Clin. Gastroenterol. Hepatol. (2018). doi:10.1016/j.cgh.2018.05.002

Lorenz Curves and Gini Coefficient Analyses Indicate Inefficiencies in Esophageal Adenocarcinoma Screening.

Hur C, Zhan T, Thrift AP, Vaughan TL & Feuer, EJ

Abstract

Rates of esophageal adenocarcinoma (EAC) have increased rapidly in the US and much of western Europe, and five year survival continues to be poor. Prevention and early detection efforts for EAC have focused on identifying persons with EAC precursor state, Barrett’s Esophagus (BE), but the survival benefit has been disappointingly low. Precise risk stratification could improve early identification of individuals who will develop EAC, allowing for such high-risk individuals to be targeted for new and evolving preventive and minimally invasive screening technologies. This underlying issue of efficiency is conceptually similar to the Gini coefficient, a widely accepted method to depict the income distribution of a group to assess wealth inequality. A Lorenz curve plots percentage of wealth against the percentage of the population; perfect wealth distribution would result in a slope of 1 and a Gini coefficient of 0 whereas severe wealth inequity results in higher Gini values closer to 1. Some examples of Gini coefficients are shown in Supplementary Figure 1. In this article, we use relative risks of stratified groups to construct Lorenz curves and quantitively analyze the distribution of EAC burden in the US population, for which a higher Gini coefficient indicates more efficient risk stratification.