A special section, "Managing Chronic Illness in Managed Care Settings," was included in the December 2003 issue of the journal Health Services Research 38(6), Part I. The section includes an introduction and six articles from projects funded under a managed care initiative by the Agency for Healthcare Research and Quality, the American Association of Health Plans Foundation (now called America's Health Insurance Plans), and the Health Resources and Services Administration. The projects examine the impact of various features of managed care on the management of chronic disease.
For example, researchers studied the impact on chronic disease management of specific payment arrangements for individual physicians; the breadth of the provider networks offered by plans; and managerial approaches used by the plans (or medical groups) to select clinicians, monitor or profile practice patterns, and encourage adherence to practice guidelines. An introduction to the section and the six papers are briefly summarized here.
Luft, H.S., and Dudley, R.A., "Measuring quality in modern managed care," pp. 1373-1383 (AHRQ grant HS10771).
In this introduction, the journal editors note that there is no single conceptually correct level to analyze the impact of managed care on chronic disease. In some instances, the variables and outcomes of interest are best conceptualized at the health plan level because they deal with issues of coverage or care coordination. In other instances, the clinic level may be most relevant, for example, when the goal is to identify care processes that could be improved.
The papers in this section address key methodological issues that must be addressed when attempting to measure and analyze performance in the complex managed care environment. These methodological insights are of interest not only to the study of managed care, but also to issues of quality measurement, patient and consumer surveys, and complex study design.
Keating, N.L., Landrum, M.B., Landon, B.E., and others, "Measuring the quality of diabetes care using administrative data: Is there bias?" pp. 1529-1545 (AHRQ grant HS09936).
Health care organizations often measure processes of care using only administrative data. However, when quality of care is measured using administrative data without patient medical record data, diabetes quality of care indicators (performance of certain tests, for example, for diabetic retinopathy and blood-sugar level) may be under-detected more frequently for elderly and black patients and the physicians, clinics, and health plans who care for them. In other words, administrative data suggest that blacks and elderly patients with diabetes receive fewer recommended tests than they actually do. This suggests that providers who care for such patients may be disproportionately affected by public release of such data or the use of these data to determine the magnitude of financial incentives. Thus, health care organizations should consider using standardized indicators collected from both administrative and medical record data, such as those developed by the Diabetes Quality Improvement Project.
Stuart, B., Singhal, P.K., Magder, L.S., and Zuckerman, I.H., "How robust are health plan quality indicators to data loss? A Monte Carlo simulation study of pediatric asthma treatment," pp. 1547-1561 (AHRQ grant HS09950).
Proportion-based quality indicators (QIs) have become benchmarks for evaluating health plan performance, for example, the percentage of plan members receiving an annual flu shot. This study examined a proportion-based QI of treatment for persistent asthma (the percent of asthma patients filling prescriptions for two or more rescue medications who also filled a controller medication) for a group of Medicaid children transitioning from fee-for-service (FFS) to managed care. The goal was to test the robustness of proportion-based QIs to loss of encounter data. The QI was based on FFS medical and prescription claims in the before-transfer period and encounter data in the followup period. The QI measure proved highly robust to most forms of encounter data loss tested. The measure declined by less than 2 percent in the presence of up to a 35 percent loss of data. The findings suggest that other proportion-based QIs are likely to reflect true levels of health plan quality in the face of incomplete data capture.
Lozano, P., Grothaus, L.C., Finkelstein, J.A., and others, "Variability in asthma care and services for low-income populations among practice sites in managed Medicaid systems," pp. 1563-1578 (AHRQ grant HS09935).
Chronic asthma care for children in managed care Medicaid plans could be improved, particularly in the areas of self-management support and information systems, according to this study. The investigators surveyed 85 managed care Medicaid practice sites about processes of asthma care. They found substantial variation in reports of asthma processes of care, particularly in the areas of self-management and information systems. Practice sites also varied significantly in availability of services targeting low-income populations, specifically relating to cultural diversity, communication, and enrollee empowerment. Very little of the site-to-site variation was attributable to the managed care organization.
Kahn, K.L., Liu, H., Adams, J.L., and others, "Methodological challenges associated with patient responses to follow-up longitudinal surveys regarding quality of care," pp. 1579-1598 (AHRQ grant HS09951).
Certain methodological challenges are associated with patient responses to followup longitudinal surveys regarding quality of care, which are outlined in this study. The researchers surveyed 30,308 patients in 1996 and 13,438 patients in 1998 as part of a 2-year longitudinal study of quality of care and health status of patients receiving care delivered by 63 physician groups across three West Coast States. They also included a subset of patients with chronic disease. In 1998, response rates were 54 percent overall (46 percent nonresponders) and 63 percent for patients with chronic disease (37 percent nonresponders). Patient demographics, health status, use of services, and satisfaction with care in 1996 were all significant predictors of response in 1998. However, process of care scores weighted for nonresponse differed from unweighted scores. This highlights the importance of applying nonresponse weights to minimize bias in estimates of care and outcomes associated with longitudinal quality of care and health outcome analyses.
Shenkman, E., Wu, S.S., Nackashi, J., and Sherman, J., "Managed care organizational characteristics and health care use among children with special health care needs," pp. 1599-1624 (AHRQ grant HS09949).
It is not certain how well managed care organizations (MCOs) are able to establish provider networks that meet the needs of children with special health care needs (CSHCN) and how MCOs assist primary care physicians in coordinating care for these children. The purpose of this study was to examine the relationship between MCOs' provider network composition (number of pediatricians, family practitioners, and pediatric subspecialists), strategies to coordinate or facilitate care for CSHCN, and the effects of reimbursement practices on access to care for 2,223 CSHCN enrolled in one of eight Florida MCOs. Three MCO indexes—pediatrician focus index, specialist focus index, and fee-for-service index—were significantly associated with children's health care use patterns. These MCO indexes, not the child-level characteristics, explained most of the variation in outpatient use rates by CSHCN in the eight MCOs.
Adams, J.L., Wickstrom, S.L., Burgess, M.J., and others, "Sampling patients within physician practices and health plans: Multistage cluster samples in health services research," pp. 1625-1640 (AHRQ grant HS09942).
A common study design in health services research is the multistage cluster sample to abstract data from medical records. The simple two-stage problem is sampling patients in physician practices to estimate a population average (for example, mean blood pressure). However, multistage design is needed if there are two or more subpopulations (for example, men and women) for whom separate estimates are required. These authors developed a simulation-based approach for designing multistage samples. They constructed a sample frame from all diabetes patients in six health plans who had evidence of chronic eye disease (glaucoma and retinopathy). They found that simulations of different study designs can uncover efficiency gains as well as inform potential tradeoffs among study goals. One of several benefits to this approach is that it can be easily adapted to problems that arise in the field (for example, when cost overruns reduce sample size or practices or physicians refuse to participate).
Return to Contents
Proceed to Next Article
540 Gaither Road Rockville, MD 20850