Optimising primary care for refugees. Findings from an Australian cluster randomised trial

The problem

The global community is experiencing unprecedented levels of human displacement - with 26.4 million refugee or asylum seekers in 2018. Australia has resettled over 180,000 refugees in the last decade.

The health and wellbeing of people of refugee background is linked with their ability to access high quality, coordinated primary care. National guidelines recommend a health assessment be offered to all new arrivals from refugee-like backgrounds soon after arriving in Australia.

We asked whether an outreach facilitation intervention could increase the conduct of comprehensive health assessments for General Practice (GP) patients from refugee backgrounds (primary outcome).  While secondary outcomes assessed refugee status identification, interpreter use and knowledge of refugee specific referral pathways, this paper focusses on the primary outcome.  

The approach

OPTIMISE was a mixed methods, practice facilitation intervention co-designed with regional consortia to improve access, integration and quality of primary care received by resettled refugees across 3 urban regions of high refugee resettlement in Australia. We used a pragmatic, cluster randomised controlled trial design with practices allocated using a stepped wedge approach (2 steps) to early or late intervention groups.

Intervention facilitators were employed by regional health services, who made three face-to-face visits and three telephone calls over 6 months to intervention practices. They used structured action plans to help practice staff improve routines of refugee care.

Routine practice data was extracted (using the PENCS CAT4™ tool) to identify patient records where demographic fields and free text indicated refugee status and whether individuals had received refugee health assessments (RHAs) in the previous 6 months.

Data were complemented by information on practice use of the national interpreter service and three surveys assessing practice structure, practitioner characteristics, and practice approaches to refugee care.  Analysis of the primary outcome used multilevel mixed effects models to account for clustering, intervention status, timepoint, practice size, early or late group and region.

Findings

31/36 practices completed the intervention. Analysis focussed on the 14,633 patients from refugee backgrounds who had first visited the practice within the previous 12 months. The mixed effects model found the proportion of these patients with a RHA in the previous 6 months increased from 19% [95%CI 19%-20%] to 27% [95%CI 27-28%] (OR 1.81 [95%CI 1.37-2.39] following the intervention. Refugee health assessment rates were also higher in practices that were larger (≥5 full-time-equivalent GPs), had received refugee health training in the last 12 months, and had a higher baseline use of RHAs.

Implications

Our pragmatic intervention is the first to show that a structured, low intensity outreach facilitation program can improve key components of primary care for refugees. That facilitators were existing staff in health care services has clear implications for future efforts to improve the quality of primary care delivered to this vulnerable population.