JMIR Publications

ABSTRACT

Background:

With the amount of available medical data in Electronic Health Records (EHRs) rapidly increasing, perspectives of learning from this data to allow for better personalized care are growing. This study explores the potential of learning from EHRs to enhance decision-making in the Dutch primary care setting, with a focus on personalized treatment options for Type-2 Diabetes Mellitus (T2DM).

Objective:

The primary objective of this study is to investigate the feasibility of employing a personalized treatment data analysis workflow in the Dutch primary context. Specifically, the aim is to adapt and apply a previously published workflow to data from the Nivel primary Care Database (Nivel-PCD), using T2DM as a use case.

Methods:

The workflow consists of various steps: patient data extraction from the Nivel-PCD; training a similarity model; generating a precision cohort of the most similar patients; and analyzing treatment options. The analysis focuses on identifying treatment options that have led to improved outcomes for the precision cohort, particularly in terms of clinical readouts for glycemic control. Statistical significance is assigned to the results that should be taken into account when health professionals utilize this technology in their decision-making processes.

Results:

The analysis of the personalized treatment reveals insights into which treatment options have statistically significant associations with improved clinical outcomes in various clinical situations. This study demonstrates the value of retrospective analysis in integrating and visualizing real evidence from clinically similar patients, aiding physicians in making informed decisions for individual patients.

Conclusions:

The added value and usefulness of these models are highly dependent on the quality and quantity of the available data. While the workflow successfully identifies treatment options with better outcomes for different clinical scenarios, its effectiveness relies heavily on the presence of a substantial and diverse patient pool. The workflow developed in this study contribute to the understanding of how personalized treatment data analysis can enhance decision-making in primary care settings for different clinical situations.