Targeting emerging risk individuals for care management interventions

Abstract

Introduction: Most care management programmes focus on managing disease to prevent further progression and control costs. To this end, predictive models used for identification of at-risk patients have typically focused on those at the top end of the proverbial population pyramid. This research defines a new at-risk population as lives at lower current morbidity levels than those usually identified for care management but who are responsible for increasing population morbidity and cost. This paper builds on work presented at the 2014 Johns Hopkins ACG System ® conference where it was awarded the inaugural Barbara Starfield Award for its proactive approach and potential to impact population health. Policy objective: Through our research, we propose that health policy includes a stronger focus towards managing lives at lower current levels of morbidity as these lives present the largest financial risk to healthcare funders in the long term. We show that to curb costs, a stronger focus is needed on preventing chronic disease and preventing multi-morbidity, rather than simply focussing on managing existing chronic conditions. Targeted population: We have constructed a predictive model that identifies individuals at lower levels of current morbidity who are at risk of moving to a higher morbidity group in the next year based on demographic, clinical and behavioural risk factors. Longitudinal analyses show that the increase in morbidity is sustained after the initial “jump” in morbidity level. The aim of the proposed care management programme is to prevent this increase in morbidity from occurring. The clinical profile of the targeted population indicates the early prevalence of chronic conditions for which self-care is expected to be consequential. Approximately 50% of the targeted population have a chronic condition requiring an intervention to manage existing disease and prevent multi-morbidity. For those without chronic conditions protecting health and preventing disease is required. Highlights: Aiming care management towards individuals at relatively low current morbidity levels demands a shift away from traditional care management which tends to generally focus on a small percentage of lives that are already highly morbid. The emerging risk group is, however, much larger and will require new innovative interventions. A randomised controlled pilot study currently testing our implementation of an intervention involving targeted and personalised communication to patients is in its third year, with some promising early results. Transferability: While intervention mechanisms are likely to differ between countries and health systems, we believe that the emerging risk principles are fully transferable, namely: a predictive model that identifies individuals at-risk early enabling more efficient programmes; and an intervention approach focused on individual behaviour change and improving healthcare use behaviour in large populations. Conclusion: Given the increase in morbidity and costs experienced by health systems worldwide, an approach aimed at preventing morbidity is required. Using predictive modelling, a new at-risk population can be identified, where risk is defined in terms of increasing morbidity and cost. This is a new approach which will require investment from funders of healthcare with the aim of realising long-term savings. This will require a shift in thinking, but we believe the evidence supporting this approach is overwhelming.