Preferences for cardiopulmonary resuscitation: physician-patient agreement and hospital resource use. The SUPPORT Investigators

Abstract

Objective: To describe the association between hospital resource utilization and physicians' knowledge of patient preferences for cardiopulmonary resuscitation (CPR) among seriously ill hospitalized adult patients.

Design: Prospective cohort study.

Setting: Five U.S. academic medical center, 1989-1991.

Patients: A sample of 2,636 patients with self- or surrogate interviews and matching physician interviews describing patient preferences for CPR, from a cohort of 4,301 patients with life-threatening illnesses enrolled in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT).

Measures: Patient, surrogate, and physician reports of preferences for resuscitation, and resource use derived from the Therapeutic Intensity Scoring System and hospital length of stay, converted into 1990 dollars.

Results: Nearly one-third of the patients preferred to forego resuscitation. Of the 2,636 paired physician-patient answers, nearly one-third did not agree about preferences for resuscitation. The physicians' views of the patients' preferences and those preferences themselves were both associated with resource use. Standardized adjusted hospital resource consumption, expressed as average cost in dollars during the enrollment hospitalization, was lowest when the physician agreed with the patient preference for a do-not-resuscitate order ($20,527), and highest when the patient did not have a preference and the physician believed the patient wanted resuscitation in the case of a cardiopulmonary arrest ($34,829). Hospital resource use was intermediate when patient-physician pairs evidenced either lack of agreement or communication, or awareness of options about resuscitation.

Conclusions: Both physician and patient preferences for CPR influence total hospital resource consumption. Physician misunderstanding of patient preferences to forego CPR was associated with increased use of hospital resources, and could have led to a course of care at odds with patients' expressed preferences in the event of cardiac arrest. Increasing physicians' knowledge of patient preferences, and increasing communication to help patients understand that options for medical care that include foregoing resuscitation efforts, might reduce hospital expenditures for the seriously ill.