A panel of 18 experts from 10 European countries had some difficulty in defining when the benefits sufficiently outweighed the potential adverse effects of various nonsteroidal anti-inflammatory drugs – with and without a proton pump inhibitor – for 144 profiles of patients with chronic rheumatic diseases.
Panelists generally gave patients with low gastrointestinal or cardiovascular risks the full range of NSAID options. Approximately one-third of the patient-drug matches labeled "inappropriate" by panelists applied to the use of a nonselective NSAID without a PPI, the panelists wrote in Annals of the Rheumatic Diseases (Ann. Rheum. Dis. 2011;70:818-22).
When scoring patient profiles, panelists took into account seven clinical variables: age of 65 years or older, history of upper gastrointestinal problems, use of anticoagulants, use of systemic corticosteroids, intermittent or continuous treatment pattern, cardiovascular risk, and the use of low-dose aspirin (for those patients with cardiovascular risk). Against these variables, panelists considered 10 treatment options: ibuprofen, diclofenac, naproxen, celecoxib, etoricoxib, and each of these drugs plus a PPI. They did not consider costs when making their treatment recommendations, according to Dr. G.R. Burmester of the department of rheumatology and clinical immunology, Charité Medical University Berlin, and his coauthors.
For patients with the lowest gastrointestinal and cardiovascular risks, a nonselective NSAID (ibuprofen, diclofenac, or naproxen) was deemed appropriate. As gastrointestinal risks increased, the cyclooxygenase-2 (COX-2) inhibitors celecoxib and etoricoxib alone or a nonselective NSAID plus PPI were considered appropriate. In cases of high gastrointestinal risk and low to average cardiovascular risk, panelists rated ibuprofen/diclofenac plus PPI, or a COX-2 inhibitor plus PPI, as the most appropriate options. For patients with both high gastrointestinal and cardiovascular risks, avoidance of all NSAIDs was recommended, with the use of diclofenac, naproxen, celecoxib, or etoricoxib plus PPI deemed acceptable if necessary.
In January 2008, the panel established the appropriateness rating of treatment options on a 1-9 scale, with 1 as "inappropriate" and 9 as "appropriate." As defined by the RAND/UCLA appropriateness method, a "treatment had to be considered appropriate if the expected benefits exceeded the potential negative consequences by a sufficient margin." However, the panel did not define "sufficient," which led to most of the disagreement on scoring.
In June 2008, after the panel evaluated the first-round practice ratings and revised the evaluation measures, the second round of rating began with the 144 patient profiles and 10 treatment choices. A treatment with a median score of 7-9 was categorized as appropriate, whereas treatments with a median score of 1-3 were labeled inappropriate and those scoring 4-6 were categorized as uncertain.
An online tool (www.e-hims.com/Sensar) shows the corresponding panel treatment recommendation for various patient data.
Panelists were experts in the fields of rheumatology, orthopedics, cardiology, clinical pharmacology, gastroenterology, family medicine, and geriatrics.
All panelists disclosed receiving honoraria from Pfizer, which supported the study with an unrestricted educational grant. Eleven panelists disclosed other relationships with pharmaceutical companies, including Pfizer.