Proposed Checklist of Hospital Interventions to Decrease the Incidence of Healthcare‐Associated Clostridium difficile Infection
Background. The incidence and severity of Clostridium difficile infection (CDI) are increasing, and previously described interventions for controlling the spread of CDI are not easily generalized to multiple healthcare institutions.
Objective. We tested prevention and treatment bundles to decrease the incidence of CDI and the mortality associated with CDI at our hospital.
Design. Observational before‐after study of adult patients admitted to a tertiary care, university‐affiliated hospital during the period from January 2004 through December 2008.
Methods. In January 2006, we launched an educational campaign and introduced a prevention bundle—a series of specific processes aimed at preventing the transmission of C. difficile among hospitalized patients, including enhanced isolation practices, laboratory notification procedures, and steps coordinating infection control and environmental services activities. In April 2006, we implemented a treatment bundle—a set of hospital‐wide treatment practices aimed at minimizing the risk of serious CDI complications. We tracked quarterly incidence rates and case‐fatality rates for healthcare‐associated CDI cases at our hospital. Our main outcome was the healthcare‐associated CDI incidence rate, measured as the number of healthcare‐associated cases of CDI per 1,000 patient‐days.
Results. We followed patients for a total of 1,047,849 patient‐days. The healthcare‐associated CDI incidence rates fell from an average of 1.10 cases per 1,000 patient‐days before intervention to 0.66 cases per 1,000 patient‐days after intervention. This statistically significant decrease amounts to a 40% reduction in incidence after the intervention.
Conclusions. Our intervention was successful in reducing the incidence of CDI at our hospital. On the basis of our experience, we propose the use of a checklist of hospital interventions to decrease the incidence of healthcare‐associated CDI.
Received April 20, 2009; accepted June 10, 2009; electronically published September 14, 2009.
The incidence and severity of Clostridium difficile infection (CDI) are increasing.1,2 The number of hospitalized patients in the United States for whom CDI was listed as a discharge diagnosis more than doubled during the period from 2000 through 2005 (ie, from 134,361 to 291,303 patients),3 and CDI‐associated mortality rates are estimated to have more than quadrupled during the period from 1999 through 2004 (ie, from 5.7 to 23.7 cases per million population in the United States).4 This increasing severity has been linked to the emergence of a new strain of C. difficile that has been observed in the United States and in an increasing number of other countries.5‐7 There is substantial financial burden associated with this changing epidemiology.8 A recent study9 estimates CDI‐related hospital costs to be as much as $6,326 per hospitalization. Because of the increasing incidence, severity, and costs associated with CDI, there is a substantial need for efforts to prevent exposure to and transmission of CDI, as well as to improve the treatment for patients who received a diagnosis of CDI.
Although antimicrobial use is viewed as the primary risk factor for developing CDI10 (with targeted antimicrobial use preferred),11 widespread infection prevention efforts are also necessary for reducing the spread of CDI in a healthcare setting. Recently, the Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America published recommendations for preventing CDIs in hospitals.12 These guidelines described recommended strategies for acute care hospitals, including using contact precautions; cleaning and disinfecting equipment and the environment; implementing laboratory‐based alert systems to notify infection prevention and control personnel about new diagnoses of CDI; conducting CDI surveillance; analyzing and reporting CDI data; educating healthcare personnel, housekeeping personnel, hospital administrators, as well as patients and their families about CDI; and measuring compliance with the hand‐hygiene and contact‐precaution recommendations of the Centers for Disease Control and Prevention or the World Health Organization. In addition, special approaches were described that can be considered by hospitals when the transmission of C. difficile continues despite implementation of basic practices. These special approaches include the use of hypochlorite‐based environmental cleaning agents for cleaning of rooms once occupied by patients with CDI, the preferential use of soap and water (rather than alcohol‐based hand disinfectant) for routine hand hygiene after contact with patients known or suspected to have CDI, and the initiation of antimicrobial stewardship programs.
