Dementia Diagnosis Before and After Hip Fracture in Medicare Patients

Dementia is a risk factor for hip fracture [1]. Persons with both hip fracture and dementia are medically complex and have heightened risk for post-operative complications, hospital readmission, functional impairment, and death [1]. Despite implications for healthcare quality and post-acute care reimbursement reform [2], little is known about the prevalence and trajectory of dementia diagnosis before and after hospitalization for hip fracture. We used data from a 20% sample of Medicare Fee-for-Service beneficiaries and included individuals with an incident hip fracture hospitalization between January 2016 and September 2019, who were continuously enrolled 1 year prior and 6 months post-hospitalization. We used the Bynum algorithm [3] with one-year look back window to identify individuals with a claims-based dementia diagnosis at the following-time points relative to incident admission: 90, 60 and 30 days prior and up to 180 days post hospitalization, 1 day before hospitalization, and day of hospital discharge. We describe the characteristics of three subgroups of older adults: (1) no dementia diagnosis, (2) dementia diagnosis within 180 days of fracture and no prior diagnosis (3) dementia diagnosis prior to fracture. The exact date of new dementia diagnoses made in the hospital and in a skilled nursing facility (SNF) are unknown, so we used the date of admission for diagnoses made in these settings. Between-group differences of means and proportions for the three dementia diagnosis groups were tested using analysis of variance (means) and chi-squared tests (proportions). The study was approved by the institutional review board at the Johns Hopkins Bloomberg School of Public Health. Among 65,519 older adults hospitalized for hip fracture, the proportion with a dementia diagnosis increased from 28.5% pre-fracture to 42.3% at 180-days post fracture (Figure 1). People with (versus without) pre- or post-fracture dementia were 5 years older and had longer lengths of hospital stay (Table 1). A higher proportion of people with (versus without) pre- or post-fracture dementia had a post-operative delirium diagnosis, were dual eligible, discharged to a skilled nursing facility (SNF) and died within 180 days. Compared with post-fracture dementia, people with pre-fracture dementia had more chronic conditions and a greater proportion were living in a long-term care facility. The 180-day cost of care was highest for people with post-fracture dementia ($55,577) compared with people with pre-fracture dementia ($46,957) and no dementia ($45,553). Higher costs were driven by higher inpatient, SNF, and hospital readmission cost. Our study finds that 14% of Medicare FFS beneficiaries who suffer a hip fracture incur a new diagnosis of dementia at, or 6 months following hospitalization with over a third of diagnoses occurring while hospitalized. One possible explanation for the increase in new diagnosis of dementia after hip fracture is that patients who fracture had underlying dementia that went unrecognized until the time of the fracture. Unrecognized dementia is common in older adults, and 36% of new dementia diagnoses in our study occurred during hospitalization, suggesting that dementia was likely present but unrecognized prior to the hospitalization. Some of the dementia cases may have been identified when the patients experienced postoperative delirium, which is common among older adults with hip fractures [5] and especially in those with underlying cognitive impairment. Delirium is a known risk factor for cognitive decline following fracture for which mitigating interventions are needed. The rate of delirium was lower than documented incidence in hospitalized older adults [6] which may be due to inaccurate coding or failure to recognize. Our findings are consistent with growing evidence that fractures are associated with incident dementia diagnosis [7], highlighting the need for cognitive screening for older adults who are hospitalized for hip fractures. We also observe higher healthcare cost among older adults with a post-fracture dementia diagnosis, which underscores the importance of care coordination initiatives such as orthogeriatric programs and age friendly hospitals for providing high-quality, cost-effective care during hospitalization and after discharge. Study findings are also important in the context of evolving Medicare payment policy around dementia and post-acute care. Recent payment reform for SNFs introduced higher payment for patients meeting clinical criteria including dementia [2]. Change in reimbursement policy may incentivize inpatient and post-acute facilities to identify more dementia cases. The impact of this policy on hospital coding and discharge and intake decisions of hospitals and SNFs is unknown. The sensitivity of identifying dementia diagnosis using Medicare data is low which is problematic for policy and health services research that relies on this data. Future work on improving the validity of dementia diagnosis in claims data and clinical guidelines for dementia screening in this patient population are warranted. Lisa Reider: study concept and design, interpretation of data and preparation of paper. Joseph F. Levy: study concept and design, analysis, interpretation of data and preparation of paper. Yiqing Kuang: analysis, preparation of paper. Esther S. Oh: interpretation of data and preparation of paper. The authors declare no conflicts of interest.