The Comparative Effectiveness of Ciprofloxacin and Metronidazole for an Initial Episode of Pouchitis: A Propensity Matched Study

Abstract

Objectives:

There is minimal evidence regarding the comparative effectiveness of individual antibiotics in the treatment of pouchitis. We sought to evaluate the comparative effectiveness of ciprofloxacin monotherapy, metronidazole monotherapy, and combination therapy (ciprofloxacin and metronidazole) in the treatment of an initial episode of pouchitis after ileal pouch-anal anastomosis (IPAA) for ulcerative colitis (UC).

Methods:

We performed a retrospective cohort study in TriNetX, a global federated research network of electronic health records. Primary outcomes were failure of initial antibiotic therapy and the development of recurrent pouchitis in the first 12 months after an initial episode of pouchitis. One-to-one propensity score matching (PSM) was performed for age, sex, race, primary sclerosing cholangitis, nicotine dependence, obesity, and previous exposure to tumor necrosis factor inhibitors between the cohorts.

Results:

Among 271 patients who developed pouchitis (mean age at IPAA 35.8 years, male sex 57%) and were treated with ciprofloxacin, metronidazole, or combination therapy, 190 (70%) developed recurrent pouchitis. After PSM, there was no significant difference in the odds of early relapse or non-response with ciprofloxacin compared to metronidazole monotherapy (aOR 0.56, 95% CI 0.23–1.34) or when either monotherapy was compared to combination therapy. There was also no significant difference in odds of recurrent pouchitis when comparing patients treated with ciprofloxacin to metronidazole monotherapy (adjusted OR [aOR] 0.86, 95% CI 0.40–1.84) or either monotherapy to combination therapy.

Conclusions:

In this retrospective cohort study, we demonstrated no significant difference in the real-world effectiveness of ciprofloxacin, metronidazole, or combination therapy for the initial episode of pouchitis.

INTRODUCTION

Pouchitis is the most common complication affecting patients after restorative proctocolectomy with ileal pouch-anal anastomosis (IPAA).^1–3 Our understanding of the relative effectiveness of individual therapies for the treatment of pouchitis and particularly the comparative effectiveness of these therapies is limited. Although an estimated 48% of patients will develop pouchitis within the first 2 years after IPAA,⁴ the evidence to support treatment decisions regarding initial antibiotic therapy choices is largely based on small randomized controlled trials (RCTs)⁵ or empiric choices.

When examining the available literature regarding the treatment of acute pouchitis, perhaps the most significant head-to-head comparative efficacy study to date was an RCT where 7 patients received ciprofloxacin and 9 patients received metronidazole.⁶ Other studies of effectiveness have also been limited by sample size including evaluations of metronidazole vs. budesonide enemas,⁷ as well as other antibiotics and probiotics compared to placebo.^{8, 9} Although antibiotics are often well tolerated in the recommended short initial courses, an improved understanding of the real-world comparative effectiveness of antibiotic choices is necessary. An estimated 19% of patients experience recurrent pouchitis in the short-term following an initial episode of pouchitis, which has been associated with greater costs and healthcare utilization.^{4, 10} Effective therapy that treats symptoms of pouchitis, and perhaps decreases the recurrence of pouchitis and subsequent progression to chronic pouchitis, remains a key objective in the treatment of patients with an IPAA.

Real-world data sources such as the TriNetX global federated research network offer the opportunity to analyze large sample sizes that are indicative of current care patterns in the United States. To improve our understanding of the comparative effectiveness of ciprofloxacin and metronidazole for an initial episode of pouchitis after restorative proctocolectomy with IPAA for UC, we performed a retrospective cohort study using TriNetX. In these analyses, we had two primary outcomes: 1) early relapse or non-response in the evaluation of the initial antibiotic therapy chosen and 2) the development of recurrent pouchitis in the first 12 months after an index pouchitis episode. Recurrent pouchitis was specifically chosen as an outcome given the emphasis on effective therapy for an initial episode of pouchitis, given that patients with pouchitis in the first year after IPAA have demonstrated an increased risk for chronic inflammatory conditions of the pouch.^{11, 12}

METHODS

Data Source

A retrospective cohort study utilizing the multi-institutional research network TriNetX (Cambridge, MA, USA) was performed. TriNetX is a global federated research network which provides real-time access to de-identified electronic health records (EHRs) of more than 85 million patients within 56 health care organizations (HCOs) in the United States. The de-identification process is determined and done at a network-level and attested to through a formal determination by a qualified expert as defined in the HIPAA Privacy Rule. TriNetX obfuscates patient counts <11 to ensure patient anonymity. Clinical variables are derived directly from EHRs of included HCOs as well as retrieved through a built-in natural language processing system that extracts variables from clinical documents. Robust quality assurance is achieved at the time of extraction from EHRs before inclusion in the database, in a systemic and standardized format. The interface only provides aggregate counts and statistical summaries to protect patient health information and ensures that the data remain de-identified at all levels of data retrieval and dissemination.

Study Participants and Cohorts

A real-time search and analysis of the US Collaborative Network in the TriNetX platform was conducted from January 1, 2007 to March 6, 2023. The ulcerative colitis (UC) cohort was identified using International Classification of Disease, Tenth Revision, Clinical Modification (ICD-10-CM) codes for UC (K51*) plus RxNorm codes for at least one UC-related medication which included mesalamine, sulfasalazine, balsalazide, olsalazine, azathioprine, 6-mercaptopurine, methotrexate, infliximab, adalimumab, golimumab, vedolizumab, ustekinumab and tofacitinib. Complex case definitions for identification of an IBD cohort which include ≥ 1 ICD-10-CM code plus a relevant IBD-related prescription have been shown to have ≥ 80% positive predictive value and ≥ 85% specificity.¹³ The TriNetX database has been used previously in prior published studies in patients with IBD.^{14, 15} Patients undergoing proctocolectomy with IPAA were identified using Current Procedural Terminology (CPT) codes (Supplementary Table 1). The final stage of surgery was defined as a CPT code for an IPAA or CPT code for an ileostomy takedown following CPT code for total abdominal colectomy or total proctocolectomy (with or without IPAA). We then identified patients with pouchitis using a previously validated case-finding definition within administrative claims data.^{4, 16} Patients were required to have a RxNorm code for ciprofloxacin and/or metronidazole at least 30 days after undergoing IPAA surgery. A 30-day washout period was used to avoid including patients who may have received these antibiotics during the perioperative period. Given that the accuracy of determining the antibiotic regimen for the index episode of pouchitis was critical, we also required patients to have an ICD-10-CM code for pouchitis (K91.850). The antibiotic regimen was determined by the RxNorm code for ciprofloxacin and/or metronidazole which occurred within 2 weeks of pouchitis ICD-10-CM code. TriNetX allows establishing temporality between different groups within a cohort which enabled us to determine the index antibiotic regimen. The index antibiotic regimen was divided into patients who received ciprofloxacin only, metronidazole only and combination therapy with ciprofloxacin and metronidazole. A two-week period was used given the recommended duration of antibiotic treatment for acute pouchitis.¹⁷ Patients who were switched to a different antibiotic within two weeks of the index regimen for pouchitis were excluded from the study as this would indicate either an allergic reaction or intolerance due to a side effect.

Outcomes

The primary objective of the study was to assess the comparative effectiveness of different antibiotic regimens used for the index episode of pouchitis. Failure of the initial antibiotic regimen was defined as a composite outcome of patients who had developed early relapse or non-response. It was not possible to distinguish between patients who had non-response and early relapse as these granular clinical variables cannot be extracted from the database. Patients who had RxNorm codes for ciprofloxacin, metronidazole, rifaximin, amoxicillin and/or amoxicillin-clavulanate after 2 weeks but within 4 weeks of the index antibiotic regimen were deemed to have met the outcome of failure of the initial antibiotic regimen.

