- Email: [email protected]
Prevalence of Clostridium Difficile Colonization at Admission to Rehabilitation
1086
ORIGINAL ARTICLE
Prevalence of Clostridium Difficile Colonization at Admission to Rehabilitation Christina Marciniak, MD, David Chen, MD, Adam C. Stein, MS, Patrick E. Semik, BA ABSTRACT. Marciniak C, Chen D, Stein AC, Semik PE. Prevalence of Clostridium difficile colonization at admission to rehabilitation. Arch Phys Med Rehabil 2006;87:1986-90. Objectives: To assess the prevalence of intestinal colonization with Clostridium difficile (C. difficile) at admission to acute rehabilitation and to identify risk factors associated with colonization. Design: Case-control study. Participants: Consecutive admissions to 2 rehabilitation units (spinal cord injury, brain injury and stroke). Setting: Free-standing acute rehabilitation facility. Interventions: Rectal swabs for culture for C. difficile were obtained at admission and cytotoxin assay performed on all culture positive specimens. Rates of colonization with cytotoxic C. difficile were calculated. Charts were reviewed for medical and demographic factors that may have predisposed patients to colonization, and for possible symptoms at the time of admission. Main Outcome Measures: Percentage of patients with culture and cytotoxin assay positive for C. difficile. Frequency of specific patient characteristics that could predispose to C. difficile colonization. Results: Of admission stool samples, 16.4% tested positive for C. difficile; none of these patients had been identified as colonized before admission. No patients were discordant for C. difficile positivity on culture and presence of a toxigenic strain. No medical or demographic factors were associated with increased risk of colonization, including age (t52⫽⫺.748, P⫽.458, not significant [NS]), diarrhea within 24 hours of admission (21 test⫽.001, P⫽.973 [NS]), or use of oral or intravenous antibiotics at admission (21 test⫽.044, P⫽.834 [NS]). Conclusions: Patients admitted to acute rehabilitation may have an elevated rate of intestinal colonization with C. difficile without having clinical symptoms. No medical or demographic characteristics were found to be predictive of colonization, however, most of the patients admitted had more than 1 factor that may have increased their susceptibility to infection with this organism. Inadvertent transfer of this organism within the rehabilitation setting may occur because asymptomatic colonization is not recognized. Key Words: Clostridium difficile; Rehabilitation. © 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
From the Department of Physical Medicine & Rehabilitation, Northwestern University Feinberg Medical School, Chicago, IL (Marciniak, Chen, Semik); Rehabilitation Institute of Chicago, Chicago, IL (Marciniak, Chen, Semik); and University of Illinois, Chicago, IL (Stein). Presented to the American Academy of Physical Medicine & Rehabilitation, October 2005, Philadelphia, PA. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Christina Marciniak, MD, Rehabilitation Institute of Chicago, 345 E Superior St, Chicago, IL 60611. 0003-9993/06/8708-10557$32.00/0 doi:10.1016/j.apmr.2006.03.020
Arch Phys Med Rehabil Vol 87, August 2006
LOSTRIDIUM DIFFICILE (C. Difficile), a gram-positive, spore-forming anaerobe, is the most common source C of infectious diarrhea in hospitalized acute care and acute
rehabilitation patients.1,2 Diarrhea associated with antibiotic use is common, and 15% to 20% of such cases may be caused by C. difficile.3 Though this organism is a component of normal fecal flora in a small percentage of adults (up to 3%), it may be detected in from 10% to 30% of hospitalized adults without diarrhea who have been treated with antibiotics, or who have received chemotherapeutic agents.4,5 Increases in hospital length of stay (LOS) and costs result when acute care patients develop a C. difficile infection.6 There are similar increases in LOS in hospitalized patients in acute rehabilitation who develop nosocomial infections, including C. difficile.1,7 Significant morbidity, including functional status and, at times, mortality, are associated with this infection.1 Of concern is the increasing incidence of symptomatic C. difficile infection among hospitalized patients in the United States, from 0.68% to 1.2%, including an increase in serious disease.8 It is not clear, however, whether this is the result of increased detection of these cases.8 Because of the medical and economic consequences resulting from C. difficile infection, implementation of methods to reduce spread of this organism in hospital settings is extremely important. Transmission between patients is particularly a risk in a rehabilitation environment where the focus is on encouraging group activities and socialization. For the same reasons, controlling the spread of the organism is more difficult in the rehabilitation setting without compromising the rehabilitation milieu. While infection control measures can be implemented after the infection has been identified in symptomatic patients, inadvertent spread may occur when colonization is unrecognized.9 Asymptomatic colonization with C. difficile can result when the normal intestinal tract bacterial flora are altered, generally from use of antibiotics; proliferation of toxigenic strains of C. difficile may occur in a colonized patient.10 Patients admitted to acute rehabilitation are most frequently transferred from inpatient hospital settings and often have comorbid conditions that have been associated with greater risk for C. difficile infection during acute care.1,11 Studies have identified the incidence of asymptomatic infection with C. difficile and incidence on admission to acute care, however, this has not previously been studied in an acute rehabilitation population.6 The occurrence of symptomatic C. difficile–associated diarrhea among inpatients at our institution, despite persistent efforts to control its spread, prompted our infection control committee to study whether colonization that is not identified when a patient is admitted, contributes to the incidence of symptomatic disease, such as might be seen when antibiotics are initiated in treatment. Another goal of the study was to identify at admission any medical and demographic characteristics that may be predictive of C. difficile colonization.
