Case Study: Improving Patient Outcomes for a Healthcare Provider

How DSIGHTS helped a major hospital system predict patient readmissions and improve the quality of care.

The Challenge

A large hospital system was facing a high rate of patient readmissions, which was not only costly but also indicated a lower quality of care. The hospital needed a way to identify patients at high risk of readmission so that they could provide targeted interventions and improve patient outcomes. Their existing methods for identifying high-risk patients were manual and subjective, and they were not effective at predicting readmissions.

Our Solution

DSIGHTS was engaged to develop a predictive analytics solution to identify patients at high risk of readmission. Our team of data scientists and healthcare experts worked closely with the hospital to understand their patient population, clinical workflows, and data systems. We then designed and implemented a custom machine learning model that could predict the likelihood of readmission for each patient.

Technology Stack

The solution was built on a secure and compliant technology stack that included:

• Python: For data analysis, model development, and automation.
• PyTorch: For building and training deep learning models.
• Google Cloud Platform (GCP): For secure and compliant cloud infrastructure, including data storage, processing, and model deployment.
• Databricks: For collaborative data science and large-scale data processing.

The Process

Our process involved several key steps:

1. Data Collection and Anonymization: We collected and anonymized electronic health record (EHR) data for thousands of patients, ensuring compliance with all privacy regulations.
2. Feature Engineering: We created new features from the EHR data to improve model accuracy, such as patient demographics, clinical history, and medications.
3. Model Development and Training: We developed and trained a deep learning model to predict the likelihood of readmission for each patient.
4. Model Evaluation and Interpretation: We evaluated the performance of the model and used model interpretation techniques to understand the factors that were most predictive of readmission.
5. Integration and Deployment: We integrated the model with the hospital's EHR system so that clinicians could see the readmission risk score for each patient in real-time.

The Results

The new predictive analytics solution delivered significant improvements for the hospital, including:

• 20% reduction in patient readmissions: This resulted in significant cost savings and improved patient outcomes.
• Improved quality of care: The solution enabled the hospital to provide targeted interventions to high-risk patients, which improved the quality of care and patient satisfaction.
• Increased efficiency: The automated solution freed up clinicians to focus on providing care to patients.

Conclusion

By leveraging the power of predictive analytics and machine learning, DSIGHTS was able to help the hospital system reduce patient readmissions and improve the quality of care. This case study demonstrates the potential of AI to transform the healthcare industry and improve the lives of patients.