Recurrent pouchitis was defined by RxNorm codes for ciprofloxacin, metronidazole, tinidazole, rifaximin, amoxicillin and/or amoxicillin-clavulanate after 4 weeks but within 12 months of index pouchitis episode. Patients who required antibiotics within 4 weeks of index antibiotic regimen were excluded as this would represent early relapse or non-response. Of note, in the prior validation of the case-finding definitions for pouchitis in administrative claims, excluding competing common diagnoses that might require antibiotics such as pneumonia or urinary tract infection did not improve the sensitivity or specificity of the definitions for pouchitis.¹⁶ Secondary outcomes included time to development of recurrent pouchitis.

Statistical analysis

All statistical analyses were conducted using the TriNetX software using the browser-based real-time analytics feature, TriNetx Live (TriNetX LLC, Cambridge, MA). Baseline characteristics of cohorts were described using means, standard deviations (SDs), and proportions. Co-variates based on demographics, co-morbid diseases, laboratory parameters and historical IBD medication use were identified. One-to-one (1:1) propensity score matching was performed to balance the following co-variates between groups: age, sex, race, primary sclerosing cholangitis, nicotine dependence, obesity, and previous exposure to tumor necrosis factor inhibitors. These co-variates were chosen as their presence or absence has been associated with an increased risk of pouchitis.^{4, 18, 19} TriNetX platform utilizes input matrices of the user-identified covariates to conduct logistic regression analysis to obtain propensity scores for all individual subjects. The propensity scores generated are used to match patients using greedy nearest-neighbor algorithms with a caliper width of 0.1 pooled standard deviations. TriNetX randomizes the order of rows to eliminate bias resulting from nearest-neighbor algorithms. After propensity-score matching, the risk of each outcome was calculated and expressed as adjusted odds ratios (aOR) with 95% confidence intervals (CIs). Two-sided p-values <0.05 were considered statistically significant. Kaplan-Meier curves for antibiotic-free survival were generated with log-rank tests after propensity score matching. Hazard ratios (HRs) and associated CIs, together with the test for proportionality were calculated using R’s survival package. The numbers are validated by comparing them with output from SAS version 9.4 (SAS Institute, Cary, NC, USA).

RESULTS

We identified 271 patients who had undergone restorative proctocolectomy with IPAA for UC and subsequently developed an initial episode of pouchitis that was treated with ciprofloxacin monotherapy, metronidazole monotherapy, or a combination of ciprofloxacin and metronidazole. Among the entire cohort, the mean age at the time of IPAA was 35.8 years (SD 14.5 years) and 57% of patients were of male sex (Table 1) with all four US census regions represented (Northeast 38%, Midwest 23%, South 27% and West 13%).

Table 1.

Demographics and clinical characteristics of the patients developing pouchitis in TriNetX

	Patients with Pouchitis N=271
	mean	Standard Deviation
Age of pouch surgery	35.8	14.5
	n	%
Male sex	149	57
Race and/or ethnicity
White	222	85
African American	18	7
Hispanic or Latino	24	9
Co-morbid diseases
Primary sclerosing cholangitis	13	5
Hypertension	40	15
Hyperlipidemia	24	9
Diabetes mellitus	23	9
Ischemic heart disease	15	6
Chronic lower respiratory diseases	33	12
Obesity	30	11
Nicotine dependence	15	6
Alcohol use disorder	12	4
UC-related medications #
Infliximab	113	42
Adalimumab	65	24
Vedolizumab	51	19
Ustekinumab	12	4
Tofacitinib	19	7
Azathioprine	84	31
Mercaptopurine	45	17
Methotrexate	25	9

Ulcerative colitis (UC)

^#Patient could have been on more than one medication

When evaluating the specific antibiotic regimens utilized, 77 (28%) patients utilized ciprofloxacin alone, 84 (31%) metronidazole alone, and 110 (41%) a combination of approach. In comparing the overall demographics and clinical characteristics of the three treatment approaches utilized, patients utilizing ciprofloxacin monotherapy were more likely to be of female sex when compared to patients utilizing metronidazole monotherapy (53% vs. 35%, P=0.01) and were more likely to use an anti-TNF therapy prior to colectomy (82% vs. 54%, P<0.001), Table 2. Patients utilizing combination antibiotics were also more likely to use an anti-TNF therapy prior to colectomy than those using metronidazole monotherapy (89% vs. 58%, P<0.001). There were no other significant differences noted prior to propensity score matching.

