Click for next page ( 156


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 155
c Glossary and Acronyms GLOSSARY Adjusted, adjustment In a statistical analysis, the process of manipulating or stratifying the values of independent variables so as to minimize their confounding influence on the relationship or association between an independent variable of interest and the dependent variable. Association A correlation or relationship that may or may not be causal, for example when events occur more frequently together than one would expect by chance alone.* Bias "Any systematic error in the design, conduct or analysis of a study that results in a mistaken estimate of an exposure's effect on the risk of disease" (Gordis, 1996, p. 183~. Bivariate analysis A statistical method to characterize the relationship between an independent variable that measures an exposure or treatment (e.g., a potential cause) and a dependent variable that measures an outcome or effect. Causality A relationship that may exist between an exposure or treatment (cause) and an outcome (effect), depending in part on the strength of the association between exposure or treatment and outcome. *Adapted from the Academy for Health Services Research and Health Care Policy glossary at http: //www.academyhealth.org/publications.glossary.pdf, accessed February 4, 2002. 155

OCR for page 155
156 CARE WITHOUT COVERAGE: TOO LITTLE, TOO LATE Chronic disease A disease that has one or more of the following characteristics: is permanent; leaves residual disability; is caused by nonreversible physiological damage; requires special training of the patient for rehabilitation; or may be expected to require a long period of supervision, observation, or care.* Confidence interval (CI) A numeric range estimated with a specific degree of confidence or probability to include a value. Conventionally reported confidence intervals are ranges in which the actual value of the estimated variable can be expected to fall 95 or 99 percent of the time, corresponding to probabilities that a difference or significant result is due to chance of 5 percent and 1 percent, respectively (p <.05; p <.01~. In reporting quantified results throughout this report (e.g., odds ratios or relative risks), if the confidence interval is not given, point estimates have at least a 95 percent probability of being statistically significant. Confidence intervals are given for findings reported with lesser levels of statistical . ~ slgmilcance. Confounder A variable that is associated with an exposure or treatment of interest and, as a result, influences the relationship between the exposure or treatment and an outcome. The ability to adjust or analytically control for the presence of a confounder depends on how well this variable is measured. Cost sharing Any provision of a health insurance policy that requires the insured individual to pay some portion of medical expenses. The general term includes deductibles, copayments, and coinsurance.* Covariate A variable that is related to or associated with the study variablets) of interest. Cross-sectional Describes a research study in which measurements are collected and comparisons made among populations at one point in time. Experimental Describes a study design, for example, a randomized clinical trial, where researchers use a defined study population, randomly assign members of the population to exposure or treatment and control groups, control the timing of the exposure or treatment, and influence the timing of measurements. Hazard ratio, Cox proportional hazard rate A comparative measure of the strength of a relationship or association between an exposure, intervention, or *Adapted from the Academy for Health Services Research and Health Care Policy glossary at http: //www.academyhealth.org/publications.glossary.pdf, accessed February 4, 2002.

OCR for page 155
APPENDIX C 157 treatment (e.g., measured in terms of one or more independent variables) and an outcome (e.g., measured in terms of a dependent variable) over time for a defined study population divided into exposure and control groups. For example, to estimate the influence of an exposure on the length of time until death (mortality over time), the use of a Cox proportional hazard rate or hazard ratio allows for multivariate analysis ofthe incremental or proportional difference that each unit of time would be expected to make in increasing or reducing the risk of mortality. In this example, the mortality hazard ratio, or incidence of mortality estimated at one moment in time, is defined as the relative risk of mortality for persons in an exposure group compared with the relative risk of mortality for members of a control group. Health-related quality of life (HRQOL) A research construct developed by the Centers for Disease Control and Prevention to help monitor progress in achieving national health objectives. It has been used in the Behavioral Risk Factor Surveillance System surveys since 1993 and, since 2000, in the National Health and Nutrition Examination Survey. Its core element consists of four ques- tions that encompass general self-reported health status, the number of unhealthy days within a recent time period (e.g., the month before the interview) for both physical and mental dimensions, and restricted activity days. Incidence A measure of the probability of a disease or an outcome's occurrence, defined as the number of new cases within a defined time period for a specific population divided by the total number in the population (Gordis, 1996~. Longitudinal Describes a research study in which measurements are collected and comparisons made among populations over time. Medically indigent Persons who cannot afford needed health care because of insufficient income and/or lack of adequate health insurance. Indigent care con- sists of health services provided to the poor or those unable to pay. Since many indigent patients are not eligible for federal or state programs, the costs that are covered by Medicaid are generally recorded separately from indigent care costs.* Multivariable or multivariate analysis A statistical method to characterize the relationship among at least two independent variables that measure exposures or treatments (e.g., potential causes) and a dependent variable that measures an out- come or effect. *Adapted from the Academy for Health Services Research and Health Care Policy glossary at http: //www.academyhealth.org/publications.glossary.pdf, accessed February 4, 2002.

