Measuring psychological change during cognitive behaviour therapy in primary care: a Polish study using 'PSYCHLOPS' (Psychological Outcome Profiles).

Background: Psychological outcome measures are evolving into measures that depict progress over time. Interval measurement during therapy has not previously been reported for a patient-generated measure in primary care. We aimed to determine the sensitivity to change throughout therapy, using 'PSYCHLOPS' (Psychological Outcome Profiles), and to determine if new problems appearing during therapy diminish overall improvement.
Methods: Responses to PSYCHLOPS, pre-, during- and post-therapy were compared.
Setting: patients offered brief cognitive behaviour therapy in primary care in Poland.
Results: 238 patients completed the pre-therapy questionnaire, 194 (81.5%) the during-therapy questionnaire and 142 the post-therapy questionnaire (59.7%). For those completing all three questionnaires (n = 135), improvement in total scores produced an overall Effect Size of 3.1 (2.7 to 3.4). We estimated change using three methods for dealing with missing values. Single and multiple imputation did not significantly change the Effect Size; 'Last Value Carried Forward', the most conservative method, produced an overall Effect Size of 2.3 (1.9 to 2.6). New problems during therapy were reported by 81 patients (60.0%): new problem and original problem scores were of similar magnitude and change scores were not significantly different when compared to patients who did not report new problems.
Conclusion: A large proportion of outcome data is lost when outcome measures depend upon completed end of therapy questionnaires. The use of a during-therapy measure increases data capture. Missing data still produce difficulties in interpreting overall effect sizes for change. We found no evidence that new problems appearing during therapy hampered overall recovery.