Proactive IT-assisted CGA (i-CGA) in care homes improves adherence to preferred place of care & death, hospitalisation & mortality rates

Introduction: 

Primary care-based frailty identification and proactive comprehensive geriatric assessment (CGA) remains challenging. Our Devon-based Primary Care Network has developed and introduced an innovative, community-based IT-assisted CGA (i-CGA) process, which includes advance care planning (ACP). We wished to see if this process could improve effectiveness of ACP in residential care home (CH) residents.

Methods/Intervention:

1) GPs clinically assessed all CH residents for frailty.

2) Proactive i-CGAs completed using our IT-assisted CGA tool, which prompts to review/consider/address:

• Previous i-CGA-related entries
• Traditional CGA-domains/risks
• High-risk drugs/deprescribing
• ACP discussions (hospitalisation/resuscitation/place of death preferences)

3) ACPs shared with relevant healthcare services/Out-Of-Hours.

Interim analysis focused on adherence to ACP-documentation in severely frail residents, comparing groups:

• i-CGA (1-year post-i-CGA completion)
• Control (1-year post-frailty diagnosis, no i-CGA, usual care)

Results:

i-CGA group: 196 residents(16 mild, 69 moderate,111 severe frailty)

Control group: 100(13 mild,31 moderate,56 severefrailty ).

No significant baseline differences.

Advance care planning:

• i-CGA: 100% residents had documented resuscitation decisions. 97% (191/196) preferred to "allow a natural death. Patients with severe frailty: 85%(94/111) preferred not to be hospitalised. 55%(52/94) died, 90%(47/52) in their CH.
• Control:  72%(72/100) had documented resuscitation decisions or which 97% in this group (70/72) preferred to "allow a natural death". Patients with severe frailty: 29%(16/56) had no hospitalisation preferences documented and in this group 25%(4/16) died in hospital.

Hospitalisation in residents with severe frailty:

• i-CGA: compared to the preceding year, unplanned hospitalisation rates fell:0.86 to 0.68/person years alive.
• Control: Unplanned hospitalisations increased:0.87 to 2.05/person years alive.

Survival: significant group mortality difference was seen at one year: 55%(62/111) severely frail i-CGA residents died compared to 77%(43/56) controls, p=0.0013.

Conclusions:

Proactive primary care-led i-CGA in severely frail CH residents promotes up-to-date discussions regarding preferred place of care and death. Most prefer not to be hospitalised, despite traditionally high rates of unplanned admissions. Our i-CGA/ACP process improves adherence to preferences, reduces unplanned hospitalisations and mortality rates. Progressive i-CGA completion and annual/opportunistic reviews should confer progressive benefits.