Abstract
Dear Editor in Chief,
We read with interest the correspondence from Gilbert-Kawai and Syratt regarding the role of socioeconomic status in ICNARC’s risk prediction modelling. 1
At ICNARC, we have had a long-standing interest in the contribution of socioeconomic status and deprivation to patient outcomes. In 2004, we published on data from the first 5 years of the Case Mix Programme using Carstairs deprivation scores, and found a significant socioeconomic gradient of greater deprivation being associated with higher hospital mortality for admissions following elective surgery, but not for emergency surgical or non-surgical admissions. 2 In 2010, using the Index of Multiple Deprivation, we found a similar socioeconomic gradient in hospital mortality across all admission types. 3
With this knowledge, deprivation was one of the factors we explored when we came to redevelop our risk prediction model for adult critical care in 2015. 4 We were therefore surprised when deprivation made only a very small contribution to model performance and did not reach the threshold for inclusion in the final risk model. We have continued to re-evaluate deprivation as a potential predictor in our regular updates to the risk model, and also when modelling other outcomes, such as longer-term mortality. We are always aiming to improve our models to ensure comparisons of providers are as fair as possible, and we will therefore keep the inclusion of deprivation under regular review.
We agree with McHenry et al. 5 that it is vital that we continue to seek to understand the complex interactions between socioeconomic status and critical illness, and to address the structural inequalities that perpetuate these.
