Abstract
Healthcare should be judged by its equity as well as its quality. Both aspects depend not only on the characteristics of service delivery but also on the research and innovation that ultimately shape them. Conducting a fully-inclusive evaluation of the relationship between enrolment in primary research studies at University College London Hospitals NHS Trust and indices of deprivation, here we demonstrate a quantitative approach to evaluating equity in healthcare research and innovation. We surveyed the geographical locations, aggregated into Lower Layer Super Output Areas (LSOAs), of all England-resident UCLH patients registered as enrolled in primary clinical research studies. We compared the distributions of ten established indices of deprivation across enrolled and non-enrolled areas within Greater London and within a distance-matched subset across England. Bayesian Poisson regression models were used to examine the relation between deprivation and the volume of enrolment standardized by population density and local disease prevalence. A total of 54593 enrolments covered 4401 LSOAs in Greater London and 10150 in England, revealing wide geographical reach. The distributions of deprivation indices were similar between enrolled and non-enrolled areas, exhibiting median differences from 0.26% to 8.73%. Across Greater London, enrolled areas were significantly more deprived on most indices, including the Index of Multiple Deprivation; across England, a more balanced relationship to deprivation emerged. Regression analyses of enrolment volumes yielded weak biases, in favour of greater deprivation for most indices, with little modulation by local disease prevalence. Primary clinical research at UCLH has wide geographical reach. Areas with enrolled patients show similar distributions of established indices of deprivation to those without, both within Greater London, and across distance-matched areas of England. We illustrate a robust approach to quantifying an important aspect of equity in clinical research and provide a flexible set of tools for replicating it across other institutions.
Highlights
Equity of care is a central tenet of medicine
Distributional comparisons We compared the distributions of deprivation indices of Layer Super Output Areas (LSOAs) that included at least one enrolled patient vs none, quantifying the magnitude and significance of any difference for each index, and reporting its direction based on the difference in medians
For all LSOAs within Greater London (Table 1 and Figure 4), the distributions were very similar on visual inspection; statistically indistinguishable in Employment Deprivation, Health Deprivation and Disability, and Income Deprivation Affecting Children Index; and significantly different in all others
Summary
Equity of care is a central tenet of medicine. Its pursuit has conventionally focused on the structures of healthcare delivery[1] rather than the research and innovation that precede them[2]. In our pursuit of closer individuation, widens the field of factors brought to bear on clinical decision making, the effects of such heterogeneities may be magnified, potentially increasing the disparities between those from whom the guiding intelligence is drawn and those to whom it is merely applied[5]. How should equity of care be promoted in innovation? A subpopulation need not be defined by demographics alone: its distinctive characteristics may span a wide array of interacting factors only complex generative statistical models, given sufficient data, could adequately summarise. The problem is analogous to that of estimating heterogeneous treatment effects[7], and is just as difficult to solve
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
