1436 CLINICAL OUTCOMES OF OLDER HOSPITALISED ADULTS LIVING WITH FRAILTY IN NORTHERN TANZANIA: A PROSPECTIVE MULTICENTRE STUDY

Abstract Introduction As global populations age, healthcare systems are facing challenges posed by multimorbidity, disability and geriatric syndromes. In high-income countries, frailty is a strong predictor of poor hospital outcomes. Comprehensive Geriatric Assessment is effective but resource-intensive and unavailable in sub-Saharan Africa where specialist geriatric training and allied health infrastructure are limited. Aim To establish clinical outcomes of older adults with frailty admitted to hospital in northern Tanzania. Methods All adults aged ≥60 years admitted to medical wards at four hospitals were invited to participate. Participants were screened for frailty using the Clinical Frailty Scale (CFS). The primary outcome was inpatient death, with secondary outcomes including length of stay, 30-day readmission and delirium (confirmed using the Confusion Assessment Method [CAM]). Outcomes for frail (≥5 on CFS) and non-frail participants (1-4 on CFS) were compared. Results Over 6 months, 308/540 patients admitted participated. Reasons for non-participation included death (n=34) and discharge (n=159) before researcher attendance. Mean age of participants was 74.9 years and 154 (50.1%) were female. Of these, 205 (67%) participants had a CFS ≥5. 21 (14.9%) frail participants died, compared with 5 (6.4%) in the non-frail group (Chi-squared, p=.095). Length of stay and re-admission rates were higher in frail participants, but differences were not statistically significant. Delirium was diagnosed in 35 (17%) frail participants, compared with 4 (4%) in the non-frail group (Fisher’s Exact test, p=<.001). Conclusion Frailty in older adults admitted to hospitals in northern Tanzania is common and associated with significantly higher rates of delirium. Mortality, readmissions, and length of stay were higher in the frail group, but differences did not reach statistical significance. Type II Error (exacerbated by selection bias from non-inclusion of individuals who were discharged, or died, early in their admission) may explain this. Participants will now be followed-up for 12-months to assess outcomes longitudinally.