Cloud based DevOps Framework for Identifying Risk Factors of Hospital Utilization

2025-04-18

AI Summary

This study addresses the poor reproducibility and limited scalability of analyses on sparse healthcare datasets—such as NHANES—when investigating hospitalization risk factors (e.g., diabetes, obesity, cardiovascular disease). We propose the first modular, generalizable, cloud-native DevOps analytics framework. Methodologically, it integrates CI/CD pipelines (GitHub Actions/Jenkins), cloud platforms (AWS/Azure), the Python scientific stack, data version control, and synthetic data generation to enable automated NHANES data updates, hybrid modeling, and end-to-end analysis. Key contributions include: (1) substantially improved analytical reliability and reproducibility; (2) seamless co-scaling of real and synthetic data; and (3) empirical validation of cross-domain transferability to other sparse-data domains—including environmental science and cybersecurity—establishing a general-purpose infrastructure for multi-domain health data analytics.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Natural Language Processing: Safety and Robustness
• Humans and AI: Crowd Sourcing and Human Computation

Application Category

• Security and Privacy: Data transparency and provenance
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Ethical and legal aspects of web-scale data analysis, uses and collection practices

Abstract

A scalable and reliable system is required to analyze the National Health and Nutrition Examination Survey (NHANES) data efficiently to understand hospital utilization risk factors. This study aims to investigate the integration of continuous integration and deployment (CI/CD) practices in data science workflows, specifically focusing on analyzing NHANES data to identify the prevalence of diabetes, obesity, and cardiovascular diseases. An end-to-end cloud-based DevOps framework is proposed for data analysis which examines risk factors associated with hospital utilization and evaluates key hospital utilization metrics. We have also highlighted the modular structure of the framework that can be generalized for any other domains beyond healthcare. In the framework, an online data update method is provided which can be extended further using both real and synthetic data. As such, the framework can be especially useful for sparse dataset domains such as environmental science, robotics, cybersecurity, and cultural heritage and arts.

Problem

• Analyze NHANES data to identify hospital utilization risk factors
• Integrate CI/CD practices in data science for healthcare analysis
• Propose scalable cloud DevOps framework for multi-domain application

Innovation

• Cloud-based DevOps framework for NHANES data analysis
• Modular structure adaptable to non-healthcare domains
• Online data update method with real and synthetic data