EHRs can Exacerbate being Unseen, Unheard, and Unbelieved

 

The #EndGenderBias Survey Summary Report was just released by the Australian Government and National Women’s Health Advisory Council, and it was damning. About 2,700 people weighed in to say that two-thirds of women have experienced health care related gender bias or discrimination, which impacted women’s lives in feelings of shame and self doubt, financial burden, lost career opportunities, and delayed diagnosis and treatment. That bias manifests itself as not being believed or heard or able to make choices or access the right care.

My husband Brian spoke about an aspect of this in his Medinfo keynote: testimonial injustice is a type of epistemic injustice “when someone is ignored, or not believed, because of their sex, sexuality, gender presentation, race, disability, or, broadly, because of their identity.”

A woman from the survey described her experience

Within the first few moments of meeting me, [the doctor] had decided I was a “waste of time”, a hysterical woman with a non-urgent problem who was being dramatic.

Both women and men go unheard when seeking help for autoimmune disease, but as many as 4 out of 5 patients with autoimmune disease are women who have consistently not been believed and told their symptoms are due to anxiety or are made up. Meghan O’Rourke’s memoir, “The Invisible Kingdom: Reimagining Chronic Illness,” describes the years she suffered from an unrecognized autoimmune disease and her painful interactions with a healthcare system that ignored and condemned her for "complaining". Chapter 8 is titled “The Woman Problem” and starts with this

Of the nearly one hundred women I interviewed, all of whom were eventually diagnosed with an autoimmune disease or another concrete illness, more than 90 percent had been encouraged to seek treatment for anxiety or depression by doctors who told them nothing physical was wrong with them.

New research by Howard Chang and others shows a biological reason for the overactive immune response in women: A molecule made by one X chromosome (called Xist) in every female cell can generate antibodies to a woman’s own tissues. Why are we just discovering this?

“... for several decades, we’ve used a male cell line as the standard of reference…So, all of a female patient’s anti-Xist-complex antibodies — a huge source of women’s autoimmune susceptibility — go unseen.”

What does this mean for digital health? We are writing a grant proposal with the working title “Informing future patient-centred information exchange by mapping primary care information and communication needs, co-designing consent workflows, and testing emerging technical solutions.”  It’s a mouthful and is motivated by the following premise: The positive changes government is making towards information sharing will be limited if we don’t understand the information needs of the patients and their primary care providers, and the changes will be blocked if patients do not have control over who sees what information. 

What goes into a patient’s electronic health record has a sort of permanence that follows them from provider to provider and has the potential for transmitting bias. The arrogant but common assumption that negative test results indicate there is no physical explanation for a patient’s symptoms stays in the record and often affects the quality of care that patients subsequently receive. Stigmatizing language in patient medical records can take many forms:

(1) questioning patient credibility (2) expressing disapproval of patient reasoning or self-care (3) stereotyping by race or social class (4) portraying the patient as difficult (5) emphasizing physician authority over the patient.

One study showed that exposure to a note with stigmatizing language was associated with more negative attitudes towards the patient and that reading the stigmatizing language note was associated with less aggressive management of the patient’s pain.

Patients are right to want the right information getting to the right person in the process of their care, as I noted in my previous blog post, but they are also right to be concerned about certain types of information or ways they are described in the notes transmitting even more bias than they already face in the healthcare system. 

As with everything, digital health can make things worse--like transmitting bias and stigma. But it can also make things better. As we write our grant proposal, I'm asking myself what we already know about patient's information sharing concerns--they aren't static or the same across all cultures--and how feasible it is to support customized consent in terms of technical considerations, legal frameworks, and workflow. I'd love to hear what you know about this topic!