Criteria for Selection

In selecting indicators of the quality of medical care for evaluation, OTA considered the perspectives of consumers, the medical profession, research, and policy. As indicated in table 3-6, OTA attempted to incorporate indicators perceived to be valid by consumers and by those in the medical, research, and policy communities. Each of these groups is using certain indicators to assess quality, often without thorough evaluation of the indicators' validity. Subjecting such indicators to intensive examination could validate their appropriateness or elucidate problems with their use.

Since OTA's task is to evaluate indicators of quality that consumers could use to choose physicians and hospitals, the public's requirements for information received high priority. People are most likely to face decisions about medical care for the conditions that have the highest incidence and prevalence in the United States. The most common causes of physician office visits, hospitalizations, disability days, and death were the basis of the entries in tables 3-4 and 3-5. As one would expect, the most frequent afflictions vary

by age and sex. In addition, the circumstances and type of medical condition influence how consumers choose providers. One survey organization reported that, on average, 22 percent of consumers selected a hospital on their own, without their physicians' advice; in cases involving accident or injury, however, 33 percent chose the hospital independently. People were also more likely to act on their own in choosing a hospital for general tests and treatment (29 percent) and for illness and maternity (27 percent) than for surgery (17 percent) (320).

Also important in OTA's selection was that the indicators taken together relate to the aspects of care that are important to people (see table 3-1 and ch. 2). People have reported being particularly concerned about humaneness and communication of information, including information on primary prevention (392).

Other considerations in selecting indicators to evaluate hinged on the state of medical knowledge. Given current information and technology, certain events, such as maternal deaths, should

occur only rarely, and their occurrence often raises concern about the quality of care. Especially in the past 50 years, medical advances have enabled providers to intervene in the natural progression of many medical conditions, to restore function or to prevent further decline. But most techniques, even well-accepted ones, have not been well evaluated, and many may lack efficacy. Consequently, it is reasonable to restrict evaluations of quality to the application of technologies with demonstrated efficacy and to conditions with efficacious interventions.

By drawing indicators from the different research approaches used to evaluate quality (structure, process, and outcome), OTA hoped to gain insight into advantages and disadvantages of each approach. To ensure the feasibility of its own research, OTA limited its analysis to indicators for which sufficient published and unpublished information existed to support an evaluation.

Reflecting the interest of Congress and other policymakers, OTA paid particular attention to indicators that quality assessors are using or considering, especially for public programs. Also in line with policy interests, OTA wished to target conditions or interventions where quality problems are likely because of overuse or underuse of particular procedures.

Indicators Selected for Evaluation

OTA selected the following eight categories of indicators for intensive evaluation:2

• hospital mortality rates, for the overall institution, by department, and by condition or procedure;

• adverse events that affect patients, as exemplified by nosocomial (institutionally acquired) infections in hospitals;

⚫ formal disciplinary actions by State medical boards, sanctions recommended by utilization and quality control professional review organizations (PROS) and imposed by the U.S. Department of Health and Human Services (HHS), and malpractice compensation; • evaluations of physicians' performance for a specific condition, as exemplified by physicians' care for hypertension;

• volume of services performed in hospitals and by physicians;

⚫ scope of hospital services, with particular emphasis on emergency services, cancer care, and neonatal intensive care units;

• physician specialization; and patients' assessments of their care.

2App. A contains more information about the selection process.

Table 3-6.-Considerations in Selecting Indicators of Quality for OTA Evaluation

Consumer interests:

• High-frequency conditions or reasons for seeking

care

• Indicators together cover range of what is important to people

• Indicators together relate to general population, particular age-sex categories, and vulnerable groups Medical interests:

• Conditions for which medical care can alter the natural history

• Events that should not occur

• Conditions or interventions where quality problems are likely from overuse or underuse of particular procedures

• Indicators perceived as valid by medical community Research interests:

• Information available to support an evaluation • Indicators that relate to different approaches to assessing quality (structure, process, and outcome) Policy interests:

• Indicators frequently considered to assess quality • Indicators being used to assess quality SOURCE: Office of Technology Assessment, 1988.

Taken together, these eight indicators relate to a range of medical providers, types of medical care, aspects of care, approaches to quality assessment, and sources of data (see table 3-7). Hospital mortality rates and scope of hospital services apply only to hospitals, and physician specialization applies most directly to physicians. Five of the indicators-adverse events, disciplinary actions and malpractice compensation, evaluation of physicians' performance for a specific condition, volume of procedures, and patient ratings— could apply to both physicians and hospitals. This report does not explicitly consider indicators of quality for HMOs and other alternative delivery systems; however, quality assessors could use these indicators to evaluate physicians and hospitals associated with such organized delivery systems as well as physicians and hospitals operating more independently.

All but one of the eight indicators evaluated in this report pertain to the evaluation of general rather than condition-specific care. Only the evaluation of physicians' performance through hypertension screening and management pertains to a specific condition, but the evaluation of other indicators touches on age- and sex-specific conditions for which people frequently seek care. The

analysis of hospital mortality rates examines mortality rates for specific departments, such as neonatal intensive care units; and the analysis of volume of procedures examines procedures for several specific conditions, such as appendectomy, hysterectomy, coronary artery bypass graft, total hip replacement, prostatectomy, and acute myocardial infarction. Whether a hospital's scope of services is adequate depends on what medical conditions the hospital treats. Although this report does not explore them in depth, some adverse events, such as maternal death, relate to specific conditions.

Each of the indicators that OTA chose for evaluation is associated with 1 or more of the 10 specific aspects of medical care that were listed in table 3-2. As shown in table 3-7, hospital mortality rates, adverse events, State disciplinary actions, PRO/HHS sanctions, and malpractice compensation could result from deficiencies in any of several aspects of care. Patients' assessments are associated with a number of matters of particular concern to consumers: the responsiveness of a provider to urgent situations, the personal respect or humaneness accorded a patient, the communication of desired information, and the performance of primary preventive activities. Review of the care given for hypertension would give information on almost the entire range of medical care aspects.

The eight indicators encompass the range of approaches to assessing quality: structure, process, and outcome. Two indicators-hospital mortality rates and adverse events that affect patientsenumerate undesirable effects on patient health. Both pertain almost exclusively to physiologic health and physical function. State disciplinary actions, PRO/HHS sanctions, and malpractice compensation are indicators that straddle the process and outcome categories; patients or colleagues may undertake malpractice and disciplinary actions because of providers' negligence in the provision of medical care, but the allegedly negligent behavior may attract notice because of adverse effects on patients' health or satisfaction. The review of physicians' care for a specific medical condition, such as hypertension, entails scrutinizing aspects of the medical care process. Three indicators-volume of procedures provided

Table 3-7.-Issues Addressed by the Indicators Selected for OTA Evaluation