Comparing patient outcomes after CT scans by raceethnicity and morbidly-matched people scientists from USC Viterbi and NYU Langone have found that treating COVID-19 patients not just requires multidimensional approaches on one end and also a common set of challenges at the other. In a stinging critique of recent scholarship they point out that in the midst of a worldwide pandemic analytics are largely focused on comparing outcomes in ethnic minorities. They point out that only one in six cases are attributed to intentional efforts by the healthcare provider although this figure does not mean that physicians are exceptionally motivated to change their clinical behavior because of race or ethnicity.

You have to become extremely good at scouring through an individuals history and adapting to it. At the end of the day sometimes it could save a detailed frame or elicit a meaningful number of comments. An individuals blood clotting and environmental exposures are essential to understanding hemorrhagic stroke said Viterbi the Alphonse I. and Rochelle B. Sternman Professor of Medicine and director of the USC Norris Comprehensive Cancer Center.

The critique takes aim at the narrow focus on ethnicity in therapy planning. In the lab of Kristine Muller the Nirrfan K. Viterbi Professor of Education Policy and Management at NYU Langone they translated data from individual studies including one published in 2019 in JAMA Network Open. The Cornell investigators then compared outcomes since March 1 2020 among people enrolled in the health systems long-planned postgraduate program in ethnically diverse populations. They looked at outcomes for people who were white and who were black in theory: 6422 black and 1433 white patients.

For this work the scientists compared outcomes for hospitals-matched across raceethnicity-with outcomes among the matching patients for all clinical categories-i.e. 391 on whom data was available from both forecasting and database-based systems.

Viterbi emphasized that quantitative comparison is just one data type of analysis and that qualitative comparisons may be far more important. But she noted that as her team team demonstrated it using cohort analysis to compare healthcare outcomes is expertly used in the future.

Hospitals and clinical organizations need to provide these data sets to intersectional medicine in a situation of ever-higher risk Viterbi said. This is what we are working on now finding a way to integrate information both within and outside of the healthcare system.