DOWNGRADING CERTAINTY IN EVIDENCE FOR PROBIOTIC MEDICINE IS PARTIALLY UNCORRECT
MAURIZIO KOCH, LUCIO CAPURSO
we read with a great interest the paper on AGA Clinical Practice Guidelines on the Role of Probiotics in the Management of Gastrointestinal Disorders by (1).
We agree that clinical research on probiotics lacks or avoids some basic rules of Evidence Based Medicine. Indeed, we have identified large methodological flaws in probiotic medicine (2).
The dream of microbiome-based medicine requires a fresh approach — an ecological and evolutionary understanding of host-microbe interactions (3). So, Guidelines are highly welcome.
However, let us comment on two clinical recommendations the paper offers.
The first is on the use of probiotics in the prevention of Clostridioides difficile infection. We think that the conclusion on “conditional recommendation, low quality of evidence” is particularly severe, and inappropriate.
Certainty can be rated up for:
- Large magnitude of effect
- Dose-response gradient
- All residual confounding would decrease magnitude of effect (in situations with an effect)
- The authors considered to downgrade the overall certainly of the evidence from Moderate to Low, as presented in the Cochrane review (5), due to unclear or high risk of bias in most of the trials across all domains for all outcomes assessed. May we suggest to give a look to the parallel meta-analysis by Nicole T. Shen and colleagues (6). They present a nice table of potential bias for each paper (their Table 2). The paper comes to the same clinical conclusions, but identifies high level of bias (H) only in 7 out of 19 papers, and only for one type of bias, the Incomplete outcome data (attrition bias). For the rest of the potential bias,
As stated in What is Grade (4) the reasons for rating certainty in evidence up or down are:
Certainty can be rated down for:
- Risk of bias
- Publication bias