The Road to Unintended Consequences Is Paved With Good Intentions*
Though most currently practicing physicians were trained in the era of the do-not-resuscitate (DNR) order, it is a relatively recent development in the history of medicine. After the adoption of closed-chest cardiac massage as a routine medical treatment for cardiac arrest in 1960, it became evident that the practice may not necessarily be in the best interest of every patient (1). With a very low survival rate to hospital discharge in even the healthiest patients, it was clear that those with more advanced disease may indeed prefer to opt out of a practice that would ultimately provide them with little to no benefit (2–4). At this point in medical history, however, the imperative to preserve and extend life superseded any concept of patient autonomy to refuse intervention. This extension of “being alive,” however, was balanced with neither the futility of modern medicine nor the patient’s personal concept of “life” especially in the face of incurable disease. In 1976, Quinlan versus New Jersey changed the paradigm (5). This case pitted 21-year-old Quinlan’s parent’s wish to remove their daughter from life-supporting devices against her physician’s refusal to honor this request citing their duty to preserve and extend life. The court eventually ordered that the hospital honor the parent’s request, and thus began the discussion on patient autonomy and the refusal of life-sustaining care as a legitimate option for patients. In 1983, the President’s Commission for the Study of Ethical Problems in Medicine concluded that having the option of a DNR order is an acceptable practice; however, it was recommended that the default order continue to be full resuscitation in the absence of documentation otherwise (6). Previously, the concept of patient autonomy had been subordinate to the absolute value of human life. The Quinlan case marked the transition to acceptance of a relative rather than absolute value to human life. This ultimately allowed the patient to act as the arbiter, able to weigh their medical circumstances against their own ostensible quality of life (a morally correct principle).
The article by Fuchs et al (7) in this issue of Critical Care Medicine demonstrates with some clarity that there continues to be misapplication of this very simple order. During the time period studied (2001–2008), the presence of a DNR order on admission to the ICU was identified as an independent risk factor for 28-day mortality (7). After matching for severity of illness, the patients in the DNR cohort were found to have a significantly higher mortality rate than patients who had the default full resuscitation order upon admission to the ICU. At first glance, these results may simply reinforce that DNR was correctly applied to terminal patients wishing to avoid care in their moribund state. However, a more subtle examination of the data supports an apparent modification of the intensity of the provided medical care, that is, fewer laboratory and radiologic studies were ordered on those patients in the DNR cohort. One possible confounder of these data is that it exists in a time period predating the approval and use of Medical Orders for Life-Sustaining Treatment, which serves to specifically delineate a patient’s further wishes in limiting care beyond cardiac rest scenarios. It may be that DNR during this time period is simply a marker for these other wishes to limit care. In fact, a higher proportion of the DNR cohort went on to “comfort care only.” Regardless, these findings are concerning because not only were the stated wishes of the patients (DNR) possibly misapplied, it appears as though the unintended consequences included increased mortality.