Trial results underestimate anticoagulant bleeding risk

By Eleanor McDermid

A study in The BMJ demonstrates that estimates of major bleeding risk in patients taking anticoagulants in clinical trials are unlikely to reflect the reality of clinical practice.

Researcher Shirley Wang (Brigham and Women's Hospital and Harvard Medical School, Boston, Massachusetts, USA) and team say that the restricted population recruited to clinical trials probably accounts for this discrepancy. For example, the exclusion of patients with risk factors for bleeding and inclusion of patients with known tolerance to warfarin.

"These factors resulted in distributions of patient characteristics and range in baseline risks quite different from what might be observed in routine care", say the researchers.

"That the RE-LY trial included 'persistent' [warfarin] users further speaks to the difficulty in trying to use trial data to inform real world practice decisions."

The team found that thromboembolism occurred at an annual rate of 1.7 per 100 patients among 21,934 patients with atrial fibrillation (AF) who started anticoagulation therapy (30% dabigatran) in routine practice.

Observed thromboembolism rates were very similar to those reported in clinical trials, when patients were divided in low-, medium- and high-risk groups according to their CHADS₂ scores.

However, patients' major bleeding risk (4.6 per 100 patients/year) was markedly higher than that reported in the clinical trials, particularly for those taking warfarin. The largest underestimate was by 4.0 major bleeds per 100 patient-years in patients with HAS-BLED risk scores of 3 or higher.

The researchers then developed thromboembolism and bleeding risk models based on clinical practice data. Included bleeding factors were those in the HAS-BLED score: age older than 65 years, hypertension, alcohol abuse, abnormal renal or liver function, previous stroke, history of or predisposition to bleeding, use of drugs that increase bleeding risk, and international normalised ratio.

This approach proved more accurate for assessing bleeding risk, giving predicted rates that were much closer to the observed rates than the clinical trial rates had been.

The model predicted more major bleeding in warfarin- than dabigatran-treated patients at all risk levels, with the difference most marked at high bleeding risk. Predicted major bleeding rates from clinical trials ranged from 3.3 to 4.7 per 100 person-years for warfarin-treated patients with a high-risk HAS-BLED score, whereas the model predicted a rate of 5.8 per 100 person-years.

"Our results suggest that, once trials have determined efficacy of a drug, analyses of observational data of patients treated in routine practice can be of great value for providing tailored estimates of risk for relevant benefit or safety outcomes under alternative treatment strategies", say the researchers.

They add that such models can "be adaptively updated as practice patterns change, indications for use expand, or new treatments emerge for which there are no head-to-head trials."

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.