Health information equity: Rebalancing healthcare collections for racial diversity in UK public service contexts

2.1Decolonisation and collection management

Decolonisation has been a topic of LIS discourse for over a decade, requiring both acknowledgement of coloniality and action to dismantle resulting power structures, thereby de-centralising white normativity (Crilly & Everitt, 2022, p. xxiii; Hartland & Larkai, 2020). The movement has more recently impacted on medical education (Lokugamage et al., 2020), where clinical knowledge bases have deeply colonial roots (Roberts, 2011, p. 98). Decolonisation of healthcare information can target both deficits in representation and the resultant inequitable outcomes through addressing and challenging inequitable power structures; developing more inclusive curricula, norm values, and reference materials to encourage person-centred care for more diverse populations (Lokugamage et al., 2020).

Collection policy sets resource parameters for both the present and future (Fieldhouse & Marshall, 2012, p. vii). Shifts in community needs, or in communities themselves, necessitate libraries and information services to adapt (Gerber, 2017). The UK’s ethnic and racial diversity is forecast to continue increasing (Otu et al., 2020), predicating ongoing evolution in public services such as healthcare. While health inequality cannot be eradicated through improved representation of diversity (Louie & Wilkes, 2018), collection policies can give operational clarity to equitable aims, such as decolonisation, creating opportunities for tangible change.

A key purpose of healthcare information services is education, both in training and throughout a clinician’s career (Weightman et al., 2009). Gerber notes Continuing Medical Education (CME) and librarianship’s shared objective of knowledge dissemination as a unifying purpose for information collections (2017). This shared purpose may be pivotal in dissemination of knowledge now known to have been deficient in earlier career training, building on the benefits and impact of decolonisation in medical education.

Medical libraries have been pioneers in using impact assessments to measure and report their efficacy in practice (Robinson, 2010, p. 212; Brettle et al., 2016). Positive impact depends on provision of accurate and adequate resources, however. Information for healthcare is conventionally produced by a narrow group of authors through ‘traditional’ publishing routes (Robinson, 2010, p. 72). Leung and López-McKnight assert LIS’s ‘foundational role’ in gatekeeping knowledge (2021, p. 7), with traditions and processes upholding legacy power structures, potentially excluding marginalised voices (Inefuku & Roh, 2016; Delgado Bernal & Villalpando, 2002). Instantiation of more diverse, representative, and therefore impactful, information in the healthcare knowledge base may require LIS innovation. To this end, Herther (2020) advocates for information professionals to influence vendors and creators, while Gwynn et al. (2019) propose LIS involvement in authorship.

2.2Social justice, a LIS value

LIS’ long established social justice discourse is foundational to pursuit of racial equity through collections and resources; a contemporary iteration of the ‘common good’ ethos espoused by Lankes (2016, p. 105). Information professionals have, however, been criticised for avoiding difficult choices under cover of neutrality (Gibson et al., 2017), with Hudson asserting that neutrality makes libraries ‘sites for the perpetuation of white supremacy’ (2017). Morales et al. call for library services to prioritise both theoretical and practical social justice, action alongside intention, to counter this (2014), while Clarke (2019) encourages libraries and librarians to ‘self-decolonise’.

Critical librarianship has been central to contemporary LIS social justice discourse and reflective activism (Garcia, 2015). However, critiques of the movement include perceptions of social media contingency, disconnect from practice, performativity, exclusion, and inaccessibility (Ferretti & MICA, 2020; Nicholson & Seale, 2018, p. 2). To be an effective vehicle for inclusion, critical librarianship must be inclusive itself and must deliver action, not just reflection. Greater potential for social justice may lie in engagement with organisational policies than the individual advocacy considered typical of critical librarianship (Nicholson & Seale, 2018, p. 2).

