We used modified Cochrane Review strategies to search ten databases (Medline, Embase, CAB Abstracts, Global Health, PsychInfo, Cochran Library, Scopus, Academic Search Complete, CINAHL, and ProQuest Central) [8–12]. The full search strings for each database are available in Appendix Table 1. Non-English language studies were excluded, and the publication date was restricted to January 2000 to April 2024. The search was conducted in April 2024 and returned 3416 abstracts; all articles were downloaded into EndNote X9.3.3 (Philadelphia, PA).

Four authors (C.J.G., A.F.A., X.R., and J.Z.) conducted a preliminary review of abstracts with at least two authors assessing each full article for relevance. Articles were eligible for inclusion if they met both of the following: 1) vaccine coverage data for rural or both rural and urban adult populations were reported and 2) individual, interpersonal, organizational, or society-level factors contributing to coverage differences was assessed. To the authors knowledge, this is the first review of its kind that addresses rural adult immunization. Therefore, we did not apply a pre-established definition of rurality; instead, we utilized the definitions provided in each individual study. Consensus was required for inclusion and disagreements were settled by a third reviewer (M.C.L. or J.M.N.). References for included articles were manually checked to identify relevant studies not captured in the initial search.

Of the 3416 abstracts returned by our search, eight duplicative references were removed. An additional 3174 were excluded after preliminary review for the following reasons: abstract unavailable; duplicate abstract; epidemiological study (e.g., studies estimating prevalence, morbidity, and mortality); unlicensed (e.g., HIV) or nonroutine (e.g., anthrax or cholera) vaccines; methodological study; no within-study comparison of urban and rural populations or data not stratified by rurality; not a study of differences in vaccine receipt; adults were not the subject of the study or separate data on adults was not available; study not based in the United States; and study did not report original data. Many excluded abstracts fell into multiple exclusion categories (Fig. 1).

The full texts of 235 abstracts were obtained and assessed for eligibility. Following the review, 190 additional studies were excluded: 44 were not stratified by rurality, 34 did not assess reasons associated with observed differences, and 112 were excluded for reasons noted above. One additional study was identified during the manual bibliography review. A standardized template was used for extracting the included articles.

As of the time the review was conducted, the Advisory Committee on Immunization Practices (ACIP) recommended the following vaccinations to adults ages 19 years or older [13]: COVID-19; influenza; respiratory syncytial virus (RSV); tetanus, diphtheria, pertussis (Tdap); pneumococcal; serogroup A,C, W, Y meningococcal, serogroup B meningococcal; Haemophilus influenza type b; inactivated polio; measles, mumps rubella (MMR); varicella; human papillomavirus (HPV); Mpox; hepatitis A; hepatitis B, and Zoster vaccines. Individual recommendations may vary based on health status, age, and the presence of additional risk factors. Additionally, although not routinely recommended, Tdap, MMR, HPV, varicella vaccines may be indicated for certain adults if no record of childhood vaccination is available.

This review is grounded in the COM-B model for behavior change which outlines three essential conditions for voluntary behavior: capability, opportunity, and motivation [7]. Capability pertains to an individual’s ability to take action. Opportunity involves external factors that facilitate action. Motivation encompasses both conscious decision-making and emotional responses. Both Capability and Opportunity conditions are further defined by either psychological or physical abilities associated with an action; Motivation can be described as reflective intentions or automatic impulses. Findings are organized and summarized by COM-B model condition and sub-level.

Two coauthors (C.J.G. and A.F.A.) independently reviewed each non-intervention study and extracted author name(s), title, publication year, funding source, methods, study population demographics, rurality definition, data metrics and analysis approach, and results.

We used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to review studies on interventions to increase vaccine uptake in rural adult populations [14]. Included studies assessed immunization of adult populations 18 years of age or older using either single-level or multilevel interventions for any ACIP-recommended adult vaccination [13], and reported vaccination uptake or changes in uptake over time by rurality.

Articles were not restricted by study type (e.g., randomized controlled trial, cohort, longitudinal), intervention type, population evaluated, or whether a priori intention to reduce rural and urban differences in adult vaccination coverage was evaluated. We used the following PICOT question components to assess and compare included intervention studies: Population (P) rural and urban adult populations in the U.S. eligible for vaccination; Intervention (I) Providers, systems, or patient-specific intervention; Comparison (C) No intervention; Outcome (O) Difference in vaccination coverage or vaccine uptake by rurality; Time (T) Before and after intervention.

