The authors note that “whites enjoy a survival advantage over blacks in nearly all cancer sites, regardless of the stage at which disease is detected. Relative 5-year survival rates for all cancers diagnosed between 1983 and 1990 were 56% for white Americans and 40% for black Americans. Studies have demonstrated that blacks receive less intense treatment than whites for several highly prevalent cancers.”

This study focused on two outcome variables -- surgical treatment and survival time -- and four independent variables -- age, sex, income, and race. The authors used data collected by the National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program. Analyses focused on black and white individuals 75 years of age and younger (n=5189) diagnosed between 1978 and 1982 with stage I non-small cell lung cancer in 3 SEER sites (Detroit, San Francisco, and Seattle-Puget Sound).

“Whites were 20% more likely to receive surgery than Blacks and 31% more likely to survive 5 years. A logistic regression analysis of race and income as predictors of surgical treatment indicates a strong statistical relationship between race and surgical treatment, with Blacks appearing less likely to receive such treatment. After adding median family income to the model, race remained statistically significant, but the magnitude of the coefficient was approximately half that observed in the previous model (without income). For the Cox proportional hazards model, the race variable was associated with a higher mortality risk. Blacks had a risk ratio of 1.278 relative to Whites, equivalent to a nearly 30% greater chance of dying in any given month following diagnosis. When income was included in the model, race remained statistically significant although weaker than in the model omitting income.”

“Race proved a less consistent predictor of treatment and survival across SEER sites. Coefficients for race in equations predicting surgical treatment and including income among the independent variables were statistically significant only for patients from San Francisco.”

In discussion, the authors conclude, “These findings are consistent with the disturbing possibility that low SES and Black patients die needlessly because they do not receive a widely available treatment of significant potential benefit. Alternative explanations are plausible (including biological host factors and poor general health and adverse health behavior)… The differences reported here in effects of race across geographic regions are more consistent with a social than a biological explanation…The apparently independent impact of race on receipt of surgery noted here raises concern that parts of the health care system may treat members of some minority groups differently from nonminorities.”