These interventions are complex and involve changes in practices and policies that require difficult behavioral changes on the part of healthcare workers at multiple points in care. The guidelines did not specify methods for successful implementation. For this reason, they have proven difficult to implement for the entire hospital population.11,13,14 For quality improvement, the grouping of practices in bundles and in checklists has proven to be a highly successful strategy in multiple medical domains.15‐17 The principle is to identify for teams of healthcare workers the critical actions that must be performed daily, assigning specific responsibilities to individual personnel and ensuring that the information contained in the guidelines is translated into behavioral changes.
In late 2005, routine infection control surveillance for CDI at our hospital revealed an upward trend in CDI rates, and an additional investigation demonstrated an increase in the number of patients with severe infection necessitating colectomy or intensive care as well as an increase in the number of deaths attributable to CDI complications. Therefore, in January of 2006, we initiated an intervention that involved the implementation of an educational campaign, a prevention bundle, and a treatment bundle all aimed at decreasing the incidence and severity of CDI at our hospital.
Methods
Baseline Condition
Brigham and Women’s Hospital is a 750‐bed tertiary care, university‐affiliated teaching hospital in Boston, Massachusetts. Prior to the intervention, its hospital‐wide infection control policies for patients with diarrhea due to CDI were consistent with the Centers for Disease Control and Prevention recommendations and included the use of contact precautions (ie, single‐patient rooms, alcohol‐based hand hygiene, and the donning of gowns and gloves prior to patient contact). Positive C. difficile toxin test results were communicated to patient‐care staff by the hospital’s microbiology laboratory staff, and the threshold for suspecting and initiating workup for CDI was variable, depending on the judgment of individual clinicians. Surveillance for all incident cases of healthcare‐associated CDI was performed routinely by the hospital’s infection preventionists. A case of healthcare‐associated CDI was defined as a hospitalized patient whose first positive C. difficile toxin test result was at least 3 days after hospital admission or within 4 weeks of a previous discharge from the hospital, unless there were intervening hospitalizations at other healthcare facilities. Healthcare‐associated CDI rates were calculated using a denominator of patient‐days, excluding patient‐days attributed to newborns in the neonatal intensive care unit. At baseline, there was no set of treatment guidelines implemented, no mechanism to confirm adherence to prevention guidelines, and no specific expectation for soap‐based hand washing or for cleaning of rooms with a hypochlorite‐based disinfectant.
Intervention
Our intervention included 3 components: an educational campaign, a prevention bundle, and a treatment bundle. By use of this intervention, we aimed to increase prevention adherence and add enhanced infection prevention practices—including washing hands with soap and water after patient contact and cleaning patient rooms with a hypochlorite‐based disinfectant—for all patients with clinically suspected CDI, in addition to all patients with laboratory‐confirmed CDI. We also implemented basic treatment guidelines for the first time. The prevention and treatment bundles assigned specific responsibilities to individual healthcare personnel.
Educational Campaign
When introducing our prevention bundle, we initiated an educational outreach campaign designed to teach nurses, physicians, physician assistants, environmental services personnel, and hospital leaders about the increasing incidence and severity of CDI in our hospital and to encourage them to increase their level of suspicion for this diagnosis and to promptly initiate appropriate diagnostic testing, isolation precautions, and treatment. In addition, we emphasized the importance of consistent adherence to hand hygiene and isolation precautions when caring for patients with presumed or confirmed CDI.
Prevention Bundle
We designed a CDI prevention bundle that specified individual responsibilities for physicians, physician assistants, nurse practitioners, floor nurses, microbiology staff, infection control practitioners, and environmental services personnel. This prevention bundle included specific infection control practices, laboratory notification procedures, and steps to be taken in coordinating infection control and environmental services that aimed to decrease the transmission of C. difficile between patients (ie, a prevention checklist; Figure 1). The bundle begins with “provider suspicion,” which is defined as the ordering of a stool C. difficile toxin test (Figure 1).
Figure 1. Clostridium difficile infection checklist at Brigham and Women’s Hospital.
