Should we eradicate Helicobacter pylori before prescribing an NSAID? Result of a placebo-controlled study

(1) To determine the prevalence of gastrointestinal (GI) symptoms in patients with and without Helicobacter pylori infection and treated with non-steroidal anti-inflammatory drugs (NSAIDs) and (2) to estimate the impact of H. pylori eradication on these symptoms.

This was a multicentric, community-based, randomized, case-control study. Patients presenting with a rheumatic disorder motivating the prescription of an NSAID for at least 2 wks were stratified in two groups (H. pylori-positive and H. pylori-negative) by a serological doctor test and H. pylori-positive patients divided further into two subgroups, receiving either an eradication treatment (group 1) or a placebo (group 2). The main outcome measure was the prevalence of GI symptoms estimated in groups 1 and 2 and in noninfected patients (group 3) at weeks 2, 6, and 12.

Among H. pylori-negative patients (n=145), GI symptoms were present in 42.6%, 21.4%, and 10.0% at weeks 2, 6, and 12, respectively. In groups 1 and 2, GI symptoms were present in 57.7% and 40.7%, respectively, at week 2 (p= 0.03); 24.7% and 23% at week 6 (p= 0.85); and 9.4% and 17.3% at week 12 (p= 0.13). The prevalence of GI symptoms at week 2 was similar in group 2 and in the H. pylori-negative group (p= 0.77). The highest prevalence of symptoms at week 2 in group 1 was essentially due to diarrhea. The prevalence of GI symptoms was the same for groups 1 and 3 at week 12, and higher in group 2, but the difference did not reach statistical significance.

The short-term (6 wks) GI tolerance of conventional NSAIDs does not differ whether or not the patients are infected by H. pylori. The tendency observed for the medium term (12 wks) deserves to be confirmed.