Showing papers in "Health Services Research in 1995"

Assessing the impact of continuous quality improvement/total quality management: concept versus implementation.

Abstract: Objective This study examines the relationships among organizational culture, quality improvement processes and selected outcomes for a sample of up to 61 U. S. hospitals. Data sources and study setting Primary data were collected from 61 U. S. hospitals (located primarily in the midwest and the west) on measures related to continuous quality improvement/total quality management (CQI/TQM), organizational culture, implementation approaches, and degree of quality improvement implementation based on the Baldrige Award criteria. These data were combined with independently collected data on perceived impact and objective measures of clinical efficiency (i.e., charges and length of stay) for six clinical conditions. Study design The study involved cross-sectional examination of the named relationships. Data collection/extraction methods Reliable and valid scales for the organizational culture and quality improvement implementation measures were developed based on responses from over 7,000 individuals across the 61 hospitals with an overall completion rate of 72 percent. Independent data on perceived impact were collected from a national survey and independent data on clinical efficiency from a companion study of managed care. Principal findings A participative, flexible, risk-taking organizational culture was significantly related to quality improvement implementation. Quality improvement implementation, in turn, was positively associated with greater perceived patient outcomes and human resource development. Larger-size hospitals experienced lower clinical efficiency with regard to higher charges and higher length of stay, due in part to having more bureaucratic and hierarchical cultures that serve as a barrier to quality improvement implementation. Conclusions What really matters is whether or not a hospital has a culture that supports quality improvement work and an approach that encourages flexible implementation. Larger-size hospitals face more difficult challenges in this regard.

The geographical accessibility of hospitals to the aged: a geographic information systems analysis within Illinois.

Abstract: OBJECTIVE. This article uses geographic information systems and their related tools to empirically measure and display the geographic accessibility of the aged population to hospital facilities within Illinois. DATA SOURCES AND STUDY SETTING. Geographic accessibility of Illinois' aged population is measured from each of the state's 10,796 census block groups to the state's 214 hospital facilities. Block group demographic compositions and centroids are obtained from 1990 census files. Hospital coordinates are obtained by the authors. STUDY DESIGN. Of five alternative measures of accessibility considered, empirical estimates are obtained for two: choice set and minimum distance. Access to both general hospitals and the subset having specialized geriatric facilities is measured with special attention to differences in accessibility between the aged within metropolitan statistical areas (MSAs) and those outside MSAs. Cumulative accessibility distributions and their summary statistics provide a basis of comparison among subgroups. DATA COLLECTION AND EXTRACTION. Geographic information systems (GIS) and their related tools are used as a means of efficiently capturing, organizing, storing, and retrieving the required data. Hospitals and census block groups are geocoded to specific locations in the database, and aspatial attributes are assigned to the hospitals and block groups. The GIS database is queried to produce shaded isarithm and point distribution maps that show the location of hospitals relative to surrounding aged populations. CONCLUSION. The vast majority of Illinois' aged population is within close proximity to hospital facilities. Eighty percent (1,147,504 persons) of the aged in Illinois are within 4.8 miles (7.7 km) of a hospital and 11.6 miles (18.7 km) of two hospitals. However, geographic accessibility differences between the aged living in MSAs and those living outside MSAs to hospitals offering geriatric services are substantial; but there is no evidence that the aged's geographical accessibility to hospitals is less favorable than that of the general population. Detailed accessibility measures permitted by geographic information system technology call into question the continued use of crude empirical accessibility measures.

Recruitment and retention of subjects for a longitudinal cancer prevention study in an inner-city black community.

Abstract: Recruiting and retaining subjects for longitudinal prevention trials is challenging. The inherent difficulties are compounded when the trial is to take place in a low-income minority community, since prevention is a low priority among residents of such communities, and research is viewed with suspicion. We present our experiences in attempting to recruit and retain low-income black women living in inner-city Atlanta for a trial of an educational intervention to promote screening for cancer. The intervention was conducted in the home by trained lay health workers. We found that recruitment was more successful when we recruited directly from the community than when we recruited from the patient registry of a primary health care center. The attrition rate over an 18-month period was high. Among members of the intervention group, those retained in the study tended to be wealthier and better educated and were more likely to be married and employed than those who dropped out. It seems probable that women of lower socioeconomic status found our intervention to be intrusive or burdensome. Among the controls, socioeconomic factors did not discriminate between those who completed the study and those who did not; loss to follow-up in this group was associated only with younger age. In conducting research of this type in low-income minority communities, special attention must be given to issues of recruitment and retention if the validity of the study is to be preserved.

Impact of varying panel membership on ratings of appropriateness in consensus panels: a comparison of a multi- and single disciplinary panel.

Abstract: OBJECTIVE. The objective of the study was to examine the appropriateness ratings for the use of spinal manipulation for low back pain of a multidisciplinary panel of medical and chiropractic physicians, and those of a panel composed only of chiropractic physicians. DATA SOURCES. The study analyzed data from two consensus panels conducted at RAND in 1990 and 1991. STUDY DESIGN. The study design followed that of the traditional RAND consensus panels. Nine individuals comprised each panel, and each panelist was asked to rate, on a nine-point scale, the indications for spinal manipulation twice, the first time alone and the second time jointly with the panel. DATA COLLECTION. The ratings of the panelists from both groups, for both round one and round two, were collated and compared. PRINCIPAL FINDINGS. While both panels were more likely to rate the indications as inappropriate than appropriate, the single disciplinary panel was more likely to rate an indication as appropriate than the multidisciplinary panel. CONCLUSION. The composition of a panel clearly influences the ratings and those who use a given procedure in practice, in this case manipulation, are more likely to rate it as appropriate than those who do not use the procedure.

