# Research at a glance ## Research at a glance ## Statistical Methodology Professor Sharon-Lise Normand is widely regarded for her significant contributions advancing statistical methods in health care research. Her work addresses complex methodological challenges in provider profiling, causal inference, and meta-analysis, focusing on rigorous solutions that are both practically relevant and influential in the evaluation of medical care and policy. ### Provider Profiling In provider profiling, Professor Normand has developed innovative statistical approaches for assessing and comparing the quality of care delivered by hospitals and clinicians. Her research accounts for differences across patient populations and institutional practices, supporting more meaningful and equitable comparisons that inform public reporting and policy initiatives. [Normand SL, et al., 1997 chevron\_right](https://www.jstor.org/stable/2965545) [Bronskill SE, et al., 2002 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/11933034/) [Normand SL, et al., 2016 chevron\_right](https://doi.org/10.1146/annurev-statistics-022513-115617) ![Bivariate comparison of cardiac surgeon performance. ](/sites/g/files/omnuum9296/files/styles/hwp_1_1__480x480/public/2025-09/Bronskill2002_log%20odds%20%282%29.png?h=1c01df14&itok=jW7g9AgL) (Bronskill SE, et al., 2002) ### Causal Inference Professor Normand has advanced the field of causal inference with novel methods tailored for complex health care data and systems. Her work addresses the realities of confounding, multiple treatment options, and heterogeneity, resulting in robust methodologies that are now widely adopted in comparative effectiveness research and health policy evaluation. She continues to play a key role in the development of modern analytic tools in this domain. [Tchernis R, et al., 2005 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/15887310/) [Poulos J, et al., 2024 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/38314950/) [Adhikari S, et al., 2019 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/33568877/) ![Average coverage probability for the ATE over all 15 pairwise comparisons and 1000 simulated datasets](/sites/g/files/omnuum9296/files/styles/hwp_1_1__480x480/public/2025-10/Poulos2024_figure2.jpg?itok=Lc0W4ba-) Average coverage probability for the ATE over all 15 pairwise comparisons and 1000 simulated datasets. Estimator: (pink) TMLE‐multi.; (yellow) TMLE‐bin.; (green) IPTW‐multi.; (blue) IPTW‐bin.; (purple) G‐comp. (Poulos J, et al., 2024) ### Meta-Analysis Through highly influential work in meta-analysis, Professor Normand has expanded the field’s methodological toolkit, enabling more precise and reliable synthesis of complex and incomplete evidence. Her research addresses issues such as rare adverse events, incomplete data, and heterogeneous study designs, enabling more reliable and interpretable conclusions. These advances are widely applied in clinical investigations and health policy settings, and have set new standards for reporting and interpreting meta-analytic findings, directly impacting the quality and transparency of evidence in medicine and health policy. [Normand SL, 1999 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/10070677/) [Bhaumik DK, et al., 2012 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/23734068/) [Gomes M, et al., 2016 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/27090498/) ![Estimated study-specific and pooled estimates for lidocaine example. ](/sites/g/files/omnuum9296/files/styles/hwp_1_1__480x480/public/2025-10/Normand1999_Figure10%282%29.jpg?h=20807ada&itok=qBcvBDZU) (Normand SL, 1999) ## Clinical Applications ### Quality of Care In the area of health care quality, Professor Normand’s research examines variations in outcomes such as mortality and readmission rates, offering frameworks to evaluate performance at the hospital and system level. Her work informs the development of benchmarks, accountability tools, and national quality standards, as well as efforts to reduce disparities in care. [Krumholz HM, et al., 2009 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/19690309/) [Shahian DM, et al., 2010 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/21175315/) [Normand SL, et al., 2022 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/35120879/) ![Estimates of observes-to-expected (O/E) mortality ratios and corresponding 95% confidence intervals (CIs) for discordant centers.](/sites/g/files/omnuum9296/files/styles/hwp_1_1__720x720_scale/public/2025-10/Normand2022_Figure4.jpg?itok=fyTkRwbQ) Estimates of observes-to-expected (O/E) mortality ratios and corresponding 95% confidence intervals (CIs) for discordant centers. (Normand SL, et al., 2022) ### Serious Mental Illness Professor Normand’s research on serious mental illness addresses patterns of care, utilization, and persistent disparities for people with conditions such as schizophrenia. Her studies highlight the role of race, ethnicity, and geography in influencing access to evidence-based treatment and guide ongoing efforts to improve equity and outcomes for vulnerable populations. [Horvitz-Lennon M, et al., 2015 chevron\_right](https://pmc.ncbi.nlm.nih.gov/articles/PMC4600368/) [Horvitz-Lennon M, et al., 2012 chevron\_right](https://pmc.ncbi.nlm.nih.gov/articles/PMC3666934/) [Horvitz-Lennon M, et al., 2009 chevron\_right](https://pubmed.ncbi.nlm.nih.gov/19648189/) ![Distribution of County-Level Minority–White Disparities, by State (2002–2008). (Horvizt-Lennon, 2015)](/sites/g/files/omnuum9296/files/styles/hwp_1_1__480x480/public/2025-10/HorvitzLennon2015_Figure2.jpg?itok=OKhCPt6_) Distribution of County-Level Minority–White Disparities, by State (2002–2008). (Horvitz-Lennon M, et al., 2015) ## Current Projects [### Medicare Policies on Beneficiaries with Dementia ](/medicare-policies-beneficiaries-dementia) [### Improving quality and value of serious mental illness care in Medicaid ](/improving-quality-and-value-serious-mental-illness-care-medicaid) [### Co-Investigator ](/co-investigator) [### Robust Learning Approaches ](/robust-learning-approaches-assessing-effects-and-effect-heterogeneity-real-world) [### Causal Inference Approaches for Profiling ](/modern-analytics-improve-quality-and-outcome-assessments-following-congenital-heart-surgery) [ Past Projects arrow\_circle\_right ](/sts-congenital-heart-surgery-database-analysis)