Adult vaccination coverage in the United States: A database analysis and literature review of improvement strategies

ABSTRACT

Despite vaccines being instrumental in reducing vaccine-preventable disease, adult vaccination rates in the United States (US) are below optimal levels. To better understand factors affecting vaccination rates, we analyzed trends in adult vaccination coverage using data from the Behavioral Risk Factor Surveillance System (BRFSS) and conducted a targeted literature review (TLR) on interventions to improve adult vaccination rates in the US. Both the BRFSS analysis and the TLR focused on influenza; pneumococcal disease; tetanus and diphtheria or tetanus, diphtheria, and acellular pertussis; herpes zoster; and human papillomavirus vaccination for US adults aged 18–64 years. The TLR additionally included hepatitis A and hepatitis B vaccination. Vaccination coverage rates (VCRs) and changes in VCRs were calculated using the 2011–2019 BRFSS survey data. For the TLR, the MEDLINE and MEDLINE In-Process databases were searched for articles on vaccination interventions published between January 2015 and June 2021. The BRFSS analysis showed that changes in VCRs were generally modest and positive for most states over the study period. The TLR included 32 articles that met the eligibility criteria; intervention strategies that improved adult vaccination outcomes incorporated an educational component, vaccination reminders or reinforcement at the point of care, or authorized non-clinician members of the healthcare team to vaccinate. Furthermore, interventions combining more than one approach appeared to enhance effectiveness. The strategies identified in this TLR will be valuable for policymakers and stakeholders to inform the development and implementation of evidence-based policies and practices to improve adult vaccination coverage.

KEYWORDS:

• Adult
• vaccinations
• vaccination programs
• immunization
• health policy

Introduction^a

Vaccines have been instrumental to public health by reducing the burden of infectious diseases globally, and widespread immunization is a major success story of modern medicine. Routine childhood vaccinations have tremendously reduced the impact and burden of communicable diseases; however, adult vaccination programs have not seen a similar level of success regarding vaccine uptake.^1–4 In the United States (US), the Healthy People 2020 vaccination goals were 80% for influenza among adults aged 18–64 years and 60% for pneumococcal disease among at-risk adults aged 18–64 years;⁵ however, national vaccination coverage rate (VCR) estimates were markedly lower than these targets, with data from the 2017–2018 National Health Interview Survey in the US⁶ indicating 46% of adults aged ≥19 years received an influenza vaccine and just 23% of at-risk adults aged 19–64 years received a pneumococcal vaccine. Vaccinating adults is critical to reduce the risk for vaccine-preventable diseases, especially for those with immunocompromising health conditions;⁷ addressing the gap between target and achieved adult VCRs remains a priority.

Previous studies have shown that adult VCRs are variable across states,^2,8 suggesting that state-level factors may have an important impact on vaccination coverage. In an analysis of data from the 2011–2014 Behavioral Risk Factor Surveillance System (BRFSS), interstate variability in VCRs remained high after adjusting for individual characteristics (e.g., income, education level).² Additionally, a study using 2015–2017 BRFSS data found a significant association between adult VCRs and state-level factors, including insurance coverage and participation in adult immunization information systems (IIS).⁸ However, state-level data regarding the facilitators and initiatives for improving adult vaccination coverage remain limited.

Further research is needed to inform the development of evidence-based policies and practices that could improve VCRs and help reach target goals in adults aged 18–64 years in the US. Accordingly, we conducted a mixed methods study with the objectives of better understanding trends in national- and state-level adult vaccination coverage over time and identifying potential strategies to improve upon these trends. This study comprised a quantitative analysis of US adult vaccination coverage using BRFSS data and a targeted literature review (TLR) of interventions aimed at improving adult vaccination in the US, with a focus on adults aged 18–64 years. The objectives of this study were to examine the trends in national- and state-level VCRs, evaluate changes over time in state-level VCRs, and identify state-level factors associated with increases in adult VCRs in the US.

Materials and methods

We focused on the following routine vaccinations, which are recommended by the Advisory Committee on Immunization Practices (ACIP) for the US adult population and which vary based on factors such as an individual’s age and the presence of an immunocompromising condition or other risk factors⁹: seasonal influenza; pneumococcal disease; tetanus and diphtheria (Td) or tetanus, diphtheria, and acellular pertussis (Tdap); herpes zoster (HZ); and human papillomavirus (HPV).¹⁰ Additionally, interventions related to hepatitis A¹¹ and hepatitis B¹² vaccination were included in the TLR.

BRFSS analysis

Study population

We evaluated vaccination coverage data for adults aged 18–64 years who participated in any of the 2011–2019 BRFSS surveys. To be included in the study, BRFSS survey participants must have responded to at least one vaccination question and met the age criteria and lived in the US at the time of the survey. Vaccination coverage was computed for the influenza, pneumococcal, Td/Tdap, HZ, and HPV vaccines; vaccine-specific subgroups of interest (e.g., age groups, at-risk groups) were defined according to the Centers for Disease Control and Prevention (CDC) criteria and recommendations.^10,13

BRFSS data

The BRFSS consists of core component questions that are asked in all states without modification, optional modules that states may select for inclusion without modification, and state-added questions. Vaccine-specific BRFSS survey questions and their inclusion in the BRFSS survey from 2011–2019 are presented in Table S-1 and Table S-2, respectively (Supplementary Materials). In the present analysis, VCRs were calculated in accordance with methodology used by the CDC to analyze BRFSS data.¹⁴ National- and state-level influenza and pneumococcal VCRs were calculated for each year in 2011–2019, whereas national- and state-level HZ and Td/Tdap VCRs were calculated for the years the questions pertaining to each vaccine were included as a core component. Because the HPV vaccine module was optional in all years, state-level VCRs were calculated for the states that included the optional module in 2018 or 2019 and in at least one other year, at least three years earlier.

For each vaccine with available BRFSS data, change in national- and state-level coverage was defined as the difference in vaccination coverage between the latest and earliest years divided by the number of years over which that difference was computed. Change in coverage was calculated using the years 2011 and 2019 for the influenza and pneumococcal vaccines; 2013 and 2019 for Td/Tdap; and 2014 and 2017 for HZ. In 2019, New Jersey was unable to collect enough data to meet the minimum requirements for inclusion in the aggregate BRFSS data; therefore, change in coverage was calculated using the year 2018 as the latest year, where applicable. The availability of data on the HPV vaccine varied by state; therefore, change in coverage for each state was calculated from the earliest year that the state included the optional HPV module to the latest year.

Statistical analysis

All vaccination coverage measures were summarized descriptively using point estimates and associated 95% confidence intervals. To produce coverage estimates representative of the overall US population, we used the analysis weights provided by BRFSS using the BRFSS ranking-weighting methodology.

TLR of interventions to improve vaccination coverage

Our TLR was conducted using guidelines from the Cochrane Collaboration and from the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA),^15,16 with the goal of identifying studies that describe adult vaccination interventions throughout the US. The search strategy encompassed searches of the MEDLINE and MEDLINE In-Process (via the PubMed platform) electronic medical databases. These electronic databases were searched for articles published in January 2015–June 2021; available in English; and that described adult vaccination interventions within the US. Full search terms are presented in Table S-3 (Supplementary Material). Screening was performed by a single reviewer in two phases using predefined inclusion and exclusion criteria (Table S-4, Supplementary Material): (1) screening of article titles and abstracts and (2) screening of full-text articles. Data were thematically categorized by intervention type.

Results

BRFSS analysis

Change in vaccination coverage

In general, most states demonstrated increases in VCRs across all five vaccine types, although decreases in vaccination coverage were observed in some states for the seasonal influenza vaccine (Louisiana, South Dakota, Tennessee, West Virginia), pneumococcal vaccine (Alabama, Arizona, Idaho), and HZ vaccine (Delaware, Idaho, Nevada, Oregon, Washington) (Figures 1–2 and Figure S-1, Supplementary Material). The change per year in vaccination coverage across states ranged from −0.53 to 1.51 for seasonal influenza vaccination among adults aged 18–64 years, −0.27 to 1.29 for pneumococcal vaccination among adults aged 18–64 years at increased risk of pneumococcal disease, 0.95 to 3.78 for Td/Tdap vaccination among adults aged 18–64 years, and −1.67 to 4.63 for HZ vaccination among adults 60–64 years. For HPV vaccination among adults aged 18–49 years, the change per year in vaccination coverage ranged from 0.45 to 3.02 among females and 0.47 to 2.31 among males. Trends in vaccination coverage per year for influenza, pneumococcal, Td/Tdap, HZ, and HPV vaccines are presented in Supplemental Figures S-2 to S-6.

Figure 1. Change per year in influenza and pneumococcal vaccination coverage among adults in the United States.

BRFSS = Behavioral Risk Factor Surveillance System.

Change in vaccination coverage was calculated for seasonal influenza among adults aged 18–64 years and for pneumococcal among adults aged 18–64 years at increased risk of pneumococcal disease, using the years 2011–2019.

Change in coverage per year was calculated as the difference in vaccination coverage between the latest and earliest years divided by the number of years over which that difference was computed. In 2019, New Jersey was unable to collect enough data to meet the minimum requirements for inclusion in the aggregate BRFSS data. Therefore, change in vaccination coverage for New Jersey was computed from 2011 to 2018.

Increased risk for pneumococcal disease was defined as a self-report of one or more of the following: being a current smoker; currently having asthma; ever having had diabetes, myocardial infarction, angina, or coronary heart disease; having chronic obstructive pulmonary disease, emphysema, or chronic bronchitis; or having any cancer except skin cancer.

Figure 2. Change per year in Td/Tdap and HZ vaccination coverage among adults in the United States.

BRFSS = Behavioral Risk Factor Surveillance System; HZ = herpes zoster; Td = tetanus and diphtheria; Tdap = tetanus, diphtheria, and acellular pertussis.

Change in vaccination coverage was calculated for Td/Tdap among adults aged 18–64 years using the years 2013–2019. Change in vaccination coverage was calculated for HZ among adults aged 60–64 years using the years 2014–2017.

Change in coverage per year was calculated as the difference in vaccination coverage between the latest and earliest years divided by the number of years over which that difference was computed. In 2019, New Jersey was unable to collect enough data to meet the minimum requirements for inclusion in the aggregate BRFSS data. Therefore, change in vaccination coverage for New Jersey was computed from 2011 to 2018.

TLR of interventions to improve vaccination coverage

Our search identified 302 records; 188 records were excluded after title and abstract screening and 82 records were excluded after full-text review (Figure 3), resulting in 32 publications^17–48 meeting the inclusion criteria. The included articles were heterogenous in terms of study design, interventions, and outcomes reported. Interventions reported in these 32 articles included multifaceted interventions (i.e., >1 intervention type utilized), patient or provider education, policy changes, vaccination reminder tools, and other interventions (i.e., free workplace vaccination, incentive and penalty-based strategies, and the adoption of Medicare Part D Tier-6). The interventions were distributed across the US (Figure 4), and settings included clinics, pharmacies, community-based settings, a hospital, and a long-term care facility.

Figure 3. PRISMA flowchart.

PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

^a The category “Other” includes duplicate references.

Figure 4. Regional distribution of interventions included in the TLR.

TLR = targeted literature review.

