Colonic Lavage for the treatment of Severe C. Difficile Infections: The London Protocol
Author(s):
Kalvin Lung, University of Western Ontario; Biniam Kidane, University of Western Ontario; Greig McCreery, University of Western Ontario; Michael Ott, University of Western Ontario; Neil Parry, University of Western Ontario; Daryl Gray, University of Western Ontario; Chris Vinden, University of Western Ontario; Ken Leslie, University of Western Ontario; Tina Mele, University of Western Ontario
Background: Severe Clostridium difficile infection (CDI) is associated with high mortality & may require subtotal colectomy. Based on encouraging results of colonic irrigation via ileostomy (“Pittsburgh protocol” [PP]), we developed a non-invasive protocol of colonic irrigation via nasojejunal (NJ) tube (“London protocol” [LP]) to use in poor surgical candidates with severe CDI. Our objective was to compare patient outcomes with these 2 treatment approaches.
Hypothesis: We hypothesized that colonic lavage via the London Protocol is safe, effective treatment of severe CDI patients.
Methods: This was a retrospective cohort study of consecutive CDI patients refractory to medical therapy referred for surgical consultation. In addition to antibiotics, the LP group received 8L of polyethylene glycol over 48 hours via NJ tube. The PP and LP were used at our tertiary hospital in the last 2 years. Consulting surgeons decided whether to use PP or LP and proceeded to colectomy if they felt that PP or LP had failed. Patients who went directly to colectomy over the last 5 years were used as a comparison group. Intention-to-treat analysis, Fisher’s exact test & ANOVA were used.
Results: Thirteen, 9 and 17 patients underwent LP, PP and direct colectomy, respectively. There were no significant differences between groups with respect to age, sex, ASA class, ICU admission, vasopressor use, presence of peritonitis, hypotension, tachycardia, WBC >20 or <4, mean lactate or creatinine at time of surgical consultation (p>0.05). More PP patients failed treatment & required colectomy compared to LP (5/9 vs 1/13, p=0.02). In-hospital mortality rates were 15% (2/13), 44% (4/9) and 41% (7/17) for the LP, PP and direct colectomy groups, respectively (p=0.26). Of the 2 mortalities in the LP group, both had early deviations from protocol. One patient had active care withdrawn by family the day after LP was initiated and died that day. The 2nd patient started LP at a delayed stage, could not tolerate NJ irrigation on the 1st day of LP and thus, had immediate colectomy & died within 48 hrs. Mortality rates, excluding these 2 deaths, were significantly lower in LP group (p=0.03).
Conclusions: Colonic irrigation via NJ tube in the context of the LP appeared to be safe and effective in management of severe CDI. We found no significant differences in disease severity between groups. Per-protocol analysis shows mortality reduction with use of this approach. Intention to treat analysis shows that this approach does not lead to increased mortality.