Evidence reversal: Towards awareness of the phenomenon in library and information science

1. The role of evidence in LIS practice

Evidence-Based Medicine (EBM) is now very established in Western medicine. The EBM approach promotes the integration of individual clinical expertise, research evidence, and patient preferences for the improvement of decision-making (Sackett et al., 1996). This integration has had positive outcomes which led other fields to consider the same approach. Of mention, evidence-based library and information practice (EBLIP) was initiated by adapting EBM into evidence-based librarianship (EBL) (Booth, 2002; Crumley & Koufogiannakis, 2002; Eldredge, 2000). However, while transposing the concept and framework of EBM to EBLIP, challenges encountered by EBM might also become apparent to EBLIP.

In brief, EBLIP aims “to improve the profession of librarianship by asking questions as well as finding, critically appraising and incorporating research evidence from library science (and other disciplines) into daily practice” (Crumley et al., 2002, p. 62). In other words, librarians and information specialists are encouraged to gather and evaluate the available information to inform their practice. This constant use of updated information introduces an important notion of fluidity in practice. More specifically, a librarian might be doing something for many years before a study demonstrates what s/he is doing to be inessential. Therefore, a practice might change over time with new information emerging from research in the field.

EBLIP is a growing field with a peer-reviewed, open access journal EBLIP (https://journals.library.ualberta.ca/eblip/index.php/EBLIP/about) available to librarians and information specialists to submit original research and commentaries on library and information practice. As the body of evidence in library and information science (LIS) grows and more information becomes available, multiple sources of evidence can quickly become difficult to manage and incorporate into practice, especially when the evidence is contradictory. In addition, when evidence from LIS is missing, the gaps may be filled with evidence from other disciplines that is not necessarily applicable, creating another layer of complexity (Crumley et al., 2002).

With the increasing availability of scholarly work in LIS, contradictions between the evidence generated and the evidence available are bound to emerge. When this occurs, librarians and information specialists may witness the occurrence of a phenomenon recently described as ‘Evidence Reversal’ (ER). In the work described below, the authors want to share their experience in health information research and raise awareness of librarians and information specialists with respect to ER.

2. What is evidence reversal?

Before moving forward, the term ‘evidence’ here refers to the facts and information indicating whether a proposition is valid (Simpson & Weiner, 1989). An ER occurs when original information supporting a claim is contradicted by newer information derived from a study of higher quality, meaning the study was more rigorous and less biased (Sutton et al., 2018). As a concept, ER is applicable to research and scientific inquiry in all types of fields such as in public health (Sutton et al., 2018).

An example of ER in medicine is the prescribing of hormone replacement therapy (HRT) to otherwise healthy postmenopausal women. Based on the findings of observational research, in the 1990s, many physicians recommended this intervention to prevent cardiovascular disease. Subsequently, a randomized trial (a study of higher quality) concluded that HRT was ineffective (Boardman et al., 2015; Manson et al., 2003). Ergo, many women stopped HRT as this information integrated into clinical practice. Similarly, we contend that new and contradictory evidence in library science may result in a change in practice.

Also from the medical field, we learn that the lack of awareness of ER is a major obstacle to optimizing evidence-based practice. ER in medicine, also referred to as medical reversal (MR) is gaining attention, but only a small fraction of PhD scientists and health professionals are aware of this phenomenon, librarians and information specialists included (Prasad & Cifu, 2015). As a result of MR, harms have been associated with clinical practices that should never have been adopted in the first place. For a list of some of these practices and harms, see Prasad et al. (2015).

3. Why is evidence reversal a relevant concept for library and information science?

This phenomenon is relevant to librarians and information specialists for two main reasons. First, knowledge of ER may improve the services offered by librarians and information specialists to researchers, students, and practitioners working in academic settings and in health disciplines, although the authors recognize that librarians and information specialists are a resource for professionals and researchers across a plethora of fields, educating and providing services to all ‘evidence-seekers’. For example, researchers who produce systematic reviews of research findings can request the help of a librarian. When such reviews are used to inform policy, or any other decision, being aware of this phenomenon might affect the outcome. As this is not limited to one particular field, the range of potential beneficial impact is substantial. Thus, knowing about ER will help librarians and information specialists guide and support their users who stumble across conflicting evidence.

Second, librarians and information specialists will very likely be confronted with at least one, if not many ERs in their own discipline and career. Results from an empirical research have confirmed that librarians gather various types of evidence from different sources to be better informed before making a decision (Koufogiannakis, 2012). As one source is typically not enough (Koufogiannakis, 2012), many sources are used. This increases the opportunity to find contradicting evidence. In addition, if LIS researchers use methods and engage in studies which are more prone to bias, this might lead to a greater number of ERs (Crumley et al., 2002). If that is the case, being aware of, and understanding the phenomenon of ER may allow librarians and information specialists to optimize how they deal with conflicting evidence before integrating it into practice.

In addition, knowledge about the phenomenon will become increasingly important as EBLIP becomes more prevalent. Indeed, EBLIP not only encourages library and information professionals to gather and critically appraise evidence before incorporating it into practice, but it also invites them to conduct high quality research (Crumley et al., 2002). Understanding of the concept of ER may lead LIS researchers to produce the best possible evidence considering their context and find solutions to apparent contradictions. Lastly, understanding ER is important for educators of future LIS professionals for all the reasons mentioned above.

In all, evidence is a fluid currency. It can be inconsistent and contradictory, leading to ER. An understanding of this phenomenon may improve the services librarians and information professionals provide to all ‘evidence-seekers’. It is therefore valuable to be aware of this phenomenon, as librarians and information specialists progressively build upon EBLIP.