Addressing Skin Colour Bias in Clinical Assessments and Guidelines

CONTEXT

Skin colour bias can show up in multiple ways in the healthcare system, including in clinical training texts, assessments and guidelines. These biases affect patient safety, outcomes and experience. A number of incidents have highlighted this issue in recent years:

• Mohammad died at three months. A range of factors contributed to his death, including the way existing systems for triage do not always take into account the colour of a patient’s skin. The report of the national safety investigation concluded that ‘Assessment of skin colour and people being asked if patients are ‘pale’ or if their lips have ‘turned blue’ are not useful approaches for patients with non-white skin. In addition, descriptions of rashes focus on redness or paleness, both of which are more difficult to recognise in a patient with non-white skin’ [1]

  
  

• Elliana had a diagnosis of jaundice delayed in part because she was Black. A report into her experience noted that national guidance doesn’t ‘contain information on how to address the challenges of detecting jaundice in newborn babies with black or brown skin’ [2]

Cases such as these have increased awareness in NHS trusts of the need to review clinical guidelines and assessment tools. The following are some examples of project work and reflections in paediatric care from one London NHS trust.

IMPROVEMENT WORK

Quality Improvement Programme: Improving identification and responsiveness to the acutely deteriorating patient

The Trust’s Quality Improvement (QI) Programme on patient deterioration is made up of multiple projects including nursing observations, detection of deterioration, escalation and treatment of deterioration, and the role of parental concern. A key element of the observations work relates to changes in patient skin tone and/or colour as an indicator of deterioration that may be missed in patients with black or brown skin. Patient deterioration assessments have typically used language that describes what white skin can look like during clinical deterioration (e.g. ‘pale’ or ‘ashen’).

The Programme’s working group has been trialling new ways to support clinicians assessing potential changes in skin colour and tone. These have included asking parents/carers to identify their child’s ‘normal’ skin tone using a colour chart (similar to those in the cosmetic industry). However, feedback indicated some issues with this option:

• skin colour and/or tone can change for a variety of reasons, such as lighting (even a sunny day can make a difference to how skin is visually perceived)

• complexity is introduced by preferences for fair skin within some cultures, and positive connotations with whiteness

• there can be some translation and language complexities in describing changes to skin tone and/or colour when a patient is deteriorating

The work is evolving. At the moment, the Trust is testing an approach worked out with patients and their parents/carers. A QI Coach on the project describes it this way: “Ultimately, the feedback from young people was ‘focus on me when I’m well and when I’m not well, and the best people to ask about this difference are my parents and carers’. So the question we ask now is, ‘Has anything changed since we last saw your child?’. This avoids a lot of the difficulties with attempting to have standard terms for skin colour or tone change.”

Clinical guidelines workstream

At about the same time, the Trust was reviewing its clinical guideline on the detection and management of jaundice in neonates based on recently updated national guidance. The national guidance said that clinical recognition of jaundice could be more difficult in babies with ‘darker’ skin tones.

For the Trust’s Clinical Quality Assurance Lead, this was a concern. “While it is commonly used wording, it does raise the question: ‘darker than what?’ There is an invisible white norm in both the language and assessment approaches, that I wanted us to address directly as part of our guideline review process – for this particular guideline and for all going forward,” they said.

They brought this concern to the Trust’s Clinical Guidelines Committee. The committee members were receptive and eager to discuss the implications of the term ‘darker’ in relation to skin, with a view to improve the clarity of the guidance. The Clinical Quality Assurance Lead said, “As a white person, I’m conscious of my responsibility to raise matters related to colour bias, but also to do so in a way that engages and opens up discussion. I found our committee members keen to explore this. The discussion amongst colleagues was important, because as a committee we’re trying to ensure that all guideline wording is factual, descriptive and useful to clinical staff trying to make the best decision at the point of care and especially so when under pressure. So, in the case of these guidelines, which instruct clinicians to look for a yellowing of the patient as part of assessment for jaundice, we needed to think about what this means for patients with black and brown skin.”

A common response to this is to add an addendum saying that there may be difference when assessing darker skin – this was deemed insufficient. After an engaged discussion, the committee settled on the guideline wording for this intensive care setting:

Do not rely on visual inspection alone to estimate the bilirubin concentration. This is especially important to avoid missing jaundice in babies with black or brown skin, as guidance has previously not provided specific advice on jaundice presentation in black and brown skin to ensure optimal identification in all patients.

