Healthcare News & Insights

Issues with software program highlight importance of overcoming implicit bias

Hospitals must make sure patients of all backgrounds receive the same treatment and attention from clinical staff when they’re admitted for care. While most doctors and nurses don’t intentionally treat patients differently, implicit bias could unknowingly affect their actions. And technology they’re relying on for assistance may not be helping matters. 

A popular algorithm in a software program commonly used by health systems to determine whether patients need follow-up care after a hospital stay showed bias against black patients, according to a recent study in Science magazine.

The study, which was conducted by a health system in Boston, the University of Chicago and the University of California (Berkeley), examined how the program determined which patients at a New York hospital were at the highest risk of becoming critically ill.

Instead of looking at a patient’s actual diagnoses or treatments, the algorithm in the program made decisions based on how much money was spent on each patient’s care.

When researchers compared the financial-based risk score to the actual health of each patient, they found that black patients were significantly sicker than white patients with the same risk score. This meant that these patients were assigned to the wrong risk category and should’ve received additional follow-up services to improve their health outcomes.

In many cases, the researchers said, less money is spent on treating black patients for various conditions. This is due to systemic differences in how patients from different backgrounds are treated in health care, as proven in many previous studies. Therefore, an algorithm that only looks at costs doesn’t give the full picture of patients’ actual health needs.

Next steps for fighting implicit bias

Response to this research was swift. State regulators reached out to the software vendor and asked it to improve the algorithm, according to an article on

The company was responsive and promised to fix the issues so the program can paint a more accurate picture of patients’ actual health risk. In addition, the study’s researchers are working with other vendors to prevent similar problems with newer software programs before they’re released.

Even with these software tweaks, implicit bias may impact providers’ treatment decisions. It’s crucial for hospitals to support clinicians in overcoming these issues, as discussed in an article from Medical Economics.

For starters, it’s important to acknowledge that everyone has different types of explicit bias – and that’s not necessarily a bad thing. It only becomes a problem when it affects a clinician’s behavior or recommendations for patients. Acknowledging this reality is key to making the conscious decision to overcome them in medical decision making.

Training is also critical. Lessons should focus on getting past stereotypes and biases to see patients as people, and using that insight to better include them in decisions about their treatment. Communication training can also be beneficial, so clinicians can relate to patients on a more personal level.

Taking the time to talk to patients and get to know them – and the unique circumstances that may impact their health and recovery – is an important tool in fighting implicit bias. Ultimately, working to overcome these biases will help hospitals and providers meet the goal of providing high-quality care to everyone who walks through the facility’s doors.

Subscribe Today

Get the latest and greatest healthcare news and insights delivered to your inbox.