Driving R adoption in an NHS information service, barriers and solutions

Abstract

Healthcare Evaluation Data (HED) is an online NHS benchmarking using national data. We provide web-based interactive reporting tools for organisations to compare performance, featuring various statistical modelling approaches and live manipulation of large datasets. We have relied on SAS for many of our models, but are transitioning to R for modelling, analysis work and developing support material. This talk will explain how we are moving from one keen user, to wider adoption, focussing on Setting up R in a ‘locked-down’ NHS environment Best practise for building statistical models on large data sets with limited hardware Spreading use of R through package development and training Encouraging R use for analysis and ad-hoc reporting Open-source principles, information governance and intellectual property Barriers we have faced include scepticism about security risks, unhelpful hardware and network settings, a perception that R was not fit for production, fear of losing SAS at the ‘standard,’ and the learning curve for new R users. We have embarked on a structured training program for our teams and have set up a local R user group. We are aiming to further engage with the R community as user confidence grows.

Date