At the Trust, all neonates have a bilirubin test to confirm diagnosis of jaundice in addition to visual assessment. The new wording in the guideline names the rationale for testing, which includes the increased risk of missing jaundice in babies with black or brown skin, and it provides further detail to improve the effectiveness of assessment.

CONCLUSION

“The naming of the historic emphasis on white skin colour and tone in assessment tools felt important in the explanation of why corrective attention to detail is essential to avoid harm from missing clinical information,” said the Clinical Quality Assurance Lead.

There is no consensus in the healthcare sector as to what terms to use. Some terms may be more applicable to certain circumstances, so it’s not straightforward, as is evidenced in the Trust’s deteriorating patient work described above. National organisations connected with and supporting the NHS use a range of terms, including ‘darker skin’, ‘non-white skin’, and ‘black and brown skin’. All of these terms have their pros and cons. The table below (put together by the Clinical Quality Assurance Lead to aid consideration of these issues) outlines what these may be.

OPTION	PROS	CONS
Darker skin	Acknowledges the gap that can contribute to People of Colour receiving poorer care than white people.	Doesn’t name the white reference point that underpins it (ie ‘darker’ than what?).
Black and brown skin	Acknowledges the gap that can contribute to People of Colour receiving poorer care than white people. Names skin colours and tones being referred to	Doesn’t explicitly include olive, yellow etc skin colours and tones. ‘Black’ is used for a range of dark skin colours and tones so it may not function as a literal descriptor in a clinical setting (though it is commonly understood)
Non-white skin	Acknowledges the gap that can contribute to People of Colour receiving poorer care than white people. Names whiteness.	Centres whiteness so any other skin colours and tones are ‘other’.
Patients/People of colour	Acknowledges the gap that can contribute to People of Colour receiving poorer care than white people.	‘People of Colour’ is typically a US term that is less familiar in the UK.
BAME patients (ie 'Black, Asian, and Minority Ethnic')	Acknowledges the gap that can contribute to People of Colour receiving poorer care than white people.	The term ‘BAME’ has limitations that are increasingly recognised, and it is no longer used by a growing number of public bodies. ‘Minority ethnic’ is relative to context [ie those deemed 'minority’ are in most cases global majority]

Following these discussions another member of staff, who leads on clinical audit, contacted the authors of the original guidance to query the use of the word ‘darker’. Their response reflected the complexity of the issue. “It’s clear there are no easy answers,” says the Clinical Quality Assurance Lead. “We’re trying to address an issue built into the structure of our society, an historic bias, and it starts with acknowledging it.”

KEY LESSONS AND STRATEGIES

This is not an issue that lends itself to pithy lessons or neat summaries. The main takeaway is that building a new system free from racism is an ongoing process: it requires all of us to constantly interrogate the words, norms, and structures we use, usually without thinking. Bearing this in mind, here are some reflections from the Trust as they continue their journey.

Framing

• It can be helpful to approach conversations about bias or discrimination in terms of improving care quality. Grounding the conversation in the aspiration for great healthcare helps people to engage with it, even if anti-racism work is less familiar to them.

• Use impactful patient stories to illustrate the real-world consequences of bias in healthcare to make the issue more relatable and urgent. When discussing the assessment of children with sepsis, the QI Coach said: “Everybody wanted to see it working optimally... So, we framed this by asking, ‘Do you feel you are as confident as you want to be when assessing a patient? Do you feel you have everything you need (recording procedures, guidance, etc)?’”

Continuous improvement

• Recognise that perfect solutions may not exist initially, but continuous effort and adaptation is crucial. Implement small tests of changes (a QI approach) while keeping the broader goal in mind. “The health inequalities issue is pressing, so don’t allow complexity around language etc to prevent action,” says the QI Coach. “We had a parallel inquiry running around the changes while still having challenging conversations about language and terminology. We were not satisfied with the language we settled on, but that did not prevent us from progressing iteratively with the work.”

Skills development and support

• The engagement and commitment of senior leaders is essential to ensure prioritisation of this patient safety issue in already crowded trust agendas.

• Understand the terrain: provide training to help staff understand and navigate the complexities of skin colour bias in clinical settings. Create the conditions in which staff can engage without fear of judgment for their levels of knowledge in this area. Support staff to become familiar and to develop their capabilities to spot bias and know how to respond to it. Anti-racism training in the NHS plays an important part: “Development opportunities provided by brap and others enabled me to find the entry point and approaches that better enable people to engage in the dialogue,” says the Clinical Quality Assurance Lead. “People may still feel uncomfortable, but the conversations start to feel less threatening, more familiar, and we develop our competencies to act effectively to improve care.”