Secondary oral vancomycin prophylaxis (OVP) has been used in adults with a history of Clostridioides difficile infection (CDI) while receiving systemic antibiotics to prevent CDI recurrence. However, this practice has not been studied in pediatric patients. The objective of this study was to assess the utility of secondary OVP in pediatric patients with previous CDI who received subsequent antibiotic exposure.
A multicampus, retrospective cohort evaluation was conducted among patients aged ≤18 years with any history of clinical CDI and receiving systemic antibiotics in a subsequent encounter from 2013–2019. Patients who received concomitant OVP with antibiotics were compared with unexposed patients. The primary outcome was CDI recurrence within 8 weeks after antibiotic exposure. Infection with vancomycin-resistant enterococci and risk factors for CDI recurrence were assessed.
A total of 148 patients were screened, of which 30 and 44 patients received OVP and no OVP, respectively. Patients who received OVP had greater antibiotic use and hospital lengths of stay. The incidence of CDI recurrence within 8 weeks of antibiotic exposure was significantly lower in patients who received OVP (3% vs 25%; P = .02) despite this group having notably more risk factors for recurrence. There were no vancomycin-resistant enterococci infections in any patients within either group. After adjustment in a multivariable analysis, secondary OVP was associated with less risk of recurrence (odds ratio, 0.10; 95% confidence interval, 0.01–0.86; P = .04).
Secondary OVP while receiving systemic antibiotics reduces the risk of recurrent CDI in pediatric patients with a history of CDI.
Evidence supporting oral vancomycin as secondary prophylaxis against Clostridioides difficile infection (CDI) is limited to retrospective studies in adult cohorts. The role of secondary prophylaxis in the pediatric patient population is unknown.
This retrospective study is the first to assess oral vancomycin prophylaxis in pediatric patients with previous CDI and subsequent antibiotic exposure. Secondary prophylaxis was found to be protective against CDI recurrence when children were receiving systemic antibiotics.
Clostridioides difficile infection (CDI) is among the most common infections encountered in the United States. Rates of relapse and recurrence are estimated at 25% and can cause significant morbidity and mortality.1 Oral vancomycin is often prescribed as secondary prophylaxis in adults as a means to reduce the risk of CDI recurrence during concomitant antibiotic therapy. However, this recommendation is not officially endorsed by the Infectious Diseases Society of America, and additional research is needed.2 In retrospective studies, it has been suggested that there may be a benefit to administering secondary oral vancomycin prophylaxis (OVP) in adult patients with previous CDI.1,3–5
Pediatric patients have been largely excluded from previous studies. As a result, the impact of secondary OVP in children with a history of CDI is not well understood. Approximately 30% to 40% of infants are colonized with C difficile and are largely asymptomatic.6,7 However, after the first year of life, rates of asymptomatic colonization decrease, and the prevalence of colonization parallels that of adult patients, with a rate of ∼3% to 8%.8 Children with previous C difficile are at a much higher risk for subsequent infections compared with uninfected control patients, and up to 30% of C difficile patients experience recurrence.9 In the pediatric oncology population, approximately one-third of children are colonized with C difficile, and a large proportion experience prolonged shedding.10 Epidemiological studies have revealed that the incidence of CDI in children has increased from 2.6 cases per 1000 admissions to 4.6 cases through the years and is associated with increased mortality, longer lengths of stay, and higher costs.11–14 Despite these concerns, current CDI treatment guidelines do not comment on the role of secondary OVP in pediatric patients receiving concomitant antibiotics.2,15
The use of OVP is not without consequences. Oral vancomycin has been associated with profound effects on the gastrointestinal microbiota, including colonization with vancomycin-resistant enterococci (VRE) as well as subsequent, more rapid recolonization with C difficile, after oral vancomycin withdrawal.16,17
At New York University Langone Health (NYULH), guidelines for CDI management established in October 2016 recommend the use of secondary OVP for any pediatric patient with a history of CDI while receiving systemic antibiotics. However, the ultimate decision to institute OVP is determined on a case-by-case basis after accounting for patient-host factors, CDI risk associated with present antibiotics, and time since last CDI occurrence. The heterogeneous practice and lack of formal guidance warrant additional research into the utility of secondary OVP as an option for pediatric patients with a history of CDI while receiving concomitant antibiotics.
This was a retrospective cohort study performed within the NYULH system, which comprises the Manhattan campus (Tisch Hospital, Kimmel Pavilion, and Hassenfeld Children’s Hospital), NYU Langone Hospital – Brooklyn, NYU Langone Orthopedic Hospital, and various ambulatory sites. Data collection occurred for patients evaluated at any of these locations between January 2013 through December 2019. This study was approved by the NYULH Institutional Review Board.
