Stefan Elbers

37 Evaluating IMPT programmes over time Characteristics of the Included Studies Study Characteristics. We included 37 case series, (i.e. longitudinal studies with prospectively collected data on one cohort of patients), 19 RCTs, 4 N-RCTs and 5 other types of study design. In total, these studies provided data on 74 cohorts with a median final follow-up measurement of 12 months (range 12 - 120). Pretreatment sample sizes of the cohorts ranged from 10 to 2089, with a median of 102. After filtering out cohorts that reported no dropout in the study flow (assuming a complete case analysis) the median dropout ratio for the 53 remaining cohorts was 8.56 (range 0 – 42.11) at post treatment and 16.67 (range = -3.85 – 62.17) at final follow-up, using the posttreatment sample size as reference. Six of the 65 studies were evaluated as low risk of bias, indicating that in the majority of the cohorts we identified at least one factor that threatened the internal validity. Statistical analysis and attrition (67.69%), unclear inclusion criteria (41.54%), incomplete inclusion of participants (61.54%), and incomplete reporting of clinical information of the participants (32.31%) were the most frequent reasons for assigning an unclear or high risk of bias. Appendix 1 shows a summary of the risk of bias assessment. Patient Characteristics. The majority of the included patients were treated in European or North-American countries (94.6%). The mean distribution of sex was 68.8% female ( SD = 18.1%) The average mean age of the cohorts was 44.39 years ( SD = 4.47). The median pain duration prior to treatment, reported in 40 of the included cohorts, was 76.8 months (range = 16 – 217.2). Generally, patients were referred to the programme by their primary care physician or medical specialist. Table 1 provides a summary of the patient characteristics. The multimedia appendix includes a more extensive and searchable table.