Integration: the quest for true equity, beyond accessibility

Abstract

Integration: the quest for true equity, beyond accessibilityPlanning the location of healthcare services and funding them appropriately is complex, as it requires several measures that pull in different directions. Normally, more complex services tend to be placed in larger population nucleus, without full consideration of how the population uses each resource. This favors accessibility from a wider area and efficiency, but goes against equity, as it implies longer travel time for part of the population. Also, it complicates integrated care, as the same person may have to be using geographically separate resources, some more local, some more distant, which are harder to coordinate.We posit that we need tools to help discover and evaluate these tradeoffs, and plan for resource allocation + treatment plans. These tools should, at the very least,1) Be able to ingest data from several tiers of the healthcare systems (acute hospitals, emergency services, primary care, intermediate care…), and analyze them jointly.2) Analyze how citizens with different profiles are actually using the different resources, including travel distance, frequency of use, and continuity of attention.3) Analyze in detail the variability within a given coarse classification (such as DRG or ICD codes): Why some patients in the same DRG and for the same procedure leave their zones and why others stay? Are these differences justified by variability within DRG and patient history? Or are they arbitrary patient choices? Or differences in protocol among different centers?We describe our tool ALOE which was co-designed with the Catalan Health Service for this purpose. The tool monitorizes attraction to different services and flows from different geographical areas, and shows and describes the anomalies in flows: Significant numbers of people that go where they shouldn’t, and how these people differ from similar people that stay in their zones.We will describe a few actual cases studied using the tool. Two involve eating disorder problems and autistic spectrum disorders, respectively in a subregion of Catalonia where there seemed to be unjustified disparity in the centers where people were being treated. In another study, we discovered that a number of heart failure patients in a medium-size town had a tendency to go to a larger tertiary hospital over 80 km away, instead of being treated locally. It turns out that these were the people who had neoplasic antecedents, and they routinely went to the tertiary hospital where they are regularly checked for more complex attention.In still another study, we discovered a (partial) explanation why two otherwise similar hospitals had widely differing lengths of stay after knee replacement surgery (4 days vs. 8 days). It turned out that the 4-day hospital is close to a large tertiary hospital, and that most “complex” patients went there, while the 8-day hospital took care of all knee replacement cases in its territory, both the easy and the difficult ones.A conclusion of the use of the tool is that such in-depth analysis is impossible, or highly time-consuming, to perform with the traditional indicator-based business intelligence tools.