Each week our PhD students have their own Journal Club. Apparently they recently developed an interest in http://www.reflectionsipc and asked whether I was interested in getting their input. What a great idea, I thought, and here is the first one. Josephine van Dongen discussed a Scientific Reports paper on acute gastroenteritis (AGE) due to rotavirus. A global burden among children, but it can be prevented effectively by vaccination (still not recommended in our country!). Rotavirus discriminates: if your genes encode your cells to have a “secretor status” or being “Lewis positive” your infection risk increases (there even is a meta-analysis, for those still in doubt). And you are more likely to have these genotypes (at least secretor) if you’re Asian. Whether genotypes also predispose for more severe infection is unknown, and that’s what the study was about.
First thing to do in research is to define a specific research question. For this Taiwanese study that question was: is genotype associated with severity of infection? And while having fun they also aimed to asses vaccine effectiveness (question 2). They included 68 cases and 133 controls and found that – yes – different (previously established!) genetic factors were associated with rotavirus AGE (OR 28.6 for secretor status and OR 16.8 for Lewis). And the estimated vaccine effectiveness was 90.3% (95% CI 78.1-95.7%). Rotavirus immunization was recorded in 8 (12%) cases and 77 (58%) controls.
But what about the burning research question? Among the cases 13 had moderate and 55 had severe AGE. Obviously “ … two independent genetic factors (secretor and Lewis) are associated with increased risk of moderate-to-severe rotavirus AGE …”, which logically follows from only including patients with moderate-to-severe AGE and determining their genetic factors (you don’t need a control group for that). Surprisingly, though, was that these numbers were far too low to draw any conclusion on the association between genotype and AGE severity. Too bad, after all that work the question still unanswered.
When discussing the reasons for not answering the research question the authors state that the scoring system of AGE severity was not sensitive enough, but it is not clear for what. Apparently, it was not discriminative enough among children hospitalized for AGE. More to a surprise was ”that their study was not primarily designed to evaluate this (genes and severity) outcome”. That seems to validate the second rule in research: design your study in such a way that you can answer your research question.
Some PhD candidates felt that the study did answer the research question, if they “wildly interpreted” that question. The majority, though, still felt that life goes on without an answer (at least to this question).
So for fellow researchers, starting or long begun, all research starts with defining a (very specific) research question, followed by designing the study that answers the question, in order to prevent “wild interpretations”of your results.
Guest blogger this week: Fien van Dongen, who studies the effects of rotavirus vaccination in immunocompromised children