Most behavior change communication guidance for healthcare professionals in general practice is not clearly supported by relevant cited evidence, according to a systematic review published in BMC Medicine.
Across 1,163 specific communication recommendations drawn from 84 guidance sources, only 233 (20%) were supported exclusively by relevant cited evidence. Although 677 recommendations (58%) included at least one citation, fewer than half of the 3,640 total citations (39%) were considered relevant to the specific claim they were cited to support.
The findings suggest that many recommendations may not be transparently linked to evidence, though the researchers note that some documents may have drawn on evidence without explicitly connecting citations to individual statements.
Methods
Schwarze-Chintapatla and colleagues conducted a systematic review without meta-analysis (SWiM) following Cochrane-recommended processes and reported findings according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
Two search approaches were used: electronic database searches of MEDLINE, Embase, PsycINFO, CINAHL, Scopus, and Web of Science (through June 2023), and hand-searching of national guidelines, websites, training resources, and professional magazines (through August 2023). Eligible sources provided verbal health behavior change communication guidance targeting UK general practice health care professionals or were country-nonspecific.
Two reviewers independently screened titles, abstracts, and full texts. One reviewer extracted individual communication recommendations—defined as statements providing guidance on how to verbally communicate with patients about health behaviors—and identified all associated citations. Citations were assessed for relevance by one reviewer in discussion with a second; 100 randomly selected evidence documents were evaluated in duplicate.
Evidence was classified as relevant if it addressed behavior change communication in a clinical setting and involved an adult population.
Key Findings
The 84 included guidance sources comprised 33 national clinical practice guidelines, 16 journal articles, 15 websites, and 20 professional magazines. Recommendations addressed diet and physical activity (434 recommendations), general health behaviors (471), alcohol use (116), smoking cessation (102), and sexual health (40).
Citation practices varied substantially by source type. National guidelines and journals included more citations per recommendation than websites and magazines. However, national guidelines had the lowest proportion of relevant citations (38%). Magazines had the highest proportion of relevant citations (62%), followed by journals (55%) and websites (51%).
Variation also emerged across health behavior categories. Alcohol-related recommendations were most likely to include citations (83%) and the highest proportion of relevant citations among cited documents (60%). In contrast, smoking cessation recommendations, despite a 73% citation rate, had the lowest proportion of recommendations supported exclusively by relevant citations (11%). Diet and physical activity recommendations generated the largest number of citations overall (2,183), yet only 27% were relevant to the specific claims.
Of the 677 recommendations that included citations, 431 (64%) contained at least one relevant citation; however, 198 of those also relied on citations deemed irrelevant to the specific recommendation. Only 233 cited recommendations (34%) were supported entirely by relevant evidence. Sixty-one recommendations (5%) relied solely on irrelevant citations, and 486 (42%) included no citations.
Systematic reviews were the most frequently cited evidence type, followed by qualitative studies and quantitative studies. Eighteen recommendations explicitly claimed an evidence base without providing a supporting citation, and 39 cited evidence described as inference rather than direct support.
Limitations
The researchers note that guidance sources listing references at the end of a document without linking them to specific recommendations were coded as uncited, even though such documents may have drawn on evidence. As a result, the findings reflect transparency and citation practices rather than the total absence of underlying evidence.
Hand-searching may have missed some website-based documents, though systematic omission was considered unlikely. Books cited as evidence could not be assessed for relevance. The review was limited to UK-applicable sources, which may restrict international generalizability.
Importantly, the review assessed citation presence and relevance—not the methodological quality of the cited evidence itself—an area the researchers identified for future study.
Implications
The researchers note that effective communication strategies can be counterintuitive and may conflict with commonly taught approaches. They suggest that the absence of clearly linked, relevant citations in training materials and guidelines may contribute to primary care professionals’ reported difficulty conducting behavior change conversations.
“Policymakers and training developers should consider providing citations wherever specific claims are made, to increase transparency and facilitate assessment of the quality of the supporting evidence,” the researchers wrote.
The researchers reported no competing interests.
Source: BMC Medicine
