How Kodak Gold Contributed to Bias in Dermatology

Interviewer: Temitayo A. Ogunleye, MD; Interviewee: Jenna Lester, MD


September 30, 2020

A recent review looked at all articles describing skin manifestations associated with COVID-19 — 46 articles with a total of 130 clinical images — and found none that documented dermatologic conditions in people with dark skin. This was despite the disproportionate incidence of the disease in Black, Latinx, and Native American/American Indian populations of color.

Temitayo A. Ogunleye, MD

What is going on? Temitayo Ogunleye, MD, an assistant professor of dermatology at University of Pennsylvania and a Medscape Dermatology advisory board member, spoke with lead investigator Jenna Lester, MD, a dermatologist and director of the Skin of Color Clinic at University of California, San Francisco, about the implications of her research.

Ogunleye: What prompted you to do this study in the first place?

Jenna Lester, MD

Lester: It was actually driven by frustration. While we're recognizing these COVID-related dermatologic manifestations, we're still trying to figure out their clinical significance. We've learned what changes could be an early sign of infection and what might occur in people who are otherwise asymptomatic and should be tested. But in the process, we're leaving out the group of people — dark-skinned populations — who are most heavily impacted by COVID. This is an injustice to the people who could benefit the most if found to have an early, visible sign of the disease on their skin. For example, pernio-like lesions and erythema, both of which have been seen in COVID patients, are harder to identify in darker skin.

As dermatologists, we know that the skin is the biggest organ and one that has the advantage of being visible. We partner with patients because they can see what we can see. We can explain what skin changes mean, and having examples to show them is really powerful. Because doctors typically respond best to numbers, I recognized that we needed data to prove that this lack of representation of persons of color in our COVID documentation was in fact true.

Can you tell us the key findings from your research?

We included any article that described the cutaneous manifestations of COVID — and also included a photograph — and was published over a period of 5 months. Then we categorized each image using the Fitzpatrick Skin Type Scale. We found that the majority of the photos were of people with light skin; there were no images at all of patients with dark skin, Fitzpatrick type V or VI. A handful, about 6%, depicted skin changes in patients with Fitzpatrick type IV skin.

Were you surprised by the findings?

No. I was not surprised at all, for a couple of reasons. First, many of the referrals that I had been getting to evaluate possible COVID manifestations were from primary care doctors. And as we both know, erythema, hyperpigmentation, and discoloration can be difficult for even a trained dermatologist to pick up on. So if these patients with skin changes potentially suggestive of COVID are presenting to their primary care doctor who could not determine what they were, perhaps because they had never seen them in someone with darker skin, it means those same clinicians are not likely to document these rashes. I suspect that a lot of these photos were sourced from primary care.

The other reason for my lack of surprise is that I have looked at this issue before. In a study I did in 2019 looking at images in dermatology textbooks, my co-authors and I found that there was a pretty dramatic underrepresentation of skin of color overall. Another analysis of images in core dermatology textbooks, published earlier this year by one of your colleagues at the University of Pennsylvania, Jules Lipoff, MD, showed that the percentage of images of dark skin ranged from as low as 4% to a high of about 18%.

So I wonder whether that makes us less likely to look for things in patients with darker skin, except for those conditions that we're taught are more common in people with darker skin, like keloids, vitiligo, or certain types of hair loss. I wonder whether we just don't think of other conditions because all the photos we ever see of psoriasis or rosacea, for example, are of people with lighter skin.

You bring up a good point about the cyclical nature of this process. If you don't see skin conditions in darker skin, then you don't know what to look for and so you never look for it.

Exactly. What if a Black patient says, for example, "This looks different on my skin; do you think this could be related to COVID?" And their doctor looks at their skin and says no. Not because it isn't, but because they haven't seen that before. Then they don't take a picture of it and we'll never know what that might have been.

The disturbing thing I have found is that there is an overrepresentation of dark skin in images of sexually transmitted infections. So based on what you are taught, you begin to create these powerful cognitive biases in your brain as a clinician and you start to put people in categories: "This person is more likely to get this because I've seen a lot of photos of it."

Take the example of a 40ish Black woman with a cough who is presumed to have sarcoidosis, because we've all been taught that. But what we have not been taught, at least not as definitively, is that the highest prevalence of sarcoidosis is in Nordic countries, where there are not many Black people.

So these loops teach you things that don't necessarily represent reality. We are taught to recognize patterns but the patterns that we've created are not necessarily valid and carry the biases of the people who decided they were important for us to learn.

So the underrepresentation of persons of color in images depicting skin changes in COVID-19 is, in your perspective, a continuation of this pattern of bias?

I definitely think it is. It's important to understand the history of photography and the development of color film in order to contextualize that question. Kodak Gold was one of the first color films made. To assist photographers, the company developed a Shirley card which could be used for color balancing in developing this film. That card, which was distributed to film development labs across the country, depicted a fair-skinned, White woman as the standard for how people should appear in the photograph. Film technology was built around this idea.

As a result, people with darker skin were never portrayed accurately and did not have a lot of detail to their features or their skin. A number of famous photographers boycotted by not using the film.

