Maarten van Egmond

88 Chapter 5 used, taking into account the heterogeneity of patients across studies. If meta-analysis was not possible, a narrative overview of the findings was presented, including tabular summaries of extracted data. Where data were missing, the authors attempted to contact the study authors. An intention- to-treat analysis was conducted where possible. Otherwise, data were analyzed as reported. Loss to follow-up information was documented and assessed as a source of potential bias. Subgroup analysis has been executed on type of surgery, time of intervention (pre- or postoperatively), type and duration of intervention (mono- or multidisciplinary, consultation, monitoring, training), type of communication technology, the healthcare provider and the method of implementation (as new care for something that did not exist before or in addition to existing care). Due to the small number of studies included in each subgroup, heterogeneity was not assessed, and it was not possible to assess reporting bias using funnel plots. Selective outcome reporting was assessed using the Cochrane risk of bias tool. 23