ABSTRACT
Older patients have been remarkably underrepresented in bleeding risk cohorts. Thus, the PRECISE-DAPT (Derivation and validation of the predicting bleeding complications in patients undergoing stent implantation and subsequent dual antiplatelet therapy) and Academic Research Consortium for High Bleeding Risk (ARC-HBR) scores are not validated in older adults. Therefore, we sought to evaluate the PRECISE-DAPT and ARC-HBR scores in an exclusively older population and assess the prognostic value of a truly simplified clinical evaluation (SCE), consisting of only 3 binary clinical variables (hemoglobin <11 g/100 ml, previous bleeding, and anticipated use of anticoagulants). This is a retrospective analysis of the prospective single-center older-HCD registry. Consecutive patients aged ≥75 years who underwent percutaneous coronary intervention from 2012 to 2019 were included. The primary end point was postdischarge bleeding at 12 months of follow-up, defined according to the Bleeding Academic Research Consortium 3 or 5 criteria. A total of 693 patients with a mean age of 81 (±4.4) years were included in the study and 60 patients (6.8%) met the primary end point. The PRECISE-DAPT and ARC-HBR scores did not significantly predict postdischarge bleeding in the Cox regression models (hazard ratio 1.65 [0.78 to 3.42] and 1.46 [0.72 to 4.24], respectively), whereas the SCE outperformed both scores (hazard ratio 2.47, 1.34 to 4.49). All 3 scores exhibited a moderate discriminatory potential, as determined by a receiver-operating characteristic curve analysis (areas under the curve 0.601, 0.621, and 0.616, respectively), with no significant differences between them. The SCE showed an Integrated Discrimination Improvement of 0.25, p = 0.02 (SCE vs ARC-HBR) and 0.24, p = 0.01 (SCE vs PRECISE-DAPT), with an Net Reclassification Improvement of 6.54%, p = 0.37 and 7.12%, p = 0.43, respectively. In conclusion, the PRECISE-DAPT score and ARC-HBR criteria showed insufficient predictive value in older adults. A truly SCE consisting of 3 easily accessible variables not only provides equal discriminatory potential but also demonstrates superior predictive value, as determined by Cox regression models. This makes it a highly appealing tool for risk stratification, pending its evaluation in larger prospective studies.