Because of the increase in the incidence of CDI at Brigham and Women’s Hospital and reports of the increasing severity of the disease, we elected to incorporate some infection prevention practices recommended for outbreak settings into special and more stringent isolation precautions that we termed “contact precautions plus” (CPP), including cleaning rooms with a hypochlorite‐based disinfectant after patient is discharged from hospital and emphasizing hand hygiene with soap and water after contact with patients with CDI. The CPP signs were brightly colored, facilitating both staff awareness and compliance (Figure 2).
Figure 2. Contact precautions plus sign, as developed at Brigham and Women’s Hospital for healthcare workers caring for patients with Clostridium difficile infection.
The infection control practices in the prevention bundle were initiated when a provider suspected that a patient may have CDI. Providers were then encouraged to start CPP, to order a C. difficile toxin test, and to discontinue all nonessential antimicrobials as well as all antiperistaltic medications.
The patient’s nurse was responsible for obtaining a stool specimen for C. difficile toxin testing, for moving the patient to a single‐patient room, and for placing a CPP sign outside the doors of the patient’s room. In keeping with the CPP requirements, the nurse must ensure that adequate supplies of gloves and gowns are easily accessible outside of the patient's room, must place a stethoscope inside the patient's room for the patient's use only, and must urge all staff to wash their hands with soap and water after having had contact with the patient with suspected or confirmed CDI.
The next set of practices in the prevention bundle involved clinical laboratory notification procedures. The microbiology laboratory staff were responsible for calling the relevant hospital floor to notify the patient’s nurse verbally after each positive C. difficile toxin test result. In addition, a list of positive C. difficile toxin test results was generated each day from the clinical microbiology database and automatically sent to the hospital’s infection preventionists.
The infection preventionists were responsible for checking all microbiology results for positive C. difficile test results each weekday. Once informed of a patient’s positive test result, the infection preventionist must contact the patient’s floor to verify that the patient is in a single‐patient room and that the CPP sign is posted on the patient’s door. The preventionist then must activate a C. difficile flag in the electronic medical record to alert anyone accessing the patient’s clinical information to the patient’s CDI status. In addition, the preventionist must notify environmental services management each day about all patients on CPP. As a final step, environmental services personnel are trained to look for the CPP sign prior to cleaning a room after the patient is discharged from the hospital and to use a hypochlorite‐based cleaning agent to clean rooms that have a CPP sign posted.
Treatment Bundle
A CDI treatment bundle (Figure 1) was created to standardize the treatment of patients with severe CDI at our institution and to provide guidelines for when to consider consulting the surgery department regarding a patient's need for a colectomy. The guidelines contained in the treatment bundle were the result of a multidisciplinary effort involving infectious disease physicians, general surgeons, infection control preventionists, and pharmacy staff. These guidelines recommended the specific actions for physicians, physician assistants, and nurse practitioners to take based on categories of CDI (including mild, moderate, and severe disease) determined by clinical and laboratory information. In particular, for patients meeting the criteria for severe CDI, the guidelines urged providers to obtain immediate consultations with the departments of infectious diseases and general surgery. The overall goal of these guidelines was to encourage timely and effective medical treatment and prompt surgical treatment, if appropriate, based on consultation with the surgery department, as a means of reducing the risk for serious complications and mortality resulting from CDI.
Implementation
We launched our educational campaign and prevention bundle in January 2006. The treatment bundle was introduced in April 2006.
Statistical Methods
We analyzed the aggregate number of cases of CDI and the aggregate number of patient‐days over 20 different 3‐month time periods. We defined 2 study time periods: the preintervention period (from January 2004 through March 2006) and the postintervention period (from April 2006 through December 2008). We analyzed the data for differences in outcomes between the pre‐ and postintervention periods. Our primary outcome was the rate of healthcare‐associated CDI (estimated as the number of cases of healthcare‐associated CDI per 1,000 patient‐days) over each of the 20 different 3‐month time periods. Our secondary outcome was the CDI case‐fatality rate (estimated as the number of deaths due to healthcare‐associated CDI divided by the number of cases of healthcare‐associated CDI, times 100) over each of the 20 different 3‐month time periods. By means of Poisson regression analysis, log‐linear models were used to determine the rates over time, as proposed by Holford18 and Laird and Olivier.19 In the log‐linear models, we tested for trends over time and for changes before and after implementation of the checklists using Wald statistics. We also controlled for the possible confounding effects of an aggregate Charlson score (as a proxy for severity of illness) during the time period. A P value of less than .05 was considered to be statistically significant. All tests were 2‐sided.
