JMIR Publications

ABSTRACT

Background:

Anticoagulation therapy with Heparin is a frequent treatment in intensive care units, which is monitored by activated partial thromboplastin clotting time (aPTT). It has been demonstrated that reaching an established anticoagulation target within 24 hours is associated with favorable outcomes. However, patients respond to heparin differently and reaching the anticoagulation target can be challenging. Machine learning algorithms may potentially support clinicians with improved dosing recommendations.

Objective:

This study evaluates a range of machine learning algorithms on their capability of predicting the patients’ response to heparin treatment. In the present analysis, we apply, for the first time, a model that considers time series.

Methods:

We extracted patient demographics, laboratory values, dialysis and ecmo treatments, and scores from the hospital information system. We set up a regression task for predicting the aPTT laboratory values 24 hours after continuous heparin infusion and evaluate seven different machine learning models. We consider all data before and within the first 12 hours of continuous heparin infusion as features and predict the aPTT value after 24 hours.

Results:

The distribution of aPTT in our cohort of 5926 hospital admissions is highly skewed. While most patients show aPTT values below 75 s, some outliers show much higher aPTT values. A recurrent neural network that consumes time series of features shows the highest performance on the test set.

Conclusions:

A recurrent neural network that uses time series of features instead of only static and aggregated features, shows the highest performance of predicting aPTT after heparin treatment.