Multifaceted interventions

Seven articles reported on multifaceted interventions to improve vaccine uptake (Table 1), with approaches such as education, reminder tools, and workflow modification; all articles reported improvements in vaccination outcomes following the intervention.^{19,22,24,31,34,40,46} Loskutova et al.³¹ conducted a study with multicomponent interventions (n = 43 primary care physicians) to improve adult vaccination rates for influenza, pneumococcal, and HZ vaccines. The intervention included clinical decision support provider reminders, provider vaccination performance goals; vaccination education materials for patients, providers, and staff; monthly newsletters; and patient education visual aids, while the control group received clinical decision support provider reminders only. The authors observed significantly increased rates of influenza vaccination (p ≤ .01) and HZ vaccination (p ≤ .001) in both the reminders-only group and the multicomponent intervention group. However, pneumococcal vaccination rates did not significantly differ (p = .3) from those at baseline in the at-risk adult multicomponent intervention group.³⁶

Table 1. Key study characteristics for multifaceted interventions.

Article	Study Population	Intervention	Comparator	Setting	State	Vaccine Type	Significant Improvement in Vaccination Outcomes*
Loskutova et al.^Citation31	Adults (n = 43 primary care physicians)	Clinical decision support provider reminders, provider vaccination performance goals, provider materials for patient and staff vaccination education, provider education, monthly newsletter, visual aids for patient education	Clinical decision support provider reminders only	Health facility	North Carolina	Influenza, pneumococcal, HZ	Yes (p < .05)
Broderick et al.^Citation24	Adults with rheumatoid arthritis(n = 228)	Provider educational session on EMR-based alerts, EMR vaccination alerts for patients, and personalized encouragement	After the EMR-based alerts intervention	Health facility	New York	Influenza	Yes (p < .05)
Baker et al.^Citation22	Adults with rheumatoid arthritis(n = 1,255)	Electronic quality measurement, computerized point-of-care clinical decision support tools, individual performance feedback to physicians, and outreach to patients	Before intervention	Health facility	Illinois	Influenza, pneumococcal, HZ	Yes (p < .05)
Ofstead et al.^Citation34	Adults(n = 726)	Educational programming and materials, vaccination tracking mechanisms, and worksheets to assist with program implementation during 2014–2015 season	2013–2014 season	Long-term care facility	Illinois	Influenza	Yes (p < .02)
Wilson et al.^Citation46	Adults with a rheumatic condition(n = 1,298 during a 1-year baseline data collection period; 17–35 during the 13-week intervention)	Interprofessional approach incorporating provider education, reinforcement at the point of care, and workflow simplification	Before intervention	Health facility	Illinois	Pneumococcal	p values not reported
Nowalk et al.^Citation40	Adults (n = 25 primary care practices; 70,549 patients)	Provider education and 1:1 coaching for each practice	Wait-list control	Health facility	Pennsylvania and Texas	Tdap	Yes (p < .05)
Bluml et al.^Citation19	Adults (n = 1,080)	Practice model following the principle-centered approach	Before intervention	Pharmacy	Washington	Pneumococcal, HZ, HPV, Tdap	Yes (p < .05)

EMR = electric medical record; HPV = human papilloma virus; HZ = herpes zoster; Tdap = tetanus, diphtheria, and acellular pertussis.

*Indicates a significant increase, as defined by the reported article, in at least one vaccination outcome for at least one vaccine type. The corresponding p value thresholds for statistical significance are indicated in parentheses. Vaccination outcomes include vaccination rates, number of vaccines administered, time to vaccination, participant readiness to receive vaccination, percent of study participants up to date with vaccination, and others.

In a study by Broderick et al.²⁴ (n = 228), an intervention comprised of a provider educational session on electronic medical record (EMR)–based alerts, EMR vaccination alerts for patients, and personalized encouragement was significantly associated with less delay in time to vaccinate (p = .038). Baker et al.²² (n = 1,255) implemented a multicomponent system-level intervention that consisted of electronic quality measurement, computerized point-of-care clinical decision support tools, individual performance feedback to physicians, and patient outreach in adults with rheumatoid arthritis. Over the 1-year intervention period, the authors found no improvement in influenza vaccination rates; however, the authors reported significant increases in pneumococcal (p = .002) and HZ vaccination coverage (p = .01).²⁶ Ofstead et al.³⁴ found significantly increased influenza vaccination rates (p < .01) among nursing care staff and their family members following a multicomponent intervention that consisted of educational programming and materials, vaccination tracking, and worksheets to assist with program implementation.

Wilson et al.⁴⁶ (n = 1,298 during a 1-year baseline data collection period; n = 17–35 during a 13-week intervention) implemented an interprofessional approach intervention that incorporated provider education, education reinforcement at the point of care (i.e., ACIP guideline infographics displayed in clinic rooms), and workflow simplification; the authors observed improvements in pneumococcal vaccination rates (3% at baseline; 23% during the final 4 weeks of the 13-week intervention), although p values were not reported. Nowalk et al.⁴⁰ (n = 25 primary care practices; 70,549 patients) found that a provider education and 1:1 coaching intervention significantly increased Tdap vaccination rates compared with a control group that did not receive this intervention (p < .001). Lastly, Bluml et al.¹⁹ evaluated an innovative practice model intervention in community pharmacies following the principle-centered approach; briefly, when patients presented for influenza vaccinations, this approach consisted of pharmacists using IIS data to identify unmet vaccination needs for eight additional vaccine types and provide patient education. Of 1,080 people who had an influenza vaccine, unmet vaccination needs identified using the IIS included unmet needs for the pneumococcal conjugate vaccine (PCV; n = 409); pneumococcal polysaccharide vaccine (PPSV; n = 14); HPV vaccine (n = 16); and Tdap vaccine (n = 483). Following the intervention, the authors observed a significant difference in up-to-date PCV and Tdap vaccination rates (p < .001); however, there were no significant differences in PPSV or HPV vaccination rates. Overall, all the multifaceted interventions reported in this TLR included an education component, reminders, and/or reinforcement at the point of care. Collectively, these studies suggest that vaccine education and vaccination reminders, alone or in combination, may be an effective means to improve adult vaccination rates.

