|Eugene Y. H. Yeung*, Jason G. Gore, Edward V. Auersperg
Ridge Meadows Hospital, 11666 Laity Street, Maple Ridge, British Columbia, Canada V7X 2G5
|Corresponding Author: Eugene Y. H. Yeung Email: [email protected]|
|Related article at Pubmed, Scholar Google|
Introduction: Meropenem and piperacillin-tazobactam have similar indications, spectrum of antimicrobial actions, and cost. When selecting between two antibiotics with similar efficacy, one may want the antibiotic with the least harm, such as C. difficile associated diarrhea (CDAD).
Aim & Objectives: This study tested the hypothesis that patients on piperacillintazobactam had a lower incidence of CDAD than patients on meropenem.
Methods: A retrospective analysis was performed that included patients who received meropenem or piperacillin-tazobactam during their admissions to Ridge Meadows Hospital, Canada from September 2007 to August 2009. This study had a subgroup analysis on patients with risk factors of developing CDAD: male, over 65 years old, staying longer than 28 days in healthcare settings, receiving concurrent risk factor medications, not on Saccharomyces boulardii in the previous two months, and not on oral metronidazole, intravenous metronidazole, or oral vancomycin in the previous two months.
Results: There were 168 patients in the meropenem group and 122 patients in the piperacillin-tazobactam group. No significant difference was found between meropenem and piperacillin-tazobactam with respect to incidence of CDAD (3.57% and 4.92%, respectively; p=0.5676), two-month in-hospital mortality (34.52% and 36.89%, respectively; p=0.7102), and composite outcome of CDAD and two-month in-hospital mortality (37.50% and 40.16%, respectively; p=0.7142). All subgroup analyses showed no difference in incidence of CDAD between the two antibiotics.
Conclusions: There was no evidence to support patients on piperacillin-tazobactam had a lower incidence of CDAD or mortality than patients on meropenem. Further research is still needed to help selecting the safest antibiotic for patients.
|meropenem, piperacillin-tazobactam, Clostridium difficile, diarrhea, mortality, Saccharomyces boulardii|
|Clostridium difficile is the most clinically relevant bacteria known to cause antibioticassociated diarrhea and is responsible for 15-25% of cases of antibiotic-associated diarrhea.1 The incidence of C. difficile associated diarrhea (CDAD) in hospitalized patients is approximately 8% and accounts for 20-30% of cases of hospital-acquired diarrhea.2|
|Meropenem and piperacillin-tazobactam are time-dependent killing, β-lactamaseresistant antibiotics that are active against gram-negative aerobes, gram-positive aerobes, and anaerobic bacteria.3,4 Both meropenem and piperacillin-tazobactam are indicated for treatment of lower respiratory tract, intra-abdominal, gynecologic, and skin infections.5,6 Both drugs have similar costs, and review of their respective drug monographs suggests that adverse effect profiles are very similar.5,6 Because both antibiotics have similar indications and adverse effect profiles, selecting one antibiotic over the other can be difficult. When selecting between two antibiotics with similar efficacy, one may want the antibiotic with the least harm, such as CDAD. Knowing the prevalence of CDAD with meropenem and piperacillin-tazobactam may help health institutions in reducing incidence of antibiotic associated diarrhea.|
|It has been demonstrated that piperacillin-tazobactam inhibits growth and toxin production of C. difficile.7 In two institutions, CDAD rates increased by 200% or more during a shortage of piperacillin-tazobactam.8,9 However, piperacillintazobactam has also been associated with CDAD. In one institution, the incidence of CDAD decreased by 47% during the same piperacillin-tazobactam shortage period.10 However, there is currently no information on the prevalence of CDAD in subjects exposed to meropenem.|
|The present study was designed to test the hypothesis that patients on piperacillintazobactam had a lower incidence of CDAD than patients on meropenem. This study also compared the incidence of in-hospital mortality within two months of meropenem or piperacillin-tazobactam use.|
