Clinical Study Designs for Antibiotic-Associated Diarrhea: Moving Forward

IPA AdminIPA In The Lab

Josh Baisley, B.Sc., Associate Director of Clinical Trials, Nutrasource

Recent clinical research in the area of probiotics has led to advancements which have helped shape the future of probiotic clinical trials. One area that has presented substantial learning opportunities in the field is the study of antibiotic associated diarrhea (AAD) and probiotic intake.
AAD is a common side effect of many antibiotics, especially those that affect anaerobic bacteria. AAD occurs in about 5-39% of patients taking antibiotics.1 Incidence varies based on the vulnerability of patients. Factors affecting patient vulnerability include age, duration of hospitalization, underlying medical conditions, the specific antibiotic treated with, as well as dose and treatment regimen. The clinical presentation of AAD, which ranges from mild diarrhea to fulminant pseudomembranous colitis (a severe form of colitis), can present within a week of starting the antibiotic up to two months after the end of treatment.2
Antibiotics disrupt the gut microbiota, disturbing the composition and function, which leads to an environment where overgrowth of toxigenic bacteria can occur which can induce diarrhea. Clostridium difficile is reported to cause 10-20% of all cases of AAD and perhaps as high as 30%.3 Other mechanisms exist including the loss of beneficial bacteria (and their metabolic activities), which leads to an accumulation of high molecular weight carbohydrates in the colon potentially resulting in osmotic diarrhea.4
In the case of AAD clinical research, probiotics are studied as an adjunct supplement or food product to prevent AAD. Research in this area has been ongoing for nearly 50 years studying the effects of various probiotics.5 A quick search of clincialtrials.gov yielded more than 60 listings of studies researching the use of probiotics with antibiotic therapy.
Over time, as researchers gain more knowledge, clinical study designs have evolved, with the earliest studies conducted in hospital settings.5 Science, and specifically clinical trial design, must adapt as we learn more about confounding factors as well as the makeup and variability of the gut microbiome. Difficulties in interpreting results of AAD studies have occurred due to several confounding factors, including the definition of diarrhea, course of antibiotic treatment, duration of treatment, concomitant medications, illness being treated or other concurrent illnesses, dietary changes (especially in hospital setting) and susceptibility of the probiotic to the antibiotics used in treatment, among many other variables.
Care must be taken with respect to any clinical trial design to understand potential confounding factors and steps taken to address these when writing the protocol. Whether it be limiting the type and dose of antibiotic therapy, choosing subjects under treatment for a specific common illness, or excluding subjects with specific confounding underlying medical conditions. Confounding factors may also be addressed through stratification as part of the randomization process to ensure balance between groups or through statistical modelling in the statistical analysis plan, ensuring important covariates are defined and included in the analytical plan prior to database lock and analysis.
Defining endpoints, how it/they will be collected, measured and analyzed is critical to study design. All endpoints under consideration should be disclosed in the protocol and treatment of these endpoints defined in the statistical analysis plan. In addition to comparing groups with respect to time of onset, duration, and/or presence of diarrhea, it may be prudent to also beyond diarrhea specifically and test other gastrointestinal complaints. If this approach is taken, the protocol must specify this and statistical analysis plan include how this will be scored and/or analyzed prior to analysis.
If the primary endpoint of the study design is the incidence rate of AAD (the percentage of subjects that will get AAD within the specified treatment period), the definition of diarrhea should be clearly defined and medically accepted, and methods for data collection/assessment and analysis need to be clearly stated. Further, the incidence rate of AAD with the specific antibiotic must be factored in the sample size calculation in order to mitigate the risk of equivocal results. Additional design aspects could be considered to minimize the required sample size such as using an enriched population. In the case of AAD, an enriched population could be a group of subjects known to get AAD when treated with a specific antibiotic or a group of subjects known to be more susceptible to AAD (e.g. subjects under the age of 6 and subjects over the age of 50). Choosing a group of subjects known to get AAD could be achieved through review of medical history. Another approach to defining an enriched population is to test subjects for response to a particular antibiotic. However, in the context of studies on AAD and the microbiome this may not be feasible for a number of reasons including duration of onset of AAD (up to 2 months after administration of the antibiotic), time for the microbiota to revert back to baseline levels, and possible ethical issues surrounding multiple courses of antibiotic therapy during a short period of time..
As more data is shared on these studies in peer reviewed publications and open discussions on clinical study limitations occur, new study designs will evolve incorporating best practices and lessons learned from previous research, leading to novel approaches in studying probiotics for their efficacy in AAD and other indications.
1 McFarland LV. Epidemiology, risk factors and treatments for antibiotic-associated diarrhea. Dig Dis 1998; 16:292-307.
2 Barbut F. Managing antibiotic associated diarrhoea. BMJ. 2002; 324(7350): 1345–1346.
3 Issa I and Moucari R. Probiotics for antibiotic-associated diarrhea: Do we have a verdict? World J Gastroenterol 2014; 20(47): 17788-17795.
4 Högenauer C, Hammer HF, Krejs GJ, Reisinger EC. Mechanisms and management of antibiotic associated diarrhea. Clin Infect Dis 1998; 27: 702-710.
5 D’Souza AL, Rajkumar C, Cooke J, Bulpitt CJ. Probiotics in prevention of antibiotic associated diarrhoea:meta-analysis. BMJ. 2002; 324: 1361-1366.