METHODS Patients admitted consecutively to 2 inpatient units (spinal cord injury [SCI], brain injury and stroke) of our free-standing acute rehabilitation facility over a 3-month period were
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
screened for C. difficile colonization. Rectal swabs were sent for anaerobic culture for C. difficile within a day of inpatient admission. Specimens with cultures positive for C. difficile were evaluated by enzyme immunoassay for C. difficile cytotoxin, which detects both A and B cytotoxin. Subtyping analysis was performed on any positive cultures and on any C. difficile cultures of patients who developed infection after hospitalization. Molecular strain typing of the C. difficile isolates was performed by both pulsed field gel electrophoresis and restriction endonuclease analysis. Deoxyribonucleic acid (DNA) fragments were extracted, stained, and visualized under ultraviolet light. The resulting band patterns were visually compared to determine if the isolates were genetically similar. Institutional review board approval was given for hospital chart review for patients screened during the infection control admission study. We used the admission medical history and physical examination to acquire information on the patients’ status prior to their admission, including comorbid conditions, the rehabilitation diagnosis, infections, including a history of C. difficile, and antibiotic use. In addition, we identified specific types of medications that the patient had been taking at the time of admission: intravenous or oral antibiotics, intravenous or oral steroids, and gastric acid suppressing drugs. We reviewed the nursing admission form for charting of a history of loose or liquid stools or diarrhea prior to admission; nursing electronic documentation of bowel movement characteristics was also recorded for these specific stool consistencies within the 24 hours after admission. A patient was considered to have diarrhea if there was at least 1 documented loose or liquid bowel movement, using these descriptive terms, within 24 hours after admission. The development of symptomatic C. difficile with treatment was also recorded. The characteristics of patients that were positive for C. difficile infection by culture and cytotoxin at admission were compared with those that were negative. Potential risk factors were evaluated for any significant differences between the groups, including the following characteristics: rehabilitation impairment category, infections experienced during acute care, antibiotic use at admission, the transferring hospital, and use of specific medication classes. Groups were also compared for acute care LOS and age with the t test for equality of means. Differences between groups were considered significant for P values less than or equal to .05. RESULTS Fifty-five consecutive patients admitted to these units were screened; rectal swabs were taken and sent for C. difficile culture. One patient’s sample was rejected by the lab because it was submitted in an inappropriate media. Another sample was collected at day 5; that culture was positive for C. difficile, however, the patient was omitted from further analyses because we could not be certain of his status at admission. Nine of the remaining 54 patients tested positive for C. difficile. Their specimens were subsequently evaluated by cytotoxin assay and all 9 were positive for C. difficile cytotoxin. None of these patients had been identified as being colonized with C. difficile before their admission. One patient was taking metronidzole for C. difficile at the time of admission. This patient’s culture was negative at admission. Our study sample included 34 men and 20 women, ranging in age from 20 to 79 years. Age and sex differences between those admitted with and without C. difficile colonization were not significant (NS) (t52⫽⫺.748, P⫽.458, [NS]; 21 test⫽.254, P⫽.973 [NS], respectively). Six patients were reported either by the referring facility or by the patient to have loose or liquid stools at transfer. Eleven
1087
patients had at least 1 loose or liquid stool within 24 hours after admission. Stool consistency either before or during the 24 hours subsequent to admission was not predictive of C. difficile colonization (21 test⫽.033, P⫽.856 [NS]; 21 test⫽.001, P⫽ .973 [NS], respectively.) Ninety-six percent of all admissions had at least 1 factor previously reported in the literature to predispose to C. difficile, and 79.7% had 2 or more risk factors. Patients were frequently on gastric acid suppressive agents at transfer. Nine were on histamine2 (H2) blocking agents and 9 were on proton-pump inhibitors (PPIs) when admitted. Nine patients were also on antibiotics at admission. None of these agents was associated with increased colonization (table 1). Patients were also frequently treated for infections while in acute care, with pneumonias, urinary tract infections, and wounds reported most frequently. Neurogenic bowel dysfunction was reported in 32 patients, but this was not associated with increased C. difficile disease. Patients testing both positive or negative for C. difficile had similarly long lengths of stay in acute care (see table 1). Patients identified as positive for toxigenic C. difficile were isolated, but not immediately treated with antibiotics. Four patients subsequently required antibiotics for C. difficile because of gastrointestinal symptoms. Two patients who were initially negative became symptomatic during their hospitalization and tested positive for C. difficile. Subtyping analysis for specific DNA banding patterns failed to find a nosocomial transmission pattern in the 2 patients who developed symptoms and positive C. difficile studies after initial negative studies. DISCUSSION Nosocomial infection is associated with a longer LOS and a smaller improvement in functional status for patients admitted to acute rehabilitation.1 This same detrimental effect seen after C. difficile infection may be in part due to its significant morbidity, including profuse diarrhea, which limits participation in the therapy program.7 C. difficile–associated diarrhea is the result of severe inflammation and increased secretory volume of the colon as the result of induced prostaglandin production.12 In addition, cytoskeletal protein regulation is altered, ultimately leading to cell death. Mucosal permeability increases, with resultant fluid secretion. The typical clinical manifestations of symptomatic C. difficile infection include diarrhea, abdominal pain, low-grade fever, and leukocytosis, although fevers as high as 40°C and leukocytosis to 50⫻109/L have been reported.3,13 In addition to colitis, infection with C. difficile may result in liver abscess, bacteremia, sepsis, splenic and pancreatic abscesses, peritonitis, small bowel enteritis, and bone and prosthetic joint infection.14-17 We were unable to demonstrate a correlation with a history of diarrhea around the time of admission in our patient population. Symptoms such as abdominal pain would likely not be a sensitive measure in the patients we screened in this study because of their sensory and cognitive deficits. Patients in hospital geriatric and rehabilitation wards have the highest incidence of C. difficile–associated diarrhea.2 Prevalence of C. difficile as a source of diarrhea in acute rehabilitation is as high as 39%.18 This may be related to difficulty in implementing strategies to decrease nosocomial transmission within a rehabilitation hospital that encourages patient interactions. C. difficile is most commonly transmitted through the hands of hospital personnel or through handling contaminated objects.19 Isolation of patients and the use of gloves can reduce spread of the infection, but early identification of infected patients is important so that limiting their personal contacts can help to control spread.20 Arch Phys Med Rehabil Vol 87, August 2006
1088
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak Table 1: Comparison of Patient Characteristics With and Without C. Difficile Colonization at Admission C. Difficile Culture at Admission
Negative
Positive
No. of patients Primary rehabilitation diagnosis SCI Brain injury SCI and brain injury Other Average LOS in acute care ⫾ SD (d) Sex Female Male Age at admission ⫾ SD (y) Infections in acute care Urinary tract infection Pneumonia Wound History of neurogenic bowel dysfunction Surgery in acute care Medication use Antibiotics at admission H2 blocker at admission PPI at admission Diarrhea at transfer Diarrhea 24h after transfer
45 (83.3)
9 (16.6)
18 (40) 12 (27) 4 (9) 11 (24) 30.44⫾20.96
4 (44) 2 (22) 0 (0) 3 (33) 29.44⫾18.05
16 (36) 29 (64) 46.98⫾17.90
4 (44) 5 (56) 52.11⫾23.18
Statistical Test
12 test⫽.061, NS 12 test⫽.077, NS 12 test⫽.864, NS 12 test⫽.254, NS t52⫽.133, NS 12 test⫽.254, NS
t52⫽⫺.748, NS
14 (31) 17 (38) 4 (9) 26 (58) 33 (73)
3 (33) 1 (11) 1 (11) 6 (67) 7 (78)
12 test⫽.017, NS 12 test⫽2.66, NS 12 test⫽.044, NS 12 test⫽.245, NS 12 test⫽.077, NS
18 (40) 13 (29) 19 (42) 6 (13) 11 (24)
4 (44) 2 (22) 6 (67) 1 (11) 2 (25)
12 test⫽.061, NS 12 test⫽.166, NS 12 test⫽1.80, NS 12 test⫽.033, NS 12 test⫽.001, NS
NOTE. Values are n (%) unless otherwise indicated. Abbreviation: SD, standard deviation.