Patient or provider education

Eight articles reported on interventions involving patient or provider education (Table 2).^{17,20,25,26,32,35,39,47} A majority reported improved vaccination rates following the intervention,^{20,26,32,35,39} and those that did not observe increased vaccination rates reported higher participant readiness to receive vaccination.^17,25 To improve vaccination rates in a community pharmacy setting, Brackett et al.¹⁷ assessed the use of motivational interviewing–based interventions (n = 19 motivational interview encounters), in which two pharmacists conducted collaborative, person-centered interviews with eligible patients to elicit intrinsic motivation to receive vaccination. Vaccination rates did not differ between the motivational interviewing intervention and control groups for the pneumococcal (n = 5), HZ (n = 6), or hepatitis B (n = 13) vaccines (p values not reported); however, based on questions assessing participant readiness to receive vaccination (5-point Likert scale), there was a significant improvement in readiness to receive both vaccinations (hepatitis B: p = .001; pneumococcal: p = .033). In a study utilizing a patient-centered, educational iBook-based app explaining the benefits of antenatal influenza and pertussis vaccination, Chamberlain et al.²⁵ (n = 300) found higher vaccination rates in the intervention group; however, these increases were not significant (influenza: p = .38; Tdap: p = .27). An integrated approach by Clark et al.⁴⁷ (n = 2,505) using a 2-phased intervention consisting of staff education coupled with audit, feedback, incentive, and patient education to increase pneumococcal vaccine demand significantly increased pneumococcal vaccination rates compared with the control group (p < .0001).

Table 2. Key study characteristics for patient or provider education interventions.

Article	Study Population	Intervention	Comparator	Setting	State	Vaccine Type	Significant Improvement in Vaccination Outcomes*
Brackett et al.^Citation17	Adults (n = 19)	Motivational interviewing	No motivational interviewing	Pharmacy	Georgia	Pneumococcal, HZ, hepatitis A/B, Tdap	Yes (p < .05)
Chamberlain et al.^Citation25	Pregnant adults (Influenza vaccine: n = 300; Tdap vaccine: n = 291)	iBook-based app explaining the benefits of antenatal influenza and pertussis vaccination	No educational materials	Health facility	Georgia	Influenza, Tdap	No
Clark et al.^Citation47	Adults eligible for pneumococcal vaccination (n = 2,505 patients in the intervention)	Phase 1: staff education coupled with audit, feedback, and incentives; Phase 2: increased patient demand through education	Preintervention	Health facility	Virginia	Pneumococcal	Yes (p < .0001)
Healy et al.^Citation39	Pregnant women (n = 6,577)	Email immunization information and faculty meetings (ACOG toolkit)	Before ACOG toolkit implementation	Health facility	Texas	Tdap	p values not reported
Ciemins et al.^Citation26	Adults (n = 508 primary care providers)	Learning collaborative approach with in-person meetings, monthly webinars, and site visits	No learning collaborative matches	Health facility	Multistate	Influenza, pneumococcal	Yes (p < .05)
Olanipekun et al.^Citation35	Adults with a diagnosis of heart failure, unspecified, or heart failure-associated conditions (n = 281)	Vaccination information and physician recommendation	Did not receive information or recommendations	Health facility	Georgia	Influenza	Yes (p < .05)
Lu et al.^Citation32	Adults (n = 4,305)	Doctor visit where the patient: (i) received vaccination recommendation, and offered vaccination, (ii) received recommendation, not offered vaccination, (iii) did not receive recommendation, not offered vaccination	Did not visit doctor or healthcare professional	Healthcare provider	All states	Influenza	Yes (p < .05)
Cataldi et al.^Citation20	Females aged 9–26 years (98 clinics included; n = NR)	HPV fact sheets, parent education website, pictures of diseases caused by HPV, decision aid or HPV vaccination and HPV communication training for healthcare professionals	None	Health facility	Colorado	HPV	p values not reported

ACOG = American College of Obstetricians and Gynecologists; HPV = human papilloma virus; HZ = herpes zoster; NR = not reported; Tdap = tetanus, diphtheria, and acellular pertussis.

*Indicates a significant increase, as defined by the reported article, in at least one vaccination outcome for at least one vaccine type. The corresponding p value thresholds for statistical significance are indicated in parentheses. Vaccination outcomes include vaccination rates, number of vaccines administered, time to vaccination, participant readiness to receive vaccination, percent of study participants up to date with vaccination, and others.

Healy et al.³⁹ (n = 6,577) observed an increase in antenatal Tdap vaccination from 36% to 61% after the implementation of an intervention comprised of physician education about pertussis illness in infants, Tdap vaccination recommendations, and the American College of Obstetricians and Gynecologists (ACOG) toolkit (resources to support healthcare professional communication with pregnant women about the importance of Tdap vaccination); Black women were significantly less likely (p < .001) to receive a vaccination compared to other race/ethnicity groups. Ciemins et al.²⁶ (n = 508 primary care providers in the intervention arm) found that a learning collaborative approach intervention, consisting of in-person meetings, monthly webinars, and site visits, resulted in significant improvements in influenza and pneumococcal vaccination rates among healthcare organizations participating in the intervention (p < .01) compared with matched control healthcare organizations.

