Objectives: To assess the effect of a multicomponent advance care planning intervention directed at nursing home social workers on identification and documentation of preferences for medical treatments and on patient outcomes.
Design: Controlled clinical trial.
Setting: New York City nursing home.
Participants: One hundred thirty-nine newly admitted long-term care residents.
Intervention: Nursing home social workers were randomized to the intervention or control groups. The intervention consisted of baseline education in advance care planning that incorporated small-group workshops and role play/practice sessions for intervention social workers; structured advance care planning discussions with residents and their proxies at admission, after any change in clinical status, and at yearly intervals; formal structured review of residents' goals of care at preexisting regular team meetings; "flagging" of advance directives on nursing home charts; and feedback to individual healthcare providers of the congruence of care they provided and the preferences specified in the advance care planning process. Control social workers received an educational training session on New York State law regarding advance directives but no additional training or interventions. Subjects were enrolled from January 9, 2001 through May 25, 2003 and followed for 6 months after enrollment.
Measurements: Nursing home chart documentation of advance directives (healthcare proxies, living wills) and do-not-resuscitate orders; preferences for artificial nutrition and hydration, intravenous antibiotics, and hospitalization; and concordance of treatments received with documented preferences were compared for residents assigned to intervention and control social workers.
Results: Intervention residents were significantly more likely than residents in the control group to have their preferences regarding cardiopulmonary resuscitation (40% vs 20%, P=.005), artificial nutrition and hydration (47% vs 9%, P<.01), intravenous antibiotics (44% vs 9%, P<.01), and hospitalization (49% vs 16%, P<.01) documented in the nursing home chart. Control residents were significantly more likely than intervention residents to receive treatments discordant with their prior stated wishes. Two of 49 (5%) intervention residents received a treatment in conflict with their prior stated wishes (one hospitalization, one episode of intravenous antibiotics), compared with 17 of 96 (18%) control patients (P=.04).
Conclusion: This generalizable intervention directed at nursing home social workers significantly improved the documentation and identification of patients' wishes regarding common life-sustaining treatments and resulted in a higher concordance between patients' prior stated wishes and treatments received.