2001) and the American Managed Behavioral Health Association. The federal government also has adopted a framework through its State Outcomes Measurement and Management System (described below) (SAMHSA, undated-b). These efforts are in addition to performance measure sets that address health care overall and include some M/SU performance measures, such as NCQA’s HEDIS and measures used by VHA (see Appendix C).
All of these efforts have tackled two enduring and related problems that are encountered in all performance measurement efforts: (1) the tension between having measures of high validity, reliability, and ease of calculation and having a broader set of measures that is more representative of the populations and conditions of interest; and (2) the difficulty of achieving consensus on the measure set across all stakeholders (Hermann and Palmer, 2002). In addition to these problems, conceptualizing a framework for M/SU health care is more complex than doing so for general health care for the reasons discussed below.
More-diverse stakeholders The larger number of disciplines licensed to diagnose and treat M/SU problems and illnesses relative to those licensed to diagnose and treat general health conditions potentially requires the involvement of a greater number of stakeholder groups in a consensus process. Moreover, as discussed earlier, M/SU health care involves both specialty and general medical providers. In addition, the involvement of the education, juvenile and criminal justice, and child welfare systems as payers and providers of M/SU services means performance measures selected for M/SU health care must be determined with input from these stakeholders, who are not typically involved in general health care. Consumer advocates also have been very active in shaping the delivery of M/SU health care, again with implications for the numbers and diversity of stakeholders in a consensus process.
Difference between the public and private sectors Although general health care is delivered in both the public and private sectors, in M/SU health care the public sector serves a population with a clinical profile much different from that of the population served by the private sector—most often those with severe and chronic illnesses. Thus, measures that may be meaningful to private-sector stakeholders may be less useful to those in the public sector. In NCQA’s HEDIS measures for general health care, for example, some measures15 are designated for calculation for Medicaid populations but not for privately insured populations (NCQA, 2004b). This practice may need to be employed more widely for M/SU health care. Even measures appropriate for multiple populations may need to be reported separately.