The goal of this study was to assess whether stage of disease explains the race differences in survival of breast cancer in the recently published literature. Studies published between 1968 and 1983 were reviewed, and 18 were selected for inclusion in this analysis. The authors rated the "reliability" of each study based on the specificity of the measure used for stage of illness and the uniformity of the quality of care measures. That is, studies that used better measures for stage of illness were considered most reliable, and studies that included patients who were most likely to receive uniform quality of care were considered most reliable. Reliability scores also accounted for whether the study included measures of other factors of prognostic significance (age, SES, and nutritional status).

The authors then correlated their reliability scores with the reported racial differences in survival rates, showing those studies with the largest racial disparities in survival had the lowest reliability scores. That is, 61% of the studies showed very small racial differences, and the average reliability for these studies was 8 of 10 possible points. This finding indicates that the studies that used the better stage of disease measures, that were unlikely to be confounded by quality of care, and that included other prognostic factors in their analyses were unlikely to find racial disparities from breast cancer. On the other had, 28% of the studies found large racial differences, and the average reliability score for these studies was only 3.4.

Thus, these authors made the assumption that race differences are explained primarily by race differences in stage of disease and quality of care and might also be influenced by race differences in other prognostic factors. The published studies tend to support this pattern. The authors conclude, "efforts directed at early detection and improvements in the quality of care delivered are likely to reduce the excess breast cancer mortality by women of low SES, regardless of race."