LAUSR

Background The UK has higher rates of epilepsy deaths among children and young people than comparable countries, and poor care quality is a contributing factor a significant proportion of epilepsy-related deaths. Admission rates are widely used as a proxy measure of care quality in other paediatric conditions. For epilepsy, admission rates vary 4–5 fold between areas in England and the national clinical audit of paediatric epilepsy care (Epilepsy 12) shows wide variation in measures of care quality. It is not known whether better audit performance is associated with lower admission rates. Objective To investigate the degree to which better performance in the national clinical audit is associated with lower admission rates for children and young people (0–19 years) with epilepsy. Methods Independent variables:12 unit-level quality indicators including access to paediatric, specialist nurse and paediatric neurologist expertise.(Round 1=2009–12; Round 2=2014) n=144 units. Dependent variable: unit-level epilepsy admission rate (admissions calculated from Hospital Episode Statistics (HES 2011/12–2013/14); estimated unit catchment populations derived from age-specific, all-cause admissions and population data). Analyses Cross–sectional, multi–level regression models of Round 2 audit data and contemporaneous admission rates. Difference–in–difference analyses comparing time trends in performance measures and admission rates between 2011/12 and 2013/14. All analyses excluded tertiary centres. Results In 2013/14, matched data were available on 8685 epilepsy-related admissions across 144 units. In the longitudinal model, increased involvement of paediatric neurologists was associated with a small relative increase in admission rates over time (β=0.13 (95% CI: 0.03 to 0.22), p=0.007). There was no significant relationship between paediatric neurologist involvement and admission rates in the cross-sectional model (β=−0.04 (−0.14–0.05)). No other perfomance indicator from the Epilepsy12 audit was associated with adjusted, unit-level admission rates on either cross-sectional or longitudinal models. Conclusions Neither cross-sectional nor longitudinal models showed any consistent relationship between performance on the national clinical audit and unit-level admission rates for epilepsy among children and young people. This may reflect inconsistency in how the performance indicators are interpreted and measured, as well as the limitations of admission rates as a proxy for good seizure and quality of life among children and young people with epilepsy.