AIM: To assess adherence with the the Society for Healthcare Epidemiology of America (SHEA)/ the Infectious Diseases Society of America (IDSA) guidelines for management of Clostridium difficile (C. difficile)-associated disease (CDAD) at a tertiary medical center.

METHODS: All positive C. difficile stool toxin assays in adults between May 2010 and May 2011 at the University of Maryland Medical Center were identified. CDAD episodes were classified as guideline adherent or non-adherent and these two groups were compared to determine demographic and clinical factors predictive of adherence. Logistic regression analysis was performed to assess the effect of multiple predictors on guideline adherence.

RESULTS: 320 positive C. difficile stool tests were identified in 290 patients. Stratified by disease severity criteria set forth by the SHEA/IDSA guidelines, 42.2% of cases were mild-moderate, 48.1% severe, and 9.7% severe-complicated. Full adherence with the guidelines was observed in only 43.4% of cases. Adherence was 65.9% for mild-moderate CDAD, which was significantly better than in severe cases (25.3%) or severe-complicated cases (35.5%) (P < 0.001). There was no difference in demographics, hospitalization, ICU exposure, recurrence or 30-d mortality between adherent and non-adherent groups. A multivariate model revealed significantly decreased adherence for severe or severe-complicated episodes (OR = 0.18, 95%CI: 0.11-0.30) and recurrent episodes (OR = 0.46, 95%CI: 0.23-0.95).

CONCLUSION: Overall adherence with the SHEA/IDSA guidelines for management of CDAD at a tertiary medical center was poor; this was most pronounced in severe, severe-complicated and recurrent cases. Educational interventions aimed at improving guideline adherence are warranted.

Clostridium difficile (C. difficile) is the major infectious cause of nosocomial diarrhea and can cause prolonged hospital stays, renal failure, toxic megacolon, and death[1-3]. In 1995, the Society for Healthcare Epidemiology of America (SHEA) published a clinical position paper on C. difficile-associated disease (CDAD)[4]. Based on data from small, randomized, controlled studies showing no outcome-difference when comparing metronidazole and vancomycin, the 1995 position paper considered them equally effective; however, it stated, “metronidazole may be preferred to reduce the risk of vancomycin resistance among other organisms in hospitals”.

Updated clinical practice guidelines for the management of CDAD in adults were published in 2010 by SHEA and the Infectious Diseases Society of America (IDSA)[5]. The 15-year interval between the two sets of recommendations was marked by dramatic changes in CDAD epidemiology and outcomes, with increases in prevalence, severity, and therapy resistance; emergence of hypervirulent strains may have contributed to these trends[6-8]. Additionally, new data suggested vancomycin might be superior for CDAD treatment in some cases. Zar et al[9] prospective, randomized, comparative efficacy study of metronidazole vs vancomycin demonstrated superiority of vancomycin for the treatment of severe CDAD. These results influenced the 2010 SHEA/IDSA guidelines that recommended vancomycin as first-line treatment for severe CDAD, while maintaining a recommendation for metronidazole in mild-moderate cases. These guidelines recommend treating an initial recurrence in the same manner as the initial episode, and a second recurrence with vancomycin in a tapering and/or pulsed regimen[5].

The 2010 SHEA /IDSA recommendations promote significant clinical practice changes. Since adherence to the guidelines may affect patient outcomes and infection control, we sought to determine adherence with the updated SHEA/IDSA CDAD guidelines at a tertiary care medical center.

The Institutional Review Board of the University of Maryland Baltimore approved this study and waived the requirement for informed consent. All positive C. difficile stool tests (Quick Check A/B Toxin Assay; Wampole Laboratories, Princeton, New Jersey) in adults between May 2010 and May 2011 at the University of Maryland Medical Center were retrospectively identified. Medical charts were reviewed for demographics, clinical information, and adherence to CDAD guidelines.

Classifications defined in the updated 2010 SHEA/IDSA guidelines were used. These guidelines define mild-moderate CDAD as the presence of a white blood cell count ≤ 15000/mm³ and a serum creatinine level ≤ 1.5 times the premorbid level. Conversely, severe CDAD is defined by the presence of a white blood cell count ≥ 15000/mm³ or a serum creatinine level ≥ 1.5 times the premorbid level. Severe-complicated CDAD is defined by the presence of hypotension, shock, ileus, or megacolon. According to the guidelines, the correct treatment for mild-moderate CDAD is metronidazole 500 mg orally three times per day for 10-14 d. For treatment of severe CDAD, recommended treatment is oral vancomycin 125 mg four times per day for 10-14 d. For severe-complicated CDAD, the recommended treatment is oral vancomycin 500 mg four times per day in addition to intravenous metronidazole 500 mg every eight hours. If complete ileus exists, then rectal administration of vancomycin should be considered. Patients with a first recurrence are recommended to receive the same treatment as per their initial episode. For a second recurrence, vancomycin in a tapered and/or pulsed regimen is recommended.

Specific data collected included age, gender, disease severity as defined by the 2010 SHEA/IDSA guidelines, location of treatment (stratified into outpatient, hospital ward or intensive care unit), non-CDAD antibiotic treatment during the month preceding diagnosis, presence of immunosuppression, if the episode was a recurrence, 30 d mortality, and agent selection and dosage of CDAD treatment. CDAD episodes were classified as guideline adherent if treatment provided was with the correct agent(s) at the correct dosage(s). If one of these parameters was not in accordance with the guidelines, then the treatment regimen was deemed non-adherent. Partial adherence was defined as the patient receiving the correct antibiotic, but at the wrong dose. Patients stratified into adherent and non-adherent groups were compared to determine demographics and clinical factors predictive of guideline adherence. Logistic regression analysis was performed to assess the effect of multiple predictors on guideline adherence (SAS, version 9.2).

