A Person-Centred Approach to Performance Measurement in the Health Svstem

Abstract

Health systems strive to improve health outcomes in the populations they serve. In Australia, a national health system performance framework supports this aim. Review of performance measures showed a focus on organisational activity rather than outcomes for people. South Australia (SA) also set strategic targets for improved healthy life expectancy as influenced by: premature mortality; health related quality of life (HRQoL); and, potentially preventable hospitalisation (PPH). There are unmet information needs and capacity for improvement in the application of each of these measures. Aims This thesis aims to help inform system improvement by reorienting performance measurement toward outcomes of importance to people receiving healthcare – so called ‘person-centred’ measures. The thesis aims to provide empirical examples that help: i. Reframe premature mortality measures to account for survival time from disease detection until death; ii. Extend morbidity measurement to describe a person’s self-reported state of health; and, iii. Enhance enumeration of people experiencing PPH in emergency departments (EDs) and as admitted inpatients. Methods Four studies stem from the candidate’s projects in SA: monitoring summary population health; piloting an advanced cancer data system; steering the first Aboriginal specific population survey; and, quantifying individuals experiencing PPH. Study one introduces a new method that quantifies mortality related cancer burden using an example based on cancer registrations among Aboriginal and non-Aboriginal cohorts matched one-to-one on sex, year of birth, primary cancer site and year of diagnosis. Cancer burden is expressed as the PREmature Mortality to IncidencE Ratio (PREMIER), the ratio of years of life expectancy lost due to cancer against life expectancy years at risk at time of cancer diagnosis for each person. Study two presents the first, self-reported HRQoL utility results by Aboriginal South Australians. Population weighted HRQoL was measured using SF-6D and SF-12 version 2 in face-to-face interviews. Analyses describe relationships between HRQoL and respondent characteristics, and the characteristics of interviewees completing HRQoL questions. Studies three and four consider ED and inpatient PPH respectively. Those studies extend current reporting practices by shifting analyses from PPH as a proportion of activity, to a person-centred approach counting individuals experiencing PPH and the frequency of their events. Both studies draw on person-linked public hospital records within a period prevalence study design. Study three compares ED presentations among Refugee and Asylum Seeker Countries of birth (RASC); Aboriginal; those aged 75 years or more and all other adults. Study four determines disparities in rates, length of stay (LOS) and hospital costs of PPH for chronic conditions among Aboriginal and non-Aboriginal people. Results Study one included records for 777 Aboriginal people diagnosed with cancer from 1990 to 2010. Aboriginal people (n=777) had 57% (95%CI 52%-60%) more scope for improved cancer mortality outcomes two years after diagnosis compared to non-Aboriginal people of equivalent age, sex, diagnosis year and cancer site. PREMIER informs interventions by identifying people with greatest capacity to benefit from earlier detection, treatment and reduced premature mortality. Study two showed substantial variation in self-reported HRQoL among 399 Aboriginal people in 2010/11. For example, average SF-6D results varied from 0.82 (95%CIs 0.81-0.83) among those with no chronic conditions to 0.63 (95%CIs 0.59-0.67) where 3 or more conditions were reported. Comparatively less responding to HRQoL questions was evident among people speaking Aboriginal languages, in non-urban settings, and with multi-morbidities. Further developing culturally safe, self-reporting HRQoL instruments may improve participation by vulnerable and health compromised community members. Study three’s comparisons among adult residents attending EDs in 2005–2006 to 2010–2011 showed greatest disparities in GP-Type presentations among people from RASC compared to non-Aboriginal residents aged less than 75 years (423.7 and 240.1 persons per 1,000 population respectively). Study four’s inpatient PPH for chronic conditions showed Aboriginal people experienced more first-time events compared to others (11.5 and 6.2 per 1,000 persons per year respectively) and substantially longer, total length of stay (11.7 versus 9.0 days). Improved understanding of peoples’ PPH informs tailored services addressing primary healthcare needs. Conclusion The studies assembled in this thesis help align performance measurement with outcomes for people and provide support for system improvement and health reform. While the labour-intensive collaborations necessary may limit development, current opportunities for advancing research within government agencies are discussed. Australia’s health system performance measures remain underdeveloped. This thesis contributes to addressing that need by focussing attention on the people the system exists to serve – effectively, efficiently and equitably.