In a patient education–based intervention in patients with heart failure or associated conditions, Olanipekun et al.³⁵ found that patients who received influenza vaccination information and recommendations from their physician were significantly more likely to be vaccinated against influenza than those who did not (p < .05). Furthermore, patients with heart failure who received information and recommendations from a cardiologist had the greatest odds of being vaccinated (adjusted OR, 8.2 [95% CI, 4.8–16.4]; p < .05). Lu et al.³² observed that influenza VCRs were significantly higher (p < .05) in adults who received both a provider recommendation and were offered vaccination during their visit (VCR, 66.6%; n = 1,080) compared with adults who received only a recommendation without an offer (VCR, 48.4%; n = 414) or those who did not receive a recommendation or an offer (VCR, 32.0%; n = 1,450).

Lastly, a communication aids and training intervention in a modeling study by Cataldi et al.²⁰ (n = 98 clinics) led to an estimated additional 5,218 vaccinated patients and 43 cervical cancer cases prevented among the 10 clinics with the highest predicted number of cervical cancer cases prevented. However, the predicted number of cases of cervical cancer prevented by the intervention may be underestimated in this study due to time horizon differences and the analysis not accounting for people aging in (or out) of the age range of eligibility for routine HPV vaccination. In summary, education-based interventions were generally associated with improved vaccination outcomes; however, the wide variety of intervention methods and types of outcomes reported made comparisons across specific intervention types infeasible.

Policy change

Five articles reported on vaccination-related policy interventions (Table 3),^{21,27,33,36,44} and three of these studies reported significant improvements in VCRs.^21,27,44 Tan et al.²¹ (vaccine-eligible adult sample size not reported) implemented a standing order protocol (SOP) intervention at five health facility sites, which authorized nonphysician healthcare professionals to assess unmet vaccination needs and administer vaccines. Following SOP implementation, the authors observed increases in the number of influenza vaccinations given at all sites; however, this only resulted in a statistically significant increase in the influenza vaccination rate at two sites (p < .01). Furthermore, one site experienced an increased number of influenza vaccinations but a significant decrease in the vaccination rate (p < .01); however, the authors noted that this decrease was due to the site inheriting patients from another closed practice. Study authors also reported significant increases in pneumococcal vaccination coverage (p < .01); significantly higher HPV VCR in male participants (from 12% to 24%; p < .01); a significant increase in Tdap vaccination rates among patients aged 19–64 years (p < .01) in 4 of the 5 health facility sites; greater HZ vaccination rates, although not statistically significant; and no significant differences in hepatitis B vaccination rates in adults aged ≥19 years.²⁵

Table 3. Key study characteristics for policy change interventions.

Article	Study Population	Intervention	Comparator	Setting	State	Vaccine Type	Significant Improvement in Vaccination Outcomes*
Tan et al.^Citation21	Adults (number of adult patients per health facility site during the baseline year ranged from 2,000–7,500+; n = NR for the eligible study population)	SOP implementation	Prior to SOP implementation	Health facility	All states	Influenza, pneumococcal, HPV, HZ, Tdap, hepatitis B	Yes (p < .01)
Patel et al.^Citation36	Adults (n = 1,002,994 adults aged 18–64 years)	Pharmacy-based immunization services	None	Pharmacy	All states	Influenza, pneumococcal	No
Drozd et al.^Citation27	Adults (n = NR)	Policies allowing pharmacists to administer seasonal influenza vaccinations	After policy changes	Pharmacy	All states	Influenza	Yes (p < .001)
McConeghy and Wing^Citation33	Adults (n = NR)	Pharmacy-based immunization statutes: (i) physician prescription requirements, (ii) state-wide protocol agreements, (iii) independent pharmacist authority, or (iv) pharmacist-physician collaborative practice agreements	Before statutes	Pharmacy	All states	Influenza	No
Tak et al.^Citation44	Adults aged ≥60 years (n = NR)	Pharmacist authorization to give HZ vaccine without prescription order	HZ vaccine requires prescription order	Pharmacy	All states	HZ	Yes (p < .05)

HPV = human papilloma virus; HZ = herpes zoster; NR = not reported; SOP = standing order protocol; Tdap = tetanus, diphtheria, and acellular pertussis.

*Indicates a significant increase, as defined by the reported article, in at least one vaccination outcome for at least one vaccine type. The corresponding p value thresholds for statistical significance are indicated in parentheses. Vaccination outcomes include vaccination rates, number of vaccines administered, time to vaccination, participant readiness to receive vaccination, percent of study participants up to date with vaccination, and others.

Drozd et al.²⁷ (sample size not reported) studied the impact of policy changes allowing pharmacist-administered seasonal influenza vaccinations and found that these policy changes were associated with significant increases in influenza vaccination rates ≥6 years after the policy change (p ≤.001). Tak et al.⁴⁴ reported significantly higher HZ vaccination rates in states where pharmacists were authorized to give HZ vaccinations without a prescription order compared with states with prescription order requirements (p = .0022). Conversely, Patel et al.³⁶ (n = 1,002,994 adults aged 18–64 years) evaluated pharmacy-based vaccination services using data on the availability of the live attenuated influenza vaccine annually for 8,466 pharmacies during 2006–2010 as a proxy for pharmacy-based vaccination service availability. The authors found no significant association between influenza or pneumococcal vaccination rates and the availability of pharmacy-based vaccination services. McConeghy and Wing³³ (sample size not reported) did not observe a significant change in adult influenza vaccination rates following the adoption of flexible pharmacy-based immunization statutes by states (OR, 0.9; 95% CI, −0.3 to 2.2). Overall, policy interventions that provided pharmacists, nurses, and other healthcare personnel authorization to administer vaccines, including the implementation of SOPs, where allowed by state law, generally resulted in increased vaccination rates. These findings suggest that policies that expand vaccination authorization to non-clinician care team members may positively impact adult vaccination rates.

