The extent to which disruptive surgical or medical events impact mortality and function is critical for anticipatory planning and informing goal-aligned care. Using Health and Retirement Study data (2008-2018), we employed propensity score matching to compare the impact of hospitalization for hip fracture (a surgical event) or pneumonia (a medical event) among people with dementia to two groups: (1) people with dementia who did not experience these events; and (2) people without dementia who experienced an event. Dementia status was determined using validated cognitive assessments (Hurd method); hip fracture and pneumonia were identified from Medicare claims. Outcomes were 1-year mortality and function, defined as a summary score of requiring assistance with 6 ADL's and 5 IADL's, with higher scores indicating better function. Among people with dementia, predicted 1-year mortality was higher among those with hip fracture (35.4%) versus those without hip fracture (14.8%), with similar patterns for pneumonia (49.6% vs. 13.0%). Among people with dementia, function declined abruptly at time of hip fracture (-2.09 [95% CI -2.94, -1.25]) and continued to decline after (-0.48 [95% CI -0.87, -0.09]). There were similar patterns for pneumonia (drop at time of pneumonia of -1.49 [95% CI -2.0, -0.97] and after -0.05 [95% CI, -0.29, 0.19]). Compared to people without dementia with hip fracture, people with dementia had higher 1-year mortality (35.4%) versus people without dementia (24%), with similar patterns for pneumonia (49.6% vs. 39.7%). Function stabilized for people without dementia after hip fracture (-0.03, 95% CI -0.22, 0.16), which was significantly different than people without dementia (p < 0.0001). Function improved for people without dementia after pneumonia (0.13, 95% CI 0.03, 0.24), but was not statistically different than for people with dementia (p = 0.17). Disruptive events such as hip fracture or pneumonia substantially alter the clinical trajectories of people with dementia.
Read full abstract