Treatment Outcomes |
Published: October, 2001
|
Treatment outcomes are reported as "effect size," a type of standardized change score calculated by dividing change during treatment (raw score difference in scores from start to end of treatment) by the standard deviation (variability of change scores) of the outcome measure. An effect size of zero means no change on average resulted from psychotherapy. The average effect size for psychotherapy studies is estimated to be .82, while minimal treatment controls effect size averages .42. 1The overall combined effect size for the PBH treatment sample is .36. While on the surface this appears to be a moderate effect size at best, this conclusion is misleading. The probability of improvement with treatment is proportional to the severity of distress at intake. Patients with more symptoms are much more likely to show improvement than those with minimal distress.
The PBH sample contains a large percentage of cases with minimal levels of distress as measured by the outcomes questionnaires. Since psychotherapy studies often select for individuals that meet certain diagnostic criteria, the subjects are more likely to have high levels of symptoms and thus show more improvement. If the PBH sample is reanalyzed after excluding cases with intake scores below 44, the effect size is .8, comparable to that found in psychotherapy studies.
The ALERT system avoids the problem of the dependency of effect size on the severity of the sample by using the Outcome Index statistic. The Outcome Index is case mix adjusted in that it represents the deviation from a target outcome, and thus is independent of the magnitude of the effect size per se.
1Lambert, M.J. Weber, F.D. & Sykes, J.D. (1993, April) Psychotherapy versus placebo. Poster presented at the annual meeting of the Western Psychological Association, Phoenix.
Back to Top
  | 