Vaccination reminder tools

Nine articles reported on interventions using vaccination prompts and reminder tools (Table 4).^{18,23,29,30,38,41,43,45,48} In a study by Klassing et al.³⁰ (n = 210), the intervention groups received either a phone call or a letter referencing the up-to-date CDC immunization schedule and guidelines after picking up a prescription at a pharmacy. The authors reported that the influenza vaccination rate was significantly greater (p = .02) in the letter recipient group, but found no significant differences (p = .76) in pneumococcal vaccination rates between the control and intervention groups. Benedict et al.²³ (n = 15,115) found that adults receiving influenza vaccination reminders were 1.15 (95% CI, 1.06–1.25) times more likely to be vaccinated than adults not receiving reminders. Szilagyi et al.³⁸ (n = 164,205) found that vaccination reminders resulted in a small but significant increase in influenza vaccination rates between the three intervention groups combined (participants received 1, 2, or 3 portal messages) compared with the control group with no message for adults aged 18–64 years (p = .001). Interestingly, the authors did not observe any dose-dependent increase in vaccination rates in the participants that received two or three messages compared with those that received one message.

Table 4. Key study characteristics for vaccination reminder tool interventions.

Article	Study Population	Intervention	Comparator	Setting	State	Vaccine Type	Significant Improvement in Vaccination Outcomes*
Klassing et al.^Citation30	Adults with possible diagnosis of asthma and/or COPD (n = 210)	In-store advertising and flyers advertising on-site immunizations when picking up prescriptions, received either a phone call or a letter referencing the 2014 CDC immunization schedule and guidelines	Adults who did not receive a phone call or letter	Pharmacy	Kansas	Influenza, pneumococcal	No
Szilagyi et al.^Citation38	Adults (n = 164,205)	Patients received 1, 2, or 3 portal messages that included (i) information on influenza and the influenza vaccine; (ii) recommendation to receive an influenza vaccine; (iii) a website link to input influenza vaccinations received elsewhere into the UCLA Health System record; and (iv) a link to a webpage containing information about influenza vaccine and video testimonials	Patients who did not receive portal messages	Health facility	California	Influenza	Yes (p < .017 for primary analyses; p < .05 for all other analyses)
Grivas et al.^Citation29	Adults (n = NR)	An EMR/ordering notification system, reminders placed in physical charts and workstations, stickers affixed to patient’s medication list, and patient-focused reminders	Historical controls (2005–2011 Influenza seasons)	Health facility	Michigan	Influenza	Yes (p < .0001)
McAdam-Marx et al.^Citation48	Adults eligible for pneumococcal vaccination (n = 18,851 aged 19–64 years; n = 12,057 aged ≥65 years)	Best practice alerts with and without workflow redesign	Health maintenance notifications	Mobile outreach	Ohio	Pneumococcal	Yes (p < .05)
Erlandson et al.^Citation41	Adults with HIV (n = 1,478)	Weekly e-mail to provider with eligible patients, patient alerts, and enhanced prompts for vaccine (Part 1); an order for HZV placed on the patient’s chart before their clinic appointment (Part 2)	Preintervention	Health facility	Multistate and Canada	HZ	p Values not reported
Sheth et al.^Citation43	Adults with rheumatoid arthritis (n = 1,823)	Electronic identification of vaccine eligibility and electronic best practice alert	Before intervention	Health facility	Pennsylvania	HZ	Yes (p < .0001)
Hechter et al.^Citation18	Adults with diabetes (n = 116,217)	Electronic reminders and alerts	No electronic reminders or alerts	Health facility	California	Hepatitis B	Yes (standardized difference < 0.1)
Benedict et al.^Citation23	Adults (n = 15,115)	Reminders (included mail, email, phone call, text message)	No reminders	Mobile outreach	All states	Influenza	Yes (p < .05)
Burns et al.^Citation45	Adults with HIV (n = 99)	Virtual clinic, mailed letters with vaccine recommendations and information	None	Mobile outreach	Ohio	Pneumococcal	p Values not reported

CDC = Centers for Disease Control and Prevention; COPD = chronic obstructive pulmonary disease; EMR = electric medical record; HZ = herpes zoster; HZV = herpes zoster virus; NR = not reported; Tdap = tetanus, diphtheria, and acellular pertussis; UCLA = University of California, Los Angeles.

*Indicates a significant increase, as defined by the reported article, in at least one vaccination outcome for at least one vaccine type. The corresponding p value or standardized difference thresholds for statistical significance are indicated in parentheses. Vaccination outcomes include vaccination rates, number of vaccines administered, time to vaccination, participant readiness to receive vaccination, percent of study participants up to date with vaccination, and others.

Two studies used a combination of patient and provider reminder tools, and both reported improved vaccination coverage. Grivas et al.²⁹ implemented an intervention during the 2011–2012 and 2012–2013 influenza seasons using an EMR notification system, laminated reminders placed in physical charts and workstations, stickers affixed to patients’ medication reconciliation list, and patient-focused reminders (e.g., educational information on the benefits of influenza vaccination) (2012–2013 season only). The intervention resulted in relative influenza vaccination rate increases of 37.6% (2011–2012 season) and 56.1% (2012–2013 season) compared with historical controls (2005–2011 seasons), although p values and sample sizes were not reported. Burns et al.⁴⁵ (n = 99) used an EMR-based intervention to assess patients’ pneumococcal vaccination status and notify patients and physicians of the recommended pneumococcal vaccine (i.e., PCV, PPSV). The authors found that the EMR-based intervention improved pneumococcal vaccine coverage among HIV-positive veterans at the Veterans Affairs facility that conducted the intervention (n = 22/81 eligible veterans) compared with another Veterans Affairs medical center (n = 10/84 eligible veterans; p < .05).