Table 2.

Covariates included in the propensity score matching between two pouchitis cohorts based on type of antibiotic regimen used for index pouchitis

	Before propensity score matching			After propensity score matching
	Ciprofloxacin (n = 77)	Metronidazole (n = 84)	P value	Ciprofloxacin (n = 51)	Metronidazole (n = 51)	P value
	(mean +/− SD)	(mean +/− SD)		(mean +/− SD)	(mean +/− SD)
Age	36.6 +/− 15.5	39.5 +/− 15.2	0.23	38.1 +/− 16.8	37.7 +/− 16.7	0.90
	n (%)	n (%)		n (%)	n (%)
Female sex	41 (53.2%)	29 (34.5%)	0.01	24 (47%)	23 (45%)	0.84
White	62 (80.5%)	71 (84.5%)	0.50	42 (82.3%)	42 (82.3%)	1
Primary sclerosing cholangitis	10 (12.9%)	10 (11.9%)	0.83	10 (19.6%)	10 (19.6%)	1
Nicotine dependence	10 (12.9%)	10 (11.9%)	0.83	10 (19.6%)	10 (19.6%)	1
Obesity	17 (22%)	16 (19%)	0.63	12 (23.5%)	10 (19.6%)	0.63
TNFi	63 (82%)	45 (53.5%)	0.0001	32 (62.7%)	33 (64.7%)	0.79
	Ciprofloxacin mono (n = 77)	Combination (n = 110)	P value	Ciprofloxacin mono (n = 67)	Combination (n = 67)	P value
	(mean +/− SD)	(mean +/− SD)		(mean +/− SD)	(mean +/− SD)
Age (mean +/− SD)	36.6 +/− 15.5	37.4 +/− 13.9	0.72	36.8 +/− 15.9	35.1 +/− 14.6	0.53
	n (%)	n (%)		n (%)	n (%)
Female sex	41 (53.2%)	50 (45.4%)	0.29	31 (46.2%)	34 (50.7%)	0.60
White	62 (80.5%)	92 (83.6%)	0.58	55 (82%)	55 (82%)	1
Primary sclerosing cholangitis	10 (12.9%)	10 (9%)	0.39	10 (14.9%)	10 (14.9%)	1
Nicotine dependence	10 (12.9%)	18 (16.3%)	0.52	10 (14.9%)	10 (14.9%)	1
Obesity	17 (22%)	25 (22.7%)	0.91	10 (14.9%)	10 (14.9%)	1
TNFi	63 (82%)	75 (89.2%)	0.19	47 (92.1%)	48 (94.1%)	0.61
	Metronidazole mono (n = 84)	Combination (n = 110)	P value	Metronidazole mono (n = 51)	Combination (n = 51)	P value
Age (mean +/− SD)	39.5 +/− 15.2	37.4 +/− 13.9	0.30	38.5 +/− 14.5	39.1 +/− 14	0.80
	n (%)	n (%)		n (%)	n (%)
Female sex	29 (34.5%)	50 (45.4%)	0.12	28 (43%)	24 (36.9%)	0.47
White	71 (84.5%)	92 (83.6%)	0.86	55 (84.6%)	55 (84.6%)	1
Primary sclerosing cholangitis	10 (11.9%)	10 (9%)	0.52	10 (15.3%)	10 (15.3%)	1
Nicotine dependence	10 (11.9%)	18 (16.3%)	0.38	10 (15.3%)	10 (15.3%)	1
Obesity	16 (19%)	25 (22.7%)	0.53	15 (23%)	12 (18.4%)	0.51
TNFi	45 (58.4%)	75 (89.2%)	<0.001	39 (76.4%)	35 (68.6%)	0.38

TNF inhibitor (TNFi)

In the assessment of risk of early relapse or non-response to the initial antibiotic therapy used in the treatment of pouchitis, there were no significant differences in risk of early relapse when comparing patients treated with ciprofloxacin monotherapy compared to metronidazole monotherapy (aOR 0.56, 95% CI 0.23–1.34, Table 3). Similarly, there were no significant differences when comparing either monotherapy strategy to combination therapy.

