Age-Stratified Treatment Response Rates in Hospitalized Patients with Clostridium difficile Infection Treated with Metronidazole

DISCUSSION

Clostridium difficile is the most common cause of infectious diarrhea in hospitalized patients and the most common cause of death due to gastroenteritis in the United States (2). Metronidazole is commonly used for the treatment of CDI and is guideline recommended for an initial episode of mild-to-moderate CDI (3, 11). Observational and clinical trials have observed a decreased clinical cure rate with patients treated with metronidazole; however, the reasons for this decreased response rate are unclear (4–6, 8). In this study, continued use of antibiotics, severe CDI, and a high Horn's index were predictive of patients at higher risk for refractory CDI when given metronidazole. The strengths of the study include a large sample size, the participation of multiple centers, and strict definitions for CDI and refractory CDI that allowed comparisons from previous studies. These results would indicate that risk factors such as continued use of antibiotics, severity of CDI, and underlying illnesses are more predictive than advanced age in determining who may fail metronidazole therapy. These results could be used to help determine who may be eligible for CDI therapies other than metronidazole.

Advanced age is a major risk factor for CDI and poor outcomes of CDI, including recurrent CDI and death (12, 13). Although advanced age as a risk factor for acquisition of CDI has been demonstrated, the reasons why advanced age increases the risk of CDI and poor outcomes of CDI are relatively unknown. We previously demonstrated that a high Horn's index, i.e., a severity score of underlying acute and chronic illnesses in patients with CDI, was associated with increased hospital costs and length of stay (10). This study builds upon these previous findings by demonstrating that the severity of underlying illness as measured by Horn's index is also associated with increased risk of metronidazole treatment failure. Continued use of antibiotics creates a thriving environment for spore germination by eliminating the gut microflora, resulting in the persistence of CDI despite adequate treatment. Higher rates of CDI treatment failures have been demonstrated in the past in patients given concomitant antibiotics (14), treated with fidaxomicin or vancomycin (15). As many hospitalized patients receive unnecessary antibiotics for an extended duration, antimicrobial stewardship efforts to decrease unnecessary concomitant use of non-CDI antibiotics may improve treatment outcomes of patients with CDI (16, 17). The reduction in these concomitant antibiotics may also help justify the use of more-expensive, but more-effective, agents for the treatment of CDI. Taken together, these results suggest that although advanced age is a good surrogate marker for underlying severity of illness and other age-related host factors, a more appropriate determination of risk factors for CDI treatment failures may help develop modifiable interventions or help identify a unique high-risk population. With the mounting evidence of the inferiority of metronidazole in this regard, it is also likely that metronidazole should be reserved as a second-line agent or used only for very mild disease. Changing treatment patterns in patients with CDI will likely take active involvement by antimicrobial specialists or stewardship teams. Our group has previously demonstrated that the use of oral vancomycin, even in guideline-recommended situations, is relatively low without active intervention (18). While the use of oral metronidazole conflicts with guideline-based recommendations for treatment of severe CDI, this treatment pattern reflects real-world practice at many institutions.

This study has many limitations. In our study, we designated refractory CDI as continued signs and symptoms of CDI after 7 days of anti-CDI treatment. We did this in order to compare our results to those of previous studies evaluating refractory CDI. However, many patients experience significant sequelae of CDI beyond this time period, including recurrent CDI and postinfectious irritable bowel syndrome (19–21). The long-term sequelae of refractory CDI will require further study. This was an observational retrospective study, and not all information was available for a small number of patients. However, we used consistent definitions from past retrospective studies and were able to retrieve these data for the majority of patients. Also, all study hospitals were located in Houston, TX, and therefore treatment standards and patient characteristics may not reflect practices around the United States or in other countries. We chose to measure the composition of comorbid conditions (Charlson index) and underlying disease severity (Horn's index). Assessment of specific comorbidities and refractory CDI will require further study. Lastly, we were unable to identify or assess for strains of C. difficile in this study.

In conclusion, hospitalized elderly patients with Clostridium difficile infection treated with metronidazole had increased refractory CDI rates compared to younger patients. This was likely due to increased underlying severity of illness, severity of CDI, and increased rates of concomitant antibiotic use. These results may help identify patients that may benefit from alternative C. difficile treatments other than metronidazole.