C. difficile can be acquired asymptomatically within hospitals and transmitted to other patients.9 After colonization is identified, treatment is not recommended because metronidzole and vancomycin do not reliably eradicate spore carriage in asymptomatic carriers.3 Although the patients identified in our study were not treated unless they became symptomatic, precautions were taken to prevent spread of the organism immediately after their positive cultures were reported. The U.S. Center for Disease Control and Prevention (CDC) recommends that patients with C. difficile disease be restricted to a private room or cohorted with patients with the same organism. Dedicated equipment is also recommended. For hand hygiene, handwashing with soap and water, as opposed to alcohol gel, is believed to be more effective because of C. difficile spores’ resistance to the alcohol gels. Hypochloritebased disinfectant should be used to clean environmental surfaces.21 In an inpatient rehabilitation setting, patients are often treated in common areas; this could potentially result in patient-to-patient contact and organism transmission. A previous study at our institution, however, found that while surfaces within patients’ rooms were frequently colonized with a drugresistant organism found on stool culture, common areas were much less likely to be contaminated, even when patients with this organism frequented these areas.22 Thus, although our institution does follow CDC guidelines regarding private room assignment for patients known to be harboring toxigenic C. difficile, they are not restricted to their rooms for therapy unless they are incontinent or unregulated in their bowel regimens. Therapy equipment is cleaned after treatment with the same recommended solutions and therapists always wear gloves. Because fewer patients on these units with symptomatic disease were noted during this study, when isolation was employed on identification of C. difficile colonization, earlier identification of colonized patients may be more important. Arch Phys Med Rehabil Vol 87, August 2006
Testing for C. difficile can be done through culture of the organism as well as detection of the cytotoxin.23 Cultures generally take at least 3 days to produce results and toxins may degrade during transport and become inactive.20 Strategies that use both toxin and cultures should result in increased yield. Toxin immunoassays may have a false-negative rate of from 10% to 20% and thus sensitivity is improved by sending a second stool sample on the next day.20 Stool cultures, however, detect nontoxigenic strains. Sensitivities of the anaerobic culture technique we use are approximately 95%. The decreased sensitivity of the toxin assay did not affect our ability to identify colonization at admission, because all culture-positive patients were positive on the initial cytotoxic assay. An important risk factor for development of C. difficile diarrhea is antibiotic exposure.10 Antibiotic-associated alterations in flora result in overgrowth of C. difficile. In the gut, antibiotic associated C. difficile disease may occur after a few days or up to 2 months after drug exposure. Though clindamycin is the antibiotic most frequently implicated, due to resistant strains of C. difficile, other agents such as second and third generation cephalosporins, penicillins, and floroquinolones, may also produce symptomatic disease.24-26 Other related predisposing factors include duration of antibiotic use, with treatment courses of fewer than 3 days being less associated than longer treatment.10 We were unable to determine whether antibiotic use during rehabilitation was associated with C. difficile colonization. Also, because of the low frequency of specific antibiotic types in use at admission, we could not evaluate the effects of different antibiotic types. The frequency of infections while in acute care was extremely high in this population, however, with most patients having antibiotic exposure at some point during their acute care stay. Because information was not available in all cases about specific antibiotics used during acute care, we could not study preadmission use of specific types of antibiotics as a risk factor.
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
Among patients in inpatient acute rehabilitation, those with SCI have a significantly higher rate of nosocomial infection, including C. difficile.1 LOS appears to influence infection rate; 21% of patients in 1 study developed C. difficile after admission at a median of 12 days after admission. Incidence increases significantly 4 weeks after entering the hospital.11 Our patients generally had acute care LOS that exceeded this parameter. Advanced age has also been noted to be a risk factor; in a Swedish study, persons older than 60 years were 20 to 100 times more likely to have positive assays for toxin.2 In contrast to these reported associations, we did not find that age or the diagnosis of SCI, compared with a brain injury diagnosis, were predisposing factors, but did confirm the prior finding that SCI subjects receiving acute care are commonly colonized. Use of PPIs increases the risk of developing C. difficile infection.27 Concurrent antibiotic use, cytotoxic chemotherapy, and PPIs result in a much greater risk, with an odds ratio (OR) of 43.2.27 A recent Canadian cohort and case control study28 found adjusted ORs of 2.1 and 2.7 with the use of PPIs. Although we considered these medication factors on admission, we were unable to demonstrate an increased risk of colonization with the current use of antibiotics or acid-suppressive agents, including PPIs. This study was limited in that information on the acute care stay was obtained from the rehabilitation admission history and physical examination and thus may not have identified all acute care treatments. In addition, as mentioned previously, the use of a single swab likely resulted in our underestimating the colonization rate. Also, there was not a standard definition for “loose” or “liquid” stools as documented by the nursing staff, however, this nursing documentation is what is used clinically in the decision to begin laboratory evaluation of C. difficile infection. Finally, this study was conducted on units where the patients tended to have higher acute care LOS before hospitalization for rehabilitation, and therefore may not be representative of other types of rehabilitation patients, for example, those with orthopedic conditions. At our facility, a C. difficile culture costs $25 and a toxin assay costs $18. Although it was not considered in this study, it may be cost effective to screen patients with admission characteristics that put them at high risk, because infections adversely affect attendance at therapy sessions.7 Such screening would include the patient groups confirmed in this study to be frequently colonized. We were, unfortunately, unable to distinguish subgroups of people within these rehabilitation populations of lower risk. It would be important to document in further studies the cost effectiveness of screening strategies, and to determine the potentially different rates of colonization in other inpatient rehabilitation populations that are less likely to have multiple risk factors, for example, orthopedic patients. CONCLUSIONS Patients admitted to acute rehabilitation may be colonized with asymptomatic intestinal C. difficile. While previously identified risk factors were not predictive of colonization in this study, most patients had more than 1 factor that may have increased their susceptibility to colonization. Inadvertent transfer of the organism within the rehabilitation unit may occur if this colonization is not recognized. References 1. Mylotte JM, Graham R, Kahler L, Young BL, Goodnough S. Impact of nosocomial infection on length of stay and functional improvement among patients admitted to an acute rehabilitation unit. Infect Control Hosp Epidemiol 2001;22:83-7.
