Reference: Klaas Heyland, M.F., Patrick Buehr, and Christian P. Braegger, No Advantage for Antibiotic Treatment Over Placebo in Blastocystis hominis–positive Children With Recurrent Abdominal Pain.
Journal of Pediatric Gastroenterology and Nutrition, 2012. 54(5): p. 677–679.
Brief Summary: Blastocystis hominis (B. hominis) is a parasite sometimes found in the stools of healthy people and those with diarrhea which can cause abdominal pain or other gastrointestinal problems. The main objective of this study was to determine if the antiobiotic trimethoprim-sulfamethoxazole (TMP/SMX) can be used to effectively treat children with recurrent abdominal pain (RAP) who are also positive for Blastocystis hominis. Forty patients were recruited for the investigation and randomly assigned to either the TMP/SMX group or the placebo group. The main outcome measure was the difference in Pain Index (PI) from initial treatment with TMP/SMX compared to 7 days later.
Results: There was no significant reduction in PI between the treatment and placebo group. A treatment with Metronidazole did reduce the PI index.
Implications for Practice: Do not prescribe the the antiobiotic TMP/SMX for children with recurrent abdominal pain who are positive for Blastocystis hominis.
Discussion: Negative results are just as important as positive results, so I am glad to see that the author’s decided to move on with a publication on this. I am further curious as to why the TMP/SMX anti-biotic did not work from a biological perspective. There are some signaling pathway databases that could be used to look this up; this could be an interesting follow up study perhaps.
At the end, the author’s suggested that since the antibiotic did not work, this is evidence that B. hominis has no causative role in recurrent abdominal pain (RAP). It seems as if an observational study could easily verify this and would be interesting to follow up with.
Since this study showed that the antibiotic TMP/SMX does not work, it would be interesting to see if the alternative drug – metronidazole – does significantly work vs. a placebo.
Commentary on Statistics and Study Design: Overall, the study design looked good. It was good to see that the investigators under-took a double blind study. Also, I liked how the author’s included a table on the demographic properties of the placebo vs. the treatment group. This is a really good idea, because it shows – from an actual visual standpoint - that potential confounding factors were balanced between the treatment and placebo groups. I’m assuming the authors did an exhaustive search of any potential confounding factors here.
The author’s had to exclude any patients from the investigation with a past history of a gastrointestinal or systematic disease. When you have a potential confounding factor which could severely hinder the investigation such as these ones, excluding subjects with these criteria is probably the best possible option, because it is obviously not possible to really block on these types of confounding factors – especially if the sample size is small.
The statistical technique used was the Mann-Whittney U Test (also known as the Wilcoxan Rank Sum Test), and this is an appropriate technique to use since – with a small sample size – it may not be feasible to assume a normal distribution.
A big thanks to our friends in Switzerland for doing this investigation!
No comments:
Post a Comment