But it wasn't until chocolate manufacturers and the furniture industries decided that Kodak Gold film was not showing their brown products in enough detail that Kodak was forced to change.

Photo courtesy of Dr Lorna Roth, Concordia University, Montreal, Canada. Reproduced with permission of Eastman Kodak Company.

It took these big industries saying "we're not going to use your film anymore" to spur the development of multicultural Shirley cards which included images of Asian, White, and Black women (a Latina was added later). That didn't happen until 1995. By then, digital photography had already started to take off. Unfortunately, that advance built off of the original color film and still harbored some of the same issues.

So in addition to concerns about clinicians not recognizing a skin change in a darker-skinned person, if they do recognize it and do decide to photograph it, it just might not come out right. So then it's possible that clinicians just decide to stop taking photos.

This is another example of structural racism — things that are just baked into the system about which people are unaware. I fear we'll continue to perpetuate these unconscious biases with the development of augmented intelligence or various algorithms.

I think that this history — which I was unaware of — highlights what happens when White skin becomes the standard. It certainly explains the issue we've both seen of clinicians not recognizing erythema on dark skin.

That's a big one. And I think it's a big one because it's a shared presenting sign of a lot of different rashes. It can also be a sign of a dermatologic emergency that requires rapid recognition and treatment. Erythema can be very subtle, especially in skin of color. It is one of the things that we use to grade severity of psoriasis, and as a result of it not being appreciated in people with darker skin, those patients have not been included in a lot of the clinical trials.

Erythema seen with granulomatous rosacea in a woman with Fitzpatrick skin type VI. Photograph courtesy of Susan C. Taylor, MD (Philadelphia, Pennsylvania).

There has even been discussion of whether erythema is something we should deem to be an important finding in people with darker skin.

Maybe one of the problems is terminology. "Erythema" connotes pink or red coloration, and that does not really show up on dark skin as overtly.

Exactly. Emollient use is also different in people of color. So the "classic" scales of psoriasis often aren't present in someone who uses a lot of lotion.

In addition to the different clinical appearance of many common skin conditions, cultural practices might be different in different groups, which may also alter your differential diagnoses and treatment recommendations — for instance, moisturizer use, shampooing frequency, or use of different hair styling products.

I totally agree. Hair is another big one. Identifying broken hairs, short hairs, texture changes, and variability in the size of the hair shaft is different in coiled, Black hair and is not really applicable to people with more textured hair patterns. People of different ethnicities and cultures do different things. Standards that we traditionally use, such as the hair-pull test, don't work quite as well in different groups.

I think that speaks to the changes that we also need to make in educating both dermatologists and non-dermatologists in diagnostic criteria and clinical findings.

Dermatology is the second least diverse specialty. And that means our experts — faculty, lecturers, mentors — are not likely to be people of color. You don't know what you don't know. You only have your own experiences. If you don't have people who can explain why something may be different for a different group of people, you don't have an opportunity to broaden your understanding. You and I noticed because we know what it's like to have this type of hair. But it's not something that other people who don't share the same hair type would have the same perspective on.

It is even more reason to make sure that our dermatology workforce mirrors our population. About 3% of the dermatologists in the US are Black; the numbers are equally bad for Latinx and indigenous dermatologists. If you don't have enough people in your immediate environment who can help broaden your perspective, that becomes a problem that can harm patients.

If we don't have a diverse group of people in the field who are contributing to the literature, changing some of the ways that we think about practice, then everyone is just going to keep doing the wrong thing.

We need to build our evidence and pay more attention to the racial breakdown of patients included in studies. We are recognizing that we have blind spots but we don't have people or data that can change that. If what we are publishing overrepresents a certain group, how will we ever learn to do things differently?

I recognize the fact that this is not a new idea. Others have worked on these issues for a long time. Your colleague, Susan Taylor, a professor at University of Pennsylvania and a founder of the Skin of Color Society, used her energy and frustration about this issue and published a whole textbook on Dermatology for Skin of Color. When I tried to publish my first paper on this issue, I got a lot of pushback, but I'm happy that now it's something that everyone is talking about. So just in the span of a couple of years, the acceptance of this idea has changed dramatically.

I hope that this is the start of sustainable, systemic changes that can have a real impact.

Let me ask a technical question. Is the excuse of not knowing how to photograph darker skin just that — an excuse? Or is that concern truly valid?

That's a great question. There are certain techniques that one can employ. And yes, it can be more challenging if you are not used to taking photos of people with darker skin, but it certainly can be learned.

I think that an intelligent group of people who have faced things before that felt insurmountable could continue to push to figure out how to do it. But it is something that does require skills, and part of it is because there's this bias built into the camera, so you have to make up for that.

I think every single dermatologist has that ability, and you just have to know that it's worthwhile to do because your patient is that valuable.

Temitayo A. Ogunleye, MD, is an assistant professor at University of Pennsylvania and maintains a full-time clinical practice in Philadelphia. She has published numerous articles on nuances of dermatology in persons of color for both professional and lay publications.

Jenna Lester, MD, is an assistant professor in the Department of Dermatology at University of California, San Francisco Health where she is the director of the Skin of Color Clinic.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.