Inter-observer variability in the reporting of dynamic evacuation proctography

Abstract

To determine the degree of inter-observer variability between two reporters of dynamic evacuation proctography (DEP). The videotapes of 136 patients who had undergone the investigation of DEP in a 1-year period were retrospectively and independently reported by two of the authors. The authors recorded the presence or absence of an anterior rectocoele (as well as size-small, medium or large), an intussusception, a rectal prolapse and any degree of non-relaxing puborectalis syndrome (NRPS). The results were analysed using the κ statistic. I.W.I. reported 18% of the proctograms as normal. S.G.S. reported 26% of the proctograms as normal. κ scores for agreement on the presence or absence of rectal prolapses, NRPS, intussusception and anterior rectocoeles between the two reporters were 0.79, 0.70, 0.54 and 0.82, respectively. Usefulness of an investigation is dependent upon its reproducibility and hence on the degree of inter-observer variability in its reporting. Our study shows that DEP is a useful investigation for delineating the mechanics of the pelvic floor and that consistency of reporting in our Department is good.