OCR Text |
Show Uses of Hospital Discharge Data for Population-Based Health Assessment by Robert T. Rolfs, M.D. and Wu Xu, Ph.D. Introduction Assessment, one of the three principal functions of public health, involves data collection, management, and most importantly, analysis and dissemination of the resulting information to support and guide policy development and activities to assure the health of the population (Institute of Medicine, 1988). Assessment activities should identify health problems, pinpoint segments of the population with disproportionate health problems, provide a basis for prioritizing those problems, and guide and monitor the effectiveness of interventions intended to improve or protect the population's health (Kepple and Freedman, 1995). Population-based data are usually needed for assessment. Population-based data generally pertain to a known population that is defined by residence within geographic or political boundaries1 (Last, 1988). Epidemiology provides the approach and methods to analyze population-based health data for public health. Different kinds of data have been used for public health assessment. Some of the most commonly used have been vital statistics (deaths and births), notifiable disease reports, and surveys such as the Behavioral Risk Factor Surveillance System (Centers for Disease Control [CDC], 1993). Despite the importance of clinical medical care to the health of a population, public health has infrequently used clinical care data as a source of information about the population. Hospital discharges are an important source of data about clinical care. Hospital discharge data record patients' demographic characteristics and diagnoses, procedures performed, source of payment, and other elements describing a hospitalization in a standardized manner. While hospital discharges from a single hospital (among several serving a population) are not population based, a data set that includes the discharges from all the hospitals serving a population can be a population-based assessment tool. This paper will discuss and provide examples of some population based uses of hospital discharge data. Determinants of Hospitalization and of Hospitalization Rates The measurement of epidemiologic interest is a rate (number of events divided by the number of persons in the population at risk during a defined time period), rather than just an event or a number of events. However, hospital discharge rates do not directly measure the occurrence of an illness (or other health-related condition). Hospitalization is more appropriately considered as a type of treatment. Thus, hospitalization rates are determined by the following factors: (1) the occurrence (incidence and prevalence) of health conditions that might be treated in a hospital; (2) whether, when, and where individuals seek treatment; and (3) whether health care providers decide an individual is to be treated in a hospital. The occurrence, severity, and timing of health conditions are affected by factors that include: the environment, age and sex composition, socioeconomic status of the population, and the prevalence of risk behaviors such as smoking, diet, and exercise. For some conditions, whether an individual seeks treatment is discretionary and likely to be affected by that individual's health beliefs, including perceptions of risk and of the health care system. For most conditions, when and where a patient seeks care will affect the likelihood of hospitalization. For example, a patient with asthma or congestive heart failure who seeks care early when symptoms are mild may be able to receive outpatient treatment and avoid hospitalization. For such conditions, access factors (including insurance coverage) are likely to be important. Finally, health care providers, jointly with the patient, determine whether the patient needs treatment and whether that treatment should be provided in a hospital. Those decisions may be further influenced by health care system factors such as the supply of physicians and of hospital beds (Ginsburg et al., 1983; Rice, 1983), the specialty mix of physicians in the community, by reimbursement levels and mechanisms (e.g., capitated versus fee-for-service), and by other financial incentives such as physician ownership of treatment facilities (e.g., an outpatient surgery center). (Hillman et al., 1990; Hillman et al., 1989) Variation in practice style may also be important, possibly due to one or more influential clinicians who espouse a particular approach to treatment. (Chassin, 1993) The determinants of hospitalization are summarized in Figure 1 (see page 35). Users of hospital discharge data are often interested in one or a few of these factors that 26 Hospital Discharge Data |