Abstract
RATIONALE & OBJECTIVE: A response "no" (SQ-No) to the surprise question (SQ) of whether a clinician would be surprised if a dialysis patient died in the next 6 months is associated with a higher risk of all-cause death. It is uncertain what domains are intuitively assessed with the SQ. We hypothesized that the SQ would assess the patient's frailty, malnutrition, or patient-perceived health-related quality of life in a cohort of patients on maintenance hemodialysis. STUDY DESIGN: Cohort study. SETTING & PARTICIPANTS: A multicenter study including 994 patients on maintenance hemodialysis in Japan. PREDICTORS: (1) SQ answered by nurses; (2) frailty by modified Cardiovascular Health Study criteria; (3) malnutrition as evaluated by Geriatric Nutritional Risk Index (GNRI); and (4) patient-perceived health-related quality of life examined by the 36-Item Short Form Health Survey (SF-36) physical component summary (PCS). OUTCOMES: All-cause mortality. ANALYTICAL APPROACH: Cox proportional hazard models. RESULTS: Median age and dialysis vintage were 66 and 5.9 years, respectively, 35.8% were women, and 39.6% had diabetic kidney disease. The prevalence of SQ-No and frailty was 19.7% and 45.9%. Median GNRI and SF-36 PCS scores were 96.3 and 36.9, respectively. During the 5-year follow-up, 247 patients died. SQ-No, being frail, low GNRI, and low SF-36 PCS were each significant predictors of a higher risk for mortality independent of potential confounders. SQ-No remained a significant predictor after further adjustment for frailty or GNRI, but SQ-No was no longer significant when adjusted for SF-36 PCS. LIMITATIONS: We did not assess the agreement of responses to the SQ between different raters. CONCLUSIONS: The predictive ability of the SQ was closely related to SF-36 PCS in hemodialysis patients. Nurses' answer to the SQ appears to assess the physical domain of patient-perceived health-related quality of life rather than objectively assessed frailty or malnutrition.