Academic journal article American Journal of Psychotherapy

Practice-Based Evidence: 45 Years of Psychotherapy's Effectiveness in a Private Practice

Academic journal article American Journal of Psychotherapy

Practice-Based Evidence: 45 Years of Psychotherapy's Effectiveness in a Private Practice

Article excerpt

Of 2,259 patients seen during 45 years of private practice, outcome data was produced for 1,599 cases. The mean (SD) number of sessions per case was 18.82 (29.89). The dropout rate was 18.76%. Of all treated cases with outcome data 4 (0.25%) were rated as Much Worse; 11 (0.69%), Worse; 497 (31.08%), No Change from Intake; 546 (34.15%), Improved; and 541 (33.83%), Much Improved. The mean (SD) pre-/post-treatment effect size (ES) was 1.90 (1.61), the median was 1.62, and the range was from -2.91 to +15.22. Patients and parents of minors rated outcomes more positively than the therapist did. Outcome varied significantly across diagnostic categories. There was a significant, positive relationship in length of treatment and outcome. The therapist's effectiveness did not improve across the years. Years with the largest patient caseloads or the greatest proportion of patients with managed-care insurance tended to show the poorest outcomes.1

KEYWORDS: practice-based evidence; private practice; treatment effectiveness; treatment effect size; evidence-based practice


About 60 years ago Eysenck (1952) reviewed the published research on adult psychotherapy and concluded, "The figures fail to support the hypothesis that psychotherapy facilitates recovery from neurotic disorder" (p. 323). Five years later Levitt (1957) reviewed the research on child and adolescent psychotherapy and reached a verdict similar to that of Eysenck: "... the results of the present study fail to support the view that psycho- therapy with 'neurotic' children is effective" (p. 195). These two articles triggered many criticisms from throughout the world, but these two psychologists updated their literature reviews and came to the same conclusions made in their first articles (Eysenck, 1961; Levitt, 1963).

In spite of such negative reviews of research on psychotherapy, psy- chotherapists continued to practice. On rare occasions they reported their results. For example, Heilbrunn (1966) evaluated her outcomes from 17 years of practicing psychoanalysis and psychoanalytic therapy and pub- lished her results in the American Journal of Psychotherapy. She claimed that 77 of 173 patients (i.e., 45%) improved; however, she excluded more than 80 patients seen for less than 20 sessions. When I read her paper a few months before becoming a licensed psychologist in California, I resolved to do something similar. That is the purpose of the present paper.

Throughout the 1950s and 1960s researchers were responding to the challenges posed by Eysenck and Levitt. In 1970 Meltzoff and Kornreich reviewed that research. They concluded that well-designed and controlled research had demonstrated very positive outcomes from psychotherapy. They also reviewed research on characteristics of patients and therapists that contribute to positive treatment outcomes, and on patient-therapist relationship variables that make a difference. Although their review pro- vided encouragement to psychotherapists in all work settings, it did not reveal what kind of outcomes were obtained by therapists in private practice. It did not identify brief outcome measures suitable for repeated administrations to gauge patient change across time.

Seven years later Smith and Glass (1977) introduced a quantitative approach for performing literature reviews of controlled treatment-out- come studies on adults. They called it "meta-analysis." Three years later Smith, Glass, and Miller (1980) expanded and updated the previous review. Although many methodologists criticized their approach, many other psychologists (and researchers from many other disciplines) adopted and adapted meta-analysis for reviewing research findings. Many investi- gators have performed meta-analyses of controlled treatment-outcome research on psychotherapy for children and adolescents. In analyzing the findings from 27 meta-analyses of child, adolescent, and adult psychother- apy research, the mean and standard deviation effect size (ES) was 0. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.