|This study was a retrospective analysis of all patients who received meropenem or piperacillin-tazobactam during their admissions to Ridge Meadows Hospital, Canada from September 2007 to August 2009. There was no protocol in place in choosing between the two antibiotics. The data was obtained using the health record software Meditech Version 3.26 (Westwood, Massachusetts, USA) and patients’ chart records in the hospital. CDAD was defined as having more than 2 documented unformed or watery stools in 24 hours plus a positive assay for toxin B in a stool sample.2,11 The episode of CDAD was attributed to the use of meropenem or piperacillin-tazobactam if the C. difficile toxin was found during, or within two months of, the antibiotic therapy.|
|The following patients were excluded from the study: patients under 18 years of age, patients who received both piperacillin-tazobactam and meropenem in a two-month period, patients who developed CDAD in the two months prior to the use of meropenem or piperacillin-tazobactam, and patients with gastrointestinal tract colonization of pathogens also known to cause infectious diarrhea (Figure 1).|
|The primary outcome of this study was the incidence of CDAD. The secondary outcomes were two-month in-hospital mortality, and composite outcome of CDAD or two-month mortality. This study also looked at a subgroup of patients with risk factors for developing CDAD: male gender12,13, above 65 years of age at the end of meropenem or piperacillin-tazobactam therapy14, prolonged (> 28 days) stay in health-care settings in the previous 6 years15, use of risk factor medications that were given within two months of meropenem or piperacillin-tazobactam (antineoplastic agents16, tacrolimus17, histamine 2-receptor antagonists18, proton pump inhibitors18, non-steroidal anti-inflammatory drugs19, and systemic antibiotics except vancomycin and metronidazole), and no exposure to oral Saccharomyces boulardii, oral vancomycin, oral metronidazole, or intravenous metronidazole within two months prior to meropenem or piperacillin-tazobactam use.|
|The incidence of CDAD with piperacillin-tazobactam was predicted to be 7%. This was based on a study that compared incidence of CDAD in patients on piperacillintazobactam and cefotaxime (7% vs. 53%).20 There is no published data on the incidence of CDAD with meropenem. We estimated that the CDAD incidence rate of meropenem was between that of the piperacillin-tazobactam and cefotaxime (7% vs. 53%). Because meropenem is active against the clinical isolates of C. difficile21, we did not foresee that its incidence of CDAD could be as high as 53%. The incidence of CDAD with meropenem was arbitrarily set at 22%, which was 15% higher than the estimated piperacillin-tazobactam rate (7%). A sample size calculation with dichotomous outcome variables, a two-tailed alpha of 5%, and a beta of 80% was used.22 A minimum of 86 patients was needed per group to detect a difference of 15% between the two groups.|
|The two-sided Fisher’s exact test was used to compare data from two treatment groups. The criterion of statistical significance was set at p < 0.05. Statistical tests were performed using the computer software GraphPad Prism version 4.00 for Windows, (GraphPad Software, San Diego, California USA).|
|Medical records were reviewed for a total of 333 patients (Figure 1). Thirty-five patients in the meropenem group and 32 patients in the piperacillin-tazobactam group were excluded. There were a total of 168 meropenem-treated patients and 122 piperacillin-tazobactam-treated patients included in the analysis. The sample size was sufficient to detect an effect difference of about 12% between the two groups. The demographic data showed that the piperacillin-tazobactam group had a higher percentage of males than the meropenem group (Figure 2; 61.48% and 47.62%, respectively; p = 0.0235).|
|Table 2 showed no significant difference between meropenem and piperacillintazobactam groups with respect to incidence of CDAD (3.57% and 4.92%, respectively; p = 0.5676), two-month in-hospital mortality (34.52% and 36.89%, respectively; p = 0.7102), and composite outcome of CDAD and two-month inhospital mortality (37.50% and 40.16%, respectively). In each subgroup analysis in Table 3, the piperacillin-tazobactam group had a slightly higher CDAD incidence than the meropenem group. However, the difference did not reach statistical significance.|
