SUMMARY. Journal of Clinical Pharmacy and Therapeutics, 2015, 40, doi: /jcpt.12329

Transcription

1 Journal of Clinical Pharmacy and Therapeutics, 2015, 40, doi: /jcpt Evaluation of a Clostridium difficile infection management policy with clinical pharmacy and medical microbiology involvement at a major Canadian teaching hospital S. S. T. Yeung* BSc, (Pharm), ACPR, MSc, J. K. Yeung,, BSc, (Pharm), ACPR, Pharm D, T. T. Y. Lau,, Pharm D, ACPR, FCSHP, L.A. Forresterk MA, MSc, T. S. Steiner,** MD, W. R. Bowie ** MD, FRCPC and E. A. Bryce, MD, FRCPC *BC Provincial Academic Detailing Service,Fraser Health Authority,Vancouver,BC,, Lower Mainland Pharmacy Services, Fraser Health Authority, Vancouver Coastal, Providence Health Care, Provincial Health Services Authority, Vancouver, BC, Pharmaceutical Sciences, Vancouver General Hospital (VGH), Vancouver, BC, Faculty of Pharmaceutical Sciences,University of British Columbia (UBC),Vancouver, BC, Division of Infectious Diseases, Department of Medicine, Faculty of Medicine, UBC, Vancouver, BC, kinfection Control, Vancouver Coastal Health (VCH), Powell River, BC, **Division of Infectious Diseases, VGH, Vancouver, BC, Division of Medical Microbiology and Infection Control, VCH, Vancouver, BC, and Department of Pathology & Laboratory Medicine, Faculty of Medicine, UBC, Vancouver, BC, Canada Received 20 July 2015, Accepted 28 September 2015 Keywords: antibiotics, antimicrobial stewardship, clinical pharmacy, clostridium difficile infection, evidence-based pharmacotherapy, guideline adherence, infection control, medical microbiology SUMMARY What is known and objective: Clostridium difficile infection (CDI) represents a spectrum of disease and is a significant concern for healthcare institutions. Our study objective was to assess whether implementation of a regional CDI management policy with Clinical Pharmacy and Medical Microbiology and Infection Control involvement would lead to an improvement in concordance in prescribing practices to an evidence-based CDI disease severity assessment and pharmacological treatment algorithm. Methods: Conducted at a tertiary care teaching hospital, this two-phase quality assurance study consisted of a baseline retrospective healthcare record review of patients with CDI prior to the implementation of a regional CDI management policy followed by a prospective evaluation post-implementation. Results and discussion: One hundred and forty-one CDI episodes in the pre-implementation group were compared to 283 episodes post-implementation. Overall treatment concordance to the CDI treatment algorithm was achieved in 48 of 141 cases (34%) pre-implementation compared with 136 of 283 cases (481%) post-implementation (P = 001). The median time to treatment with vancomycin was reduced from five days to one day (P < 001), with median length of hospital stay decreasing from 30 days to 21 days (P = 001) post-implementation. There was no difference in 30-day all-cause mortality. What is new and conclusion: A comprehensive approach with appropriate stakeholder involvement in the development of clinical pathways, education to healthcare workers and prospective audit with intervention and feedback can ensure patients diagnosed with CDI are optimally managed and prescribed the most appropriate therapy based on CDI disease severity. Correspondence: Dr J. K. Yeung, Lower Mainland Pharmacy Services, # Heather Pavilion, Level D, Vancouver General Hospital, Vancouver, BC V5Z1M9, Canada. Tel.: ; fax: ; Janice.Yeung@vch.ca WHAT IS KNOWN AND OBJECTIVE Clostridium difficile infection (CDI) presents as a spectrum of disease, ranging from asymptomatic carriage to severe complications such as pseudomembranous colitis, toxic megacolon, septic shock and death. 1 CDI is a significant concern for healthcare institutions as it is associated with increased morbidity, mortality, prolonged hospitalization and increased costs. 1 The Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA) clinical practice guidelines for CDI in adults recommend patients to be appropriately stratified and treated based on disease severity. 2 Metronidazole and vancomycin remain the mainstays of CDI treatment. 2 For mild-to-moderate CDI cases, metronidazole is the preferred agent, whereas vancomycin is the drug of choice in severe and fulminant cases. 2 Newer agents such as fidaxomicin have not yet gained widespread use, but may be considered in patients with first or second episodes of CDI who are at high risk of relapse and recurrence. However, its high cost prohibits routine use and its use for salvage cases is unfounded at this time due to lack of controlled clinical studies. 3 To ensure patients with CDI are optimally managed and treated, a regional CDI management policy was developed at our institution. This management policy is a comprehensive document that includes an evidence-based CDI disease severity assessment and pharmacological treatment algorithm, patient monitoring parameters and infection prevention and control procedures for managing patients diagnosed with CDI (Appendix S1). This policy was developed collaboratively by a number of departments within our institution, including Medical Microbiology and Infection Control, Clinical Pharmacy, Infectious Diseases, Gastroenterology, General Surgery, Critical Care, Leukaemia/Bone Marrow Transplant and Family Practice. 2 The intent of the CDI management policy is to resolve gaps in clinical practice, with the specific goals of ensuring early patient isolation, rapid treatment initiation based on appropriate CDI disease severity assessment and prompt identification of patients requiring escalation of treatment to vancomycin and/or surgical intervention. (Appendix S1). The CDI disease severity assessment 2015 John Wiley & Sons Ltd 655

