Project Summary

One of PCORI’s goals is to improve the methods that researchers use for patient-centered outcomes research. PCORI funds methods projects like this one to better understand and advance the use of research methods that improve the strength and quality of comparative effectiveness research.

This research project is in progress. PCORI will post the research findings on this page within 90 days after the results are final.

What is the project about?

A randomized controlled trial, or RCT, is often the best way to learn if one treatment works better than another. RCTs assign patients to different treatments by chance. When RCTs are not feasible, researchers can conduct non-randomized studies. But non-randomized studies are more easily affected by bias. Bias occurs when unmeasured factors affect the results of a study. Bias makes it hard to know if the study results are due to the treatment or the unmeasured factors.

To address bias, researchers can use statistical methods called instrumental variables, or IVs. In this study, the research team is using machine learning to develop methods to improve current IV designs. In machine learning, computers use data to learn how to perform different tasks with little or no human input.

How can this project help improve research methods?

Results may help researchers when considering ways to reduce bias when using IV methods in non-randomized studies.

What is the research team doing?

The research team is improving IV methods in three ways. First, the team is using machine learning to create statistical methods that can estimate results with less bias. The team is testing the methods using data on surgical treatments for medical emergencies such as a burst appendix.

Second, the research team is increasing the accuracy of analyses by developing IV methods that use information on patient traits. The team is testing the methods using data on the use of life support for newborns with low birth weights.

Third, the research team is developing more accurate IV methods for use with continuous data. Continuous data are data that have an infinite number of possible values, such as height or weight. The team is testing the methods by estimating the effects of treatment for patients having emergency surgery at trauma centers.

Research methods at a glance

Design Element Description
Goal To develop robust estimation methods for IV designs, including estimation methods for covariate-assisted bounds and IV designs with a continuous instrument
Approach Machine learning

Project Information

Luke Keele, PhD
The Trustees of The University of Pennsylvania
Improved Statistical Methods for IV Designs in CER

Key Dates

December 2021
July 2026


Award Type
Health Conditions Health Conditions These are the broad terms we use to categorize our funded research studies; specific diseases or conditions are included within the appropriate larger category. Note: not all of our funded projects focus on a single disease or condition; some touch on multiple diseases or conditions, research methods, or broader health system interventions. Such projects won’t be listed by a primary disease/condition and so won’t appear if you use this filter tool to find them. View Glossary
State State The state where the project originates, or where the primary institution or organization is located. View Glossary
Last updated: September 26, 2023