Tom Love, Director, Sapere Research Group
Thursday, 18 May 2017, 12.30 – 1.30 pm
GBLT4, Lecture Theatre 4, Old Govt Building, Pipitea Campus,
Managing demand for unscheduled care is a major problem for healthcare systems, with frequent attenders often singled out as exceptional. Complex systems, which may include healthcare, typically include apparently exceptional events within their natural, heavy-tailed, distributions. We tested the hypothesis that the "problem" of frequent attenders could be understood as a property of complex systems. We examined statistical features of complex systems in the distribution of unscheduled care episodes per patient in three ways. (1) Systematic review of published distributions of attendance; (2) Fitting heavy tailed inverse power law distributions to attendances at a primary care out of hours service in Scotland (724,921 patients) and an Emergency Department in New Zealand (60,106); (3) Examination of the distribution of bursts in patients’ use of unscheduled care within these datasets. The review included data from 25 emergency department studies and two Primary Care Out of Hours studies. Despite heterogeneity of settings, all studies showed similar, heavy tailed, distributions of attendance numbers. Both the primary care and emergency department datasets showed good fit to an inverse power law for between 3 and 30 episodes per year. Consultations occurred in bursts, the distribution of which was also heavy tailed. Unscheduled care displays statistical features of complex systems. Frequent attending patients are a universal feature of such systems and should be considered as part of a system rather than in isolation. Measures to optimise healthcare systems should address system-wide behaviours rather than targeting individual cases.
Tom Love is a health economist working with Sapere Research Group. He is an Adjunct Senior Research Fellow with the Department of Primary Health Care and General Practice at the University of Otago Wellington. He has a particular interest in primary care economics and policy, and in phenomena arising from practice variation.
You are welcome to bring your colleagues
RSVPS are not required
For further information contact: firstname.lastname@example.org