Creating Price Indexes for Measuring Productivity in Mental Health Care

Economists have long suggested that to be reliable, a preferred medical care price index should employ time-varying weights to measure outcomes-adjusted changes in the price of treating an episode of illness. In this article, we report on several years of research developing alternative indexes for the treatment of the acute phase of major depression, for the period 1991-1996. The introduction of new treatment technologies in the past two decades suggests well-known measurement issues may be prominent in constructing such a price index.
We report on the results of four successively refined methods to measure price changes. In contrast with Bureau of Labor Statistics (BLS) price indexes, we find prices decline and productivity increases. In Method 1, we employ a publicly available retrospective claims database to create BLS-like price indexes for two discrete inputs used in the treatment of depression: antidepressant drugs and physician's services. In Method 2, we move to an episode-based price index. As a first attempt to adjust for outcome variations, we limit the analysis to episodes receiving one of seven guideline-level treatments, each of which have been shown to have equivalent efficacy in clinical trials. In Method 3 we expand the number of episodes included and allow for finer adjustments in expected outcomes. We assign expected outcomes to episodes based on an expert panel's rating of the effectiveness of the treatment received. In Method 4, our preferred price index, we use hedonic regression to adjust also for changes in patient characteristics that add to treatment complexity.
We find that our preferred method suggests a decline in the AAGR of treatment for depression of 2.7 percent, which is in contrast to a price increese of 2.6 percent per year when using BLS-like methods. This research suggests constructing episode-based, outcomes-adjusted price indexes is a complex and cumbersome task but an important one for public policy decisions. Although it may not be sensible or practical for the BLS to produce such an index on a monthly basis, it is important that policy analysts use episode-based outcomes-adjusted price indexes when evaluating productivity based on National Health Accounts.