Two coauthors (C.J.G. and A.F.A.) independently reviewed each intervention study and extracted the author(s), year of publication, funding source, study design and setting, timing, study population, vaccine(s) of interest, intervention(s), comparison groups, outcomes, results, strengths and weaknesses of the study, and level of evidence (Table 2). To assess the risk of bias, each study was independently evaluated based on the U.S. Preventive Services Task Force hierarchy of research design [15]. This hierarchy includes Level I (randomized controlled trials), Level II (controlled trials without randomization, cohort studies, or case-control studies), and Level III (opinions or case reports). Additionally, studies were graded for internal validity using the following criteria: initial assembly of comparable groups; maintenance of comparable groups; measurement; intervention definition; outcomes; analysis [15]. Validity scores were defined as “good” (meets all criteria), “fair” (does not meet all criteria), or “poor” (contains a fatal flaw). Quality factors such as research design and methodological features, including the use of comparable groups, measurements, potential confounders, analytical approach, and outcome ascertainment, were assessed. Due to the variability in study designs and reporting of outcomes, we did not conduct a quantitative analysis or compute summary outcome measures.

Nineteen (42.2 %) of the 45 included studies assessed COVID-19 vaccination in rural and urban, or rural-only adult populations; most (16 of 19; 84.2 %) studies focused on either Opportunity or Motivation-level factors (Table 1). The second most commonly evaluated adult vaccine was influenza (14 of 45, 31.1 %), with six studies (6 of 14, 42.9 %) assessing physical opportunities to vaccination. The Capability level had the fewest number of studies across all vaccines for which publications were available (9 of 45, 20.0 %).

We identified six studies addressing psychological capability for vaccination among rural adults [16–21]. One study of rural school employees found high MMR vaccine knowledge, acceptance, and perceived effectiveness [16]. In contrast, knowledge of HPV disease or vaccine was highly variable among the general population, with rural adults demonstrating lower knowledge as compared to their urban counterparts [17]. For physical capability, transportation challenges such as long travel distances were a predominant barrier to vaccine access among rural but not urban adults and resulted in lower COVID-19 [18,19] and HPV [20] vaccination coverage. However, a study assessing barriers to flu vaccination during a natural disaster observed veterans living in rural areas had higher vaccine coverage as compared to urban areas [21].

Two studies identified cost as a substantial barrier to adult vaccination uptake within rural areas [18,22]. Health insurance status influenced access to vaccinations regardless of rurality, with insured individuals showing significantly higher vaccination coverage [23–27]. Among the uninsured, pregnant women in rural areas were less likely to receive flu or Tdap vaccinations compared to their urban counterparts [23,24]. However, pregnant women in rural areas with public insurance exhibited higher vaccination coverage for both flu and Tdap vaccines compared to publicly insured women living in urban areas [24].

Nine studies identified access to healthcare providers was directly linked to vaccination coverage regardless of rurality [23,25,28–34]. In comparison to urban adults, low numbers of physicians per capita correlated with decreased COVID-19 vaccination coverage among rural adults [25,28]. Among rural physicians, obstetricians were less likely than family physicians to report available flu vaccines and often lacked appropriate vaccine storage units [32]. Rural adults were less likely to report receipt of provider recommendation for vaccination as compared to their urban counterparts [22,31,33]. Within rural Latino communities, language barriers and inadequate cultural competency of healthcare professionals (HCPs) further complicated access to COVID-19 vaccinations [34].

Five studies identified community acceptance as a crucial factor in shaping vaccination behaviors in rural populations, specifically by poverty status [35,36] and health professional shortage area [37–39]. Within rural communities, individuals hesitant to get the COVID-19 vaccine often felt that medical professionals and important people in their lives disapproved of them receiving the vaccine [35]. Similarly, younger rural women were more likely to receive the HPV vaccine if they believed that someone in their family supported their decision [36,38]. Additionally, “Spheres of Silence” surrounding HPV vaccination were observed within rural Appalachian communities for both provider and community contexts; discussions regarding HPV vaccination were further complicated by maternal [39] but not paternal [38] influence. Community health resources, such as the number of non-religious and nonprofit organizations per 1000 residents were associated with increased COVID-19 vaccination intention among rural adults; conflicting findings were observed for adults residing in urban areas and dependent on metropolitan statistical area distinctions [27].

Eight studies among rural populations found mistrust of the government and deeply held religious beliefs were significant reasons cited by patients in rural communities for declining COVID-19 vaccination [18,34,40–44] and HPV vaccination [45]. For HPV vaccine, lack of trust was compounded by a “Protection Dilemma”, where skepticism toward vaccine advertising and negative perceptions of pharmaceutical companies contributed to low adult vaccine uptake within rural Appalachian communities [39]. Effective messaging strategies tailored to specific religious beliefs or that prioritized open conversations while avoiding fear-based language was found to be essential when addressing these concerns among adults living in health professional shortage areas [46].

Four studies found intent to vaccinate was highly correlated with the perceived benefit of vaccination regardless of rurality [20,47–49]. Within rural communities, lack of intent to complete HPV vaccination was associated with the increased likelihood of not finishing the full vaccine series [20]. Intent to receive flu vaccination was moderately correlated with the belief that vaccination is effective, but no correlations were found with the belief that vaccines are not effective [48]. Within rural communities, low perceived risk of infection or severe disease corresponded with reduced intent to receive COVID-19 [18] and HPV vaccination [39].

