Background: Patients with Alzheimer’s disease and related dementias (ADRD) account for 3.2 million hospital admissions per year; they have over three times more hospitalizations than persons without cognitive impairment. Hospitalization in these patients is a seminal event that often leads to delirium, lasting functional and cognitive impairment, institutionalization, premature death, increased resource consumption, and family caregiver distress. This study aimed to evaluate the relationship between ADRD and in-hospital mortality.
Methods: A retrospective chart review across 7 hospitals (3 tertiary and 4 community) within a large integrated health system compared patients 65 and older with a previous diagnosis of ADRD vs a random sampling of patients 65+ with no previous diagnosis of ADRD. Primary outcomes included: in-hospital mortality; secondary outcomes included: incidence of delirium, length of stay (LOS), utilization (tests, consults, hospital cost, and Medicare cost), discharge home, and 30-day readmission. The multivariable regression model included: the presence of dementia, age, gender, race, marital status, arrived from, hospital type (tertiary vs. community), service type (medicine vs. surgical), ICU, DNR status, and comorbidity index (0, 1-2, 3-4, over 4)
Results: The ADRD group (n=6195) had a median age of 84.6, 63% were female and 65% white, , while the non-ADRD group (n=7660) were younger with an median age of 77.0, 54% female, and 62% white. In the univariate analysis, patients with ADRD were more likely to receive psychoactive medications (13% vs. 1.7%), restraints (6.6% vs. 2.5%), and 1:1 observations (14% vs. 3%). Patients with ADRD as compared to those without ADRD were found to have: a longer LOS (5 days vs 4 days; sub-distribution hazard ratio was 1.23, p<0.0001), higher incidence of in-hospital mortality (estimated cumulative incidence of being discharged alive at 5 days after admission was 51.5% [95% CI: 50.6, 52.3] vs. 58.9% [95% CI: 58.0, 59.8]), and 3% increased risk of 30-day readmissions (95% CI: 2, 4.5, p<0.0001). In a multivariable regression model, ADRD, age, marital status, service, arrived from, ICU, DNR, and comorbidity index were associated with mortality. Dementia (0.55, p<0.0001) and arrived from home (0.41, p<0.0001) were protective against mortality. Increased age (0.014, p=0.01), being married (0.26, p=004), admission to medicine (0.72, p<0.0001), ICU (1.54, p<0.0001), DNR (3.26, p<0.0001), and comorbidity index 1-2 (0.29, p=0.037), 3-4 (0.379, p=0.0088), >4 (0.39, p=0075) were associated with increased mortality.
Conclusions: When controlling for illness severity during the acute hospitalization, ADRD was protective for in-hospital mortality. Further studies are needed to understand how to best optimize hospital care for persons with ADRD.