International studies report a decline in mortality following non–ST-elevation myocardial infarction (NSTEMI). Whether this is due to lower baseline risk or increased utilization of guideline-indicated treatments is unknown.Objective
To determine whether changes in characteristics of patients with NSTEMI are associated with improvements in outcomes.Design, Setting, and Participants
Data on patients with NSTEMI in 247 hospitals in England and Wales were obtained from the Myocardial Ischaemia National Audit Project between January 1, 2003, and June 30, 2013 (final follow-up, December 31, 2013).Exposures
Baseline demographics, clinical risk (GRACE risk score), and pharmacological and invasive coronary treatments.Main Outcomes and Measures
Adjusted all-cause 180-day postdischarge mortality time trends estimated using flexible parametric survival modeling.Results
Among 389 057 patients with NSTEMI (median age, 72.7 years [IQR, 61.7-81.2 years]; 63.1% men), there were 113 586 deaths (29.2%). From 2003-2004 to 2012-2013, proportions with intermediate to high GRACE risk decreased (87.2% vs 82.0%); proportions with lowest risk increased (4.2% vs 7.6%; P= .01 for trend). The prevalence of diabetes, hypertension, cerebrovascular disease, chronic obstructive pulmonary disease, chronic renal failure, previous invasive coronary strategy, and current or ex-smoking status increased (all P < .001). Unadjusted all-cause mortality rates at 180 days decreased from 10.8% to 7.6% (unadjusted hazard ratio [HR], 0.968 [95% CI, 0.966-0.971]; difference in absolute mortality rate per 100 patients [AMR/100], −1.81 [95% CI, −1.95 to −1.67]). These findings were not substantially changed when adjusted additively by baseline GRACE risk score (HR, 0.975 [95% CI, 0.972-0.977]; AMR/100, −0.18 [95% CI, −0.21 to −0.16]), sex and socioeconomic status (HR, 0.975 [95% CI, 0.973-0.978]; difference in AMR/100, −0.24 [95% CI, −0.27 to −0.21]), comorbidities (HR, 0.973 [95% CI, 0.970-0.976]; difference in AMR/100, −0.44 [95% CI, −0.49 to −0.39]), and pharmacological therapies (HR, 0.972 [95% CI, 0.964-0.980]; difference in AMR/100, −0.53 [95% CI, −0.70 to −0.36]). However, the direction of association was reversed after further adjustment for use of an invasive coronary strategy (HR, 1.02 [95% CI, 1.01-1.03]; difference in AMR/100, 0.59 [95% CI, 0.33-0.86]), which was associated with a relative decrease in mortality of 46.1% (95% CI, 38.9%-52.0%).Conclusions and Relevance
Among patients hospitalized with NSTEMI in England and Wales, improvements in all-cause mortality were observed between 2003 and 2013. This was significantly associated with use of an invasive coronary strategy and not entirely related to a decline in baseline clinical risk or increased use of pharmacological therapies.