OCR for page 155
58 CARE WITHOUT COVERAGE: TOO LITTLE, TOO LATE Observational Describes a research study with a nonexperimental design, in which researchers gather observations or measurements while not intentionally affecting the conditions of exposure, the treatment of the study population, or the . . timing ot measurements. Odds ratio (OR) A comparative measure of the strength of a relationship or association between an exposure or treatment and an outcome for two popula- tions, where the baseline incidence of the outcome in these groups may not be known. In this report, it is the relative odds of either (1) developing the outcome for an uninsured group, compared with the odds for an insured group, or (2) having been uninsured for a group with an outcome, compared with the odds for a control population. For example, if the odds of receiving a Pap test are 2:1 in a group of uninsured women (i.e., two of every three women, or 67 percent, receive the test) and the odds are 4:1 in a group of women with insurance (i.e., four of every five women, or 80 percent, receive the test), the odds ratio of uninsured compared to insured women is 0.5 (2:1/4:1~. The OR is not a good estimate of the relative risk (the probability of been screened in the uninsured group divided by the probability of being screened in the insured group) because screening is not a rare event. Predictor, independent predictor ~ , 7 ~ variable (e.g., that measures an exposure or treatment) that is shown to be likely to influence or predict the value of a dependent variable (e.g., an outcome). in a statistical analysis. an independent Prevalence A measure of how common a disease or condition is within a population, defined as the number of cases in the population at a specified time divided by the number of persons in the population at that same time (Gordis, 1996~. Quasi-experimental Describes a research study, for example, a natural experi- ment, whose design combines experimental and nonexperimental aspects. Typi- cally, researchers cannot control the timing of the intervention or exposure whose effects are being measured, or the random assignment of a defined group of study subjects, but they can influence the timing of measurements. Randomized trial Describes a research study in which the members of a defined group of subjects are randomly assigned to at least two groups for the purpose of analysis: a treatment or intervention group and a control group. Relative risk (RR) A comparative measure of the strength of a relationship or association between an exposure, intervention, or treatment and an outcome for a defined study population, where the baseline incidence of the outcome is known. It is expressed as the ratio of two risks, namely, the rate of a disease or condition of interest in the treated portion of the population, divided by the rate in an

OCR for page 155
APPENDIX C 159 untreated or control portion of the population. A value of one means that the rates in both portions are the same.* Selection bias In research studies, a systematic error in analysis that results when study subjects are not assigned randomly among treatment and control groups. Statistically significant See definition of Confidence interval. ACRONYMS AMI acute myocardial infarction ASOC Survey of Aging, Status and the Sense of Control BRFSS Behavioral Risk Factor Surveillance System CABG coronary artery bypass graft CBE clinical breast exam CDC Centers for Disease Control and Prevention CI confidence interval CPS Current Population Survey ED emergency department EPO erythropoietin ESRD end-stage renal disease FFS fee for service FOBT fecal occult blood test FPL federal poverty level HAART highly active antiretroviral therapy HCSUS HIV Cost and Services Utilization Study HCUP Healthcare Cost and Utilization Project HMO health maintenance organization HRQOL health-related quality of life ICU intensive care unit IOM Institute of Medicine ISS injury severity score *Adapted from the Academy for Health Services Research and Health Care Policy glossary at http:/ /www.academyhealth.org/publications.glossary.pdf, accessed February 4, 2002.

OCR for page 155
160 CARE WITHOUT COVERAGE: TOO LITTLE, TOO LATE LOS length of stay MEPS Medical Expenditure Panel Survey NCHS National Center for Health Statistics NHANES National Health and Nutrition Examination Survey NHIS National Health Interview Survey NMES National Medical Expenditure Survey NNRTI nonnucleoside reverse transcriptase inhibitor NRMI National Registry of Myocardial Infarction OR odds ratio PI protease inhibitor PTCA percutaneous transluminal coronary angioplasty RR relative risk RW}F The Robert Wood Johnson Foundation SCHIP State Children's Health Insurance Program SEER Surveillance, Epidemiology, and End Results Program SES socioeconomic status SMI severe mental illness UCSF University of California at San Francisco USRDS United States Renal Data System VA Department of Veterans Affairs