Analytical Methods and Reports

The Center for Health Statistics (CHS) collects and analyzes vital records and hospitalization data. Within CHS, The Linkage and Integrated Data Analysis unit and the Data Science and Engineering unit routinely perform linkage and analyses on a variety of topics that interact with these data.

We publish findings, methodological advancements and analyses via DOH reports, posters, videos, and peer-reviewed studies. Several areas of interest have been established with the intent of disseminating information to interested audiences.

Data Linkage Methodology

The Linkage and Integrated Data Analysis unit (LIDA) specializes in large scale machine learning data linkages. The unit emphasizes equitable representation and maximal accuracy in linkage practices.

2026

Improved Birth/CHARS Linkage: CHINCHILLA vs. UW BERD (PDF)

2025

Hyperblock (PDF)

Record Deduplication: Reference Pipeline and Common Pitfalls (PDF)

2024

Supervised Machine Learning Linkage: A Demo and Evaluation of Project ECHIDNA (PDF)

Introduction to Machine Learning, Vimeo Video

2023

Prioritizing Equitable Representation, Sustainability, and Accuracy: The Deployment of Machine Learning Linkage Strategies During the COVID-19 Pandemic (PDF)

2023 CSTE Outstanding Poster Award Winner: Improving COVID-19 Case and Immunization Record Linkage via Non-Probabilistic Machine Learning-Based Classification (PDF)

Data Science, Statistics, and Analyses

CHS performs independent and collaborative analyses that involve vital record and hospitalization data. Reports produced by data scientists span multiple research topics.

2024

2024 NAPHSIS Poster: Hyperdimensional Change Detection for Novel Data Set Exploration and Continuous Unsupervised Monitoring (PDF)

2023

Excess Deaths During the COVID-19 Pandemic and 2021 Heat Dome (PDF)

Understanding the Washington Excess Mortality in 2020 and 2021 Report (PDF)