations. RESEARCH TEAM The research team consisted of the following members: Health Effects Panel Peter Orris, MD, MPH • Professor of Occupational and Environmental Health Sciences, University of Illi- nois School of Public Health, Cook County Hospital • Director of Occupational Health Services Institute, Great Lakes Center for Occu- pational and Environmental Safety and Health, University of Illinois • Chief of Service, Occupational and Environmental Medicine, University of Illi- nois at Chicago Hospital and Medical Center • President, Medical Staff, Cook County Hospital • Secretary/Treasurer, Journal of Public Health Policy • Member of Medical Advisory Committee of International Brotherhood of Teamsters • Author of multiple publications relating to public health topics and reviewer and participant in editorial boards of a range of professional journals related to pub- lic health topics Susan Buchanan, MD, MPH • Interim Program Director, Occupational Medicine Residency, UIC College of Medicine • Author, several publications relating to Occupational Health • Reviewer, American Journal of Industrial Medicine, 2004 120

121 Health Effects Panel Members • Leslie Stayner, PhD. – Professor and Director, Epidemiology and Biostatistics, University of Illinois, Chicago School of Public Health – Served as Chief of Risk Evaluation branch, National Institute for Occupational Safety and Health, Education and Information Division, and in several other career positions relating to risk evaluation – Contributing Editor to Journal of Industrial Medicine, and involved in a wide range of professional activities relating to industrial health • Eric Garshick, MD, MOH – Assistant Professor of Medicine, VA Boston Healthcare System, Channing Lab- oratory, Brigham and Women’s Hospital, Harvard Medical School – Served as Advisor, World Health Organization; International Program on Chem- ical Safety, Environmental Health Criteria for Diesel Fuel and Exhaust Emis- sions, Geneva Switzerland – Served as Consultant, U.S. EPA Science Advisory Board, Clean Air Scientific Advisory Committee Diesel Emissions Health Document • William Marras, PhD. – Co-Director, Institute of Ergonomics, Ohio State University – Professor, Department of Physical Medicine, Biomedical Engineering Center, Ohio State University – Associate Editor, Human Factors • Natalie Hartenbaum, MD, MPH – President and Chief Medical Officer of OccuMedix, Inc. – Adjunct Assistant Professor of Emergency Medicine/Occupational Medicine at the University of Pennsylvania – Editor-in-Chief of CDME (Commercial Driver Medical Examiner) Review Fatigue Effects Panel Alison Smiley, PhD. • President of Human Factors North, Inc., a Toronto-based human factors and engi- neering consulting company; and a Canadian Certified Professional Ergonomist (CCPE) • 30 years experience in measurement of human performance, and human factors consulting, specializing in driver behavior, transportation safety, and shift work • Senior specialist in assessment of work-rest schedules, shift work, hours of work and worker rest for transportation industries (railways, coast guard and marine vessels, trucking, etc.) and for nuclear power plant and manufacturing operations • Project manager for several Transport Canada projects involving literature review and development of experimental protocols related to fatigue and minimum recov- ery periods for CMV drivers • Forensic consultant with expertise on car and truck driver fatigue and shift- scheduling issues • Consultant to both Canadian and U.S. governing bodies on trucking industry hours of service regulations

Dianne Davis, M.Eng. • Associate Consultant, Human Factors North, Inc. • Over 10 years experience conducting human factors analyses in a variety of dif- ferent domains such as the safety of driver examination tests, the study of fatigue and truck driving, way-finding, and the design of medical mobile devices and online shipping tools ORGANIZATION AND PRESENTATION OF PART II: REVIEW OF REFERENCES RELATED TO PUBLIC COMMENTS This supplemental references review is presented in two sections: Health Effects and Fatigue Effects. The Health Effects section has the following subsections: • Executive Summary • Process and Methodology • Review of References • Reference Summaries The Fatigue Effects section has the following subsections: • Selection Criteria • Executive Summary • Reference Summaries Note: The sections are presented in a format prescribed by the FMCSA HOS Regu- latory Review Team subsequent to the publication of the initial statement of work. 122

123 HEALTH EFFECTS REFERENCES REVIEW EXECUTIVE SUMMARY The purpose of Part I of this synthesis was to provide infor- mation that clearly discussed in a scientific, experimental, qualitative, and quantitative way the relationship between the hours a person works, drives, and the structure of the work schedule (on-duty/off-duty cycles, time on task, especially time in continuous driving, sleep time, etc.) and the impact on the health of truck drivers. The issues of cardiovascular disease, cancer, musculoskeletal disorders, gastrointestinal disorders, reproductive health effects, and the effects of vibration and noise were reviewed in this part. Part II covers additional articles on lung cancer and the health effects of short sleep duration. In Part II, the research team reviewed additional available information provided with respect to data relevant to driver health potentially associated with the 2003 HOS Regulations. FMCSA provided a list of 266 references included in com- ments on the proposed rules. After review of the titles or abstracts, 256 were judged to cover matters that were either previously reviewed in Part I or were outside the scope of Part II, which focused solely on health effects, and not on the fatigue effects discussed in a later section of this review. The research team focused on references which utilized a scien- tific approach in analyzing new data. Ten references were chosen and summarized based on the relevance of the article to the health effects of the changes in HOS for truck drivers. This review excluded articles that evaluated the health impact of shift work. As stated in Part I, the literature indicates that lung cancer is likely caused by exposures to diesel exhaust and the longer that exposure lasts the more likely it is that a cancer will develop. Additional evidence for this association was found in the additional material reviewed yet tempered with respect to truck drivers with the understanding that substantial expo- sure misclassification may have occurred in job-derived exposure estimates. An additional article (not included in Part I) on the rela- tionship of whole body vibration (WBV) and self-reported low back pain again identified an association, although the association was weak. Finally, an additional study (not included in Part I) sug- gested that, in contrast to prior results, short sleep duration, as low as 4.5 hours, may not affect mortality. Yet, the litera- ture on sleep duration and health concludes that there is suf- ficient reason to be concerned about a possible link between long hours and physical health outcomes such as cardiovas- cular disease and diabetes. Several investigations of the acute biochemical effects of acute sleep restriction supported this concern due to the increase in risk factors for diabetes and obesity and an additional study noted that self-reported daily short sleep duration was associated with an increase in the rate of diabetes. PROCESS AND METHODOLOGY Literature Search Source and Terms FMCSA provided a list of 266 references cited in public comments on the advanced NPRM. Selection Criteria The original list provided by the FMCSA contained 266 articles. Those not deemed appropriate by title (not relating to health effects) were eliminated, although the fatigue sec- tion of Part II includes those references relating to fatigue effects. The preliminary draft list then included 73 articles. Those not pertaining to health effects of truck driving or a related exposure, duration of work shift or duration of sleep were then removed from the list. Several citations taken from lay press websites and those already reviewed in Part I were also eliminated. Articles on shift work were likewise not included unless they specifically addressed short sleep duration. The remaining ten references were summarized by a pri- mary reviewer based on the validity of the methodology, the relevance of the studied population to truck driving, and the quality of the statistical analysis of health outcomes. These were abstracted and summarized in the format prescribed by FMCSA. The three end points covered by these new refer- ences were (1) lung cancer, (2) WBV, and (3) the health effects of short sleep duration. REVIEW OF REFERENCES The summaries are divided into the following subsections: lung cancer, WBV, and health effects of short sleep duration.

Lung Cancer Three articles addressed the association of lung cancer and exposure to diesel exhaust. A case-control study using a survey of the general population in Sweden (Gustavsson et al. 2000) found a positive association as well as a dose- response relationship. Two additional reports were reviewed that evaluated the strength of the methodologies of existing studies on the rela- tionship of truck driving to lung cancer. An extensive feasi- bility study conducted by the Health Effects Institute Diesel Epidemiology Working Group (Garshick et al. 2002) demon- strates with its pilot data that long-haul drivers have a low exposure to diesel exhaust compared with dock workers and suggests that new diesel technology may be an explanation. They also note that estimating exposure based on job alone, as is done in many studies, may give highly misclassified expo- sure assignments. Similarly, in an analysis of the risk assess- ment data by the same panel (HEI 1999), recommendations were made regarding Quantitative Risk Assessment (QRA) based on the current studies available. The research team con- cluded that the railroad worker cohort study (Garshick 1988, a secondary reference in Part I) has limited utility for quali- tative risk assessment, but that the Steenland article (1998) also reviewed for Part I is potentially useful based on a QRA approach. The report, Diesel and Health in America: the Lingering Threat, from the Clean Air Task Force in February 2005 mod- eled the health effects of the current diesel fleet on the U.S. population as a whole (Schneider and Hill 2005). Despite the necessary assumptions inherent in this type of report, this study concludes that increasing the weekly working hours of drivers is likely to increase their risk of cancer with the under- standing that new trucks, through changes in engine and cab design, will mitigate this affect due to reduction in exposure. WBV Effects In addition to the articles reviewed for Part I on the effects of WBV, one article reviewed for Part II (Palmer et al. 2003) showed a weak association between estimated vibration dose and low back pain. However, this was a survey of the general population and vibration dose was estimated from self report of vehicle use; both factors may explain the weaker associa- 124 tion compared with previously reviewed studies of exposed workers. Health Effects of Short Sleep Duration A large study using data from 1.1 million questionnaires (Kripke et al.) investigated the mortality associated with short sleep duration and found that the best survival was experi- enced by those reporting a usual sleep duration of 7 hours. When controlled for co-variants, the mortality risk associated with short sleep all but disappeared, suggesting in contrast to prior studies that short sleep duration, as low as 4.5 hours, may not affect mortality. A review of the literature on sleep duration and health (Alvarez 2004) concludes that the metabolic changes asso- ciated with short-term sleep deprivation may provide a potential mechanism for the effects of long-term sleep depri- vation on health. An extensive review of the literature on long working hours by the United Kingdom Safety Labora- tory (White 2003) only briefly includes the health effects of long hours. It concludes that there is sufficient reason to be concerned about a possible link between long hours and physical health outcomes such as cardiovascular disease and diabetes. Three studies addressed the metabolic changes which may result from acute sleep restriction. A controlled study of sleep restriction in healthy volunteers (Spiegel 1999) showed car- bohydrate intolerance, and cortisol and thyrotropin abnor- malities similar to those seen in patients with type 2 diabetes and aging. On the other hand, a survey study of sleep dura- tion with a sleep-lab component (Taheri 2004) did not find an association between sleep duration and insulin or glucose. However, two hormones which control appetite were both found to be abnormal in sleep-deprived participants. Finally, large survey study of self-reported sleep duration and incidence of diabetes (Ayas et al. 2003), found a signifi- cantly elevated risk of diabetes in those sleeping ≤5 hours per night which disappeared when elevated Body Mass Index (BMI) (a known risk factor for diabetes) was not controlled for. As BMI or obesity may itself be associated with short sleep duration based on the studies of acute sleep deprivation, this may not be an appropriate control as the effect may be mediating rather than confounding.

125 Susan Buchanan The Impact of Daily Sleep Duration on Health: A Review of the Literature. Alvarez, G.G., Ayas, N.T. Prog Cardiovasc Nurs 2004;19(2): 56–59. A healthy amount of sleep is paramount to leading a healthy and productive lifestyle. Although chronic sleep loss is common in today’s society, many people are unaware of the potential adverse health effects of habitual sleep restriction. Under strict experimental con- ditions, short-term restriction of sleep results in a variety of adverse physiologic effects, including hypertension, activation of the sympathetic nervous system, impairment of glu- cose control, and increased inflammation. A variety of epidemiologic studies have also sug- gested an association between self-reported sleep duration and long-term health. Individu- als who report both an increased (>8 h/d) or reduced (<7 h/d) sleep duration are at modestly increased risk of all-cause mortality, cardiovascular disease, and developing symptomatic diabetes. Although the data are not definitive, these studies suggest that sleep should not be considered a luxury, but an important component of a healthful lifestyle. This was not a scientific study. It was a review article, but not a methodic literature review. To discuss the physiologic effects of short-term and long-term sleep restriction and exam- ine the relationship between sleep restriction or sleep excess and a variety of health out- comes such as all-cause mortality, coronary heart disease (CHD), and diabetes. Driver Health (General) On short-term sleep restriction: mentions Spiegel’s article (reviewed for Part II) showing impaired glucose tolerance and reduced leptin, etc., with acute sleep deprivation as well as another study by Meier-Ewert showing increased C-reactive protein with 4.2 hours of sleep per night. “Although the magnitude of the physiologic changes found in these short-term studies was modest, they provide a potential mechanism whereby long-term sleep restriction may affect long-term health. However, the experiments described above predominantly studied young, healthy subjects and it is not known if similar changes would be found in older subjects.” (There were no page numbers in e-file.) On long-term sleep restriction: mentions the Nurses Health Study by Ayas showing a sig- nificantly increased risk of cardiovascular disease with ≤5 hours of sleep per night. Another article by Ayas using the same cohort (reviewed for Part II) which examined the risk of dia- betes, was noted for its lack of significant association after adjusting for multiple co-variates. Also noted was the fact that these studies were done on women only so they should not apply to men. They were also criticized for using self-reports of sleep duration and for pos- sible residual confounding. Longer sleepers: several studies note increased mortality, not useful for our purposes. This article is a review of the literature on health effects of sleep deprivation. It supports the findings of other studies reviewed for Part II, specifically the Spiegel and Ayas articles. REFERENCE SUMMARIES** Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Findings Directly Related to HOS (include page references): Reviewer’s Notes: **Summaries are provided in the order submitted by the researchers.

Susan Buchanan A Prospective Study of Self-Reported Sleep Duration and Incident Diabetes in Women. Najib T. Ayas, MD, David P. White, MD, Wael K. Al-Delaimy, MD, PhD., Joann E. Manson, MD, DRPH, Meir J. Stampfer, MD, DRPH, Frank E. Speizer, MD, Sanjay Patel, MD, Frank B. Hu, MD, PhD. Diabetes Care 26:380–384, 2003. Short-term sleep restriction results in impaired glucose tolerance. To test whether habitu- ally short sleep duration increases the risk of developing diabetes, we studied a cohort of 70,026 women enrolled in the Nurses Health Study, without diabetes at baseline, and who responded to a question about daily sleep duration in 1986. Subjects were followed until 1996 for the diagnosis of diabetes (1,969 cases). Long and short sleep durations were asso- ciated with an increased risk of diabetes diagnosis. The relative risks (RRs) for short (slept <5 h per day) and long (slept >9 h per day) sleepers were 1.57 (95% CI 1.28–1.92) and 1.47 (1.19–1.80), respectively. After adjustment for BMI and a variety of confounders, the RR was not significantly increased for short sleepers (1.18 [0.96–1.44]) but remained modestly increased for long sleepers (1.29 [1.05–0.59]). We then performed a similar analysis using only symptomatic cases (n = 1,187). Adjusted RRs for symptomatic diabetes were mod- estly elevated in both short (1.34 [1.04–1.72]) and long (1.35 [1.04–1.75]) sleepers. Our data suggest that the association between a reduced self-reported sleep duration and dia- betes diagnosis could be due to confounding by BMI or sleep restriction may mediate its effects on diabetes through weight gain. Sleep restriction may be an independent risk fac- tor for developing symptomatic diabetes. The Nurses Health Study Cohort was used. In 1986, subjects were asked about total hours of sleep. Between 1986 and 1996, the incidence of diabetes was assessed with a supple- mentary questionnaire (to confirm the diagnosis). Relative risk for diabetes was calculated in the different categories of exposure (hours of sleep per night), adjusted for age. Multi- variate analysis included age, smoking status, hypertension, alcohol consumption, physical activity, menopausal status, family history of diabetes, and hyperlipidemia. BMI was adjusted for in secondary analysis. To assess the relationship between self-reported sleep duration and the diagnosis of diabetes. 70,026 nurses in the United States Nurses Old criterion for diabetes was used (FBS < 140). Current diagnostic criteria are stricter and would have resulted in more cases of diabetes, if used. Driver Health (General) Those who slept ≤5 or 6 hours per night showed significantly elevated relative risks (adjusted for age.) After adjusting for covariates, results remained significant but attenu- ated. After adjustment for BMI, the short sleepers were no longer at risk. Results were modest and disappeared when adjusted for BMI, a known risk factor for dia- betes. A modest study (due to the results, not the design). 126 Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

127 Susan Buchanan Quantitative Risk Assessment of Lung Cancer Risk from Diesel Exhaust Exposure in the U.S. Trucking Industry: A Feasibility Study. Garshick, E., Smith, T.J., Laden, F. Health Effects Institute Diesel Epidemiology Working Group 2002. pp. 115–149. The objectives of this study were to test the feasibility of identifying a population exposed to diesel exhaust in which small to moderate excesses in lung cancer could be estimated with reasonable precision and to develop a strategy to provide quantitative estimates of cur- rent and past exposures. We chose to assess the feasibility of designing an epidemiologic study based in the U.S. trucking industry. With cooperation of the Motor Freight Carriers Association (the trucking industry trade association) and the International Brotherhood of Teamsters (Teamsters Union), 4 large unionized national trucking companies agreed to par- ticipate in the feasibility study. We obtained samples of personnel, payroll, and truck inven- tory records and interviewed long-term employees, record managers, and senior manage- ment. The types of retirement records available from 2 large Teamsters union pension funds were determined. A pilot questionnaire was mailed to 526 employees at one terminal to obtain information on smoking behavior and job history. Short-term variations in exposure were assessed by measurement of air quality in truck cabs, loading docks, and yards in 2 large urban terminals and 4 small rural terminals. Measurements included elemental car- bon (EC*) and organic carbon (OC) particles 2.5 µm or smaller in diameter, and respirable particulate clusters 2.5 µm or smaller in aerodynamic diameter (PM2.5). The OC collected in high-volume area samples was further analyzed to assess the extent to which particles collected in the loading dock area came from diesel vehicles. Past studies and outside expo- sure databases were reviewed. Major determinants of exposure included an individual’s job title, terminal size, and termi- nal location. A gradient of exposure was identified. Smoking behavior did not differ between long-haul drivers and other workers. In 1985, the number of male union workers at the 4 companies whose job history could be characterized was 55,750, and in 1999, it was 72,666. A retrospective cohort study of workers from the cooperating trucking com- panies and the Teamsters union alive in 1985 with mortality assessed through 2000 would have a greater than an 80% power to detect a relative risk of lung cancer of 1.25 to 1.29 attributable to diesel exposure. Thus, epidemiologic studies can be designed to study the occurrence of lung cancer and to estimate past exposures to diesel exhaust among employ- ees of the trucking industry. The four largest unionized trucking companies in the U.S. were used to assess the quality of records of personnel, equipment, payroll, operations, and retirement in order to recon- struct job and exposure histories. Exposure was assessed at 2 large terminals in an urban area and at 4 small terminals in rural New England. Personnel database assessment included work history records and Teamsters Union employment records. Assessment of company data of truck fleets included inventory and maintenance records, vehicle purchasing, and retirement. Terminals were assessed for design, operations, and location (urban, rural). Field tests, including both personal and fixed location sampling, were conducted to obtain some limited data on current exposures to diesel exhaust. To determine the feasibility of designing an epidemiologic study to assess lung cancer risk from long-term exposure to diesel exhaust for the purpose of hazard identification and risk assessment based on exposure estimates. Thirty workers in the pilot exposure measurement part of the study. Four trucking compa- nies with a total of 72,666 workers were included in the feasibility assessment. 526 employ- ees participated in the pilot questionnaire part of the study. Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size:

Trucking industry, both short and long haul This is a feasibility study only. The data collected were for use in estimating feasibility and not for investigating the association between exposure to diesel exhaust and lung cancer. Working Conditions (Environmental except sleeper berth) Long-haul drivers had the lowest and dock workers had the highest exposure levels to organic carbon, PM2.5 and elemental carbon (n = 30). However, when looking at correla- tion between the levels, “Assuming that EC represents diesel exposure in these settings, these results show that PM 2.5 would be a poor marker of diesel exposure and the OC would add little information (although the sampling numbers are small.)” (p. 131) Regarding area samples: “These data show that even in rural locations the existence of diesel sources near the terminal can increase exposure levels.” (p. 132) Comparison to NIOSH Health Hazard Evaluation exposure data (Zaebst 1989): “Exposures of long haul drivers were much lower in our Atlanta data than in the Zaebst data, which may reflect the effect of new diesel technology.” (p. 134) Regarding estimation of previous exposures: “For this cohort, the results showed that job alone might give highly misclassified exposure assignments. This definition should be refined by including terminal characteristics and formulating job-terminal exposure categories for historical periods.” (p. 138) Power calculations were presented for 3 exposure scenarios and demonstrated the feasibil- ity of detecting a significant trend across exposure groups with relative risks of 1.27–1.33 at 80% power. This is a feasibility study only. The data collected were for use in estimating feasibility and not for investigating the association between exposure to diesel exhaust and lung cancer. 128 Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

129 Susan Buchanan Occupational Exposure and Lung Cancer risk: A Population-Based Case-Referent Study in Sweden. Am J Epidemiol. 2000 Jul 1;152(1):32–40. Gustavsson, P., Jakobsson, R., Nyberg, F., Pershagen, G., Jarup, L., Scheele, P. This case-referent study investigated the lung cancer risk from occupational exposure to diesel exhaust; mixed motor exhaust; other combustion products, asbestos, metals, oil mist, and welding fumes. All cases of lung cancer in males aged 40 to 75 years old among stable residents of Stockholm County, Sweden, were identified from 1985 to 1990. Refer- ents were selected as a stratified (age, inclusion year) random sample. Information on life- time occupational history, residency, and tobacco smoking was obtained from the study subjects or from next of kin. Response rates of 87% and 85% resulted in 1,042 cases and 2,364 referents, respectively. Occupational exposures were assessed by an occupational hygienist who coded the intensity and probability of each exposure. Risk estimates were adjusted for tobacco smoking, other occupational exposures, residential radon, and envi- ronmental exposure to traffic-related air pollution. For the highest quartile of cumulative exposure versus no exposure, the relative risk was 1.63 (95% confidence interval (CI): 1.14, 2.33) for diesel exhaust, 1.60 (95% CI: 1.09, 2.34) for combustion products, and 1.68 (95% CI: 1.15, 2.46) for asbestos. Dose-response analyses indicated an increase in lung cancer risk of 14% per fiber-year/ml for asbestos exposure. No increased risk was found for the other exposure factors. An overall attributable proportion of 9.5% (95% CI: 5.5, 13.9) was estimated for lung cancer related to diesel exhaust, other combustion products, and asbestos. The population included all men aged 40 to 75 years old who were residents of Stockholm County, Sweden, at any time between 1985 and 1990. All cases of lung cancer diagnosed in that time period were identified from the regional cancer registry. Controls were ran- domly selected from population registers and matched for age and year of inclusion. A postal questionnaire was sent to all study subjects and included information on lifetime occupational history, residential history, smoking habits and other risk factors for lung can- cer. An industrial hygienist assessed the intensity and probability of exposure to occupa- tional exposure factors for every subject on a case-by-case basis and based the assessment on personal contacts, personal experience, and reports of exposure levels specific for occu- pation. Nitrogen dioxide was used as an indicator for exposure to diesel exhaust. The relative risks of developing lung cancer were estimated by unconditional regression. Smoking was controlled for. Environmental levels of nitrogen dioxide were used to esti- mate non-occupational exposure to air pollutants from road traffic. Indoor radon exposure was estimated using geographic data on ground levels of radon. To investigate the lung cancer risk from occupational as well as environmental exposures, using detailed individual exposure data. This paper focused on lung cancer risk in relation to seven occupational exposure factors: diesel exhaust, mixed motor exhaust, combustion products, asbestos, metals, oil mist, and welding fumes. Results regarding environmental exposures were published separately. 1,042 cases and 2,364 controls General population Exposure estimates performed by only one industrial hygienist Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations:

