Predicting Mortality of Older Residents in Long-Term Care Facilities: Comorbidity or Care Problems?

Objective: Accurate prediction of life expectancy in long-term care facilities (LTCFs) is important, but previous studies emphasized demographic characteristics, disease diagnosis, or comorbidity. The purpose of this study was to evaluate the roles of geriatric care problems and comorbidity in predicting 12-month mortality in LTCFs. Design: Prospective, observational. Setting: Veterans Care Home. Participants: Residents of Banciao Veterans Care Home. Measurements:A minimum data set (MDS) was implemented, and resident assessment protocol (RAP) triggers were collected as geriatric care problems. Comorbidity of the residents was evaluated using Charlson's comorbidity index (CCI). Results: A total of 559 residents (mean age = 80.9 +/- 5.3 years, all males) were successfully followed, and 50 residents (7.9%) died during the study period. Compared with survivors, deceased subjects had a higher sum of RAP triggers (4.9 +/- 2.0 versus 4.1 +/- 2.0, P = .004) and CCI (1.2 +/- 1.2 versus 0.7 +/- 0.9, P = .014), and were more likely to be hospitalized (1.6 +/- 1.9 versus 0.4 +/- 0.9, P < .001) and visit the emergency department (0.9 1.2 versus 0.5 +/- 1.2, P = .012). Moreover, deceased subjects were more prone to have cognitive loss, urinary incontinence, and behavioral symptoms than survivors (P all < .05). A Cox proportional hazards model showed that both CCI (HR = 1.44, 95% CI: 1.13-1.82, P = .003) and the sum of RAP triggers (HR = 2.03, 95% CI: 1.08-3.82, P = .028) were significantly associated with 12-month mortality. Conclusion: Independent of comorbidity, the sum of geriatric care problems is a significant predictor of 12-month mortality in a veterans care home. Further intervention studies are needed to evaluate whether elimination of these care problems can improve survival in the long-term care setting. (J Am Med Dir Assoc 2010; 11:567-571)