Maddalena Marini, Ph.D.

The Hidden Mind

Does Implicit Bias Affect Clinical Decision Making?

The potential hidden costs of implicit bias in the health care setting.

Posted May 23, 2019

This post was authored in collaboration with Enrico Vescovo, a medical student at Ferrara University (Italy).

The medical profession strives for equal treatment of all patients. This pledge is one of the core points of the Declaration of Geneva, which clearly states that a physician, like all the other medical professionals, must treat every human being equally regardless of social class, ethnicity, religion or ideology. However, despite this principle and the efforts of health care professionals to respect it, disparities in health care have been documented in our society. In 2002 the Institute of Medicine (IOM), a part of the National Academy of Sciences, reviewed about 600 studies investigating the relation between medical outcomes (i.e., diagnoses and treatments) and demographics (i.e., age, sex and race) of patients. Results showed that Black Americans and other minority groups receive less effective medical care than White Americans.

What are the causes of disparities in health care?

In the last 20 years, research has shown that our mind contains a large set of biases about social groups. These biases are called implicit or unconscious biases as we can be unaware of their presence. Scientific studies showed that in our everyday interactions with others, we automatically process information associated with their social characteristics—such as race, gender and age—and we often behave according to associated biases “stored” in our mind instead of factual evidence. For example, upon meeting an Italian woman for the first time we automatically assume that she is a good cook. Similarly, when we first meet a Chinese man we think that he is a hardworking person.

Researchers have shown that people unconsciously endorse social stereotypes and attitudes regardless of their own gender, race/ethnicity, age and even profession. Physicians and medical students hold implicit biases, too. A study conducted in a sample of more than 2,000 medical doctors showed that they exhibit implicit weight biases—i.e., preferences for thin people over obese people—as strong as those of people with no medical degree.

Might implicit social biases affect clinical decision making and physicians’ behaviors? Do they unintentionally contribute to health care disparities in our society?

First of all, let’s simplify in three steps how a physician usually manages a clinical case in order to understand how implicit social biases may operate and influence his/her medical performance.

Step 1: the Encounter. The physician meets the patient and collects information about his/her present health condition and medical history.

Step 2: the Elaboration and diagnosis. The physician processes the information provided by the patient and hypothesizes potential causes of symptoms. In this step the physician can prescribe medical examinations to test the formulated hypotheses.

Step 3: the Diagnosis and Treatment Decision. The physician, on the basis of the information obtained in the two previous steps, defines a diagnosis and an appropriate treatment for the patient.

It is easy to imagine how implicit social biases could potentially occur in all these steps and affect a physician’s journey towards an appropriate diagnosis and therapy.

For example, studies have documented that physicians with stronger implicit racial bias show markers of poor visit communication and interpersonal care for Black patients in comparison to White patients. These factors might lead to the collection of a different amount of information from the two groups of patients and affect a physician’s consideration of the patient's medical condition (Step 1), the causes of their illness and the choice of clinical tests to perform (Step 2) and the definition of a diagnosis and a treatment given to the patients (Step 3).

In addition, it has been also shown that implicit bias can operate directly in the phase of the diagnosis and treatment, influencing clinical decision making. For example, studies found that physicians’ implicit racial bias predict their decisions in using or not using specific medical procedures such as thrombolysis for myocardial infarction. Similarly, a study of pediatricians showed that pain treatment recommendations—such as the quantity or type of drugs prescribed to the patients—is associated with the strength of their implicit racial biases.

The implicit social biases are thus a plausible cause of the health care disparities. However, this issue requires more research. Although available findings have suggested that physicians’ implicit social biases predict aspects of medical care, there are still too few such studies to allow firm conclusions. A potential problem in this field of research is the attitude of medical institutions that, like many other organizations, are not always willing to participate in studies that might reveal the presence of social biases.

Social biases represent one of the challenges that our society needs to address in order to improve the well-being of its citizens and offer them equal opportunity, fairness and justice. People’s health should be protected by objective medical analysis, not swayed by personal social preferences.


Sabin, J. A., Marini, M. & Nosek, B. A. Implicit and explicit anti-fat bias among a large sample of medical doctors by BMI, race/ethnicity and gender. PLoS One 7, e48448 (2012).

Cooper, L. A. et al. The associations of clinicians’ implicit attitudes about race with medical visit communication and patient ratings of interpersonal care. Am. J. Public Health 102, 979–987 (2012).

Green, A. R. et al. Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients. J. Gen. Intern. Med. 22, 1231–1238 (2007).

Janice A. Sabin, J., Brian A. Nosek, B. A., Anthony G. Greenwald, A. & Frederick P. Rivara, F. P. Physicians’ Implicit and Explicit Attitudes About Race by MD Race, Ethnicity, and Gender. J. Health Care Poor Underserved 20, 896–913 (2009).