1089
2. Karlstrom O, Fryklund B, Tullus K, Burman LG. A prospective nationwide study of Clostridium difficile-associated diarrhea in Sweden. The Swedish C. difficile Study Group. Clin Infect Dis 1998;26:141-5. 3. Fekety R. Guidelines for the diagnosis and management of Clostridium difficile-associated diarrhea and colitis. American College of Gastroenterology, Practice Parameters Committee. Am J Gastroenterol 1997;92:739-50. 4. Kimmey MB, Elmer GW, Surawicz CM, McFarland LV. Prevention of further recurrences of Clostridium difficile colitis with Saccharomyces boulardii. Dig Dis Sci 1990;35:897-901. 5. Kyne L, Farrell RJ, Kelly CP. Clostridium difficile. Gastroenterol Clin North Am 2001;30:753-77, ix-x. 6. Kyne L, Sougioultzis S, McFarland LV, Kelly CP. Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea. Infect Control Hosp Epidemiol 2002;23: 653-9. 7. Yablon SA, Krotenberg R, Fruhmann K. Diarrhea in hospitalized patients. Am J Phys Med Rehabil 1992;71:102-7. 8. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg 2002;235:363-72. 9. McFarland LV, Mulligan ME, Kwok RY, Stamm WE. Nosocomial acquisition of Clostridium difficile infection. N Engl J Med 1989;320:204-10. 10. Wistrom J, Norrby SR, Myhre EB, et al. Frequency of antibiotic-associated diarrhoea in 2462 antibiotic-treated hospitalized patients: a prospective study. J Antimicrob Chemother 2001;47:43-50. 11. Clabots CR, Johnson S, Olson MM, Peterson LR, Gerding DN. Acquisition of Clostridium difficile by hospitalized patients: evidence for colonized new admissions as a source of infection. J Infect Dis 1992;166:561-7. 12. Alcantara C, Stenson WF, Steiner TS, Guerrant RL. Role of inducible cyclooxygenase and prostaglandins in Clostridium difficile toxin A-induced secretion and inflammation in an animal model. J Infect Dis 2001;184:648-52. 13. Kelly CP, Pothoulakis C, LaMont JT. Clostridium difficile colitis. N Engl J Med 1994;330:257-62. 14. Sakurai T, Hajiro K, Takakuwa H, Nishi A, Aihara M, Chiba T. Liver abscess caused by Clostridium difficile. Scand J Infect Dis 2001;33:69-70. 15. Freiler JF, Durning SJ, Ender PT. Clostridium difficile small bowel enteritis occurring after total colectomy. Clin Infect Dis 2001;33:1429-31; discussion 32. 16. Jacobs A, Barnard K, Fishel R, Gradon JD. Extracolonic manifestations of Clostridium difficile infections. Presentation of 2 cases and review of the literature. Medicine (Baltimore) 2001;80: 88-101. 17. Garcia-Lechuz JM, Hernangomez S, Juan RS, Pelaez T, Alcala L, Bouza E. Extra-intestinal infections caused by Clostridium difficile. Clin Microbiol Infect 2001;7:453-7. 18. Yablon SA, Krotenberg R, Fruhmann K. Clostridium difficilerelated disease: evaluation and prevalence among inpatients with diarrhea in two freestanding rehabilitation hospitals. Arch Phys Med Rehabil 1993;74:9-13. 19. McFarland LV, Surawicz CM, Stamm WE. Risk factors for Clostridium difficile carriage and C. difficile-associated diarrhea in a cohort of hospitalized patients. J Infect Dis 1990;162: 678-84. 20. Bartlett JG. Clinical practice. Antibiotic-associated diarrhea. N Engl J Med 2002;346:334-9. 21. Centers for Disease Control. Clostridium difficile: information for healthcare providers. August 2004, updated July 22, 2005. Available at: http://www.cdc.gov/ncidod/dhqp/id_CdiffFAQ_HCP.html#5. Accessed March 12, 2006. Arch Phys Med Rehabil Vol 87, August 2006
1090
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
22. Trick WE, Temple RS, Chen D, Wright MO, Solomon SL, Peterson LR. Patient colonization and environmental contamination by vancomycin-resistant enterococci in a rehabilitation facility. Arch Phys Med Rehabil 2002;83:899-902. 23. Fekety R, Shah AB. Diagnosis and treatment of Clostridium difficile colitis. JAMA 1993;269:71-5. 24. McCusker ME, Harris AD, Perencevich E, Roghmann MC. Fluoroquinolone use and Clostridium difficile-associated diarrhea. Emerg Infect Dis 2003;9:730-3. 25. Bates CJ, Wilcox MH, Spencer RC, Harris DM. Ciprofloxacin and Clostridium difficile infection. Lancet 1990;336:1193.
Arch Phys Med Rehabil Vol 87, August 2006
26. Climo MW, Israel DS, Wong ES, Williams D, Coudron P, Markowitz SM. Hospital-wide restriction of clindamycin: effect on the incidence of Clostridium difficile-associated diarrhea and cost. Ann Intern Med 1998;128(12 Pt 1):989-95. 27. Cunningham R, Dale B, Undy B, Gaunt N. Proton pump inhibitors as a risk factor for Clostridium difficile diarrhoea. J Hosp Infect 2003;54:243-5. 28. Dial S, Alrasadi K, Manoukian C, Huang A, Menzies D. Risk of Clostridium difficile diarrhea among hospital inpatients prescribed proton pump inhibitors: cohort and case-control studies. CMAJ 2004;171:33-8.