All pediatric patients aged ≤18 years with a positive diagnostic test result for C difficile by the Xpert C difficile polymerase chain reaction (PCR)–based assay (Xpert CD assay; Cepheid, Sunnyvale, CA) during the aforementioned time period were identified by using a laboratory report generated from the electronic health record. Patients were then screened to identify those who (1) met clinical criteria for CDI during their initial episode, (2) required an inpatient or outpatient visit after their initial CDI episode (this was considered the study encounter), and (3) were prescribed at least 24 hours of systemic antibiotics during the study encounter. After this, patients were classified into the group that received OVP or the group that did not on the basis of whether OVP was prescribed along with systemic antibiotics during the study encounter. If multiple encounters were eligible for secondary OVP, only the first encounter during the study period was included.
Patients were excluded if they (1) were <1 year of age, (2) were not treated for the initial CDI episode, (3) received active CDI treatment rather than OVP during the study encounter, or (4) died of non-CDI causes during the follow-up period.
Patient characteristics, medication histories, and potential confounders were collected from the electronic health record as follows: demographic data; descriptors of previous CDI episodes, including dates of occurrences; isolation of North American pulsed-field gel electrophoresis type 1 (NAP1) strains; previous hospitalizations; antibiotic use by type, timing, and duration; hospital lengths of stay; concomitant proton pump inhibitors (PPIs), histamine-2 receptor antagonist (H2RA), and probiotic use; and OVP details, including prescribing locations and treatment durations.
CDI was defined as documentation of at least 3 unformed stools per day with or without radiographic or histopathologic findings revealing colitis and a positive test result for C difficile via the Xpert C difficile PCR-based assay in a stool specimen. In accordance with institutional guidelines, testing was discouraged in those <12 months of age because of high rates of asymptomatic colonization and only conducted in those with clinical symptoms of CDI. Testing algorithms did not change over the study period.
Active CDI treatment was determined on the basis of a confirmed diagnosis of CDI and use of enteral vancomycin doses of 10 mg/kg (up to 125 mg per dose for nonsevere CDI and 500 mg per dose for severe or fulminant CDI) every 6 hours and/or intravenous or enteral metronidazole doses of 7.5 mg/kg (up to 500 mg per dose) every 6 to 8 hours. Severe CDI was classified by elevations in white blood cell counts or serum creatinine, as stated by the Infectious Diseases Society of America guidelines, or the presence of any of the following: toxic megacolon, perforation, colectomy, hypotension, or sepsis.2
OVP was defined as enteral vancomycin doses of 10 mg/kg (up to 125 mg per dose) every 12 hours during concomitant antibiotic use. OVP duration was intended to continue while on systemic antimicrobial agents and for 5 days after completion of antimicrobial agents (extended prophylaxis tail), but practice varied, and duration was ultimately left to the discretion of the provider. Antibiotic risk was classified as high, medium, or low on the basis of previously published data.18–20 High-risk antibiotics included clindamycin, fluoroquinolones, third- and fourth-generation cephalosporins, carbapenems, and β-lactam β-lactamase inhibitors (BLBLIs); medium-risk antibiotics included macrolides, sulfamethoxazole-trimethoprim, penicillins, aminopenicillins, and first-generation cephalosporins; and low-risk antibiotics included tetracyclines, aminoglycosides, and vancomycin.18–20
The primary end point was the rate of CDI recurrence within 8 weeks after initiation of systemic antibiotics. Patients were considered to have recurrence if they had any history of CDI (as defined previously) and developed CDI requiring treatment within 8 weeks after initiation of antibiotics during the study encounter. If patients in either group developed CDI recurrence, they were considered to have met the primary end point and subsequently censored. Secondary end points included severity of CDI recurrence, time to CDI recurrence, and VRE infection from any site within 8 weeks after initiation of OVP. Univariable and multivariable regression analyses were conducted to identify risk factors for CDI recurrence.
Initial univariable comparisons between recipients of OVP and nonrecipients were conducted by using χ2 or Fisher’s exact test for categorical variables and the Mann–Whitney U test for continuous variables. Two-tailed P ≤ .05 denoted statistical significance. In addition, patients who developed and did not develop CDI recurrence were compared. Univariable and multivariable logistic regression were used to determine risk factors that were associated with CDI recurrence. All variables with a univariable P ≤ .20 were considered for inclusion in the multivariable model. The Hosmer–Lemeshow goodness-of-fit test was used to test the power of the model. All analyses were performed by using SPSS, version 25 (IBM SPSS Statistics, IBM Corporation).
A total of 148 patients with an initial CDI episode were screened for inclusion, as shown in Fig 1. Of these, a total of 74 patients were excluded, mostly because of the lack of subsequent encounter with systemic antibiotic use after the initial CDI episode. After exclusion, 74 patients remained, of which 30 patients received OVP and 44 patients did not receive OVP.