2 and pharmacological treatment algorithm, which is based on the SHEA and IDSA clinical practice guidelines, provides guidance on risk-stratifying patients and initiating appropriate treatment with either metronidazole or vancomycin based on disease severity (Fig. 1). 2,4 The CDI management policy was disseminated institution-wide via print (i.e. newsletters and pocket reference cards) and made readily accessible through the institution s intranet (Appendix S2). To facilitate and encourage compliance among prescribers, Clinical Pharmacy and Medical Microbiology and Infection Control involvement was simultaneously instituted. When a CDI screening test is ordered for a patient, imbedded links within the institution s patient information systems program immediately prompt the prescriber to place the patient in contact isolation and to review the CDI management policy posted on the intranet (Appendices A2 and A3). Clinical Pharmacy and Medical Microbiology and Infection Control practitioners are notified of all positive C. difficile results, and clinical pharmacists are responsible for assessing all inpatient CDI treatment orders for adherence to the disease severity assessment and pharmacological treatment algorithm. If prescribed treatments are not concordant to the treatment algorithm, clinical pharmacists are required to intervene and contact prescribers to optimize CDI therapy. As a result of prompt implementation of contact precautions and appropriate treatment, it is expected that patients will be at reduced risk for acquiring nosocomial CDI, extended durations of CDI and associated symptoms, and mortality due to CDI. Our study objective was to assess whether the implementation of a regional CDI management policy with Clinical Pharmacy and Medical Microbiology and Infection Control involvement would lead to an improvement in concordance in prescribing practices to a CDI disease severity assessment and pharmacological treatment algorithm in patients diagnosed with CDI at a major Canadian teaching hospital. METHODS This was a two-phase quality assurance study comparing the concordance in prescribing practices to a CDI disease severity assessment and pharmacological treatment algorithm before and after the implementation of a CDI management policy with Clinical Pharmacy and Medical Microbiology and Infection Control involvement. Conducted at a 830-bed tertiary care teaching Fig. 1. Clostridium difficile infection (CDI) disease severity assessment and pharmacological treatment algorithm. # In patients unable to mount a blood cell count (WBC) response >15 000/mm 3, an increasing WBC with pronounced left shift may also be considered in these criteria; threshold of >15 000/mm 3 is based on expert opinion. ^May change to vancomycin if patient intolerant to metronidazole. * Doses of 125 to 500 mg may be considered; appropriate dose has not been established in clinical trials. Vancomycin IV is not effective for the treatment of CDI. Tapering regimens may vary considerably, as clinical data are limited. Note: Physician assessment for perforation risk is required prior to rectal tube placement. Prophylactic treatment for patients on antibiotics who have previously had C. difficile is not recommended. Consider Infectious Diseases consult. 656

3 hospital in Canada, this study consisted of (i) a baseline retrospective healthcare record review (1 April 2008 to 31 March 2009) of a random sample of patients with CDI prior to the implementation of a regional CDI management policy followed by (ii) a prospective evaluation of patients with CDI post-implementation. This post-implementation prospective review occurred over two periods, with the first being a pilot phase from 1 September 2009 to 31 March 2010 immediately after the initial implementation of the policy and the second from 1 March 2011 to 31 March 2012 when Clinical Pharmacy follow-up of CDI patients became a mandatory requirement. Prescribed treatments for patients diagnosed with CDI were assessed for their concordance to the treatment algorithm and compared between the pre-implementation and post-implementation study groups. Treatment concordance was defined as the patient being prescribed the correct drug, dose, route, frequency and duration based on CDI disease severity. CDI severity was assessed and determined by the patient s baseline complete