• Tiffany A Radcliff, Aram Dobalian, and Cari Levy.
• VA Eastern Colorado Healthcare System, HSR and D TREP for Long-term Care Research, Denver, CO 80220, USA. Tiffany.Radcliff@uchsc.edu
• J Am Med Dir Assoc. 2007 Feb 1; 8 (2): 91-7.

ObjectivesLittle is known about whether advance directives impact inpatient care for a condition with clear treatment guidelines. The goal of this research was to determine the association between limitation of aggressive treatment (LAT) orders and guideline adherence for acute myocardial infarction (AMI).DesignSecondary examination of data from the national Cooperative Cardiovascular Project (CCP) baseline data. We used seemingly unrelated regression to correct for potential selection bias between patients with and without LAT orders and to determine whether such orders predict guideline adherence for several treatments related to acute myocardial infarction.SettingThe setting included 4111 short-term non-federal acute care hospitals in the United States.ParticipantsParticipants were 147,475 AMI cases with complete data abstracted from inpatient hospital charts, representing most fee-for-service Medicare patients who were hospitalized with AMI between February 1994 and July 1995.MeasurementsAdherence to guidelines for treating acute myocardial infarction, including aspirin, Beta blockers, and reperfusion via thrombolytics or PTCA.ResultsPatients with LAT orders are less likely to receive care in accordance with guidelines when controlling for other factors that may explain a lower likelihood of guideline adherence. After adjustment for selection effects, we found a lower predicted probability that patients received more invasive treatments.ConclusionPatients with LAT orders appear to receive care that is less aggressive and less congruent with acute myocardial infarction care guidelines compared with patients without such orders. Quality improvement measures will need to take this difference into account and ensure that physicians are not penalized for complying with patient care preferences.

36 keywords...

hide…