Table 3.

Risk of early relapse or non-response based on the antibiotic regimen for the index pouchitis episode before and after propensity score matching

	Before propensity score matching		After propensity score matching
	N (%)	OR (95% CI)	N (%)	aOR (95% CI)
Ciprofloxacin only	14 (18.1%)	0.55 (0.26–1.17)	11 (18.6%)	0.56 (0.23–1.34)
Metronidazole only	24 (28.5%)		17 (28.8%)
Ciprofloxacin only	14 (18.1%)	0.62 (0.30–1.27)	13 (17.3%)	0.77 (0.34–1.74)
Combination	29 (26.3%)		16 (21.3%)
Metronidazole only	24 (28.5%)	1.11 (0.59–2.11)	17 (23.6%)	0.80 (0.38–1.70)
Combination	29 (26.3%)		20 (27.7%)

Among all patients that developed pouchitis, 190 patients (70.1%) developed recurrent pouchitis. Median time to recurrent pouchitis was 72 days (IQR 29–359). After propensity score matching, there was no significant difference in the odds of developing recurrent pouchitis when comparing patients treated with ciprofloxacin monotherapy to those patients treated with metronidazole monotherapy (adjusted OR [aOR] 0.86, 95% CI 0.40–1.84, Table 4). There was also no significant difference in the odds of developing recurrent pouchitis when comparing ciprofloxacin monotherapy to combination therapy with ciprofloxacin and metronidazole (aOR 0.70, 95% CI 0.32–1.47) and when comparing metronidazole monotherapy to combination therapy (aOR 0.84, 95% CI 0.38–1.87).

Table 4.

Risk of recurrent pouchitis based on the antibiotic regimen for the index pouchitis episode before and after propensity score matching

	Before propensity score matching		After propensity score matching
	N (%)	OR (95% CI)	N (%)	aOR (95% CI)
Ciprofloxacin monotherapy	51 (66.2%)	0.83 (0.42–1.61)	38 (64.4%)	0.86 (0.40–1.84)
Metronidazole monotherapy	59 (70.2%)		40 (67.7%)
Ciprofloxacin monotherapy	51 (66.2%)	0.73 (0.39–1.38)	43 (66.1%)	0.70 (0.32–1.47)
Combination	80 (72.7%)		48 (73.8%)
Metronidazole monotherapy	59 (70.2%)	0.84 (0.47–1.65)	42 (70%)	0.84 (0.38–1.87)
Combination	80 (72.7%)		44 (73.3%)

When comparing the time to development of recurrent pouchitis, there were also no differences when comparing the three treatment strategies (ciprofloxacin monotherapy vs. metronidazole monotherapy, P=0.79; ciprofloxacin monotherapy vs. combination therapy, p=0.48; metronidazole monotherapy vs. combination therapy, P=0.92, Figure 1). In Cox Proportional Hazards Models, there were also no significant differences when comparing the 3 strategies: ciprofloxacin monotherapy vs. metronidazole monotherapy aHR 0.85 (95% CI, 0.55–1.35), ciprofloxacin monotherapy vs. combination therapy aHR 0.86 (95% CI, 0.57–1.30), metronidazole monotherapy vs. combination therapy aHR 1.02 (95% CI, 0.67–1.55).

Figure 1.

