Bates, D.W., Evans, R.S., Murff, H., and others (2003, March). "Detecting adverse events using information technology." (AHRQ grant HS11046). Journal of the American Medical Informatics Association 10(2), pp. 115-128.
Although patient safety is a major concern, most health care organizations rely on spontaneous reporting, which detects only a small minority of adverse events. Chart review can detect such adverse events in research settings but is too expensive for routine use. These authors reviewed methodologies for detecting adverse events using information technology (IT), reports of studies that used these techniques to detect adverse events, and study results for specific types of adverse events. Their review revealed that IT techniques, such as event monitoring and natural language processing, provide an inexpensive way to detect certain types of adverse events in clinical databases. These approaches already work well for some types of adverse events, including adverse drug reactions and hospital-induced infections, and are in routine use in a few hospitals. It also appears likely that these techniques will be adaptable in ways that allow detection of a broad array of adverse events, especially as more medical information becomes computerized.
Bernard, D., and Selden, T.M. (2003). "Employer offers, private coverage, and the tax subsidy of health insurance: 1987 and 1996." International Journal of Health Care Finance and Economics 2, pp. 297-318.
Economists have long been interested in the effects of tax-based subsidies on private health insurance coverage. The authors examined this relationship using pooled data from the 1987 National Medical Expenditure Survey (NMES) and the 1996 Medical Expenditure Panel Survey (MEPS). Their main tax price elasticity estimates for employer offers and for private coverage are near the mid-point of the existing literature. However, these estimates may mask substantial differences in tax-price responsiveness across subsets of workers. Their more disaggregated analysis revealed tax price responsiveness to be significantly above average for low-income workers, workers with low health risks, and workers in small firms—precisely those groups whose continued participation in employment-related risk pooling is of greatest policy concern.
Reprints (AHRQ Publication No. 03-R031) are available from the AHRQ Publications Clearinghouse.
Eisenman, D.P., Cunningham, W.E., Zierler, S., and others (2003, February). "Effect of violence on utilization of services and access to care in persons with HIV." (AHRQ grant HS08578). Journal of General Internal Medicine 18, pp. 125-127.
Based on a national sample of adults with HIV or AIDS being seen in primary care, nearly 21 percent of women, 12 percent of homosexual/bisexual men, and 8 percent of remaining men (presumably heterosexual) have been physically abused within an important relationship since HIV diagnosis. These researchers analyzed data from the HIV Cost and Services Utilization Study to determine the association of violence, assessed at baseline, with use of and access to health care at followup among gay/bisexual male, heterosexual female, and heterosexual male HIV/AIDS patients. They found that male gay/bisexual violence victims had increased odds of reporting emergency department visits, going without needed medical care because of expense, and having poor ability to access medical specialists. The researchers call for more research to further clarify the association of violence with health care among gay/bisexual men with HIV/AIDS.
Gill, C.J., and Mularski, R.A. (2003, January). "Haemophilus aphrophilus purulent pericarditis and tamponade." (AHRQ National Research Service Award training grant T32 HS00060). Infections in Medicine 20, pp. 31-33.
These authors describe a clinical case of a healthy man with chest pain and prolonged fever, which highlights the difficulties of diagnosing purulent pericarditis when it is caused by a relatively avirulent pathogen. The usual clinical course of purulent pericarditis is acute and severe. This case of purulent pericarditis was caused by an unusual pathogen, Haemophilus aphrophilus, an organism more commonly seen in the context of slowly progressive bacterial endocarditis. In this case, the man's blood cultures demonstrated H. aphrophilus, but echocardiographic studies revealed only small amounts of pericardial fluid and no valvular vegetations. The researchers report a sequence of events that eventually led to the diagnosis of purulent pericarditis and the therapeutic steps taken to manage the problem. They note that a contrast-enhanced computerized tomographic scan could have suggested the diagnosis earlier than echocardiography. This case study reinforces the contention that successful therapy depends on early recognition of this disease and prompt surgical drainage, with antibiotic therapy playing a secondary role.
Grabowski, D.C. (2002). "A multi-part model approach to examining Medicaid payment methods and nursing home quality." (AHRQ National Research Service Award training grant T32 HS00084). Health Services & Outcomes Research Methodology 3, pp. 21-39.
The issues of escalating costs and problems with quality have dominated nursing home policy discussions over the last three decades. The incentive to provide nursing home quality depends, in part, on whether the State Medicaid reimbursement system is prospective or retrospective in nature. This author examined the effects of State-level Medicaid reimbursement methods on the provision of quality in the context of bed constraint regulations that may influence market tightness. He constructed a three-part estimation strategy around the idea that a change in the Medicaid reimbursement method may affect both a facility's payer mix and the provision of quality. Across a range of quality measures, this multi-part
model did not show nursing home quality to be significantly higher under a retrospective reimbursement system than under a prospective-based system of reimbursement. This finding held regardless of whether the analysis was isolated to those markets with the tightest supply of beds or those homes caring for predominantly Medicaid residents.
Groessl, E.J., Kaplan, R.M., and Cronan, T.A. (2003, February). "Quality of well-being in older people with osteoarthritis." (AHRQ grant HS09170). Arthritis & Rheumatism 49(1), pp. 23-28.