Hudson critiques a singular focus on staff diversity as limiting social justice impact in LIS, exhorting the profession to instead adopt multi-pronged ‘anti-racisms’ (2017). Morales et al. (2014) call for LIS to address both bodily representation and structures of power and privilege, with a parallel discourse occupying the UK medical domain, Wong et al. documenting a shift ‘from a diversity-oriented to a decolonial framework’ in medical education (2021). Meghji (2021) notes CRT’s use in the legal domain as a tool for ‘an institution that was supposedly race-neutral … to work against its structural racism’, a positioning that suggests its potential as a framework for LIS decolonisation, mirroring the shift from a narrow focus on diversity.

2.3CRT as a library and information tool

Stauffer notes CRT’s surprisingly low profile in LIS, given its utility as a framework for analysis (2020), notably useful where provisions of law are for equality of opportunities and access, not outcomes (Meghji, 2021), as has been the case in healthcare. However, Dunbar notes a recent, gradual adoption in LIS discourse (2021), while Leung and López-McKnight propose CRT as ‘a central philosophy’ for LIS (2021, p. 7) and enumerate tenets core to contemporary understanding of CRT as relevant in the information context (2021, p. 13).

CRT was introduced to UK academia through education scholarship (Warmington, 2020), an association that continues (Meghji, 2021). Gibson et al., however, find that LIS education itself provides ‘little to no exposure to CRT’ (2018). Inclusion of CRT in LIS curricula may serve to further integrate the LIS and education domains, enabling information professionals to undertake more impactful social justice and decolonisation initiatives, particularly in environments with an educational remit such as in healthcare.

LIS, science, and medicine share practices of categorisation and classification. Roberts attests that science lends racial classifications false legitimacy (2011, p. 27). Delgado and Stefancic (2017, p. 9) and Bonilla-Silva (1997) note the dehumanising effect of such hierarchies and categorisations, notwithstanding their falsity and contingency. It is difficult to reconcile such foundational misconceptions and harms with expectations of scientific and informational integrity. However, Delgado and Stefancic establish a cornerstone of CRT in the assertion that adoption of colour-blindness, and avoidance of categorisations, may impede elimination of systemic racism and resultant disparities, proposing instead an ‘aggressive colour-conscious’ approach (2017, p. 27).

Race is a necessary element of health data, with racial categorisations exposing under-representation, risks, and inequities (Otu et al., 2020; Yearby, 2021). The trend and the necessity are for increasing ‘acquisition and utilisation of disaggregated [race] data’ (Lester et al., 2020), in order for research and resources to become more representative. Otu et al. herald the treatment of race in UK COVID-19 data as indicative of the future, emphasising connections between race, socioeconomic determinants, and health disparities (2020). Saini, however, cautions risks of race-based causation assumptions (2019, p. 214). Not only data standards, but ethical frameworks for nomenclature, data management, critical analysis and appraisal must be applied, with CRT serving as a potential reference point.

2.4Literature analysis summary

This topical review of the LIS and interdisciplinary scholarly communications landscape has established that collection and resource management form key elements of decolonisation practice, and that collection policy is a powerful signal for standards and priorities driving equity and social justice. Strong links between healthcare information provision and education, from training to CME, have been established, with decolonisation emerging as a trend in both medical education and the information domain. Longstanding imperatives for the pursuit of social justice through information services and librarianship have been identified in the literature, with objectives of completeness and accuracy, in terms of diverse racial representation, asserted as a matter of both information ethics and collection efficacy. CRT has been appraised as an emerging LIS discourse and has been found to be an appropriate tool for analysis in scholarship and decolonisation initiatives, such as in collection management, given its acknowledgement in the LIS literature, links to education discourse, utility in disambiguating and justifying racial categorisations for information management, and the aptitude of CRT tenets for naming and examining a breadth of systemic and structural racisms.

4.1Survey results

The full questionnaire with response data is available (O’Driscoll, 2022, appendix E).

All 32 survey respondents reported being involved in collection management and resource dissemination and promotion, validating the suitability of the sample population, illustrated in Fig. 1.

Figure 1.

Collection management involvement [n= 32].

