Measurement and interpretation of patient-reported outcomes in surgery: an opportunity for improvement.

BACKGROUND

Surgery may have a profound effect on patients' health-related quality of life (QOL). To be optimally useful, trials that seek to guide clinical decision making should measure outcomes that are important to patients and report the results in a clinically meaningful way. We sought to explore how researchers currently measure and interpret QOL in surgical trials, using gastric cancer as a case study.

METHOD

We performed a systematic review of randomized controlled trials (RCTs) of gastric cancer surgery published between 1966 and 2009 that included at least one patient-reported outcome (PRO). Investigators assessed trial eligibility and extracted data in duplicate using standardized forms, then resolved disagreements by consensus.

RESULTS

Our search identified 87 RCTs of gastric cancer surgery, of which 11 (13%) included at least one PRO. Ten RCTs measured one or more validated PROs, although six also included ad hoc measures. All manuscripts presented the results as raw scores and nine of the 11 trials identified a statistical difference between groups. All 11 manuscripts prominently reported the PRO results in the abstracts and conclusions, but only one discussed the clinical significance of the differences between groups.

CONCLUSIONS

Most RCTs of gastric cancer surgery do not include measures of QOL and those that do suffer from important limitations. RCTs would be more useful to surgeons and patients if authors measured PROs and utilized existing approaches to present the results of PROs in ways that provide an intuitive sense of the magnitude of effects.