Less NSAIDs for Arthritis Equals More Falls



SNOWMASS, COLO. – The guideline-endorsed demotion of nonsteroidal anti-inflammatory drugs in favor of narcotic analgesics for chronic pain has led to a marked increase in falls, fractures, and other bad outcomes among elderly arthritis patients.

"I think the real take-home message here is that current guidelines for the treatment of pain should be revisited," Dr. Bruce N. Cronstein asserted at a symposium sponsored by the American College of Rheumatology.

Dr. Bruce N. Cronstein

Since the cyclo-oxygenase-2 (COX-2)-selective NSAID rofecoxib (Vioxx) was famously taken off the market in late 2004 because of a scandal related to cover-up of an increased risk of myocardial infarction (MI), prescriptions for narcotic analgesics in elderly patients with arthritis have risen sharply. This trend accelerated after the 2007 publication of an American Heart Association (AHA) scientific statement on the treatment of chronic pain in patients with or at increased risk for heart disease (Circulation 2007;115:1634-42). The AHA guidelines elevated short-term use of narcotic analgesics to first-tier status alongside aspirin, acetaminophen, and tramadol, while demoting both COX-2-selective and nonselective NSAIDs to second-tier status.

The AHA recommendations have been embraced by geriatrics groups and other medical societies. Notably, however, the American College of Rheumatology has declined to get on board, observed Dr. Cronstein, Dr. Paul R. Esserman professor of medicine at New York University.

The AHA guidelines, in his view, are poorly done. The problem is one of tunnel vision. The guidelines focus on the evidence of increased cardiovascular risk associated with nearly all NSAIDs without considering how the drugs stack up in terms of overall safety – noncardiovascular as well as cardiovascular – compared with the other major analgesic group: narcotic analgesics. And it turns out that the NSAIDs look pretty good in comparison, the rheumatologist added.

"You’re trading off falls and fractures for MIs – and it turns out that in patients over age 65, the mortality from hip fracture is significantly greater than it is for MI," said Dr. Cronstein, who is also director of the Clinical and Translational Science Institute, which is a partnership between New York University and the New York City Health and Hospitals Corporation.

Dr. Cronstein cited a large Medicare study conducted by investigators at Brigham and Women’s Hospital, Boston. They examined the comparative safety of analgesics in elderly arthritis patients and concluded that narcotic analgesics come up short.

The investigators sifted through the population of Medicare beneficiaries in Pennsylvania and New Jersey to identify elderly patients with rheumatoid arthritis or osteoarthritis (OA) who were started on a nonselective NSAID, a COX-2-selective NSAID, or a narcotic analgesic during 1999-2005. They came up with 4,280 propensity score–matched patients in each of the three groups. The subjects’ mean age was 80 years, and 85% were women. Patients with prior cancer or nursing-home use were excluded from the study in order to avoid data skewing.

The composite incidence of fractures of the hip, pelvis, humerus, or radius in this National Institutes of Health–funded study was 26 per 1,000 person-years among patients on nonselective NSAIDs, 19 with COX-2-selective NSAIDs, and 101 with opioids.

While it’s not really surprising that opiate analgesics should be associated with increased risk of falls and fractures, another finding in this study proved unexpected: Intriguingly, the composite cardiovascular event rate was 77 per 1,000 person-years with nonselective NSAIDs, 88 per 1,000 with COX-2-selective NSAIDs, and 122 in the narcotic analgesics group.

The patients taking opioids had a 77% greater risk of cardiovascular events and those taking COX-2-selective NSAIDs had a 28% greater risk than did patients on nonselective NSAIDs, according to findings from a multivariate Cox regression analysis. The fracture risk was 4.47-fold greater with narcotic analgesics than with NSAIDs. The gastrointestinal bleeding risk was 40% lower in the COX-2-selective NSAID group than in the other two study arms. The all-cause mortality risk was 87% greater in the narcotic analgesic group than with nonselective NSAIDs, while COX-2-selective NSAIDs were not associated with increased risk (Arch. Intern. Med. 2010; 170: 1968-76).

Dr. Cronstein also presented highlights of a not-yet-published study that documented the changing pattern of analgesic prescriptions for elderly patients with OA since 2005. He conducted the study with coinvestigators at the Geisinger Health Plan in Danville, Pa.

During 2001-2004, 24% of elderly patients with OA in the Geisinger system were on a narcotic-only for pain relief, 15% were on a narcotic analgesic plus a COX-2-selective NSAID, and 13% were on COX-2-selective NSAID monotherapy. But in 2005-2009, 56% of patients were on monotherapy with a narcotic analgesic, 9% were on a narcotic plus a COX-2-selective NSAID, and 2% were on COX-2-selective NSAID monotherapy.


Next Article: