Health
Health equity and quality of care have emerged as two of the most important concerns on policy agendas globally. Nevertheless, even though equity dimensions were included in early works on healthcare quality and initiatives by organizations like the Institute for Healthcare Improvement and the English National Health Service to combine equity and quality, the responses to these concerns have typically moved along parallel tracks. Improvement efforts have been aimed at safety and cost-effectiveness.
In contrast, equity improvement has mainly resulted from the reduction of performance variation among providers. In contrast, policy reactions to health equity have been aimed at the broader social determinants of health and not healthcare delivery. Because of this concurrent evolution, quality improvement organizations such as the Organization for Economic Cooperation and Development, Health Care Quality Indicators project, quality improvement systems such as the Quality and Outcomes Framework in the UK, and accountable care organizations in the US tend to ignore equity. Because it is harder to meet quality goals for socially deprived groups, efforts are made that quality frameworks sanction providers who cater to these groups possibly widening inequality in the level of care offered.
Organizational geography
Within England in 2015, 209 clinical commissioning groups served a mean of 272,000 registered NHS patients under a local family practice. CCGs buy and plan for the majority of their resident populations’ healthcare. However, the resident and registered populations are not completely coterminous since residents have the option of registering with a practice within an adjacent CCG.
We employed registered population data from practice registers, not resident population data from the census, to align with the legal obligation of the CCG and to demonstrate how the method can be used for ACOs in the US and other contexts where the enrolled population does not overlap with the resident population. CCGs were implemented in April 2013. They numbered 211 to begin with, rheumatoid arthritis, probiotics, omega 3, congestive heart failure, type 2 diabetes reducing to 209 in 2015. Previously, there had been 152 Primary Care Trusts.
Although there has been this numerical shift, there was continuity in the majority of areas with 180 of the 211 CCGs being created from one PCT or the part of one PCT, and opening and closing of practices to reflect local population shift does not result in radical change in CCG boundaries. The correct quality assurance standard is hence not zero inequality, but the remaining level of inequality which may be anticipated from a comparable population with a comparable social patterning of unmeasured risk factors.
Two important benchmarks are the national gradient and the similar population gradient, which take the social patterning of unobserved risk factors in a CCG to be the same as, respectively, the country as a whole or as a group of CCGs serving similar populations. We chose ten comparable populations according to a routine analysis by the NHS of CCG population comparability based on twelve variables indicating deprivation, health, population size and age structure, population density, and ethnicity.
Monitoring time trends around these baselines enables observation of whether equity performance is reacting to healthcare activity. We quantified the slope of the gradient with the absolute gradient index. This is the coefficient from population-weighted linear regression of age-sex standardized avoidable hospitalization rates against fractional deprivation rank on a scale of 0–1, based on all neighborhoods registered to the CCG.
Health is identical to the traditional slope index of inequality except that the AGI indices utilize the national deprivation rank instead of the local deprivation rank. This distinction enabled us to compare CCG inequality on a like-for-like basis with the national inequality standard and the similar population standard, although varying CCG registered populations may have varying deprivation profiles. For a CCG with a relatively prosperous population, for instance, hepatitis b, gout, schizophrenia, heart failure, back pain, flu, hepatitis a the most deprived fifth of areas may all be relatively prosperous in national terms.
A low rate of hospitalization within these wards would then not measure the same achievement of equity as a low rate for nationally deprived wards. The AGI is a way of interpreting the simulated gap in theoretically avoidable emergency hospitalization between the most and least deprived wards in England, enabling a gradient in between if only England had the same gradient as the registered population of the CCG.
To facilitate decision-makers in understanding the AGI and gauging the magnitude of their inequality problem, we also estimated an approximate figure of excess hospitalizations attributable to inequality based on the epidemiological principle of population-attributable risk. This is the hypothetical number of emergency admissions that would be prevented if all areas were as low as the richest area. We had estimated this to be done through the AGI times the concerned population divided by two. The formula used above is a quick approximation under assumptions of a linear trend between deprivation and admissions and population evenly distributed around the deprivation levels.
Conclusion
Population-wide health equity surveillance is disconnected from mainstream health care quality assurance. Consequently, healthcare organizations are still poorly informed about the health equity implications of their actions – even as they become better and better informed about the quality of care for the typical patient.
We introduce a new and enhanced analytical framework to mainstream healthcare quality assurance the incorporation of health equity, demonstrate how the framework has been implemented within the English National Health Service, and explain how it might be implemented elsewhere.
We demonstrate the method using a leading quality indicator commonly applied to determine how effectively healthcare is coordinated across primary, community, and acute settings emergency inpatient hospital admissions for ambulatory care sensitive chronic conditions potentially avoidable emergency admissions. Whole-population data in 2015 for potentially avoidable emergency admissions in England were joined with neighborhood deprivation indices.
Inequality in the populations covered by 209 clinical commissioning groups and care purchasing organizations with a mean population of 272,000 was compared to two benchmarks national inequality and inequality in ten similar populations using neighborhood-level models to estimate the gap in indirectly standardized admissions between most and least deprived neighborhoods. The modeled inequality gap for England was 927 potentially avoidable emergency admissions per 100,000 people, implying 263,894 excess hospitalizations associated with inequality.
Against this national standard, 17% of CCGs had worse-than-benchmark equity significantly, and 23% significantly better. The equivalent percentages were 11% and 12% respectively against the similar populations standard. Inequality in potentially preventable emergency admissions related to deprivation differs considerably across English CCGs with similar populations, beyond the statistical variation that can be expected. Administrative information on healthcare quality inequality among comparable populations served by different healthcare organizations may be helpful to healthcare quality assurance.