Nationwide prediction of type 2 diabetes comorbidities

Abstract

Identification of individuals at risk of developing disease comorbidities represents an important task in tackling the growing personal and societal burdens associated with chronic diseases. We employed machine learning techniques to investigate to what extent data from longitudinal, nationwide Danish health registers can be used to predict individuals at high risk of developing type 2 diabetes (T2D) comorbidities. Leveraging logistic regression-, random forest- and gradient boosting models and register data spanning hospitalizations, drug prescriptions and contacts with primary care contractors from >200,000 individuals newly diagnosed with T2D, we predicted five-year risk of heart failure (HF), myocardial infarction (MI), stroke (ST), cardiovascular disease (CVD) and chronic kidney disease (CKD). For HF, MI, CVD, and CKD, register-based models outperformed a reference model leveraging canonical individual characteristics by achieving area under the receiver operating characteristic curve improvements of 0.06, 0.03, 0.04, and 0.07, respectively. The top 1,000 patients predicted to be at highest risk exhibited observed incidence ratios exceeding 4.99, 3.52, 1.97 and 4.71 respectively. In summary, prediction of T2D comorbidities utilizing Danish registers led to consistent albeit modest performance improvements over reference models, suggesting that register data could be leveraged to systematically identify individuals at risk of developing disease comorbidities.

Predicting adverse outcomes due to diabetes complications with machine learning using administrative health data

“…It was not our goal for this model to be used for individual level patient care. Our model was trained on data from over 1.5 million patients from Ontario, which is among one of the most diverse populations in the world and, to our knowledge, one of the largest prediction modelling studies that takes into account multiple types of diabetes complications [22][23][24][25][26][27][28][29][30][31][32][33][34][51][52][53] . Our model was also wellcalibrated and showed good discrimination.…”

“…Many prognostic models have been developed for diabetes complications in the clinical setting [22][23][24] , including more recent applications of machine learning approaches [25][26][27][28][29][30][31][32][33][34] . These models generally have made use of rich suites of features (e.g., body mass index, smoking status, biomarkers ranging from commonly ordered lipids to extensive genetic panels) extracted from electronic medical records (EMRs) 25,27,[31][32][33] or clinical trials 28,30 .…”

Development and Validation of a Machine Learning Model Using Administrative Health Data to Predict Onset of Type 2 Diabetes

“…Such an approach could be feasible in countries such as the UK, Australia, New Zealand, and the Scandinavian countries, which have large, administrative databases suitable for linkage. 53 , 54 , 55 , 56 , 57 Furthermore, this approach could also be deployed in populations covered under a singular health insurance system, such as Medicare or private insurers. 58 …”

An Explainable Multimodal Neural Network Architecture for Predicting Epilepsy Comorbidities Based on Administrative Claims Data

“…Machine learning models to predict individualized comorbidity risks of diseases different from epilepsy have recently been published (e.g., Dworzynski et al (2020) and Noh et al (2020)) using clinical routine data from the Danish national registry and hospital electronic health records, respectively. For epilepsy, Glauser et al (2020) proposed an ML model for psychiatric comorbidities based on survey data from 122 patients.…”