Objectives: To determine (i) factors which predict whether patients hospitalised with acute myocardial infarction (AMI) receive care discordant with recommendations of clinical practice guidelines; and (ii) whether such discordant care results in worse outcomes compared with receiving guideline-concordant care.
Design: Retrospective cohort study.
Setting: Two community general hospitals.
Participants: 607 consecutive patients admitted with AMI between July 1997 and December 2000.
Main outcome measures: Clinical predictors of discordant care; crude and risk-adjusted rates of inhospital mortality and reinfarction, and mean length of hospital stay.
Results: At least one treatment recommendation for AMI was applicable for 602 of the 607 patients. Of these patients, 411(68%) received concordant care, and 191 (32%) discordant care. Positive predictors at presentation of discordant care were age > 65 years (odds ratio [OR], 2.5; 95% CI, 1.7-3.6), silent infarction (OR, 2.7; 95% CI, 1.6-4.6), anterior infarction (OR, 2.5; 95% CI, 1.7-3.8), a history of heart failure (OR, 6.3; 95% CI, 3.7-10.7), chronic atrial fibrillation (OR, 3.2; 95% CI, 1.5-6.4); and heart rate >/= 100 beats/min (OR, 2.1; 95% CI, 1.4-3.1). Death occurred in 12.0% (23/191) of discordant-care patients versus 4.6% (19/411) of concordant-care patients (adjusted OR, 2.42; 95% CI, 1.22-4.82). Mortality was inversely related to the level of guideline concordance (P = 0.03). Reinfarction rates also tended to be higher in the discordant-care group (4.2% v 1.7%; adjusted OR, 2.5; 95% CI, 0.90-7.1).
Conclusions: Certain clinical features at presentation predict a higher likelihood of guideline-discordant care in patients presenting with AMI. Such care appears to increase the risk of inhospital death.