The efficacy of primary care for vulnerable population groups.

Abstract: This article reviews the existing literature on the efficacy of primary care with an emphasis on the evaluation of primary care for vulnerable populations: groups whose demographic, geographic, or economic characteristics impede or prevent their access to health care services. A significant portion of the literature derives from studies of poor and underserved populations. However, to construct a more complete evaluation of primary care services, the authors cite literature that has examined both advantaged and disadvantaged populations. Even then the literature is incomplete, at best. The article describes a definition of primary care suitable for policy analysis and formulation, reviews evidence on the efficacy of care that meets that definition, and concludes that widespread use of primary care services is likely to result in improved patient satisfaction and health status.

Racial/ethnic differences in health care utilization of cardiovascular procedures: a review of the evidence.

Abstract: Studies based on a variety of primary data sets have consistently demonstrated that African Americans are about half as likely as whites to receive interventional therapy for coronary artery disease. Neither disease severity per se nor access to hospitals performing these procedures accounts for this finding. Likewise, available measures of income and ability to pay, including insurance status, do not explain the differences. Subtle personal factors, including physician bias and the willingness of patients to accept referral for surgery, may be important but have not as yet been measured. These findings present a challenge to the presumed equality in access to health care within the United States medical system. A new generation of health services research studies will be required to provide definitive reasons for this important disparity in treatment.

The effects of hospital-physician integration strategies on hospital financial performance.

Abstract: STUDY QUESTION This study investigated the longitudinal relations between hospital financial performance outcomes and three hospital-physician integration strategies: physician involvement in hospital governance, hospital ownership by physicians, and the integration of hospital-physician financial relationships DATA SOURCES AND STUDY SETTING Using secondary data from the State of California, integration strategies in approximately 300 California short-term acute care hospitals were tracked over a ten-year period (1981-1990) STUDY DESIGN The study used an archival design Hospital performance was measured on three dimensions: operational profitability, occupancy, and costs Thirteen control variables were used in the analyses: market competition, affluence, and rurality; hospital ownership; teaching costs and intensity; multihospital system membership; hospital size; outpatient service mix; patient volume case mix; Medicare and Medicaid intensity; and managed care intensity DATA COLLECTION/EXTRACTION Financial and utilization data were obtained from the State of California, which requires annual hospital reports A series of longitudinal regressions tested the hypotheses PRINCIPAL FINDINGS Considerable variation was found in the popularity of the three strategies and their ability to predict hospital performance outcomes Physician involvement in hospital governance increased modestly from 1981-1990, while ownership and financial integration declined significantly Physician governance was associated with greater occupancy and higher operating margins, while financial integration was related to lower hospital operating costs Direct physician ownership, particularly in small hospitals, was associated with lower operating margins and higher costs Subsample analyses indicate that implementation of the Medicare prospective payment system in 1983 had a major impact on these relationships, especially on the benefits of financial integration CONCLUSIONS The findings support the validity of hospital-physician financial integration efforts, and to a lesser extent the involvement of physicians in hospital governance The results lend considerably less support for strategies built around direct physician ownership in hospitals, particularly since PPS implementation RELEVANCE/IMPACT These findings challenge prior studies that found few financial benefits to hospital-physician integration prior to PPS implementation in 1983 The results imply that financial benefits of integration may take several years after implementation to emerge, are most salient in a managed care or managed competition environment, and vary by hospital size and multihospital system membership

Hospital adoption of medical technology: an empirical test of alternative models.

Abstract: OBJECTIVE. This study examines hospital motivations to acquire new medical technology, an issue of considerable policy relevance: in this case, whether, when, and why hospitals acquire a new capital-intensive medical technology, magnetic resonance imaging equipment (MRI). STUDY DESIGN. We review three common explanations for medical technology adoption: profit maximization, technological preeminence, and clinical excellence, and incorporate them into a composite model, controlling for regulatory differences, market structures, and organizational characteristics. All four models are then tested using Cox regressions. DATA SOURCES. The study is based on an initial sample of 637 hospitals in the continental United States that owned or leased an MRI unit as of 31 December 1988, plus nonadopters. Due to missing data the final sample consisted of 507 hospitals. The data, drawn from two telephone surveys, are supplemented by the AHA Survey, census data, and industry and academic sources. PRINCIPAL FINDING. Statistically, the three individual models account for roughly comparable amounts of variance in past adoption behavior. On the basis of explanatory power and parsimony, however, the technology model is "best." Although the composite model is statistically better than any of the individual models, it does not add much more explanatory power adjusting for the number of variables added. CONCLUSIONS. The composite model identified the importance a hospital attached to being a technological leader, its clinical requirements, and the change in revenues it associated with the adoption of MRI as the major determinants of adoption behavior. We conclude that a hospital's adoption behavior is strongly linked to its strategic orientation.