Patient baseline characteristics, comorbidities, previous CDI episodes, hospitalization descriptions, previous antibiotic use, antibiotic use during the study encounter, and concomitant medication use are reported in Table 1. The groups were similar in age and comorbidities, but more patients who received OVP were male. Malignancy and immunosuppressed states were common, comprising nearly half of both cohorts. Most patients had experienced only 1 CDI episode before antibiotic exposure, with the majority of previous episodes occurring within the previous 6 months. Hospital length of stay was significantly longer in patients who received OVP (median 13 days vs 6.5 days; P = .009). Previous antibiotic use within 3 months of the study encounter was common in both groups; however, patients who received OVP received more fluoroquinolones and carbapenems. Of note, only 1 patient received fidaxomicin for CDI treatment within 3 months of the study encounter, but this patient did not receive OVP. Concomitant metronidazole or PPI use was infrequent, whereas H2RA and probiotic use were more common and similarly distributed between cohorts.
During the study encounter, the majority of patients in both groups received high-risk antibiotics, with third- and fourth-generation cephalosporins given most often. In terms of overall antibiotic exposure, patients who received OVP received more classes of antibiotics (median 2 classes versus 1.5 classes; P = .004), longer duration of antibiotics (median 12 vs 8 days; P = .004), and longer duration of high-risk antibiotics (median 11.5 vs 5.5 days; P = .003) compared with patients who did not receive OVP. Use of low-risk antibiotics was also significantly higher in OVP recipients (47% vs 21%; P = .02), with systemic vancomycin as the primary antibiotic prescribed.
OVP was provided for a median of 12 days for the 30 patients who received OVP, and the time to OVP initiation after starting antibiotics was a median of 1 day. Patients were mostly admitted in the inpatient setting at the time of OVP initiation (28 of 30, 93%), whereas OVP completion occurred more commonly in the outpatient setting (18 of 30, 60%). Ten patients (33%) received a short tail of OVP after completion of antibiotics, for a median of 5 days.
Outcomes are summarized in Table 2. In patients who did not receive OVP, 11 patients (25%) experienced CDI recurrence within 8 weeks after antibiotic exposure compared with only 1 patient (3%) in the group that received OVP (P = .02). The majority of recurrences after antibiotic exposure were classified as nonsevere and occurred at a median of 19 days in patients who did not receive OVP. There were no VRE infections from any site within 8 weeks after antibiotic exposure in either group. In patients who received OVP, the only recurrence occurred in the inpatient setting after 6 days of antibiotic therapy. At onset, the patient experienced multiple watery stools accompanied with worsening abdominal pain, which prompted PCR testing that ultimately resulted as positive. There was an initial delay of 1 day between antibiotic and OVP initiation, and all OVP doses were administered before recurrence. No laxatives or bowel regimens were present at the time of or preceding the diagnosis. The patient had several risk factors for recurrence, including malignancy, recent hospitalization, previous receipt of third-generation cephalosporins, and receipt of a third-generation cephalosporin while on OVP.
Univariable and multivariable analyses were conducted to assess for predictors of CDI recurrence after antibiotic exposure (Table 3). On univariable analysis, receipt of OVP was associated with reduced risk of CDI recurrence (P = .02), but no other risk factors predicted recurrence. In the multivariable logistic regression model, after controlling for recent hospitalization, duration of antibiotics, and previous high-risk antibiotic use, receipt of OVP remained protective against CDI recurrence (odds ratio, 0.10; 95% confidence interval, 0.01–0.86; P = .04).
Findings from this retrospective observational study suggest that OVP is effective in preventing CDI recurrence in pediatric patients, and ∼5 patients would need secondary OVP to prevent 1 recurrence. Likewise, OVP was not associated with any VRE infections. The majority of our patients had comorbid conditions, particularly malignancy, and considerable previous and concurrent antibiotic exposure, thus representing a high-risk population for CDI acquisition and recurrence.13,21 Notably, patients who received OVP were at much higher risk for recurrence, as compared with those who did not, for several reasons. Patients receiving OVP had heavier exposures and longer courses of systemic broad-spectrum antibiotics. Additionally, patients who received OVP had longer hospital stays, representing a sicker population at baseline with an increased risk for nosocomial acquisition of C difficile. Despite the presence of these risk factors, the potentially beneficial effect of OVP in preventing CDI recurrence was still able to be revealed.