blood count (CBC), electrolytes, serum creatinine (SCr) and specified clinical criteria (Fig. 1). Study objective The primary objective of this study was to assess whether the implementation of a regional CDI management policy with Clinical Pharmacy and Medical Microbiology and Infection Control involvement would lead to an improvement in concordance in prescribing practices to a disease severity assessment and pharmacological treatment algorithm in patients diagnosed with CDI. Secondary objectives were to compare the median time from C. difficile identification to initiation of pharmacological treatment, rates of CDI-related complications, length of hospital stay and allcause mortality at 30 days after CDI diagnosis, transfer or discharge (whichever occurred first) between the pre- and postimplementation study groups. Phase I - pre-implementation To assess the appropriateness of CDI treatment prior to the implementation of a CDI management policy, a retrospective healthcare record review of a randomly selected cohort of patients diagnosed with CDI (using a random number sequence generated in Microsoft Excel, Microsoft, Redmond, WA, USA) was conducted. Adult inpatients with diarrhoea and at least one positive test for C. difficile (by tcdb PCR) over a one-year period (1 April 2008 to 31 March 2009) were identified from our institution s Infection Control surveillance database and evaluated. Canadian Nosocomial Infection Surveillance Program (CNISP) definitions were used throughout the study, 2 and patients diagnosed with new or relapsed CDI were included. Patients were excluded if CDI was diagnosed as an outpatient, if patients were palliative, or if medical records were incomplete or unavailable to allow for assessment of CDI disease severity by the study investigators. Patient information collected included baseline characteristics and demographics, initial white blood cell count (WBC), SCr, albumin, vital signs at time of CDI diagnosis, time from stool specimen identification of C. difficile to initiation of metronidazole or vancomycin treatment and initially empirically prescribed pharmacological CDI treatment including drug, dose, dosing regimen, route and duration of therapy. The lead study investigator individually assessed each study patient to determine CDI severity and concordance of the prescribed treatment to the treatment algorithm. Previous and concurrent use of antibiotics, promotility and antiperistaltic agents, histamine-2 receptor antagonists (H2RA), and proton-pump inhibitors (PPI) were documented. Clinical patient outcomes including the occurrence of CDI-related complications, admission to the intensive care unit (ICU) for any indication, length of stay and all-cause mortality at 30 days after CDI diagnosis, transfer or discharge (whichever occurred first) were collected by Medical Microbiology and Infection Control. CDI-related complications captured included toxic megacolon, total or partial colectomy, gastrointestinal bleed, bowel perforation and secondary bacteremia. Probiotic use and compliance to specific infection control procedures and monitoring could not be reliably collected and were therefore not evaluated in this study. Just prior to the implementation of the CDI management policy, educational sessions to familiarize all clinical pharmacists with the policy and treatment algorithm were provided by the lead study investigator. Medical Microbiology and Infection Control reports identifying patients with PCR-positive CDI were generated daily to alert clinical pharmacists of all inpatient cases of CDI and a monitoring form (Appendix S4) was developed to assist pharmacists in their assessment and documentation. Clinical pharmacists were required to complete a form for each CDI case identified and resolve any noted deviations from the treatment algorithm with the medical team or prescriber. Phase II - post-implementation After completion of the retrospective review, adult inpatients with diarrhoea and at least one positive test for C. difficile (by tcdb PCR) were prospectively evaluated over a 7-month pilot phase from 1 September 2009 to 31 March 2010 and again over a 13-month period from 1 March 2011 to 31 March 2012 when follow-up of all inpatients diagnosed with CDI by clinical pharmacists became a mandatory requirement. Patient information and clinical outcomes collected, and exclusion criteria were identical to that of the pre-implementation phase. Information was gathered from patient healthcare records, the Infection Control surveillance database and the CDI monitoring forms completed by Clinical Pharmacy. The lead study investigator again individually assessed each study patient to determine CDI severity and concordance of the prescribed CDI treatment to the treatment algorithm and reconciled this with the information collected on the CDI monitoring forms completed by the clinical pharmacists. Data were entered in Microsoft Excel 2007, and statistical analyses were performed using SPSS (version 17, Windows, SPSS Inc., Chicago, IL, USA). Student t-tests and Fisher s exact tests were used to assess the differences between means and differences in proportions, respectively. Nonparametric tests were employed when assumptions for parametric tests were not met. P-values were corrected for multiple comparisons. This study was approved by the University of British Columbia Clinical Research Ethics Board and the Vancouver Coastal Health Research Institute. RESULTS AND DISCUSSION Between 1 April 2008 through to 31 March 2009, a total of 391 CDI episodes were identified at our institution through the Infection 657