|Overall, the present study showed that piperacillin-tazobactam did not have a significantly lower incidence of CDAD than meropenem (Table 2). The results should not warrant the use of one antibiotic over the other to reduce the incidence of CDAD. The two-month in-hospital mortality rate was also not significantly different between the meropenem and piperacillin-tazobactam groups (Table 2). The lack of significance was speculated to be due to a lack of statistical power. The composite outcome of incidence of CDAD or two-month in-hospital mortality was measured to improve the statistical power. The results should not warrant the use of one antibiotic over the other to reduce C. difficile related mortality. It is important to note that the severity of diseases could be different between the two antibiotic groups, and thereby offset the difference in mortality rates, though this possibility was not explored in this study.|
|Factors that increased the risk of developing CDAD included the following: advanced age, prolonged duration of stay in healthcare settings, concurrent risk factor medications, and no previous exposure to oral S. boulardii, oral vancomycin, oral metronidazole, or intravenous metronidazole. The demographic data showed no significant difference in the above noted risk factors between the meropenem and piperacillin-tazobactam groups (Table 1). It suggested that the lack of difference of CDAD rates between the two groups was unlikely due to the confounding variables. However, the meropenem group had significantly more male patients than the piperacillin-tazobactam group. Two studies found that male patients were more prone to have CDAD than female patients, although the mechanism is still unclear.12,13 In contrast, two studies showed female patients were more prone to have CDAD than male patients.23,24 It is inconclusive therefore, whether gender is a risk factor for CDAD.|
|In any events, the subgroup analysis showed no significant difference in incidence of CDAD between the meropenem and piperacillin-tazobactam groups in male patients (Table 3). The subgroup analysis also showed no significant difference in incidence of CDAD between the two antibiotics in patients at higher risk of developing CDAD. It suggested that selecting piperacillin-tazobactam over meropenem did not reduce the incidence of CDAD in high-risk patients.|
|It is interesting to note that the incidence of CDAD in the piperacillin-tazobactam group was lower than the rate reported in the study by Settle et al (4.92% and 7%, respectively).20 Differences in settings, demographics, and study designs might have contributed to the difference in the incidence. It is important to note that the study by Settle et al. was conducted in geriatric wards. In the piperacillin-tazobactam group in the current study, the incidence of CDAD in elderly patients was 6.67% (Table 3), which was comparable to the rate in the study by Settle et al.|
|The present study had its limitations. The first limitation was that there was a relatively small sample size (a total of 290 patients after exclusion), and thus lacked statistical power to detect small differences between the meropenem and piperacillintazobactam groups. The absolute difference in incidence of CDAD between the two groups was 1.35% (Table 2), which was much smaller than the estimated difference used in the sample size calculation (15%). It appeared that the current study overestimated the difference between the incidence of CDAD with meropenem and piperacillin-tazobactam. To detect an absolute difference of 1.35%, a sample size of 3500 patients would be needed per group. This would require the patient data in the hospital in the last 40 years. The hospital lacks the available data.|
|The second limitation was that the present study was performed at a single centre. We had no access to patient medical information prior to admission. Moreover, this study was retrospective, and thus could not eliminate all confounders. For example, it did not screen each patient for all possible pathogens for infectious diarrhea. Nasogastric tube feeding25, peri-partum (4 weeks before to 4 weeks after delivery)11, serious underlying illness-comorbidities, immune-compromising conditions, and gastrointestinal surgery and disorders have also been related to development of CDAD.26 Therefore, the incidence of CDAD in this study might not be exclusively related to meropenem or piperacillin-tazobactam usage alone. Despite that, the objective of this study was to investigate incidence of antibiotic associated diarrhea but not diarrhea caused by antibiotics. Increase in incidence of CDAD might still suggest an indirect relationship with meropenem or piperacillin-tazobactam.|
|The third limitation was that the present study used C. difficile toxin B test and the number of unformed or watery stool to diagnose CDAD. The toxin test was reported to be only 70-80% sensitive.27 Other published diagnostic methods include positive stool culture for C. difficile, positive assay for toxin A in a stool sample, characteristics of C. difficile infection on colonic biopsy, and pseudomembranous colitis observed on lower gastrointestinal endoscopy.27 In addition, certain diets and medications might have caused constipation and masked the number of loose stools per day, which is important for the diagnosis of CDAD. Therefore, the incidence of CDAD observed in the current study could be an underestimate.|
|Based on the information in the current study, there was not enough evidence to support that piperacillin-tazobactam had a lower incidence of CDAD than meropenem. The study lacked evidence to suggest a significant difference in the incidence of in-hospital mortality. The study also failed to show significant difference in incidence of CDAD between the two antibiotics in high-risk patients. The results should not warrant the use of one antibiotic over another to prevent CDAD. Further research, involving larger sample size and prospective study design, is still needed to help selecting the safest antibiotic for patients. Knowing the prevalence of CDAD can help health institutions in establishing a protocol in choosing between the two antibiotics.|
|Using the current incidence rate of CDAD in the meropenem and piperacillintazobactam groups (3.57% and 4.92%, respectively), researchers can perform a prospective randomized controlled trial to confirm the results of the current study. A sample size calculation with dichotomous outcome variables, a two-tailed alpha of 5%, and a beta of 80% can be used.22 A minimum of 3500 patients would be needed per group to detect an effect size difference of 1.35%. In addition, the new trial should be a multi-centred study with a sensitive diagnostic test and exclusion of confounders described above. It is necessary to reduce the incidence of this life-threatening diarrhea, which is a huge financial burden for the healthcare system.|
|The authors would like to thank Michael Wasdell for his assistance in statistical analysis, Gary Arnold for the retrieval of health records, and the Fraser Health Pharmacy Residency Program Research Committee for their feedback on the research design.|
|None of the authors has any competing interests|
|E.Y.H.Y was the lead researcher in this study, responsible for data collection and analysis and preparation of the manuscript. J.G.G. and E.V.A. are the supervisors in the study, responsible for revision of the study design and editing of the manuscript.|
|1. Fekety R. Guidelines for the diagnosis and management of clostridium difficileassociated diarrhea and colitis. american college of gastroenterology, practice parameters committee. Am J Gastroenterol. 1997;92(5):739-750.
2. Lentino JR. Merck manual: Clostridium difficile–Induced diarrhea [online]. https://www.merck.com/mmpe/sec14/ch178/ch178d.html. Updated 2009 Aug 1. Accessed Dec 20, 2010.
3. Lowe MN, Lamb HM. Meropenem: An updated review of its use in the management of intra-abdominal infections. Drugs. 2000;60(3):619-646.
4. Gin A, Dilay L, Karlowsky JA, Walkty A, Rubinstein E, Zhanel GG. Piperacillintazobactam: A beta-lactam/beta-lactamase inhibitor combination. Expert Rev Anti Infect Ther. 2007;5(3):365-383.
5. Compendium of Pharmaceuticals and Specialties, ed. Merrem product monograph. Mississauga (ON): AstraZeneca Inc; 2007 Feb 27.
6. Compendium of Pharmaceuticals and Specialties, ed. Tazocin product monograph. Markham (ON): Wyeth Canada; 2008 Aug 20.
7. Baines SD, Freeman J, Wilcox MH. Effects of piperacillin/tazobactam on clostridium difficile growth and toxin production in a human gut model. J Antimicrob Chemother. 2005;55(6):974-982.
8. Wilcox MH, Freeman J, Fawley W, et al. Long-term surveillance of cefotaxime and piperacillin-tazobactam prescribing and incidence of clostridium difficile diarrhoea. J Antimicrob Chemother. 2004;54(1):168-172.
9. Alston WK, Ahern JW. Increase in the rate of nosocomial clostridium difficileassociated diarrhoea during shortages of piperacillin-tazobactam and piperacillin. J Antimicrob Chemother. 2004;53(3):549-550.
10. Mendez MN, Gibbs L, Jacobs RA, McCulloch CE, Winston L, Guglielmo BJ. Impact of a piperacillin-tazobactam shortage on antimicrobial prescribing and the rate of vancomycin-resistant enterococci and clostridium difficile infections. Pharmacotherapy. 2006;26(1):61-67.
11. Centers for Disease Control and Prevention (CDC). Severe clostridium difficileassociated disease in populations previously at low risk--four states, 2005. MMWR Morb Mortal Wkly Rep. 2005;54(47):1201-1205.
12. Koh TH, Tan AL, Tan ML, Wang G, Song KP. Epidemiology of clostridium difficile infection in a large teaching hospital in singapore. Pathology. 2007;39(4):438-442.
13. Manian FA, Aradhyula S, Greisnauer S, et al. Is it clostridium difficile infection or something else? A case-control study of 352 hospitalized patients with new-onset diarrhea. South Med J. 2007;100(8):782-786.
14. Pituch H. Clostridium difficile is no longer just a nosocomial infection or an infection of adults. Int J Antimicrob Agents. 2009;33 Suppl 1:S42-S45.
15. Buchner AM, Sonnenberg A. Epidemiology of clostridium difficile infection in a large population of hospitalized US military veterans. Dig Dis Sci. 2002;47(1):201-207.
16. Fekety R, McFarland LV, Surawicz CM, Greenberg RN, Elmer GW, Mulligan ME. Recurrent clostridium difficile diarrhea: Characteristics of and risk factors for patients enrolled in a prospective, randomized, double-blinded trial. Clin Infect Dis. 1997;24(3):324-333.
17. Sharma AK, Holder FE. Clostridium difficile diarrhea after use of tacrolimus following renal transplantation. Clin Infect Dis. 1998;27(6):1540-1541.
18. Dial S, Delaney JA, Barkun AN, Suissa S. Use of gastric acid-suppressive agents and the risk of community-acquired clostridium difficile-associated disease. JAMA. 2005;294(23):2989-2995.
19. Dial S, Kezouh A, Dascal A, Barkun A, Suissa S. Patterns of antibiotic use and risk of hospital admission because of clostridium difficile infection. CMAJ. 2008;179(8):767-772.
20. Settle CD, Wilcox MH, Fawley WN, Corrado OJ, Hawkey PM. Prospective study of the risk of clostridium difficile diarrhoea in elderly patients following treatment with cefotaxime or piperacillin-tazobactam. Aliment Pharmacol Ther. 1998;12(12):1217-1223.
21. Jamal WY, Mokaddas EM, Verghese TL, Rotimi VO. In vitro activity of 15 antimicrobial agents against clinical isolates of clostridium difficile in kuwait. Int J Antimicrob Agents. 2002;20(4):270-274.
22. Stolley PD, Strom BL. Sample size calculations for clinical pharmacology studies. Clin Pharmacol Ther. 1986;39(5):489-490.
23. Asha NJ, Tompkins D, Wilcox MH. Comparative analysis of prevalence, risk factors, and molecular epidemiology of antibiotic-associated diarrhea due to clostridium difficile, clostridium perfringens, and staphylococcus aureus . J Clin Microbiol. 2006;44(8):2785-2791.
24. Palmore TN, Sohn S, Malak SF, Eagan J, Sepkowitz KA. Risk factors for acquisition of clostridium difficile-associated diarrhea among outpatients at a cancer hospital. Infect Control Hosp Epidemiol. 2005;26(8):680-684.
25. Surawicz CM, Elmer GW, Speelman P, McFarland LV, Chinn J, van Belle G. Prevention of antibiotic-associated diarrhea by saccharomyces boulardii: A prospective study. Gastroenterology. 1989;96(4):981-988.
26. Hookman P, Barkin JS. Clostridium difficile associated infection, diarrhea and colitis. World J Gastroenterol. 2009;15(13):1554-1580.
27. Peterson LR, Robicsek A. Does my patient have clostridium difficile infection? Ann Intern Med. 2009;151(3):176-179.
|Figure 1: Flow diagram of patient selection.|
|Table 1: Demographic Characteristics of Patients Treated with Meropenem and Piperacillin-Tazobactam|
|Table 2: Analysis of Clostridium difficile Associated Diarrhea Incidence and Mortality Rates|
|Table 3: Subgroup Analysis of Patients with Risk Factors for Clostridium difficile Associated Diarrhea|