The objective of this study was to examine the sensitivity and validity of the Quality of Well-Being Scale (QWB) as a measure of health-related quality of life (HRQOL) in older people with osteoarthritis (OA). OA can involve degenerative changes, which sometimes lead to stiffness, swelling, and deformity that negatively affect functioning and quality of life. The study involved 363 elderly people with OA in a Southern California health maintenance organization. The majority of participants were white (92.3 percent) and retired (75.2 percent), and more than 64 percent were women. The researchers administered the QWB, the Arthritis Impact Measurement Scale (AIMS), a sensitive measure of arthritis symptoms, and other health measures, and obtained data on self-reported health status and health care use at baseline and 1 year later. The QWB scale asks questions about 27 different symptom complexes over the 6 previous days and assesses functioning during that time in three areas: mobility, physical activity, and social activity. Results showed that the QWB is comprehensive and incorporates many aspects of symptoms and functioning that affect quality of life in people with OA. The QWB score was also related to health care costs, to a depression-specific measure, and to self-rated health. These findings support the validity of the QWB for applications in studies involving patients with OA.
Horn, S.D., Torres, A., Wilson, D., and others (2002, October). "Development of a pediatric age- and disease-specific severity measure." (AHRQ Contract 290-95-0042). Journal of Pediatrics 141, pp. 496-503.
The current severity of illness scoring systems available in pediatrics are based on statistical regression models constructed to explain variation in a single outcome, in-hospital death. However, death is uncommon in hospitalized children. Thus, mortality prediction may have limited utility. The researchers adapted the adult Comprehensive Severity Index (CSI) for hospitalized children and evaluated the ability of the CSI to predict common pediatric outcomes. The CSI is a disease-specific severity system that provides a consistent method to define grades of severity for patient historical factors, physiologic parameters, and laboratory results. They evaluated CSIÕs predictive power by using retrospective data collected from 16,496 randomly selected children admitted to 10 hospitals in 1995 and 1996. Admission CSI score predicted mortality well and discriminated well within 9 case-mix groups with 10 or more deaths. Maximum CSI score explained the variation in length of stay and cost within 32 case-mix groups. CSI had better predictability than the Pediatric Risk of Mortality, a pediatric severity scoring system.
Huie, S.A., Hummer, R.A., and Rogers. R.G. (2002). "Individual and contextual risks of death among race and ethnic groups in the United States." Journal of Health and Social Behavior 43, pp. 359-381.
The authors used the 1986-1997 National Health Interview Survey (NHIS)—National Death Index linked data file to examine the effects of individual and contextual factors on black-white and multiple Hispanic subgroup differences in adult mortality. They used a new, innovative area—the very small area—as the contextual unit of analysis. They found that excess mortality risks for all racial and ethnic groups considered were associated with not only individual characteristics, but also neighborhood characteristics. In addition, the percentage of foreign-born individuals in a neighborhood was found to be protective of Hispanic subgroup mortality for Puerto Rican, Mexican American, and other Hispanic adults in the 45-74 age category. The authors cite the need for additional research to examine more thoroughly the pathways through which neighborhood factors affect mortality among many Hispanic subgroups and the role of nativity as a protective factor for older adult Hispanic mortality.
Reprints (AHRQ Publication No. 03-R022) are available from the AHRQ Publications Clearinghouse.
Neumann, P.J., and Levine, B. (2002). "Do HEDIS measures reflect cost-effective practices?" (AHRQ grant HS10709 and National Research Service Award training grant T32 HS00063). American Journal of Preventive Medicine 23(4), pp. 276-289.
The Health Plan Employer Data and Information Set (HEDIS) performance measures, used widely to assess the quality of care in health plans in the United States, generally reflect cost-effective practices. However, in a number of cases, practices may not be cost effective for certain subgroups, according to this study. The researchers examined the cost-effectiveness evidence for each of the 15 "effectiveness of care" measures in HEDIS 2000. They searched two databases of economic evaluations and two published lists of cost-effectiveness ratios in health and medicine through 1998 for cost-effectiveness ratios of similar interventions and target populations. They also searched for important interventions with evidence of cost-effectiveness—that is, less than $20,000 per life year (LY) or quality-adjusted life year QALY), which are not included in HEDIS. Evidence was available for 11 of the 15 HEDIS measures. Cost-effectiveness ranged from cost savings to $660,000/LY gained. Numerous non-HEDIS interventions also had some evidence of cost effectiveness, particularly interventions to promote healthy behaviors.
Ray, W.A., Thapa, P.B., and Gideon, P. (2002). "Misclassification of current benzodiazepine exposure by use of a single baseline measurement and its effects upon studies of injuries." (AHRQ grant HS10384). Pharmacoepidemiology and Drug Safety 11, pp. 663-669.
To properly measure exposure to benzodiazepines and other medications taken intermittently requires more than a single baseline measurement. Exposure needs to be tracked on a day-by-day basis, according to this study. Previous studies have defined benzodiazepine exposure status from a single baseline measurement, and these have not consistently reported increased risk of unintentional injuries due to impaired psychomotor function. These authors used the medication records of 2,510 elderly Tennessee nursing home residents identified in a prior study of antidepressants and falls to determine both baseline use and current use of benzodiazepines. They identified falls from nursing home incident reports and medical records. The 666 baseline benzodiazepine users had current use on 45 percent of followup person-days; baseline non-users had current use for 4 percent of days. Misclassification of drug exposure increased with length of followup and with quintile of fall risk.
Return to Contents
Current as of May 2003
AHRQ Publication No. 03-0034
Internet Citation:
Research Activities newsletter. May 2003, No. 273. AHRQ Publication No. 03-0034. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/research/may03/
540 Gaither Road Rockville, MD 20850