Studies of process--outcome correlations in medical care evaluations: a critique

The validity of process evaluations of medical care has been challenged by a number of studies which purport to show that process and outcome measures are unrelated. However, each of the studies had numerous methodological flaws which biased their results against finding a relationship: either their outcome measures had questionable validity, their research designs were inappropriate, or the statistical analyses were poorly conceived. Better studies have found significant, although modest, correlations between process and outcome measures. Since the validity of outcome measures has never been determined, there is little reason at present for believing that outcome measures are more valid than process measures.