Which physicians limit their Medicaid participation, and why.

Abstract: OBJECTIVE. This study identifies factors differentiating Medicaid participating physicians who accept all Medicaid patients from those limiting their Medicaid participation. DATA SOURCES. Data come from periodic telephone surveys of random samples of physicians conducted by the American Medical Association (AMA). STUDY DESIGN. Surveys conducted in 1990-1993 were pooled to form a sample of 4,188 Medicaid-participating office-based physicians. Respondents were classified as accepting all Medicaid patients or as limiting their Medicaid participation. Descriptive statistics are used to examine differences between these groups with respect to selected personal, practice, community, and reimbursement variables. Logistic regression analysis is used to identify factors associated with physicians accepting all Medicaid patients or limiting their Medicaid participation in some way. DATA COLLECTION METHODS. Survey data were supplemented with 1990 census data, 1990 AMA Physician Masterfile data, and 1989 data on physician payment levels. PRINCIPAL FINDINGS. Less than half of Medicaid-participating physicians and only about one-third of participating primary care physicians accept all Medicaid patients. Higher Medicaid fees are associated with physicians participating fully, but the marginal effects of changes in fees on the probability of physicians participating fully is small. CONCLUSIONS. Increases in Medicaid reimbursement aimed at primary care physicians or those in underserved areas may convert limited participants into full participants and, in so doing, improve the access of Medicaid eligibles to care. The increases in payment level needed to increase the proportion of physicians participating fully would be substantial, however, and may not be politically feasible.

Small area variations in health care delivery in Maryland.

Abstract: OBJECTIVE: Our purpose is a descriptive analysis of variations in hospital use among small areas of Maryland. DATA SOURCE: The data are Maryland patient discharge records from acute care hospitals for 1985-1987 and small area population estimates by age, gender, race, and income. FINDINGS: The common finding was excess geographic variability among Maryland's 115 areas. The hypothesis of uniform rates was rejected for most DRGs, including low-variation mastectomy and hernia repair. Clustering of high-use rates occurred in neighboring areas for orthopedic, vascular, and elective procedures. Admission rates for most nondiscretionary procedures and medical DRGs were reduced in affluent areas while discretionary surgery increased with income level. Elective procedures had extreme variation and were related to income. Coronary artery disease rates declined with income while coronary artery procedure rates increased, indicating that access and patient selection were factors in the use of coronary bypass and angioplasty. CONCLUSIONS: The issue is not the ubiquitous variation among small areas but its extent and identification of geographic patterns. Hospital use is related to demography, morbidity, medical resources, access, selection for care, and physician practice patterns. Heterogeneity of these factors ensures that uniform delivery of health care rarely holds. There is little evidence that incidence of surgical disease is the main source of variation in use of discretionary surgery. Rather, variations reflect differing medical opinion on appropriate use. Without evaluation, excessive use cannot be distinguished from underservice. Morbidity explains the variability of nondiscretionary surgery and conditions related to lifestyle. Access plays an important role for discretionary surgery. Geographic analysis can identify variation and relate incidence to socioeconomic and specific local effects. Hospital data do not permit direct assessment of appropriate care. Understanding the reasons for variation requires information beyond incidence data. The challenge is to identify and explain small area variations or to fix them.

Mental health care utilization in prepaid and fee-for-service plans among depressed patients in the Medical Outcomes Study.

Abstract: OBJECTIVE: We compare mental health utilization in prepaid and fee-for-service plans and analyze selection biases DATA SOURCE: Primary data were collected every six months over a two-year interval for a panel of depressed patients participating in the Medical Outcomes Study, an observational study of adults in competing systems of care in three urban areas (Boston, Chicago, and Los Angeles) STUDY DESIGN: Patients visiting a participating clinician at baseline were screened for depression, followed by a telephone interview, which included the depression section of the NIMH Diagnostic Interview Schedule Patients with current or past lifetime depressive disorder and those with depressed mood and three other lifetime symptoms were eligible for this analysis We analyze mental health utilization based on periodic patient self-report ANALYTIC METHODS: We use two-part models because of the presence of both nonuse and skewness of use Standard errors are corrected nonparametrically for correlations across observations due to clustered sampling within participating physicians and repeated observations on the same individual PRINCIPAL FINDINGS: The average number of mental health visits was 35-40 percent lower in the prepaid system, adjusted and unadjusted for observed differences in patient characteristics, including health status Utilization differences were concentrated among patients of psychiatrists, with only minor differences among patients of general medical providers Analyzing the effect of switches that patients make between payment systems over time, we found some evidence of adverse selection into fee-for-service plans based on baseline utilization, but not based on utilization at the end of the study In particular, after adjusting for observed patient characteristics and health status, patients switching out of prepaid plans had higher baseline use than predicted, whereas patients switching out of fee-for-service had lower use than predicted Switching itself appears to be related to an immediate decline in utilization and was not followed by an increase or "catch-up" effect CONCLUSIONS: The absence of the commonly found "catch-up" effect following switching and the significant decrease in utilization during the switching period suggests an interruption in care that does not occur for patients staying within a payment system This finding emphasizes the need for integrating new patients quickly into a system, an issue that should not be neglected in the current policy discussion

The roles of race and socioeconomic factors in health services research.

