Chapter 8 192 Variations in the results of RCTs on IMR may be due to differences in the patient population, sample size, type of control group, duration of IMR, level of IMR exposure (completion), and model fidelity (11) (Table 1). Comparing fidelity and completion Regarding the impact of IMR, our study found indications of the relevance of fidelity and the level of completion. A minimum threshold of exposure to IMR may be required for treatment effects to occur (8) (Chapter 5). Therefore, in addition to the percentage of participants exposed to >50% of the scheduled sessions, the total number of sessions attended is relevant for evaluating treatment efficacy. We suggest that the negative results of two IMR trials (8-10, 12) (Table 2) may be related to their low completion rates (Table 1 and Chapter 5). There is evidence regarding the predictive validity of fidelity from programs in the former National EBP Project in the United States (13), including ACT (14-18), IPS (19), and prior to our study (Chapter 5) also IMR (7, 20). In our study, assessed using the IMR Fidelity Scale, the average fidelity was slightly lower than the weighted mean of six studies reported in an IMR review (11). This might be partly due to the relatively high degree of refinement of the applied assessment procedure because, in the present study, the assessment procedure included all components of the protocol (13, 21). The average fidelity measured using IT-IS in our study was also lower than that measured in an earlier RCT (8). However, this was markedly higher than that reported in another study on implementing IMR in community practice (22). Previously, it was suggested that model fidelity in clinical trials was higher than in naturalistic contexts (22). However, we suggest that the implementation support in the current study was limited compared to what is apparent in clinical trials (22). In the current trial, clinicians were not highly trained, nor were they included in the study because of their demonstrated skills, nor did they receive implementation support in the form of session feedback or consultation. However, "interested clinicians" (11) were included, and they did indeed receive supervision, and one-time fidelity monitoring. This may explain why our IT-IS scores were lower than those in one RCT (8) but higher than those in community practices (22). Our results regarding the implementation levels of the different elements measured with the IT-IS are largely in line with the results of a previous study (22), in which five elements with the lowest fidelity—medication management, weekly action planning, action plan follow-up, cognitive-behavioral techniques, and significant other involvement—were among the seven elements with the lowest
RkJQdWJsaXNoZXIy MjY0ODMw