ORIGINAL ARTICLE
Prevalence of Clostridium Difficile Colonization at Admission to Rehabilitation Christina Marciniak, MD, David Chen, MD, Adam C. Stein, MS, Patrick E. Semik, BA ABSTRACT. Marciniak C, Chen D, Stein AC, Semik PE. Prevalence of Clostridium difficile colonization at admission to rehabilitation. Arch Phys Med Rehabil 2006;87:1986-90. Objectives: To assess the prevalence of intestinal colonization with Clostridium difficile (C. difficile) at admission to acute rehabilitation and to identify risk factors associated with colonization. Design: Case-control study. Participants: Consecutive admissions to 2 rehabilitation units (spinal cord injury, brain injury and stroke). Setting: Free-standing acute rehabilitation facility. Interventions: Rectal swabs for culture for C. difficile were obtained at admission and cytotoxin assay performed on all culture positive specimens. Rates of colonization with cytotoxic C. difficile were calculated. Charts were reviewed for medical and demographic factors that may have predisposed patients to colonization, and for possible symptoms at the time of admission. Main Outcome Measures: Percentage of patients with culture and cytotoxin assay positive for C. difficile. Frequency of specific patient characteristics that could predispose to C. difficile colonization. Results: Of admission stool samples, 16.4% tested positive for C. difficile; none of these patients had been identified as colonized before admission. No patients were discordant for C. difficile positivity on culture and presence of a toxigenic strain. No medical or demographic factors were associated with increased risk of colonization, including age (t52⫽⫺.748, P⫽.458, not significant [NS]), diarrhea within 24 hours of admission (21 test⫽.001, P⫽.973 [NS]), or use of oral or intravenous antibiotics at admission (21 test⫽.044, P⫽.834 [NS]). Conclusions: Patients admitted to acute rehabilitation may have an elevated rate of intestinal colonization with C. difficile without having clinical symptoms. No medical or demographic characteristics were found to be predictive of colonization, however, most of the patients admitted had more than 1 factor that may have increased their susceptibility to infection with this organism. Inadvertent transfer of this organism within the rehabilitation setting may occur because asymptomatic colonization is not recognized. Key Words: Clostridium difficile; Rehabilitation. © 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation
From the Department of Physical Medicine & Rehabilitation, Northwestern University Feinberg Medical School, Chicago, IL (Marciniak, Chen, Semik); Rehabilitation Institute of Chicago, Chicago, IL (Marciniak, Chen, Semik); and University of Illinois, Chicago, IL (Stein). Presented to the American Academy of Physical Medicine & Rehabilitation, October 2005, Philadelphia, PA. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Christina Marciniak, MD, Rehabilitation Institute of Chicago, 345 E Superior St, Chicago, IL 60611. 0003-9993/06/8708-10557$32.00/0 doi:10.1016/j.apmr.2006.03.020
Arch Phys Med Rehabil Vol 87, August 2006
LOSTRIDIUM DIFFICILE (C. Difficile), a gram-positive, spore-forming anaerobe, is the most common source C of infectious diarrhea in hospitalized acute care and acute
rehabilitation patients.1,2 Diarrhea associated with antibiotic use is common, and 15% to 20% of such cases may be caused by C. difficile.3 Though this organism is a component of normal fecal flora in a small percentage of adults (up to 3%), it may be detected in from 10% to 30% of hospitalized adults without diarrhea who have been treated with antibiotics, or who have received chemotherapeutic agents.4,5 Increases in hospital length of stay (LOS) and costs result when acute care patients develop a C. difficile infection.6 There are similar increases in LOS in hospitalized patients in acute rehabilitation who develop nosocomial infections, including C. difficile.1,7 Significant morbidity, including functional status and, at times, mortality, are associated with this infection.1 Of concern is the increasing incidence of symptomatic C. difficile infection among hospitalized patients in the United States, from 0.68% to 1.2%, including an increase in serious disease.8 It is not clear, however, whether this is the result of increased detection of these cases.8 Because of the medical and economic consequences resulting from C. difficile infection, implementation of methods to reduce spread of this organism in hospital settings is extremely important. Transmission between patients is particularly a risk in a rehabilitation environment where the focus is on encouraging group activities and socialization. For the same reasons, controlling the spread of the organism is more difficult in the rehabilitation setting without compromising the rehabilitation milieu. While infection control measures can be implemented after the infection has been identified in symptomatic patients, inadvertent spread may occur when colonization is unrecognized.9 Asymptomatic colonization with C. difficile can result when the normal intestinal tract bacterial flora are altered, generally from use of antibiotics; proliferation of toxigenic strains of C. difficile may occur in a colonized patient.10 Patients admitted to acute rehabilitation are most frequently transferred from inpatient hospital settings and often have comorbid conditions that have been associated with greater risk for C. difficile infection during acute care.1,11 Studies have identified the incidence of asymptomatic infection with C. difficile and incidence on admission to acute care, however, this has not previously been studied in an acute rehabilitation population.6 The occurrence of symptomatic C. difficile–associated diarrhea among inpatients at our institution, despite persistent efforts to control its spread, prompted our infection control committee to study whether colonization that is not identified when a patient is admitted, contributes to the incidence of symptomatic disease, such as might be seen when antibiotics are initiated in treatment. Another goal of the study was to identify at admission any medical and demographic characteristics that may be predictive of C. difficile colonization.
