br Post intervention outcomes br The changes in
3.1. Post-intervention outcomes
The changes in empowerment, distress, and quality of life from pre-intervention to post-intervention/ post waiting did not differ significantly between the intervention group and the control group (see Table 3).
3.2. Follow-up outcomes
The increase of empowerment scores and quality of life and the decrease of distress in the intervention group from pre-interven-tion to follow-up assessment differed significantly from the respective difference scores of the control group (see Table 4). Empowerment, quality of life, and distress improved in the intervention group but not in the control group.
3.3. Inverse probability weighting analyses
In order to investigate whether participant dropout biased the findings, observations were re-weighted by the inverse of the probability of being observed. A logistic regression was conducted with distress, gender, disease stage, and education entered in the model to determine predictors of dropout at post-intervention and follow-up. Education at follow-up was the only significant predictor (Odds Ratio = .40; B = 0.93; p = .04), which indicates that SC 236 patients with lower educational attainment were more likely to drop out of the study. Weighted regressions were conducted with the outcome measures entered as the dependent variables and group allocation entered as the independent variable. Table 5 summarizes the results of the inverse-probability weighted (IPW) analyses. We noted a few small differences between the complete case and inverse probabil-ity weighted results. At the post-intervention assessment point, levels of distress had decreased significantly in the intervention group, while prior analyses revealed no differences. No other significant differences were noted at post-intervention/ post-wait time for either group. At follow-up, distress remained significant. However, quality of life was no longer significant and empower-ment almost reached statistical significance (p = .06). These differences to the complete case analyses might be due to the fact that individuals with lower education were slightly more likely to
Demographic and illness-related characteristics of the sample at baseline.
Characteristic Intervention group Control group X2
Age groups in years
not complete the study. As such, participants with lower educational attainment, who did not drop out, had a higher weight when data was re-weighted. Note that the direction of the associations and the general conclusions from the inverse-probability weighted analyses remained the same as those in the complete case analyses. The only differences are in the standard errors of the effects and some slight differences in magnitude.
4. Discussion and conclusion
The findings of the present study indicate that a self-help workbook can be efficacious in improving psychological well-being
Treatment adherence assessed at post intervention (26 n 29).
a Of the four patients, who indicated that hydrophytic leaves spent on average 0 min per week reading or practicing the prescribed intervention, one indicated having read 80%, two indicated that they read 100% of the workbook, and one did not answer the first question. All four of these participants positively endorsed items 3, 4, 8, and 9.
Differences between intervention and control group in changes from baseline to post-intervention/ post-waiting for empowerment, distress, and quality of life.
Baseline M (SD) Post M (SD) dIG of M Baseline Post M (SD) dCG of M
Notes: d = difference between score at baseline and at post intervention/ post 6 weeks waiting; IG = intervention group; CG = control group; CI = confidence interval.