Comparison of time to an episode of recurrent pouchitis among patients treated with ciprofloxacin or metronidazole for an initial episode of pouchitis after ileal pouch-anal anastomosis for ulcerative colitis

Legend: A: Ciprofloxacin monotherapy vs. Metronidazole monotherapy; B: Ciprofloxacin monotherapy vs. Ciprofloxacin and Metronidazole combination therapy; C: Metronidazole monotherapy vs. Ciprofloxacin and Metronidazole combination therapy

DISCUSSION

In this large, geographically representative cohort we performed a study of the comparative effectiveness of the most common approaches to the treatment of an initial episode of pouchitis. Although antibiotics are often well-tolerated by most patients, particularly in the short courses required for the treatment of acute pouchitis, evaluating the real-world effectiveness of these approaches has important implications. We designed our study to assess primary non-response or failure of an antibiotic therapy for the initial episode of pouchitis. However, by also assessing the rate of recurrent pouchitis as a primary outcome, we have assessed an outcome that reflects quality of life and symptom burdens for individual patients as well as burdens to the healthcare system including costs and healthcare resource utilization. Additionally, if early recurrent pouchitis leads to chronic inflammatory conditions of the pouch, the effectiveness of an initial therapy for pouchitis is paramount. Perhaps reassuringly, we have demonstrated that although both are effective, neither of these approaches seem superior, and thus patient preference and tolerance may in fact most properly inform initial treatment decisions when choosing between these regimens.

The traditional recommendations for the use of ciprofloxacin and metronidazole in the treatment of pouchitis have been founded in either observational studies or RCT that were performed in the early decades of IPAA surgery.^{6, 7, 20–23} In a RCT of 16 patients receiving ciprofloxacin or metronidazole for the treatment of acute pouchitis, both ciprofloxacin and metronidazole were associated with a significant reduction in the total PDAI score as well as symptom, endoscopic and histologic subscores of the PDAI.⁶ When comparing the 2 therapies directly, ciprofloxacin was associated with a significantly greater decrease in total PDAI score, as well as symptom and endoscopic subscores. Additionally, 3 patients (33%) treated with metronidazole experienced adverse effects attributed to antibiotic therapy. Our findings bolster this RCT that has led to the recommendation that ciprofloxacin or metronidazole should be the preferred antibiotic therapy for an initial course of pouchitis,^{17, 24} providing a larger sample size for comparison and demonstrating similar efficacy overall when comparing ciprofloxacin monotherapy, metronidazole monotherapy, and a combination approach using both antibiotics. Although there was no significant difference when comparing the risk of early relapse among patients treated with ciprofloxacin monotherapy or metronidazole monotherapy, the relative difference in aOR and accompanying 95% CI warrants consideration of underlying type II error. We did not perform a formal sample size calculation given the use of real-world data, and large prospective studies would be necessary to further evaluate this potential. Additionally, given the nature of administrative claims data, we could not directly assess patient tolerability, however given the established nature of each of these approaches in the treatment of pouchitis it is unlikely we would have identified new patient symptoms in the short initial treatments of pouchitis that are recommended.

Antibiotics remain the foundation of initial treatment for acute pouchitis. This initial treatment approach is based in a clinical assumption that the pathogenesis of pouchitis is presumed to involve abnormal immune responses to the commensal microbiota in those individuals with a genetic predisposition for inflammatory conditions of the pouch.^{25, 26} The contribution of the microbiota to pouch inflammation is further supported by the corresponding response to antibiotic therapy as well as additional methods of microbiota manipulation such as the use of probiotic therapy^27–30 in the treatment of pouchitis. Several authors have demonstrated a correlation of certain microbial groups with disease activity markers of inflammation,^{26, 31} and the development of pouch inflammation is rare prior to the restoration of continuity. For each of these reasons, understanding the most effective antibiotics (or alternative approaches) in the treatment of an initial episode of pouchitis remains a critical goal in the management of patients after an IPAA for UC.

