RACIAL BIAS IN MEDICAL DECISION MAKING

Doctors rely on guidelines developed by medical associations, nonprofit organizations, and government agencies to help them decide when to recommend surgery or specialized care. Recent studies show that many of these statistically based analyses are racially biased and can result in Black and Hispanic patients receiving inferior care.

A study in the New England Journal of Medicine examined 13 algorithms used in cardiology, obstetrics, and other specialties. One uses levels of creatinine, a waste product made by muscles, to estimate kidney function. Kidneys remove waste from the body, so if a patient has less creatinine, he or she is presumed to have better kidney function. Because studies have shown that Black people on average have higher creatinine levels than White people do (perhaps because, as studies have also found, Blacks have higher muscle mass on average), the algorithm “race-corrects” by lowering their level. Thus, Black people appear to be healthier and are less likely to be referred to a specialist for further care.

The rationale for such adjustments is based on studies showing correlations between race and certain medical outcomes. But in some cases, other factors that might be the true drivers of the correlation, such as socioeconomic status, were discounted. In others, the studies have since been proved to be incorrect.

While it may sometimes be appropriate to include race in medical diagnostics, the researchers caution that many commonly used algorithms need to be revised to remove these statistical biases.