Withdrawal of mechanical ventilation in anticipation of death in an intensive care unit: Cook D, Rocker G, Marshal J, et al. N Engl J Med 2003;349:1123-32

Abstract

ConclusionHigh likelihood of poor cognitive functional outcome, the use of vasoactive drugs, and physician perception of low patient survival in the intensive care unit (ICU) or physician perception that the patient did not wish life support were the strongest determinants of withdrawal of mechanical ventilation in an ICU. SummaryMechanical ventilation is the most frequently withdrawn support when a patient is believed to have no realistic chance for survival. Patients 18 years of age or older receiving mechanical ventilation and expected to be in an ICU for at least 72 hours were studied. The study involved 15 university-affiliated medical-surgical ICUs: 11 in Canada, 2 in the United States, 1 in Sweden, and 1 in Australia. The relationship between baseline patient physiologic characteristics, multiple organ dysfunction scores, patient decision-making ability, forms of life support used, the use of “Do Not Resuscitate” orders, physician prediction of patient status, and patient preferences about life support were examined with regard to association with withdrawal of mechanical ventilation.The study included 851 patients. Of these, 539 (63.3%) were weaned from the ventilator, and 482 were discharged from the hospital; 146 died while receiving ventilatory support; and in 166 (19.5%) mechanical ventilation was withdrawn. Of these, 160 died in the hospital and 6 were discharged from the hospital. Factors associated with withdrawal of mechanical ventilation included the need for vasoactive drugs (hazard ratio, 1.78; 95% confidence interval [CI], 1.20-2.66; P = .004), physician prediction of less than 10% patient survival in the ICU (hazard ratio, 3.49; 95% CI, 1.39-8.79; P = .002), doctor prediction of severe cognitive dysfunction (hazard ratio, 2.51; 95% CI, 1.28-4.94; P = .004), and physician perception that the patient did not desire life support (hazard ratio, 4.10; 95% CI, 2.57-6.81; P < .001). Age, severity of illness, previous functional status, and severity of organ dysfunction were not independently associated with withdrawal of mechanical ventilation. CommentThe article calls into question the concept that patient physiologic characteristics, such as age and organ dysfunction, are the main determinants of the decision to withdraw mechanical ventilation. While it is encouraging that physicians are focusing on the patient's perceived wishes for life support, it must be noted that in most cases these perceptions are guided by physician intuition or those expressed by family members. Certainly physician and family members' perception of the patient's desires may differ from those of the patient. The article points out the continued need for clarification of patient wishes through the advanced directive process.