Results
The surveillance data encompassed 1,047,849 patient‐days, from January 2004 to December 2008. The preintervention period included 431,264 patient‐days, and the postintervention period included 616,585 patient‐days.
The incidence rate of healthcare‐associated CDI decreased from an average of 1.10 cases per 1,000 patient‐days (95% confidence interval [CI], 1.00–1.21) during the preintervention period to 0.66 cases per 1,000 patient‐days (95% CI, 0.60–0.72) during the postintervention period. This reduction was sustained for 21 months and amounted to a 40% decrease in the rate of healthcare‐associated CDI (incidence rate ratio, 0.60 [95% CI, 0.52–0.68]; 
) after the implementation of the prevention bundle (Table 1 and Figure 3).
Figure 3. Incidence rates of healthcare‐associated Clostridium difficile infection (CDI) among patients hospitalized at Brigham and Women’s Hospital (excluding newborns in the neonatal intensive care unit).
During the course of our study, the number of C. difficile toxin tests sent to the microbiology laboratory increased significantly from the preintervention period (rate, 28.0 tests per 1,000 patient‐days [95% CI, 27.5–28.5]) to the postintervention period (rate, 32.1 tests per 1,000 patient‐days [95% CI, 31.7–32.6]). There was a 15% increase in the rate of C. difficile toxin testing (testing rate ratio, 1.15 [95% CI, 1.12–1.17]; 
) after the implementation of the prevention bundle (Table 2).
Among patients with CDI, there was no statistically significant decrease in the probability of dying from CDI (Table 1). The case‐fatality rate was 2.52% (95% CI, 1.44%–4.39%) before the prevention bundle was introduced and 2.22% (95% CI, 1.16%–4.21%) after the prevention bundle was introduced. The relative risk of dying from CDI during the postintervention period, compared with the preintervention period, was 0.88 (95% CI, 0.37–2.11; 
). Because of the small number of deaths due to healthcare‐associated CDI, we would have had adequate power to detect a statistically significant difference only if there had been a relative risk of 0.1 or less in the probability of dying from CDI during the postintervention period, compared with during the preintervention period.
During the course of our study, the medical acuity of our hospitalized patients' CDI increased, as reflected by the aggregate Charlson scores over time (Table 3). There was a significant increase in the aggregate Charlson score over time. The Spearman rank correlation between time period and Charlson score was 
(
for trend). In addition, on the basis of Poisson regression analysis, the aggregate Charlson score was not a statistically significant predictor of CDI over time (
) (Table 3).
Discussion
In our study, the implementation of our CDI educational campaign, prevention bundle, and treatment bundle was associated with a statistically significant, approximately 2‐year decrease in the incidence rate of healthcare‐associated CDI. This decrease occurred despite an increase in patient acuity among hospitalized patients at our institution, as reflected by the upward trend in aggregate Charlson scores over time. There were no major changes in CDI‐related infection control practices other than the ones included in our intervention.
We did not observe a significant decrease in CDI‐related mortality. However, the number of CDI‐related deaths was small, and we were underpowered to detect a statistically significant difference.
Although our educational programs included a discussion about exposure to antimicrobials as a major risk factor for CDI, we did not seek to restrict the use of high‐risk antimicrobials because of the intensity of resources that this would have required. A number of CDI prevention efforts reported in the literature have required inclusion of antimicrobial restriction as a component of the intervention, to demonstrate improvement in outcomes.13,20‐23 Antimicrobial stewardship programs are resource intensive,24 particularly when restricting the commonly used antimicrobials that are associated with increased risk for CDI, such as fluoroquinolones. An intervention such as ours that does not require antimicrobial restriction has the advantage of requiring fewer institutional resources.
