Process of forgoing life-sustaining treatment in a university hospital: an empirical study

Abstract

Objectives: The difficult decision to forgo (withhold or withdraw) life-sustaining treatment has received extensive commentary. Little attention has been paid to how physicians do, and should, care for dying patients once this decision is made. This study describes the characteristics of patients who forgo treatment, determines the range and sequential process of forgoing treatment, and suggests ethical and public policy implications.

Design: The charts of all patients who died at the University of Minnesota Hospital during a 2-month period were reviewed. The patient information that was collected included age and sex, diagnoses, mental status, location in the hospital length of hospital stay, method of payment, the timing of the first decision to forgo treatment, and the range and sequence of interventions forgone.

Setting: All ICUs and general wards in a 586-bed tertiary care university hospital.

Patients: All patients who died at the University of Minnesota Hospital during May and June 1989.

Main results: A total of 52 (74%) of 70 patients who died had some intervention withheld or withdrawn before death. Those patients in whom treatment was forgone were more likely to have an underlying malignancy or impaired mental status and longer hospital stays. Thirty-two (62%) of 52 patients who declined some treatment were in an ICU; 26 (50%) of 52 patients required mechanical ventilation. On average, 5.4 separate interventions were forgone per patient. Resuscitation and/or endotracheal intubation were generally the first measures withheld; once a patient required a ventilator, withdrawing the ventilator was a late decision. Precise methods of ventilatory and vasopressor withdrawal varied considerably among patients.

Conclusions: Forgoing life-sustaining treatment is not a single decision but it often occurs in a sequential manner over several days. A strict analysis of the benefits and burdens of various interventions may be inadequate in deciding what interventions are appropriate in the care of the dying patient.