Aim: Early phase studies are essential to evaluate new technologies prior to randomized evaluation. Evaluation is limited, however, by inconsistent measurement and reporting of outcomes. This study examines outcome reporting in studies of innovative colorectal cancer surgery.
Methods: Systematic searches identified studies of invasive procedures treating primary colorectal adenocarcinoma. Included were a random sample of studies which authors reported as 'new' or 'modified'. Outcomes were extracted verbatim and categorized using an existing framework of 32 domains relevant to early phase studies. Outcomes were classified as 'measured' (where there was an explicit statement to that effect or evidence that data collection had occurred) or 'mentioned but not measured' (where outcomes were discussed but data collection was not evident). Patterns of identified outcomes are described.
Results: Of 8373 records, 816 were potentially eligible. Full-text review of a random sample of 218 studies identified 51 for inclusion of which 34 (66%) were 'new' and 17 (33%) were 'modified'. Some 2073 outcomes were identified, and all mapped to domains. 'Anticipated disadvantages' were most frequently identified [660 (32%) outcomes identified across 50 (98%) studies]. No domain was represented in all studies. Under half (944, 46%) of outcomes were 'measured'. 'Surgeon's/operator's experience of the innovation' was more frequently 'mentioned but not measured' [207 (18%) outcomes across 46 (90%) studies] than 'measured' [17 (2%) outcomes, 11 (22%) studies].
Conclusion: There is outcome reporting heterogeneity in studies of early phase colorectal cancer surgery. The adoption of core outcome sets may help to resolve these inconsistencies and enable efficient evaluation of surgical innovations.
© 2020 The Association of Coloproctology of Great Britain and Ireland.