Our intervention relied on increasing provider suspicion for CDI. The significant increase in C. difficile toxin testing that we observed after the intervention suggests that the educational campaign and prevention bundle effectively lowered the threshold for hospital providers to suspect the diagnosis of CDI among patients with diarrhea and to consider diagnostic testing.
Under pressure from payers who may no longer reimburse for cases of CDI and other healthcare‐associated infections, providers may be pushed to limit toxin testing and other documentation of CDI. It is important to emphasize, however, that there is potentially a dangerously high cost to be paid—both in lives lost and money wasted—in decreasing documentation of healthcare‐associated CDI and, therefore, foregoing the opportunity to decrease the spread of this serious infection. We believe that there is a substantial benefit associated with increasing provider suspicion and C. difficile toxin testing and with confirming cases of CDI. We urge hospital administrators to adopt a more active and transparent approach to reducing the incidence and severity of CDI by introducing this hospital‐wide prevention and treatment intervention.
There were a number of limitations to our study, including its observational before‐after design. The decrease in incidence of healthcare‐associated CDI, however, was temporally related to the implementation of the intervention, took place during a period of continued increasing incidence of CDI based on national data,10 and was sustained over 21 months. The decreasing rates of CDI that we noted after the implementation of our intervention are even more striking because of the more complete ascertainment of cases of CDI that would be expected with an increased frequency of C. difficile toxin testing.
An additional limitation was our inability to correlate compliance with specific practices with the decrease in CDI rates, because we did not formally monitor adherence to the components of the bundles. We did, however, as indicated above, achieve a statistically significant increase in C. difficile toxin testing after implementation of the intervention. Future work regarding measurement of compliance with specific practices in the bundles is warranted.
Further research to evaluate the cost‐effectiveness of our prevention and treatment bundles would be useful. We did not gather specific cost data for the cases of CDI included in our study, although we would suspect that the cost of preventive equipment such as gowns and gloves would be small relative to the cost savings from the decrease in incidence. The use of more single‐patient rooms may represent an opportunity cost.
Our findings suggest that the guidelines from the Society for Healthcare Epidemiology of America and the Infectious Diseases Society of America can be effectively implemented as prevention and treatment bundles and can help to substantially reduce the incidence of CDI. We recognize that adherence to the practices included in our bundles could be improved, and one approach to do so is to distribute these bundles as a checklist (Figure 1), with individual sections for specific healthcare workers. Checklists offer a convenient method for confirmation that necessary actions have been performed by the teams of healthcare workers. One means of implementing the CDI checklist and tracking compliance with its steps is to print the checklist on stickers that are checked off by physicians, physician assistants, nurse practitioners, or registered nurses, as appropriate, and placed in the patient’s medical chart. The microbiology, infection control, and environmental services portions of the checklist could be printed on posters that hang in the relevant hospital departments where they could be confirmed daily. Alternatively, the portions of the prevention and treatment checklists aimed at providers could be built directly into electronic ordering systems, creating systems‐based forcing functions that directly improve quality of care.
Our CDI checklist, which combines critical multidisciplinary and systems‐oriented approaches,14 is a simple tool to aid hospitals in tackling the devastating problem of CDI. This tool should be tested on a larger scale to evaluate its generalizability.
Acknowledgments
Potential conflicts of interest. All authors report no potential conflicts of interest relevant to this article.
References
- 1. McDonald LC, Killgore GE, Thompson A, et al. An epidemic, toxin gene‐variant strain of Clostridium difficile. N Engl J Med 2005; 353:2433–2441.
- 2. McDonald LC, Owings M, Jernigan DB. Clostridium difficile infection in patients discharged from US short‐stay hospitals, 1996–2003. Emerg Infect Dis 2006; 12:409–415.
- 3. Zilberberg MD, Shorr AF, Kollef MH. Increase in adult Clostridium difficile–related hospitalizations and case‐fatality rate, United States, 2000–2005. Emerg Infect Dis 2008; 14:929–931.
