Major bleeding events and risk stratification of antithrombotic agents in hemodialysis: results from the DOPPS

Abstract

Benefits and risks of antithrombotic agents remain unclear in the hemodialysis population. To help clarify this we determined variation in antithrombotic agent use, rates of major bleeding events, and factors predictive of stroke and bleeding in 48,144 patients in the Dialysis Outcomes and Practice Patterns Study (DOPPS) phases I-IV. Antithrombotic agents including oral anticoagulants (OACs), aspirin (ASA), and anti-platelet agents (APAs) were recorded along with comorbidities at study entry, and clinical events including hospitalization due to bleeding were then collected every 4 months. There was wide variation in OAC (0.3-18%), APA (3-25%), and ASA use (8-36%), and major bleeding rates (0.05-0.22 events/year) among countries. All-cause mortality, cardiovascular mortality, and bleeding events requiring hospitalization were elevated in patients prescribed OACs across adjusted models. The CHADS2 score predicted the risk of stroke in atrial fibrillation patients. Gastrointestinal bleeding in the past 12 months was highly predictive of major bleeding events; for patients with previous gastrointestinal bleeding, the rate of bleeding exceeded the rate of stroke by at least twofold across all categories of CHADS2 score, including patients at high stroke risk. Appropriate risk stratification and a cautious approach should be considered before OAC use in the dialysis population.

Conflict of interest statement

All other authors have no conflicts of interest to report.

Figures

Figure 1

Use of antithrombotic agents, by DOPPS phase and country. APA -other than aspirin (clopidogrel, ticlopidine, dipyridamole, pentoxifylline); OAC includes warfarin (85% of total OAC) and a few other products (e.g. fluindione in France); DOPPS study period: phases 1–2 (1998–2004) and 3–4 (2005–10) with N=48,144 patients.

Figure 2

Variation in facility use of antithrombotic agents, by geographic region. N=20,475 patients from DOPPS phases 3 and 4 (2005–2010) in facilities with at least 7 patients with medication use records.

Figure 3

Rates of bleeding events and stroke by country. GI=gastrointestinal, VA=vascular access. N=39,440 patients from DOPPS phases 1,2 and 3. *Other bleeding events include epistaxis, subdural hematoma, evacuation of hematoma, hemoptysis and hematuria. **Stroke includes hospitalization or death due to stroke.

Figure 4

Stroke rate and bleeding rate by CHADS2 score and history of GI bleed. Stroke risk categories are based on CHADS2 categories used commonly in the non-dialysis population: 0=low, 1=moderate, ≥2=high. N=37,657 (N=4,026 atrial fibrillation) patients from DOPPS phases 1, 2 and 3. NOTE: Patient medication usage was classified hierarchically (OAC >APA >ASA for patients taking ≥one medication class. See text for details).

Figure 4

Stroke rate and bleeding rate by CHADS2 score and history of GI bleed. Stroke risk categories are based on CHADS2 categories used commonly in the non-dialysis population: 0=low, 1=moderate, ≥2=high. N=37,657 (N=4,026 atrial fibrillation) patients from DOPPS phases 1, 2 and 3. NOTE: Patient medication usage was classified hierarchically (OAC >APA >ASA for patients taking ≥one medication class. See text for details).