Supplementary Material for: General Practitioner coordinated multidisciplinary care improves long-term survival following stroke with variation by impairment

Background: Australian Medicare funded policies to support primary care physicians (GPs) to coordinate multidisciplinary care (MDC) with other healthcare providers have potential to benefit survivors of stroke/transient ischemic attack (TIA). However, the effectiveness of these policies is unknown. We aimed to determine the population effect of such policies in improving long-term outcomes following stroke/TIA, by impairment groupings.

Methods: Target trial emulation using observational data within a cohort of community-dwelling adults with stroke/TIA from the Australian Stroke Clinical Registry (January 2012-June 2015, 42 hospitals). Person-level Medicare, pharmacy, aged care, death, and hospital records were linked. The exposure was ≥1 Medicare GP MDC claim 6-18 months post-stroke. Outcomes were survival and hospitalisations at 19-30 months. Impairment group (minimal, moderate, severe) was classified by latent class analysis of EQ-5D-3L questionnaire data obtained 90-180 days post-stroke. Analysis comprised multivariable, multilevel survival analysis with inverse probability treatment weights (42 covariates).

Results: The cohort comprised 7,255 people with stroke (42% female, median age 71 years, 24% TIA, impairment: 39% minimal, 32% moderate, 29% severe, 29% had a MDC claim). More claims occurred with each increasing level of impairment group: minimal 22%; moderate 30%; severe 37%. Twelve-month mortality was reduced in those with ≥1 MDC claim (compared to those without) in the minimal (adjusted Hazard Ratio (aHR):0.50, 95% CI:0.27, 0.91) and severe (aHR:0.65, 95% CI:0.46, 0.91) impairment groups, but not in the moderate group (aHR:1.31, 95% CI:0.86, 1.99). Compared to those without a claim, hospital presentations were greater in the minimal (aHR:1.30, 95%CI:1.06, 1.59) and moderate impairment groups (aHR:1.40, 95%CI:1.23, 1.60) but not the severe group (aHR:1.05, 95%CI:0.85, 1.30).
Conclusions: Government policy incentives for GP-coordinated MDC were effective at the population level at improving long-term survival outcomes, in those with minimal and severe impairments.