Overlooked: Implicit Bias in Health Care

Generally, implicit racial bias tends to draw a disproportionate amount of attention in healthcare research and policy. Rightly so: research has shown that patients of color face receive a lower quality of care than white people, even when socioeconomic status is considered.

However, this has come at the expense of attention to other important factors. In the context of health care, a recent systematic literature review identified 42 articles that measured implicit bias among healthcare professionals in-patient care. While 27 studies examined racial/ethnic bias, only 10 studies focused on gender, age, weight, marital status, or other demographic factors.²

In this article, I use a sociological lens to discuss the existing body of research on implicit biases related to gender, body size, and marital status among healthcare professionals, linking these biases to prevailing social norms. By doing so, I hope to draw attention to the fact that race is not the only implicit bias that we must address in healthcare settings. 

Over 40% of the U.S. adult population is obese,³ and negative attitudes about weight are prevalent throughout the country. In our society, obese and overweight individuals are often perceived as lazy, weak-willed, unsuccessful, unintelligent, and lacking self-discipline. These harmful stereotypes promote widespread prejudice and discrimination. In fact, weight discrimination is nearly just as common in America as racial discrimination.

In contrast to race, however, anti-fat bias is often perceived as more socially acceptable because one’s weight is believed to be under one’s control. In reality, body size is influenced by factors such as an individual’s environment and genetic makeup. Still, our culture promotes that overweight individuals should be blamed for their size.⁴ Obese individuals often face the consequences of these stigmas in a variety of aspects of their life including, but not limited to, the workplace, educational institutions, and healthcare settings. 

Anti-fat bias is deeply ingrained among healthcare professionals. In one study, physicians indicated that they react to obesity with feelings of discomfort, reluctance, or dislike. Additionally, they associated obesity with poor hygiene, noncompliance, hostility, and dishonesty.⁵ Similar findings have been demonstrated in other studies on physicians and medical students.⁶ I do recognize that many of these studies are from over two decades ago. However, these findings should direct future research towards evaluating current anti-fat attitudes. 

Several studies have indicated that anti-fat attitudes among healthcare professionals can negatively impact clinical judgment, diagnosis, and quality of care. For instance, physicians who hold such attitudes indicate that they themselves do not expect treatments will succeed when the patient is overweight.⁷ Physicians also express that poor patient compliance and motivation is a common frustration during obesity treatments, resulting in a reduced emphasis on communicating information that can promote lifestyle changes.⁸

This can lead to a self-fulfilling prophecy: physicians may put less effort into treating obese patients or communicating with them, resulting in poorer outcomes and reinforcing the physician’s original attitudes. In one study that used vignettes to assess clinical judgments, it was found that mental health workers tend to associate obese patients with more negative symptoms.⁹

The negative attitudes of healthcare providers towards obese and overweight patients can also lead to hesitance among these patients to seek health care.¹⁰ For example, women who are overweight are significantly less likely to obtain regular pelvic exams due to the negative body image fostered by physicians’ unwillingness to attend to these patients.¹¹

The assumption that obesity leads to worse health outcomes largely goes unquestioned. Going forward, researchers should seek to better understand how much of this is due to physiological characteristics, and how much is from the consequences of discrimination.

Social support can be characterized by the emotional, informational, and instrumental resources that people obtain from other people. Social support is strongly correlated with better health and well being.¹²

It is not uncommon for physicians to consider social support when determining a patient’s ability to handle challenging treatments. However, it is important to note that marital status is not synonymous with social support. Unmarried patients can still have strong support systems through family, friends, etc. In medical decision-making, this creates concern for marital status bias that is founded on cultural narratives, not evidence. 

In a recent article published in the New England Journal of Medicine, it was discussed how unmarried patients are less likely to survive cancer. The authors suggested that this may be due to physicians’ implicit stereotyping of patients based on marital status as an indication of the patient’s support system.¹³ Physicians are less likely to recommend surgery or radiotherapy as treatment for cancer patients who are unmarried.

Marital status also strongly affects clinical judgment when evaluating patient eligibility for scarce medical resources such as organ transplants.¹⁴ Divorced patients, as opposed to married patients, are less likely to receive an organ transplant because they are perceived to be less resilient, have less social support, and be “less deserving.”

