This is the first randomised placebo-controlled trial with an adequate sample size in a group of patients with SCI, which has evaluated the efficacy of LcS for the prevention of AAD and CDI. Although this specific probiotic drink could not prevent AAD in all SCI patients, finding suggests that taking LcS is associated with a lower risk of AAD in people with SCI who are on regular PPI therapy. The present study’s conclusion is different from our earlier openlabelled study. The discrepancy in the overall effect may result from the smaller size of the earlier underpowered study (158 vs 359 patients) or from differences in the study design. The present study defined AAD as 2 liquid stools (Bristol Stool Scale type 5, 6 or 7) for a 24-hour period, whereas the previous trial required 3 days. This altered definition may have led to a failure in distinction between clinically relevant AAD and loose stool due to neurogenic bowel as a result of SCI. Indeed, the definition of AAD varies widely between published studies. For example, Rajkumar et al. defined diarrhoea as 2 loose stools, Bristol 6 or 7 a day for 3 days, whereas Allen et al. and Helps et al. defined diarrhoea from 3 loose stools, Bristol 5, 6 or 7 in a single 24-hour period. The definition for CDI also varies. Using more standardised definitions of study outcomes in AAD/CDI is now improving the quality of and aiding interpretation of newer research studies, especially important for systematic reviews and meta-analyses. The inclusion of patients treated in different SCI centres could be considered a strength, but can also be regarded as a weakness. Infection control policies, AAD/CDI definition and antibiotic management were different in the participating SCI centres, thus the influence of these factors on the study results could not be excluded. The present study suggests malnutrition risk could increase the incidence of AAD in 30 days follow up (60.4% v 37.5%, p < 0.01). These results are in agreement with results of a similar published study. Previous research has suggested that consumption of oral nutritional supplements in patients with known malnutrition risk could improve clinical outcomes. Clinicians, therefore should be aware of the need to identify and treat malnutrition which will prevent malnutrition related complications. The present study’s incidence of AAD (32.1%) is similar to some previous studies in patients with SCI (14.9-36%) and seems higher than studies conducted in general populations (10.8 - 17.9%). This may be attributed to a longer follow-up period (30 days) than in many of the other published trials (often only 7-14 days); diarrhoea may occur up to 2 months after discontinuing antibiotic treatment. Current evidence remains unclear in whether probiotics could reduce the incidence of AAD/CDI in general hospitalised populations. The complexity of probiotic use is not just strain-, product-, dose- and disease specific, but also includes defining when the probiotic should be administered and the duration of its use; all need to be considered. The present study dose of a minimum of 6.5 x 109 CFU LcS was selected based on the previous trial’s data. LcS is well tolerated in clinical settings and has been used in a broad range of patients. However, dose and type of probiotic vary between published studies. For example, Allen et al. used a mixed strains probiotic (L. acidophilus CUL60, CUL21, B. bifidum CUL20, B. lactis CUL34 in 6 x 109 CFU /day), Helps et al. used a single strain probiotic (LcS in 13 x 109 CFU /day), and Rajkumar et al. used mixed strains (L. casei immunitas DN-114,001 in 10 x 109 CFU /day) as did Selinger et al (VSL#3 in 900 x 109 CFU /day). The compliance with LcS therapy in the present study was good (92.4%), with no known severe adverse events directly related to LcS. It is possible that a higher dose might have yielded a greater benefit. However, the effect of increasing dosage of probiotics should be monitored carefully as unexpected adverse events may also occur. Another important criterion for any probiotic is that the strains used survive the passage through the stomach and arrive in a viable state in the small intestine and colon. LcS has been shown to survive and be well tolerated in the upper gastrointestinal tract and reach the large intestine in a viable state. The present study does not provide support for a general use of LcS to prevent AAD in SCI patients but suggests that LcS may help to prevent AAD in SCI patients taking PPI regularly. Confirmatory studies are now needed to allow translation of this apparent therapeutic success into improved clinical outcomes. Further research should define the most appropriate target populations. Study designs using standardized definitions of AAD and CDI will help to evaluate the effect of probiotics in preventing AAD/CDI as well as in providing more homogeneous evidence for future systematic review and metaanalysis.