In this study, our primary objective was to evaluate the comparative effectiveness of ciprofloxacin and metronidazole in the treatment of acute pouchitis. However, given the understanding of the impact of the microbiota on pouch inflammation, understanding the potential for other antibiotics to treat pouchitis also remains a goal for future evaluations. Other antibiotics have been used in this setting, including rifaximin, which has been studied vs. placebo in a RCT.⁸ Additionally, other antibiotics that have demonstrated effectiveness in the treatment of chronic antibiotic dependent pouchitis such as vancomycin³² may be even more effective when utilized earlier in the treatment algorithm.

The high rate of recurrent pouchitis in our study population may reflect a referral bias in our study population, and we attempted to address underlying confounding using propensity score matching. The majority of patients in each treatment regimen received advanced therapies prior to colectomy (over 72% of patients in each regimen) and thus we were unable to evaluate the refractory state of UC as a potential high-risk state and driver of recurrent pouchitis. This remains a potential area of future research. Given that 70.1% of patients developed recurrent pouchitis, another important consideration in the treatment of an initial episode of pouchitis is whether secondary prophylaxis or prevention strategies should also be considered in this population. Although this is beyond the scope of this manuscript, future studies should continue to focus on the potential of secondary prevention strategies. Previously, multiple randomized controlled trials have assessed the potential of probiotic formulations in the prevention of relapse of pouchitis in patients with prior antibiotic-responsive pouchitis.^33–35 When considering the potential for progression to recurrent pouchitis (and thus in theory the potential need for chronic antibiotic therapy), consideration must be given to such methods of secondary prevention in addition to the potential effectiveness of primary therapy strategies.

Our study has several strengths including a well-defined cohort using validated case-finding definitions and propensity score matching to reduce the impact of confounding variables despite the retrospective nature of the study. Our study also has limitations. Our primary objective was to study the comparative effectiveness of these therapies; however, given the nature of claims data, we can only demonstrate prescription fills and cannot confirm adherence to an antibiotic regimen or evaluate adverse events related to antibiotic therapy. We did exclude patients who had an RxNorm code for another antibiotic within the 2-week period of the initial antibiotic regimen to minimize this effect. However, the inability to distinguish non-response from early relapse or adverse events directly related to antibiotic therapy is one of clinical importance, especially given prior small RCTs demonstrating the potential for improved tolerance of ciprofloxacin compared to metronidazole in this setting.⁶ There may be unmeasured confounding present, despite the use of propensity score matching, and we cannot assess for unmeasured preferences that would lead a gastroenterologist or patient to choose a specific antibiotic regimen. There is also a lack of granularity to evaluate specific activity indices such as the pouchitis disease activity index.³⁶

In conclusion, in a large, well-defined cohort of patients with acute pouchitis we demonstrated that both ciprofloxacin and metronidazole were effective for the treatment of acute pouchitis, with no significant difference in the clinical effectiveness of monotherapy with either agent, or a combination therapy approach in evaluating the rate of recurrent pouchitis. These data should be reassuring to patients and their treating providers regarding the initial choices for the treatment of acute pouchitis but may also prompt future evaluations of alternative antibiotic therapies that may be as effective or more effective in the treatment of acute pouchitis. Comparative evaluations and a continued evaluation of the true effectiveness of many of our accepted therapeutic approaches in the treatment of pouch-related disorders are needed, as confirming effectiveness or promoting alternative strategies both have the potential to improve outcomes among patients after IPAA for UC.

STUDY HIGHLIGHTS

WHAT IS KNOWN

• Ciprofloxacin and metronidazole are the most commonly recommended treatments for an initial episode of pouchitis

• There is a paucity of data evaluating the comparative effectiveness of these regimens

WHAT IS NEW HERE

• There were no differences when comparing the effectiveness of ciprofloxacin monotherapy, metronidazole monotherapy or combination therapy with ciprofloxacin and metronidazole in the treatment of initial episode of pouchitis

• Effectiveness of initial antibiotics regimens were evaluated by the odds of initial treatment failure, developing recurrent pouchitis, and time to recurrent pouchitis

• These findings suggest that therapy with either antibiotic may be appropriate and that patient preference or tolerance may drive initial antibiotic choice for the treatment of pouchitis