Respondents were only somewhat less likely to be involved in EDI work compared to collection work, as illustrated by Fig. 2, which depicts only four respondents, as not involved in EDI work, and 88% or 28 respondents as involved: 18 ‘sometimes’, and 10 ‘often’. This indicates that a majority of responding NHS information professionals are aware of racial equality as an organisational priority. One respondent commented that services ‘are becoming much more aware … the issues are being raised more within the [t]rust’, and another affirmed that their service ‘have been working with EDI networks … to build awareness of the library service and how it can support EDI inititives (sic) and causes’.

97% of respondents confirmed awareness of racial health disparities in the UK, with 47% highly aware and 50% somewhat aware, as illustrated in Fig. 3.

Figure 2.

EDI initiatives involvement [n= 32].

Figure 3.

Awareness of UK health disparities [n= 32].

Figure 4 illustrates respondent’ awareness of racial under-representation in the medical knowledge base. 78% of respondents, or 25, characterised themselves as ‘aware’ of under-representation in medical research, with 22%, or seven respondents ‘highly aware’, and 56%, or 18, ‘somewhat aware’. 72% of respondents, 23 in total, identified as aware of under-representation of racially minoritised patients in reference resources, with six,‘highly aware’ and 17 ‘somewhat aware’. Awareness was highest with regard to under-representation of black and brown skin in reference imagery, with 88% of respondents characterising themselves as aware, including 15 being ‘highly aware’, 13 ‘somewhat aware’, and only four responding as ‘not aware’, indicating a widespread acknowledgment of under-representation, and therefore racial bias, being a norm in collection resources.

Figure 4.

Awareness of under-representation of ethnically minoritised participants in research and resources [n= 32].

A majority of 29 respondents, 91% of the total, confirmed their belief that NHS information services can contribute to reducing health disparities, with 12 expressing this to be probable, and 17 as definitely possible, as depicted in Fig. 5. Two respondents registered the opinion that it is ‘probably not’ possible for information work to contribute to health equality, and one was ‘unsure’. No respondents selected the ‘definitely not’ option.

The only two respondents to assert that information services could ‘probably not’ affect racial health disparities were within the cohort of ‘not involved’ in wider EDI work at all, as depicted in Fig. 6. This may suggest that wider EDI involvement correlates with confidence in the potential for services and resources to contribute to health equity.

Figure 5.

Belief in information services’ ability to contribute to health equity [n= 32].

Figure 6.

EDI involvement correlation to belief in contribution to health equity [n= 32].

In order to better understand the perspective from which respondents were answering, experience of different EDI workplace initiatives was queried, as illustrated in Fig. 7. Here 31 respondents, 97%, indicated that they had taken part in EDI initiatives, five describing having experienced ‘a lot’, 15 ‘several’ and a further 11 ‘not many’. Only one respondent reported experiencing ‘none at all’. 90%, or 28 of the respondents, had experienced staff diversity initiatives, with two respondents reporting ‘a lot’, 14 ‘several’ and a further 12 ‘not many’. 23 respondents, or 72%, reported some experience of decolonising or diversifying collections, comprised of four with ‘a lot’ of experience, eight recounting ‘several’ experiences and 11 reporting ‘not many’. 16 respondents, 50% of the base, had attended unconscious bias training, comprising of three reporting ‘a lot’, seven ‘several’ and six ‘not many’ of these trainings. These findings suggest that EDI activity is established and somewhat commonplace in NHS settings, with most staff having participated in initiatives and trainings around staff diversity and bias.

The least established application of EDI work was in monitoring for systemic bias in research, with no respondents indicating ‘a lot’ of experience. Three respondents indicated ‘several’ experiences, seven confirmed ‘not many’ and a majority of 63%, or 20 respondents, indicated no exposure to this practice, suggesting that EDI experiences in the healthcare information environment are more focused on interpersonal equality, inclusion, and diversity than on resource-oriented applications addressing systemic racism.

Figure 7.

Breakdown of participant EDI initiative experiences [n= 32].