Several interventions focused on vaccination reminder tools for health-care providers, which were also successful in improving vaccination coverage. McAdam-Marx et al.⁴⁸ (n = 18,851) found significantly improved vaccination rates for at-risk adults aged 19–64 years (p < .001) following the implementation of a pneumococcal best practice alerts tool that reflects current guidelines, implemented with and without workflow redesign to identify unmet vaccination needs prior to a patient’s visit and document vaccination history during the visit. Erlandson et al.⁴¹ (n = 1,478) observed an 8.3% increase in HZ vaccination rates following the implementation of weekly emails to providers with eligible patients, patient alerts during their clinic visit, and enhanced prompts for vaccination. The authors found a further increase of 17.8% for HZ vaccination following orders being placed in the patient’s chart before their clinic appointment, although p values were not reported in this study to determine statistical significance.⁴⁶ Sheth et al.⁴³ reported a significant increase (p < .0001) in HZ vaccination rates after implementing electronic identification of vaccine eligibility and electronic best practice alerts. Lastly, Hechter et al.¹⁸ (n = 116,217) found that the vaccination coverage rate of the 3-dose series of the hepatitis B vaccine increased significantly (p < .0001) at an intervention site that utilized electronic reminders and alerts compared with the control site, where no alerts were sent. In summary, the incorporation of vaccination prompts and reminder tools generally improved vaccination outcomes, although some studies lacked information regarding statistical significance.

Other interventions

Three articles reported on other types of interventions: free workplace vaccination, incentive and penalty-based strategies, and the adoption of Medicare Part D Tier-6 (Table 5).^28,37,42 Gany et al.²⁸ (n = 53) reported 44% of unvaccinated drivers accepted an influenza vaccination (p value not reported) following the implementation of free workplace vaccination for taxi drivers, compared with a baseline of 17% of drivers being vaccinated. Podczervinski et al.³⁷ compared vaccination rates at baseline (n = 1,446) with vaccination rates following incentive- or penalty-based (n = 1,586 and n = 1,641, respectively) strategies to improve influenza vaccination coverage. Following the incentive-based strategy, a $25 gift card was awarded to all staff if 95% of employees received vaccination, whereas the penalty-based strategy required employees who declined vaccination to complete an online education module, undergo one-on-one counseling, and sign an attestation statement. The authors found that both strategies resulted in improved vaccination rates (p ≤.0001); furthermore, the penalty-based strategy resulted in greater improvement compared with the incentive-based strategy (p < .0001).³⁷ Lastly, Hechter et al.⁴² (sample size not reported) found that the annual HZ vaccination rate was only slightly higher after the adoption of Medicare Part D Tier-6, with no patient co-pay for HZ vaccination. In addition, the Medicare Part D annual HZ vaccination rate was not statistically different (p > .05) than the annual HZ vaccination rate among patients with commercial health plans.

Table 5. Key study characteristics for other interventions.

Article	Study Population	Intervention	Comparator	Setting	State	Vaccine Type	Significant Improvement in Vaccination Outcomes*
Basurto-Dávila et al.^Citation49	Children and adults (n = NR)	School-based vaccination clinics (Maine)	Before school-based vaccination clinics	Mobile outreach	Multistate	Influenza	No
Gany et al.^Citation28	Taxi drivers (n = 53)	A pharmacy offered free influenza vaccination at a taxi garage with a physician present	None	Community	New York	Influenza	Yes (p < .05)
Podczervinski et al.^Citation37	Adult employees (2010: n = 1,446; 2011: n = 1,586; 2012: n = 1,641)	Incentive-based strategy (2011): If 95% of all employees received a vaccination, then all staff would receive a $25 gift card. The incentive was advertised across the center with weekly vaccination rates posted on the intranet and email. Penalty-based strategy (2012): Those who opted to decline vaccination had to complete a 30 min online education module, have 1:1 counseling session and sign an attestation	Before interventions (2010)	Health facility	Washington	Influenza	Yes (p < .05)
Hechter et al.^Citation42	Adults 60 years and older (n = NR)	Adoption of Medicare Part D Tier 6 with 0 patient co-pay for HZ vaccination	No policy adoption	Health facility	California	HZ	No

HZ = herpes zoster; NR = not reported.

*Indicates a significant increase, as defined by the reported article, in at least one vaccination outcome for at least one vaccine type. The corresponding p value thresholds for statistical significance are indicated in parentheses. Vaccination outcomes include vaccination rates, number of vaccines administered, time to vaccination, participant readiness to receive vaccination, percent of study participants up to date with vaccination, and others.

Discussion

The findings from this study provide a deeper understanding of changes in adult vaccination coverage in the US as well as interventions aimed at improving adult VCRs. These findings also uncover the complex nature of adult vaccination and the importance of multiprong strategies to improve adult vaccine uptake. Despite the benefits of vaccination, adult VCRs^1–3 are considerably lower than national adult vaccination goals.⁵

While optimal adult VCRs in the US remain unmet, our analysis of BRFSS data from 2011-2019 found generally modest and positive changes in adult VCRs for most US states, with a few states demonstrating negative changes in VCRs; however, these trends may have changed due to the COVID-19 pandemic with further negative changes in vaccine uptake.⁵⁰ Additionally, in agreement with the findings of previous studies,^2,8 we observed high interstate variability in VCRs, which underscores the importance of understanding state-specific needs and barriers as well as state- and local-level interventions, policies, and programs that lead to high vaccination rates. The growing and increasingly complex CDC adult vaccination schedule in the US also highlights the importance of improving awareness of the value of adult vaccination.

Interventions to improve adult VCRs in the US mainly focused on patient or provider education, policy change, and vaccination reminder tools. The TLR findings indicate that vaccine education, reminders or reinforcement at the point of care, and authorization of non-clinician members of the healthcare team to vaccinate are all effective strategies to improve adult vaccination rates. Furthermore, policy change interventions that expand vaccination access by expanding vaccination authorization to non-clinician members of the healthcare team may also improve adult VCRs; approaches combining multiple strategies may further enhance effectiveness. Notably, these approaches may be particularly beneficial to implement in the states identified in our BRFSS analysis that did not demonstrate increases in adult VCRs over time.

