05:54 am
Negative Language in SCD Clinical Notes Linked to Bias
In a cross-sectional analysis published in JAMA Network Open, Austin Wesevich, MD, MPH, MS, and colleagues examined whether negative language in the EHR clusters around sickle cell disease (SCD) itself or around the overlapping stigmas that often affect these patients: Black race, pain, and opioid exposure.
The dataset for the study, drawn from the University of Chicago hospital, was sizable: 39,871 clinician notes related to 18,326 patients, spanning inpatient, outpatient, and emergency department encounters. Notes from January 1, 2019, through October 1, 2020, were analyzed with natural language processing and machine learning to identify seven negative descriptors used in context: aggressive, agitated, angry, nonadherent, noncompliant, noncooperative, and refuse. Patients with SCD were compared with four non-SCD groups: Black patients, patients with chronic pain, patients with opioid use disorder (OUD), and a counterfactual group of non-Black patients without chronic pain or OUD.
One obvious question for clinicians is whether biased documentation related to SCD is driven more by race, by pain, or by the stigma that follows opioid treatment. Another is whether language in the chart becomes a mechanism by which bias is transmitted from one encounter to the next.
Negative descriptors appeared in 220 of 1,443 notes for patients with SCD, or 15%. The most common negative descriptors were refuse, noncompliant, and nonadherent. By comparison, negative descriptors appeared in 14% of notes for patients with OUD, 7% for Black patients without SCD, 7% for patients with chronic pain without SCD, and 3% for the counterfactual group. Patients carrying all three stigmatizing disease-related and racial identities had the highest frequency, with negative descriptors in 19% of notes.
As Dr. Wesevich pointed out in forwarded written comments, “Our study showed that 15% of notes for patients with sickle cell disease included a negative descriptor, with the most common negative descriptors being refuse, noncompliant, and nonadherent.” That finding is hard to ignore, but the adjusted analyses are what make the study clinically important.
After adjustment for sex, age group, marital status, encounter setting, insurance, and Charlson Comorbidity Index, patients with SCD had significantly higher odds of having a negative descriptor than Black patients without SCD, patients with chronic pain without SCD, and patients in the counterfactual group, with adjusted odds ratios (aORs) of 2.46, 1.96, and 14.26, respectively. Odds were similar to those seen in patients with OUD without SCD, at an aOR of 0.76. In sensitivity analyses limited to Black patients, those with SCD still had higher odds of having a negative descriptor than Black patients with chronic pain, at an aOR of 1.97 but lower odds than Black patients with OUD, at an aOR of 0.49.
That pattern supports Dr. Wesevich’s interpretation: “Patients with sickle cell disease had higher odds of a negative descriptor in their notes than Black patients or patients with chronic pain, but they had similar odds as patients with opioid use disorder.” He added, “Our findings suggest that patients with sickle cell disease likely experience both race-based and disease-related biases. The stigma associated with opioids likely affects sickle cell care.”
The clinical implication is not abstract. Vaso-occlusive pain is one of the main reasons patients with SCD present for acute care, and those encounters often depend on rapid decisions about analgesia, trust, and adherence. This study suggests that the EHR itself can encode stigma, especially around opioid-treated pain, and carry it forward into future care.
Wesevich A, et al. JAMA Netw Open. 2026;9(4):e266458. doi:10.1001/jamanetworkopen.2026.6458
Global Lessons in Scaling AI-Based Chronic Kidney Disease Screening
Pulmonology Today Welcomes Hassan Bencheqroun, MD, EMBA, to Editorial Board
Subcutaneous Infliximab Demonstrates Consistent Endoscopic Efficacy Across Crohn’s Disease Locations
These cookies may be set through our site by our advertising partners. They may be used by those companies to build a profile of your interests and show you relevant adverts on other sites. They do not store directly personal information, but are based on uniquely identifying your browser and internet device. If you do not allow these cookies, you will experience less targeted advertising.
These cookies allow us to count visits and traffic sources so we can measure and improve the performance of our site. They help us to know which pages are the most and least popular and see how visitors move around the site. All information these cookies collect is aggregated and therefore anonymous. If you do not allow these cookies we will not know when you have visited our site, and will not be able to monitor its performance.
These cookies are necessary for the website to function and cannot be switched off in our systems. They are usually only set in response to actions made by you which amount to a request for services, such as setting your privacy preferences, logging in or filling in forms. You can set your browser to block or alert you about these cookies, but some parts of the site will not then work. These cookies do not store any personally identifiable information.