4 Control database. Using a generated random numbers list, a random sample of 141 episodes (361%) was included in the pre-implementation group. During the post-implementation pilot phase (1 September 2009 and 31 March 2010), 140 CDI episodes were identified over a 7-month period. One hundred and seventeen of these episodes (836%) were included, whereas 23 episodes were excluded because patients were palliative or medical records were incomplete. When Clinical Pharmacy involvement became a mandatory requirement, a total of 266 CDI cases were identified over a 13- month period (1 March 2011 to 31 March 2012). Fifteen episodes were excluded because 12 cases were from another hospital and three were diagnosed prior to the study period, whereas another 85 episodes were missing clinical outcome information. Overall, a total of 283 episodes were included in the post-implementation group for analysis. Baseline characteristics including age, sex and concurrent antibiotic use are summarized in Table 1 and were similar between the pre- and post-implementation groups. More cases of severe CDI were noted in the post-implementation group (P < 001). Previous antibiotic use within 6 weeks prior to CDI diagnosis was documented in 239 of the 283 patients (844%) in the post-implementation group. Previous antibiotic use was unavailable for an unknown proportion of the remaining 44 patients, and therefore, the number of patients on previous antibiotics may be higher than the identified 844%. However, previous antibiotic use appears to be similar between groups. Concordance in prescribing to the treatment algorithm improved from 48 of 141 cases (34%) in the pre-implementation group to 136 of 283 cases (481%) post-implementation (P = 001). Concordance, as determined by each individual domain for appropriate drug, dose, dosing route, regimen and treatment duration, is summarized in Table 2. When each domain is independently assessed, the correct route of administration and recommended duration of treatment appeared to improve the most post-implementation. For patients treated with metronidazole, median time from C. difficile identification to initiation of treatment was one day for both groups. For patients prescribed vancomycin, median time to initiation of treatment from CDI confirmation significantly improved from five days in the pre-implementation group to one day in the post-implementation (P < 001). Table 1. Patient demographics Pre-implementation n = 141 Post-implementation n = 283 Pre- vs. post-implementation (P-value) Median age (range) 720 (17 98) 760 (16 98) 040 Mean age (SD) Male (%) 71 (504) 115 (406) 006 CDI severity Mild/moderate (%) 126 (894) 202 (714) <001* Severe (%) 10 (71) 70 (247) <001* Fulminant (%) 5 (35) 8 (28) 077 Previous antibiotics a (%) 120 (851) 239 (844) 100 Concurrent antibiotics (%) 87 (617) 166 (587) 060 H 2 RA b use (%) 22 (156) 53 (187) 050 PPI c use (%) 68 (482) 112 (396) 010 Antiperistaltic agent use (%) 11 (78) 17 (60) 053 Motility agent use (%) 38 (270) 63 (223) 033 *P = a Antibiotic treatment in the previous 6 weeks prior to CDI diagnosis. b Histamine-2 receptor antagonists. c Proton-pump inhibitors. Table 2. Comparison of Clostridium difficile infection treatment concordance and patient clinical outcomes Pre-implementation n = 141 Post-implementation n = 283 Pre- vs. post-implementation (P-value) Overall treatment concordance (%) 48 (340) 136 (481) 001 Correct drug (%) 118 (837) 233 (823) 079 Correct dose (%) 115 (816) 223 (788) 052 Correct route (%) 112 (794) 245 (866) 007 Correct dosing interval (%) 112 (794) 216 (763) 054 Correct duration (%) 58 (411) 146 (516) 005 Median time to treatment Metronidazole (days) 1 1 NSS Vancomycin (days) 5 1 <001 All-cause mortality at day 30 (%) 16 (113) 32 (113) NSS Median length of stay (range) 30 (1 210) 21 (1 299)

