(298) Predicting Individual Risk of Muscle Disorders in Patients Eligible for Statin Treatment: Derivation and External Validation of STRATIFY-StatinMD Prediction Model

Friday, August 26, 2022

8:00 AM – 6:00 PM CEST

Location: Hall C

Publication Number: 629

Presenting Author(s)

Ting Cai, MPH
– DPhil Researcher, Nuffield Department of Primary Care Health Sciences, University of Oxford

Co-Author(s)

Constantinos Koshiaris
– Nuffield Department of Primary Care Health Sciences, University of Oxford, United Kingdom
Jennifer A. Hirst
– Nuffield Department of Primary Care Health Sciences, University of Oxford, United Kingdom
F D Richard Hobbs
– Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK, United Kingdom
James P. Sheppard
– Nuffield Department of Primary Care Health Sciences, University of Oxford, United Kingdom

Background: Concerns about serious muscle-related adverse events have posed a dilemma when considering statin prescription for prevention of cardiovascular disease, which may lead to inappropriate treatment withdrawal or hesitation.

Objectives: This study aimed to develop a prediction model for an individual’s risk of muscle disorders to support clinical decision making on statin treatment, as part of a research project for STRAtifying Treatments In the multi-morbid Frail elderlY (STRATIFY).

Methods: An observational cohort study was designed to derive the model, using electronic healthcare records from the Clinical Practice Research Datalink (CPRD) GOLD database in the UK. Males aged over 50 and females aged over 60, who were potentially eligible for statin treatment, were followed-up for hospitalisation or death with a diagnosis of muscle disorders within ten years. Patient characteristics, co-morbidities and medications including statins were selected as predictors based on the literature and expert opinions. The model was derived using the Fine-Gray sub-distribution hazards approach, taking into account the competing risk. External validation in the CPRD Aurum database examined the model’s discrimination using the C statistic and the calibration using the observed/expected ratio (O/E) and calibration slope.

Results: The cohort for model derivation included 1,785,207 patients, with a mean age of 64 and 44% females. Patients prescribed statins were predicted to have a higher risk of muscle disorders. Other important predictors that increased the risk of muscle disorders included female sex, obesity, frailty, previous muscle problems, vitamin D deficiency, mild liver diseases and concomitant myotoxic medications. External validation in 3,889,504 patients showed a good discrimination (C statistic 0.7821, 95%CI: 0.7818 – 0.7824) and calibration (O/E 0.87, 95%CI: 0.85 – 0.88; calibration slope 1.06, 95%CI: 1.05 – 1.07) of the model for the ten-year risk prediction. The model performance was similar for the prediction at one and five years.

Conclusions: This model uses routinely available patient characteristics and medical history to accurately predict an individual’s risk of muscle disorders. It could assist communication between physicians and patients about the benefit-harm trade-off of statin treatment and help make well-informed shared clinical decisions.