Evaluation of Nursing Home Referrals: A retrospective study of Nursing home admissions to acute hospital

Background Older nursing home residents (NHRs) typically present with multiple comorbidities and high frailty, necessitating careful management to avoid potentially avoidable hospital transfers . International reports indicate that between 6.8 % and 45.7 % of NHRs are transferred to acute hospitals over variable follow-up periods . Such transfers carry increased risks of in-hospital complications, including pressure ulcers, nosocomial infections, delirium, functional decline, and substantial healthcare costs. Moreover, nearly 40 % of care home residents who die in hospital do so within 24 hours of admission, suggesting that many transfers may be inappropriate. Methods To characterize referral processes and outcomes, we performed a retrospective chart analysis audit of 38 consecutive NHR admissions to our model 3 hospital over a 30 day period. Results The cohort’s mean age was 83 years. Sepsis accounted for 50% of admissions, with falls, delirium, acute kidney injury, and lower respiratory tract infections also common. 35% of referrals occurred at weekends, and 50% of weekday referrals were out-of-hours. Only 10 % underwent in-person general practitioner review and 20 % had telephone assessment before transfer. Transfer documentation was complete in 12 % of cases, incomplete in 65 %, and absent in 23 %, while 90 % had full medication lists. Advance directives precluding hospital admission existed for 13 % of admissions. In-hospital mortality within 24 hours occurred in 5 % of admissions, 15 % were discharged on oral therapy within 24 hours, and 27 % were readmitted within 30 days. Conclusion These findings highlight significant deficiencies in pre-transfer clinical review and communication, particularly during out-of-hours periods, underscoring the need to strengthen community-based care pathways, standardize documentation, and ensure timely GP involvement to reduce avoidable hospitalizations in this vulnerable population.