METHODS Patients admitted consecutively to 2 inpatient units (spinal cord injury [SCI], brain injury and stroke) of our free-standing acute rehabilitation facility over a 3-month period were
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
screened for C. difficile colonization. Rectal swabs were sent for anaerobic culture for C. difficile within a day of inpatient admission. Specimens with cultures positive for C. difficile were evaluated by enzyme immunoassay for C. difficile cytotoxin, which detects both A and B cytotoxin. Subtyping analysis was performed on any positive cultures and on any C. difficile cultures of patients who developed infection after hospitalization. Molecular strain typing of the C. difficile isolates was performed by both pulsed field gel electrophoresis and restriction endonuclease analysis. Deoxyribonucleic acid (DNA) fragments were extracted, stained, and visualized under ultraviolet light. The resulting band patterns were visually compared to determine if the isolates were genetically similar. Institutional review board approval was given for hospital chart review for patients screened during the infection control admission study. We used the admission medical history and physical examination to acquire information on the patients’ status prior to their admission, including comorbid conditions, the rehabilitation diagnosis, infections, including a history of C. difficile, and antibiotic use. In addition, we identified specific types of medications that the patient had been taking at the time of admission: intravenous or oral antibiotics, intravenous or oral steroids, and gastric acid suppressing drugs. We reviewed the nursing admission form for charting of a history of loose or liquid stools or diarrhea prior to admission; nursing electronic documentation of bowel movement characteristics was also recorded for these specific stool consistencies within the 24 hours after admission. A patient was considered to have diarrhea if there was at least 1 documented loose or liquid bowel movement, using these descriptive terms, within 24 hours after admission. The development of symptomatic C. difficile with treatment was also recorded. The characteristics of patients that were positive for C. difficile infection by culture and cytotoxin at admission were compared with those that were negative. Potential risk factors were evaluated for any significant differences between the groups, including the following characteristics: rehabilitation impairment category, infections experienced during acute care, antibiotic use at admission, the transferring hospital, and use of specific medication classes. Groups were also compared for acute care LOS and age with the t test for equality of means. Differences between groups were considered significant for P values less than or equal to .05. RESULTS Fifty-five consecutive patients admitted to these units were screened; rectal swabs were taken and sent for C. difficile culture. One patient’s sample was rejected by the lab because it was submitted in an inappropriate media. Another sample was collected at day 5; that culture was positive for C. difficile, however, the patient was omitted from further analyses because we could not be certain of his status at admission. Nine of the remaining 54 patients tested positive for C. difficile. Their specimens were subsequently evaluated by cytotoxin assay and all 9 were positive for C. difficile cytotoxin. None of these patients had been identified as being colonized with C. difficile before their admission. One patient was taking metronidzole for C. difficile at the time of admission. This patient’s culture was negative at admission. Our study sample included 34 men and 20 women, ranging in age from 20 to 79 years. Age and sex differences between those admitted with and without C. difficile colonization were not significant (NS) (t52⫽⫺.748, P⫽.458, [NS]; 21 test⫽.254, P⫽.973 [NS], respectively). Six patients were reported either by the referring facility or by the patient to have loose or liquid stools at transfer. Eleven
1087
patients had at least 1 loose or liquid stool within 24 hours after admission. Stool consistency either before or during the 24 hours subsequent to admission was not predictive of C. difficile colonization (21 test⫽.033, P⫽.856 [NS]; 21 test⫽.001, P⫽ .973 [NS], respectively.) Ninety-six percent of all admissions had at least 1 factor previously reported in the literature to predispose to C. difficile, and 79.7% had 2 or more risk factors. Patients were frequently on gastric acid suppressive agents at transfer. Nine were on histamine2 (H2) blocking agents and 9 were on proton-pump inhibitors (PPIs) when admitted. Nine patients were also on antibiotics at admission. None of these agents was associated with increased colonization (table 1). Patients were also frequently treated for infections while in acute care, with pneumonias, urinary tract infections, and wounds reported most frequently. Neurogenic bowel dysfunction was reported in 32 patients, but this was not associated with increased C. difficile disease. Patients testing both positive or negative for C. difficile had similarly long lengths of stay in acute care (see table 1). Patients identified as positive for toxigenic C. difficile were isolated, but not immediately treated with antibiotics. Four patients subsequently required antibiotics for C. difficile because of gastrointestinal symptoms. Two patients who were initially negative became symptomatic during their hospitalization and tested positive for C. difficile. Subtyping analysis for specific DNA banding patterns failed to find a nosocomial transmission pattern in the 2 patients who developed symptoms and positive C. difficile studies after initial negative studies. DISCUSSION Nosocomial infection is associated with a longer LOS and a smaller improvement in functional status for patients admitted to acute rehabilitation.1 This same detrimental effect seen after C. difficile infection may be in part due to its significant morbidity, including profuse diarrhea, which limits participation in the therapy program.7 C. difficile–associated diarrhea is the result of severe inflammation and increased secretory volume of the colon as the result of induced prostaglandin production.12 In addition, cytoskeletal protein regulation is altered, ultimately leading to cell death. Mucosal permeability increases, with resultant fluid secretion. The typical clinical manifestations of symptomatic C. difficile infection include diarrhea, abdominal pain, low-grade fever, and leukocytosis, although fevers as high as 40°C and leukocytosis to 50⫻109/L have been reported.3,13 In addition to colitis, infection with C. difficile may result in liver abscess, bacteremia, sepsis, splenic and pancreatic abscesses, peritonitis, small bowel enteritis, and bone and prosthetic joint infection.14-17 We were unable to demonstrate a correlation with a history of diarrhea around the time of admission in our patient population. Symptoms such as abdominal pain would likely not be a sensitive measure in the patients we screened in this study because of their sensory and cognitive deficits. Patients in hospital geriatric and rehabilitation wards have the highest incidence of C. difficile–associated diarrhea.2 Prevalence of C. difficile as a source of diarrhea in acute rehabilitation is as high as 39%.18 This may be related to difficulty in implementing strategies to decrease nosocomial transmission within a rehabilitation hospital that encourages patient interactions. C. difficile is most commonly transmitted through the hands of hospital personnel or through handling contaminated objects.19 Isolation of patients and the use of gloves can reduce spread of the infection, but early identification of infected patients is important so that limiting their personal contacts can help to control spread.20 Arch Phys Med Rehabil Vol 87, August 2006
1088
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak Table 1: Comparison of Patient Characteristics With and Without C. Difficile Colonization at Admission C. Difficile Culture at Admission
Negative
Positive
No. of patients Primary rehabilitation diagnosis SCI Brain injury SCI and brain injury Other Average LOS in acute care ⫾ SD (d) Sex Female Male Age at admission ⫾ SD (y) Infections in acute care Urinary tract infection Pneumonia Wound History of neurogenic bowel dysfunction Surgery in acute care Medication use Antibiotics at admission H2 blocker at admission PPI at admission Diarrhea at transfer Diarrhea 24h after transfer
45 (83.3)
9 (16.6)
18 (40) 12 (27) 4 (9) 11 (24) 30.44⫾20.96
4 (44) 2 (22) 0 (0) 3 (33) 29.44⫾18.05
16 (36) 29 (64) 46.98⫾17.90
4 (44) 5 (56) 52.11⫾23.18
Statistical Test
12 test⫽.061, NS 12 test⫽.077, NS 12 test⫽.864, NS 12 test⫽.254, NS t52⫽.133, NS 12 test⫽.254, NS
t52⫽⫺.748, NS
14 (31) 17 (38) 4 (9) 26 (58) 33 (73)
3 (33) 1 (11) 1 (11) 6 (67) 7 (78)
12 test⫽.017, NS 12 test⫽2.66, NS 12 test⫽.044, NS 12 test⫽.245, NS 12 test⫽.077, NS
18 (40) 13 (29) 19 (42) 6 (13) 11 (24)
4 (44) 2 (22) 6 (67) 1 (11) 2 (25)
12 test⫽.061, NS 12 test⫽.166, NS 12 test⫽1.80, NS 12 test⫽.033, NS 12 test⫽.001, NS
NOTE. Values are n (%) unless otherwise indicated. Abbreviation: SD, standard deviation.
