Predictive factors for postoperative constipation and continence after stapled transanal rectal resection.

BACKGROUND

Although stapled transanal rectal resection (STARR) has become an important surgical option in the treatment of obstructive defaecation syndrome, objective data about parameters that predict its success or failure are not yet available.

METHODS

Medical history, clinical and radiomorphological data were obtained prospectively from a multi-institutional STARR registry. Predictive factors for postoperative constipation (Cleveland Clinic Constipation Score, CCS) and incontinence (Cleveland Clinic Incontinence Score, CCIS) were identified using univariable and multivariable analysis.

RESULTS

Data were obtained for 181 of 201 patients in the STARR registry, with completed median follow-up of 19·4 (range 12-41) months. Although the CCS decreased significantly overall (from mean(s.d.) 16·3(4·9) to 6·7(4·1); P < 0·001), 31 patients (17·1 per cent) complained about persisting constipation. CCIS levels remained unchanged overall, but 16 patients (8·8 per cent) had new-onset faecal incontinence. Multivariable analysis revealed that rectocele (β = -0·302, P < 0·001) and intussusception (β = -0·392, P < 0·001) were independent predictors of low CCS levels, and intussusception (β = -0·216, P = 0·001) and enterocele (β = -0·171, P = 0·012) were independent predictors of low CCIS levels. In contrast, small rectal diameter (β = -0·293, P < 0·001), low squeeze pressure (β = -0·188, P = 0·005) and increased pelvic floor descent at rest (β = 0·264, P < 0·001) predicted high CCIS levels.

CONCLUSION

Factors for a favourable outcome after STARR included rectocele, intussusception and enterocele, whereas small rectal diameter, low sphincter pressure and increased pelvic floor descent were unfavourable. These findings should be integrated into the therapy algorithm for STARR.