Abstract: For decades data have been collected comparing health care in racial and ethnic groups. The use of such groups in health services research assumes that standard, reliable, and valid definitions of race and ethnicity exist and that these definitions are used consistently. In fact, race is a term often used, but ill defined. It can incorporate biological, social, and cultural characteristics of patients and can refer to both genetic and behavioral traits. Various investigators have reported differences between racial and ethnic groups in health status, disease manifestation and outcome, resource utilization, and health care access, often specifying neither a definition of race nor the measurement they used to classify their study populations. The role of race as an explanatory variable in health services research requires greater scrutiny than many researchers currently provide. Many studies use race as a proxy for other socioeconomic factors not collected in the research effort. This article explores the ambiguities about race as an explanatory variable that render such research difficult to interpret. We suggest that health services researchers focus on nonracial socioeconomic characteristics that might be both more informative and more useful in guiding policy formation.

Black/white differences in prenatal care utilization: an assessment of predisposing and enabling factors.

Abstract: OBJECTIVE. This article reports on analysis of the predisposing and enabling factors that affect black/white differences in utilization of prenatal care services. DATA SOURCES. We use a secondary data source from a survey conducted by the Michigan Department of Public Health. STUDY DESIGN. The study uses multivariate analysis methods to examine black/white differences in (1) total number of prenatal care visits, (2) timing of start of prenatal care, and (3) adequacy of care received. We use the model advanced by Aday, Andersen, and Fleming (1980) to examine the effect of enabling and predisposing factors on black/white differences in prenatal care utilization. DATA COLLECTION. A questionnaire was administered to all women who delivered in Michigan hospitals with an obstetrical unit. PRINCIPAL FINDINGS. Enabling factors fully accounted for black/white differences in timing of start of prenatal care; however, the model could not fully account for black/white differences in the total number or the adequacy of prenatal care received. CONCLUSION. Although there are no black/white differences in the initiation of prenatal care, black women are still less likely to receive adequate care as measured by the Kessner index, or to have as many total prenatal care contacts as white women. It is possible that barriers within the health care system that could not be assessed in this study may account for the differences we observed. Future research should consider the characteristics of the health care system that may account for the unwillingness or inability of black women to continue to receive care once they initiate prenatal care.

Geographic variation in rates of selected surgical procedures within Los Angeles County.

Abstract: OBJECTIVE. We explore the contribution of income and ethnicity to geographic variation in utilization of surgical procedures. DATA SOURCES/STUDY SETTING. We assessed the use of eight procedures from 1986 through 1988 among residents of Los Angeles County using data from the California Discharge Dataset, the 1980 census, and other secondary sources. Procedures chosen for evaluation were coronary artery bypass grafting (CABG), coronary artery angioplasty, permanent pacemaker insertion, mastectomy, simple hysterectomy, transurethral prostate resection (TURP), carotid endarterectomy, and appendectomy. STUDY DESIGN. The amount of inter-zip code variation for each procedure was first measured using various estimates including the analysis of variance coefficient of variation (CVA). Population-weighted multivariate regression analysis was used to model variation in age- and gender-adjusted rates of procedure use among 236 residential zip codes. PRINCIPAL FINDINGS. Highest-variation procedures were coronary artery angioplasty (CVA = .392) and carotid endarterectomy (CVA = .374). The procedures with the lowest degree of variation were cardiac pacemaker implantation (CVA = .194) and hysterectomy (CVA = .195). Variation was significantly related to income (carotid endarterectomy) and either African American or Latino zip code ethnicity for all procedures except pacemaker implantation. For all procedures except appendectomy, the direction of the effect was toward fewer procedures with lower income. However, the effect of African American or Latino population ethnicity varied. CONCLUSIONS. In this large urban area both population ethnicity and socioeconomic status are significantly associated with the geographic utilization of selected surgical procedures.

The role of race/ethnicity and social class in minority health status.

Abstract: Minority health is often considered as a unitary phenomenon; it is often assumed that the health status of minority groups in the United States is similar across groups and much worse than that for whites Yet the reality is extraordinary diversity Racial/ethnic groups differ greatly both among and within themselves with regard to health status and with regard to a large number of other indices Mortality rates around the world generally show an inverse relationship with social class While this generally holds true in the United States as well, once again we see a strong interaction with race/ethnicity However, the mediating factors between race/ethnicity and social class, and health status are not well understood Especially in the face of health care reform, a broad-based research agenda needs to be undertaken so that any restructuring of the health care delivery system is informed by empirical information

Expanding the frame of health services research in the drug abuse field.

