3
The Survey of Respirator Use and Practices: A Learning Experience
The objective of the Survey of Respirator Use and Practices (SRUP), as stated in the overview of the report, was “to provide information to develop educational interventions for specific populations and to increase the frequency and effectiveness of respirator use in the workplace.”1 The survey was designed to provide estimates of the number of establishments and employees who used respirators in a recent 12-month period by type of respirator and type of use. Types of respirators were defined as powered air-purifying respirators, nonpowered air-purifying respirators, and air-supplied respirators. More detailed classifications under these three types were also to be collected—these types were defined as voluntary use, required nonemergency use, and required emergency use.
The survey also was designed to collect data on the characteristics of the respirator program at the establishment; assessment of medical fitness to wear respirators, characteristics of respirator training at the establishment, usefulness of National Institute for Occupational Safety and Health (NIOSH) approval labels and respirator manufacturers’ instructions, substances protected against by the use of respirators, and fit-testing methods used for respirators. The target population of the survey was private-sector establishments with employment covered by unemployment insurance programs that had been judged to be acceptable for use in estimation in the sample of the 1999 Survey of Occupational Injuries and Illnesses (SOII). Public-sector establishments and the self-employed were not included in the sample coverage for this survey.
The data to address these survey objectives had never been systematically collected by mail and telephone from such a large number of establishments covering so many industries and size classes. NIOSH and the Bureau of Labor Statistics (BLS) would find that conducting this survey was a learning experience showing that successfully implementing such a survey is no easy matter.
NIOSH SURVEILLANCE STRATEGY
As stated earlier in this report, the survey was conducted under the auspices of the NIOSH surveillance initiative. In the late 1990s, NIOSH initiated several multidisciplinary initiatives to develop a hazard surveillance strategy for the agency. These initiatives included empowering an in-house team to develop options for a national hazard surveillance survey and database to update the data that had been collected in the 1970s and 1980s. The team considered three options for an ongoing hazard survey: (1) an on-site hazard survey like the previous NIOSH hazard and exposure surveys; (2) a telephone survey of management, much like a recently completed Occupational Safety and Health Administration Cost Survey;2 and (3) a number of smaller hazard surveys targeted by industry or hazard.3 After considering the pros and cons of each of these options, the team recommended the comprehensive, on-site national survey option.
Although the main purpose of the survey options was to produce information on exposure to hazards, plans were extended to include gaining information on worksite safety and health strategies to include personal protective equip-
ment practices. These rather ambitious plans were not implemented, and only a more narrowly focused effort to collect information on respirator use bore fruit.
Respirator Surveillance Team. The overall NIOSH surveillance program turned its immediate attention to the work of an internal NIOSH team—the Respirator Surveillance Team. This team conducted a study of options for surveillance of respirator use and integration of respirator surveillance into the ongoing NIOSH program of field investigations. The team had multiple objectives for the study: to inform a fiscal year 1999 respirator surveillance project, to provide information to other program teams and the certification program, and to inform other parties interested in the use of respirators to control occupational exposures.4
In consultation with staff of the respirator certification program, the Respirator Surveillance Team developed a list of questions that a respirator program might answer:
-
Of the total U.S. workforce, how many employees use respirators?
-
Of those employees using respirators, what respirator types are used?
-
For each respirator type, what hazards are they used to protect against?
-
Of those employees using respirators, how many are employed by firms with a complete written respirator program, with a designated program administrator, and with a program that includes hazard assessment, fit-testing, training, and maintenance?
-
Of those employees with access to a complete program, what fit-test methods are used?
-
Of those employees who are fit-tested, is fit-testing conducted by in-house personnel or by a private vendor of fit-testing service?
NIOSH personnel recognized that these questions were not suitable for direct administration in a field collection instrument. These are topical questions formulated to focus on the type of information to be developed.
It was further recognized by members of the Respirator Surveillance Team that the questions identified did not address a fundamental aspect of respirator protection—the acceptance of them by employees. The team had considered including provisions for exploring respirator acceptance by including an item such as: “Provide feedback to NIOSH about any aspects of respirator design or use that could be improved to promote respirator use when engineering controls are inadequate.” However, the team elected to focus only on aspects of respirator use that could be assessed objectively with help from the employer alone.5 NIOSH personnel decided to forgo direct collection from employees (i.e., respirator users) themselves, setting the stage for a significant weakness in the eventual survey.