C. difficile can be acquired asymptomatically within hospitals and transmitted to other patients.9 After colonization is identified, treatment is not recommended because metronidzole and vancomycin do not reliably eradicate spore carriage in asymptomatic carriers.3 Although the patients identified in our study were not treated unless they became symptomatic, precautions were taken to prevent spread of the organism immediately after their positive cultures were reported. The U.S. Center for Disease Control and Prevention (CDC) recommends that patients with C. difficile disease be restricted to a private room or cohorted with patients with the same organism. Dedicated equipment is also recommended. For hand hygiene, handwashing with soap and water, as opposed to alcohol gel, is believed to be more effective because of C. difficile spores’ resistance to the alcohol gels. Hypochloritebased disinfectant should be used to clean environmental surfaces.21 In an inpatient rehabilitation setting, patients are often treated in common areas; this could potentially result in patient-to-patient contact and organism transmission. A previous study at our institution, however, found that while surfaces within patients’ rooms were frequently colonized with a drugresistant organism found on stool culture, common areas were much less likely to be contaminated, even when patients with this organism frequented these areas.22 Thus, although our institution does follow CDC guidelines regarding private room assignment for patients known to be harboring toxigenic C. difficile, they are not restricted to their rooms for therapy unless they are incontinent or unregulated in their bowel regimens. Therapy equipment is cleaned after treatment with the same recommended solutions and therapists always wear gloves. Because fewer patients on these units with symptomatic disease were noted during this study, when isolation was employed on identification of C. difficile colonization, earlier identification of colonized patients may be more important. Arch Phys Med Rehabil Vol 87, August 2006
Testing for C. difficile can be done through culture of the organism as well as detection of the cytotoxin.23 Cultures generally take at least 3 days to produce results and toxins may degrade during transport and become inactive.20 Strategies that use both toxin and cultures should result in increased yield. Toxin immunoassays may have a false-negative rate of from 10% to 20% and thus sensitivity is improved by sending a second stool sample on the next day.20 Stool cultures, however, detect nontoxigenic strains. Sensitivities of the anaerobic culture technique we use are approximately 95%. The decreased sensitivity of the toxin assay did not affect our ability to identify colonization at admission, because all culture-positive patients were positive on the initial cytotoxic assay. An important risk factor for development of C. difficile diarrhea is antibiotic exposure.10 Antibiotic-associated alterations in flora result in overgrowth of C. difficile. In the gut, antibiotic associated C. difficile disease may occur after a few days or up to 2 months after drug exposure. Though clindamycin is the antibiotic most frequently implicated, due to resistant strains of C. difficile, other agents such as second and third generation cephalosporins, penicillins, and floroquinolones, may also produce symptomatic disease.24-26 Other related predisposing factors include duration of antibiotic use, with treatment courses of fewer than 3 days being less associated than longer treatment.10 We were unable to determine whether antibiotic use during rehabilitation was associated with C. difficile colonization. Also, because of the low frequency of specific antibiotic types in use at admission, we could not evaluate the effects of different antibiotic types. The frequency of infections while in acute care was extremely high in this population, however, with most patients having antibiotic exposure at some point during their acute care stay. Because information was not available in all cases about specific antibiotics used during acute care, we could not study preadmission use of specific types of antibiotics as a risk factor.
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
Among patients in inpatient acute rehabilitation, those with SCI have a significantly higher rate of nosocomial infection, including C. difficile.1 LOS appears to influence infection rate; 21% of patients in 1 study developed C. difficile after admission at a median of 12 days after admission. Incidence increases significantly 4 weeks after entering the hospital.11 Our patients generally had acute care LOS that exceeded this parameter. Advanced age has also been noted to be a risk factor; in a Swedish study, persons older than 60 years were 20 to 100 times more likely to have positive assays for toxin.2 In contrast to these reported associations, we did not find that age or the diagnosis of SCI, compared with a brain injury diagnosis, were predisposing factors, but did confirm the prior finding that SCI subjects receiving acute care are commonly colonized. Use of PPIs increases the risk of developing C. difficile infection.27 Concurrent antibiotic use, cytotoxic chemotherapy, and PPIs result in a much greater risk, with an odds ratio (OR) of 43.2.27 A recent Canadian cohort and case control study28 found adjusted ORs of 2.1 and 2.7 with the use of PPIs. Although we considered these medication factors on admission, we were unable to demonstrate an increased risk of colonization with the current use of antibiotics or acid-suppressive agents, including PPIs. This study was limited in that information on the acute care stay was obtained from the rehabilitation admission history and physical examination and thus may not have identified all acute care treatments. In addition, as mentioned previously, the use of a single swab likely resulted in our underestimating the colonization rate. Also, there was not a standard definition for “loose” or “liquid” stools as documented by the nursing staff, however, this nursing documentation is what is used clinically in the decision to begin laboratory evaluation of C. difficile infection. Finally, this study was conducted on units where the patients tended to have higher acute care LOS before hospitalization for rehabilitation, and therefore may not be representative of other types of rehabilitation patients, for example, those with orthopedic conditions. At our facility, a C. difficile culture costs $25 and a toxin assay costs $18. Although it was not considered in this study, it may be cost effective to screen patients with admission characteristics that put them at high risk, because infections adversely affect attendance at therapy sessions.7 Such screening would include the patient groups confirmed in this study to be frequently colonized. We were, unfortunately, unable to distinguish subgroups of people within these rehabilitation populations of lower risk. It would be important to document in further studies the cost effectiveness of screening strategies, and to determine the potentially different rates of colonization in other inpatient rehabilitation populations that are less likely to have multiple risk factors, for example, orthopedic patients. CONCLUSIONS Patients admitted to acute rehabilitation may be colonized with asymptomatic intestinal C. difficile. While previously identified risk factors were not predictive of colonization in this study, most patients had more than 1 factor that may have increased their susceptibility to colonization. Inadvertent transfer of the organism within the rehabilitation unit may occur if this colonization is not recognized. References 1. Mylotte JM, Graham R, Kahler L, Young BL, Goodnough S. Impact of nosocomial infection on length of stay and functional improvement among patients admitted to an acute rehabilitation unit. Infect Control Hosp Epidemiol 2001;22:83-7.
