Letter to the Editor: Editorial: Beware of Studies Claiming that Social Factors are “Independently Associated” with Biological Complications of Surgery

To the Editor,

In a recent editorial, Seth Leopold MD addressed the use of race and insurance status in orthopaedic research. As a group that prioritizes health disparities research, we take issue with many of Dr. Leopold’s statements and feel compelled to respond to his conclusion that “orthopaedic surgeons should be skeptical of papers that link social factors that have little or no physiology to them (such as insurance status and race) with biological endpoints or medical complications after surgery” [18].

We believe this statement is counterproductive because, while race is a complex concept [7], it reliably correlates with many biological endpoints highly relevant to orthopaedic outcomes. For example, although black patients have lower dietary intake of calcium and lower 25-hydroxyvitamin D levels than white patients, black patients also have higher bone mineral density and less frequently suffer osteoporosis and fragility fractures [6, 13, 24].

Furthermore, we cannot ignore the reality that, in the United States, race is closely linked to factors that increase the risk of health complications [28]. Economic stability, health literacy, nutrition, and healthcare access are all tied to race through institutional racism [29]. These race-correlated factors have a documented impact on physiological variables [17] and are associated with a range of health outcomes [22, 30], including stress [3], depression [26], mental health [23], hypertension [5, 8], coronary heart disease [11, 31], diabetes [25], cancers [9, 14, 19, 21], and end-stage renal disease [2]. Race-correlated factors also influence term birth and infant mortality [1, 20], smoking and alcoholism [4, 15], and the quality of health care [12]. The impact of institutional racism on health has been clearly documented for two decades, yet the situation is not improving in surgical fields [16]. We should not abandon health disparities research for fear that the findings will be used to justify discrimination. Indeed, this research is vital to provide solutions for discrimination, including risk-adjustments for bundled payments to improve access to care.

Because social variables are typically absent from medical records, race and insurance status may be used as proxy variables that represent these complex factors and studies can and should control for these variables. Mathematically, controlling for proxy variables reduces treatment effect bias compared to omitting unmeasured confounders entirely [10, 27]. To reject articles because they control for some confounding proxy variables would be negligent. Statements in journal articles that improperly infer causality may mislead naïve readers, but we have confidence in orthopaedic surgeons to understand the limitations of multivariable analyses and appropriately interpret results.

We should not let a fear of misapplication invalidate race and social determinants as contributors to health inequities. Race-associated health outcomes disparities represent a national healthcare crisis, and our healthcare journals should put a spotlight on this crisis instead of hiding it. Race is a critical social determinant which, as a proxy variable, summarizes a great number of factors that should be of interest to any researcher and clinician. Accordingly, we hope Dr. Leopold’s concern about imperfect correlations will not distract from the necessary work of informing physicians and policymakers how to best care for their patients.