- 4. Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficile–related mortality rates, United States, 1999–2004. Emerg Infect Dis 2007; 13:1417–1419.
- 5. Pépin J, Valiquette L, Alary ME, et al. Clostridium difficile–associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ 2004; 171:466–472.
- 6. Pépin J, Valiquette L, Cossette B. Mortality attributable to healthcare‐associated Clostridium difficile–associated disease during an epidemic caused by a hypervirulent strain in Quebec. CMAJ 2005; 173:1037–1042.
- 7. Loo VG, Poirier L, Miller MA, et al. A predominantly clonal multi‐institutional outbreak of Clostridium difficile–associated diarrhea with high morbidity and mortality. N Engl J Med 2005; 353:2442–2449.
- 8. Dubberke ER, Reske KA, Olsen MA, McDonald LC, Fraser VJ. Short‐ and long‐term attributable costs of Clostridium difficile–associated disease in nonsurgical patients. Clin Infect Dis 2008; 46:497–504.
- 9. Song X, Bartlett JG, Speck K, Naegeli A, Carroll K, Perl TM. Rising economic impact of Clostridium difficile–associated disease in adult hospitalized patient population. Infect Control Hosp Epidemiol 2008; 29:823–828.
- 10. Blossom DB, McDonald LC. The challenges posed by reemerging Clostridium difficile infection. Clin Infect Dis 2007; 45:222–227.
- 11. McDonald LC. Confronting Clostridium difficile in inpatient health care facilities. Clin Infect Dis 2007; 45:1274–1276.
- 12. Dubberke ER, Gerding DN, Classen D, et al. Strategies to prevent Clostridium difficile infections in acute care hospitals. Infect Control Hosp Epidemiol 2008; 29:S81–S92.
- 13. Gerding, DN, Johnson S, Peterson LR, Mulligan ME, Silva J Jr. Clostridium difficile–associated diarrhea and colitis. Infect Control Hosp Epidemiol 1995; 16:459–477.
- 14. Goldmann DA, Weinstein RA, Wenzel RP, et al. Strategies to prevent and control the emergence and spread of antimicrobial‐resistant microorganisms in hospitals. JAMA 1996; 275:234–240.
- 15. Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter‐related bloodstream infections in the ICU (published correction appears in N Engl J Med 2007; 356:2660). N Engl J Med 2006; 355:2725–2732.
- 16. Haynes AB, Weiser TG, Berry WR, et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med 2009; 360:491–499.
- 17. Institute for Healthcare Improvement. Healthcare‐associated infections: reducing healthcare‐associated infections. Available at: http://www.ihi.org/IHI/Topics/HealthcareAssociatedInfections/. Accessed March 26, 2009.
- 18. Holford, TR. The analysis of rates and survivorship using log‐linear models. Biometrics 1980; 36:299–306.
- 19. Laird N, Olivier D. Covariance analysis of censored survival data using log‐linear analysis techniques. J Am Stat Assoc 1981; 76:231–240.
- 20. Pear SM, Williamson TH, Bettin KM, Gerding DN, Galgiani JN. Decrease in nosocomial Clostridium difficile–associated diarrhea by restricting clindamycin use. Ann Intern Med 1994; 20:272–277.
- 21. Muto CA, Blank MK, Marsh JW, et al. Control of an outbreak of infection with the hypervirulent Clostridium difficile B1 strain in a university hospital using a comprehensive “bundle” approach. Clin Infect Dis 2007; 45:1266–1273.
- 22. Valiquette L, Cossette B, Garant MP, Diab H, Pépin J. Impact of a reduction in the use of high‐risk antibiotics on the course of an epidemic of Clostridium difficile–associated disease caused by the hypervirulent NAP1/027 strain. Clin Infect Dis 2007; 45:S112–S121.
- 23. Fowler S, Webber A, Cooper BS, et al. Successful use of feedback to improve antibiotic prescribing and reduce Clostridium difficile infection: a controlled interrupted time series. J Antimicrob Chemother 2007; 59:990–995.
- 24. Dellit TH, Owens RC, McGowan JE, 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; 44:159–177.