Background/Purpose: The most common group of patients with chronic use of NSAIDs include those with rheumatic diseases such as osteoarthritis (OA), rheumatoid arthritis (RA) and ankylosing spondylitis (AS). Serious GI complications such as bleeding, perforation and obstruction occur in about 1.5% of chronic NSAID users, while about 15% of patients report dyspepsia daily. Guidelines recommend co-therapy with PPIs or misoprostol and/or the use of COX-2 selective inhibitors in chronic NSAID users at high risk for GI events. Most such patients do not receive PPI co-therapy. We investigated the incidence of new GI events and physician compliance of co-prescribing PPIs to chronic NSAID users with RA, OA and AS at a large community hospital.

Methods: We conducted a retrospective cohort study using a simple random sample of 125 patients with OA, RA, AS who were started on NSAIDs between 1/1/2014- 12/31/2014 taken from ECW. Data was collected on 89 patients after excluding for age <30, duration of NSAID use <60 days, or no information on NSAID prescription. The study was divided as: 1) Inclusion period (1/1/2014– 12/31/2014) to identify NSAID prescription and PPI co-prescription. 2) Follow-up period (1/1/2015- 12/31/2015) to assess duration of NSAID/PPI use, and new GI events. 3) History period (1/12013 – 12/31/2013) to assess GI Risk Score (1 point each for Age>65 years, previous PPI use, GI symptoms, H.Pylori/EGD, aspirin, steroids, anticoagulation). Primary outcome measured was new GI event as dyspepsia, PUD, GERD or GI bleeding. We calculated relative risks of GI events by NSAID-PPI overlap time (none, some or complete), both unadjusted and adjusted for previous NSAID use, GI symptoms, cardiac disease, h.pylori/EGD, aspirin, steroids, and anticoagulant use.

Results: 13% of patients were on a PPI throughout NSAID use, 16% were on PPI for some time and 71% had no overlap of PPI. Only 17% were either co-prescribed or were already on a PPI at the time of NSAID prescription. There was a greater risk of GI event with greater raw number of days of NSAID use (Crude RR 1.002, p=0.007). After adjusting for covariates, those on PPI only some of the time were 4.67 times more likely to have a GI event as compared to those on complete overlap (p=0.003). Physicians were 5.31 times more likely to co-prescribe a PPI for patients with 1 or more GI risk factors. (Crude RR 5.31, p=0.007). Physicians co-prescribed PPIs 27% of the time in patients with high GI risk score as compared to 5% for patients with no GI risk factors (Crude p=0.002). Physicians were 5.31 times more likely to co-prescribe a PPI for patients with 1 or more GI risk factors. (Crude RR 5.31, p=0.007).

Conclusion: PPI co-therapy appears to have a protective effect on GI risk associated with chronic NSAID use in RA, OA and AS patients. Chronic NSAID users with higher GI risk factors are more likely to receive a PPI co-prescription, indicating that physicians consider patient’s GI risk level when prescribing NSAIDs. However, 63% of patients who had GI risk factors did not receive a PPI co-prescription indicating incomplete adherence to clinical practice guidelines that recommend gastro-protection for at-risk chronic NSAID users. More awareness about PPI prophylaxis is needed among physicians prescribing NSAID for chronic use.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/incidence-of-gastrointestinal-events-and-physician-compliance-of-co-prescribing-proton-pump-inhibitors-in-chronic-nsaid-users-with-osteoarthritis-rheumatoid-arthritis-and-ankylosing-spondylitis/