To the Editor: Certain senators announced on August 13 that the end-of-life provision would be dropped from the healthcare reform bill. Senator Charles Grassley was quoted as saying that the provision would be dropped because of fears that it would be misinterpreted or implemented incorrectly. The provision to include Medicare funding for advance directive (AD) discussions between practitioners and patients roused active and, at times, vitriolic discussions on radio talk shows and social networking sites, but the concept of ADs is based on America's core ethical principle of autonomy. Although Nancy Cruzan and Terri Schiavo brought national attention to the issue for a brief time, recent data suggest that, although only approximately 30% of adults have completed an AD,1–3 93% of adults would like to discuss ADs with their physician.4 The reality is that these conversations are time consuming, incompatible with 20-minute appointments, and not billable. The healthcare reform bill would have allowed physician compensation for an AD discussion every 5 years, provided guidelines, and suggested a mechanism to track the quality of these discussions. The bill stated (p. 430) that AD consultations could take place more frequently if there were significant changes in a patient's health status. Those opposed to the end-of-life provision interpreted this to mean that discussing ADs with one's physician would hasten death.
Between 2003 and 2005, 464 patients were interviewed at an index hospitalization and asked whether anyone had discussed ADs with them. A concurrent chart review documented the presence or absence of completed ADs in the medical record (do-not-resuscitate orders (the majority of which are written 3 days before death5), and other broader types of ADs (living will, durable power of attorney, a form such as Five Wishes)). One hundred twenty-three (26.5%) of these patients have died. Using logistic regression, no association was found between having had an AD discussion or the presence of ADs in the medical record and death at 1 year or death over the complete follow-up period after adjusting for age and disease severity (Table 1). Although this study was not powered to detect small differences in mortality or risk of death, it provides some data to inform the current national debate that favors honoring patient wishes to have AD discussions and confirming that there is no evidence that these discussions or completing an AD lead to harm.
|Variable of Interest||Adjusted Odds Ratio|
(95% Confidence Interval)
|Death at 1 Year||Death 2003–2009|
|Model of AD Discussions||(c=0.81)||(c=0.82)|
|AD Discussions||1.2 (0.69–2.24)||1.6 (0.93–2.59)|
|Model of AD on chart||(c=0.83)||(c=0.82)|
|AD on chart||1.4 (0.61–3.03)||1.1 (0.51–2.25)|