5 In the pre-implementation group, there were two cases of ICU admissions, three cases of toxic megacolon, three cases of total/ partial colectomy, seven cases of gastrointestinal bleed, one case of bowel perforation and three cases of secondary bacteremia. In the post-implementation group, there were four cases of ICU admission, three cases of toxic megacolon, five cases of total/partial colectomy, seven cases of gastrointestinal bleed, one case of bowel perforation and one case of secondary bacteremia. There was no statistical difference in CDI-related complications between the groups. All-cause mortality at day 30 after CDI diagnosis was similar between the pre- and post-implementation groups and not statistically significant. There was a statistically significant reduction in median length of hospital stay from 30 days in the preimplementation group compared to 21 days post-implementation (P = 001). DISCUSSION During the four years of this study ( ), the CDI rate remained relatively stable at our institution, averaging 114 per 10,000 patient-days in , 122 in and 114 in These rates were higher than the CDI rate of 697 per 10,000 patient-days reported by the CNISP for Western Canada during the study period. 5 This may be partly accounted for by the use of the more sensitive PCR molecular testing at our laboratory which has been reported in the literature to increase case detection by approximately 35% compared to standard enzyme immunoassay methods. 6 Our program prompted clinical pharmacists to intervene within 24 h and pharmacist recommendations made to optimize CDI therapies were generally well received and accepted by prescribers. Overall treatment concordance, defined as correct drug, dose, route, dosing regimen and duration, was 34% pre-implementation and statistically improved to 481% post-implementation of the CDI management policy. Incorrect treatment duration accounted for the majority of algorithm discordance. Although concordance rates improved only modestly, it is essential to note that rates are reflective of initial empiric regimens prescribed when all patient laboratory data may not have been available to assess CDI severity. At our institution, there is an automatic stop date of 7 days if a specific antibiotic treatment duration is not indicated and this likely contributed to the discordance in treatment duration. Our institution continues to work to improve concordance rates by prescriber educational sessions, the use of preprinted orders and periodical reminders to ensure appropriate CDI treatments are prescribed. It is however important to note that all CDI positive patients are assessed prospectively by clinical pharmacists, and all non-concordant cases are intervened upon to optimize their treatments. As a result, all patients at our institution are on appropriate CDI treatments within 48 hours of their CDI results. There appeared to be an increase in the number of patients with severe or fulminant CDI post-implementation. The epidemiology of CDI has been changing since the early 2000s where a more hypervirulent strain, North American PFGE type 1 or NAP1/BI/ 027 (NAP1), is causing more severe disease and relapses compared with other C. difficile strains. 2,7 Genotyping of CDI episodes were only selectively performed at our institution as part of a national surveillance study; therefore, we were unable to determine the prevalence of the NAP1 strain during the study period. The postimplementation pilot phase was not randomized and depended on the voluntary participation of clinical pharmacists. Pharmacists may have selectively monitored and followed patients with more severe disease during this time immediately post-implementation, when clinical pharmacy follow-up of all CDI patients was not yet mandatory. This may have contributed to the increased number of severe CDI cases observed post-implementation, but cannot be ascertained. Those patients requiring escalation to treatment with vancomycin were treated more promptly after implementation of the CDI management policy with Clinical Pharmacy involvement. Median time to treatment with vancomycin significantly improved from 5 days to 1 day post-implementation. This improvement in time to treatment with vancomycin in patients with more severe CDI may have contributed to the shorter median length of hospital stay observed in our study, but again cannot be ascertained. Our study did not demonstrate an effect on mortality, likely due to our limited sample size and short follow-up period. Other treatment options or the use of newer agents such as fidaxomicin were not evaluated in our study. Adherence to specific infection control procedures and monitoring were also not evaluated, and we are therefore unable to assess or quantify the impact these interventions may have contributed to patient outcomes. The number of cases of CDI reviewed in our study was obtained through random sampling of CDI patients during the study period and did not include all consecutive CDI patients. This random selection of patients is a limitation of the study, but we believe the sampling accurately reflects the changes associated with the increasing prevalence of the NAP-1 strain of CDI. Despite having more severe cases, our CDI-associated complication rates decreased and are in part due to the more rapid initiation of appropriate vancomycin treatment associated with our pharmacy and infection control CDI initiative. We acknowledge that the experiences of our program occurred from 2008 to 2012; however, we believe that our approach of reminding clinicians of infection control measures through electronic prompts and having clinical pharmacists assess all CDIpositive inpatients for treatment optimization is a novel and innovative initiative. This multidisciplinary strategy can be adopted by other institutions to enhance the management of CDI cases and improve patient care. WHAT IS NEW AND CONCLUSION The implementation of mandatory Clinical Pharmacy assessment of all CDI patients follows the concept that prospective audit and feedback is the most effective antimicrobial stewardship intervention in changing prescribing practices. 