Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Conclusions The committee reached several interrelated conclusions. First, with some critical caveats, the HCFA data bases can be used to compare and contrast patterns of care, both diagnostic and therapeutic, across a variety of commu- nities, care settings, and providers. Thus, the use of specific treatment modalities that have been documented to be effective in clinical trials in younger patients, which therefore can be expected to be routinely employed in older patients, can be analyzed. The committee pointed out those clinical issues, particularly the use of thrombolytic agents, angioplasty, and bypass surgery alone or in combination that are most in need of ongoing surveillance. Second, the use of HCFA data sets is limited by several vexing problems that will hinder effectiveness research on the Medicare population, at least for AMI. These limitations include the variability in diagnostic coding of AMI, the lack of relevant ambulatory data both before and after hospitaliza- tion, and the inadequacy of coding for common interventions and new inter- ventions as they are put into practice. Third, although the most drastic outcome of AMI death - is often documented even if it occurs outside the hospital, the data set lacks good outcome measures such as long-term functional health status. This major limitation can be overcome by determined effort, which clearly includes a commitment to collecting a broad set of health-related quality-of-life infor- mation at least in part directly from patients. Fourth, the common use of risk stratification methods at various phases in the diagnosis and treatment of AMI provides opportunity for natural experiments in effectiveness where clinical trials would be impossible or too 43
44 EFFECTIVENESS INITIATIVE expensive. With improvements in databases and recording of outcomes as cited above, the effectiveness of these stratification methods in identifying the most appropriate care for Me elderly should be readily testable. In some caseswhen conclusions cannot be drawn from, say, the HCFA a~ninis- trative data or effectiveness research projects alone effectiveness analy- ses may indicate the need for specific RCI~s, and the committee judged that this would be a valuable input into improving health care delivery and health services research over the long run. Analysis of observational natural experiments based on risk stratification of patients by clinicians must include adjustment for risk of patients selected to receive different treatments, including adjustments related to comorbidity. Methods must be developed to identify and adjust for death or severity of illness and comorbidity in groups of patients prior to the assignment of treatment plans, perhaps based on surrogates such as prior hospitalization or by random samples of these groups from the Medicare data base. Without such methodological attention in advance, comparisons of outcomes with different treatments will be suspect. Substantial reductions in mortality over the last 25 years have not changed the status of AMI as the leading cause of death. The availability of powerful new drugs and technologies holds promise for even further improvements, but their effectiveness in the elderly is often unclear. This workshop highlighted the value of available patient and administrative records in evaluating the effectiveness of these measures in the elderly. Moreover, ii provided direction to future research in assessing the long-term outcomes of AMI and in identifying those diagnostic and treatment modali- ties most appropriate for clinical trials. This research will be essential in ensuring survival and high quality of life for many elderly in the face of increasing health care costs.