Development and application of quality assurance methods for interventions in randomised controlled trials of surgical oncology: the ROMIO study (a comparison of minimally invasive and open oesophagectomy)

TL;DR

This study develops and tests quality assurance methods for a cancer surgery trial comparing minimally invasive and open oesophagectomy procedures.

Contribution

The paper introduces feasible quality assurance measures for surgical interventions in pragmatic cancer trials.

Findings

All centers met entry criteria, and most surgeons submitted required videos.

Intra-operative photographs were submitted for most procedures but only a small fraction were complete.

CRFs showed better adherence to surgical components than photographs.

Abstract

Results of RCTs are criticised because the quality assurance (QA) of surgical interventions is not considered. This is particularly true in cancer trials, because higher standards of surgery may confer more favourable outcomes. Although methods for surgical QA exist, it is unclear how to operationalise and report them in the context of pragmatic cancer trials. We describe the development and application of QA processes to an RCT comparing laparoscopically assisted (LAO) and open oesophagectomy (OO) in patients with localised oesophageal cancer.

Three QA measures were developed in Phase 1 and tested for feasibility in Phase 2: (i) centre/surgeon entry criteria, (ii) agreement of key components of LAO/OO, and (iii)monitoring adherence to intervention protocols using CRFs and intra-operative photographs.

All centres met entry criteria and 30/31 Phase 2 surgeons submitted two videos. Although photos were received for 88.8% of procedures, only 44(14.9%) were complete. Adherence to key intervention components (abdominal/thoracic nodal clearance, hiatal dissection) was consistently reported as better in CRFs than that observed in the intra-operative photographs.

Embedding QA measures into pragmatic surgical cancer RCTs is feasible, and provides important data about the quality of interventions. Methods to streamline data collection and analyses are needed prior to widespread use.

Figures