Aims Health insurance claims (HIC) databases in the Netherlands capture unselected patient populations, which makes them suitable for epidemiological research on sex differences. Based on a HIC database, we aimed to reveal sex differences in heart failure (HF) outcomes, with particular focus on co‐morbidities and medication. Methods and results The Achmea HIC database included 14 517 men and 11 259 (45%) women with a diagnosis treatment code for chronic HF by January 2015. We related their sex, co‐morbidities, and medication adherence (medication possession rate >0.8) with the primary endpoint (PE) of all‐cause mortality or HF admission during a median follow‐up of 3.3 years, using Cox regression. Median age of men and women was 72 and 76 years, respectively. Prevalence of co‐morbidities and use of disease‐modifying drugs was higher in men; however, medication adherence was similar. At the end of follow‐up, 35.1% men and 31.8% women had reached the PE. The adjusted hazard ratio for men was 1.25 (95% confidence interval: 1.19–1.30). A broad range of co‐morbidities was associated with the PE. Overall, these associations were stronger in women than in men, particularly for renal insufficiency, chronic obstructive pulmonary disease/asthma, and diabetes. Non‐adherence to disease‐modifying drugs was related with a higher incidence of the PE, with similar effects between sexes. Conclusions In a representative sample of the Dutch population, as captured in a HIC database, men with chronic HF had a 25% higher incidence of death or HF admission than women. The impact of co‐morbidities on the outcome was sex dependent, while medication adherence was not.

Background Accurately predicting which patients with chronic heart failure (CHF) are particularly vulnerable for adverse outcomes is of crucial importance to support clinical decision making. The goal of the current study was to examine the predictive value on long term heart failure (HF) hospitalisation and all-cause mortality in CHF patients, by exploring and exploiting machine learning (ML) and traditional statistical techniques on a Dutch health insurance claims database. Methods Our study population consisted of 25,776 patients with a CHF diagnosis code between 2012 and 2014 and one year and three years follow-up HF hospitalisation (1446 and 3220 patients respectively) and all-cause mortality (2434 and 7882 patients respectively) were measured from 2015 to 2018. The area under the receiver operating characteristic (ROC) curve (AUC) was calculated after modelling the data using Logistic Regression, Random Forest, Elastic Net regression and Neural Networks. Results AUC rates ranged from 0.710 to 0.732 for 1-year HF hospitalisation, 0.705–0.733 for 3-years HF hospitalisation, 0.765–0.787 for 1-year mortality and 0.764–0.791 for 3-years mortality. Elastic Net performed best for all endpoints. Differences between techniques were small and only statistically significant between Elastic Net and Logistic Regression compared with Random Forest for 3-years HF hospitalisation. Conclusion In this study based on a health insurance claims database we found clear predictive value for predicting long-term HF hospitalisation and mortality of CHF patients by using ML techniques compared to traditional statistics.