Chapter 5 94 recommendations decreased over time [24]. All these results emphasize the importance of tailored made care and target services. Post hoc power analysis The sample size of this study was calculated for the main study (manuscript in preparation) and thus not specifically for this study. If no difference is found between groups, this may be the result of a type II error [25]. To ensure that this is not the case, a post hoc power analysis was performed. The results showed that the study had sufficient power to demonstrate any significant difference between the different measurement times, if any. Strengths and limitations The strengths of this study are inclusion of patients from different physiotherapy practices and from patient-organizations in the Netherlands and Belgium. In this way, a representative sample of COPD patients attending a PR program participated in this study, which enhances the generalizability of the study results. It must be taken into account, however, that these are patients who receive rehabilitation for a prolonged time (70% of patients in this cohort followed PR for ³ 1 year). This is important to notice because it is plausible that in patients who have been attending a PR program for a prolonged time, and in whom no intervention is offered in terms of adherence, adherence will be relatively stable. This is in contrast to patients who have just started attending a PR program. Another strength of the study is that missing data was examined and imputed, although multilevel analysis can deal with missing data [26]. Multiple imputation was used in this process. Methodologists currently regard multiple imputation as a state-of-the-art technique because it improves accuracy and statistical power relative to other missing data techniques [21]. This study also has some limitations. This study was conducted during the COVID19 pandemic and may have affected motivation of patients for rehabilitation. This was demonstrated by the study by Menting et al. In their research group, one in five people reported having modified their self-management during the pandemic, both positively and negatively, out of fear of becoming ill [27]. If patients have been more consistently adherent or non-adherent, this may have led to a slightly biased result; adherence is less constant over 12 months than this study has shown. However, not comparing well, previous studies also came up with similar results. Furthermore, the results showed that there were some patients with variation in adherence. This variation was not further explored. For this study, this was not necessary, as the aim was to understand mean adherence over 12 months. Chapter 6 is trying to understand the factors that predict adherence in this population and thus may be able to explain the variation in adherence.
RkJQdWJsaXNoZXIy MjY0ODMw