1089
2. Karlstrom O, Fryklund B, Tullus K, Burman LG. A prospective nationwide study of Clostridium difficile-associated diarrhea in Sweden. The Swedish C. difficile Study Group. Clin Infect Dis 1998;26:141-5. 3. Fekety R. Guidelines for the diagnosis and management of Clostridium difficile-associated diarrhea and colitis. American College of Gastroenterology, Practice Parameters Committee. Am J Gastroenterol 1997;92:739-50. 4. Kimmey MB, Elmer GW, Surawicz CM, McFarland LV. Prevention of further recurrences of Clostridium difficile colitis with Saccharomyces boulardii. Dig Dis Sci 1990;35:897-901. 5. Kyne L, Farrell RJ, Kelly CP. Clostridium difficile. Gastroenterol Clin North Am 2001;30:753-77, ix-x. 6. Kyne L, Sougioultzis S, McFarland LV, Kelly CP. Underlying disease severity as a major risk factor for nosocomial Clostridium difficile diarrhea. Infect Control Hosp Epidemiol 2002;23: 653-9. 7. Yablon SA, Krotenberg R, Fruhmann K. Diarrhea in hospitalized patients. Am J Phys Med Rehabil 1992;71:102-7. 8. Dallal RM, Harbrecht BG, Boujoukas AJ, et al. Fulminant Clostridium difficile: an underappreciated and increasing cause of death and complications. Ann Surg 2002;235:363-72. 9. McFarland LV, Mulligan ME, Kwok RY, Stamm WE. Nosocomial acquisition of Clostridium difficile infection. N Engl J Med 1989;320:204-10. 10. Wistrom J, Norrby SR, Myhre EB, et al. Frequency of antibiotic-associated diarrhoea in 2462 antibiotic-treated hospitalized patients: a prospective study. J Antimicrob Chemother 2001;47:43-50. 11. Clabots CR, Johnson S, Olson MM, Peterson LR, Gerding DN. Acquisition of Clostridium difficile by hospitalized patients: evidence for colonized new admissions as a source of infection. J Infect Dis 1992;166:561-7. 12. Alcantara C, Stenson WF, Steiner TS, Guerrant RL. Role of inducible cyclooxygenase and prostaglandins in Clostridium difficile toxin A-induced secretion and inflammation in an animal model. J Infect Dis 2001;184:648-52. 13. Kelly CP, Pothoulakis C, LaMont JT. Clostridium difficile colitis. N Engl J Med 1994;330:257-62. 14. Sakurai T, Hajiro K, Takakuwa H, Nishi A, Aihara M, Chiba T. Liver abscess caused by Clostridium difficile. Scand J Infect Dis 2001;33:69-70. 15. Freiler JF, Durning SJ, Ender PT. Clostridium difficile small bowel enteritis occurring after total colectomy. Clin Infect Dis 2001;33:1429-31; discussion 32. 16. Jacobs A, Barnard K, Fishel R, Gradon JD. Extracolonic manifestations of Clostridium difficile infections. Presentation of 2 cases and review of the literature. Medicine (Baltimore) 2001;80: 88-101. 17. Garcia-Lechuz JM, Hernangomez S, Juan RS, Pelaez T, Alcala L, Bouza E. Extra-intestinal infections caused by Clostridium difficile. Clin Microbiol Infect 2001;7:453-7. 18. Yablon SA, Krotenberg R, Fruhmann K. Clostridium difficilerelated disease: evaluation and prevalence among inpatients with diarrhea in two freestanding rehabilitation hospitals. Arch Phys Med Rehabil 1993;74:9-13. 19. McFarland LV, Surawicz CM, Stamm WE. Risk factors for Clostridium difficile carriage and C. difficile-associated diarrhea in a cohort of hospitalized patients. J Infect Dis 1990;162: 678-84. 20. Bartlett JG. Clinical practice. Antibiotic-associated diarrhea. N Engl J Med 2002;346:334-9. 21. Centers for Disease Control. Clostridium difficile: information for healthcare providers. August 2004, updated July 22, 2005. Available at: http://www.cdc.gov/ncidod/dhqp/id_CdiffFAQ_HCP.html#5. Accessed March 12, 2006. Arch Phys Med Rehabil Vol 87, August 2006
1090
CLOSTRIDIUM DIFFICILE COLONIZATION, Marciniak
22. Trick WE, Temple RS, Chen D, Wright MO, Solomon SL, Peterson LR. Patient colonization and environmental contamination by vancomycin-resistant enterococci in a rehabilitation facility. Arch Phys Med Rehabil 2002;83:899-902. 23. Fekety R, Shah AB. Diagnosis and treatment of Clostridium difficile colitis. JAMA 1993;269:71-5. 24. McCusker ME, Harris AD, Perencevich E, Roghmann MC. Fluoroquinolone use and Clostridium difficile-associated diarrhea. Emerg Infect Dis 2003;9:730-3. 25. Bates CJ, Wilcox MH, Spencer RC, Harris DM. Ciprofloxacin and Clostridium difficile infection. Lancet 1990;336:1193.
Arch Phys Med Rehabil Vol 87, August 2006
26. Climo MW, Israel DS, Wong ES, Williams D, Coudron P, Markowitz SM. Hospital-wide restriction of clindamycin: effect on the incidence of Clostridium difficile-associated diarrhea and cost. Ann Intern Med 1998;128(12 Pt 1):989-95. 27. Cunningham R, Dale B, Undy B, Gaunt N. Proton pump inhibitors as a risk factor for Clostridium difficile diarrhoea. J Hosp Infect 2003;54:243-5. 28. Dial S, Alrasadi K, Manoukian C, Huang A, Menzies D. Risk of Clostridium difficile diarrhea among hospital inpatients prescribed proton pump inhibitors: cohort and case-control studies. CMAJ 2004;171:33-8.