Skip to main content

30-10-2011 | Cardiology | Article

CRUSADE helps predict long-term mortality in older NSTEMI patients


Free abstract

MedWire News: The long-term mortality risk for older individuals who have experienced non-ST-segment elevation myocardial infarction (NSTEMI) can be predicted using a model and risk score derived from CRUSADE trial data, US researchers report.

The model and risk score could serve as a valuable risk adjustment tool to inform in-hospital and discharge treatment decisions for NSTEMI patients aged 65 years or older, write Matthew Roe (Duke Clinical Research Institute, Durham, North Carolina, USA) and co-workers in the American Heart Journal.

They linked 43,239 NSTEMI patients (age ≥65 years) from the CRUSADE registry with Centers for Medicare and Medicaid Services data to track all-cause mortality over a median follow-up period of 453 days.

Patients were divided randomly into a model derivation cohort (n=34,640) and a model validation cohort (n=8599). Mortality rates for the whole study population at 1, 2, and 3 years were 24.4%, 33.2%, and 40.3%, respectively.

The researchers identified 22 variables that were independently associated with long-term mortality in multivariable analysis; a full model including these 22 variables had a c-statistic of 0.754 in the derivation sample, and 0.744 in the validation sample.

They then selected the 13 most clinically and statistically significant variables to create a simplified long-term mortality risk score. The variables were age, serum creatinine, systolic blood pressure, signs of heart failure on presentation, heart rate, weight, prior heart failure, hematocrit, troponin ratio, prior stroke, diabetes, gender, and prior peripheral artery disease.

This simplified risk score showed comparable discrimination with the initial model in the derivation and validation cohorts, with c-statistics of 0.734 and 0.727, respectively, note the authors.

"We found that long-term outcomes are 'predictable' based on demographic, clinical, and laboratory variables present at the time of initial hospital presentation," they write.

"We [also] devised a simplified long-term mortality risk score that performed reliably in predicting the probability of 1-year mortality among older patients with NSTEMI," they add.

Roe et al conclude that the CRUSADE model and risk score can be used to target interventions designed to mitigate the high risk for long-term mortality among older patients with NSTEMI and facilitate comparisons of long-term provider outcomes.

By Piriya Mahendra

Related topics