Working Conditions (Environmental except sleeper berth) “Slightly increased relative risks were observed in the highest estimated exposure to mixed motor exhaust and to intermediate exposure to diesel exhaust.” (p. 6 of 14) “Increased risks of lung cancer were noted in the highest quartiles of cumulative exposure to diesel exhaust, combustion products and asbestos. The risk associated with exposure to diesel exhaust was affected neither by adjustment for smoking habits nor by adjustment for exposure to combustion products and asbestos.” (p. 6 of 14) “A positive dose-response relationship was noted for diesel exhaust.” (p. 7 of 14) “The present findings add further evidence for an association between diesel exhaust and lung cancer” (p. 9 of 14) This is a study adding to the literature on the association between exposure to diesel exhaust and lung cancer. Typical biases in questionnaire studies were controlled for and risk esti- mates for lung cancer adjusted for smoking, radon exposure, and ambient nitrogen dioxide levels (as an indicator of road traffic air pollution). 130 Findings Directly Related to HOS (include page references): Reviewer’s Notes:

131 Susan Buchanan Executive Summary, Diesel Emissions and Lung Cancer: Epidemiology and Quantitative Risk Assessment. Health Effects Institute 1999. n/a Not given in executive summary (1) to review the epidemiologic data that form the basis of current quantitative risk assess- ments, (2) to identify data gaps and sources of uncertainty, (3) make recommendations about the usefulness of extending or conducting further analyses of existing data sets, and (4) make recommendations for the design of new studies that would provide a stronger basis for risk assessment. “It [the panel] was not charged to evaluate either the broad toxicologic or epidemiologic literature concerning exposure to diesel exhaust and lung cancer for haz- ard identification purposes, which has been done by others.” (p. 1) n/a n/a This was a limited evaluation of the literature. Driver Health (General) “At present, the railroad worker cohort study (Garshick et al. 1988), though part of a larger body of hazard identification studies, has very limited utility for QRA of lifetime lung can- cer risk from exposure to ambient levels of diesel exhaust . . .” (p. 4) “The investigators’ analysis of the teamster data reported an exposure-response relation (Steenland et al. 1998) that may be useful for QRA; this relation will be better understood with further exploration of uncertainties and assumptions, particularly those relating to the reconstruction of past exposures and the selection of controls.” (p. 4) This summarizes the work of the Diesel Epidemiology Project of the Health Effects Institute (HEI) which set out to address the 4 issues listed above. It does not evaluate the evidence associating diesel exhaust with lung cancer, but it does comment on studies (Steenland and Garshick) reviewed for the HOS literature search in Part I. Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

Susan Buchanan Mortality Associated with Sleep Duration and Insomnia. Kripke, D.F., Garfinkel, L., Wingard, D., Klauber, M., Marler, M. Arch Gen Psych 59. 131–136. 2002. Background: Patients often complain about insufficient sleep or chronic insomnia in the belief that they need 8 hours of sleep. Treatment strategies may be guided by what sleep durations predict optimal survival and whether insomnia might signal mortality risks. Methods: In 1982, the Cancer Prevention Study II of the American Cancer Society asked participants about their sleep duration and frequency of insomnia. Cox proportional haz- ards survival models were computed to determine whether sleep duration or frequency of insomnia was associated with excess mortality up to 1988, controlling simultaneously for demographics, habits, health factors, and use of various medications. Results: Participants were more than 1.1 million men and women from 30 to 102 years of age. The best survival was found among those who slept 7 hours per night. Participants who reported sleeping 8 hours or more experienced significantly increased mortality hazard, as did those who slept 6 hours or less. The increased risk exceeded 15% for those reporting more than 8.5 hours sleep or less than 3.5 or 4.5 hours. In contrast, reports of “insomnia” were not associated with excess mortality hazard. As previously described, prescription sleeping pill use was associated with significantly increased mortality after control for reported sleep durations and insomnia. Conclusions: Patients can be reassured that short sleep and insomnia seem associated with little risk distinct from comorbidities. Slight risks associated with 8 or more hours of sleep and sleeping pill use need further study. Causality is unproven. The Cancer Prevention Study of the American Cancer Society (CSPII) provided data from 1.1 million participants who completed health questionnaires in 1982. Survival or date of death was ascertained 6 years later. 32 covariates were entered into the models of sleeping duration and mortality risk. To explore whether sleep duration predicts mortality. 636,095 men and 480,841 women General population n/a Driver Health (General) “Among both women and men, the best survival was experienced by those reporting a usual sleep duration of 7 hours.” (p. 5 of 10) “ Reported sleep had to be less than 3.5 hours among women and less than 4.5 hours among men for the added risk associated with short sleep to exceed 15%.” (p. 5 of 10) “Comparison of the 32-covariate models with the simplified CPSII models and the less- controlled CPSI tabulations showed that most mortality risk associated with short sleep could be explained by comorbidities.” (p. 6 of 10) Short sleep duration of as low as 4.5 hours does not affect mortality. This study offers an opposing view to the studies showing adverse health effects of sleep deprivation. 132 Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

133 Susan Buchanan The relative importance of whole body vibration and occupational lifting as risk factors for low-back pain. K.T. Palmer, M.J. Griffin, H.E. Syddall, B. Pannett, C. Cooper, D. Coggon. Occup Environ Med 2003;60:715–721. Aims: To explore the impact of occupational exposure to whole body vibration (WBV) on low back pain (LBP) in the general population and to estimate the burden of LBP attribut- able to occupational WBV in comparison with that due to occupational lifting. Methods: A questionnaire including sections on WBV at work, LBP, and potential risk factors was mailed to a community sample of 22,194 men and women of working age. Sources and durations of exposure to occupational WBV were ascertained for the past week and per- sonal vibration doses (eVDV) were estimated. Analysis was confined to subjects reporting exposures in the past week as typical of their work. Associations of LBP with eVDV, driv- ing industrial vehicles, and occupational lifting were explored by logistic regression and attributable numbers were calculated. Results: Significant associations were found between daily lifting of weights greater than 10 kg at work and LBP, troublesome LBP (which made it difficult to put on hosiery), and sciatica (prevalence ratios 1.3 to 1.7); but the risk of these outcomes in both sexes varied little by eVDV and only weak associations were found with riding on industrial vehicles. Assuming causal associations, the numbers of cases of LBP in Britain attributable to occupational WBV were estimated to be 444,000 in men and 95,000 in women. This compared with an estimated 940,000 male cases and 370,000 female cases of LBP from occupational lifting. Conclusions: The burden of LBP in Britain from occupational exposure to WBV is smaller than that attributable to lifting at work. A postal survey was sent to 22,194 adults in Britain regarding exposure to vibration and health. Subjects were selected at random from patient lists of general practices and from members of the armed services randomly selected from pay records. Exposure to vibration was assessed by asking about driving or riding any of a checklist of vehicles. Duration of exposure was also assessed. Logistic regression was adjusted for non-occupational expo- sure to WBV. To assess the burden of LBP caused by exposure to WBV and to compare it with LBP caused by heavy lifting. Questionnaires were sent to general public and armed services in Britain. 4,250 men and 3,061 women were included in the final analysis. n/a Exposure was assessed by questionnaire only. May have response bias; those with back pain or exposure to vibration were more likely to complete the survey. Working Conditions (Environmental except sleeper berth) Very weak association found between estimated vibration dose and presence of low back pain. “The data on WBV do not provide strong evidence to suggest a cause-effect relation (and do not distinguish the risks of sitting and driving from those of merely sitting).” (p. 717) “Use of industrial vehicles in general did not appear to confer an increased risk in men.” A significant association was found in women, although many fewer women had these exposures. Studies using populations of exposed workers instead of just the general population show stronger associations between WBV and low back pain. This may be due to the multiple possibilities for bias inherent in a cross-sectional survey of the general population. Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

Susan Buchanan Impact of sleep debt on metabolic and endocrine function. Spiegel, K., Leproult, R., Van Cauter, E. Lancet 1999;354:1435–1439. Background: Chronic sleep debt is becoming increasingly common and affects millions of people in more-developed countries. Sleep debt is currently believed to have no adverse effect on health. We investigated the effect of sleep debt on metabolic and endocrine func- tions. Methods: We assessed carbohydrate metabolism, thyrotropic function, activity of the hypothalamo-pituitary-adrenal axis, and sympathovagal balance in 11 young men after time in bed had been restricted to 4 h per night for 6 nights. We compared the sleep-debt condi- tion with measurements taken at the end of a sleep-recovery period when participants were allowed 12 h in bed per night for 6 nights. Findings: Glucose tolerance was lower in the sleep-debt condition than in the fully rested condition (p < 0.02), as were thyrotropin con- centrations (p<0.01). Evening cortisol concentrations were raised (p = 0.0001) and activity of the sympathetic nervous system was increased in the sleep-debt condition (p < 0.02). Interpretation: Sleep debt has a harmful impact on carbohydrate metabolism and endocrine function. The effects are similar to those seen in normal aging and, therefore, sleep debt may increase the severity of age-related chronic disorders. Eleven healthy young men spent 16 nights in the clinical research center. For the first 3 nights, they spent 8 hours in bed; for 6 nights, they were in bed for 4 hours; and the last 7 nights, they were in bed for 12 hours. They were assessed for carbohydrate metabolism and hormonal profiles and compared sleepiness, sympathovagal balance, and saliva-free cortisol concentrations in all three conditions. To evaluate the metabolic and hormonal variables in people in whom sleep had been restricted and extended. 11 healthy volunteers n/a Very small population. (n = 11) Limited population (healthy young men) Driver Health (General) “During the sleep-debt condition, responses were consistent with a clear impairment of car- bohydrate tolerance.” (p.1437) “The normal rise in thyrotropin at night was strikingly decreased in the sleep-debt condi- tion compared with that in the sleep-recovery condition . . .” (p. 1438) “Based on the analysis of the concentrations of free cortisol in saliva, the rate of decrease of free cortisol concentrations between 1600 h and 2100h was about six times slower in the sleep-debt condition than in the sleep-recovery condition.” (p. 1438) “Therefore, although the primary function of sleep may be cerebral restoration, sleep debt also has consequences for peripheral function that, if maintained chronically, could have long-term adverse effects on health.” “The metabolic and endocrine alterations seen during the sleep-debt condition therefore mimic some of the hallmarks of aging, which suggests that chronic sleep loss could increase the severity of age-related pathologies, such as dia- betes and hypertension.” (p. 1438) Changes in metabolic function seen during sleep deprivation in this study are similar to that seen in patients with type 2 diabetes. Changes in cortisol levels with sleep deprivation are similar to that seen in normal aging. 134 Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

135 Susan Buchanan Short Sleep Duration Is Associated with Reduced Leptin, Elevated Ghrelin, and Increased Body Mass Index. Shahrad Taheri, Ling Lin, Diane Austin, Terry Young, and Emmanuel Mignot Plos Med. 2004 December; 1(3): e62. Sleep duration may be an important regulator of body weight and metabolism. An associ- ation between short habitual sleep time and increased body mass index (BMI) has been reported in large population samples. The potential role of metabolic hormones in this asso- ciation is unknown. Methods and Findings: Study participants were 1,024 volunteers from the Wisconsin Sleep Cohort Study, a population-based longitudinal study of sleep disor- ders. Participants underwent nocturnal polysomnography and reported on their sleep habits through questionnaires and sleep diaries. Following polysomnography, morning, fasted blood samples were evaluated for serum leptin and ghrelin (two key opposing hormones in appetite regulation), adiponectin, insulin, glucose, and lipid profile. Relationships among these measures, BMI, and sleep duration (habitual and immediately prior to blood sampling) were examined using multiple variable regressions with control for confounding factors. A U-shaped curvilinear association between sleep duration and BMI was observed. In persons sleeping less than 8 h (74.4% of the sample), increased BMI was proportional to decreased sleep. Short sleep was associated with low leptin (p for slope = 0.01), with a predicted 15.5% lower leptin for habitual sleep of 5 h versus 8 h, and high ghrelin (p for slope = 0.008), with a predicted 14.9% higher ghrelin for nocturnal (polysomnographic) sleep of 5 h versus 8 h, independent of BMI. Conclusion: Participants with short sleep had reduced leptin and ele- vated ghrelin. These differences in leptin and ghrelin are likely to increase appetite, possi- bly explaining the increased BMI observed with short sleep duration. In Western societies, where chronic sleep restriction is common and food is widely available, changes in appetite regulatory hormones with sleep curtailment may contribute to obesity. All employees aged 30 to 60 yr of four state agencies in south central Wisconsin were mailed a survey on sleep habits, health, and demographics in 1989. Mailed surveys were repeated at 5-yr intervals. A stratified random sample of respondents was then recruited for an extensive overnight protocol including polysomnography at baseline. Collection of morning, fasted blood levels of leptin, insulin, ghrelin and adiponectin were added to the protocol in 1995. A 6-day diary of sleep duration was added as part of a protocol to assess daytime sleepiness after the initiation of the cohort study. Multiple regression was used to evaluate the relationship of age, sex, and BMI on hormones. Partial correlations adjusted for age, sex, and BMI were calculated for hormones, with and without control of other potential confounders. The relationships between hormones and sleep were evaluated using multiple linear regression after control for potential confounders including age, sex, BMI, SDB (sleep disordered breathing), and morningness tendencies. To investigate the associations among sleep duration (acute and habitual), metabolic hor- mones, and BMI in the population-based Wisconsin Sleep Cohort Study. Two key oppos- ing hormones in appetite regulation, leptin and ghrelin, play a significant role in the inter- action between short sleep duration and high BMI. Leptin is an adipocyte-derived hormone that suppresses appetite. Ghrelin is predominantly a stomach-derived peptide that stimu- lates appetite. Other mediators of metabolism that may contribute include adiponectin and insulin. Adiponectin is a novel hormone secreted by adipocytes and is associated with insulin sensitivity. 1,024 participants in the overnight study and blood sample. 720 in the diary portion. General population Baseline response rate was 51%. Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations:

Driver Health (General) “We found a significant U-shaped curvilinear relationship between average nightly sleep and BMI after adjustment for age and sex. The minimum BMI was predicted at 7.7 h of average nightly sleep. The most striking portion of the curve was for persons sleeping less than 8 h, where increased BMI was proportional to decreased sleep.” (p. 4 of 11) In the multiple regression model, there was a significant increasing trend in leptin for aver- age nightly sleep duration. There was a significant decreasing trend in ghrelin with total sleep time. (Low leptin and elevated ghrelin are usually associated with increased appetite.) There was no significant correlation between sleep duration (acute or chronic) and serum adiponectin, insulin, and glucose. “[These findings] also represent the first demonstration of a correlation between peripheral hormone levels and both self-reported (questionnaire and diary data) and polysomno- graphically determined sleep amounts in a general population sample.” (p. 5 of 11) “When controlling for BMI, we found no significant correlation between insulin, glucose, or adiponectin levels and various measures of sleep duration.” (p. 7 of 11) Study shows association between sleep duration and obesity, and sleep duration and hor- mones which increase appetite. No association was found with sleep duration and glucose or insulin levels. 136 Findings Directly Related to HOS (include page references): Reviewer’s Notes:

137 Susan Buchanan Working Long Hours. White J, Beswick J. United Kingdom Health and Safety Labora- tory. 2003. n/a Literature review. Literature searches were conducted by the HSE Information Centre search team. The team searched the following databases: OSH-ROM, RILOSH; HSELINE; CISDOC: NIOSHTIC2; Medline; Psychlit; EMED; and Healsafe. The key words used were: long working hours; working time; fatigue; health; safety; work-life balance; acci- dents; psychological effects; task; industry. Shiftwork was specifically omitted. In order to keep the review to a manageable size, it was decided to concentrate solely on articles relat- ing to long working hours, and not those relating to shiftwork, and on literature from the last 10 years. Articles from academic journals as well as articles from health and safety related and other journals were retrieved. The articles included in this review were mostly selected on the basis of their abstract. The findings of the review reflect these constraints. To review the literature on the relationship between long working hours and fatigue, health and safety, and work-life balance outcomes. n/a n/a n/a Driver Health (General) Several studies investigating ‘karoshi’ in Japan were reviewed. Karoshi refers to a syn- drome of cardiovascular attacks attributable in part to the long-hours culture of the Japa- nese. Few other health effects were included in this write-up: one study assessing the effect of long hours on immunity and one showing higher rates of diabetes were mentioned. “There appears to be a link between working long hours and cardiovascular disorder but several factors (e.g., existing medical conditions or insufficient sleep) may mediate this link.”(p. 20) “There appears to be evidence linking working long hours with poor lifestyle behaviors and other physical health problems, such as lowered immunity and diabetes mellitus.” (p. 20) “From the available evidence, there is sufficient reason to be concerned about a possible link between long hours and physical health outcomes, especially for hours exceeding 48 to 50 per week. However, samples were not very diverse, as much research seems to focus on men in Japan.” (p. 20) This literature review reinforces the notion that long work hours may contribute to risk of cardiovascular disease. However, other health effects of long hours were barely mentioned. Most of this extensive report was devoted to fatigue, performance, and lifestyle issues, not to health effects. Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references): Reviewer’s Notes:

Peter Orris Diesel and Health in America: the Lingering Threat. Schneider, C.G., Hill, L.B., Clean Air Task Force, Boston, MA, February 2005. The U.S. Environmental Protection Agency has issued regulations that will require dra- matic reductions in emissions from new diesel vehicles starting in 2007—but only the new ones. These regulations, to be phased in over the next quarter century, apply only to new engines. The lifespan of the average diesel vehicle is nearly 30 years. Many diesels are driven over a million miles. Because of this longevity, we will be left with the legacy of pollution from dirty diesel vehicles for decades to come. Pollution from dirty diesels on the road now can be dramatically reduced using a combination of cleaner fuels, retrofit emis- sion controls, rebuilt engines, engine repowerings, and accelerated purchase of new, cleaner vehicles. The Clean Air Task Force commissioned Abt Associates, to quantify the health impacts of fine particle air pollution from America’s diesel fleet. Using this information, we were able to estimate the expected benefits—in lives saved—from an aggressive but fea- sible program to clean up dirty diesel buses, trucks, and heavy equipment across the United States. It then reviews the degree to which diesel vehicles increase the level of fine particle pollution in the air we breathe, and recommends reduction measures that will save thou- sands of lives each year. Collected available environmental particulate data, reviewed the literature as to health effects, modeled the amount of exposure from the current diesel truck fleet in the United States and calculated the human and economic costs of not putting in available protective technologies across the board. To review the relationship between diesel exhaust and health impacts on the population as a whole and estimate the human and economic costs of continued inaction with respect to utilizing available technology to reduce exposure. n/a n/a Modeling exercise based on partial data allowing debate as to the magnitude of the quanti- tative conclusions reached. Driver Health (General) • Reducing diesel fine particle emissions 50% by 2010, 75% by 2015, and 85% by 2020 would save nearly 100,000 lives between now and 2030. • Fine particle pollution from diesels shortens the lives of nearly 21,000 people each year. This includes almost 3,000 early deaths from lung cancer. • Tens of thousands of Americans suffer each year from asthma attacks (over 400,000), heart attacks (27,000), and respiratory problems associated with fine particles from diesel vehicles. These illnesses result in thousands of emergency room visits, hospi- talizations, and lost work days. Together with the toll of premature deaths, the health damages from diesel fine particles will total $139 billion in 2010. • Nationally, diesel exhaust poses a cancer risk that is 7.5 times higher than the com- bined total cancer risk from all other air toxics. • In the United States, the average lifetime nationwide cancer risk due to diesel exhaust is over 350 times greater than the level U.S. EPA considers to be “acceptable” (i.e., one cancer per million persons over 70 years). • Residents from more than two-thirds of all U.S. counties face a cancer risk from diesel exhaust greater than 100 deaths per million population. People living in eleven urban 138 Reviewer: Complete Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings Directly Related to HOS (include page references):

139 counties face diesel cancer risks greater than 1,000 in a million—one thousand times the level EPA says is acceptable. • The risk of lung cancer from diesel exhaust for people living in urban areas is three times that for those living in rural areas. This literature review is consistent with the conclusion that increased weekly working hours and the probable concomitant increased exposure to diesel exhaust is likely to contribute to an increased risk of cancer. It is also consistent with the conclusion that this impact is likely mitigated by changes in cab and engine design currently incorporated in new vehicles. Reviewer’s Notes:

FATIGUE EFFECTS REFERENCES REVIEW SELECTION CRITERIA The 25 references included in this part were derived from a total of 266 references included in comments on the advanced NPRM. Of the 266 references, 86 pertained to health effects and were passed on to Dr. Peter Orris for evaluation. The remaining 180 articles were rated on a scale of 1 to 4 for rel- evance, with 34 given a score of 1, 25 a score of 2, and the remaining 121, scores of 3 or 4. The most relevant (score of 1) 34 articles on performance, crash risk, and fatigue were selected for review, of which 25 could be obtained. Articles considered most relevant were those involving epidemiolog- ical studies of CMV crash risk or field studies of performance of commercial drivers in relation to fatigue issues such as daily and weekly hours, time of day, and short sleep, or stud- ies of non-CMV drivers showing the effects of sleep loss and comparing sleep loss and alcohol impacts. In Part I, no crash studies were reviewed. The reasons for not reviewing the remaining articles sug- gested by commentators included the following: • The article was not published as a report of a recognized agency or in a peer-reviewed journal (e.g., a website only or popular magazine). • The article was very general in nature (e.g., Sleep and Circadian Disturbances in Shift Work: Strategies for Their Management). • The article was not sufficiently relevant to the task of CMV driving and to the issue of fatigue and health (e.g., A Photograph-Based Study of the Incidence of Fatal Truck Underride Crashes in Indiana; Census of Fatal Occupational Injuries Summary; Effects on Per- formance of High and Low Energy-Expenditure During Sleep Deprivation). EXECUTIVE SUMMARY Fatigue, Time of Day, and Performance The impacts of shift schedule on subjective fatigue and performance were measured in a field study by Williamson et al. (2004). In addition to permanent day shift and night shift drivers, drivers working alternating weeks of day and night shifts participated in the study. Fifty-four drivers were measured repeatedly over a 2-week period. Actigraph data were also collected to provide objective measures of the tim- ing and quality of sleep. The researchers found that while the night shifts made drivers feel more tired than day shifts, they did not produce significantly poorer performance on tests of PVT and Mackworth Clock, “suggesting that night drivers can manage their fatigue.” Over a typical workweek of five consecutive 10- to 12-hour shifts, there was a significant increase in subjective ratings of fatigue by all drivers. Night shift drivers worked longer shifts than day shift driv- ers and spent much more of their working time driving than day shift drivers “which might predict that night shifts would be more tiring than day shifts.” However, the authors suggest that night shift drivers may have performed as well as day shift drivers as they may be “especially tolerant of fatigue or skilled at managing fatigue” and because they organize their sleep differently (e.g., napping in the hours before their first shift of the week) which may partly explain how they could maintain performance. For example, the authors noted that night drivers “endeavored to capitalize on the sleep propen- sity influences of the circadian rhythm by getting as much sleep as they could as close as possible to the early morning circadian trough when sleep is most likely.” All drivers had restricted sleep (4 to 6 hours) and worked long hours (50 to 55 hours arranged in five 10- to 12-hour shifts), which the authors believe may have overshadowed any circadian differences. However, it could be argued this would be more likely to exacerbate them. They note a prob- lem of missing data. Also all tests were done when the vehi- cle was stopped which may have re-alerted drivers. Other studies which have used performance measures that were integral to the driving task, such as lane tracking control and critical incidents, have found poorer performance at night, compared with during the day, most notably the U.S.-Canada study by Wylie et al. 1997, which compared different driving schedules as well as the study of long-haul single and team drivers by Dingus et al. 2001. Nonetheless the Williamson et al. (2004) findings are surprising given other studies where the PVT was sensitive to circadian effects (Dinges et al. 1997). Scheduling Flexibility and Fatigue Williamson et al. (2004) report further analysis of their 1996 study (Williamson et al. 1996) in which 27 commercial drivers participated in each of three work practices: staged trip driving (two drivers from different points of origin meet mid-trip and exchange loads), flexible trip driving (single 140

141 driver, trip scheduled without reference to HOS regulations) and single trip driving (single driver, within HOS regulations). A range of fatigue measures were used including performance tests, physiological and subjective measures. The authors found that “a 10–12 hour trip is tiring no matter how the work is organized, and that the effects of accumulated fatigue may overshadow the effects of fatigue on a single 10–12 hour trip.” “Differences between driving types were not sufficient to account for changes in fatigue or performance in this study. All drivers reported more fatigue over the trip, but not all driv- ers showed poorer performance. It seems that the 12-hour trip is relatively immune to any effects of differences in work practices. It is possible that studying such relatively short trips will not provide clear findings . . . It is certainly note- worthy that flexible trips produced no worse an outcome than either of the other two ways of doing the same trip. In fact, a more exhaustive evaluation of flexibility, where drivers have the opportunity to learn about manipulating the timing of work and rest during several trips, might reveal that flexibil- ity is of benefit in managing fatigue. . . . It should be noted, however that flexible drivers did tend to select work-rest schedules which were quite similar to the regulated working hours. It would be interesting to determine whether this sim- ilarity persists when the trip is longer.” Decrements in performance may be more detectable when driving performance is measured as opposed to the alerting situation of stopping to carry out a special performance test as was done in this study. The findings of this study suggest that drivers do not make use of flexible schedules in a manner that reduces fatigue. Fatigue management programs may assist, especially if not only drivers but also dispatchers and managers are involved. A limitation is suggested by the Williamson study that indi- cates little difference in fatigue effects on different schedules if sleep is restricted and hours are long. Regulated and unregulated HOS regimes were also com- pared using survey data by Hartley (1999). The Hartley et al. (1996) study on the impact of fatigue on heavy vehicle driv- ers in Western Australia (where there are no HOS regula- tions), was compared with Williamson and Feyer’s 1992 study, which contained a comparable survey of drivers work- ing under HOS in the eastern states. Drivers whose hours were not regulated were no more likely to exceed the HOS regulations that the drivers for whom those regulations were enforced. Drivers whose hours were not regulated were less likely to consider fatigue a problem than those whose hours were regulated. This may be due to less fatigue or to less awareness of the association of fatigue with poorer perfor- mance and increased crash risk. The authors list the follow- ing problems with HOS: • HOS prescribe what a driver should be capable of doing (i.e., no flexibility). • HOS regulations do not inform organizations about fatigue and safety. • HOS regulations permit no discretion for different freight tasks and environments. • HOS do not take account of the influence of the driver’s circadian cycle. • HOS regulations take no account of time zone changes. • HOS may restrict access to sleep. • There is no commercial incentive to restrict driving to HOS. The numerous factors affecting fatigue and the difficulty of regulating hours of work to minimize it are discussed by Moore-Ede and Schlesinger (2005). They argue that risk relat- ing to work and rest hours is multi-factorial and that simplis- tic regulations based on only one or two factors have limited value in minimizing this risk. Over 30 factors that determine level of sleepiness and fatigue-related accident risk are listed, with the most important being circadian phase followed by the number of consecutive hours spent continuously awake since the previous sleep episode. Other important factors are the length of the sleep episode, the quality of sleep, job work- load, and moment to moment stimulants or depressants of alertness. The authors use two case studies to illustrate the problem with the current (11-hour driving) HOS regulations. In particular, they note the disincentive for drivers to take a nap when they are tired as daytime naps are only allowed to be excluded from a driver’s hours on duty in certain situations when the nap is followed by driving, which is in turn imme- diately followed by an extended period of rest. The authors suggest that the efficacy of “alternative, less punitive, risk management strategies based upon the science of fatigue management” should be demonstrated to provide the basis of HOS regulation. The authors suggest alternative paradigms to the current work-rest regulations: fatigue management pro- grams, fatigue risk models, alertness monitors. Long Weekly Hours and HOS Violations A number of studies have shown that despite the lengthy hours allowed by HOS regulations (60 hours in 7 days or 70 hours in 8 days), significant numbers of drivers work even longer. Beilock et al. (1995) use self-reported data to estimate the frequency of HOS violation-inducing schedules for a sam- ple of 498 long-distance drivers. Twenty-six percent of the drivers, assuming average legal speed limits of 55 mph, were found to have violation-inducing schedules. Drivers most likely to have these schedules included solo drivers, drivers hauling refrigerated loads, regular route drivers, and those with longer current trip distances. The findings indicate that the very large majority of long-distance drivers have more than 40-hour work weeks (82%, assuming average maintained speeds of 50 mph), and extremely lengthy work weeks are common. For example, assuming 50-mph average maintain- ing speed, half the drivers work more than 65 hours weekly and one-quarter work over 81 hours.

142 as the baseline. The accident risk of driving between 4:00 and 6:00 p.m. was significantly higher (approximately 60% higher) than that of the baseline. This was attributed to two effects: evening rush hour and reduced alertness because of a low cir- cadian period for some drivers. The accident risks from mid- night to 2:00 a.m., 6:00 to 8:00 a.m., and 8:00 to 10:00 p.m. were significantly higher than during the baseline. Two peri- ods involve night driving; the other involves part of the dawn period. Rest breaks, particularly those taken before the 6th or 7th hr of driving, appeared to lower accident risk significantly for many times of day. Park et al. (2005) used pre-existing crash data from the 1980s and measurements from the Driver Fatigue and Alert- ness Study (DFAS) conducted in the mid-1990s. The total sample size was 5,050 drivers (i.e., 954 accident-involved drivers and 1,506 non-accident drivers in 1984; 887 accident drivers, and 1,604 non-accident drivers in 1985). The research appears to use a larger data set but similar methods to the Lin et al. study described previously. The study explores “whether a more detailed examination of time of day of driving, par- ticularly over multiple days, indicates associations with crash risk.” Night and morning driving and irregular schedules with primarily night and early morning driving, have significantly elevated crash risk of 20 to 70%, 30 to 80%, respectively, compared with daytime driving. A case-control approach was used in New Zealand to deter- mine the effect of work schedule variables on crash risk (Frith 1994). A ‘case’ group of drivers and heavy vehicles involved in crashes (1988 to 1990) were compared with a ‘con- trol’ group of drivers and vehicles (1992 to 1993). The crash- involved drivers were 2.6 times more likely, as compared with non-crash involved drivers, to have driven 8 or more hours since the last compulsory 10-hour off-duty period (as recorded in the log book). However, no other scheduling vari- ables were found to be associated with crash risk. Drivers involved in crashes tended to be younger than control drivers. Once trip length is controlled for exposure, three crash studies show an increase in crash risk with hours of driving. Campbell (2002) concludes that the relative risk of fatigue given involvement in a fatal accident “gradually increases during the first 8 hours, doubles during the ninth hour and is higher by a factor of 6 by the 12th hour.” Lin et al. shows that “accident risk increases significantly after the 4 hr, by approximately 50 percent or more, until the 7th hr. The 8th and 9th hr show a further increase, approximately 80 and 130 percent higher than the first 4 hr.” Frith (1994) shows crash involved drivers to be 2.6 times more likely than non-crash involved drivers to have driven 8 or more hours. With respect to hours driving and crash risk, these studies are consistent with earlier studies by Jones and Stein (1985) and Harris (1978). Using a case-control approach to examine the relative risk associated with long hours of driving, Jones and Stein (1987) found that tractor-trailer drivers who drove in excess of 8 hours, who violated log book regulations, and who were aged 30 and younger had an increased risk of crash Hertz et al. (1990) found for 130 long-haul tractor trailer drivers that at “assumed trip speeds of 40 mph and 50 mph, 90% and 51% of the drivers, respectively, were in violation of the hours of service rules by more than one hour.” Long Hours, Time of Day, and Crash Risk Studies of the impact of long hours and time of day on crash risk are methodologically challenging for a number of rea- sons. First, the distribution of trip lengths is such that there are more 4-hour trips than 8-hour or 12-hour trips. Consequently, there will be more crashes associated with 4-hour trips than with 12-hour trips. This is a different issue from the risk per 4-hour trip or per 12-hour trip. Thus, it is important to control for exposure in order to determine per trip risk. Secondly, time of day and long hours both impact fatigue, and it is difficult to separate these impacts. Campbell (2002) reports that more than 25% of the accidents occurred in the first hour, and two- thirds in the first 4 hours, and that “only about 4 percent of all medium and heavy truck drivers involved in a fatal crash reported driving more than 8 hours at the time of the accident.” The authors note that this pattern is driven “by exposure, not risk” as the “nature of the exposure distribution will always keep the number of accidents after many hours driving a small proportion of the total, even with dramatic increases in risk with hours driving.” When differences in exposure were con- sidered the authors found the following: 1. “The relative risk of fatigue given involvement in a fatal accident follows the circadian rhythm.” 2. The relative risk of fatigue given involvement in a fatal accident “gradually increases during the first 8 hours, doubles during the ninth hour and is higher by a factor of 6 by the 12th hour.” Lin et al. (1994) formulated an elapsed time-dependent logistic regression model to assess the safety of motor carrier operations. The data were obtained from a national less-than truckload firm that operated coast-to-coast with no sleeper berths. The total number of observations used for modeling was 1,924 cases, of which 694 were accidents and 1,230 were non-accidents. The model “estimates the probability of hav- ing an accident at time interval, t, subject to surviving (i.e., not having an accident) until that time.” The model was then tested with data from trips involving and not involving crashes from trucking company operations. Analysis showed that driving time had the strongest direct effect on crash risk. Accident risk increased “significantly after the 4 hr, by approximately 50 percent or more, until the 7th hr. The 8th and 9th hr show a further increase, approximately 80 and 130 percent higher than the first 4 hr.” Drivers with more than 10 years of driving experience had the lowest accident risk. Daytime driving, particularly around noon, was associated with significantly lower risk of an accident and was defined

involvement. In particular, the relative risk of crash involve- ment for drivers who reported a driving time in excess of 8 hours was almost twice (i.e., almost 100% higher) that for drivers who had driven fewer hours. Lin et al. 1994 found a 50% increase in risk after 4 hours of driving. In samples of dozing driver and single-vehicle crashes, Harris (1978) found that the changeover to more accidents than expected from fewer accidents than expected, occurred after about 5 hours of driving. With respect to time of day of driving, Park et al. found that night and morning driving and irregular schedules with pri- marily night and early morning driving, have significantly elevated crash risk of 20 to 70%, 30 to 80%, respectively, compared with daytime driving. Lin et al. found some asso- ciation with time of day. However, as noted by the authors, there was no control for exposure, hence the finding that the highest accident risk occurred between 4:00 and 6:00 p.m., peak hour with respect to traffic volume. Time of day was controlled for exposure in the earlier study reported by Harris (1978). In a sample of single-vehicle crashes, the circadian effect was evident as 46% of the acci- dents occurred between midnight and 0800, almost “2.5 times as many as would be expected from the exposure data (19%). In a sample of “dozing driver” crashes, approximately 70% occurred between midnight and 0800. In contrast, approxi- mately 25% of the accidents for the multi-vehicle crash sam- ple occurred between midnight and 0800. Time of day of driving has also been shown to impact crash risk of passenger car drivers. When exposure is accounted for, as it has been in three studies in different countries, of single- vehicle passenger car crashes, without alcohol involvement, a very strong association with time of day is found, with 13 to 25 times the risk of a crash per mile driven in the 2:00 to 4:00 a.m. period as compared with during typical working hours (see Smiley 2002 for a summary). The effect of cumulative shifts in a sequence on crash risk has received little attention. One study by Jovanis and Kaneko (1990) examines this through an analysis of carriersupplied accident and nonaccident data for a 6-month period in 1984. The data were obtained from a “pony express” type opera- tion, which operates coast to coast with no sleeper berths. Cluster analysis was used to identify nine distinct patterns of driving hours over a 7-day period. The driving patterns of drivers who had an accident on the 8th day were compared with drivers who had no accident on the 8th day. An increased crash risk was found for night but not day drivers after 3 to 4 days of driving. Industrial shift schedules are typically more rigid than CMV driver schedules and studies using these are helpful in illu- minating risks associated with various shift features. In a meta-analysis, Folkard and Lombardi (2004) examined stud- ies of injuries and accidents which occurred in industrial set- tings and related them to the time of day; to the point within the shift system that they occurred; and to the shift features such as type of shift, length of shift, and number of succes- 143 sive shifts. (There were about 20 studies of accident and injuries referenced, and only a sub-sample of relevant stud- ies could be used for each analysis.) There was a highly sig- nificant main effect of shift, in that risk increased by 18.3% on afternoon shifts and by 30.4% on night shifts relative to the morning shift. There was also a consistent trend in accident risk over four successive nights. On average, risk was about 6% higher on the second night, 17% higher on the third night, and 36% higher on the fourth night. Although the effect was not signif- icant, there was also increased risk, though to a lesser degree, associated with successive day shifts. Risk was about 2% higher on the second morning/day, 7% higher on the third morning/day, and 17% higher on the fourth morning/day shift than on the first shift. These findings are consistent with the Jovanis and Kaneko (1990) findings for truck drivers, in that cumulative shifts at night had a greater impact on crash risk than cumulative shifts worked during the day. Folkard and Lombardi found that risk also increased with the length of the shift. Relative to 8-hour shifts, 10-hour shifts were associated with a 13.0% increase, and 12-hour shifts with a 27.5% increase in risk. These findings are also consis- tent with findings for CMV drivers, showing increases in crash risk after work exceeds 4 or 5 hours. Sleep Restriction and Performance A number of studies have shown that CMV drivers, espe- cially long-haul drivers suffer from sleep restriction and thus the impact of sleep restriction on performance is a concern. Van Dongen et al. (2003) studied effects of chronic and total sleep restriction on cognitive performance and found that “chronic restriction of sleep periods to 4 h or 6 h per night over 14 consecutive days resulted in significant cumulative, dose- dependent deficits in cognitive performance on all tasks.” Lapses in behavioral alertness and reductions in working mem- ory performance in the 4 h condition reached levels equiv- alent to those observed after 2 nights without sleep. After 14 days of sleep restriction, cognitive throughput performance was equivalent to that observed after 1 night without any sleep. The authors note that the study results do not support the notion of “core” and “optional” sleep.” Belenky et al. (2003) viewed the core versus optional sleep issue slightly differently. They found that 7 days of sleep restriction degraded psychomotor vigilance performance in a sleep-dose dependent manner. With mild to moderate sleep restriction (7 and 5 hours time in bed), performance initially declined and, after a few days, appeared to stabilize at a lower-than-baseline level for the remainder of the sleep restric- tion period. In contrast, with severe sleep restriction (3 hours time in bed) performance declined continuously across the sleep restriction period, with no apparent stabilization of per- formance. Thus the 5 to 7 hours time in bed might be consid- ered “core” sleep, in that performance stabilized, but was not

at a level to maintain maximum performance. Three days of recovery sleep did not restore performance to baseline levels for subjects with mild to moderate sleep restriction (5 or 7 hours time in bed). Lenne et al. (1997) looked at the effects of sleep depriva- tion, time of day, and driving experience on driving simula- tor performance. Lane position and speed variability were significantly higher following sleep deprivation. Circadian effects were shown in that performance steadily improved across the day between 0800 and 2000, after both normal sleep and sleep deprivation. Inexperienced drivers had higher reaction times than experienced drivers in both sleep-deprived and non-sleep deprived conditions.” The results of these studies suggest that there are cumula- tive performance consequences to limiting sleep to even as much as 6 hours per day, and that the effects of sleep depri- vation may be more for inexperienced drivers. Sleep Restriction, Time of Day, and Crash Risk NTSB (1995) focused on the sleep patterns of the 96 hours preceding 107 single-vehicle heavy truck crashes in which the driver survived. Fifty-eight percent of the crashes were fatigue-related. Fatigue was considered a probable cause of the crash if the driver was estimated to have been on duty for more than 15 consecutive hours (the current legal limit), and if the driver’s performance involved non-professional, irra- tional actions such as failure to brake or make appropriate steering maneuvers. A statistical analysis determined that the most important measures predicting a fatigue-related crash in this sample were the “duration of the last sleep period, the total hours of sleep obtained during the 24 hours prior to the crash and the split-sleep patterns.” Studies described above indicate the negative effect of restricted sleep on performance. The NTSB study provides evi- dence that the performance changes resulting from restricted sleep can lead to crashes. Fatigue, Sleep Restriction, and Crash Risk Kecklund et al. (1999) examined 79 rail crashes, finding that approximately 17% were potentially related to fatigue or sleepiness. Indicators of suspected fatigue were considered to be present when one of the following three criteria appeared in combination with the fourth criterion: • “The driver admitted or the investigator observed fatigue. • Time of the accident (between 3:00 a.m. and 6:00 a.m.). • Lack of sleep (less than 5 hours sleep) or a shift being preceded by a brief period of off-duty time (less than 11 hours). • Accidents or incidents characterized by missed signals, lack of attention or loss of memory. It is known that this type of event is frequently triggered by fatigue. “ 144 Only 4% of the accidents were fatigue related according to first criterion. However, this figure rose to 17% if all criteria were applied. While the majority of accidents occurred during the first 3 working hours, the authors note that length of trip was not controlled for exposure, and that the sample was small. “The statistically significant analysis determined that the most important measures in predicting a fatigue-related acci- dent in this sample are the duration of the last sleep period, the total hours of sleep obtained during the 24 hours prior to the accident, and split sleep patterns.” Rail accidents, like truck accidents associated with fatigue, are characterized by non-performance and are related to time of day and restricted sleep. In less than one-quarter of the crashes did the engineer admit fatigue, suggesting only a small number of police reports indicate fatigue as the cause. Alcohol vs. Prolonged Wakefulness Arnedt et al. (2001) compared effects of alcohol with those of prolonged wakefulness on a simulated driving task. They found that performance following 19 and 22 hours of wake- fulness (measured at 0230 and 0500) was equivalent to 0.05 and 0.08% BAC, respectively (measured during the day). Roehrs et al. (2003) looked at the effects of sleep loss (0, 2, 4, and 8 hours of sleep loss) as compared with those of ethanol ingestion in 32 adults (ages 21 to 35). “The study was conducted in a mixed design with a between-subject factor, ethanol or sleep loss, and a within-subject factor, dose of either ethanol or sleep loss.” The authors found that “sleep loss was more potent than ethanol in its sedative effects but comparable in effects on psychomotor performance. Ethanol produced greater memory deficits, and subjects were less aware of their overall performance impairment.” These two studies provide an appreciation of the impact of sleep loss in terms of the effects of alcohol, the effects of which on crash rates are well known. These studies suggest that CMV drivers working long hours after restricted sleep may be as impaired with respect to driving performance as drivers at the legal limit of alcohol. Fatigue, Drugs, and Crash Risk National Transportation Safety Board (NTSB) (1990) found “Thirty-three percent of the fatally injured drivers in 182 acci- dents tested positive for alcohol and other drugs of abuse.” In approximately 8 of the 168 cases, urine samples were used to detect drugs of abuse. In the remainder of the 168 cases, blood samples were used at NTSB’s sensitivity thresholds, which have substantially lower cutoff concentrations than the DOT drug testing regulations, making it more likely that pos- itive test results would be found. A concern is whether the lev- els found indicate behavioral impairment due to the drug, or merely its presence. This is because presence can be detected in blood many hours after consumption, and in urine, days after

consumption, in many cases long after effects on behavior can be detected. Alcohol/drugs positive results were more likely to be present on or after the weekend, and were no more likely to be found in fatigued drivers than in drivers not so designated. The authors found that there was a strong association between violation of the federal HOS regulations and drug use. In addition, there was a significant relationship between drug positive test results and a shipment deadline for the load being carried. However, some of the drugs considered in this study are stimulants, which have been demonstrated to improve performance. The presence of an illegal drug cannot be considered as “impairing” unless there is evidence in the performance testing literature that the drug in the quantity found actually does impair performance. Impact of System Issues on CMV Driver Fatigue Virtually all studies of fatigue focus on the driver; few stud- ies have looked at fatigue from a system perspective. One such study is by Braver et al. (1999) who looked at the role of ship- per demands. Dispatchers were interviewed and reported that shippers rarely requested tight delivery schedules. However, there is a possibility that dispatchers may have responded to questions about tight delivery schedules according to typical driver work schedules rather than HOS regulations. In partic- ular, the authors note, that the study “did not attempt to quan- tify how a dispatcher defined ‘more than enough time,’ ‘just enough time,’ or ‘not enough time’ to pickup and deliver.” The majority of dispatchers said that time allotted per ship- ment for non-driving duties was up to the driver. Of those giv- 145 ing a quantitative estimate, most expected 2 to 4 hours. Driv- ers were interviewed and approximately 20% reported penal- ties from their motor carriers for late deliveries. At the conference, Truck Safety: Perceptions and Reality, the attendees concluded that current HOS regulations in Canada and the United States are too narrowly focused to reduce the incidence of driver fatigue in truck accidents. There is a need to establish a comprehensive set of standards that reflect all types of driver fatigue for different driving situa- tions. The group also felt that low driver wages and lack of empowerment compelled drivers to drive longer hours with- out necessary rest ( Saccomanno et al. 1995). The above documents consider broader system issues ver- sus individual trucker decisions. The Braver et al. study is limited in not defining “tight delivery schedules.” A possi- ble indicator of this is the dispatcher’s estimate of time required for trucker’s other duties of 2 to 4 hours. It would be interesting to compare dispatcher, driver, and fatigue expert opinions on what constitutes an appropriate delivery schedule. Additional References Dinges, D.F., Pack, F., Williams, K., Gillen, K.A., Powell, J.W., Ott, G.E., Aptowicz, C., and Pack, A.I. (1997) Cumu- lative sleepiness, mood disturbance, and psychomotor vig- ilance performance decrements during a week of sleep restricted to 4–5 hours per night. Sleep, 20(4): 267–7. Smiley, A. Chapter 6: Fatigue and driving. In Human Fac- tors and Traffic Safety, Paul Olson and Robert Dewar (eds.), Lawyers & Judges Publishing Company, Tucson, Arizona. 2002.

146 Dianne Davis, Alison Smiley Arnedt, J.T., Wilde, G.J.S., Munt, P.W., and Maclean A.W. “How do prolonged wakeful- ness and alcohol compare in the decrements they produce on a simulated driving task?” (2001). Accident Analysis and Prevention, Vol. 33, 337–344. This report looks at a comparison of the effects of alcohol with those of prolonged wake- fulness on a simulated driving task. Eighteen males (18 to 35 years old) with BACs of 0.00, 0.05 and 0.08% drove a simulator for 30 minutes. Dependent measures included subjective sleepiness before and after the driving task as well as simulator measures (e.g., tracking, speed deviation, etc.). Tracking variability, speed variability and off-road events increased with BAC. In contrast, speed deviation (calculated as the deviation from the speed limit) decreased as a result of subjects driving faster. With increased BAC, ratings of sleepiness increased and were higher following the driving task. The authors compared the results of this study with a previous study by Arnedt and Maclean (1996), examining simulated driv- ing performance during one night of prolonged of wakefulness. They found that “mean tracking, tracking variability, and speed variability 18.5 and 21 h of wakefulness produced changes of the same magnitude as 0.05 and 0.08% blood alcohol concentrations, respec- tively.” In addition, they found that alcohol consumption produced changes in speed devi- ation and off-road occurrences of greater magnitude than the corresponding levels of pro- longed wakefulness. Eighteen males between 19 and 35 years old, who met the inclusion criteria (e.g., neither extreme morning nor evening type) were recruited for this study at Queen’s University in Kingston, Ontario. On the nights before the first experimental day and for the duration of the study, subjects were instructed to go to bed between 2300 and 0100 and to get up between 0700 and 0900. The Stanford Sleepiness Scale (SSS) and a Modified Stanford Scale (MSS) were used as the subjective dependent measures for the study. Performance was assessed using a York Driving Simulator. A variety of dependent measures were col- lected using the simulator such as speed deviation, off-road incidents, etc. Subjects were given a training session prior to the start of the experiment to orient them to the driving sim- ulator as well as other aspects of the study. They were then informed of their condition allo- cation (0.00, 0.05, and 0.08% BAC) and underwent all the three test conditions at the same time of day either 1400, 1700 or 2000. Experimental days consisted of brief training ses- sions to re-familiarized subjects with the driving simulator, followed by consumption of drinks mixed so that they would attain peak BACs of 0.0, 0.05 or 0.08%. Subjects were given one-half hour to ingest the drinks, and at 30 minutes post-ingestion, completed the SSS and the MSS and then drove the simulator for 30 minutes. The scales were also com- pleted after the 30-minute simulator task. Subjects in Arnedt and Maclean’s (1996) study underwent one night of prolonged wakefulness and drove on the driving simulator at 2400, 0230, 0500, and 0730. For 0.05 and 0.08% BAC, the most comparable test times in the Arnedt and Maclean (1996) study were 0230 (18.5 h of wakefulness) and 0500 (21 h of wakefulness), respectively. Comparison of effects of alcohol with those of prolonged wakefulness on driving. 18 males (students) between 19 and 35 years of age (Note: Report does not indicate the number of subjects in Arnedt et al.’s (1996) comparison study). n/a The driving situations were limited to situations in which there was no other traffic present. The authors noted “had drivers been exposed to the perceptual and judgment demands of REFERENCE SUMMARIES Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations:

Findings: Findings Directly Related to HOS (include page references): 147 dealing with the greater hazards created by the presence of other traffic, it is possible that driv- ing performance would have differed under prolonged wakefulness and alcohol conditions.” 1. “. . . wakefulness prolonged by as little as 3 h can produce decrements in the ability to maintain speed and road position as serious as those found at the legal limits of alcohol consumption.” 2. The ratings of subjective sleepiness increased as alcohol dosage increased for the SSS and Factors 1 (‘an energic or activating factor’) and 2 (‘related to consciousness, sleepi- ness and a loss of control over remaining awake’) of the MSS. Subjects rated themselves as more sleepy after the driving task than before on each of these scales. No significant time-of-day effects were found for the subjective sleepiness measures. 3. With increased alcohol dose tracking variability, speed variability and number of off- road accidents increased. Speed deviation from the posted speed decreased as subjects drove faster. 4. Performance within the 30-minute driving period declined overall but only reached sta- tistical significance in the case of tracking variability. 5. Time-of-day effects were largely absent, expect for speed deviation, which was lowest at 1700 relative to the other two test times. 6. “Performance following 19 and 22 h of wakefulness was equivalent to 0.05 and 0.08% BAC, respectively.” 7. “Alcohol produced a more marked increase in speed than prolonged wakefulness, with a statistically significant difference between 0.08% BAC and the 05:00 h test time (21 h of wakefulness) but not the 07:30 h test time” (23.5 h wakefulness). 8. “The frequency of off-road occurrences was significantly greater in the 0.08% BAC con- dition than in the 05:00 h test time” (21 h of wakefulness) but not significantly different from the 07:30 h test time (23.5 hr wakefulness). 9. The study raised important issues “regarding driving after consuming alcohol at times of increased physiological sleepiness, namely between the 23:00–07:00 and 14:00–17:00 h” time periods as fatigue-related accidents are more likely to occur during these peak times for sleepiness. Driver Fatigue/Alertness p. 341, “. . . performance following 19 and 22 h of wakefulness was equivalent to 0.05 and 0.08% BAC, respectively.” p. 341, “. . . driving performance was in a number of respects, affected similarly by pro- longed wakefulness and by alcohol. With increasing time awake and increasing blood alco- hol level, subjects tracked increasingly to the left of the center of the lane . . .), and their tracking variability . . . and speed variability . . . increased).” Driver Health No significant findings or assumptions concerning impact on health.

Dianne Davis, Alison Smiley Beilock, R. “Schedule-induced hours-of-service and speed limit violations among tractor- trailer drivers” (1995). Accident Analysis and Prevention, Vol. 27, No. 1, 32–42. This objective of this study was to determine the frequency of “schedule-induced hours-of- service rules (HSR) and/or speed limit violations by drivers of tractor-trailers by estimat- ing the tightness of driver schedules over a specific trip.” The authors use self-report data to estimate the frequency of violation-inducing schedules for a sample of 498 long-distance drivers. Twenty-six percent of the drivers, assuming average legal speed limits of 55 mph, were found to have violation-inducing schedules. Drivers most likely to have these sched- ules included solo drivers, drivers hauling refrigerated loads, regular route drivers, and those with longer current trip distances. Drivers were interviewed at Florida Agricultural Inspection Station about their current trip (e.g., origin, location of next pick-up/drop, time/date for pick-up/drop, number of miles over previous 7 days, participation of co-driver). The authors developed an index of sched- ule tightness (DRVSPD) which consisted of the “average speed a vehicle would need to maintain to reach the next destination without being late and with the driver obeying the driving times mandated by HSR.” Previous driving time was “estimated by determining the time required to complete the driver-reported mileage driven over the previous seven days.” In the analysis, three scenarios were examined assuming average road speed limits: 55 mph, 60 mph, and 65 mph. Weekly driving time was estimated by assuming an average speed, and then determining the driving time necessary to cover the driver’s weekly mileage. While non-driving tasks (e.g., vehicle inspections) are an integral part of work for most drivers, the drivers in the sample were not questioned regarding non-driving work hours, to avoid raising their apprehension by asking too many questions related to HSR. An analysis of schedule-induced HSR and/or speed limit violations by drivers of tractor trailers. Schedules of 498 long-distance drivers Long-distance truck drivers The authors note that the potential for response bias is always present when determining the incidence of any illegal activity. While the study relied heavily on information from questioning truck drivers, the authors employed measures to mitigate the response bias problem (e.g., limited their questions). 1. Depending on the average speed limit scenario (i.e., 55, 60, 65 mph), between 17% and 30% of the drivers were found to have violation-suspect schedules and between 14% and 26% had schedules sufficiently demanding to be also judged as violation-inducing.” (Author’s Note: For a schedule to be considered violation-suspect or violation-inducing, a minimum average trip speed must be attained or exceeded. Therefore, a violation- inducing schedule would also be considered to be violation-suspect.”) 2. Using the 55 mph average speed limits, an estimated 26% of all drivers had violation- inducing schedules. Ignoring the 60- and 70-hour rules, 15% of the drivers would have had such schedules due to the demands of the current trip. For 13% of the drivers, nei- ther the 60- nor the 70-hour rules could be adhered to without violating speed limits. For 1% of the drivers (15 + 13 − 26)/2, both the current trip and the 60- or 70-hour rules result in violation-inducing schedules.” These findings suggest a lower incidence of vio- lation than was found in the Hertz study. 3. “Solo drivers were more likely than team drivers to have tight schedules due to the cur- rent trip.” 148 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

Findings Directly Related to HOS (include page references): 149 4. “Depending upon the average speed limit assumed, drivers with refrigerated loads were between 50% and 80% more likely than other drivers to have violation-suspect or violation-inducing schedules (differences statistically significant at the .01 or .05 levels). 5. “For all three average speed limits, drivers for regular route carriers were significantly more likely than drivers on irregular routings to have violation-suspect schedules (at the .01 or .05 levels). 6. Longer current trip distance was positively correlated with tight schedules. Drivers with current trips over 1,000 miles were between five and seven times more likely to have violation-inducing schedules than were drivers with trips not over 500 miles. 7. “Assuming the average roadway traveled has 55 mph speed limits, over a quarter of all drivers must violate HSR or speed limits.” 8. There is a trade-off between speed and driving time. “For example, at 45 mph average speed the average driver must drive 51 hours per week, compared with 38 hours if 60 mph is averaged. “ 9. The authors present the estimated mean total weekly work hours for various assumed average speeds, as well as the work hours at each quartile. A very large majority of long- distance drivers have more than 40-hour work weeks (82%, assuming average main- tained speeds of 50 mph). Assuming 50 mph average maintained speed, half the drivers work more than 65 hours weekly and one-quarter work over 81 hours. Driver Duration p. 37, “Depending upon the average speed limit scenario, between 17% and 30% of the drivers were found to have violation-suspect schedules and between 14% and 26% had schedules sufficiently demanding to be also judged as violation-inducing.” p. 37, “Using the 55 mph average speed limits, an estimated 26% of all drivers had violation- inducing schedules. Ignoring the 60- and 70-hour rules, 15% of the drivers would have had such schedules due to the demands of the current trip. For 13% of the drivers, neither the 60- nor the 70-hour rules could be adhered to without violating speed limits. For one per- cent of the drivers (15 + 13 − 26)/2, both the current trip and the 60- or 70-hour rules result in violation-inducing schedules.” p. 39, “There is a very strong and positive relationship between current trip distance and schedule tightness. For all three average speed limit scenarios, the simple correlation between DRVSPD and trip distance is positive (ranging between .31 and .35) and signifi- cantly different from zero at the .01 level… “For all average speed limit scenarios and for current rip, the 60/70 –hour rule, and total: the longer the journey, the greater the fre- quency of violation-suspect and violation-inducing schedules. In all cases, the differences are easily significant at the .01 level.” p. 40, “Depending upon the average speed limits assumed, drivers with current trips over 1,000 miles are between five and seven times more likely to have violation-inducing sched- ules than are drivers with trips not over 500 miles.” p. 41, “Assuming 50 mph average speed, the average driver works 58 hours in total and drives 46 hours weekly. Seventy-five percent of these drivers work over 49 hours and drive more than 39 hours per week. Half the drivers exceed 65 hours total work and, of that, drive over 52 hours. Finally, a quarter of the drivers work more than 81 hours and drive over 64 hours per week (which is not legal under HSR).” p. 41, “The findings indicate that the very large majority of long-distance drivers have more than 40-hour work weeks (82%, assuming average maintained speeds of 50 mph), and extremely lengthy work weeks are common. For example, assuming 50 mph average

150 maintaining speed, half the drivers work more than 65 hours weekly and a quarter work over 81 hours.” p. 41, “Solo drivers were found to have much higher frequencies of violation-suspect and violation-inducing schedules, than team drivers. Again assuming 55 mph average speed limits, 28% of the solo drivers, but only 11% of the team drivers, would have to violate HSR and/or speed limits to stay on schedule. Due to these differences, the analysis focused on solo drivers. Among solo drivers, groups with higher incidences of violation-inducing schedules were those driving longer trip distances, driving refrigerated loads, and driving regular route carriers.” Driver Health No significant findings or assumptions concerning impact on health.

Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references): 151 Dianne Davis, Alison Smiley Belenky, G., Wesensten, N.J., Thorne, D.R., Thomas, M.L., Sing, H.C., Redmond, D.P., Russo, M.B., and Balkin, T.J. “Patterns of performance degradation and restoration during sleep restriction and subsequent recovery: a sleep dose-response study” (2003). Journal of Sleep Research, Vol. 12, 1–12. The data reported in this study are a subset of data collected in a larger study analyzed and published as a U.S. Department of Transportation Report (Balkin et al. 2000; see Part I). This paper focuses on the findings for psychomotor vigilance task performance (PVT), sleep latency, and subjective sleepiness. According to the authors, of the measures taken in the larger study, the PVT was chosen for this paper because it was the most sensitive to the effects of sleep restriction and was the least subject to learning effects.” The purpose of the study was to empirically determine the effects of 3, 5, 7 and 9 hours of sleep over 7 con- secutive days on objective and subjective alertness and objective performance. In addition, the study looked at the extent to which 3 days of subsequent recovery sleep restored per- formance and alertness to baseline levels. Drivers had 3 days of orientation and baseline sleep in the laboratory before data collection commenced over 7 days of performance testing with 3, 5, 7, or 9 hours of sleep each night. The recovery period, that followed, lasted 4 days with 8 hours in bed each night. A wide variety of measures were utilized. Measures consisted of the PVT, the cognitive perfor- mance assessment battery, driving simulator tasks (e.g., lane tracking) as well as sleep latency, EMG, and sleepiness ratings. In addition to these measures, a number of health measures were taken (e.g., tympanic temperature, heart rate, and blood pressure). Study of effect of sleep restriction on performance. Sixty-six CMV drivers (16 females, media age = 43 years; 50 males, mean age = 37 years) CMV drivers • Study only looked at daytime driving. • Recovery sleep was restricted to 8 hours. • The trade-off for using a wide variety of measures was that the number of daily admin- istrations for each particular measure was restricted—precluding evaluation of circadian rhythms in this study. • Subjects were heterogeneous with respect to age, which may have contributed to error variance in performance measures. (See Balkin et al. (2000) in Part I.) 1. “Seven days of sleep restriction degraded psychomotor vigilance performance in a sleep- dose dependent manner. With mild to moderate sleep restriction (7- and 5-hr time in bed [TIB]), performance initially declined and, after a few days, appeared to stabilize at a lower-than-baseline level for the remainder of the sleep restriction period. In contrast, with severe sleep restriction (3-hr TIB) performance declined continuously across the sleep restriction period, with no apparent stabilization of performance. Sleep augmen- tation (9-hr TIB) had no effect on performance over the 7-day experimental period.” 2. Three days of recovery sleep did not restore performance to baseline levels for subjects with mild to moderate sleep restriction (5- or 7-hr TIB). Driver Fatigue/Alertness p. 9, “. . . it appears that the inflection point (i.e., the minimum amount of nightly sleep required to achieve a state of equilibrium in which daytime alertness and performance can

152 be maintained at a stable, albeit reduced, level) is approximately 4 hr per night. If less than 4 hr of sleep per night is obtained, daily decrements in performance capacity may be unavoidable – at least across a 7-day period of sleep restriction.” p. 10, “The present findings suggest that core sleep might best be considered as the mini- mum amount of sleep needed by the brain to achieve a state of equilibrium in which alert- ness and performance are maintained at a stable but lower-than-normal level. In this view, sleep durations that do not satisfy the core sleep requirement would, across days, result in continued degradation of alertness and performance relative to baseline, but degradation would not continue across days indefinitely—an asymptotic, stable level of reduced alert- ness and performance would eventually be achieved; and additional sleep (i.e., incremen- tal increases in the duration of sleep beyond the core requirement) would produce corre- spondingly higher, and stable, levels of alertness and performance.” p. 10, “Following chronic, mild to moderate sleep restriction (5- or 7-hr TIB), 3 days of recovery sleep (8-hr TIB) did not restore performance to baseline levels.” Driver Health No significant findings or assumptions concerning impact on health.

Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: 153 Dianne Davis, Alison Smiley Braver, E.R., Preusser, C.W., and Ulmer, R.G., “How long-haul motor carriers determine truck driver work schedules: The role of shipper demands.” (1999). Journal of Safety Research, Vol. 30, No. 3, pp. 193–204. The objective of this research was to identify determinants of drivers’ schedules. In partic- ular, the study looks at the “role of shipper demands within the load scheduling decision chain for individual drivers, as reported by dispatchers working for long-haul motor carri- ers.” Two hundred and seventy dispatchers participated in telephone interviews. Dispatch- ers’ reasons for accepting or rejecting loads from shippers were based on revenue (75%), delivery deadlines (24%) and the HOS status of the nearest driver (9%). However, dis- patchers reported that shippers ask for “sufficient time for drivers to adhere to both speed limits and hours-of-service rules.” Dispatchers were asked how the “time required to make a particular trip was determined.” Trip mileage was reported “as the key determinant of trip schedule assignments (58%)”; however, other factors were also considered, “including speed limits (27%) and past experience with particular routes (13%). The authors concluded that the results of the survey suggested that, “tight schedules cannot be attributed solely to shipper demands.” Long-haul drivers were surveyed at weigh stations in Wyoming and Tennessee about job characteristics (e.g., size of carrier, penalties for late delivery), and asked to identify who arranged their current loads. Interviewers eliminated a number of drivers (e.g., those who worked for private or terminal-to-terminal carriers), to focus on “U.S long-haul motor car- riers that make decisions concerning the acceptance or refusal of potential loads and that figure out driver delivery schedules based on a single truck, meeting the needs of an indi- vidual shipper for transport of a specific load.” Interviews were conducted with 270 of the 309 dispatchers identified by drivers. Dispatchers were asked a number of questions: “how they figured the time necessary for trips; how often shippers imposed penalties for late deliverables; what percentages of shippers asked for just enough time, not enough time, or more than enough time to pick up and deliver loads; and what factors affected their deci- sions to accept or reject loads from customers. Dispatchers were also asked if they used any computer program to estimate the time necessary for trips,” as these programs “can include criteria needed to comply with hours-of-service rules.” Focus on long-haul motor carrier dispatchers as to determinants of drivers’ schedules. 309 long-haul drivers; 270 long-haul motor carrier dispatchers Long-haul motor carriers Dispatchers reported that shippers rarely requested tight delivery schedules. However, there is a possibility that dispatchers may have responded to questions about tight delivery sched- ules according to typical driver work schedules rather than HOS regulations. In particular, the authors note, that the study “did not attempt to quantify how a dispatcher defined ‘more than enough time,’ ‘just enough time,’ or ‘not enough time’ to pickup and deliver. The drivers interviewed for this study were interviewed during summer morning hours at two specific sites. As a result, the authors note that “whether afternoon or nighttime truck traffic differs from morning traffic in terms of shippers’ requested delivery schedules, whether these sites differ from others in the United States, and whether there are seasonal variations in motor carrier characteristics at these sites cannot be assessed from existing data.” 1. Approximately 20% of drivers reported penalties (e.g., fines, suspension, demotion, reprimands) from their motor carriers for late deliveries.

