Conformal prediction enables disease course prediction and allows individualized diagnostic uncertainty in multiple sclerosis

Accurate assessment of progression and disease course in multiple sclerosis (MS) is vital for timely and appropriate clinical intervention. The transition from relapsing-remitting MS (RRMS) to secondary progressive MS (SPMS) is gradual and diagnosed retrospectively with a typical delay of three years. To address this diagnostic delay, we developed a predictive model that is able to distinguish between RRMS and SPMS with high accuracy, trained on data from electronic health records collected at routine hospital visits obtained from the Swedish MS Registry containing 22,748 patients with 197,227 hospital visits. To be useful within a clinical setting, we applied conformal prediction to deliver valid measures of uncertainty in predictions at the level of the individual patient. We showed that the model was theoretically and empirically valid, having the highest efficiency at a 92% confidence level, and demonstrated on an external test set that it enables effective prediction of the clinical course of a patient with individual confidence measures. We applied the model to a set of patients who transitioned from RRMS to SPMS during the cohort timeframe and showed that we can accurately predict when patients transition from RRMS to SPMS. We also identified new patients who, with high probability, are in the transition phase from RRMS to SPMS but have not yet received a clinical diagnosis. We conclude that our methodology can assist in monitoring MS disease progression and proactively identify patients undergoing transition to SPMS. An anonymized, publically accessible version of the model is available at . ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement KK acknowledges funding from the Swedish Research Council (grant 2021-02189), FORMAS (grants 2020-01267 and 2023-00905), NEURO Sweden, Region Uppsala (ALF-grant and R&D funds) and Åke Wiberg foundation. OS acknowledges funding from the Swedish Research Council (grants 2020-03731 and 2020-01865), FORMAS (grant 2022-00940), Swedish Cancer Foundation (project 22 2412 Pj 03 H) and the Swedish strategic initiative on e-Science eSSENCE. JB acknowledges funding from the Swedish Research Council (grant 2021-02814), the Swedish Society of Medicine (SLS-593521), the Swedish Society for Medical Research, and the Marianne and Marcus Wallenberg Foundation Authorship. PE is financially supported by the Knut and Alice Wallenberg Foundation as part of the National Bioinformatics Infrastructure Sweden at SciLifeLab. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study was approved by the Ethics Review Authority at Uppsala University (Dnr 2021-00702). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes The data used in the study cannot be shared to protect the privacy of individuals. All the data can be obtained by applying through SMSreg. All the necessary codes used are given in the GitHub repository .