As depicted in Fig. 8, 11 respondents, 34% of the total, confirmed awareness of the GMC’s 2020 statement committing to updating medical curricula to reflect the diversity of the UK population, with a further 15 being unaware and six ‘unsure’. Of those confirming their awareness or being ‘unsure’, seven reported having added resources to the collection following the statement, six reported increasing promotion of diverse materials to medical students, and nine to clinicians. One respondent commented that their service ‘are unclear how best to diversify our collection or how to evaluate it’. Five respondents confirming awareness of the statement report being unsure whether new materials were added, and five report not adding anything new, suggesting a need for both stronger organisational promotion of statements and policies pursuing equity, and better guidance on action required in response.

Survey participant comments support the analysis, reflecting confusion and frustration at the lack of amplification for such an important shift. While positive responses included: ‘Thanks for the heads up!’, ‘Pleased to see this’, others questioned the absence of communication; ‘Completely unaware … cannot find anything from any mailing lists or any level of NHS communication’, ‘Our medical education team have not made us aware of this, nor our library network’.

Figure 8.

Awareness of the GMC statement on medical curricula [n= 31].

66%, or 21 respondents, confirmed awareness of the ‘Mind the Gap’ resource, with a further nine reporting being unaware, and two unsure. 16 had promoted it to practicing medics, and five had promoted it to students, illustrated in Fig. 9.

Comments on ‘Mind the Gap’ varied from highly positive and engaged; ‘Very welcome resource flagged to our collections liaison team’, ‘We have involved the author in some of our work’, ‘Thanks for the reminder to promote to students’, ‘Requested an electronic copy & sent it to our consultant surgeons and BAME champions’, to more hesitant; ‘Linked to it on website – promotional activities limited due to pandemic and other internal factors’, ‘Have publicised it but could do better’, ‘We are aware, but haven’t utilised it to any great extent’, perhaps indicating an absence of central guidance as to how to manage such ‘grey’ resources.

Figure 9.

Awareness of the resource ‘Mind the Gap: signs and symptoms in Black and Brown skin’ [n= 32].

4.2Interview results

A thematic table of interview response data is available (O’Driscoll, 2022, appendix H), comprising participant quotes collated into themes identified as the interviews progressed, employing elements of Pickard’s constant comparative analysis (2007, p. 241). Clear topics emerged as themes relevant to the study.

Tenet: race is a social construct

There is consensus across the literature on the non-scientific basis of race, as illustrated by the human genome project (Roberts, 2011, p. 26, p. 53; Yearby, 2021). Despite this, outdated beliefs around racial characteristics persist in medicine and science (Saini, 2019, p. 73, p. 120; Massie et al., 2021). Survey findings suggest lower awareness of critical appraisal for racial bias, as discussed by Naicker (2021), compared to other EDI workstreams, indicating a disparity in confidence to interrogate race in the scientific domain. While acknowledged as arbitrary, and non-biologic, racial categorisations are now key metrics in identifying and reducing health disparities (Pérez-Stable et al., 2021). As noted by interview participants, only by naming and tracking race in health data can we move towards solutions to embedded racisms in healthcare.

Tenet: racism is normal

Institutions without policies to ensure services adapt as populations change risk reproducing systemic racism (Blume & Roylance, 2020). Interviews found only one participating service had EDI provision within collection policy, underscoring this risk. Interview participants related a lack of knowledge and guidance for collection equity, with much current EDI acquisition reactive to requests. Survey data confirms that 72% of respondents had some exposure to diversifying collections, indicating some engagement with decolonisation. There is an opportunity to harness this engagement through policy, guidance, and structured sharing of best practice in active challenge to norms of under-representation.

Persistence of racial health disparities serves as evidence of ongoing systemic racism. Otu et al. (2020) note that Public Health England (PHE)’s June 2020 report detailing heightened COVID-19 risks faced by Black and South Asian populations was unaccompanied by proposals for targeted mitigations. Publications depicting cutaneous symptoms of COVID-19 fail to show diversity of skin tones, despite these risk profiles (Lester et al., 2020). Interview Participant 3 expressed frustration with the disproportionate focus on the existence and measurement of disparities, in contrast to a dearth of resources for researched, evidenced solutions. Health disparities are reproduced through lack of action for change.