To our knowledge, this is the first analysis of CDI prophylactic strategies in the pediatric population. A potential benefit of secondary OVP in reducing the risk of CDI recurrence has been suggested in published retrospective studies in adult patients.1,3–5 Van Hise et al4 found that the incidence of CDI recurrence in their population was significantly lower in OVP recipients compared with nonrecipients (4.2% vs 26.6%; P < .001). Similar results were noted in subsequent retrospective studies in adults, although OVP dosing strategies and time frames for evaluating CDI recurrence were variable.1,3,5 The recurrence rate of 25% noted in our study’s control group is consistent with the reported incidence of CDI recurrence ranging from 10% to 40% in pediatric patients.21,22 The recurrence rate in patients who received OVP was low but consistent with rates reported in aforementioned literature.4,5 Our study population largely supports the role of OVP in patients with a single previous CDI episode, with previous literature revealing conflicted findings for the benefit of OVP relative to the number of previous CDI episodes.1,3,5
In contrast to previous studies, our inclusion criteria did not restrict patients to a specific time frame with regard to initial CDI occurrence and subsequent antibiotic exposure. This reflects current practice at our institution because providers may still consider OVP in pediatric patients with remote CDI episodes, if other risk factors for CDI recurrence are present. Regardless, most patients in our study had their latest CDI episodes within 6 months before the study encounter, indicating recent rather than remote CDI. The follow-up period in this study was limited to 8 weeks after systemic antibiotic exposure to account for the period of greatest risk for CDI recurrence, as previously described in both adults23 and pediatrics.21,22 A relatively low dose and frequency of OVP was used as consistent with our institutional guidelines (10 mg/kg twice daily up to 125 mg per dose) and previous literature revealing effectiveness of this dose of OVP in preventing recurrence.4 Although our guidelines recommend a short extension of OVP for an additional 5 days after antibiotic cessation, only 10 patients received an extended regimen, and none experienced recurrence. The ability to assess for recurrence on the basis of cessation of OVP on the last day of systemic antibiotics compared with 5 days afterward was limited by the presence of only 1 recurrence in recipients of OVP.
Use of concomitant medications, such as metronidazole, PPIs, H2RAs, or probiotics, were not considered exclusion criteria in this study. Inclusion of patients receiving concomitant medications with potential influence on CDI recurrence risk was designed to reflect real-world scenarios in which OVP is considered. The number of patients receiving metronidazole was low, so underestimation of CDI recurrence rates is unlikely. Probiotics have been found to reduce the incidence of CDI when used as prophylaxis on the basis of small trials and a meta-analysis.24,25 However, none of these concomitant medications were found to be significant predictors of CDI recurrence on univariable analysis.
There are several limitations to our study. First, the study was retrospective in nature, and investigators had access to only electronic health records within the health system. Therefore, the possibility of patients experiencing CDI recurrences or receiving antibiotics outside of the network cannot be excluded. Additionally, we could not exclude the possibility of asymptomatic colonization of C difficile, along with diarrhea due to another cause. Decisions related to diagnostic testing and antimicrobial management were made at the discretion of the patients’ providers, and data collection for the study was dependent on available documentation. In addition, more than half of the cohort completed OVP as outpatients, so missed doses were not entirely captured. The small sample size may have limited the ability to detect a difference in safety outcomes, such as VRE infection. Of note, VRE colonization could not be assessed in the study cohort because routine VRE surveillance is not performed at our institution. Moreover, fidaxomicin has been revealed to reduce the risk of CDI recurrence in children and adolescents.26 However, only 1 patient received this treatment of a previous CDI episode, so we were unable to assess its effect on recurrence rates in our cohort. Lastly, NAP1 strains of C difficile are associated with increased recurrence. However, relatively few patients were tested for or had NAP1 strains isolated, thereby limiting the ability to assess this as a risk factor for recurrence.27
OVP was clinically efficacious and well tolerated for pediatric patients at high risk for CDI recurrence by using recommended practices at our institution. Findings from this study should be validated by conducting randomized controlled trials.
We thank Anna Stachel, PhD, for her assistance with identification of potential study patients and statistical analysis.
Dr Bao collected data, conducted the initial and final analyses, and drafted the initial manuscript; Dr Lighter established and wrote the guideline and ensured the practice in the study; Dr Jen conceptualized and designed the study and designed the data collection instruments; Drs Dubrovskaya and Merchan designed the study, designed the data collection instruments, and assisted with statistical analyses; Drs Siegfried and Papadopoulos designed the study and the data collection instruments; and all authors reviewed and revised the manuscript and approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Deidentified individual participant data will not be made available.
FUNDING: No external funding.
β-lactam β-lactamase inhibitor
Clostridioides difficile infection
histamine-2 receptor antagonist
North American pulsed-field gel electrophoresis type 1
New York University Langone Health
oral vancomycin prophylaxis
polymerase chain reaction
proton pump inhibitor
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.