Prognostic factors of adherence 59 databases. No language restrictions were imposed. All reference lists of included studies were reviewed manually for eligible studies. In addition, grey literature sources (OpenGrey.eu, NARCIS.nl, DART-Europe.org, OATD.org) were searched using the term ‘exercise adherence’. We considered full text reports of cohort or cross-sectional studies and the experimental arm of randomized trials reporting prognostic factors associated with adherence to home-based exercise in patients (³18 years) prescribed individual home-based exercise therapy for a chronic disease, i.e., conditions that last one year or more and require ongoing medical attention or limit activities of daily living or both [17]. We considered all prognostic factors the authors associated with adherence. Adherence was defined as minutes of exercise completed or as number of sessions of exercise completed. For the outcome all validated and non-validated methods used for objectifying adherence were considered. If more than one publication was based on the same cohort or population reports were clustered [16]. Study selection First we screened the title and abstract for potentially relevant studies, after Covidence automatically de-duplicated search results and facilitated further study selection [18]. An exception on the two-reviewer process was protocoled if the search revealed more than 5000 hits. In that case, the title/abstract round was assessed independently by ER and AD for the first 25% of the titles/abstracts. If there was less than 5% difference between the results (i.e., in at least 95% of cases, ER and AD came to the same conclusion regarding inclusion or exclusion), ER screened the remaining titles/abstracts [19]. Otherwise, ER and AD both continued to perform the screening phase independently. Any study reviewed as “yes” or “unsure” was included to full-text review. When an abstract was identified but the full text was not available, we contacted the corresponding author via e-mail to obtain the full text. Next, we assessed the full text of all potentially relevant studies by applying the inclusion criteria. Data extraction For extracting data, we created a data extraction form based on the CHARMS checklist [20] in Covidence. Then, data was extracted by one reviewer (ER) and discussed with the other reviewer. Extracted data included information about the study: publication details (author, year, country), study design; the participants: sample size, gender and age, details of their chronic disease; information about the prognostic factors: studied variables associated with adherence with their effect estimates, and standard deviation or 95% confidence interval or p-values; and information about the outcome: used definition and measure of adherence. To structure the data, the identified prognostic factors were categorized into the five WHO-domains; 1. Patient-related, 2. Social/economic, 3. Therapy-related, 4.
RkJQdWJsaXNoZXIy MjY0ODMw