Abstract
BACKGROUND: The Renal Physicians Association's clinical practice guideline recommends that physicians address advance care planning with dialysis patients. However, data are lacking about how best to implement this recommendation. STUDY DESIGN: Quality improvement project. SETTINGS & PARTICIPANTS: Nephrologists caring for patients treated with maintenance hemodialysis at 2 dialysis facilities identified patients who might benefit most from advance care planning using the "surprise" question ("Would I be surprised if this patient died in the next year?"). QUALITY IMPROVEMENT PLAN: Patients identified with a "no" response to the surprise question were invited to participate in nephrologist-facilitated advance care planning, including completion of a Medical Orders for Life-Sustaining Treatment (MOLST) form. OUTCOMES: Change in MOLST completion rate and identification of preferences for limits on life-sustaining treatment. MEASUREMENTS: Pre- and postintervention code status, MOLST completion rate, and vital status at 1 year. RESULTS: Nephrologists answered "no" to the surprise question for 50 of 201 (25%) hemodialysis patients. Of these, 41 (82%) patients had a full-code status and 9 (18%) had a do-not-resuscitate (DNR) status. Encounters lasted 15 to 60 minutes. Following the encounter, 21 (42%) patients expressed preference for a DNR status and 29 (58%) maintained full-code status (P=0.001). The MOLST completion rate increased from 10% to 90%. One-year survival for patients whose nephrologists answered "no" to the surprise question was 58% compared to 92% for those with a "yes" answer (P<0.001). LIMITATIONS: Sample size and possible nonrepresentative dialysis population. CONCLUSIONS: Nephrologist-facilitated advance care planning targeting hemodialysis patients with limited life expectancy led to significant changes in documented patient preferences for cardiopulmonary resuscitation and limits on life-sustaining treatment. These changes demonstrate the benefit of advance care planning with dialysis patients and likely reflect better understanding of end-of-life treatment options.