How Life Sciences is Leveraging Machine Learning - 5 Use Cases in 5 Minutes

In our current edition of “5 in 5,” Mike Munsell, Panalgo's Director of Research, takes a look at five recent studies that used machine learning (ML) to analyze real-world data. These cases cover everything from identifying high-risk patients and generating personalized healthcare plans to guiding treatment decisions and informing hospital resource allocation. Different use cases, but all leveraging the power of ML to examine RWD and generate predictive insights.

Read on to see the five chosen for this edition.

1. Development and Evaluation of a Predictive Algorithm for Unsatisfactory Response Among Patients with Pulmonary Arterial Hypertension using Health Insurance Claims Data

Janssen Scientific Affairs used machine learning-based survival modeling to predict sub-optimal treatment response among patients with pulmonary arterial hypertension. From the top predictors, a simplified risk score containing seven variables was developed as a generalizable tool for identifying high risk patients who may be candidates for combination therapy. Click here for the study.

2. Prediction of Discharge Status and Readmissions after Resection of Intradural Spinal Tumors

Stanford University School of Medicine used claims data and a regularized logistic regression framework to identify patients with intradural spinal tumors at high risk of 90-day postresection readmission. Patients with the highest predicted risk from the machine learning algorithm were readmitted 7 times as often as those with the lowest predicted risk. Click here for the study.

3. Producing Personalized Statin Treatment Plans to Optimize Clinical Outcomes Using Big Data and Machine Learning

Premera Blue Cross, OptumLabs & University of Minnesota leveraged machine learning to estimate patient-level statin treatment plans that minimized the risk of discontinuation and statin-associated symptoms relative to standard practice. The results demonstrate the feasibility of using machine learning to generate proactive personalized healthcare treatment plans. Click here for the study. 

4. Risk Factors Associated with Skeletal-Related Events Following Discontinuation of Denosumab Treatment Among Patients with Bone Metastases from Solid Tumors: A Real-World Machine Learning Approach

Amgen Inc. & the University of Pennsylvania identified novel predictors of skeletal-related events after treatment discontinuation among patients with bone metastases from solid tumors. Primary risk factors included treatment duration and monthly resource utilization prior to discontinuation. Once validated in a clinical setting, the results can guide treatment persistence decisions and potentially improve patient outcomes. Click here for the study.

5. Predicting In-Hospital Length of Stay: A Two-Stage Modeling Approach to Account for Highly Skewed Data

Duke University used a two-stage modeling approach and EHR data to predict in-hospital length of stay for elective procedures. Predictions using random forest were accurate within 16 hours for patients with lengths of stay stay less than 4 days. The algorithm is intended to improve their clinical support tool for making scheduling decisions during times of constrained hospital resources. Click here for the study.

-

If you would like to learn more about how you can leverage machine learning analytics with the IHD Data Science module, contact us today.