154 2. Motor carrier dispatchers pointed to multiple sources of information to determine the time required to make a particular trip: computer programs (75%), miles to be traversed (58%), speed limits (27%), and experience (13%). 3. Two-thirds of the dispatchers said they used rules of thumb for the average speed driv- ers could travel. Thirty percent said specific routes were examined. 4. “Among all dispatchers (i.e., both using and not using rules of thumb) 18% reported using 50 mph or slower; 14% used 51–55 mph; 21% used 55–60 mph; and 14% used a speed in excess of 60 mph.” 5. “There was no association between the following variables and reports of expecting average speeds faster than 60 mph: penalties assigned for late delivery (reported by driver or dispatcher), hauling perishable products, and owner-operator status.” 6. 61% of respondents said that time allotted per shipment for non-driving duties were up to the driver. Of those giving a quantitative estimate, most expected 2 to 4 hours. 7. Few dispatchers reported penalties imposed by shippers for late deliveries although 20% of drivers reported penalties from their motor carriers for late deliveries. Approximately 60% percent of dispatchers said this never happened and 40% said it rarely happened. 8. In response to questions regarding shipper time frames, more than one-third of dis- patchers said 95% or more of their shippers gave more than enough time for deliver- ables to be made. Only 12% of dispatchers said that 10% or more of shippers give insuf- ficient time for pick up and delivery. More than 80% of dispatchers said that zero shippers request insufficient time. 9. Of the 233 of 270 dispatchers who had the authority to accept or reject loads, only 9% mentioned the HOS status of the nearest driver as a factor in their decision. 10. The authors conclude, “according to dispatchers, revenue generation is a primary deter- minant in decisions to accept or reject loads. Delivery deadlines and the HOS status of the nearest driver were cited much less frequently. Revenue, probably, is a strong influ- ence on delivery schedules in the very competitive trucking industry.” Driver Duration p. 199, “Dispatchers were asked about factors affecting their decisions to accept or reject loads (Table 5), and 233 of 270 said they had the authority to make such decisions. Among these 233 dispatchers, revenue was cited by 75%, followed by the credit rating of the ship- per (41%), the need to find a back haul (load for return trip; 26%), the delivery deadline (24%), and whether the shipper was a regular customer (19%). The hours-of-service sta- tus of the nearest driver was mentioned by 9% of respondents.” p. 201, “Federal and state efforts to decrease violation-inducing delivery schedules appear more likely to succeed if they continue to be directed primarily toward motor carriers. If government agencies start to monitor driver adherence to work hour limits effectively, then motor carriers will have no choice but to refuse shippers’ requests for unreasonable deliv- ery schedules.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references):

Dianne Davis, Alison Smiley Campbell, K. “Estimates of the prevalence and risk of fatigue in fatal accidents involving medium and heavy trucks.” 2002. University of Michigan Transportation Research Institute. This paper summarizes the results of a report on baseline estimates of the prevalence and risk of truck driver fatigue in fatal accidents that was prepared for the FMCSA in support of the HOS rulemaking. Data is presented on fatigue in fatal accidents by power unit type, trip distance, for-hire versus private carriers, time of day, and by hours driving. Fatigue was coded as a contributing factor for 511 truck drivers (1.9%) out of a total of 27,463 medium and heavy trucks involved in fatal accidents from 1991 to 1996. More than half of all fatigue-related fatal accidents involve for-hire tractors on trips of a one-way distance of more than 200 miles. “The risk of a fatigue-related fatal accident increases with trip dis- tance.” The fatigue risk “increases with hours driving in any operating environment and shows the characteristic circadian pattern.” The report is based on data from the UMTRI Trucks Involved in Fatal Accidents (TIFA) files and the 1992 Truck Inventory and Use Survey (TIUS) conducted by the Bureau of the Census. The data is used to provide estimates of the vehicle miles of travel and fatal acci- dents involving fatigue for various segments of the trucking industry. In addition, the inci- dence of fatigue accidents is combined with travel data to estimate the overall risk of fatigue in fatal accidents. The authors use the same definition of fatigue as coded in the Fatality Analysis Reporting System (FARS). Report looks at TIFA and TIUS data from 1991 to 1996. Truck driver fatigue is the depen- dent variable and power unit type, trip distance, for-hire versus private carriers, time of day, and hours driving are the independent variables for the analysis. Over 27,463 medium and heavy trucks involved in fatal accidents over the 6-year period Medium and heavy trucks The coding of fatigue is taken from the “driver-related factors” variables in FARS which relies on the original police accident report. The authors note that the coding of fatigue by state shows some large variations. In addition, the authors note that fatigue is “particularly difficult to assess, even with in-depth investigations, since there is no physical evidence of fatigue.” The authors suggest that the “prevalence of fatigue reported here is in all likeli- hood too low.” 1. Fatigue is coded as a contributing factor for 511 truck drivers (1.9%) out of a total of 27,463 medium and heavy trucks involved in fatal accidents from 1991 to 1996. 2. The prevalence of truck driver fatigue in fatal accidents by time of day for all medium and heavy trucks involved in fatal accidents from 1981 to 1996 follows the circadian pattern. 3. (p. 23) “. . . more than half of all fatigue-related fatal accidents involve for-hire trac- tor on trips with a one-way distance of more than 200 miles. The risk of a fatigue- related fatal accident increases with trip distance. Straight trucks have a substantially higher risk of a fatigue-related fatal accident when operated on trips outside the local area, as compared with tractor combinations. For-hire carriers have a greater risk of fatigue-related fatal accident involvement based on miles traveled in nearly all oper- ating environments as compared with private carriers. Finally, the fatigue risk increases with hours driving in any operating environment and shows the characteris- tic circadian pattern.” 4. The prevalence of truck driver fatigue was shown for six industry groups: Straight: Local, 50 to 200 miles, >200 miles; Tractor: Local, 50 to 200 miles, >200 miles. The Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: 155

156 greatest portion (62.3%) of the fatigue cases was in the category for tractors on trips greater than 200 miles. 5. The majority of reported fatigue occurs in the first few hours of driving as “half of all reported truck driver fatigue occurs in the first four hours of driving” for trucks involved in fatal accidents from 1981 to 1996. More than 25% of the accidents occurred in the first hour, and two-thirds in the first 4 hours. “Only about 4 percent of all medium and heavy truck drivers involved in a fatal accident reported driving more than 8 hours at the time of the accident.” The authors note that this pattern is driven “by exposure, not risk” as the “nature of the exposure distribution will always keep the number of accidents after many hours driving a small proportion of the total, even with dramatic increases in risk with hour driving.” (In other words, there are more 4-hour trips than 8-hour or 12-hour trips). Consequently there will be more accidents associated with 4-hour trips than with 12-hour trips. This is a different issue from the risk per 4-hour trip or per 12-hour trip— the per trip risk is higher for longer trips.) The authors also note that fatigue is cumula- tive and that while the amount of work and rest during the previous day and week also affect the level of fatigue during any hours of the current trip, no information on the pre- vious work schedule was available for this study. 6. The relative risk of fatigue given involvement in a fatal accident follows the circadian rhythm. 7. The relative risk of fatigue given involvement in a fatal accident gradually increases dur- ing the first 8 hours, doubles during the ninth hour and is higher by a factor of 6 by the 12th hour. Driver Fatigue/Alertness p. 3, “The reporting of fatigue was also examined for collisions where the only fatalities were truck occupants. Overall, 14% of all trucks involved in fatal accidents had one or more fatalities in the truck. Of the truck occupant fatalities, 9.5 percent (361 cases) were coded for truck driver fatigue. These fatigue cases are 70 percent of all fatigue coded for truck drivers in the TIFA data for 1991–1996. The proportion of fatigue in the non-truck fatalities is 0.6 percent.” p. 23, “Risk increases steeply after eight hours of driving. However, most of the fatigue- related fatal accidents occur during the first four hours of driving. This result may suggest that cumulative fatigue is a greater problem than hours driving on any given day, although exposure data are needed to confirm the result. The circadian pattern has a pervasive effect on both the risk and the prevalence of fatigue-related fatal accidents. The risk of a fatigue- related fatal accident is elevated by a factor of four in the early morning hours. Only driv- ing more than 10 hours produces comparable risk levels. Time of day and hours driving are the dominant risk factors. Distributions of these factors suggest that the lower risk of fatigue-related fatal accidents of private carriers operating long-haul tractors may be due to less nighttime driving and shorter driving hours.” Driver Duration p. 17, “The relative risk of fatigue gradually increases during the first 8 hours. During the ninth hour the fatigue risk is nearly double and by the 12th hour the risk is higher by a fac- tor of over 6. A pronounced increase is also shown in the fifth hour. Fatigue risk drops back below 1.0 during the sixth hour and increases with each additional hour. Aggregate risk for the second four hours is greater than the first four hours by a factor of 1.6. This pattern holds in every subset examined… While these results confirm the generally accepted fact that fatigue increases with time on duty, they also illustrate that time on duty is not the only factor. The time of day when each hour of driving takes place also influences the risk of Findings Directly Related to HOS (include page references):

fatigue. It is likely that there is a strong interaction between time of day and hours of driv- ing. The risk of fatigue when the eighth hour is driven at 4 a.m. is likely to be much higher than when the eighth hour is driven at 5 p.m.” Driver Health No significant findings or assumptions concerning impact on health. 157

158 Dianne Davis, Alison Smiley Folkard, S. and Lombardi, D.A., “Designing Safer Shift Systems.” (2004). In P. Nickel, K. Hanecke, M. Schutte and H. Grzech-Sukalo, (Eds.) Aspekte der Arbeitspsychologies aus Wissenschaft und Praxis. pp. 151–166. Lengerich: Pabst Science Publishers. This article reviews the literature on shiftwork safety and the “related evidence on sleep duration, fatigue, and performance capabilities.” It focuses on studies in which “real mea- sures of injuries and/or accidents can be related both to the time of day and/or to the point within the shift system that they occurred.” The studies look at specific features of shifts such as type of shift, length of shift, and number of successive shifts. The authors found a highly significant main effect of shift (i.e., risk increased on afternoon and night shifts rel- ative to the morning shift) and a consistent trend in accident risk over four successive nights (i.e., risk was progressively higher on 2nd, 3rd and 4th nights than first night). The authors also discuss measures of fatigue and performance capabilities and how “relative risk esti- mates might be combined or pooled in a simple manner to provide an overall estimate of the relative risk over any given span of shifts, and hence for an entire shift system.” This estimate would allow for a “comparison of the estimated risk on different shift systems and further facilitate the design of safer shift systems.” The authors used two forms of analyses to examine the trends discussed in the article. A repeated-measures analysis of variance based on the relative risk values calculated for each data set was used as well as a chi-square analysis based on the summed frequency of inci- dents, giving equal weight to injuries and accidents. Both forms of analyses were used to overcome the shortcomings associated with each form by itself. The authors note specific aspects of the data that were problematic. For example, they had to correct the data in some studies to take account of inequalities in the number of workers. In addition, while some stud- ies “give no precise details of the shift system in use, many of them involved a total of only four days on each shift before a span of rest days.” The authors also discuss the various prob- lems associated with the priori risk of accidents and injuries. Few published studies allow for an unbiased calculation of relative risk estimates of accidents and/or injuries associated with specific features of shift systems due to non-homogeneous a priori risk (e.g., number of individuals at work is not constant over 24-hour day; number of supervisors, etc.). Literature review of shiftwork safety and estimates of relative risk over a span of shifts. The number of articles reviewed was not stated. Shiftwork systems The authors note that there are “few published studies that allow for an unbiased calculation of relative risk estimates of accidents and/or injuries associated with specific features of shift systems due to non-homogeneous a priori risk.” As a result, their analyses are often based on one or two studies. They also note that “three of the studies report two separate sets of data, for different areas or types of incident, giving a total of eight data sets across the three shifts. Further, while some of the studies give no precise details of the shift system in use, many of them involved a total of only four days on each shift before a span of rest days.” 1. Both analyses yielded a highly significant main effect of shift. Based on pooled fre- quencies, risk increased on the afternoon and the night shift relative to the morning shift. 2. There is a consistent trend in accident risk over 4 successive night shifts. However, the authors are unsure what happens to risk over longer successive night shifts, as there is a paucity of data relating to this. Only two studies reported incidence rates for a span of more than 4 night shifts and both of these studies were based on a relatively small number of incidents. However, it is noteworthy that both studies reported a decrease in risk from the fourth to the fifth night shift, which was maintained until the seventh and final night shift in one of the studies. Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

159 3. The authors questioned whether the increase in risk over successive shifts is confined to the night shift, or whether it might be general to all shifts and represent an accumulation of fatigue over successive workdays. While there was no evidence of a main effect of successive shifts using the repeated measures analysis of variance, the chi-square test yielded a significant effect of successive shifts, however it was substantially smaller than that over successive night shifts. On average, risk was about 2% higher on the second morning/day, 7% higher on the third morning/day, and 17% higher on the fourth morn- ing/day shift than on the first shift. 4. Three studies that reported the trend in risk over successive hours on shift and were cor- rected for exposure in some way were used for an analysis of the main effect of time on shift. Risk increased in an approximately exponential fashion with time on shift. The repeated-measures analyses of variance yielded a highly significant main effect of time on shift. A chi-square test was not possible as each study had to correct for exposure in some way and thus combining raw frequency scores would be biased. 5. The authors estimated the relative risk of shifts of different lengths by “differentiating the trend shown in Figure 4” (i.e., the mean relative risk over hours of duty found in 3 studies) and this result is shown in Figure 5.” From Figure 5, “it is clear that variations Figure 4. The mean relative risk over hours on duty. Figure 5. The estimated relative risk on different lengths of shift.

160 in shift length from about 4 to 9 hours will have relatively little impact on overall safety.” They note, however, that the most important point is that they can “estimate the change in risk associated with shorter or longer shifts. They can estimate that relative to “8 hour shifts, 10 hour shifts are associated with a 13.0% increased risk and 12 hour shifts with a 27.5% increased risk.” 6. The authors note that this trend for hours on duty did not control for the influence of breaks during a duty period. Only one recent study examined the impact of rest breaks on the risk of incidents. Injuries in an engineering plan in which 15-minute breaks were given after each period of 2 hours of continuous work were calculated. “The risk in each 30-minute period was expressed relative to that in the first 30-minute period immediately fol- lowing the break.” The study found that risk rose substantially, and approximately lin- early, between successive breaks such that risk had doubled by the last 30-minute period before the next break. There was no evidence that this trend differed for the day and night shifts, or for the three successive periods of two hours of continuous work within a shift.” 7. The authors offer a model of how relative risk estimates might be combined or pooled in a simple manner to provide an overall estimate of the relative risk over any given span of shifts. The following figures show the estimated relative risks for different spans and lengths of day shifts (Figure 1) and night shifts (Figure 2): Figure 1. Estimated relative risks for different spans and lengths of Day shifts. Figure 2. Estimated relative risks for different spans and lengths of Night shifts.

Findings Directly Related to HOS (include page references): 161 Driver Fatigue/Alertness p. 5, “Based on these pooled frequencies, risk increased in an approximately linear fash- ion, with an increased risk of 18.3% on the afternoon shift, and of 30.4% on the night shift, relative to the morning shift…This finding suggests that when the a priori risk appears to be homogeneous across the three shifts, there is a consistent tendency for the relative risk of incidents to be higher on the afternoon shift than on the morning shift, and for it to be highest on the night shift.” p. 6, “There is also a consistent trend in accident risk over successive night shifts . . . On average, risk was about 6% higher on the second night, 17% higher on the third night, and 36% higher on the fourth night.” p. 7, “On average, risk was about 2% higher on the second morning/day, 7% higher on the third morning/day, and 17% higher on the fourth morning/day shift than on the first shift.” p. 8, “Thus while it remains a possibility that over longer spans of night shifts risk may actually start to decrease after the fourth night, there is no current evidence to indicate that this is actually the case.” p. 10, “A repeated measure analysis of variance of the relative risk values for the five data sets indicated that there was no evidence of a main effect of successive shifts. However, this may reflect the relatively small number of incidents and limitations of some of these stud- ies . . . A chi-square test based on the summed frequencies across the five studies for the four successive shifts yielded a significant effect of successive shifts. These summed values were used to estimate the risk on the successive morning/day shifts relative to the first such shift . . . On average, risk was about 2% higher on the second morning/day, 7% higher on the third morning/day, and 17% higher on the fourth morning/day shift than on the first shift . . . Clearly there is some evidence. albeit relatively inconsistent compared to the other trends reported in this chapter, that risk increased over successive morning/day shifts. However, it is important to note that this increase was substantially smaller than that over successive night shifts. Thus, there is evidence for an increase in risk over successive work- days, irrespective of the type of shift, but also evidence that this increase is substantially larger on the night shift than on the morning/day shift.” p.10, “. . . apart from a slightly heightened risk from the second to fifth hour, risk increased in an approximately exponential fashion with time on shift.” p.10, “. . . we can estimate that relative to eight hour shifts, ten hour shifts are associated with a 13/0% increased risk and twelve hour shifts with a 27.5% increased risk.” Driver Health No significant findings or assumptions concerning impact on health.

162 Dianne Davis, Alison Smiley Frith, W.J. “A case control study of heavy vehicle drivers’ working time and safety.” (1994.) Proceedings 17th ARRB Conference, Part 5:17–30. Queensland, Australia. This article describes a study on the risk of crash with respect to driving hours and other time intervals related to the driver’s working lives. A ‘case’ group of heavy vehicles involved in crashes were compared to a ‘control’ group of vehicles. The crash-involved drivers were 2.6 times more likely, as compared with non-crash involved drivers, to have driven 8 or more hours since the last compulsory 10-hour off-duty period (as recorded in the log book). They found no other “significant differences between the two groups with respect to other time intervals related to driving habits.” However, there was a “significant difference between the age distribution of the crash and control drivers with the crash drivers generally younger.” A ‘case’ group of heavy vehicles involved in crashes was compared with a ‘control’ group of vehicles. The ‘control’ group of vehicles was selected by police going to the scenes of crashes on as close as possible to the exact anniversary of the crash, at the same time of day and stopping the first heavy vehicle of a similar configuration to the crash involved vehi- cle. When possible, police selected vehicles traveling in the same direction. If, after an hour of waiting, no suitable vehicle came, they selected a vehicle of a different configuration. Matching “by configuration was possible in 143 out of 199 useable pairs.” Details of driv- ers’ hours for the ‘case’ vehicles were known from their log books. The police collected driving hour details from the ‘control’ vehicle drivers along with other vehicle and driver details. The study was carried out in New Zealand in urban and rural areas between June 1992 and July 1993. The crash dates ranged from 1988 through 1990. Comparison of heavy vehicle crashes with a control group 199 pairs of crash and control vehicles New Zealand heavy vehicles There is a lag of 2 years between the beginning of the survey of control drivers and the last of the crashes. However, the authors did not believe that there was any significant change in driving habits or the driving environment between the period during which the crashes occurred and the period over which the control survey took place. The authors note that as with other studies using log book data, the drivers’ log book sys- tem in New Zealand is open to falsification. As there were very few observations in which hours restrictions were violated, the authors suspect that falsification may have occurred and that the “probable effect of this in the data would be to shift, in both groups, some data which should rightly appear in illegal cells, into marginally legal cells.” They believe this effect “may be greater in the crash sample than the control sample.” 1. The crash-involved drivers were 2.6 times more likely, as compared with non-crash involved drivers, to have driven 8 or more hours since the last compulsory 10-hour off- duty period (as recorded in the log book). 2. Crash involvement tends to increase more steeply for smaller rigs than larger rigs. The authors note that this might reflect greater use of smaller rigs by tired drivers in con- gested environments. 3. The lengths of the following time periods related to drivers’ work were not found to be statistically significant related to crash risk: i. The total elapsed time since the driver’s last 24-hour off-duty period ii. On-duty hours worked since the driver’s last 24-hour off-duty time Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

iii. Driving hours worked since the driver’s last 24-hour off-duty time iv. On-duty hours worked ‘today’ since the driver’s last 10-hour off-duty time 4. However, the authors note “indications of a rough progression towards statistical sig- nificance as the work-related periods measured became more immediately related to the final driving period.” They note “significant changes in time in some of the above mea- surements may have been detected had a larger scale study been possible.” Driver Fatigue/Alertness p. 28 , “Crash involvement increases significantly after about 8 hours worked from the last 10 hour rest period.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references): 163

164 Dianne Davis, Alison Smiley Harris, W. “Fatigue, circadian rhythm, and truck accidents.” Theory, Operational Perfor- mance, and Physiology Correlates (ed. Mackie, R.) (1978). 133–46. NY, NY: Plenum Press. Truck accident data were analyzed to see if fatigue and circadian rhythm effects were pres- ent in truck accidents that seemed to be the result of failures in vigilance performance. Data was analyzed for three groups of drivers: dozing drivers, those who had had single-vehicle accidents, and those who had crashed into the rear end of other vehicles. The effect of fatigue was confirmed for each of the groups. The “circadian effect was observed for doz- ing drivers, about twice as many of whose accidents occurred between midnight and 0800 than in the other 16 hours of the day, and for single-vehicle accident drivers, about half of whose accidents occurred in the early morning hours.” Bureau of Motor Carrier Safety accident report data was analyzed. Data was analyzed for three groups of drivers: dozing drivers, those who had had single-vehicle accidents, and those who had crashed into the rear end of other vehicles. Different types of exposure data (e.g., expected durations of trips on which accidents occurred) were taken into account. Truck accident data analysis Bureau of Motor Carrier Safety accident data—various sample sizes Truck drivers n/a 1. The 1974 accident report data for 406 dozing truck drivers was analyzed. “Nearly twice as many accidents occurred in the second half of trips (65%) than in the first half (35%). The median trip duration was 7.5 hours; the median driving time for an accident was 4.4 hours. The fatigue hypothesis was supported by the data.” 2. The accident report data for 493 dozing truck drivers was analyzed. Analysis showed that the circadian effect was evident as the highest percentage of accidents occurred between 0400 and 0600 which corresponds to the occurrence of the lowest mean devi- ation in heart rate. In addition, “twice as many accidents occurred between midnight and 0800 (66%) than in the other 16 hours of the day (34%). Note: This effect is even more pronounced if exposure data (i.e., the relative number of trucks on the highway by time of day) are taken into account. 3. Single-vehicle accidents strongly showed the circadian effect as approximately 70% of the accidents occurred between midnight and 0800. In contrast, only 34% of other- vehicle accidents occurred in the early morning hours. 4. Two samples of randomly selected truck accident data from 1976 were analyzed: 226 single-vehicle accidents and 116 accidents where the truck crashed into the rear of another vehicle. The samples were selected irrespective of the driver’s condition at the time of the accident. The circadian effect was evident for the single-vehicle accident sample as 46% of the accidents occurred between midnight and 0800, almost “2.5 times as many as would be expected from the exposure data (19%). Approximately 70% of those drivers who checked “dozed at the wheel” occurred between midnight and 0800.” In contrast, only approximately 25% of the accidents for the other-vehicle accident sam- ple occurred between midnight and 0800. For the two samples of data, the “crossover” from “less than expected” to “more than expected” percentage of accidents occurred between the “fifth and sixth hours of driving time.” Approximately, “twice as many acci- dents occurred in the second half of the trips (67%) as in the first half (33%, irrespective of trip duration).” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

Driver Fatigue/Alertness p. 144, “The fatigue hypothesis was supported by the results of the analyses of the acci- dent report for each of the groups studied: dozing-driver, single-vehicle, and other-vehicle accidents.” p. 145, “The circadian effect was observed for both dozing drivers and the single-vehicle accident drivers.” Driver Duration p. 145, “The fatigue and circadian effects observed in the analyses of the accident data raise a question about the possible combined effects of long hours on the road and time of day. What is the relative likelihood of an accident for a driver who finds himself still on the road at 0400 or 0500 in the morning after, say, 8 hours of driving? There is some evidence that there is a combined effect from the heart rate data for relay drivers observed in the earlier study (Harris & Mackie 1972)… The evidence suggests that trips should not be scheduled so that drivers will still be on the road, after many hours of driving, at times of the day when their capabilities to attend to the driving task are at a low ebb.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references): 165

166 Dianne Davis, Alison Smiley Hartley, L. “Australian initiatives in managing fatigue in transportation.” Insurance Com- mission of Western Australia. (1999)(www.officeofroadsafety.wa.gov.au/Facts.papers/ initiatives_in_managing_fatigue.html). This paper contains a description of two Australian alternatives to the traditional hours of service regime. In addition, the paper contains a comparison of various studies on the impact of fatigue on heavy vehicle drivers. The authors compare Hartley et al.’s (1996) study on the impact of fatigue on heavy vehicle drivers in Western Australia (WA) (where there are no HOS regulations), with Williamson and Feyer’s 1992 study, which contained a comparable survey of drivers working under HOS in the Eastern States. For their study, Hartley and his colleagues interviewed 638 driv- ers and 83 transport companies at various locations in the state. A few years prior to this study, Williamson and Feyer conducted their survey in the Eastern States. Meta-analysis n/a Truck driving In comparing Williamson et al. and Hartley et al.’s data on drivers’ perception of fatigue, it should be noted that both studies asked rather different questions. The WA study used “one scale running through ‘always a problem’ to ‘never a problem.’ Williamson used two scales running through ‘every trip’ to ‘very rarely’ or ‘major problem’ to ‘no problem.’ Does HOS restrict fatigue? 1. The percentage of drivers exceeding 72 hours work in Williamson et al.’s survey is slightly in excess of the WA survey, which found 30% exceeded 72 hours work. 2. “To the extent to which the present WA data and Williamson et al.’s data are compara- ble, fewer WA drivers (4.4%) consider fatigue to be always a problem than do their East- ern States’ counterparts consider it to be a major problem or occurring on every trip (8.6 and 10.7%). And rather more WA drivers consider it to be never a problem (35%) as compared with Eastern States drivers (15%) who very rarely feel fatigued or consider it is no problem.” 3. According to the results of a comparison of the self-report data from Williamson et al. (1992) and Hartley et al. (1996), fatigue appears to be regarded as a more significant problem for the industry among Williamson’s Eastern States drivers than it does in the WA sample. 4. The authors list the following problems with HOS: • HOS prescribe what a driver should be capable of doing (i.e., no flexibility). • HOS regulations do not inform organizations about fatigue and safety. • HOS regulations permit no discretion for different freight tasks and environments. • HOS do not take account of the influence of the driver’s circadian cycle. • HOS regulations take no account of time zone changes. • HOS may restrict access to sleep. • There is no commercial incentive to restrict driving to HOS. Driver Duration p. 3, “It appears that WA drivers are no more likely to exceed weekly driving hours regu- lations than their Eastern States’ counterparts, despite enforcement of driving hours regu- lations in the Eastern States.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

p. 3, “To the extent to which the present WA data and Williamson et al.’s data are compa- rable, fewer WA drivers (4.4%) consider fatigue to be always a problem than do their East- ern States’ counterparts consider it to be a major problem or occurring on every trip (8.6 and 10.7%). And rather more WA drivers consider it to be never a problem (35%) as com- pared with Eastern States drivers (15%) who very rarely feel fatigued or consider it is no problem.” p. 3, “The Eastern States drivers were more likely to report that long driving hours and poor sleep were fatiguing than WA drivers, despite regulation of driving hours in the East- ern States.” Driver Health No significant findings or assumptions concerning impact on health. 167

168 Dianne Davis, Alison Smiley Hertz, R. P. “Hours of service violations among tractor-trailer drivers.” (1990). Accident Analysis & Prevention, Vol. 23, No. 1, 29–36. This study examines estimated HOS violations of more than 1-hour among long-distance truck drivers based on observed departure and arrival at inspection stations. Service viola- tions were estimated by observing the same long-haul tractor-trailers leaving “one inspec- tion station (Point A) and entering another (Point B), estimating their travel speed exclud- ing required rest, and estimating hours of rest taken on the basis of the observed travel time and the estimated speed.” The drivers used for the estimation were interviewed at the first inspection station and observed thereafter. Only truckers driving alone who reportedly did not plan to make an interim delivery or pickup stop prior to arrival at the second inspection station were included in the calculation. At “assumed trip speeds of 40 mph and 50 mph, 90% and 51% of the drivers, respectively, were in violation of the hours of service rules by more than one hour.” The authors estimated violations of the HOS regulations by comparing observed trip time between Washington and Minnesota “with the minimum legal amount of time it would have required to make the trip if the driver took all required rest.” The minimum legal trip time was calculated as the “sum of the driving hours between Washington and Minnesota (including time for stops other than required rest) plus the hours of required rest.” As the authors did not know actual driving hours and hours of rest they arrived at a range of esti- mates for the driving hours between Washington and Minnesota by assuming that the actual trip speed (excluding required rest but including other necessary stops) lay somewhere between 35 mph and 65 mph. The authors also made estimates of required hours of rest for each estimate of driving hours based on federal regulations allowing 10 hours of driving after 8 hours of off-duty time. To compute the hours of required rest, the number of driving hours the driver reported prior to Spokane was added to the driving hours estimated between Spokane and Minnesota. The authors added the corresponding estimates of driving time between Spokane and Minnesota to the required hours of rest to obtain the minimum legal trip time. Violations were assumed if the sum was more than 1 hour greater than the observed travel hours between the Spokane departure and the Minnesota arrival. Estimated HOS violations based on interviews and observations. 130 long-haul tractor-trailer drivers, driving alone Long-haul tractor-trailer The authors note that a “precise estimate of hours of service violations cannot be calculated without knowing the actual speed distribution of the drivers in the sample.” 1. 90% of the drivers were in violation of the HOS rules by more than 1 hour by the time they arrived at the Minnesota observation site with an assumed average trip speed of 40 mph. 2. 51% of the drivers were in violation of the HOS rules by more than 1 hour with an assumed average trip speed of 50 mph. 3. 24 hours rest was required by 81% of the drivers at an assumed speed of 40 mph com- pared with 18% of the drivers at an assumed speed of 50 mph. Driver Rest p. 34, “At an assumed trip speed of 40 mph, 16 hours of rest was required by 17% of the drivers, 24 hours of rest was required by 81% of the drivers, and 32 hours of rest was required by 2% of the drivers.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

169 p. 34, “At an assumed trip speed of 50 mph, 16 hours of rest was required by 82% of the drivers and 24 hours of rest was required by 18% of the drivers.” p. 34, “At an assumed trip speed of 40 mph, 36% of the drivers missed more than 70% of required rest, 61% missed more than 50% of the rest, and 88% missed more than 10% of the rest.” p. 34, “At an assumed trip speed of 50 mph, 9% of the drivers missed more than 70% of the required rest, 20% missed more than 50% of the rest, and 48% missed more than 10% of the rest.” p. 34, “At 40 mph, 42% of the drivers missed more than 15 hours of required rest, 81% of the drivers missed more than 5 hours of rest, and 90% of the drivers missed more than 1 hour of rest. p. 34, “At 50 mph, 7% of the drivers missed more than 15 hours of required rest, 36% of the drivers missed more than 5 hours of rest, and 51% of the drivers missed more than 1 hour of rest.” Driver Duration p. 33, “At an assumed average trip speed of 40 mph, 90% of the drivers were in violation of the hours of service rules by more than one hour by the time they arrived at the Min- nesota observation site. With an average trip speed of 50 mph, 51% of the drivers were in violation by more than one hour.” Driver Health No significant findings or assumptions concerning impact on health.

Dianne Davis, Alison Smiley Jones, I.S. and Stein, H.S. “Effect of driver hours-of-service on tractor-trailer crash involve- ment.” (1987). Arlington, VA. Insurance Institute for Highway Safety. In 1987, the Insurance Institute for Highway Safety in the United States used a case-control approach to examine the relative risk associated with long hours of driving (Jones and Stein 1987). For each large truck involved in a crash, three trucks were randomly selected from the traffic stream at the same time and place as the crash but 1 week later. A sample of 332 tractor-trailer crashes, each with one, two, or three case controls was extracted for analysis. For tractor-trailers, the authors found that driving in excess of 8 hours, drivers who violate log book regulations, drivers aged 30 and under, and interstate carrier operations were asso- ciated with an increased risk of crash involvement. In particular, the relative risk of crash involvement for drivers who reported a driving time in excess of 8 hours was almost twice that for drivers who had driven fewer hours. A case-control method of analysis was used whereby three trucks were selected and inspected for each crash-involved truck. The trucks were selected and inspected at the crash site at the same time of the day as the crash but 1 week later. In addition to information on truck weight, size, configuration, and type of carrier, data were collected on driver age and experience, hours of driving, and log book violations. Hours of driving was recorded as “the number of hours driven since the last eight hour rest period,” using a variety of sources (e.g., driver’s statement, log book record, bill of lading, current vehicle location, etc.). The study included 676 crashes involving 734 large trucks that occurred between June 1984 and July 1986. The analysis was limited to crashes involving tractor-trailers and their controls. The final subset represents 332 matched case-control data sets. Investigation of large truck crashes on interstate highways in Washington State. The study included 676 crashes involving 734 large trucks that occurred between June 1984 and July 1986. The analysis was limited to crashes involving tractor-trailers and their con- trols. The final subset represents 332 matched case-control data sets. Large trucks n/a 1. For tractor-trailers, the authors found that driving in excess of 8 hours, drivers who vio- late log book regulations, drivers aged 30 and under, and interstate carrier operations were associated with an increased risk of crash involvement. 2. The relative risk of crash involvement for drivers who reported a driving time in excess of 8 hours was almost twice that for drivers who had driven fewer hours. 3. The risk from driving long hours increases for drivers operating between 12:01 a.m. and 6:00 a.m. and 6:01 a.m. and noon. Driver Fatigue p. 15, “What is important from this study is that the relative risk of crash involvement for drivers driving more than eight hours is almost twice that for drivers with fewer hours behind the wheel… the risk from driving long hours increases for drivers operating between 12:01 a.m. and 6:00 a.m. and 6:01 a.m. and noon.” Driver Health No significant findings or assumptions concerning impact on health. 170 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

171 Alison Smiley, Dianne Davis Jovanis, P.P. and Kaneko, T. “Exploratory analysis of motor carrier accident risk and daily driving patterns.” (1990). Transportation Research Group, University of California at Davis. Research Report UCD-TRG-RR-90-10, 1990. The Jovanis and Kaneko (1990) report deals with the issue of cumulative days or hours of driving and crash rates. They report on an analysis of carrier-supplied accident and non- accident data for a 6-month period in 1984. The data were obtained from a “pony express” type operation, which operates coast to coast with no sleeper berths. Cluster analysis was used to identify nine distinct patterns of driving hours over a 7-day period. The driving patterns of drivers who had an accident on the 8th day were compared with drivers who had no accident on the 8th day. These patterns reflected times of day of most frequent onduty and driving time, the most frequent offduty times, the mean and standard deviation of the total hours onduty for the 7 days, the mean and standard deviation of consecutive hours driven per driver and the mean and standard deviation of the driving cycle. Their study indicates an increased accident risk for night drivers after 3 to 4 days of driving but less concern for daytime drivers with respect to a crash immediately following a 3- to 4-day period of driving. The data were obtained from a “pony express” type operation, which operates coast to coast with no sleeper berths. Cluster analysis was used to identify nine distinct patterns of driv- ing hours over a 7-day period. The driving patterns of drivers who had an accident on the 8th day were compared with drivers who had no accident on the 8th day. These patterns reflected times of day of most frequent onduty and driving time, the most frequent offduty times, the mean and standard deviation of the total hours onduty for the 7 days, the mean and standard deviation of consecutive hours driven per driver and the mean and standard deviation of the driving cycle. An analysis of carrier-supplied accident and nonaccident data and driving patterns. Approximately 1,600 accident- and nonaccident-involved drivers Less-than-truckload firm with no sleeper berths The authors did not include statistics on offduty times in excess of 24 hours. They assumed that a substantial recovery occurs when a driver is offduty in excess of 24 hours, after reach- ing the DOT limit of 60 hours in 7 days. 1. Analysis showed that there was an increased accident risk for night drivers after 3 to 4 days of driving but less concern for daytime drivers with respect to a crash immediately following a 3- to 4-day period of driving. 2. Accident risk on the eighth day is shown to be consistently higher for the four patterns involving infrequent driving the first 3 to 4 days followed by regular driving during the last 3 to 4 days, than the 4 patterns with the reverse arrangement. This suggests cumula- tive fatigue from driving over 3 to 4 days does occur and leads to increased accident risk. 3. When the patterns are examined in detail, it appears that drivers who begin their trips near midnight and typically end them around 10:00 a.m. face a particularly increased crash risk after driving for several consecutive days. In contrast, drivers who typically drive a regular daytime schedule (10 a.m. to 6 p.m.) show little evidence of any effect due to continuous driving. 4. Total hours of driving in a 7-day period varied between averages of 54 to 59 hours among the nine patterns identified. Summary only available—no direct quotes. Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

Dianne Davis, Alison Smiley Kecklund, G., Akerstedt, T., Ingre, M., Soderstrom, M. “Train drivers’ working conditions and their impact on safety, stress and sleepiness: a literature review, analyses of accidents and schedules.” (1999). National Institute for Psychosocial Factors and Health (IPM). Department of Public Health Sciences, Division of Psychosocial Factors, Karolinska Insti- tute, Stockholm, Sweden. This paper focuses on how train drivers’ work hours and work situations influence safety and performance, “in particular in traffic intense areas, such as commuting.” It contains a summary of the results of a literature review, an accident analysis, and an analysis of the timetable for train drivers located in the Stockholm area. The accident analysis examines 79 rail accidents, finding that approximately 17% were potentially related to fatigue or sleepiness. The authors conducted an analysis of investigations of 79 accidents and near-accidents occurring between 1980 and 1997 on the railway in Sweden. The investigations largely focused on the train driver’s role in the events leading up to the accident. The goal of the analysis was to examine the extent to which working conditions could have contributed to the accidents/near-accidents. While most investigations did not contain information on whether stress or fatigue was involved at the time of the accident, they did contain infor- mation about when the incident occurred as well as the type of incident. The authors used a combination of factors to determine if fatigue was present at the time of the accident. Indicators of suspected fatigue were considered to be present when one of the following three criteria appeared in combination with the fourth criterion: 1. The driver admitted or the investigator observed fatigue. 2. Time of the accident (the incident occurred between 3:00 a.m. and 6:00 a.m.). 3. Lack of sleep (less than 5 hours sleep) or a shift being preceded by a brief period of off-duty time (less than 11 hours). 4. Accidents or incidents characterized by missed signals, lack of attention or loss of memory. It is known that this type of event is frequently triggered by fatigue. An analysis of accident and near-accident investigations. 79 accident and near-accident investigations were analyzed Train drivers The authors note that their results must be interpreted with caution as the investigations they analyzed frequently contained relatively sparse information on the working conditions at the time of the incident. As a result, they were required to infer if fatigue or stress was pres- ent at the time of the accident (see Methodology) but admit that this involves methodolog- ical shortcomings as there is a risk of incorrect classification. In addition, the authors note that as the number of accidents is low, the results should be interpreted with caution. 1. “Three (4%) of the accidents were fatigue related according to criterion no. 1. However, this figure rose to 13 (17%) if all criteria were applied.” 2. “Five of the accidents where fatigue was suspected were due to a brief period of sleep after an extremely early morning shift, three were due to the driver being at the lowest level of the circadian rhythm, and two to a short rest period and in all likelihood a brief period of sleep.” 3. “The occurrence of incidents is at its lowest in the evening (between 10 p.m. and mid- night) and at night (between 2 a.m. and 6 a.m.). Twelve percent of the incidents occurred 172 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

173 at night, between midnight and 6 a.m. Rather more than half (54%) of them occurred during morning traffic (starting before 9 a.m.). 4. “Most incidents (62%) occurred during the dark winter months, defined here as October through March.” 5. “Seventy-five percent of the accidents occurred during the first three working hours. However, it should be noted that the sample is very small and there is serious risk that chance has influenced the result. It is reasonable to suppose that very few shifts indeed are shorter than 3 or 4 hours, although we do know that relatively many shifts are shorter than 7 hours. If it had been possible to calculate the risk of an accident occurring (i.e., take into account the number of shifts that are at least 3, 4, 5, 6, etc., hours long), the peak after 2–3 hours would probably be considerably lower, while the risk of accidents after 6–7 hours would be higher.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references):

Dianne Davis, Alison Smiley Lenne, M.G., Triggs, T.J., and Redman, J.R. (1997). “Interactive Effects of Sleep Depriva- tion, Time of Day, and Driving Experience on a Driving Task.” Sleep, Vol. 2, No. 1, 38–44. This study looked at the effects of sleep deprivation, time of day, and driving experience on a driving task. Twenty-four subjects (12 experienced and 12 inexperienced drivers) drove a driving simulator for 20 minutes at 5 different times of the day on 2 testing days. One test- ing day took place after a normal night’s sleep. The other testing day took place after 1 night of sleep deprivation (SD). Reaction time, lateral position and speed were all assessed. “The standard deviation of both lateral position and speed were significantly higher during SD. Performance steadily improved across the day between 0800 and 2000, and the absence of any sleep-by-time interactions suggests that the rhythm of driving performance across the day was similar after both normal sleep and SD. Inexperienced drivers had higher RTs than experienced drivers in both sleep-deprived and non-sleep deprived conditions.” Twelve inexperienced drivers (driver’s license less than 3 years) and twelve experienced drivers (i.e., drivers’ license for between 6 and 13 years) were recruited to participate in the study. Subjects participated in 2 testing days in which they drove the simulator for 20 min- utes at 0800, 1100, 1400, 1700, and 2000. The first testing day took place after a normal night’s sleep. The second testing day took place after 1 night of SD. Subjects completed visual analog scales (i.e., measured subjective alertness, sleepiness, motivation) immedi- ately prior to and after each session. The between-subject variables were sleep condition, time of day, and block number with each session. The within-subject variable was the level of driving experience. Simulator study of experienced and experienced drivers. 24 subjects; 12 female and 12 male (18 to 32 years of age). n/a The authors note that it could be argued that the improvements in performance observed across the day may reflect practice effects. However, they noted that their previous work suggests that this is not the case. “When driving ability was measured across the day, with testing times counterbalanced, performance was significantly more impaired late at night and in the early morning hours, with a steady improvement across the waking day.” 1. “Inexperienced drivers had higher secondary RTs than more experienced drivers. There was also an interaction between driving experience and sleep condition. Analysis of sim- ple main effects confirmed that inexperienced drivers had higher RTs than experienced drivers in both the control and SD conditions. However, the presence of the interaction suggests that inexperienced drivers were impaired to a greater extent in the SD condition.” 2. “There were no effects of driving experience for any of the driving performance measures.” 3. Sleep deprivation for up to 36 hours significantly reduced the ability of all drivers to maintain a steady position in the lane and a stable speed. 4. Performance varied significantly across the day. 5. The absence of any interactions between sleep condition and time of day suggest that the patterns of performance across the day were similar after 1 night of sleep and 1 night of SD. 6. The results of the subjective measures of alertness, sleepiness, and motivation were con- sistent with the performance data to a large extent. 7. Differences between inexperienced and experienced drivers were found with the sec- ondary RT task. 174 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

175 Driver Fatigue/Alertness p. 42, “Sleep deprivation for up to 36 hours significantly reduced the ability of both inex- perienced and experienced drivers to maintain a steady position in the lane and a stable speed. The ability to attend to additional stimuli, as measured by the secondary RT task, was also significantly reduced by SD, particularly for inexperienced drivers.” p. 42, “A major finding of this study was that performance varied significantly across the day. The absence of any interactions between sleep condition and time of day suggests that the patterns of performance across the day were similar after a night of sleep and a night of SD. During both control and SD conditions, the ability to maintain both a stable posi- tion on the road and a constant speed improved across the day, while secondary RT decreased. In particular, during SD, subjects performed better at 2000 hours (36 hours SD) than at 0800 hours (24 hours SD). Similar performance rhythms during SD were found for lexical decision, vigilance, logical reasoning, reaction time, and memory and search tasks. p. 43, “Differences between inexperienced and experienced drivers were found with the secondary RT task.” p. 43, “Performance during SD is not a monotonic function of the length of deprivation. This study has shown that performance still improved across the normal waking day fol- lowing a night without sleep. These findings are relevant to those who are placed in situa- tions where driving after a period of SD is unavoidable, as in the transportation industry.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references):

Dianne Davis, Alison Smiley Lin, T., Jovanis, P.P., and Yang, C. (1994). “Time of day models of motor carrier accident risk.” Transportation Research Record 1467. The objective of this paper was to develop a quantitative method to analyze the effect of time of day on accident risk. A time-dependent logistic regression model was formulated to assess the safety of motor carrier operations. The model “estimates the probability of hav- ing an accident at time interval, t, subject to surviving (i.e., not having an accident) until that time.” The model was then tested with data from trucking company operations. Analy- sis showed that driving time had the strongest direct effect on accident risk. The model was used with data obtained from a national less-than truckload firm that oper- ated coast-to-coast with no sleeper berths. The total number of observations used for mod- eling was 1,924 cases, of which 694 were accidents and 1,230 were non-accidents. Creation of time-dependent logistic regression models to assess the safety of motor carrier operations. 1,924 cases (i.e., 694 accidents and 1,230 non-accidents) were used for modeling Less-than-truckload; no sleeper berths While “time-dependent covariates play a key role in accident analysis,” the authors note that the “shortage of time-varying data makes it difficult for a researcher to consider fur- ther accident analysis and solutions.” For example, high traffic volume could be one of the reasons for the highest accident risk occurring between 4:00 and 6:00 p.m. The inclusion of time-varying risk factors, such as road class “could greatly improve understanding of time-related effects.” The authors also note that the “joint study of time of day and driving time is complicated because driving time intervals could cross more than one time of day. Although some rules have been provided in this research, the approach is still rough and could result in some loss of information and bias in estimation. A more advanced approach is needed to treat the cod- ing of time of day precisely and completely. 1. “Driving time has the strongest direct effect on accident risk.” 2. “The first 4 hr consistently have the lowest accident risk and are indistinguishable from each other.” 3. “Accident risk increases significantly after the 4 hr, by approximately 50 percent or more, until the 7th hr. The 8th and 9th hr show a further increase, approximately 80 and 130 percent higher than the first 4 hr.” 4. “Drivers with more than 10 years of driving experience retain consistently low accident risk; all other categories of driving experience have a significantly higher risk.” 5. “Daytime driving, particularly at the noon time (10:00 a.m. to 12:00 noon), results in a significantly lower risk of an accident.” This was defined as the baseline. 6. The accident risk of driving during 4:00 to 6:00 p.m. is significantly higher (approxi- mately 60% higher) than that of the baseline.” This may result from a combination of two effects: evening rush hour and an association with reduced alertness because of a low circadian period for some drivers. 7. “The accident risks from midnight to 2:00 a.m., 6:00 to 8:00 a.m., and 8:00 to 10:00 p.m. are also significantly higher than during the baseline (but at P < .10). Two of these peri- ods involve night driving; the other involves part of the dawn period.” 8. Rest breaks, particularly those taken before the 6th or 7th hr of driving, appear to lower accident risk significantly for many times of day. 176 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

177 Driver Duration/Crash Risk p. 7, “Driving time has the strongest direct effect on accident risk. The first 4 hr consistently have the lowest accident risk and are indistinguishable from each other. Accident risk increases significantly after the 4th hr, by approximately 50 percent or more, until the 7th hr. The 8th and 9th hr show a further increase, approximately 80 and 130 percent higher than the first 4 hr.” p. 7, “Time of day had an effect on subsequent accident risk, but the effect was not as strong as for driving experience or driving hours. Daytime driving, particularly at noon (10:00 a.m. to 12:00 p.m.), results in a significantly lower risk of an accident. Driving from 4:00 to 6:00 pm. has an accident risk about 60 percent higher than the baseline; drivers during the other three significant times of day also have accident risks about 40 percent higher than those during the baseline. These three involve night or dawn driving; two of them are asso- ciated with circadian rhythms.” p. 7, “When interactions were included, the accident risk for some times of day decrease. Particularly, most of the significant interactions fall in the sixth and seventh driving hours. Rest breaks appear to be associated generally with these risk reductions.” Driver Health No significant findings or assumptions concerning impact on health. Findings Directly Related to HOS (include page references):

Dianne Davis, Alison Smiley Moore-Ede, M. and Schlesinger, B.I. “Scientific basis for challenges to work-rest & hours- of-service regs.” (2005). Submitted to 2005 Fatigue Management in Transportation Oper- ations International Conference, FMCSA and Transport Canada, Seattle, WA. The authors argue that scientific knowledge that has accumulated in the past 30 years “under- mines the legitimacy of the 20th-century paradigm of hours-of-service and duty-rest regula- tion.” They argue that “workplace risk relating to work and rest hours is actually multi- factorial, and that simplistic regulations based upon only one or two factors have limited value in minimizing this risk.” The authors list over 30 factors that determine level of sleepi- ness and fatigue-related accident risk. They note that the most important factor is circadian phase followed by the number of consecutive number of hours spent continuously awake since the previous sleep episode. Other important factors are the length of the sleep episode, although there can be significant inter-individual variation in terms of cumulative level of sleep deprivation over the past week, the quality of sleep that is obtained, job workload, and moment to moment stimulants or depressants of alertness. The authors use the tale of “two truck drivers” to illustrate the problem with the current (11-hour driving) HOS regulations. In particular, they note the disincentive for drivers to take a nap when they are tired as day- time naps are only allowed to be excluded from a driver’s hours on duty in certain situations when the nap is followed by driving, which is in turn immediately followed by an extended period of rest. The authors suggest that the efficacy of “alternative, less punitive, risk man- agement strategies based upon the science of fatigue management” should be demonstrated to provide the basis of regulatory reform and discuss two case studies that their company, Circadian, has worked on. The authors suggest alternative paradigms to the current work- rest regulations: fatigue management programs, fatigue risk models, alertness monitors. The authors use a number of case studies to support their arguments that (1) HOS regula- tions based on only one or two factors have limited value in minimizing risk, (2) there are problems with the current regulations, (3) alternative, less punitive, risk management strate- gies should be demonstrated to provide the basis of regulatory reform, and (4) the current regulations have an impact on accident litigation. In addition to discussing the various fac- tors that determine level of sleepiness and fatigue-related accident risk, they briefly discuss alternative paradigms to HOS regulations that minimize the risks associated with excessive employee work hours and fatigue. Case studies examining the problems with current HOS, regulatory reform, and accident litigation. Also includes a very brief description of alternatives to work-rest regulations such as fatigue management programs, fatigue risk models, and alertness monitors. n/a Truck drivers Paper uses case studies to illustrate their views on the limitations of HOS regulations but do not describe in detail the scientific basis for their challenges to work-rest and HOS regulations. No specific findings are discussed. Driver Duration/Crash Risk No specific findings are discussed. Driver Health No significant findings or assumptions concerning impact on health. 178 Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

179 Dianne Davis, Alison Smiley National Transportation Safety Board. “Factors that affect fatigue in heavy truck accidents. Volume 1.” (1995). NTSB Number: SS-95/01. NTIS Number: PB95-917001. This study consists of an analysis of single-vehicle heavy truck accidents by the Safety Board to examine the role of drivers’ patterns of duty and sleep in fatigue-related heavy truck accidents. The Safety Board focused on the sleep patterns of the 96 hours preceding 107 single-vehicle heavy truck accidents in which the driver survived. Fifty-eight percent of the accidents were fatigue-related. Fatigue was considered a probable cause of the acci- dent if the driver was estimated to have been on duty for more than 15 consecutive hours (the current legal limit), and if the driver’s performance involved non-professional, irra- tional actions such as failure to brake or make appropriate steering maneuvers. A statisti- cal analysis determined that the most important measures predicting a fatigue-related acci- dent in this sample were the “duration of the last sleep period, the total hours of sleep obtained during the 24 hours prior to the accident and the split-sleep patterns.” As a result of the study, the Safety Board made safety recommendations regarding sufficient rest pro- visions (i.e., at least 8 continuous hours of sleep after driving 10 hours or being on duty 15 hours), and scheduling (i.e., prohibit employers, shippers, etc., from scheduling a ship- ment which would require that the driver exceed the HOS regulations in order to meet the delivery deadline). This study examines the factors that affect driver fatigue rather than the statistical incidence of fatigue. As a result, the Safety Board selected truck accidents that were likely to include fatigue-related accidents, such as single-vehicle accidents that tend to occur at night. As they were specifically interested in the 96-hour duty-sleep history prior to the accident, they focused only on those single-vehicle accidents where the driver survived and could recon- struct the previous 96 hours. Some 113 single-vehicle heavy truck accidents in which the driver survived were investigated. As the 96-hour duty/sleep history was not available for 6 drivers, the authors focused on the data from the remaining 107 single-vehicle heavy truck accidents. A “multivariate statistical analysis (a multiple discriminate analysis) was per- formed to simultaneously evaluate the relationship of a set of measures of the drivers’ duty and sleep times to the groupings of accidents established by investigator’s determination of probable cause (fatigue-related and nonfatigue-related accidents).” An analysis of the 96-hour duty-sleep history prior to 107 single-vehicle heavy truck accidents. 107 single-vehicle heavy truck accidents Heavy truck drivers The definition of fatigue may be unduly restrictive, given that it involves driving at least 16 hours, and drivers may be fatigued before this point. 1. Of the total, 58% had fatigue as a probable cause, while the remainder were considered not fatigue-related. 2. Nineteen of the 107 drivers stated that they fell asleep while driving. 3. The most important measures in predicting a fatigue-related accident were the duration of the last sleep period, the total hours of sleep obtained during the 24 hours prior to the accident, and split sleep patterns. (There were no clinical tests to determine if sleep dis- orders were a factor.) 4. The truck drivers in fatigue-related accidents were found to have obtained an average of 5.5 hours sleep in the last sleep period prior to the accident. This was 2.5 hours less than the drivers involved in nonfatigue-related accidents (8.0 hours). 5. Truck drivers with split sleep patterns obtained about 8 hours sleep in total in a 24hour time period; however, they obtained it in small segments, on average of 4 hours at a time. Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

6. A major study conclusion was that “driving at night with a sleep deficit is far more crit- ical in terms of predicting fatigue-related accidents than simply night-time driving.” 7. The study found that many of the accident-involved drivers did not recognize that they were in need of sleep and believed that they were rested when they were not. The authors note that “about 80 percent of the drivers involved in fatigue-related accidents rated the quality of their last sleep before the accident as good or excellent.” Driver Fatigue/Alertness p. 2, “The statistically significant analysis determined that the most important measures in predicting a fatigue-related accident in this sample are the duration of the last sleep period, the total hours of sleep obtained during the 24 hours prior to the accident, and split sleep patterns.” Driver Health No significant findings or assumptions concerning impact on health. 180 Findings Directly Related to HOS (include page references):

181 Dianne Davis, Alison Smiley National Transportation Safety Board. “Fatigue, alcohol, other drugs, and medical factors in fatal-to-the-driver heavy truck crashes. Volume 1.” (1990). NTSB Number: SS-90/01. NTIS Number: PB90-917002. This study investigated fatal-to-the-driver heavy truck accidents to assess the role that alco- hol and other drugs played. The authors looked at a 1-year period (October 1, 1987, through September 30, 1988) in eight states. The authors found that fatigue and fatigue-drug interac- tions were involved in more fatalities in this study than alcohol and other drugs of abuse alone. The authors examined 182 accident investigations involving 186 heavy trucks that resulted in 207 fatalities, conducted in California, Colorado, Georgia, Maryland, New Jersey, North Carolina, Tennessee, and Wisconsin. As one case was eliminated from the analysis, the study analyses 181 heavy truck accidents involving 185 case vehicles and drivers. Fatigue was considered a probable cause of the accident if the driver was estimated to have been on duty for more than 15 consecutive hours (the current legal limit), and if the driver’s per- formance involved non-professional, irrational actions such as failure to brake or make appropriate steering maneuvers. The Safety Board contract with the Centre for Human Toxicology (CHT) “provided for tox- icological analyses of up to 250 blood samples and 75 urine samples for the drugs in the analytic plan. Safety Board investigators obtained “biological specimens for toxicological testing by CHT in 168, or 91 percent, of the 185 case drivers. No specimens were obtained by Safety Board investigators in 17 cases. In an “additional 16 cases, specimens were of insufficient quantity to test for all drugs on the analytic plan. The Board chose to include in the analysis cases in which CHT testing was carried out for most, but not all, drugs in the analytic plan. The authors note that “the cutoff concentrations for screen and confirmation tests required by DOT regulation are substantially different from the cutoff concentrations used in this study.” “While the DOT sensitivity concentrations apply to urine tests and the NTSB con- centrations apply to blood tests, the substantially higher cutoff concentrations for the DOT drug testing regulations are a concern to the Safety Board. High cutoff concentrations are too limiting to allow for a complete performance assessment decrement. In general, urine measurement cannot be used to establish that impairment is present. A drug blood concen- tration is required. However, under certain circumstances, urine measurement may be used, although with less reliability.” Urine was used in approximately 8 of the 168 cases where biological specimens for toxicological testing were available by CHT. The remainder of the cases used blood at NTSB’s threshold levels which they noted had substantially lower cut- off concentrations as compared with DOT regulations. ”The different cutoff concentrations indicated the different purposes for which the DOT standards and this study were developed. If the DOT regulation concentrations had been used for post-accident testing in this study, many of the drug of abuse positive (DOAP) driv- ers would not have been detected.” An analysis of heavy truck accident investigations during a 1-year period. 182 accident investigations Heavy truck drivers Authors do not specify, in synopsis, how they identified when fatigue was a probable cause for an accident. Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations:

1. “Thirty-three percent of the fatally injured drivers tested positive for alcohol and other drugs of abuse.” 2. “There is a significant relationship between drug positive test results and the day of the week. Saturday, Sunday, and Monday are the days with the highest percentage of drug positive tests.” 3. “The most frequently cited accident probable cause or factor in fatal-to-the-driver heavy truck accidents was fatigue (57 cases or 31 percent), followed by alcohol and other drug impairment (53 cases or 29 percent). Of the 57 drivers who were fatigued, 19 were also impaired by alcohol and/other drugs.” 4. The authors found that there was a strong association between violation of the federal HOS regulations and drug use. In addition, there is a significant relationship between drug positive test results and a shipment deadline for the load being carried. Driver Fatigue/Alertness p. 3, “The most frequently cited accident probable cause was fatigue (57 drivers or 31 per- cent) followed by alcohol and other drug use impairment (53 drivers or 29 percent).” Driver Duration p. 3, “There is a strong association between violation of the Federal hours of service reg- ulations and drug usage.” p. 3, “There is a significant relationship between drug positive test results among profes- sional drivers and a shipment deadline for the load being carried.” Driver Health p. 3, “The driver’s medical condition caused, or contributed to 10 percent of the accidents. Over 90 percent of medical condition related accidents involved some form of cardiac inci- dent. This calls into question the effectiveness of the Federal program to assure the proper medical qualification of commercial vehicle drivers.” p. 3, “Older drivers are less likely to have tested positive for drugs, but are more likely to have had an incapacitating medical incident.” 182 Findings: Findings Directly Related to HOS (include page references):

183 Dianne Davis, Alison Smiley Park, S., Mukherjee, A., Gross, F., and Jovanis, P.P. “Safety implications of multi-day driv- ing schedules for truck drivers: Comparison of field experiments and crash data analysis.” Transportation Research Board 2005 Annual Meeting. This paper examines the “effect of multi-day driving and continuous driving (time on task) on crash risk. The study uses pre-existing crash data from the 1980s and measurements from the Driver Fatigue and Alertness Study (DFAS) conducted in the mid-1990s. The research explores “whether a more detailed examination of time of day of driving, particularly over multiple days, indicates associations with crash risk.” Night and morning driving and irreg- ular schedules with primarily night and early morning driving, have elevated crash risk of 20 to 70%, 30 to 80%, respectively, compared with daytime driving. Crash data from 1984 through 1985 from a national less-than truckload firm was used for this analysis. At the time of data collection, the company operated coast to coast, with no sleeper berths. The total sample size was 5,050 drivers (i.e., 954 accident-involved drivers and 1,506 non-accident drivers in 1984; 887 accident drivers, and 1,604 non-accident driv- ers in 1985). Multi-day driving schedules were identified using previous research as well as extracting driving schedules from the DFAS (Wylie et al. 1996). The authors used the data to create two models which were used to examine the data. Analysis of pre-existing crash and non-crash data. 5,050 drivers from a national less-than truckload firm (1984, 1985). Less-than-truckload; no sleeper berths n/a 1. The time of day of driving was significantly associated with increased crash risk. Those drivers who had predominately night and early morning schedules had 20 to 70% higher crash risk than drivers in the baseline regular daytime driving schedule. Overall, 16 of 27 night and early morning driving schedules had elevated risk. 2. Drivers with irregular schedules also have an elevated crash risk (30 to 80% higher). 3. There is a finding of “increased crash risk associated with hours driving, with risk increases of 30-over 80% compared to the first hour of driving. These increases are less than previously reported and are of similar magnitude to the risk increases due to multi- day schedules.” 4. The authors note “there is some evidence, although it is far from persuasive, that there may be risk increases associated with significant off-duty time, in some cases in the range of 24 to 48 hours. The implication is that “restart” programs should be approached with caution.” Driver Duration/Crash Risk p. 14, “Among the schedules that involved night driving and no days off immediately prior to the day of interest, 9 (schedules C1, C12, C13, C16, C17, C20, C25, C27, C38) out of 12 schedules have elevated risk. Drivers with one or two days off immediately prior to the day of interest have elevated risk in 3 (C7, C8, C32) of 7 cases; and, drivers with irregular schedules have elevated risk in 4 (C9, C39, C40, C42) of 8 cases. These detailed compar- isons further highlight the elevated risk posed by night driving compared to the baseline regular daytime driving.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

p. 15, “ There is also evidence that even as much as a 24 hour off-duty period may not be sufficient to alleviate the elevated risk of night and early morning driving. Driving sched- ules C7 to C9 (averaging about 100 drivers in each group) involve drivers with night and early morning driving and include large amounts of off-duty time one or two days prior to the day of interest; all show elevated crash risk. A similar result appears for schedule C32, although the sample size is only 19 drivers. These findings raise questions about the effi- cacy of a “restart” period (Smiley and Heslegrave, 1997); there appears to be evidence from this analysis that 24 and perhaps 48 hours may be insufficient, particularly for night and early morning driving. Further, the elevated risk associated with schedules C34 and C35 indicate that two days off duty prior to driving may actually elevate risk, compared to more regular schedules even for day time driving. This may be due to the relative unfamil- iarity of driving a heavy vehicle again, or other personal factors, but the evidence exists for those driving at night as well as during the day.” p. 15, “Examining the findings in the context of the HOS implemented in 2004 in the U.S., there appears to be support for the changes in regulations that sought more regular sched- ules. Several of the driving schedules with the highest relative crash risk (e.g., C38, C39, C40) involved irregular schedules. While the sample size in each group was small, the increase in relative risk was large and significant. Previous studies using smaller crash data sets were unable to identify this important effect.” Driver Health No significant findings or assumptions concerning impact on health. 184

185 Dianne Davis, Alison Smiley Roehrs, T., Burduvali, E., Bonahoom, A., Drake, C., and Roth, T. “Ethanol and Sleep Loss: A “Dose” Comparison of Impairing Effects.” (2003). Sleep, Vol. 26, No. 8, 981–985. This study looks at the effects of sleep loss (0, 2, 4, and 8 hours of sleep loss) as compared with those of ethanol ingestion in 32 adults (ages 21 to 35). “The study was conducted in a mixed design with a between-subject factor, ethanol or sleep loss, and a within-subject fac- tor, dose of either ethanol or sleep loss.” The authors found that “sleep loss was more potent than ethanol in its sedative effects but comparable in effects on psychomotor performance. Ethanol produced greater memory deficits, and subjects were less aware of their overall per- formance impairment.” Thirty-two adult volunteers (21 to 35 years old) were randomly assigned to a sleep loss (n = 12) or ethanol (n = 20) group. “The ethanol arm of the study was conducted in a double- blind fashion.” Sleep loss participants had 8, 6, 4, and 0 hours time in bed which produced 0, 2, 4, and 8 hours of sleep loss. Participants in the ethanol group ingested 0.0 g/kg, 0.3 g/kg, 0.6 g/kg and 0.9 g/kg ethanol from 8:30 a.m. to 9:00 a.m. after 8 hours of time in bed the previous night. “Each participant received his or her 4 doses of ethanol or sleep loss in a Latin square design with 3 to 7 days between doses.” Subjects completed the Multiple Sleep Latency Test (MSLT) (9:30 a.m., 11:30 a.m., 1:30 a.m., 3:30 p.m., and 5:30 p.m.) and a performance battery (10:00 a.m., 12:00 Noon, 2:00 p.m., and 4:00 p.m.), which con- sisted of memory, psychomotor vigilance, and divided attention tests. “The order in which subjects underwent the ethanol or sleep-loss does was determined by a Latin square design with 3 to 7 days for recovery between doses.” Study of the risks associated with sleep loss relative to risks of ethanol. 32 adult volunteers (21 to 35 years old) n/a n/a 1. Sleep loss was “more potent than ethanol in its sedative effects but comparable in effects on psychomotor performance.” 2. Ethanol “produced greater memory deficits, and subjects were less aware of their over- all performance impairment.” 3. Sleep loss was at least as potent as ethanol in its performance-impairing and amnesic effects: • Central reaction time was slowed by sleep loss with 8 hours and 6 hours of time in bed (TIB) differing from 0 hours of TIB. In addition, tracking deviations were increased by sleep loss with 8 hours of TIB differing from 0 hours of TIB. • Sleep loss produced an increase in lapses (PVT) with the 8 hours of TIB differing from 0 hours of TIB. Driver Fatigue/Alertness p. 984, “. . . at the studied doses, sleep loss is at least as potent as ethanol in its performance- impairing and amnesic effects and is significantly more potent in its sedative effects.” p. 984, “In terms of sedative effects as measured by the MSLT, sleep loss was 2.7 times more potent, meaning that 8 hours of sleep loss is equivalent to 2.16 g/kg of ethanol and 2 hours of sleep loss is equivalent to 0.54 g/kg.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

p. 984, “In vigilance and divided-attention performance, sleep loss and ethanol were equipotent. In memory, ethanol was slightly more potent.” p. 984, “Increasing sleep loss was perceived as increasingly impairing, while only the high- est ethanol does was rated as impairing.” p. 985, “ While sleep loss and ethanol produced equal impairment on the performance tests, at the low and medium ethanol doses, participants did not perceive that impairment. Only at the high dose was performance impairment perceived.” p. 983, PVT change scores for number of lapses: “Sleep loss produced an increase in lapses (F=3.19, P<.04) with the 8 hours of TIB differing from 0 hours of TIB.” p. 984, DAT measures—tracking deviations and central reaction times: “Central reaction time was slowed by sleep loss (F = 6.20, P < .002), with 8 hours and 6 hours of TIB differ- ing from 0 hours of TIB. The ingestion of ethanol did not alter central reaction time . . . Tracking deviations were increased by sleep loss (F = 4.35, P < .01), with 8 hours of TIB differing from 0 hours of TIB. Ethanol ingestion also increased tracking deviation with the 0.3-g/kg dose differing from the 0.9-g/kg dose. Both ethanol (F = 9.25, P < .01) and sleep loss (F = 4.32, P < .05) produced linear dose effects. Ethanol ingestion and sleep loss did not differ in effects on tracking deviations, which the relative potency analyses also reflected (NS).” p. 984, “While both performance tests used in this study have previously been shown to be sensitive to sleep-deprivation effects, these tests are relatively short (10 and 15 minutes), and longer tests may have revealed a greater potency of sleep loss compared to ethanol. As to MSLT sensitivity, in previous studies from this laboratory, the MSLT has consistently been found to be more sensitive to the effects of ethanol compared to performance testing.” p. 984, “Ethanol and sleep loss were equipotent in impairing psychomotor performance at the studies does. Tracking deviation on the DAT were increased to the same extent by both ethanol and both sleep loss, while the reaction-time parameters on this task did not show consistent effects. Subjects often concentrate on 1 component of the task at the expense of the other, which in this case was the tracking component. On the PVT, which does not require divided attention, reaction times were affected. But, interestingly, on this task, both ethanol and sleep loss slowed the fastest reaction times, parenthetically to the same degree and in a dose-related linear fashion, while lapses and the slowest reaction times were not consistently affected. This is not supportive of the “lapse” hypothesis of sleep-deprivation effects, which suggests lapses in performance occur as one becomes sleepier. What these data show is that best performance is degraded.” Driver Health No significant findings or assumptions concerning impact on health. 186

187 Alison Smiley, Dianne Davis Saccomanno, F.F, Shortreed, J.H., and Yu, M. (1996). “Effect of Driver Fatigue on Com- mercial Vehicle Accidents.” In Truck Safety: Perceptions and Reality (ed. Saccomanno, F. and Shortreed, J.) (1995). The Institute for Risk Research, University of Waterloo, Water- loo, Canada. A Canadian study, by Saccomanno et al. (1996), used several different databases to deter- mine accident risk associated with different driving times. Databases included police acci- dent reports, a commercial vehicle survey of driver demographics, work hours, and routes. The focus of the study was fatigue-related accidents, defined as single-vehicle accidents that occurred between midnight and 8:00 a.m. or single-vehicle accidents where the driver was recorded as being at fault. Fatigue-related accident risk was significantly higher for routes characterized by long driving times, that is, where the 85th percentile driving time was 9.5 hours or longer. There were more single-vehicle accidents at night (assumed to be associated with fatigue) than during the day. There was a higher proportion of single- vehicle accidents on routes typified by long driving times. In remote regions, the nighttime single-vehicle accident rates were particularly high—13 times greater than for more popu- lated areas in the daytime. This study used “surrogate measures of fatigue derived from the accident data and surveys of drivers who reported on their driving time, to compare accident rates between locations and times for different types of fatigue.” Four databases from the Province of Ontario were used in this analysis: Ontario motor vehicle accident data, Ontario Highway Inventory Man- agement System, Ontario Traffic Volume Information System, and Ontario Commercial Vehicle Survey (CVS). The CVS data, which provided information on commercial vehi- cles surveyed at 75 representative locations in the provincial highway network, were key to the fatigue analysis as these data provided direct evidence on the hours of driving from the last rest stop for each sampled truck driver in the traffic stream. Database analysis to compare accident rates between locations and times for different types of fatigue. 1988–1989 truck accident data Trucks n/a 1. Fatigue-related accident risk was significantly higher for routes characterized by long driving times, that is, where the 85th percentile driving time was 9.5 hours or longer. 2. There were more single-vehicle accidents at night (assumed to be associated with fatigue) than during the day. The results confirm the presence of circadian fatigue. 3. There was a higher proportion of single-vehicle accidents on routes typified by long driv- ing times. 4. In remote regions, the nighttime single-vehicle accident rates were particularly high— 13 times greater than for more populated areas in the daytime. This points to the effect of circadian fatigue on truck accident rates appearing to be additive to the effect of indus- trial fatigue. Driver Fatigue/Alertness p. 170, “Longer hours of driving without rest result in significantly higher fatigue accident rates. An appreciable increase in rates (i.e., a pronounced discontinuity in the relationship) was found to occur for more than 9.5 hours of driving without proper rest.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings: Findings Directly Related to HOS (include page references):

p. 171, “The results of the study confirm the presence of circadian fatigue in the Ontario truck accident data by indicating higher fatigue-related accident rates at nighttime as com- pared to daytime over the entire highway network (northern and southern regions).” p. 171, “The effect of circadian fatigue on truck accident rates appears to be additive to the effect of industrial fatigue. In the northern region, where longer driving is expected to result in industrial fatigue, the nighttime single vehicle accident rate is 3.3 times higher than the daytime rate. A similar relationship was found in the southern region with a ratio of 2.3 night accident rates to day accident rates.” Driver Health No significant findings or assumptions concerning impact on health. 188

189 Dianne Davis, Alison Smiley Van Dongen, Hans P.A., Maislin, G., Mullington, J.M., and Dinges, D. “The cumulative cost of additional wakefulness: Dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation.” (2003). Sleep, Vol. 26, No. 2, 117–126. Forty-eight adults participated in a laboratory study of chronic sleep restriction or total sleep deprivation. Subjects in the chronic restriction experiment were randomly assigned one of three sleep doses (4 h, 6 h, or 8 h time in bed per night), which were maintained for 14 con- secutive days. Subjects in the total sleep deprivation experiment had 3 nights without sleep (0 time in bed). Both experiments had 3 baseline (pre-deprivation) days and 3 recovery days. The researchers found that “chronic restriction of sleep periods to 4 h or 6 h per night over 14 consecutive days resulted in significant cumulative, dose-dependent deficits in cog- nitive performance on all tasks.” Lapses in behavioral alertness and reductions in working memory performance, in the 4 h condition, reached levels equivalent to those observed after 2 nights without sleep. After 14 days of sleep restriction, cognitive throughput performance was equivalent to that observed after 1 night without any sleep. Forty-eight healthy adult subjects participated in a chronic sleep restriction experiment or a total sleep deprivation experiment. Both experiments began with 1 adaptation day and 2 baseline days with 8 h sleep opportunities. In the chronic sleep restriction experiment, this was followed by randomization to 8 h, 6 h or 4 h sleep periods (time in bed ending at 0730) for 14 days. In the total sleep deprivation experiment, subjects were kept awake for 88 h. Both experiments concluded with 3 recovery days. Subjects in all experimental conditions (e.g., psychomotor vigilance task, Stanford Sleepiness Scale, Karolinska Sleepiness Scale) underwent neurobehavioural assessments every 2 h during scheduled wakefulness. In addi- tion to neurobehavioural measures, polysomnographic (PSG) recordings were made during the third baseline sleep period and during 10 of the 14 restricted sleep periods. Chronic sleep restriction and total sleep deprivation. 48 healthy adults (ages 21 to 38) n/a n/a 1. “Chronic restriction of the nocturnal sleep period to either 6 h or 4 h per day for 14 days resulted in significant cumulative performance deficits relative to the 8 h sleep period condition.” 2. “Subjects allowed an 8 h sleep period per night displayed only minor, non-significant increases in lapses of behavioral alertness over the 14 days.” . . . “In contrast, subjects in the 4 h sleep period condition displayed escalating numbers of lapses in behavioral alertness and decreasing cognitive accuracy and speed across the 14 days. The magni- tude of changes in performance over days of sleep restriction in the 6 h sleep period was between that observed in the 8 h and 4 h sleep period conditions.” 3. “In the 4 h sleep period condition, lapses in behavioral alertness and reductions in work- ing memory performance reached levels equivalent to those observed after 2 nights with- out sleep. Cognitive throughput performance after 14 days of sleep restriction was equiv- alent to that observed after 1 night without any sleep. Subjects in the 6 h sleep period condition also reached levels of impairment equivalent to those observed after 1 night of total sleep loss for lapses in behavioral alertness and working memory performance.” 4. “Chronic restriction of the nocturnal sleep period to either 6 h or 4 hr per day for 14 days resulted in a relatively small but significant build-up of subjective sleepiness, as measured with the Stanford Sleepiness Scale (SSS) relative to the 8 h sleep period condition.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

5. “Cumulative total sleep time increased near-linearly over days in the 8 h, 6 h and 4 h sleep period conditions. 6. “Cumulative sleep loss over 14 days in the 4 h sleep period condition was significantly greater than cumulative sleep loss over 3 days in the total sleep deprivation condition.” 7. The findings of this study contradict the “core sleep” hypothesis. The “core sleep” hypothesis asserts that “core” or “obligatory” sleep occupies the first part of the night and serves to “repair the effects of waking wear and tear on the cerebrum.” In this hypothesis, all sleep obtained beyond this core sleep (especially that dominated by SWS and SWA) duration is considered to be “optional” or “facultative” sleep. As SWS and SWA were conserved among sleep restriction conditions in this study, the finding that cumulative cognitive impairment developed “in cerebral functions at 4 h and 6 h time for sleep per night indicates that the current threshold of 6 h for core sleep duration can- not be correct. If 6 h sleep per day were the maximum duration of sleep required to main- tain normal cerebral functions, cumulative cognitive performance deficits should not have developed in that condition. Thus, the results from the present study do not support a functional distinction between “core” and “optional” sleep.” Fatigue/Alertness p. 122, “. . . the two modes of sleep loss yielded similar maximum deficits for PVT perfor- mance but chronic sleep restriction resulted in much greater cumulative sleep loss than did 3 days of total sleep deprivation. p. 124, “Contrary to earlier, uncontrolled studies of prolonged sleep restriction, this exper- iment yielded convergent findings of sleep dose-response effects on all three cognitive per- formance functions. Sleep periods chronically limited to 4 h and 6 h per night progressively eroded the effectiveness of psychomotor vigilance performance, working memory perfor- mance and cognitive throughput performance, providing convergent evidence for the adverse effects of chronic sleep restriction on cognitive functions . . . Claims that humans adapt to chronic sleep restriction within a few days on the other hand, are not supported by the present findings.” p. 124, “Since chronic restriction of sleep between 4 h and 6 h per night for 14 days pro- duced cognitive performance deficits comparable to those found under conditions of 1 to 2 days of total sleep deprivation, it appears that even relatively moderate sleep restriction— if sustained night after night—can seriously impair waking neurobehavioural functions in healthy young adults.” p. 124, “We conclude that the effects of sleep chronically limited to 4 h and 6 h per night on cognitive performance appear to reflect progressive neurocognitive dysfunction in sys- tems underlying sustained attention and working memory.” p. 124, “. . . unlike performance measures, sleepiness ratings appeared to show adaptation to chronic partial sleep deprivation . . . These findings for subjective sleepiness suggest that once sleep restriction is chronic, subjects either cannot reliably introspect with regard to their actual sleepiness levels, or as long as they are receiving at least approximately 4 h of sleep nightly they do not experience a sense of sleepiness anywhere near the levels found for total sleep deprivation.” p. 124, “Measures of sleep physiology were less responsive to chronic sleep restriction than were waking neurobehavioural functions. The primary effects on sleep architecture were immediate, overall reductions in the amounts of stages 1, 2 and REM sleep.” Driver Health No significant findings or assumptions concerning impact on health. 190 Findings Directly Related to HOS (include page references):

191 Dianne Davis, Alison Smiley Williamson, A., Friswell, R. and Feyer, A.M. “Fatigue and performance in heavy truck drivers working day shift, night shift or rotating shifts.” (2004). National Transportation Commission. This field study was designed to compare the impact of day and night shift rosters on sub- jective fatigue, performance, and the sleep and work of drivers. In addition to permanent day shift and night shift drivers, drivers working alternating weeks of day and night shifts participated in the study. Drivers participated for 2 weeks to “attempt to obtain a reliable sample of their work.” Over the 2 weeks, each driver was measured repeatedly. In addition to completing tests of concentration and reaction speed, drivers completed performance tests, kept diaries of their work, break, and sleep times and completed ratings of their sub- jective fatigue. Actigraph data was also collected to provide objective measures of the tim- ing and quality of their sleep. The researchers found that while the night shifts made driv- ers feel more tired than day shifts, it did not “produce significantly poorer performance, suggesting that night drivers can manage their fatigue.” Fifty-four Australian drivers participated in the study: 22 permanent day shift drivers, 21 permanent night shift drivers, and 11 rotating shift drivers. Each driver was repeatedly mea- sured over a 2-week period. Drivers completed concentration and reaction speed tests at the start of the first shift of the study fortnight (baseline), and at the end of the last shift in week 1 and week 2. Drivers also self-administered the Simple Reaction Time test as well as the Macworth Clock Vigilance task (shortened version) at the start and end of each shift dur- ing the fortnight, as well as at the start of one midshift break in each shift. In addition to these performance tests, drivers kept diaries of their work, break, and sleep times and com- pleted ratings of their subjective fatigue and quality of sleep. Actigraphy data was also col- lected to complement the self-report measures. Impact of day and night shift rosters on subjective fatigue, performance, and sleep and work on long-haul drivers. 54 male professional long-distance drivers (22 permanent day shift drivers, 21 permanent night shift drivers, and 11 rotating shift drivers) Professional long-distance drivers Two practical and methodological issues limited the current study: driver recruitment and missing data. The authors had to relax their initial recruitment specifications to select groups of drivers on the basis that they did a particular shift. Instead the final sample included “any driver working permanent day or night shifts that were rostered to be 11 or more hours long.” “As a result, in order to obtain a sufficiently large sample of drivers who did day shifts, the study also included drivers who worked rotating day shifts.” According to the authors, “missing data posed a serious problem for the current study, limiting the type of data analyses that could be conducted and the strength of conclusions that could be drawn.” 1. Over a typical workweek of five consecutive 10- to 12-hour shifts, there was a signifi- cant increase in subjective ratings of fatigue by all drivers. 2. Between the start and end of each shift within the workweek, rated fatigue also increased. Permanent night shift drivers and drivers on rotating night shifts showed greater increases than day drivers. However, it is important to note that the permanent night shift drivers had lower fatigue ratings at the start of their shift than permanent day drivers. 3. The study failed to find significantly poorer performance for night drivers as compared with day drivers. Simple reaction time test performance tended to be slower at the end of the week for all drivers. Similarly, there were no differences between day and night dri- vers on subjective fatigue ratings. The results of the PVT suggest a slowing in reaction Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

speed across the work week for night drivers while day drivers showed faster response time at the end of the week; however, this was non-significant. While the authors note that this lack of difference could be due to the testing time for night drivers, who tended to complete their end-of-week tests around dawn, this would not explain the absence of differences between rotating day drivers and night shift drivers who completed their end- of-week testing earlier in the morning during the circadian low. 4. The authors note that an alternative explanation for failing to find the predicted effect for night drivers could be that the “work-rest pressures were more important than the circa- dian influences experienced in night work.” All of the drivers worked similar long hours (e.g., 50 to 55 hours arranged in five 10- to 12-hour shifts), which may eclipse any effects due to circadian or time of day influences. Drivers had between 4 and 6 hours sleep between work shifts. The authors note that all of the drivers in this study “were being affected by restricted sleep and that any differential effects of night work may be over- shadowed by this effect.” In addition, they note that night shift drivers performed as well as day shift drivers as they are experienced professional drivers “who are well-suited to cope with the demands of the road transport industry by organizing their work-rest.” 5. Night shift drivers worked longer shifts than day shift drivers and spent much more of their working time driving than day shift drivers “which might predict that night shifts would be more tiring than day shifts.” However, the authors suggest that night shift driv- ers may have performed as well as day shift drivers as they may be “especially tolerant of fatigue or skilled at managing fatigue” and because they organize their sleep differ- ently (e.g., napping in the hours before their first shift of the week) which may partly explain how they could maintain performance. For example, the authors noted that night drivers “endeavored to capitalize on the sleep propensity influences of the circadian rhythm by getting as much sleep as they could as close as possible to the early morning circadian trough when sleep is most likely.” Driver Fatigue/Alertness p. 43, “For permanent shift drivers, 2 (group) by 2 (occasion) repeated measures MANOVA showed a significant occasion effect, with the overall subjective fatigue rating higher at the end of the working week compared to baseline. Although the mean scores suggest a greater increase in rated fatigue across the week for night shift drivers, no statistically significant difference in overall subjective fatigue was found between day shift drivers and night shift drivers.” p. 45, “Repeated measures MANOVA showed that permanent day and night shift drivers did not differ on any of the Simple Reaction Time task measures. RT tended to be higher at the end-of-week compared to baseline but there was no significant interaction on any perfor- mance measures between type shift and test occasion. These results indicate that both day and night drivers showed slowing of reaction speed over the week but to an equal extent.” p. 46, “On average, there was no significant difference between permanent day shift driv- ers and permanent night shift drivers on any of the Mackworth Clock Vigilance perfor- mance measures.” p. 48, “There were no significant differences in any PVT performance measures between the beginning of the week and the end of the selected study week. Analysis of interaction effects showed two non-significant trends for interaction between driver group and test occasion.” Driver Health No significant findings or assumptions concerning impact on health. 192 Findings Directly Related to HOS (include page references):

193 Dianne Davis, Alison Smiley Williamson, A., Feyer, A.M., Friswell, R., and Leslie, D. “Strategies to combat fatigue in the long-distance road transport industry: Stage 2 evaluation of alternative work practices.” (2004). National Transportation Commission. (Note: This study was covered in Part I in Williamson et al. (1996) “The impact of work practices on fatigue in long distance truck drivers.” Accident Analysis & Prevention, Vol. 28, No. 6, 709–71. The following presents findings not covered in Part I.) A repeated measures design was used in this study. Twenty-seven subjects participated in each of three work practices: staged trip driving (two drivers from different points of ori- gin meet mid-trip and exchange loads, within HOS regulations), flexible trip driving (sin- gle driver, trip scheduled without reference to HOS regulations) and single trip driving (single driver, within HOS regulations) A range of fatigue measures were used including performance tests, physiological, and subjective measures. The authors found that “a 10–12 hour trip is tiring no matter how the work is organized, and that the effects of accumulated fatigue may overshadow the effects of fatigue on a single 10–12 hour trip.” A sample of 27 drivers participated in a repeated measures design. Each driver participated in three work practices: staged driving (within HOS), flexible trip driving (outside HOS), and single trip driving (within HOS). The three methods were compared on a 10- to 12-hour route between Sydney and Melbourne. Each driver was assessed before the trip, on the road, and after the trip for all three work practices. A range of fatigue measures were used includ- ing performance tests (cognitive performance tests, on-road performance test), physiolog- ical (e.g., hear rate), and subjective measures (e.g., Stanford Sleepiness Scale). Monitoring of drivers’ cognitive and physiological functioning, as well as monitoring of their driving performance, were on-road measures. “The equipment was designed to obtain data in real time without interfering with the driving task, and allowing the driver to use his regular vehicle.” In addition to a questionnaire (e.g., demographics, driving experience) adminis- tered prior to the trip, drivers were asked to complete a trip diary including driving details (e.g., breaks) as well as their ratings of feelings of fatigue. A repeated measures design comparison of trips within HOS regulations: staged trip driv- ing and single trip driving; and driving outside HOS: flexible trip driving. 27 long-distance truck drivers (mean age: 38.4; driving experience: 15.9 years) Professional long-distance drivers The authors note that the analysis of the results of this study was hampered to some extent by missing data, particularly in the data collected during the trip. When only a section of data is missing in a repeated measures design, cases can be rejected and the “power of the study to detect differences between groups when differences actually exist” is weakened. As a result, this may lead to more conservative conclusions. 1. Drivers experienced higher subjective fatigue at the end of the trip compared with the beginning for all trip types. 2. While staged drivers reported higher fatigue at the beginning of staged trips compared with the other trip types, this “most likely reflects the cumulative impact of the previous week’s work which was typical of the schedules routinely worked by these drivers. Fatigue levels were also higher for staged drivers at the end of the trip compared to the ratings at the end of the trip for the other trip types, suggesting that if a driver starts the trip more tired, he is likely to be more fatigued at the end of the trip.” 3. “The results of this study show that overall there was relatively little difference between the trip types in the effects on the drivers’ performance.” Reviewers: Title: Abstract: Methodology: Scope of Work: Sample Size: Industry Sector: Major Limitations: Findings:

4. Some performance tests did show poorer performance for staged strips compared to con- trol trips. These results suggested that when doing staged trips drivers did not handle tasks requiring prolonged attention as well as when they were doing the other trip types. The conclusion cannot be drawn, however that the unique characteristics of staged driving created this inferior performance.” 5. “For the vigilance and unstable tracking tasks, drivers showed poorer performance when on staged trips than on either of the other trips.” 6. “Performance for the CFF test also revealed poorer performance by drivers on staged trips, but only for the beginning of the trip.” 7. There were no differences related to type of trip or the time in the trip that the test occurred for the simple reaction time test or for the on-road reaction time test. 8. “On staged and control trips, heart rate decreased across the trip, indicating decreasing alertness. In contrast, when on flexible trips heart rates were much slower at the begin- ning of the trip, but increasing such that by the end of the trip they had much faster heart rate than the other trip types. This suggests that when on flexible trips, drivers had lower alertness, based on the heart rate measure, at the beginning of the trip but their alertness increased by the end of the trip.” Driver Fatigue/Alertness p. 3, “Drivers experienced higher subjective fatigue at the end of the trip compared to the beginning for all trip types indicating that the experience of driving for 10 to 12 hours was tiring, no matter how the work was organized.” p. 3, “. . . drivers reported higher fatigue at the beginning of staged trips compared to the other trip types. This most likely reflects the cumulative impact of the previous week’s work which was typical of the schedules routinely worked by these drivers.” p. 3, “Fatigue levels were also higher for staged drivers at the end of the trip compared to the ratings at the end of the trip for the other trip types, suggesting that if a driver starts the trip more tired, he is likely to be more fatigued at the end of the trip.” p. 6, “Factors other than experiences during a 12 hour trip must be considered as causes of fatigue in these drivers. The results of this study suggest that factors leading to chronic fatigue, such as heavy work load over the past few days may account for differences in fatigue levels of drivers. Of the three types of trips, the staged trips were most vulnerable to the effects of work in the past week due to their order in the study. The impact of aspects of work organization was clearly revealed in the finding that on staged trips drivers were much more tired at the very start of the trip and remained so at the end of the trip. Given that drivers are fatigued by 12 hour trips, irrespective of how they are driven, it is essen- tial that they start fresh and fit to drive. Further, it seems likely that if 12 hour trips and the work organization surrounding them render drivers vulnerable to fatigue, the impact of work organization on trips involving even longer hours will also be considerable.” Driver Duration p. 5, “Differences between driving types were not sufficient to account for changes in fatigue or performance in this study. All drivers reported more fatigue over the trip, but not all drivers showed poorer performance. It seems that the 12 hour trip is relatively immune to any effects of differences in work practices. It is possible that studying such relatively short trips will not provide clear findings. It may be that longer trips are needed to assess the effect of the differences between these three work practices. It is certainly noteworthy that flexible trips produced no worse an outcome than either of the other two ways of doing the same trip. In fact, a more exhaustive evaluation of flexibility, where drivers have the 194 Findings Directly Related to HOS (include page references):

195 opportunity to learn about manipulating the timing of work and rest during several trips, might reveal that flexibility is of benefit in managing fatigue.” p. 105, “There is little evidence that allowing drivers the freedom to organize the work-rest schedules of their own trips affected their performance. Drivers on flexible trips showed few differences on any of the cognitive or on-board performance tests compared to control trips. It seems that the requirement to comply with the regulated work-rest arrangement does not enhance the drivers’ cognitive functioning or work performance, nor does it reduce the amount of fatigue that drivers report. It should be noted, however that flexible drivers did tend to select work-rest schedules which were quite similar to the regulated working hours. It would be interesting to determine whether this similarity persists when the trip is longer.” Driver Health p. 5, “Even though studies have linked general health and lifestyle to increased fatigue, the design of this study allowed us to rule out factors relating to individual drivers such as level of experience, general health and lifestyle as causes of fatigue in this group of drivers. Undoubtedly 12 hours of driving produces fatigue, but this study suggests that the fatigue does not occur because of factors like poor driver preparation, poor rest and break behav- ior, driving route, or their motivation to complete their trip.”

Next: Abbreviations used without definitions in TRB publications »
Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Commercial Truck and Bus Safety Synthesis Program (CTBSSP) Synthesis 9: Literature Review on Health and Fatigue Issues Associated with Commercial Motor Vehicle Driver Hours of Work examines literature relevant to health and fatigue issues associated with commercial vehicle driver hours of service. This literature review was specifically requested by the Federal Motor Carrier Safety Administration (FMCSA) to provide information related to its Hours of Service regulations issued in January 2004. The report contains a general literature review of the health issues from 1975 to the present, and fatigue issues from January 2004 to present, associated with commercial vehicle driver hours of service. The report also contains a literature review of references that were cited in response to a related FMCSA January 2005 Notice of Proposed Rulemaking. Strictly a literature review, the report does not contain any conclusions or recommendations.

CTBSSP Synthesis 9 Errata Sheet -- Some citation information and abstracts were inadvertently omitted from CTBSSP Synthesis 9 as published.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!