One study found as population density decreases, behaviors promoting others’ well-being increased due to a greater perceived impact, including for COVID-19 and flu vaccination [50]. However, although the perceived disease risk for COVID-19 was generally higher in rural areas compared to suburban and urban residential types, COVID-19 vaccination coverage for migrant adults remained low [51]. The perception of vaccination as a personal choice, most notably within low-income rural communities [35,45], heavily influenced individual willingness to receive adult vaccination, particularly for COVID-19 vaccination [35]. Engagement strategies within health professional shortage areas that emphasized how messages were delivered and included options significantly impacted perception by ensuring individuals felt a sense of control over their personal health decisions [46]. Concerns about perceived behavioral control were not observed within health professional shortage areas for routine adult vaccines such as Tdap, Varicella, HPV, Hepatitis A/B, and meningococcal vaccines [37].

Five studies found that fear played a significant role in shaping community attitudes toward vaccination; vaccine-hesitant individuals within rural communities often expressed heightened concerns regarding vaccine side effects associated with COVID-19 and flu vaccination [53]. Within migrant [52] and rural [54] communities, conspiracy theories also significantly influenced adult vaccine uptake including the belief that COVID-19 vaccines caused harm or were part of broader population control efforts [44,52]. Additionally, a “Threat-type Dilemma” arose concerning HPV vaccination, where rural Appalachian parents grappled with fears surrounding vaccine side effects alongside the risk of their daughters developing cervical cancer [39]. General perceptions that vaccines were painful further contributed to low adult vaccine uptake within rural communities and health professional shortage areas [30,47].

This review identified six intervention studies that addressed our PICOT question (Table 2) [50–60]. Three studies met the inclusion criteria for direct evidence (Level I or II, good or fair quality), and the remaining three studies met the inclusion criteria for indirect evidence (Level I or II or III, fair or poor quality). Included studies examined single or multilevel interventions aimed at increasing the availability of flu, pneumococcal, HPV, or COVID-19 vaccination among rural adult populations.

Three prospective studies provided direct evidence of interventions to increase vaccine uptake among rural adult populations [55–57]. The first good-quality study conducted was a randomized, controlled trial multi-site intervention grounded in the Theory of Planned Behavior. Of patients who received the intervention, 43.3 % completed the HPV vaccine series compared to 31.9 % of patients who received standard care (p = 0.03) [56]. The second good-quality study used a double-control, randomized trial design to test a social media intervention on a national sample. The findings of this study demonstrated Public Service Announcements (PSA) featuring testimonials from clinicians were more successful in increasing intentions to receive the primary COVID-19 vaccine series than two control PSAs (p = 0.015 and p = 0.01, respectively) [57]. The fair quality study used an unblinded pre-post-intervention design with a moderate sample size (n = 657) and found a significant increase in flu vaccination coverage for all patients 65 years and older presenting to the clinic post-intervention (28.0 % at baseline and 54.0 % post-intervention, p < 0.05) [55].

All three studies identified as poor quality implemented educational interventions [58–60]. A comprehensive multilevel intervention to increase vaccination among rural adults living in West Virginia demonstrated low income and low educational status were both significantly associated with low COVID-19 vaccine uptake (p = 0.005); the educational intervention increased willingness to receive vaccination by 11.4 % [59]. However, intervention effectiveness could not be determined because all participants received the intervention and no pre/post intervention design was used. Although similar results were observed for an intervention to increase HPV vaccine knowledge among adults ≥18 years and older (p < 0.05), the analyses were limited by poor study design and small sample size [60]. A non-randomized, experimental study among Medicare beneficiaries (aged ≥65 years) observed a 3.7 percentage point increase in pneumococcal vaccination claims following receipt of educational campaign materials and mailed vaccine reminders; however, the lack of a non-intervention comparison group prevented assessment of the intervention’s true effect [58].

The strengths of the studies providing direct evidence were multilevel interventions based on a health theory. Additionally, all three studies analyzed and reported results for all comparison groups; one fair-quality study measured vaccination coverage at baseline and post-intervention [55]. In contrast, the weaknesses of the indirect studies include the following: no or incomplete comparison groups, no adjustment for potential confounders, and a higher risk of potential bias (e.g., social desirability bias). Overall, identified studies evaluating interventions to increase vaccine uptake among rural adult populations were limited and did not adequately address individual or systemic barriers to vaccination.

Sixteen (35.5 %) studies defined rurality using Rural-Urban Continuum Codes [16,17,20,23–25,27–29,41,48–50,54,58,60]. The remaining 29 studies used the following classification systems or definitions: Rural-Urban Commuting Area Codes (7 of 45; 15.5 %) [18,19,21,32,43,57]; health professional shortage areas (7 of 45; 15.5 %) [31,37,38,46,47,55,56]; migrant communities (3 of 45; 6.6 %) [40,51,52]; poverty status (2 of 45; 4.4 %) [35,36]; previous research in differences of vaccination coverage (2 of 45; 4.4 %) [39,45]; self-report (2 of 45; 4.4 %) [33,42]. Six (13.3 %) studies did not specify a definition of rurality [26,30,44,53,59].