Future steps that medical decision-makers can take to minimize this bias is to ask patients questions about their social support system without bringing marital status into the picture. For example, they can ask patients if they have people who will be supporting them throughout their treatment journey, and if they have people they can talk to about important medical decisions. These sorts of questions will draw attention away from marital status and towards social support, which is the true focus.

Implicit gender bias has been shown to affect medical decision-making in the hospital setting. Surprisingly, one study found that women are less likely to receive a knee arthroplasty than men by a factor of 3. This study even accounted for cases where it was clinically appropriate for the women to receive the arthroplasty.^15,16,17

This drastic difference may be because men are stereotyped to be stronger and able to withstand more pain while also being more active, therefore benefiting more from the knee replacement than women because their condition is perceived to be worse. This pain stereotype, albeit clinically true, can have adverse consequences because it can inadvertently result in women not receiving the appropriate medications and treatment.¹⁸ 

This was also observed in another study where male coronary artery disease patients were considered to be at higher risk and thus were prescribed more aspirin and lipid-lowering medication as a mode of secondary prevention than women patients, even when all other factors were controlled for. In fact, the study concluded that while men tended to be prescribed the optimal amount of medication, women did not. This is an indication that gender bias actually contributes to a lower quality of medical care. 

It is important to note that there are many systemic factors that contribute to this gap. For example, there is a large funding gap in medical research concerning woman’s health. These often inevitably feed into individual implicit biases and vice versa.

How can we combat the effects of implicit bias in healthcare settings? Creating bias education programs for all physicians, as well as nurses, physician assistants, and social workers, can help counteract unconscious bias and stereotypes by making them aware of their disastrous consequences. Teaching people to recognize their biases can make them more aware and apt in attempting to reduce the bias themselves.

A commonly used tool to assess implicit biases is the Implicit Attitudes Test (IAT). While the race-based version of this test is most commonly used, health care workers should also take the tests assessing implicit attitudes towards gender, sexuality, and weight. This exercise will help elucidate the distinction between implicit bias and explicit endorsement.

While awareness does help in reducing the effects of bias, it is not sufficient to fully overcome the automatic activation of stereotypes and the consequential effects. To combat this, there are various strategies that the provider can implement that are low-cost, time-effective, and simple.

Practicing perspective-taking

Providers should be mandated to a perspective-taking strategy in between each patient that a physician sees. This short reflection period can allow the provider to understand the entire situation specifically from the standpoint of potential biases, and correct accordingly. This perspective-taking can involve manipulating a certain perspective through visual or audio guidance, or through a specifically designed form of the IAT. Both of these methods can guide the individual to start understanding what types of bias exist and how their decisions incorporate the bias.¹⁹

Focus on the patient’s individuality

Another effective strategy is individuating. Individuating is the practice of consciously focusing only on specific information about an individual that is relevant to medical decision-making while leaving race or gender out of thought.²⁰ This strategy helps physicians avoid filling in gaps and uncertainties in patient information with stereotype-based assumptions, as shown in a study conducted by Chapman et al on gender disparities in COPD.²¹ However, it is important to note that such a strategy may take away value from the provider-patient relationship as it forces the provider to treat the patient as a set of lab results and symptoms rather than as a human being.

Early education

In addition to training sessions for practicing providers, training should also be implemented into the curricula for medical students. Starting an annual training plan including recurring IAT examinations and a manipulative perspective-taking experience can begin to have these future physicians ready to adjust to the background of any patient they may encounter. A preemptive acknowledgment of the bias issues will allow physicians to combat this issue before it evolves into a problem in the workplace.

The wide array of literature culminates to a crucial conclusion: healthcare professionals are not immune to implicit biases, and these implicit biases extend beyond just race. After all, healthcare practitioners are still members of society, and are privy to society’s social norms and stereotypes. More recent literature will be important to understand the impact of discrimination in healthcare today because, in recent years, there have been great strides towards advocacy and awareness of implicit attitudes and stereotypes. However, the innate challenge of addressing implicit biases is that they differ from explicit attitudes, making it very difficult to achieve tangible results.

Another question worth addressing in the near future is whether or not racial implicit bias can lead to worse outcomes in comparison to other forms of implicit bias. Existing literature does not focus on comparative analyses. Answering this question can better guide future interventions, policies, and where to invest research funding. Till we answer this question, however, I urge that implicit biases concerning body size, marital status, and gender receive the same amount of attention.