8 Our CDI management policy and infection control and antimicrobial stewardship initiatives align directly with the responsibility of pharmacists to optimize drug therapies in patients to improve health-related outcomes. As a result of the successes of our program, our CDI management policy, along with Clinical Pharmacy and Infection Control practitioner involvement, has been implemented at two additional hospital sites. Improved concordance to treatment recommendations has also been demonstrated with similar CDI stewardship initiatives. 9 A recent antimicrobial stewardship program recently implemented at a rural hospital for the management of CDI not only resulted in more appropriate antimicrobial prescribing, but also a reduction in costs. 10 Our study suggests that a comprehensive approach with appropriate stakeholder involvement in the development of clinical pathways, education to healthcare workers and prospec- 659

6 tive audit with intervention and feedback can improve the management of CDI and ensure patients receive the most appropriate therapy based on disease severity. With the implementation of a CDI management policy with mandatory Clinical Pharmacy follow-up and Medical Microbiology and Infection Control involvement, concordance in prescribing practices to an evidence-based disease severity assessment and pharmacological treatment algorithm increased significantly at our institution, with an observed reduction in the median time to initiation of treatment for those patients prescribed vancomycin and a decreased median length of hospital stay in those patients diagnosed with CDI. ACKNOWLEDGEMENTS We would like to thank Dr. Mandeep Saran and Ms. Jane Yu-Chen Lin, the clinical pharmacists and the Infection Control Practitioners at Vancouver General Hospital for their assistance in data collection; the CDI Stakeholder Group of Vancouver Coastal Health Providence Health Care for the development of the regional CDI Guideline; and Dr. Diane Roscoe for implementation of the automated messages linked with all inpatient CDI orders and results. CONFLICT OF INTEREST Dr. Janice Yeung has served on an advisory board for Pfizer. Dr. Theodore Steiner has served on the advisory board for Merck, Iroko Pharmaceuticals, and Cubist and has received research funding from Cubist, Sanofi-Pasteur, and Merck. Shirley Yeung and Drs. Tim Lau, Elizabeth Bryce, Leslie Forrester and William Bowie have no conflicts of interests to declare. SOURCE OF FUNDING None. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Appendix S1 CDI Management Policy. Appendix S2 Screen display of intranet site for CDI management policy. Appendix S3 Screen display of CDI toxin assay result. Appendix S4 CDI data collection form for clinical pharmacists. REFERENCES 1. Bartlett JG. Clostridium difficile: old and new observations. J Clin Gastroenterol, 2007;41:S24 S Cohen SH, Gerding DN, Johnson S, et al. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the society for healthcare epidemiology of America (SHEA) and the infectious diseases society of America (IDSA). Infect control Hosp Epidemiol (The official journal of the Society of Hospital Epidemiologists of America). 2010;31: Mullane K. Fidaxomicin in Clostridium difficile infection: latest evidence and clinical guidance. Ther Adv Chronic Dis, 2014;5: Surawicz CM, Brandt LJ, Binion DG, Ananthakrishnan AN, Curry SR, Gilligan PH et al. Guidelines for diagnosis, treatment, and prevention of Clostridium difficile infections. Am J Gastroenterol, 2013;108: ; quiz Clostridium difficile Associated Diseases (CDAD) Surveillance Ottawa: Public Health Agency of Canada; 2013 [updated 22 Nov 2013]. Available at: p. (accessed 20 July 2014). 6. de Jong E, de Jong AS, Bartels CJ, van der Rijt-van den Biggelaar C, Melchers WJ, Sturm PD. Clinical and laboratory evaluation of a real-time PCR for Clostridium difficile toxin A and B genes. Eur J Clin Microbiol Infect Dis, 2012;31: McDonald LC, Killgore GE, Thompson A, Owens RC Jr, Kazakova SV, Sambol SP et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med, 2005;353: Dellit TH, Owens RC, McGowan JE Jr, Gerding DN, Weinstein RA, Burke JP et al. Infectious diseases society of America and the society for healthcare epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis, 2007;44: Jury LA, Tomas M, Kundrapu S, Sitzlar B, Donskey CJ. A Clostridium difficile infection (CDI) stewardship initiative improves adherence to practice guidelines for management of CDI. Infect Control Hosp Epidemiol, 2013;34: Yam P, Fales D, Jemison J, Gillum M, Bernstein M. Implementation of an antimicrobial stewardship program in a rural hospital. Am J Health Syst Pharm, 2012;69: