OP44 A framework for monitoring nhs equity performance – small area analysis of national administrative data from 2004/5 to 2011/12

Abstract

Background In 2003, tackling health inequalities was made a priority for the NHS as part of a cross-governmental strategy, and in 2012 a duty was placed on those commissioning NHS services to consider reducing inequality in healthcare access and outcomes. We assess progress between 2004/5 and 2011/12, with the aim of developing the first systematic approach to monitoring socioeconomic inequalities in NHS access and outcomes. Methods Indicators of healthcare access and outcomes at different stages of the patient pathway were constructed for all English small areas (2001 LSOAs) from 2004/5 to 2011/12 using GMS, QOF, HES and ONS mortality and population data – (1) GP supply: full time equivalent (FTE) GPs per 100,000 population, excluding registrars and retainers, need-weighted using the Carr-Hill workload adjustment, (2) primary care quality: quality and outcomes framework clinical performance, weighted by public health impact, (3) hospital waiting time: days from referral-to-treatment, allowing for patient-level casemix, (4) amenable hospitalisation: unplanned hospitalisation per 100,000 population for conditions amenable to healthcare, indirectly age-sex-standardised, (5) excess hospital stays: proportion of inpatients with excess length of stay, allowing for patient-level casemix, (6) post-hospital mortality: 12-month mortality after discharge, allowing for patient-level casemix and comorbidity, (7) amenable mortality: deaths from causes amenable to health care per 100,000 population, indirectly age-sex-standardised. Slope and relative indices of inequality were calculated through small-area-level regression using all 32,482 Index of Multiple Deprivation 2010 ranks, with regression-based tests of change over time. Equity “dashboards” were developed to communicate findings to decision makers in a concise form. Results Nationally, all unadjusted relative indices of inequality fell from 2004/5 to 2011/12 (with 95% CIs in brackets, where negative indices represent “pro-poor” inequality): (1) for GP supply from –2.2% [–2.9% to –1.6%] to –9.5% [–10.2% to –8.8%], (2) for primary care quality from 4.1% [3.6% to 4.6%] to 1.1% [0.6% to 1.6%], (3) for hospital waiting time from 3.2% [2% to 4.4%] to 2.7% [1.5% to 3.8%], (5) for excess hospital stays from 13.8% [14.7% to 12.9%] to 8% [9% to 7.1%], (6) for post-hospital mortality from 0.6% [2.3% to –1.2%] to –4.5% [–2.6% to –6.4%], and (7) for amenable mortality from 34% [36.5% to 31.4%] to 11.9% [14.6% to 9.2%]. Conclusion Socioeconomic inequality in healthcare access and outcomes in the English NHS reduced between 2004/5 and 2011/12 in both relative and absolute terms on all our indicators (unadjusted), though all indicators except GP supply and post-hospital mortality continued to exhibit “pro-rich” inequality. The main study limitations are imperfect adjustments for need and risk and measurement of socioeconomic status using neighbourhood deprivation.