About 320 positive C. difficile stool tests were identified in 290 patients (average age 57.6 years, 43.1% female). Of the cases, 95.9% were in hospitalized patients and 15.6% were identified as a recurrence. Stratified by disease severity criteria set forth by the SHEA/IDSA guidelines, 42.2% of cases were mild-moderate, 48.1% severe, and 9.7% severe-complicated. Most (80.6%) of the severe-complicated cases met this criterion due to hypotension or shock. Full adherence with the guidelines was observed in 43.4% of cases; 65.9% for mild-moderate, which was significantly better than in severe (25.3%) and severe-complicated cases (35.5%) (P < 0.001) (Figure 1). Of the severe CDAD cases, 55.3% were managed incorrectly with metronidazole. Partial adherence, where the correct drug was given at the incorrect dose, occurred in 17.8% of mild-moderate, 24.7% of severe, and 38.7% of severe-complicated cases (Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Rates of adherence with the 2010 the Society for Healthcare Epidemiology of America/the Infectious Diseases Society of America guidelines stratified by severity of Clostridium difficile-associated disease: (A) Mild-moderate, (B) Severe, and (C) Severe-complicated. Compliance was significantly better in mild-moderate vs severe or severe-complicated disease, P < 0.001.

On bivariate analysis (Table 1), factors significantly associated with adherence included disease severity, immunosuppression (IS), and documented receipt of antibiotics in the preceding 30 d. There was no difference in age, gender, hospitalization, ICU exposure, recurrence or 30-d mortality between adherent and non-adherent groups. IS patients were classified as mild-moderate more often than non-IS patients (60.0% vs 32.9%, P < 0.001). A multivariate model controlling for disease severity, prior antibiotics, IS, and recurrence status revealed significantly decreased adherence for severe/severe-complicated episodes (OR = 0.18, 95%CI: 0.11-0.30) and recurrent episodes (OR = 0.46, 95%CI: 0.23-0.95) but no significant difference for prior antibiotics or IS status.

Our results reveal poor overall adherence with the 2010 SHEA/IDSA guidelines for management of CDAD at a tertiary care academic medical center. Guideline adherence is worst in severe, severe-complicated, and recurrent CDAD. Our data suggests a lack of familiarity with current guidelines, as most providers continue to treat all initial episodes of CDAD with metronidazole, which was suggested as preferable by the 1995 SHEA clinical position paper on CDAD management. In fact, over half of our severe CDAD population, which should be treated with vancomycin, was incorrectly treated with metronidazole. This also explains the significantly improved adherence observed in mild-moderate patients whose treatment was not changed by the updated guidelines. We considered other possible causes of guideline non-adherence, such as the high cost of vancomycin and concern for vancomycin-resistance in other organisms, which has been shown to be significant in other nosocomial settings[10,11]. While the cost of branded oral vancomycin is approximately fifty-fold higher than oral metronidazole, our pharmacy routinely administers the generic intravenous formulation orally, which reduces the cost-difference dramatically[12], and makes cost concerns negligible. This finding also suggests an increased need for more intensive antibiotic stewardship, as not all incidences of non-adherence are likely due to knowledge. Antibiotic stewardship has been proposed as an effective method of increasing compliance at medical centers[13-15]. The exact impact of concerns over vancomycin resistance in other organisms on prescribing practices at our institution is unknown. We suspect this impact is small, as research on vancomycin use for CDAD has been conflicting with regards to rates of colonization and infection with resistant organisms[16-18].

Partial adherence with the guidelines, where the correct drug was chosen but an incorrect dosage was administered, occurred frequently as noted in Figure 1. The dosage of vancomycin chosen was often higher than recommended by the guidelines. While this is a form of non-adherence, it may be appropriate, as patients given the recommended dosage (125 mg four times daily) can have low fecal levels during the first day of treatment[19]. Additionally, one would expect a higher dose to be equally effective and, given poor systemic absorption, little difference in side effects.

Interestingly, we found clusters of cases in specific wards of the hospital, such as the Trauma Unit or the Cancer Center, suggesting nosocomial spread of the infection, which is a major source of morbidity, mortality, and cost from the disease[20-22]. This finding highlights the need for improved hand hygiene and infection control measures such as contact precautions.

Two patient populations expose limitations of applying the SHEA/IDSA guidelines: immunosuppressed patients and patients with end-stage renal disease. As noted in our results, immunosuppressed patients were significantly more likely to have their disease severity classified as mild-moderate. This may reflect an inability to mount a severe-defining white blood cell count. The frequent incidence of neutropenia among cancer patients in particular suggests a limitation of applying the guidelines to this group. Similarly, the end-stage renal disease population, in whom serum creatinine fluctuations cannot be used as indicators of disease severity, present a problem with applying the SHEA/IDSA guidelines.

The present work has limitations. Our study was retrospective, took place in a single tertiary medical center, and 95.9% of CDAD cases were in hospitalized patients. Therefore, our results may not be applicable to community hospitals or outpatients. Additionally, our study was not powered to detect differences in 30-d mortality. Our institution was also using a toxin assay for diagnosis at the time of data collection, which is less sensitive than PCR[23,24]. Furthermore, the study by Zar et al[9] that was used to develop the treatment guidelines has had questions raised with regards to its methodology[25,26]. Further studies that directly survey providers about their treatment practices and measure the effects of guideline adherence on mortality and morbidity factors such as length of stay and complications of CDAD are warranted.

In conclusion, our results suggest that many providers are unfamiliar with the 2010 SHEA/IDSA C. difficile guidelines. Educational interventions and antibiotic stewardship may prove beneficial to improve adherence, and potentially patient outcomes.