A CRUSADE-adjusted prospective cohort analysis of heparin dosing velocity, multivariable mediation and iatrogenic haemorrhage in acute coronary syndromes

Background

The activated partial thromboplastin time (aPTT) is the most widely used assay for monitoring continuous intravenous unfractionated heparin (UFH) in acute coronary syndromes (ACS), yet acute-phase reactants – particularly C-reactive protein (CRP) – can artefactually shorten clotting times by inducing Factor VIII and fibrinogen elevation. This phenomenon, which we term ‘aPTT Blindness’, prompts clinicians to escalate UFH doses in pursuit of a falsely shortened therapeutic target, creating a dose escalation cascade that may disproportionately expose low-weight patients to iatrogenic haemorrhage. No prior study has employed multivariable mediation analysis to quantify the proportion of CRP-associated bleeding risk that operates through this aPTT-mediated pathway, nor has any study used CRUSADE risk score adjustment to isolate protocol-attributable excess bleeding from baseline patient frailty.

Methods

We conducted a CRUSADE-adjusted prospective cohort study of 1600 consecutive adult patients with ACS (STEMI, NSTEMI, and high-risk unstable angina) who received continuous intravenous UFH, with data collected from admission through a 5-day observation window. The cohort was stratified by body weight (<65 kg, 65–85 kg, >85 kg) and CRP quartile. The primary safety end-point was major bleeding classified as BARC type 3 through 5. Inverse probability of treatment weighting balanced baseline covariates, and Fine-Gray sub-distribution hazard regression handled in-hospital death as a competing risk. A multivariable mediation analysis (Baron and Kenny framework, bootstrapped with 5000 replications) was performed to decompose the total effect of CRP on bleeding into direct and aPTT-mediated indirect effects. CRUSADE bleeding risk scores were calculated for each patient; the difference between observed and CRUSADE-predicted bleeding rates was defined as protocol-attributable excess risk.

Results

The overall BARC 3–5 bleeding rate was 6.8%, with the lowest weight group (<65 kg) experiencing an 11.4% incidence compared with 2.8% in the highest weight group (>85 kg). After CRUSADE adjustment, the excess bleeding attributable to the aPTT-chasing protocol in low-weight patients with pseudo-resistance was 4.5 percentage points (observed 11.4% vs CRUSADE-predicted 6.9%). The multivariable mediation analysis demonstrated that 47.3% (95% confidence interval [CI] 38.2%–56.8%) of the CRP-to-bleeding association was mediated through the aPTT pathway (indirect effect OR 1.20; 95% CI 1.11–1.30; p < 0.001). The interaction between high CRP (>24 mg/L) and low body weight yielded a sub-distribution hazard ratio of 2.94 (95% CI 2.08–4.15; p < 0.001) for major bleeding.

Nearly half of the CRP-associated bleeding risk in ACS is mediated through aPTT artefact – what we term CRP-driven ‘aPTT Blindness’. Even after accounting for baseline bleeding risk with the CRUSADE score, low-weight patients on aPTT-guided titration nomograms still experience a clinically meaningful excess of protocol-attributable haemorrhage. The interaction between high CRP and low body weight (SHR 2.94) was the strongest single predictor of major bleeding. These results suggest that aPTT-guided monitoring may be particularly problematic in patients with raised inflammatory markers or low body weight and support the need for comparative trials evaluating alternative monitoring strategies.