The Respirator Surveillance Team reviewed a number of options for collecting the necessary information. These options included repeating data collections along the lines of the National Occupation Exposure Survey and National Occupational Health Survey of Mining projects; a NIOSH interdivisional survey of occupational safety and health programs; surveys in collaboration with BLS; field studies; and focus groups. Considering these options, the team recommended several future actions that would lead to a survey of respirator programs.
Development of the NIOSH-BLS Working Agreement. In the end, NIOSH selected the option of commissioning a BLS survey and elected to enter into an Interagency Agreement with BLS. It did so for several reasons, not the least of which was NIOSH recognition that BLS is an unbiased, reliable, and objective source of expertise. Importantly, BLS had laid some of the groundwork for a respirator use survey by having established respondent contacts as part of its recent (1999) collection of SOII data. Finally, BLS had the methodological infrastructure to conduct such a survey. Nonetheless, this special-purpose survey constituted a new line of business for the Office of Safety, Health, and Working Conditions of BLS.
Although it is a large, multipurpose statistical agency, BLS had some limitations in conducting surveys of this type. Although well acclimated to the conduct of the large annual SOII and other regular collections, BLS had little experience in conducting special-purpose, reimbursable surveys in this field. Importantly, the agency had methodological expertise but did not have the field infrastructure to support a data collection effort of this scope and complexity. The agency would have to contract out data collection to the National Opinion Research Center at the University of Chicago.
The interagency agreement negotiated between NIOSH and BLS translated the overall objectives into a very specific statement of requirements. Data collected from the SRUP would be used by NIOSH “to identify by industrial sector and establishment size: (1) the distribution of respirator use, (2) the types of respirators used, (3) the hazards that respirators are used against, (4) the training of respirator users, (5) the extent medical examinations are used to qualify employees for respirator use for all respirator users, (6) the extent fit testing is used to qualify employees for respirator use, (7) the conductor of fit testing by users, (8) the distribution of fit test methods, (9) the training level of respirator program administrators, (10) the characteristics of respirator programs, and (11) the usefulness of NIOSH
certification label.”6 It was envisioned that the findings of the survey would permit direction and evaluation of efforts to protect employees, by providing researchers with information to develop educational interventions for specific populations to improve respirator use in the workplace.7 Those agreements were carried forward into the Office of Management and Budget (OMB) clearance documentation submitted to OMB by BLS.
The process of development of the questionnaire was iterative and involved both agencies. As in most survey development projects, the final design was the product of many compromises, not the least of which was the trade-off between cost and quality. BLS also had the benefit of ongoing advice during the development phase from program panels of the Labor Research Advisory Council (LRAC) and the Business Research Advisory Council. These panels meet regularly to review programs and make recommendations for program changes and improvements. Both panels included the SRUP in their purview in 2000 and 2001.
The LRAC’s Committee on Occupational Safety and Health Statistics was particularly helpful in the evolution of the survey. In its meeting on December 12, 2000, the LRAC committee critiqued the plans and the proposed questions (after the initial cognitive tests but prior to the field test), identified questions that could be eliminated, and prioritized questions that it would like added. Among the issues raised by the LRAC were (1) the prevalence of both “voluntary” and “routine” respirator use and how these two forms of use differ in participating establishments; (2) the prevalence of both “regular” and “emergency” respirator use and how these two forms of use differ in participating establishments; and (3) the duration of respirator use once employees have put them on (e.g., how long do employees wear respirators after putting them on and when do they take them off?).
The LRAC advised BLS of its priorities for information about respirator use in the workplace. LRAC priorities were (1) emergency response, (2) duration of use, (3) knowledge of who does the training, (4) determination of when the establishment last reviewed its program, and (5) knowledge of whether employees failed medical testing in the establishment’s respirator program.8,9
BLS reported back to the LRAC after the initial round of data collection in November 2001. A member of the LRAC again asked if it was possible to capture public-sector respirator use data. A public-sector frame can be constructed in about 30 states that gather public-sector data for the SOII, but not in the others. However, it was decided to continue to restrict this survey to the private sector.
In the end, the OMB survey approval request form, required for all federal data collections involving the private sector, represented that series of compromises based on trade-offs. The OMB survey plans were approved and the survey was sent to the field.