Clostridium difficile Infection Outcome Stratified by Severity: A Comparison of Three Indices

Clostridium difficile infection (CDI) is the leading cause of hospital-acquired infection. This colonic infection is responsible for approximately $5 billion dollars in excess healthcare costs per year in the U.S. The severity of CDI dictates both treatment options and prognosis. Depending on the severity of illness, patients can experience a spectrum of C. difficile-associated diseases including diarrhea, toxic megacolon, colon perforation, or death. Stratifying anti-CDI therapy based on severity has been shown to improve clinical outcomes. According to the Society for Healthcare Epidemiology in America and Infectious Diseases Society of America (SHEA/IDSA) and the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) guidelines, severe infections should be treated with vancomycin and non-severe treated with metronidazole. While several severity indices exist to differentiate between severe and non-severe CDI, it is unclear which index predicts the best clinical outcome. Moreover, there are significant discrepancies between severity indices and guidelines including treatment recommendations. This heterogeneity limits the practice of standardized, evidence-based medicine and may negatively impact patient care. This retrospective study compared SHEA/IDSA, ESCMID, and Zar’s severity indices to determine the index with the best prognostic value. While the SHEA/IDSA and ESCMID are standard guidelines against CDI, Zar’s index was University of Illinois Hospital & Health Sciences Systems’(UIHHSS) guideline against CDI during the study period. Each CDI case was classified as either severity index-concordant (SCT) or discordant (SDT) depending on the severity classification and treatment received according to the aforementioned severity indices. As the primary end point, the study assessed which index reduced the risk of poor outcome most significantly. Poor outcome was classified as recurrent infection, treatment failure, or 30-day mortality post CDI diagnosis. In addition, the effects of concordance on length of stay (LOS) post CDI diagnosis as well as differences in outcome between vancomycin and metronidazole were assessed. Out of the 229 CDI cases evaluated, 31% experienced poor outcome. Incidence of severe CDI ranged between 28% and 97% depending on the severity index used. After adjusting for variables, concordance to SHEA/IDSA, ESCMID, and Zar severity indices reduced the odds of poor outcome by 60%, 49%, and 66%. However, only SHEA/IDSA index produced a statistically significant reduction. Concordance to an index also reduced LOS post diagnosis and cost. After stratifying by CDI severity, there was no significant difference in the risk of poor outcome between metronidazole and vancomycin monotherapy groups. However, the incidence of poor outcome was higher in the metronidazole monotherapy group. The lack of consistency between guidelines and severity indices can complicate patient therapy and negatively affect clinical outcome. This is the first study to compare the three aforementioned severity indices. Although retrospective and limited in sample size, this study highlights the importance of following evidence-based medicine. Considering the simplicity of SHEA/IDSA’s criteria and profound reduction in outcomes, the study found the SHEA/IDSA index to have the best prognostic value.