Results in Brief

GAO found reasons for concern about current systems for monitoring
Medicare quality of care. Review methods of uncertain validity are
being used, quality of care problems identified by one set of reviewers
are not coordinated for action by others, and the accuracy of key infor-
mation is questionable. Further, HHS's strategy for developing quality
assurance methods is inadequate to meet future program needs. GAO
identified short-term efforts that could lead to significant improvements
in the problems identified. However, GAO also found that developing a
comprehensive quality assurance research base and creating a program
for incorporating this knowledge into Medicare quality assurance efforts
would require a long-term commitment that cannot be adequately sup-
ported by current resources.

Principal Findings

The Effectiveness of
Review Methods Has Not
Been Evaluated

Quality-Related Review
Activities Are Poorly
Coordinated

The effectiveness of the medical review activities of carriers, intermediaries, and PROS in identifying quality problems or positively changing physician or provider behavior has not been evaluated. The oversight of medical review activities focuses on whether they meet contract specifications, most of which relate to controlling utilization, rather than on whether these activities effectively identify quality problems or lead to improvements in medical care.

The medical review activities of carriers, intermediaries, and PROS with respect to quality of care are virtually independent. Each set of reviewers applies both its own and HCFA-developed computer and manual edits and screens to billing and medical record data. They independently build profiles of provider or practitioner performance to identify possibly problematic patterns of service delivery or patient outcomes. Profiling results are not routinely shared among PROS, carriers, or intermediaries reviewing care in the same geographic areas. Possible quality of care problems found by carriers and intermediaries are not systematically reported to either HCFA or PROS. If a patient qualifies for Medicare posthospital services and the intermediary suspects that the hospital discharge may have been premature, HCFA is notified. But if posthospital coverage is denied by Medicare, there is no mechanism, other than beneficiary complaints, for notifying either HCFA or PROS about possibly inappropriate or premature discharges. HCFA's planned reorganization of Medicare data systems provides an opportunity to substantially improve coordination of quality monitoring activities.

Executive Summary

Quality of Care Data Are of Questionable Accuracy and Not Generalizable

HHS's Strategy for
Research and Development
on Quality Assurance Is
Inadequate

Recommendations to the Secretary of HHS

The data that support medical review as well as the information generated by the reviews are of questionable accuracy. GAO's review indicates, however, that information on patient diagnosis on outpatient physician bills can prove useful in identifying possible utilization and quality problems; some carriers have independently added this information to their screening systems. Neither the independent activities of each contract review system, nor the HCFA systems for validating the accuracy of medical reviews are currently designed to generate national estimates of the incidence or distribution of quality problems. However, HHS plans for improvements in surveying long-term care facilities offer possibilities for generating useful national, longitudinal data on patients' rehabilitative and chronic care needs and associated quality problems.

HHS is supporting many useful studies addressing aspects of quality of care and, in particular, studies related to refining measures of health care outcomes. Nevertheless, GAO found no clearly defined strategy or organizational structure for integrating information on the quality of health care provided to Medicare beneficiaries or for developing the underlying methods and knowledge base to meet future needs. This strategy needs to include the development of a structured program to specify good clinical practice, incorporate that practice into standards and quality assurance methods, and test incentives for practitioners to adopt them. Without this, Medicare will continue to be unable to provide the information needed to ensure Medicare quality of care. Developing a comprehensive Medicare quality assurance program would require additional funding of research, evaluation, and quality assessment program operations.

To strengthen the process of medical review, GAO recommends evaluations of (1) the comparative effectiveness of carrier and intermediary screens and profiles as means to identify inappropriate and substandard care (p. 39), (2) the methods PROS use to review medical records and the utility of current methods for establishing their quality of care contract objectives (p. 64), and (3) the methods PROS use to review quality of care in Medicare prepaid health plans (p. 64). To improve coordination of medical review activities, GAO recommends (4) that formal guidelines be developed to coordinate the systematic and timely reporting by carriers and intermediaries to PROS and HCFA of possible quality of care problems (p. 41), and (5) that studies be initiated to assess the strengths and weaknesses of the division of responsibilities among carriers, intermediaries, and PROS for processing and screening Medicare claims data

Matter for
Consideration by the
Subcommittee

Agency Comments

and performing medical reviews to identify quality of care problems and
substandard providers and suppliers (p. 64). To improve Medicare qual-
ity of care data, GAO recommends that (6) PROS, intermediaries, and car-
riers document and report incidents in which key data elements for
monitoring quality of care are inaccurate (p. 82), (7) physicians provide
diagnostic information on part B claims submitted to carriers, who
would forward this information to HCFA for inclusion in central data files
(p. 81), and (8) the data used to evaluate PRO medical reviews include
the information necessary to generate national estimates of quality
problems (p. 83). Finally (9), GAO recommends a high priority be
assigned to developing a centralized data file including nationally repre-
sentative information on the health care status, care needs, and health
care outcomes of home health care patients as well as nursing home
residents (p. 83).

The Subcommittee on Health should consider developing legislative proposals to assign specific research and development responsibilities to a new federal entity or existing entities for (1) developing, disseminating, and coordinating activities intended to define good medical practice and develop improved quality assurance methods, and (2) incorporating this knowledge into Medicare quality assurance efforts. Possible locations for these activities are discussed in chapter 7 (p. 110).

HHS agrees in principle with all or part of most of GAO's recommendations, but has not presented specific plans for implementing changes. The agency does not agree with GAO's recommendation for addressing errors in billing, and has concerns about the possible cost of evaluating the utility of screening by claims processors to detect potential quality of care problems. HHS also disagrees with aspects of GAO's recommendation regarding the evaluation of PRO reviews of care provided in prepaid health care plans. Finally, HHS does not support the creation of a new federal entity responsible for the development of quality assurance methods, but does state that improvements in internal coordination efforts will be considered. HHS believes that a new federal entity would create problems of duplication of effort. GAO's view is that such an entity, which could be located in new or existing HHS offices, would perform key functions essential for sound quality assurance that are not currently the clear responsibility of any HHS office or agency. The full text of HHS comments and our response are presented in appendix VII.