A Multicenter Evaluation of Probiotic Use for the Primary Prevention of Clostridioides difficile Infection

Abstract

Background Primary prevention of C. difficile infection (CDI) is a priority for hospitals and probiotics have the potential to interfere with colonization and infection with C. difficile. This study evaluated the impact of a computerized clinical decision support tool (CCDS) to prescribe probiotics for primary prevention of CDI among adult hospitalized patients. Methods A CCDS tool was implemented into the electronic medical record at four hospitals prompting prescription of a probiotic preparation at the time of antibiotic prescription in high-risk patients in May 2019. Interrupted time series using segmented regression analysis was conducted to evaluate hospital-wide CDI incidence for the year pre- and post-CCDS implementation. In addition, multivariable logistic regression was used to evaluate CDI incidence in patients qualifying for probiotics in the pre- versus post-intervention periods adjusting for potential confounders. To adjust for potential differences in patients who received probiotics in the post-intervention period, propensity score matched pairs were developed to evaluate CDI risk by receipt of probiotics. Results Quarterly CDI incidence increased over time post-intervention relative to baseline trends (slope change 1.4, 95% CI 0.9-1.9). The odds ratio (OR) of CDI was 1.41 in eligible patients post-intervention compared to pre-intervention (adjusted OR 1.41, 95% CI 1.11, 1.79). Propensity score matched analysis showed that patients who received probiotics did not have lower rates of CDI compared to those who did not receive probiotics (OR 1.46, 95% CI 0.87, 2.45). Conclusions Use of probiotics for primary prevention of CDI among adult inpatients receiving antibiotics is not supported.

Influence of Probiotics on the Development of Clostridioides difficile Infection in Patients Receiving Fluoroquinolones

“…Allen et al failed to show a statistically significant benefit of probiotics in the prevention of antibioticassociated diarrhea (AAD) or CDI [18]. A recent multi-center trial by Heil et al evaluated the impact of a computerized clinical decision support tool to prescribe probiotics to highrisk patients for primary prevention of CDI [19]. Over the 13-month post intervention period, 2489 patients (16.6%) received probiotics and no difference in CDI was observed compared to the pre-intervention period.…”

Why 2 Studies That Used the Same Probiotic May Have Come Up With Different Outcomes

“…To the Editor—Two recent studies tested the same 3-strain Lactobacillaceae probiotic mixture for the prevention of Clostridium difficile infection (CDI), yet reported conflicting results [ 1 , 2 ]. In studies that test probiotics for CDI prevention, different efficacies are often reported, but typically it is due to not accounting for the strain-specific efficacy of probiotics [ 3 ].…”

“…A review of 28 QES found that limitations included low implementation of the probiotic and not accounting for changes in infection control measures or antibiotic use during the 2 time periods [ 4 ]. With these limitations in mind, we examined these 2 studies to attempt to determine why the same probiotic was effective in the Maziade et al study [ 1 ] but not in the Heil et al study [ 2 ].…”

“…Differences in the 2 studies ( Table 1 ) show different trends in CDI rates during the control period, increasing in one study [ 1 ] and decreasing in the other [ 2 ]. During the intervention period, CDI rates significantly decreased in one study [ 1 ] but increased in the other study [ 2 ].…”

“…Differences in the 2 studies ( Table 1 ) show different trends in CDI rates during the control period, increasing in one study [ 1 ] and decreasing in the other [ 2 ]. During the intervention period, CDI rates significantly decreased in one study [ 1 ] but increased in the other study [ 2 ]. Similar results were seen in the patient-level data, and adjustment for CDI risk factors resulted in nonsignificant differences the in Heil et al study [ 2 ] but significant efficacy remained in the Maziade et al study [ 1 ].…”