We leverage one of the largest de-identified linked patient databases in the U.S. to help people live healthier lives. The topics we explore are as broad as the challenges in health care today. These insights inform our action. Here’s a sample of our work.

Study spotlights

"Right size" diabetes care

Learn how we’re developing a new individualized quality measure to manage diabetes better.

Learn more

glucose meter

Stroke prevention

See how real-world data helps us optimize stroke prevention in atrial fibrillation. 

Learn more

bottles of pills


Projects in motion

Conservative therapies for low back pain
APTA, Boston University, UHC

How do initial care patterns for evaluation and treatment of new-onset low back pain, including provider type and service type, influence the risk of short- and long-term opioid use and misuse? Insights may inform new benefit and network designs related to back pain.

STI testing, infections, and treatment
Centers for Disease Control and Prevention

Using linked clinical and claims data, what are the national testing, infection and treatment rates for STIs such as chlamydia, gonorrhea and syphilis? Findings can enhance current STI surveillance and health services research methodology.

Predicting Alzheimer’s and dementia 
University of Maryland, Baltimore; OptumLabs

Can we look back at patient history — years before a dementia diagnosis — to find early warning signs? Through claims-based regression models, machine learning models and natural language processing of clinical data, we hope to predict dementia sooner.

Hearing loss costs and health outcomes 
AARP, Johns Hopkins University, OptumLabs

Looking 2–10 years after diagnosis of hearing loss — a condition that affects 2/3 of Americans over 70 — what are the costs and range of health problems for patients who go uncorrected compared to those without hearing loss?

Diffusing clinical evidence to practice
UMN, Harvard, Yale, Mayo Clinic

Looking at large RCTs and FDA safety warnings that reveal ineffective and unsafe treatments, what factors impact de-implementation across care delivery systems? This question is key for improving patient safety and outcomes while speeding up value-based change in care.

Lung cancer screening trends and impact
American Cancer Society

What is the uptake of 2013 recommendations on annual low-dose Helical Computed Tomography (LDCT) screening for lung cancer? Trends may inform interventions to promote evidence-based screening for the leading cause of cancer death in both men and women.