Development and validation of a brief assessment of normative health and health-related social needs using the Simple Segmentation Tool

Abstract

Objectives: Population segmentation provides a promising solution to address patients’ complex needs to provide “whole person” care. The primary objective of this study is to create an expert-based algorithm based on combinations of medical and social characteristics derived from the Simple Segmentation Tool (SST), that are indicative of high value health and health-related social service (HASS) needs for an elderly population. The secondary objective was to examine the association between failing to meet the HASS needs 3-months post hospital discharge suggested by the algorithm and adverse outcomes over the ensuing year. Design & setting: Based on a parsimonious set of 10 patient characteristics identified in the SST, a representative expert panel was engaged using the Modified Appropriateness Methodology (MAM). A prospective study was then performed on patients admitted to the Singapore General Hospital, using HASS needs identified at discharge and met needs at 3 months post-discharge follow-up of services received, to assess whether unmet needs were associated with higher adverse outcomes in the year following discharge. The primary outcome of interest was time to all-cause mortality over 12-months post-discharge and was assessed with Cox regression analysis. Results: The MAM exercise resulted in 12 normatively defined high value services, using a combination of patients’ medical and social characteristics based on the SST, as well as a list of means of providing those service needs. The all-cause mortality hazard ratio of having at least one unmet need versus having all needs met for individuals deemed to be chronically symptomatic at discharge was 1.949, (95% CI: 0.99 – 3.84, and p = 0.05), while for those who were either healthy or only had asymptomatic chronic conditions the all-cause mortality ratio of having at least one unmet need versus having all needs met was 0.28 (95% CI = 0.06–1.27 and p-value = 0.10). The hazard ratio for ED visits and hospital readmission were above one but did not reach level of 95% confidence level. Conclusion: The SST methodology provides a practical way to assess HASS needs that are predictive of mortality when needs are not met. It could serve as a screening tool to identify individuals who are likely to benefit from detailed care planning and follow-up.