OBJECTIVE

To assess the effectiveness of written advance directives (ADs) in the care of seriously ill, hospitalized patients. In particular, to conduct an assessment after ADs were promoted by the Patient Self-Determination Act (PSDA) and enhanced by the effort to improve decision-making in the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT), focusing upon the impact of ADs on decision-making about resuscitation.

DESIGN

Observational cohort study conducted for 2 years before (PRE) and for 2 years after (POST) the PSDA, with a randomized, controlled trial of an additional intervention to improve decision-making after PSDA (POST+SUPPORT).

SETTING

Five teaching hospitals in the United States.

PATIENTS

A total of 9105 seriously ill patients treated in five teaching hospitals.

INTERVENTIONS

The PSDA mandated patient education about ADs at hospital entry and documentation of ADs in the medical record. The SUPPORT intervention, in addition, provided a nurse to facilitate communication among patients, surrogates, and physicians about preferences for and outcomes of treatment alternatives and, when clinically appropriate, to encourage completion and utilization of ADs.

MEASUREMENTS

Interviews were conducted with patients, surrogates, and attending physicians about awareness, completion, and impact of ADs. Medical records were reviewed for discussion about preferences concerning resuscitation, timing and writing of "Do Not Resuscitate" (DNR) orders, evidence of ADs, and the use or forgoing of resuscitation at the time of death.

RESULTS

In the three cohorts, PRE, POST, and POST+SUPPORT, average age was 63. One-quarter of patients died during the initial hospitalization, one-half were dead within 6 months, and one-half were unconscious for their last 3 days. Before the PSDA (PRE), 62% were familiar with a living will, and 21% had an AD. These rates were similar for the POST and POST+SUPPORT cohorts. Just 36 (6%) of these directives were mentioned in the medical records for PRE, but a stable 35% were documented for POST, and POST+SUPPORT had an increasing rate averaging 78% (P < .001). As previously reported for PRE patients, the POST patients with and without ADs had no significant differences in the rates of medical record documentation of discussions about resuscitation (33% vs 38%, POST without AD vs POST with AD), DNR orders among those who wanted to forgo resuscitation (54% vs 58%), and attempted resuscitations at death (17% vs 9%). The POST+SUPPORT patients had similar results, with no evidence that the intervention enhanced the effect of ADs on these three measures of resuscitation decision-making. Patients with ADs more often reported that preferences about resuscitation were discussed with a physician (e.g., for POST patients, 30% for those with no AD and 43% for those with an AD, P < .05). Only 12% of patients with ADs had talked with a physician when completing the AD. Only 42% reported ever having discussed the AD with their physician. By the second study week, only one in four physicians was aware of patients' ADs.

CONCLUSIONS

In these seriously ill patients, ADs did not substantially enhance physician-patient communication or decision-making about resuscitation. This lack of effect was not altered by the PSDA or by the enhanced efforts in SUPPORT, although these interventions each substantially increased documentation of existing ADs. Current practice patterns indicate that increasing the frequency of ADs is unlikely to be a substantial element in improving the care of seriously ill patients. Future work to improve decision-making should focus upon improving the current pattern of practice through better communication and more comprehensive advance care planning.