Dominique J Monlezun1, Michael Darden2, Paul Friedlander1, Luis Balart1, Geoffrey Parker3, Joseph F Buell4. 1. Tulane Transplant Institute, Tulane University School of Medicine, New Orleans, LA. 2. Milken Institute School of Public Health, George Washington University, Washington, DC. 3. Thayer School of Engineering, Dartmouth College, Hanover, NH. 4. Tulane Transplant Institute, Tulane University School of Medicine, New Orleans, LA. Electronic address: Jbuell1@tulane.edu.
Abstract
BACKGROUND: Sources of liver transplant disparities are not understood adequately, particularly in terms of race and region. METHODS: Fixed effects multivariate logistic regression augmented by modified forward and backward stepwise regression of transplanted patients from the United Network for Organ Sharing Standard Transplant Analysis and Research database (1985-2016) was performed to assess causal inference of such disparities. RESULTS: In the study sample (N = 258,602), significant disparities in the odds of receiving a liver were found: African Americans odds ratio 1.12 (95% confidence interval, 1.08-1.17), Asians 1.12 (95% confidence interval, 1.07-1.18), females 0.80 (95% confidence interval, 0.78-0.83), and malignancy 1.18 (95% confidence interval, 1.13-1.22). Region 7 (IL, MN, ND, SD, and WI) was set as the reference level since its transplantation rate most closely approximated the sex and race-matched rate of the national post-Share 35 average. Significant racial disparities by region were identified using Caucasian Region 7 as the reference: Hispanic Region 9 (New York, West Vermont) 1.22 (1.02-1.45), Hispanic Region 1 (New England) 1.26 (1.01-1.57), Hispanic Region 4 (Oklahoma, TX) 1.23 (1.05-1.43), and Asian Region 4 (Oklahoma, TX) 1.35 (1.05-1.73). CONCLUSION: Despite numerous adjustments to liver allocation, we identified with causal inference statistics on a large dataset spanning ≥30 years there remain racial and regional overweighting.
BACKGROUND: Sources of liver transplant disparities are not understood adequately, particularly in terms of race and region. METHODS: Fixed effects multivariate logistic regression augmented by modified forward and backward stepwise regression of transplanted patients from the United Network for Organ Sharing Standard Transplant Analysis and Research database (1985-2016) was performed to assess causal inference of such disparities. RESULTS: In the study sample (N = 258,602), significant disparities in the odds of receiving a liver were found: African Americans odds ratio 1.12 (95% confidence interval, 1.08-1.17), Asians 1.12 (95% confidence interval, 1.07-1.18), females 0.80 (95% confidence interval, 0.78-0.83), and malignancy 1.18 (95% confidence interval, 1.13-1.22). Region 7 (IL, MN, ND, SD, and WI) was set as the reference level since its transplantation rate most closely approximated the sex and race-matched rate of the national post-Share 35 average. Significant racial disparities by region were identified using Caucasian Region 7 as the reference: Hispanic Region 9 (New York, West Vermont) 1.22 (1.02-1.45), Hispanic Region 1 (New England) 1.26 (1.01-1.57), Hispanic Region 4 (Oklahoma, TX) 1.23 (1.05-1.43), and Asian Region 4 (Oklahoma, TX) 1.35 (1.05-1.73). CONCLUSION: Despite numerous adjustments to liver allocation, we identified with causal inference statistics on a large dataset spanning ≥30 years there remain racial and regional overweighting.