Health professions training exposes clinicians to poverty and structural drivers of ill health in ways that many other occupations do not. The authors framed this exposure as a basis for expecting higher levels of voluntary civic action. To test that expectation, the protocol defines civic and community engagement as voluntary, personal-capacity behaviours: formal and informal volunteering, voting and political acts, activism, charitable giving, and participation in NGOs or civil-society organizations. Activities performed as part of employment, clinical duties, or as a formal representative of a professional body are treated separately and excluded from the primary behavioural analyses.
The work is a protocol; no outcomes are yet available. The study team reported their intent to produce a cross-disciplinary behavioural benchmark comparing health profession populations to public and education-matched comparators, and to appraise the certainty of evidence for each pooled estimate using GRADE.
The protocol casts a wide net. Included populations are health profession students (medical, dental, nursing, pharmacy, physiotherapy, and allied health) and practising clinicians of any age, sex, nationality, or career stage. There are no geographic or date limits and no language restrictions. The authors excluded non-health students except where they serve as comparators, administrative staff without patient-care roles, samples that mix health and non-health participants without separable data, and studies whose only measured engagement is mandatory, curriculum-based community service.
Eligible study designs are cross-sectional and survey-based studies, prospective or retrospective cohorts, and mixed-methods reports with extractable quantitative components. Qualitative-only papers, editorials, letters, and case reports are excluded.
Three comparator types are pre-specified and analysed separately: (i) general adult population or community samples; (ii) education-matched adults (tertiary-educated general-population respondents); and (iii) non-health university students or graduates. The protocol prioritises within-study comparators as the primary comparative evidence. External benchmarks — notably the World Values Survey — will be used only when both reviewers agree on close matching by country, wave, item wording or construct, and recall period (within three years of the index study’s data-collection period). External-benchmark comparisons will be reported separately and treated as a sensitivity analysis.
Because educational attainment strongly predicts civic participation, comparisons with education-matched samples are treated as the principal test of whether any differences reflect the health professions specifically or merely higher education.
Primary outcomes are clearly enumerated. The protocol will estimate pooled prevalence and odds ratios for any volunteering (binary), pooled prevalence and ORs for political participation (voting, petition signing, civic meeting attendance), and standardized mean differences for continuous behavioural civic-engagement scores (e.g., behaviour subscales). Recall-period hierarchy is defined: past 12 months preferred, then nearest to 12 months, then lifetime; for voting, turnout at the most recent national election will be extracted when available.
Secondary outcomes include NGO or civil-society participation rates, charitable-giving prevalence and mean amounts (harmonised to international dollars at purchasing-power parity for the survey year when reported), activism and advocacy participation rates, hours per month engaged, and an exploratory analysis of the association between empathy and civic engagement within health profession samples.
Seven bibliographic databases will be searched from inception: PubMed/MEDLINE, Scopus, Web of Science Core Collection, CINAHL, ERIC, PsycINFO, and Embase. The authors will supplement those searches with Google Scholar (first 200 results), the WHO Institutional Repository for Information Sharing, ProQuest Dissertations and Theses, and preprint servers (medRxiv and OSF Preprint). The full search strategy is provided as supporting information in the protocol.
No date or language limits are applied. The team intends to extract comparator estimates from external datasets where appropriate (for example, deriving WVS microdata restricted by country, wave, age range, and education level), and to deposit all extracted data and analysis code in an Open Science Framework repository upon completion.
For prevalence pooling, the reviewers will use random-effects generalized linear mixed models (GLMMs). Comparative effect sizes — odds ratios and standardized mean differences — will be pooled using restricted maximum-likelihood random-effects models coupled with Hartung-Knapp-Sidik-Jonkman confidence intervals; prediction intervals will be reported. The protocol specifies separate GRADE assessments for prevalence outcomes and for between-group comparisons.
Risk of bias will be appraised with Joanna Briggs Institute checklists matched to prevalence and comparative analyses, and with the Newcastle–Ottawa Scale for cohort studies. The study team will also appraise the measurement properties of instruments used to measure engagement and will classify instruments that mix attitudes and behaviours versus those that report behavioural subscales only.
Each behaviour extracted will be coded by the capacity in which it occurred: (a) personal-capacity engagement (private citizen), (b) institution-organised or incentivised engagement (employer or training-institution organised), (c) professional-role advocacy (actions taken as a designated representative of a professional group), or (d) unclassifiable when instruments do not specify capacity. Primary analyses will include categories (a), (b), and (d) because general-population instruments often omit capacity. Category (c) will be summarised separately and not pooled with individual political participation outcomes.
When instruments combine attitudes and behaviours, the protocol requires extraction of behavioural subscales only. If only composite scores mixing attitudes and behaviour are reported, the authors will attempt to obtain subscale data from study authors; lacking that, the study will be treated as using a mixed instrument, analysed separately, and excluded from primary behavioural syntheses.
The protocol is registered on PROSPERO (CRD420261348391). The authors stated that they received no specific funding and declared no competing interests. No datasets were generated for the protocol itself; the team plans to deposit extracted data, risk-of-bias judgements, GRADE profiles, comparator estimates (including WVS-derived figures where used), and R code in an Open Science Framework repository with a persistent DOI and a CC BY 4.0 licence when the review is completed.
Limitations described in the protocol include reliance on observational designs and heterogeneity in instruments and comparators. The protocol notes that pandemic-period data will be classified explicitly (data-collection and recall periods overlapping WHO’s COVID-19 pandemic window), and that pandemic-specific volunteering items will not be extracted. The review is at the protocol stage; no pooled estimates or certainty ratings are reported in this article.