Tenet: experiences and knowledge of racially minoritised people

Encountering inequality of representation in resources and curricula is experienced as erasure or othering by racially minoritised students and professionals (Hartland & Larkai, 2020). While interviews confirm management and HEE support for EDI work, survey results show diversity as the main EDI workstream within healthcare information services. Greater bodily diversity in the staffing and leadership of information environments benefits the knowledge base through lived experience, and may reduce barriers to change, but underlying structures and systemic inadequacies must also be addressed for inclusion to be meaningful and sustainable.

In recent years an increasingly diverse medical student body has raised awareness of systemic racism in the curriculum (Change.org, 2020; Wong et al., 2021). Mukwende’s ‘Mind the Gap’ is one example of change initiated by a minoritised student (SGUL, 2020). Similarly, LIS has seen calls for marginalised voices to be heard (Ettarh, 2018; Dunbar, 2021). Participant 3 notes the pressures of speaking out against racism and the benefits of silence to the status quo. While listening to racialised voices is key, the workload of rebalancing resources and collections must be shared across the profession, not charged to minoritised colleagues.

Tenet: interdisciplinarity

LIS inherently lends itself to interdisciplinarity, with adjacent disciplines of publishing and authorship as vehicles for change (Gwynn et al., 2019). Libraries’ capacity to collaborate, whether in establishing data standards, advocating for inclusion, or facilitation of outsider voices, can disrupt power structures (Inefuku & Roh, 2016). This study has not found that the health information sector is ignoring or blocking inclusive resources, rather that the wider information environment has not addressed deficits, and that the pace of change is slower than the pace of increasing awareness and need. To effect change an interdisciplinary approach is essential.

Tenet: critique of dominant ideologies

Science is assumed to adhere to ideologies of objectivity and evidence. Examining systemic bias in the healthcare domain necessitates questioning this. 78% of survey respondents reported awareness of racial under-representation in research and reference materials, confirming deficits in the knowledge and evidence base. While Participant 3 notes some progress in health research engagement with diverse representation, for science and medicine to be truly evidence-based research data must reflect population diversity and must include racial health disparities and potential solutions.

Critical librarianship discourse debates neutrality as an LIS domain ideology (Bales, 2017; Gibson et al., 2017). The ALA, whose guidance is often cited as the root of library neutrality, has now clarified a non-neutral position of anti-racism (ALA, 2021). It is possible that UK LIS espouses passivity more than neutrality, as illustrated by limited, reactive pursuit of EDI resources reported in interview data. Survey results demonstrate opportunities for more ambitious EDI workstreams in the health information realm, parallel to diversity work, with potential for healthcare information systems to be more active and deliberate in pursuing information equity.

Tenet: interest convergence

Bell defines interest convergence as the premise that ‘racial equality will be accommodated only when it converges with the interests of whites’ (1980). In the knowledge domain, biased or inadequate information environments are a corruption of our core purpose. Sciences, including medical and information sciences, benefit from completeness and accuracy of records (Pérez-Stable et al., 2021; Lokugamage et al., 2020). There is then interest convergence based on efficiency, efficacy, competence, and productivity, in ensuring that the full breadth of population diversity is represented in healthcare resources and collections. The LIS value of the common good exemplifies interest convergence; in reducing information inequity all groups are better served by a more effective system.

Tenet: totality of systemic racism

While addressing granular instances of systemic racism, such as under-representation in resources, is necessary and urgent, it is important that such instances are contextualised as indicative of the whole; embedded in wider structural inequity (Crilly, 2019). UK health bodies and institutions communicate intent to address health disparities through statements and policies (NHS England, 2015; PHE, 2018; Robertson et al., 2021, HEE, no date). Much operational focus however is confined to staff diversity and workplace equality. While this is important work, it limits scope to one strand of the totality of systemic racisms (Hudson, 2017). CRT as a framework, and decolonisation as a practice, enable a structured and multi-faceted approach to analysis of, and action on, a breadth of systemic racisms.