City-wide Analysis of Electronic Health Records Reveals Gender and Age Biases in the Administration of Known Drug-Drug Interactions

Rion Brattig Correia, Luciana P. de Araújo, Mauro M. Mattos, Luis M. Rocha

Go to the profile of Nature Communications
Jan 04, 2019
0
0

Received Date: 6th December 18

The occurrence of drug-drug-interactions (DDI) from multiple drug prescriptions is a serious problem, both for individuals and health-care systems, since patients with complications due to DDI are likely to re-enter the system at a costlier level. We present a large-scale longitudinal study of the DDI phenomenon at the primary- and secondary-care level using electronic health records (EHR) from the city of Blumenau in Southern Brazil (pop. ~340,000). This is the first study of DDI we are aware of that follows an entire city longitudinally for 18 months. We found that 181 distinct drug pairs known to interact were dispensed concomitantly to 12% of the patients in the city's public health-care system. Further, 4% of the patients were dispensed major DDI combinations, likely to result in very serious adverse reactions and costs we estimate to be larger than previously reported in smaller studies. DDI results are integrated into associative networks for inference and visualization, revealing key medications and interactions. Analysis of the large EHR data set reveals that women have a 60% increased risk of DDI as compared to men; the increase becomes 90% when only major DDI are considered. Furthermore, DDI risk increases substantially with age. Patients aged 70-79 years have a 34% risk of DDI when they are prescribed two or more drugs concomitantly. In contrast, this risk is less than 10% for patients under 40 years of age and negligible for children under 14. Interestingly, a null model demonstrates that age and women-specific risks from increased polypharmacy far exceed expectations in those populations. This suggests that social and biological factors are at play. Finally, we demonstrate that machine learning classifiers accurately predict patients likely to be administered DDI given their history of prescribed drugs, gender, and age (MCC=.7,AUC=.97) . These results demonstrate that accurate warning systems for known DDI can be devised for health-care systems. Implementation can lead to substantial reduction of DDI-related adverse reactions and health-care savings.

Read in full at arXiv.

This is an abstract of a preprint hosted on an independent third party site. It has not been peer reviewed but is currently under consideration at Nature Communications.

Go to the profile of Nature Communications

Nature Communications

Nature Research, Springer Nature