A plea for consistency in the reporting of surgical series illustrated with an analysis of 51 follow-up reports of pulmonary metastasectomy in colorectal carcinoma

Abstract

The data contained in the follow-up studies of surgical series should ideally be reported in such a way that they are comparable from one follow-up study to another and amenable to amalgamation to draw statistically robust inferences. The authors provide an analysis of 51 reports of pulmonary metastasectomy for colorectal cancer to illustrate the problem. It is in the nature of follow-up studies that data are more plentiful where they concern explicit descriptors of the patient when assessed for surgery and the conduct of the operation itself. Data may be sparse in other respects, because although of interest in a subsequent analysis of outcome, they were not of direct clinical importance at the time of surgery and so were not recorded. The authors cannot surmise about unrecorded data or use them in any analysis. The authors constructed a "wish-list" of items and tested the 51 publications against it. They also suggest that the form in which data are presented should include data summaries, which can be amalgamated. For example, means can be aggregated, provided the numerator is known, but medians cannot. In an ideal world, anonymized patient-level data should be retained and remain accessible for future researchers, but there are substantial obstacles for saving and sharing data of this kind. A check list for reports of pulmonary metastasectomy series is offered as a template.