Patient outcomes are the ultimate indicator of the quality of care received. These important measures reflect the extent to which providers are delivering high- or low-quality care, as well as the functioning of the broader health care system. However, outcomes measures are often difficult to use for quality improvement purposes. Multiple confounders are associated with health outcomes, such as patient adherence, societal factors, and the long time frames required to yield significant results. Health services researchers have wrestled for decades with models for statistical adjustment that can protect against holding providers accountable for such confounding determinants of outcome. Much progress has been made in understanding simple adjustments, such as for age, as well as more difficult ones, such as for comorbidity. In addition, more is now understood about the policy implications of the choices of adjusters. For example, race may correlate with outcome. A mortality measure that adjusts for race in effect excuses the care system from responsibility for race-related factors—a decision with profound policy implications.
The committee believes consideration should be given to measures of two outcomes of care particularly salient to patients: disease-specific mortality and pain control.
Disease-specific mortality The committee proposes the following 30-day and 1-year disease-specific mortality measures for consideration: acute myocardial infarction, coronary artery bypass graft, percutaneous coronary intervention, and end-stage renal disease. Sufficient epidemiologic work has been done on these measures to permit both appropriate adjustments for demographics and comorbidities and widespread practical adoption. As always, consideration must be given to the available sample sizes and to the expression of risks in terms of confidence intervals. It is likely that small hospitals and individual practitioners simply have too few relevant cases to be included in a disease-specific mortality measurement strategy.
Pain control Qualitative studies have demonstrated that recognition and treatment of pain are important priorities for patients receiving palliative care; however, validated measures of pain control are not yet available for widespread use (Lynn, 2000). The committee proposes that the NQCB consider which measures of pain assessment and changes in pain management are the most promising candidates for implementation, with evidence for their reliability and validity. (For more discussion of the evidence base and areas for further research with respect to pain control, see Appendix J.)