Abstract: OBJECTIVE. This article examines the distribution of heavy drug users across health and social service agencies in a community, and ways in which organizational and social policy factors influence pathways to services. DATA SOURCES/STUDY SETTING. Data are from the Community Epidemiology Laboratory, a project that includes comparable surveys of a wide variety of client, service provider, and general population groups in a single northern California county. STUDY DESIGN. The design is a cross-sectional analysis of patterns of service use and referral by heavy drug users distributed across a variety of service settings and in the general population. DATA COLLECTION. In-person, structured interviews by trained interviewers were conducted using comparable instruments, measures, sampling strategies, and fieldwork procedures. PRINCIPAL FINDINGS. The specialty drug treatment system serves only a small proportion of heavy drug users in the community. Large proportions of drug users are found in criminal justice, primary health, and welfare agencies. Patterns of service encounter and referral suggest that drug treatment clients typically have been in jail or on welfare prior to attending treatment, and are far less likely to have been referred to or from treatment by health providers. CONCLUSIONS. Health services research on drug abuse should expand its frame of reference to include services outside the specialty treatment sector. Drug treatment facilities are somewhat remote from other agencies in community service networks and are organizationally dependent on criminal justice and welfare systems. Further research should investigate both the costs and benefits of screening and providing services at earlier points of institutional involvement with drug abusers and the implications of interorganizational dependencies among criminal justice, welfare, and drug agencies for providers and clients.

The diffusion of innovation in AIDS treatment: Zidovudine use in two New Jersey cohorts

Abstract: OBJECTIVE. This study investigates patterns of utilization of zidovudine (ZDV) by gender, race, risk group, and other respondent characteristics following approval of this treatment. STUDY POPULATION. Longitudinal observational data were used on a demographically diverse population participating in New Jersey's Medicaid waiver program for persons with symptomatic HIV disease. DATA EXTRACTION METHODS. Claims data were merged with administrative data on demographic characteristics, risk group, and functional status. Periods of ZDV utilization were determined by analysis of pharmacy claims. DESIGN. The proportion of respondents ever using ZDV (treatment incidence) and the proportion of time on ZDV among users (treatment persistence) were analyzed for a cohort enrolling in 1987 and 1988, and for a cohort enrolling in 1989 and 1990, with follow-up of utilization through August 1992. For each cohort, bivariate analyses were used to compare incidence and persistence by patient subgroup; logistic regression was used to investigate the predictors of incidence in a multivariate model; and OLS regression was used to analyze proportion of time on ZDV among those with any ZDV use. PRINCIPAL FINDINGS. For the 1987-1988 cohort, substantial race, gender, and risk group differences in utilization were observed, even though all participants in this Medicaid population had financial coverage for ZDV treatment. Treatment incidence was significantly lower for blacks than for others in bivariate comparison (45 percent versus 63 percent had any use of ZDV) and in a logistic regression controlling for a variety of demographic and health status indicators (relative risk .46, CI .31 a variety of demographic and health status indicators (relative risk .46, CI .31 to .69). Treatment persistence differences were also substantial in the 1987-1988 cohort: among ZDV users, women, blacks, and injection drug users (IDUs) had significantly less persistence in use, and the gender and risk group differences were significantly in a multivariate model. In the 1989-1990 cohort, however, both incidence and persistence of treatment converged: no significant differences were observed across demographic groups. CONCLUSIONS. Less-advantaged subgroups lagged in access to this new therapy, suggesting the presence of nonfinancial barriers to care. However, these initial differences subsequently converged. RELEVANCE/IMPACT. Socioeconomic differences have been observed in access to newly introduced treatments for a variety of diseases, reflecting nonfinancial as well as financial barriers to care. Such differences may or may not disappear as use of therapies becomes institutionalized. Monitoring patterns of treatment initiation as well as persistence of treatment over time, using merged data from claims and administrative files, can provide important information on the diffusion of treatments and the extent to which initial disparities are or are not reduced over time.

Measuring hospital mortality rates: are 30-day data enough? Ischemic Heart Disease Patient Outcomes Research Team.

Abstract: OBJECTIVE. We compare 30-day and 180-day postadmission hospital mortality rates for all Medicare patients and those in three categories of cardiac care: coronary artery bypass graft surgery, acute myocardial infarction, and congestive heart failure. DATA SOURCES/COLLECTION. Health Care Financing Administration (HCFA) hospital mortality data for FY 1989. STUDY DESIGN. Using hospital level public use files of actual and predicted mortality at 30 and 180 days, we constructed residual mortality measures for each hospital. We ranked hospitals and used receiver operating characteristic (ROC) curves to compare 0-30, 31-180, and 0-180-day postadmission mortality. PRINCIPAL FINDINGS. For the admissions we studied, we found a broad range of hospital performance when we ranked hospitals using the 30-day data; some hospitals had much lower than predicted 30-day mortality rates, while others had much higher than predicted mortality rates. Data from the time period 31-180 days postadmission yield results that corroborate the 0-30 day postadmission data. Moreover, we found evidence that hospital performance on one condition is related to performance on the other conditions, but that the correlation is much weaker in the 31-180-day interval than in the 0-30-day period. Using ROC curves, we found that the 30-day data discriminated the top and bottom fifths of the 180-day data extremely well, especially for AMI outcomes. CONCLUSIONS. Using data on cumulative hospital mortality from 180 days postadmission does not yield a different perspective from using data from 30 days postadmission for the conditions we studied.

Conducting an assessment of health needs and resources in a racial/ethnic minority community.

