Bleeding Risk Scores Provide Mixed Results, Especially in Intermediate-Risk Patients

December 12, 2011

December 12, 2011 (San Diego, California)— Four major bleeding risk scores used in clinical practice identified patient risk differently, with each of the screening tools having a limited ability to predict risks of major bleeding and clinically relevant nonmajor bleeding, according to the results of a new study [1]. Overall, the bleeding risk scores were able to identify patients at high risk for bleeding but had limited ability to identify patients at intermediate risk, report investigators.

Dr Alejandro Lazo-Langne

"There are at least five risk scores that have been published, in populations ranging from patients with thromboembolic disease to those with atrial fibrillation," said senior investigator Dr Alejandro Lazo-Langner (University of Western Ontario, London). "Most of the scores, however, have been derived for atrial fibrillation. The most current guidelines suggest that we evaluate patient risk of bleeding before prescribing anticoagulants, but for the last couple of years we've been noticing that the different risk scores classify patients differently."

Speaking with heartwire , Lazo-Langner said the guidelines recommend using the HAS-BLED or HEMORR2HAGES risk scores to help guide clinicians about potential risks of bleeding in patients treated with anticoagulant therapy. The risk scores, however, were developed with different methodologies, and while they each emphasize similar risk factors, the weight accorded to each risk factor differs.

In their study, which was presented this week at the American Society of Hematology 2011 Annual Meeting, the researchers evaluated the Outpatient Bleeding Risk Index (OBRI), the Contemporary Bleeding Risk Model (CBRM), HAS-BLED, and HEMORR2HAGES in a retrospective cohort study that included 321 consecutive patients enrolled at a single academic medical center. Of the patients, 57% were male and 72.6% had atrial fibrillation.

Overall, the incidence rates for major bleeding and clinically relevant nonmajor bleeding were 3.7 and 11.2 events/100 patient-years, respectively. The incidence rates differed significantly when assessing bleeding using the four bleeding risk scores, however, with the risk scores identifying 2.6 major bleeding events/100 patient-years using HAS-BLED in intermediate-risk patients to 6.62 major bleeding events/100 patient-years using the CBRM risk score. In intermediate-risk patients, the HAS-BLED score identified 9.07 major and clinically relevant bleeding events/100 patient-years, whereas the CBRM identified 16.12 events/100 patient-years.

The predictive ability of each risk score was assessed using the C statistic. For predicting major bleeding, the C statistic for OBRI, CBRM, HEMORR2HAGES, and HAS-BLED was 0.606, 0.714, 0.735, and 0.672, respectively. For predicting major bleeding and clinically relevant nonmajor bleeding, the C statistic was 0.549, 0.591, 0.613, and 0.587, respectively.

"What we found with the four scores is that they all classify patients differently," said Lazo-Langner. "It really doesn't make any sense if it's the same patient population. Only in the very high-risk categories do we see better performance." The four risk scores all include age, renal function, comorbidities, and a previous history of bleeding, "which is what we have been using for years when assessing risk of bleeding in patients," added Lazo-Langner.

While the scores are intended to be an aid for physicians not at the forefront of anticoagulant management, such as the general practitioner, the differential performance of the four risk scores makes things more difficult. Moreover, in patients it is often hard to predict who bleeds on anticoagulant therapy, given that a lot of events are random, such as accidental falls.

"A lot of the variables are actually very easy to assess in clinical practice," said Lazo-Langner. "The problem with the risk scores is that they try to assign a weight to each variable, and that's what we don't exactly know how to do. When we're prescribing anticoagulants, the patients we're concerned about are the patients with a history of frequent falls, or the elderly patient with cognitive issues. A patient with cognitive issues might have no other risks, but we still might be quite concerned about them."


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: