A pilot cluster randomised trial of an evidence-based intervention to reduce avoidable hospital admissions in nursing home residents (Better Health in Residents of Care Homes with Nursing - BHiRCH-NH study)

Abstract Background: Nursing home residents are often frail and most have complex healthcare needs. Hospitalisation is distressing to the person, their family and nursing home staff and costly for health services. Early detection of changes in residents’ health may reduce admissions. Our aim was to pilot a complex intervention with implementation support to improve early detection and treatment for four common Ambulatory Care Sensitive Conditions (ACSCs); urinary tract and respiratory infections, chronic heart failure and dehydration. Methods: Pilot cluster randomised controlled trial in 14 nursing homes (7 intervention, 7 control). The intervention involved 1) “stop and watch (S&W)” early warning tool for changes in physical health, 2) condition- specific care pathways and 3) Situation, Background, Assessment and Recommendation (SBAR) to structure communication with primary care. of Intervention implementation was supported by Practice Development Champions (PDCs) in the care home, a Practice Development Support Group (PDSG) and regular telephone coaching with external facilitators. We collected data on nursing homes (quality ratings, size, ownership), residents, their family carers and staff demographics during the month prior to intervention and subsequently, numbers of admissions, accident and emergency visits and unscheduled GP visits monthly for 6 months during the intervention period. We collected data on how the intervention was used, healthcare resource use and quality of life data for economic evaluation. We assessed participant recruitment and retention and whether a full trial was warranted. Results: We recruited 14 nursing homes, 148 staff, 95 family carers and 245 residents. One nursing home closed prior to the intervention starting and another withdrew during the data collection phase. We retained the majority of participants recruited (95%). 15% of residents had an unplanned hospital admission for one of the four study conditions during follow-up. We were able to collect sufficient data on study questionnaires (all were over 96% complete). No home implemented intervention tools as planned. Only 16 S&W forms and 8 Care Pathways were completed. There was no evidence that the intervention was harmful.Conclusions: Study recruitment, retention, data collection and processes were effective but the intervention was not implemented in practice, therefore a full trial is not warranted. Trial registration: ISRCTN74109734 (registered 6/11/2017) https://doi.org/10.1186/ISRCTN74109734