Abstract: This article examines strategies and methodologic issues for researchers to consider when conducting community-based research within a racial/ethnic minority community. Members of minority communities have considerable skepticism about the health care system and researchers who work under its auspices. To facilitate quality research, it is necessary to build a mutually beneficial partnership between the community and researchers. Suggested strategies for accomplishing this goal, such as seeking out information on the social and political forces shaping the community and developing the community's capacity to undertake research of this type, are described. Methodologic issues include the importance of community input in defining the minority population group and its leadership, the benefits and limitations of conducting comparative analysis, and the need for measurement tools and techniques that are culturally and socially appropriate. Minority and nonminority researchers must make a concerted effort to gain knowledge of and respect for a community whose culture, values, and beliefs may differ from their own.

Relationship between physician specialty and the selection and outcome of ischemic stroke patients.

Abstract: Objective This study compares the initial characteristics, in-hospital treatment, and outcomes of stroke patients managed by neurologists versus non-neurologists. Data sources/study setting An inception cohort of 146 patients hospitalized with their first stroke at a private, a university, or a Veterans Affairs hospital. Study design A prospective, hospital-based, multisite cohort study design is used to ascertain the patients' initial 180-day in-hospital and postdischarge experience. Data collection/extraction methods Data were collected on patient demographics, disease severity, and risk factors for stroke; diagnostic tests and pharmacologic management; one-month and six-month all-cause and cause-specific mortality; and physical and functional impairment. Principal findings The 88 patients (60 percent) admitted to the neurology services of the study hospitals had better prognostic profiles, i.e., lower likelihood of having either completed stroke or cardiac comorbidity, and were less likely to die within one and six months of stroke onset. Conclusions Substantial systematic differences in the types of patients managed by neurologists versus other specialists may preclude comparison of outcomes of care.

The effect of an Rx-to-OTC switch on medication prescribing patterns and utilization of physician services: the case of vaginal antifungal products.

Abstract: OBJECTIVE. We examined the impact of over-the-counter (OTC) availability of vaginal antifungal products, beginning in January 1991, on medication prescribing patterns and utilization of physician services. DATA SOURCES AND STUDY SETTING. Data on utilization of health care services and prescription medications by female members (ages 11 and older) of the Fallon Community Health Plan (FCHP), a group model health maintenance organization and a component of the Fallon Health Care System of central Massachusetts. The census for such individuals increased from 49,551 in January 1990 to 67,365 in December 1992. DESIGN. Time-series analyses were employed to assess changes in prescribing patterns of vaginal antifungal products and physician visits for vaginitis from January 1, 1990 through December 31, 1992. Monthly numbers of prescriptions for vaginal antifungal products and physician visits per 100 members were measured. Monetary savings relating to the prescription-to-OTC switch were also estimated. DATA COLLECTION METHODS. The computerized management information system of FCHP contains records on utilization of all health care services and prescriptions filled, collected as part of routine fiscal activities. We identified all vaginally administered products on the FCHP formulary used for the treatment of vaginal candidiasis and determined the number of prescriptions filled for these agents during each month of the study period. We also identified the number of physician office visits characterized by the ICD-9-CM code 616.10 ("vaginitis and vulvovaginitis, unspecified") occurring during each month of the study period. PRINCIPAL FINDINGS. For the one-year period after OTC availability of vaginal antifungal products (January 1991 through December 1991), we estimated that the number of prescriptions dispensed for these products was reduced by 6.42 per 100 female FCHP members ages 11 and older. Physician visits for vaginitis were reduced by 0.66 per 100 members. Estimated savings to the Fallon Health Care System for the one-year period following OTC availability were $42,528 in medication costs and $12,768 to $25,729 for costs associated with physician visits, depending on use of laboratory testing in patient evaluations. CONCLUSIONS. The findings of this study suggest that the prescription-to-OTC switch of vaginal antifungal treatments reduced health care costs to the insurer in the managed care setting. These favorable effects on costs for the insurer need to be weighed against shifts in medication costs to consumers and potential adverse consequences to the patient relating to errors in self-diagnosis.

Longitudinal patterns of medical service use and costs among people with AIDS.

Abstract: OBJECTIVE. This study examines the effect of race, HIV transmission group, and decedent status on the use and cost of inpatient and outpatient care among people with AIDS. DATA SOURCES. Data come from 914 people with AIDS who were receiving services in nine cities across the United States in 1990-1991 and who indicated that a hospital clinic was their usual source of care. Review of hospital medical and billing records provided data on use and costs of medical services over an 18-month period. Vital status was determined from hospital records and death certificates. STUDY DESIGN. Data from each respondent were aggregated into three-month intervals, beginning with the last quarter of data and working backward. Regression analyses using random-effect models and generalized estimating equations were conducted to assess temporal patterns of inpatient and outpatient use and costs. PRINCIPAL FINDINGS. Inpatient utilization and costs were higher for decedents than for nondecedents. However, differences between decedents and nondecedents varied as a function of race. Nonwhites had more inpatient use and higher costs than whites, but lower outpatient use, and these differences were greater among decedents. Inpatient nights and costs rose sharply in the six months prior to death. Outpatient use and costs did not display as strong a temporal trend. CONCLUSIONS. Much of the cost of treating HIV infection is concentrated in the period immediately preceding death. The intensity of service use in the terminal period should be considered when developing estimates of annual costs of care and when designing programs to provide community-based treatment.

Beyond class, race, and ethnicity: deprivation and health in Britain.

