IT-assisted comprehensive geriatric assessment for residents in care homes: quasi-experimental longitudinal study.

BACKGROUND: Frailty interventions such as Comprehensive Geriatric Assessment (CGA) can provide significant benefits for older adults living with frailty. However, incorporating such proactive interventions into primary care remains a challenge. We developed an IT-assisted CGA (i-CGA) process, which includes advance care planning (ACP). We assessed if, in older care home residents, particularly those with severe frailty, i-CGA could improve access to advance care planning discussions and reduce unplanned hospitalisations. METHOD: As a quality improvement project we progressively incorporated our i-CGA process into routine primary care for older care home residents, and used a quasi-experimental approach to assess its interim impact. Residents were assessed for frailty by General Practitioners. Proactive i-CGAs were completed, including consideration of traditional CGA domains, deprescribing and ACP discussions. Interim analysis was conducted at 1 year: documented completion, preferences and adherence to ACPs, unplanned hospital admissions, and mortality rates were compared for i-CGA and control (usual care) groups, 1-year post-i-CGA or post-frailty diagnosis respectively. Documented ACP preferences and place of death were compared using the Chi-Square Test. Unplanned hospital admissions and bed days were analysed using the Mann-Whitney U test. Survival was estimated using Kaplan-Meier survival curves. RESULTS: At one year, the i-CGA group comprised 196 residents (severe frailty 111, 57%); the control group 100 (severe frailty 56, 56%). ACP was documented in 100% of the i-CGA group, vs. 72% of control group, p < 0.0001. 85% (94/111) of severely frail i-CGA residents preferred not to be hospitalised if they became acutely unwell. For those with severe frailty, mean unplanned admissions in the control (usual care) group increased from 0.87 (95% confidence interval ± 0.25) per person year alive to 2.05 ± 1.37, while in the i-CGA group they fell from 0.86 ± 0.24 to 0.68 ± 0.37, p = 0.22. Preferred place of death was largely adhered to in both groups, where documented. Of those with severe frailty, 55% (62/111) of the i-CGA group died, vs. 77% (43/56) of the control group, p = 0.0013. CONCLUSIONS: Proactive, community-based i-CGA can improve documentation of care home residents' ACP preferences, and may reduce unplanned hospital admissions. In severely frail residents, a mortality reduction was seen in those who received an i-CGA.