Chapter 9 168 more. Therefore, adherence in this group was expected to be constant since no intervention was provided to improve adherence. The prospective cohort study showed indeed that adherence was constant over a period of 12 months in this cohort. Also, measured health outcomes (exercise capacity and health related quality of life) remained constant in this same period. It is interesting to note that the patients who were adherent for some time remained so and had better health outcomes than those who were non-adherent. If there is a causal relationship between adherence and treatment outcomes [8], then these findings confirm the importance of adherence to health outcomes. Now that adherence over time was known in this cohort of COPD patients, the prediction model could be developed. However, to avoid selection bias and overfitting as much as possible, it was important to use a prespecified model. To create a prespecified model, it was necessary to know what predictors are already known from the literature. In chapter 4, a systematic review and meta-analysis were performed to identify what variables are already known in literature to be associated with adherence to home-based exercise therapy in patients with chronic diseases. To organize and interpret the evidence, WHO's five domains were used: 1. Patientrelated, 2. Social/economic, 3. Therapy-related, 4. Condition-related, and 5. Health system factors [3]. High-quality evidence supported that higher exercise adherence was predicted by Perceived Behavioral Control (PBC). Moderate-quality evidence supported that higher exercise adherence was predicted by higher self-efficacy, having an exercise history, being motivated, having higher education, and having less comorbidities. Low-quality evidence supported that higher exercise adherence was predicted by better physical health, having fewer depressive symptoms, and being less fatigued. The included studies mostly assumed that adherence is a patient-driven problem. The WHO's therapy-related and health system factor domains hardly appeared. On the one hand, this made the results found applicable to the Dutch situation (external validity); on the other hand, it confirms that more research should be done on the influence of the health system (which may present differences between countries) and therapy-related factors on adherence. In addition, it is generally accepted in paramedic professions that the treatment regimen alone cannot fully determine the patient's outcome [9]. The relationship between patient and therapist (therapeutic alliance) has been viewed as an important determinant of treatment outcome and is considered central to the therapeutic process [10]. Therefore, we chose to consider alliance when developing the prediction model. Because adherence in this cohort was constant over a 12-month period, the study discussed in Chapter 6 developed a prediction model for the probability of adherence that is predictive of this time period. The potential predictors identified in chapter 4, alliance and the Theory of Planned Behavior (TPB) were used at the basis for creating a prespecified model. The final model included four predictors; intention, depression, MRC-score and alliance. To make the prediction model easy
RkJQdWJsaXNoZXIy MjY0ODMw