The authors concluded that patients with ulcerative colitis appear to be at higher risk of pouch-related fistulation and pouchitis after restorative proctocolectomy than patients with familial polyposis coli. The reliance upon biased, retrospective comparative observational studies means that any conclusions are likely to be suggestive rather than definitive.

To evaluate adverse events and functional outcomes after restorative proctocolectomy in patients with familial adenomatous polyposis coli (FAP) compared to patients with ulcerative colitis (UC).

MEDLINE, EMBASE and the Cochrane Library were searched for studies published between 1980 and 2005; the search terms were reported. No language restrictions were applied. In addition, studies were tracked using the ‘Related articles’ function and references were screened.

Studies that compared adverse events and functional outcomes after restorative proctocolectomy between patients with FAP and UC were eligible for inclusion. Results data for any of the following outcomes had to be either presented or calculable: short-term adverse events including wound infection, 30-day post-operative all-cause mortality, anastomotic leak, pelvic sepsis, failure to construct pouch intra-operatively, and small bowel obstruction; long-term adverse events including pouch failure, pouchitis, post-operative pouch-related fistulation; and functional outcomes evaluated at similar times in both groups, including frequency of defecation every 24 hours and at night, faecal incontinence, stool seepage, day and night time pad use, antidiarrhoeal medication, quality of life and sexual function.

Where reported, the mean age of the patients ranged from 26.8 to 40 years (age range: 5 to 77) and the duration of follow-up ranged from 1.2 to 186 months.

Two reviewers independently selected the studies.

A modified version of the Newcastle-Ottawa Scale was used to assess the quality of randomised and non-randomised studies using items for patient selection, comparability of the treatment groups and outcome assessment. Studies scoring 7 or more stars were considered to be high quality.

The authors did not state how the validity assessment was performed

Two reviewers independently extracted the data. For each study, the numbers of patients with outcomes of interest were reported for each treatment group, together with the odds ratio (OR) or weighted mean difference (WMD) and 95% confidence interval (CI).

Pooled ORs (for dichotomous data) and WMDs (for continuous data) were calculated, along with 95% CIs, using random-effects models. Studies with zero events in each treatment group were excluded from the analysis. A value of 0.5 was added to cells with zero events in one treatment group.

Statistical heterogeneity was tested using the χ² statistic. Sensitivity analysis was undertaken by analysing the following subgroups: sample size greater than 200 patients; studies published during 1995 or later; studies scoring 6 or more stars; studies reporting histologically defined pouchitis; and studies reporting pouchitis in cases where the pouch-anal anastomosis was hand sown. The potential for publication bias was explored graphically using a funnel plot.

Fifteen independent studies (5,199 patients: 782 with FAP and 4,417 with UC) were included. There was one prospective non-randomised study (n=104) and one retrospective case-matched study (n=78); the rest were retrospective studies.

In 14 studies the groups were matched on one or more characteristics; nine matched for pouch design and method of pouch-anal anastomosis; and eight matched for use or omission of diverting-loop ileostomy. Eight studies scored 6 or more stars for validity.

Adverse events.

Patients with UC were significantly more likely to develop post-operative fistulas (10.4% versus 4.8%; OR 2.31, 95% CI: 1.43, 3.73, p<0.001; 8 studies, n=1,766; no significant heterogeneity) and pouchitis (30.1% versus 5.5%; OR 6.44, 95% CI: 3.21, 12.93, p<0.001; 12 studies, n=2,843; significant heterogeneity was found, p=0.05) than patients with FAP. There was no statistically significant difference between patients with UC and FAP in post-operative mortality (7 studies, n=1,519), anastomic leak or stricture (5 studies, n=1,047), pelvic sepsis (8 studies, n=2,309), wound infection (4 studies,n=401), incisional hernia (4 studies, n=379), small bowel obstruction (6 studies, n=1,303), pouch failure, (9 studies, n=3,031), sexual dysfunction (3 studies, n=954) or post-operative urinary dysfunction (1 study, n=110).

Functional outcomes.

Patients with FAP reported significantly lower stool frequency per 24 hours than patients with UC (WMD 0.99, 95% CI: 0.21, 1.76, p=0.01; 7 studies, n=1,274; significant heterogeneity was found, p<0.001). There were no statistically significant differences between patients with UC and FAP for other functional outcomes.

A funnel plot based on rates of pouchitis showed some evidence of publication bias.

Larger studies showed an increased rate of pouchitis in UC compared with FAP patients. Higher quality studies showed no statistically significant differences between patients with UC and FAP for stool frequency per 24 hours or rates of pouch failure, but showed significantly higher rates of fistula formation, pouchitis and anastomotic leak in UC patients. Recent studies showed no significant difference between UC and FAP patients in stool frequency, pouch failure and anastomotic leak. Studies that reported histologically confirmed pouchitis showed a statistically significant increase in pouchitis in UC compared with FAP patients.

The results of other subgroup analyses were also reported.

Findings suggest that patients with UC are at higher risk of pouch-related fistulation and pouchitis after restorative proctocolectomy than patients with FAP, but there were no differences in the rates of other adverse events. Further research is required.

The review question was defined in terms of the intervention, participants and outcomes. The inclusion criteria relating to study design were broad, which seems appropriate in view of the studies identified. Several relevant sources were searched and attempts were made to minimise language bias. No attempts to minimise publication bias were reported, but the potential for publication bias was assessed. Validity was assessed, although only the composite score was presented; this makes it difficult to independently comment on the reliability of the evidence presented. Methods were used to minimise reviewer error and bias during the review process. Appropriate methods were used for the meta-analyses, heterogeneity was assessed, and various predefined subgroup analyses were conducted. Evidence was based on predominantly potentially biased, retrospective comparative observational studies; stronger evidence would come from prospective matched studies. The reliance upon retrospective studies means that any conclusions are likely to be suggestive rather than definitive.

Practice: The authors stated that patients with UC who are considering restorative proctocolectomy should be advised about the increased risk of pouchitis.

Research: The authors stated the need for comparative studies to evaluate long-term patient satisfaction, pouch failure, quality of life and the impact of increased stool frequency in patients with UC.

Royal College of Surgeons of England.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.