Abstract: The concepts of class, race, and ethnicity figure prominently in health services research in Britain. Occupational class has been employed for nearly a century to investigate social inequalities in health and access to care. More recently, researchers have identified differences in health status and utilization between ethnic groups. This article examines how these constructs are defined in Britain and identifies some key research associated with them. It also draws attention to the considerable problems in using class and ethnicity to stratify the population. The authors conclude that a new approach that directly measures individuals' material and social resources needs to be developed.

Conceptualizing race in economic models of medical utilization: a case study of community-based elders and the emergency room.

Abstract: There is no consensus on the appropriate conceptualization of race in economic models of health care. This is because race is rarely the primary focus for analysis of the market. This article presents an alternative framework for conceptualizing race in health economic models. A case study is analyzed to illustrate the value of the alternative conceptualization. The case study findings clearly document the importance of model stratification according to race. Moreover, the findings indicate that empirical results are improved when medical utilization models are refined in a way that reflects the unique experiences of the population that is studied.

Understanding the factors behind the decision to purchase varying coverage amounts of long-term care insurance.

Abstract: OBJECTIVE. This article examines the factors related to an individual's decision to purchase a given amount of long-term care insurance coverage. DATA SOURCE AND STUDY SETTING. Primary data analyses were conducted on an estimation sample of 6,545 individuals who had purchased long-term care (LTC) insurance policies in late 1990 and early 1991, and 1,248 individuals who had been approached by agents but chose not to buy such insurance. Companies contributing the two samples represented 45 percent of total sales during the study year. STUDY DESIGN. A two-stage logit-OLS (ordinary least squares) choice-based sampling model was used to examine the relationship between the expected value of purchased coverage and explanatory variables that included: demographic traits, attitudes, risk premium, nursing home bed supply, and Medicaid program configurations. DATA COLLECTION. Mail surveys were used to collect information about individuals' reasons for purchase, attitudes about long-term care, and demographic characteristics. Through an identification code, information on the policy designs chosen by these individuals was linked to each of the returned mail surveys. The response rate to the survey was about 60 percent. PRINCIPAL FINDINGS. The model explains about 47 percent of the variance in the dependent variable-expected value of policy coverage. Important variables negatively associated with the dependent variable include advancing age, being married, and having less than a college education. Variables positively related include being male, having more income, and having increasing expected LTC costs. Medicaid program configuration also influences the level of benefits purchased: state reimbursement rates and the presence of comprehensive estate recovery programs are both positively related to the expected value of purchased benefits. Finally, as the difference between the premium charged and the actuarially fair premium increases, individuals buy less coverage. CONCLUSIONS. An important finding with implications for policymakers is that changes in Medicaid policy affect the decisions of consumers regarding the acquisition of private LTC policies as well as the level of protection chosen. This is particularly important to states interested in pursuing public-private partnerships in long-term care financing.

Impact of prior and current alcohol use on use of services by patients with depression and chronic medical illnesses.

Abstract: OBJECTIVE. Alcohol use often co-occurs with other major chronic conditions, but its effect on health care utilization in this context is not understood. This study examines the impact of alcohol consumption on health care use by patients with chronic medical conditions or depression, or both. DATA SOURCES/STUDY SETTING. Data came from the Medical Outcomes Study, an observational study of patients from the offices of general medical providers and mental health specialists in three U.S. cities. STUDY DESIGN. Longitudinal data spanning four years for outpatient general medical visits and outpatient mental health visits were analyzed using a two-part model to assess the impact of alcohol use disorder, problem drinking, and current and past alcohol consumption on health care use by patients, controlling for patient demographics and health status. DATA COLLECTION/EXTRACTION METHODS. Data were collected from 2,546 adult patients with hypertension, diabetes, heart disease (congestive heart failure or myocardial infarction), and/or current major depression or subthreshold depression using periodic, self-report surveys detailing health care utilization and health status information. PRINCIPAL FINDINGS. Current alcohol consumption increases outpatient doctor visits, and problems related to current drinking decrease outpatient mental health visits. CONCLUSIONS. Patterns of alcohol consumption have an impact on both mental health and overall health care use by patients with chronic medical conditions or depression.

Local markets and systems: hospital consolidations in metropolitan areas.

Abstract: OBJECTIVE. This study examines the formation of local hospital systems (LHSs) in urban markets by the end of 1992. We argue that a primary reason why hospitals join LHSs is to achieve improved positions of market power relative to threatening rivals. DATA SOURCES/DATA COLLECTION. The study draws from a unique database of LHSs located in and around metropolitan statistical areas (MSAs). Data were obtained from the 1991 AHA Annual Hospital Survey, updated to the year 1992 using information obtained from multiple sources (telephone contacts of systems, systems lists of hospitals, published changes in ownership, etc.). Other measures were obtained from a variety of sources, principally the 1989 Area Resources File. STUDY DESIGN. The study presents cross-sectional analyses of rival threats and other factors bearing on LHS formation. Three characteristics of LHS formation are examined: LHS penetration of urban areas, LHS size, and number of LHS members located just outside the urban boundaries. LHS penetration is analyzed across urban markets, and LHS size and rural partners are examined across the LHSs. PRINCIPAL FINDINGS. Major hypothesized findings are: (1) with the exception of the number of rural partners, all dependent variables are positively associated with the number of hospitals in the markets; the rural partner measure is negatively associated with the number of hospitals; (2) the number of doctors per capita is positively associated with all but the rural penetration measure; and (3) the percentage of the population in HMOs is positively associated with local cluster penetration and negatively associated with rural system partners. Other findings: (1) average income in the markets is negatively associated with all but the rural penetration measure; (2) LHS size and rural partners are both positively associated with nonprofit system ownership; and (3) they are also both negatively associated with the degree to which their multihospital systems are geographically concentrated in a single state. CONCLUSIONS. The findings generally support the argument that LHS formation is the product of hospital providers attempting to improve positions of power in their local markets.

