Development and validation of a prognostic model for predicting the risk of allopurinol-induced severe cutaneous adverse reactions: a retrospective new-user cohort study using linked primary care, hospitalisation, and mortality data
Edoardo Cipolletta, Georgina Nakafero, Davide Rozza, Satveer K. Mahil, Catherine Smith, Richard D Riley, Abhishek Abhishek
Abstract
BACKGROUND: Allopurinol, the most prescribed urate-lowering drug, is a known cause of severe cutaneous adverse reactions. We aimed to develop and validate a model to assess the risk of allopurinol-induced severe cutaneous adverse reactions in adults newly prescribed allopurinol. METHODS: In this retrospective new-user cohort study, we developed and validated a prognostic model using primary care, hospitalisation, and mortality data extracted from the UK Clinical Practice Research Datalink (CPRD) primary care database, for the period Jan 1, 2001, to March 29, 2021. Data from CPRD Aurum was used for model development and data from and CPRD GOLD was used for model validation. Adults (aged ≥18 years) residing in England who were newly prescribed allopurinol were followed up for 100 days to assess whether a severe cutaneous adverse reaction was recorded in hospitalisation or mortality records. Risk predictors included in the model were age, sex, ethnicity, chronic kidney disease stage, initial allopurinol dose, ischaemic heart disease, and heart failure. The primary outcome was to predict the 100-day risk of allopurinol-induced severe cutaneous adverse reactions in people newly prescribed allopurinol. We developed the model using multivariable Cox regression and pseudo-values, followed by penalisation and external validation. We assessed calibration, discrimination, and clinical utility in the risk range of 0·0001 to 0·003. People with lived experience of allopurinol use or gout were not involved in developing this research question, but will be involved in the dissemination of results. FINDINGS: was 0·44 (95% CI 0·20-0·62), and Harrell's C was 0·79 (95% CI 0·71-0·88). The model had clinical utility across the prespecified risk range. INTERPRETATION: We developed and validated a prognostic model for the 100-day risk of an allopurinol-induced severe cutaneous adverse reaction with good predictive performance and clinical utility. This model could be used to inform the choice of urate-lowering drugs. FUNDING: University of Nottingham.