Improving care for chronic conditions may take longer than expected: Evaluation of the Health Links chronic care flexible funding model of care

Abstract

Introduction: Current models of care and funding systems in Australia for managing chronic disease are largely designed to be responsive to single episodes of acute care. The Victorian Department of Health developed a flexible funding model which aimed to improve care for patients at high risk of unplanned hospital admissions, many of whom had chronic and complex conditions. The initiative investigated whether the model could remove several barriers that inhibit integrated models of care and promote innovation to produce better outcomes for patients. Four hospital and health services participated in the trial with three acting as control hospitals. Each of the participating hospitals undertook their own unique intervention, ranging from redesigns of existing services to advanced coaching models using tele-care guides.
 Methods: We undertook an analysis of HealthLinks enrolled intervention patient outcomes compared to a usual care patient cohort. In addition, we compared outcomes from participating individual health services compared to control health services. Outcomes measured were inpatient length of stay, number of inpatient admissions and 30-day readmissions, and emergency department (ED) length of stay and number of ED presentations.
 Administrative datasets of enrolled patients were converted into monthly panels where outcomes were summed to a count per patient per month for the period 12-months prior to the trial start date for participating health services and the 1st of July 2016 for control health services and for 24 months after. 
 Results: A total of 2,400 Healthlinks enrolees received an intervention model of care during the two year trial out of 49,000 enrolees.
 Compared to patients from control hospitals there were significantly fewer ED presentations per month (-0.1) and shorter ED lengths of stay (-40 mins), but more admissions per month( 0.03) and longer inpatient lengths of stay (0.2 days) for patients at intervention hospitals. A similar pattern was observed for individual hospital compared to controls.
 Patterns of hospital use varied by health service for intervention patients compared with usual care patients. Compared with usual care patients there were significant increases in ED length of stay at all hospitals; significantly fewer 30-day readmissions per month (–0.004 ) at Hospital C; significantly more hospital admissions per month at hospitals B (0.2) and C (0.1) ; and significantly shorter inpatient lengths of stay for hospital b (-2 days) but longer for hospital D (0.5 days) compared with usual care patients. 
 Discussion: Most outcomes measured in this evaluation showed little effect for those patients directly involved in an intervention apart from hospital and ED length of stay measures for one health service. Patients from flexibly funded health services seemed to have reduced ED presentations and length of stay in the ED; however, it is unclear whether this is because of flexible funding or differences in outcome trajectories before flexible funding commenced. There is an emerging view that the effects on patient outcomes from chronic disease models of care may be more apparent in the longer term, suggesting health policy takes a similar view when trialling new initiatives.