The FPs who completed our survey attributed racial/ethnic differences slightly more to environment and gender differences slightly more to genetics. FPs’ views about the etiology of racial/ethnic and gender differences in health were strongly and positively correlated. However, the proportion of FPs who strongly endorsed either genetics or environment as the sole cause of health differences was quite low, suggesting that the vast majority of FPs appreciate that common conditions are etiologically complex.
FPs who were practicing in teaching environments, those who reported being more innovative and those who were younger were less inclined to attribute health differences to genetics. These results may reflect these physicians’ more recent and/or greater exposure to genetic curricula that increasingly acknowledges growing scientific evidence for the complex influences of environment and genetics on health outcomes.
While our data show that the FPs surveyed view race/ethnicity and gender as providing clinically relevant information, relatively few of the physicians in our sample rated gender or race/ethnicity as essential to clinical decision-making. Moreover, our results suggest that FPs who regard race/ethnicity and gender as important in making their clinical decisions are not more likely than their peers to view race/ethnicity or gender as proxies for genetic differences. Our data reveal positive correlations between practice environment (that is, urban areas and areas with >10% poverty) and FPs’ ratings of the importance of race/ethnicity in clinical decisions. Further research is needed to better understand these relationships.
Not surprisingly, our results indicated that FPs consider multiple patient characteristics when making their clinical decisions. However, given the widely acknowledged importance of family history as a tool for ascertaining clinically relevant genetic, environmental and psychosocial factors [30
], it was surprising that FPs rated the importance of family history as no more important than race/ethnicity and gender in their clinical decision-making. It is possible that the time and effort required to collect and update family history offsets some of its value as a clinical tool, particularly if one considers the ever increasing number of things that FPs are expected to do as part of routine clinic visits [33
]. Race/ethnicity and gender may be viewed as quick, albeit imprecise, indicators of medically relevant genetic, environmental and psychosocial factors. The evaluation of recent efforts by the Surgeon General and others to foster widespread adoption of more rigorous family history taking, as well as the development and evaluation of new, modified versions of family history tools [34
], may help give information about what is needed to improve the perceived clinical utility of family history.
Lastly, although not a primary focus of this study, our results raise questions about physicians’ attitudes regarding insurance status that could be explored in future research. Insurance status was rated as least important of all the characteristics, though still more important than not, and these ratings were not correlated with ratings of the importance of other patient characteristics. Further research is needed to understand why and how physicians use insurance status in clinical decision-making, and how these practices might affect the quality of care for patients with no or poor insurance coverage.
Like any study, these results must be considered with some caveats. We developed our own single measures of perceptions about race/ethnicity and gender differences and the role of genes and environment. Questions assessed clinical decisions under the umbrella of nonspecific clinical decision-making. Race/ethnicity and gender may be more important for some clinical decisions [35
] than others. FPs’ ratings of the importance of a given factor in their clinical decisions may not represent that factor’s actual influence on decision-making, and would not capture physicians’ implicit use of race/ethnicity and gender in their clinical decision-making. Previous research into the providers’ contribution to health disparities suggests that patient race/ethnicity has an implicit, unconscious effect on FPs’ clinical decision-making [18
]. Questions regarding the importance of genetics and environment in health differences relied on the terms genetics and environment without defining either, leaving room for respondents to interpret them differently. Also, the scenario dealing with racial/ethnic health differences described disease diagnosis and outcome, while the scenario dealing with gender health differences described disease incidence and clinical characteristics. This asymmetry may limit the comparability of FPs’ responses to the 2 scenarios. Lastly, we recognize the limitation of combining race and ethnicity as one descriptor. Each term has different meanings that vary in use and construed meaning, and we cannot characterize how FPs interpreted the combined terminology. Future research should be conducted to refine measures of these constructs and to understand how the explicit beliefs studied here relate to the implicit influence of patient characteristics on physician behavior.
We cannot say whether FPs’ ratings of the importance of race/ethnicity and gender are appropriate or inappropriate, nor can we say whether those who reported placing substantial importance on these characteristics use them to their patients’ benefit. Further research is needed to link these practices to meaningful clinical outcomes in order to gain understanding of their effects.
Our response rate of 10% was low. However, the 1,035 FPs who completed the survey were not notably different from the overall membership of the AAFP. Additionally, our sample was not sufficiently racially or ethnically diverse to use physician self-reported race/ethnicity as an analytic variable. This should be explored further; other studies querying FPs’ beliefs about health differences have shown significant differences by FPs’ self-identified race/ethnicity.
In conclusion, this study is among the first to provide empirical evidence of a large group of physicians’ beliefs relevant to the current discussions about physicians’ attitudes regarding race/ethnicity and gender in health outcomes and clinical decision-making. How these perceptions will evolve with the increasing availability of genetic susceptibility testing and pharmacogenomics, and how, if at all, these perceptions will shift and influence clinical decision-making is worth continued study if we are to better understand and shape integration of genomic medicine into primary care practice.