The impact of prenatal exposure to cocaine on newborn costs and length of stay.

Abstract: OBJECTIVE: Our intention is to determine newborn costs and lengths of stay attributable to prenatal exposure to cocaine and other illicit drugs. DATA SOURCES AND STUDY SETTING: All parturients who delivered at a large municipal hospital in New York City between November 18, 1991 and April 11, 1992. STUDY DESIGN: A cross-sectional analysis used multivariate, loglinear regressions to analyze differences in costs and length of stay between infants exposed and unexposed prenatally to cocaine and other illicit drugs, adjusting for maternal race, age, prenatal care, tobacco, parity, type of delivery, birth weight, prematurity, and newborn infection. DATA COLLECTION/EXTRACTION METHODS: Urine specimens, with linked obstetric sheets and discharge abstracts, provided information on exposure, prenatal behaviors, costs, length of stay, and discharge disposition. PRINCIPAL FINDINGS: Infants exposed to cocaine or some other illicit drug stay approximately seven days longer at a cost of $7,731 more than infants unexposed. Approximately 60 percent of these costs are indirect, the result of adverse birth outcomes and newborn infection. Hospital screening as recorded on discharge abstracts substantially underestimates prevalence at delivery, but overestimates its impact on costs.

Impact of Medicare payment reductions on access to surgical services

Abstract: OBJECTIVE. This study evaluates the impact of surgical fee reductions under Medicare on the utilization of surgical services. DATA SOURCES. Medicare physician claims data were obtained from 11 states for a five-year time period (1985-1989). STUDY DESIGN. Under OBRA-87, Medicare reduced payments for 11 surgical procedures. A fixed effects regression method was used to determine the impact of these payment reductions on access to care for potentially vulnerable Medicare beneficiaries: joint Medicaid-eligibles, blacks, and the very old. DATA COLLECTION/EXTRACTION METHODS. Medicare claims and enrollment data were used to construct a cross-section time-series of population-based surgical rates from 1985 through 1989. PRINCIPAL FINDINGS. Reductions in surgical fees led to small but significant increases in use for three procedures, small decreases in use for two procedures, and no impact on the remaining six procedures. There was little evidence that access to surgery was impaired for potentially vulnerable enrollees; in fact, declining fees often led to greater rates of increases for some subgroups. CONCLUSIONS. Our results suggest that volume responses by surgeons to payment changes under the Medicare Fee Schedule may be smaller than HCFA's original estimates. Nevertheless, both access and quality of care should continue to be closely monitored.

Identifying complications and low provider adherence to normative practices using administrative data.

Abstract: OBJECTIVE. This study investigated whether unexpected length of stay (LOS) could be used as an indicator to identify hospital patients who experienced complications or whose care exhibited low adherence to normative practices. DATA SOURCES AND STUDY SETTING. We analyzed 1,477 cases admitted for one of three medical conditions. All cases were discharged from one of nine participating Department of Veterans Affairs (VA) hospitals from October 1987 through September 1989. Analyses used administrative data and information abstracted through chart reviews that included severity of illness indicators, complications, and explicit process of care criteria reflecting adherence to normative practices. STUDY DESIGN. We developed separate multiple linear regression models for each disease using LOS as the dependent measure and variables that could be assumed present at the time of admission as explanatory variables. Unexpectedly long LOS (i.e., discharges with high residuals) was used to target complications and unexpectedly short LOS was used to target cases whose care might have exhibited low adherence to normative practices. Information gleaned from chart reviews served as the gold standard for determining actual complications and low adherence. PRINCIPAL FINDINGS. Analyses of administrative data showed that unexpectedly long LOS identified complications with sensitivities ranging from 40 through 62 percent across the three conditions. Positive predictive values all were at greater than chance levels (p < .05). This represented substantial improvement over identification of complications using ICD-9-CM codes contained in the administrative database where sensitivities were from 26 through 39 percent. Unexpectedly short LOS identified low provider adherence with sensitivities ranging from 33 through 45 percent with positive predictive values all above chance levels (p < .05). The addition to the LOS models of chart-based severity of illness information helped explain LOS, but failed to facilitate identification of complications or low adherence beyond what was accomplished using administrative data. CONCLUSIONS. Administrative data can be used to target cases when seeking to identify complications or low provider adherence to normative practices. Targeting can be accomplished through the creation of indirect measures based on unexpected LOS. Future efforts should be devoted to validating unexpected LOS as a hospital-level quality indicator. RELEVANCE/IMPACT. Scrutiny of unexpected LOS holds promise for enhancing the usefulness of administrative data as a resource for quality initiatives.