By P.F. Anderson and LaTeesa James
PubMed special queries are master search strategies on common, important, but challenging health topics, which are usually peer-reviewed. Disability certainly qualifies as a common, important health topic that is challenging to search – and yet, no PubMed special query exists for it.
This oversight is concerning. Put simply, it’s important that disability as a topic is easily identifiable in the scholarly record. While there are specialized databases for the topic, such as NARIC’s REHABDATA, the global research community is most likely to begin with PubMed. Currently, structured vocabulary terms related to disability are scattered throughout PubMed’s medical subject heading (MeSH) trees, often without cross references to facilitate discovery, forcing discovery to largely depend on the unstructured language of whatever words were chosen by the authors to describe their research topics. This lack of consensus and consistency across professional disability terms, categories, and definitions contributes to a lack of comparability, discoverability, and access for disability data as well as research.
As research partners with the University of Michigan’s Center for Disability Health and Wellness, we decided to address this gap. It’s been an adventure, and the deeper into the project we go, the more complicated it becomes. This article describes just a few of the challenges we’ve encountered.
Disability is TOO BIG
At the outset of this project, we examined the PubMed search filter for cancer for inspiration, as it is one of the largest, with over 1,200 search statements. Then we started seeking definitions of disability with clear boundaries of the topic and with some general indication of consensus as to the authority and utility of the definition.
Eventually we realized that because of the specific medical context of cancer, compared to the breadth of disability across all aspects of life, disability as a concept is much, much larger than cancer. The idea of disability can include every part of the body or mind, and the impacts of disability can touch on every part of lived experience. Should we define it by functionality, diagnoses, interventions, needs, outcomes and impacts, environments, or something else? Do we include chosen language as well as professional? Legislation and accommodations as well as experiences?
There are countless definitions of disability, what it is, what it isn’t, what it means, how it impacts someone’s life or the lives of people near to them, who’s allowed inside disability communities or allowed to claim disability identities, who isn’t, and who is permitted to make these decisions. So, how can you define disability in a way that has boundaries tight enough to keep the search results on topic, but which is still broad enough to make it possible to discover these myriad aspects of the topic? Search experts call this sensitivity versus specificity, and have fierce debates over the best ways to focus a topic so that it maximizes the usefulness of the results.
Also, the idea of disability is a moving target, shifting rapidly over recent years as new perspectives and insights have revealed deep connections between ableism and racism, sexism, ageism, classism, heterosexism, antisemitism, and more, with stigma and discrimination impacting many aspects of life. Look at the intersections of ableism with police violence, climate change, and disaster planning, not to mention the COVID pandemic. Further, there can be specific types of disability language that are unique to a country or a culture.
Disability language is TOO MUCH
In our work, we identified many definitions of disability, but no consensus. We tried breaking disability down into categories — physical, sensory, developmental, learning, cognitive, psychological, behavioral, emotional — as a strategy utilized by the health disparities team, only to find the categories blend and blur, and different experts and authorities list different categories with different language to describe them. Are Parkinson’s disease or multiple sclerosis considered to be physical disabilities or cognitive ones? When is a person with a disability diagnosis functioning well enough that it isn’t disabling? These were topics of discussion in our research team.
Healthcare disability experts often, of necessity, focus on the human body and mind for these definitions of disability, which may leave out such critical disability concepts as education, employment, accessibility, accommodations, assistive technologies, the legislation framing these concepts, and the regional or national contexts constraining these, not to mention how these shape policy and practice. Similarly, data sources containing disability information to support policy formation typically use the language that was standard when the data first began being collected, creating a lack of common language for the same concepts between resources and databases. This consistency of terminology over time provides internal reliability in the dataset, but can lead to other challenges.
Disability language has SO MUCH STIGMA
Language surrounding disability is ever changing, and often associated culturally with stigma. There are plenty of examples where disability language has been used to stereotype, demoralize, and ostracize people of the disability community. In an attempt to combat this, communities change language to new terms not currently stigmatized, and then the broader community creates stigma around the new terms.
Librarians and informationists encounter complications in designing expert search strategies to gather literature on topics containing historically divisive or contested concepts. This is particularly important in evidence synthesis methods, such as systematic reviews, when research on these concepts includes outdated and now offensive language. According to library informationist Whitney Townsend, “Some searches, due to the nature of their topics, must include antiquated, non-standard, exclusionary, and offensive terms in order to identify all published literature on that topic.” Stigmatizing and offensive disability language, like racist language, may need to be included in order to gather all of the medical literature on a topic over time.
Researchers and editors sensitive to the impacts of this language on their audience and the acceptance of their research findings may hesitate or reject the idea of using such terms within the search strategies. To remedy this problem, Townsend and her team developed a sample statement for inclusion in the methods, outlining why it may be necessary and methodologically sound practice for antiquated or offensive terms to be included in the search strategy, and framing this with best practices for publication to mitigate exposure to the problematic language and further to contextualize it appropriately through efforts to acknowledge, explain, protect, and ultimately include the reader. Statements like these can help research teams and others to understand the importance of including stigmatizing language in evidence syntheses. Leaving out offensive language does a disservice to the disability community in that new research would be based on incomplete information, and could result in negative impacts for diagnosis or treatment.
The challenges we describe here illustrate some of the ways in which historic systems and technologies create structures locking us into the technoableism of the past, and perpetuating it as we attempt to move into the future. The difficulty of shifting these archaic infrastructures is increasingly evident, as highlighted by the recent decision of the NIMHD to exclude persons with disabilities as a health disparity population, and the pushback to this decision from the communities of disability researchers, policy makers, and stakeholders. NIMHD explained their decision as rooted in the lack of a consensus definition as to what disability means, and with that the problems described here contribute to a circle of ongoing harm.
P.F. Anderson is an Emerging Technologies Informationist for the Health Sciences at the University of Michigan.
LaTeesa James is an Informationist for Cardiovascular Medicine, Epidemiology, Kinesiology, Obstetrics/Gynecology, Medical Education, and Sanford House at the University of Michigan.
