Research

Algorithms are being increasingly used in public health and medical research in treatment planning, public health surveillance, and medical decision-making. While there is a large literature highlighting the prevalence of algorithmic bias, it is difficult to identify how and to what extent patients are provided with disparate recommendations resulting from these algorithms. Disparate recommendations refer to the differences in treatments for a particular individual or group. They can sometimes lead to an increase in mortality or adverse outcomes for these groups/individuals, ultimately leading to less-than-optimal care. It is also difficult to recognize when disparate recommendations result in outcome disparities.
Consequently, it is challenging to mitigate algorithmic bias because (1) it is hard to determine at which point in the model development and decision-making process algorithmic bias occurs, and (2) it is unclear how these algorithmic biases should be mitigated to improve patient outcomes. My research aims to detect and explain when bias occurs and mitigate any algorithmic bias that is present, not only leading to improved fairness within care but also improving optimality. I not only focus on personalized health outcomes within maternal care but also population-level health outcomes within opioid surveillance. I leverage tools from statistics and machine learning, particularly fair machine learning and interpretability.

Submitted Papers

A.K.McNealey, M.E.Meredith, G.G Garcia, S.L. Boulet, K.K. Stanhope, M.H. Platner, L.N. Steimle."Recommendations for vaginal birth following cesarean using historic and race-blind risk calculators" (Under Review)

Publications

Soto, Esteban A., Andrea Hernandez-Guzman, Alexander Vizcarrondo-Ortega, McNealey, A. and Lisa B. Bosman. "Solar Energy Implementation for Health-Care Facilities in Developing and Underdeveloped Countries: Overview, Opportunities, and Challenges." Energies 15, no. 22 (2022): 8602.

Book Chapters

McNealey, A., Marrero, W.J., Steimle, L.N., Garcia, GG.P. (2023). Optimizing Interpretable Treatment and Screening Policies in Healthcare. In: Pardalos, P.M., Prokopyev, O.A. (eds) Encyclopedia of Optimization. Springer, Cham. https://doi.org/10.1007/978-3-030-54621-2_866-1

Presentations

A.K.McNealey, M.E.Meredith, G.G Garcia, L.N. Steimle, K.K. Stanhope, M.H. Platner, S.L. Boulet. "Race-Blind

Calculator Still Disadvantages Black Patients”. Graduate Oral Presentation, 1st Place in Data Science,Physiology,       and Health. Emerging Researchers National Conference in STEM, Washington D.C. March 14-17th, 2024.

A.K.McNealey, M.E.Meredith, G.G Garcia, L.N. Steimle, K.K. Stanhope, M.H. Platner, S.L. Boulet. "Favorability for vaginal birth after Cesarean among Black women using a new race-blind risk calculator." Poster presentation, Southeast Regional Clinical and Translational Science Conference, Pine Mountain,GA., February 27-29 2024.

McNealey, A. “Predicted likelihood of successful vaginal birth after Cesarean using a race-unaware calculator in a safety net hospital”. Georgia Tech Health Systems - The Next Generation Forum 2023

McNealey, A. Optimizing Solar Energy ROI: A Case for System-Level Monitoring Dashboards to Validate Performance Warranties. Purdue University Fall Undergraduate Research Expo. (2022)

McNealey, A. Solar Energy Applications for Healthcare Facilities in Underdeveloped Countries. Institute of Industrial and Systems Engineers Mid-Atlantic Regional Conference Paper Competition. (2022)

Conference Proceedings 

McNealey, A. Esteban A. Soto, and Lisa B. Bosman. "A Case for Solar Energy System Dashboards to Validate Performance Warranties."(2022). Proceedings of the 7th North American International Conference on Industrial Engineering and Operations Management, Orlando, Florida, USA, June 12-14, 2022