Do not resuscitate orders and the cost of death

Arch Intern Med. 1993 May 24;153(10):1249-53.


Background: The appropriate role of cardiopulmonary resuscitation in the hospital continues to be a topic of interest to physicians and patients alike. The use of do not resuscitate (DNR) orders reflects a growing expression of autonomy by patients to refuse medical treatment, and also a growing recognition of its futility in many circumstances by physicians. Although it has been suggested that wider use of advance directives will lead to a reduction in health care costs near the end of life, little empiric data exist to support this prediction. This study was designed to ascertain the rates of DNR orders and their associated costs.

Methods: A retrospective chart review was conducted on the hospital records of 852 of 953 hospital deaths that occurred in a referral hospital. Data were collected on resuscitation status, timing of DNR orders, participants in decision making, and physician and hospital charges.

Results: Of the 852 records reviewed, 625 (73%) had a DNR order at the time of death. The use of DNR orders for patients who died ranged from 97% of those on an oncology service to 43% of deaths on cardiology services. One hundred seven patients (17%) had the DNR order before admission. Of 512 patients who had a new DNR order in the hospital, approval was obtained from the patient in only 19%. Patients who died with a DNR order had longer hospital stays (median, 11.0 days) compared with those who died without a DNR order (6.0 days). The time from DNR order to death was 2 days overall with 2.0 days for medical patients and 1.0 day for surgical patients. Average charges for each patient who died were $61,215 with $10,631 for those admitted with a DNR order, and $73,055 for those who had a DNR order made in hospital.

Conclusion: This study demonstrates high variability in the use of DNR orders between various medical and surgical services. These range from a high of 98% on an oncology service to a low of 43% on cardiology. Most patients have a DNR order at the time of death, but these typically occur late in the course of the hospital stay. Death in the hospital is costly and total hospital and professional charges are significantly lower when a patient is admitted with an established nonresuscitation order compared with those for whom a DNR is established while in the hospital. This study provides a basis against which to measure the impact of efforts such as the Patient Self-Determination Act of 1990 to increase the use of advance directives, as well as monitor their effect on health care expenditures.

MeSH terms

  • Advance Directives / legislation & jurisprudence
  • Aged
  • Costs and Cost Analysis
  • Fees, Medical / statistics & numerical data
  • Female
  • Health Care Costs / statistics & numerical data*
  • Hospital Bed Capacity, 500 and over
  • Hospital Mortality
  • Humans
  • Length of Stay / economics
  • Male
  • Ohio
  • Resuscitation Orders*
  • Retrospective Studies