A general practice intervention for people at risk of poor health outcomes: the Flinders QUEST cluster randomised controlled trial and economic evaluation.

Objective: To determine whether a multicomponent general practice intervention cost-effectively improves health outcomes and reduces health service use for patients at high risk of poor health outcomes.

Design, Setting: Clustered randomised controlled trial in general practices in metropolitan Adelaide.

Participants: Three age-based groups of patients identified by their general practitioners as being at high risk of poor health outcomes: children and young people (under 18 years), adults (18-64 years) with two or more chronic diseases, and older people (65 years or more).

Intervention: Enrolment of patients with a preferred GP, longer general practice appointments, and general practice follow-up within seven days of emergency department and hospital care episodes. Intervention practices received payment of $1000 per enrolled participant.

Main Outcome Measures: Primary outcome: change in self-rated health between baseline and 12-month follow-up for control (usual care) and intervention groups.

Secondary Outcomes: numbers of emergency department presentations and hospital admissions, Medicare specialist claims and Pharmaceutical Benefits Scheme (PBS) items supplied, Health Literacy Questionnaire scores, and cost-effectiveness of the intervention (based on the number of quality-adjusted life-years [QALYs] gained over 12 months, derived from EQ-5D-5L utility scores for the two adult groups).

Results: Twenty practices with a total of 92 GPs were recruited, and 1044 eligible patients participated. The intervention did not improve self-rated health (coefficient, -0.29; 95% CI, -2.32 to 1.73), nor did it have significant effects on the numbers of emergency department presentations (incidence rate ratio [IRR], 0.90; 95% CI, 0.69-1.17), hospital admissions (IRR, 0.90; 95% CI, 0.66-1.22), Medicare specialist claims (IRR, 1.00; 95% CI, 0.91-1.09), or PBS items supplied (IRR, 0.99; 95% CI, 0.96-1.03), nor on Health Literacy Questionnaire scores. The intervention was effective in terms of QALYs gained (v usual care: difference, 0.032 QALYs; 95% CI, 0.001-0.063), but the incremental cost-effectiveness ratio was $69 585 (95% CI, $22 968-$116 201) per QALY gained, beyond the willingness-to-pay threshold.

Conclusions: Our multicomponent intervention did not improve self-rated health, health service use, or health literacy. It achieved greater improvement in quality of life than usual care, but not cost-effectively.

Trial Registration: Australian New Zealand Clinical Trials Registry, ACTRN12617001589370 (prospective).