While the heterogeneity across studies included in our TLR complicated the synthesis and interpretation of findings, we were able to extract key factors that impacted intervention success. In the multifaceted intervention implemented by Loskutova et al.³¹ influenza vaccination rates significantly increased in both the multicomponent intervention group and the control group comprising clinical decision support provider reminders, which suggests that reminders were effective regardless of whether provided alone or in combination with other interventions. This underscores the importance of enhanced EMR capabilities in supporting vaccination for both provider and patient reminders. In the policy change intervention assessed by Tan et al.²¹ vaccination rates generally increased modestly following the implementation of SOPs, but study sites that used SOPs as a foundation for additional interventions found greater success. Additionally, all the multifaceted interventions presented here reported improvements in at least one vaccination outcome, and many of these studies combined patient or provider education with reminders or reinforcement at the point of care. These patient-engaging strategies exemplify a personalized approach to address patient queries directly, which offers a personalized advantage over other strategies such as policy changes and vaccination reminder tools.

Lack of knowledge and awareness about adult vaccination among both healthcare providers and patients form barriers to improving adult vaccination rates in the US.^51,52 These identified barriers align with our TLR findings, where the interventions that significantly improved vaccination rates often included an education or reminder component. Indeed, the simplest solution is often the most effective, and our TLR findings suggest that ensuring timely and accurate reminders, along with vaccination information that is easily understood by patients and providers, will drive improvements in adult VCRs. These findings are corroborated by an evaluation of evidence-based strategies to increase adult vaccination rates conducted by the Community Preventive Services Task Force in 2015.⁵³ Indeed, the Community Preventive Services Task Force recommends a combined approach to healthcare system–based interventions that aims to increase demand for vaccinations (e.g., patient education or reminders) and targets enhanced vaccine access, healthcare providers, or healthcare systems (e.g., provider reminders or SOPs). Similarly, the US Department of Health and Human Services aims to improve vaccine knowledge, access, and use through strategies such as increasing education regarding the benefits of vaccination and strengthening data infrastructure (e.g., IIS).⁵⁴

Adult vaccination is a public health priority that diverse stakeholders will need to address through interventions that include continuing education and community engagement;⁵⁵ however, additional research should focus on adapting these interventions to optimally improve rates for each type of vaccination across diverse patient populations and practice types. In the intervention implemented by Szilagyi et al.³⁸ patient portal reminders significantly increased influenza vaccination in patients aged 18–64 years but did not have a significant impact on vaccination rates in patients aged 0.5–17 years and ≥65 years. Furthermore, the authors observed racial and ethnic differences in the intervention outcomes, reporting that portal reminders were effective only in patients who were White and non-Hispanic. These findings suggest that tailored approaches for specific populations should be explored to address the nuanced barriers they face and optimize VCRs.

The impact of intervention setting on vaccination rates is also an important factor to consider following the COVID-19 pandemic, as complimentary settings were utilized for COVID-19 vaccination. Notably, most interventions identified in our TLR were implemented in healthcare facilities, with fewer interventions in community settings. Our TLR findings indicate an association between pharmacist authorization to vaccinate and higher vaccination rates,^27,44 and similarly, systematic reviews have found that pharmacy-based vaccination programs may improve access and increase VCRs, which may be attributable to their convenient locations within communities.^27,44,56,57 Future research may provide further insight regarding the feasibility and value of expanding convenient vaccination settings, such as pharmacies, for broader adult vaccination. Additionally, future research should also consider the role that employers may play in supporting adult vaccination.

The findings of this study are subject to several limitations. First, BRFSS data are self-reported and therefore subject to recall bias. The BRFSS does not collect information about vaccine eligibility; in the present study, we assumed that everyone who reported receiving a vaccine was eligible to receive it. Additionally, we used BRFSS data collected through 2019, which do not reflect any potential changes to adult HPV vaccination rates following the 2019 ACIP update that recommended catch-up vaccination for all persons through age 26 years and shared clinical decision-making regarding HPV vaccination for adults aged 27–45 years.⁵⁸ Prior to this update, HPV catch-up vaccination was recommended for females through age 26 years and males through age 21 years. Furthermore, for risk-based pneumococcal vaccination recommendations, limited data on the number of adults classified as “at-risk” makes understanding the optimal coverage rate challenging. While the Healthy People 2020 objectives include adult vaccination goals for several diseases, including influenza, pneumococcal, and HZ,⁵ additional guidance would be beneficial for determining optimal coverage rates for adults for all recommended vaccines, by age and risk group (Table S-5, Supplemental Material).⁵⁹ Nevertheless, the present study provides valuable information regarding changes in adult VCRs over time, as well as lessons learned from intervention studies. Finally, in our TLR, the search strategy was restricted to studies published in 2015–2021; this restriction could have missed studies published before 2015 that may still provide useful data despite reporting on older information. Similarly, geographical restriction to the US may overlook insightful information about vaccination strategies of other countries, which may be applicable to the US, as well as limit the global generalizability of these TLR findings. Lastly, the COVID-19 pandemic may have impacted health seeking behaviors and opportunities as well as attitudes and beliefs toward vaccination, which could alter the degree to which some of these interventions were effective as well as potentially bring to light other interventions that may be effective.

Conclusion

Healthcare and public health professionals, policymakers, and other critical health system decision makers should consider the strategies identified in this literature review, especially in states that have not demonstrated increased VCRs over time, when implementing practice at the point of clinical care and when developing programs and policies to support the adult vaccination ecosystem and, ultimately, improve adult vaccination rates. This mixed methods study advocates the development of evidence-based policies and practices aimed to improve adult vaccination coverage rates.

Contributors

ALE, AB, and MP contributed to the conception and design of this study. AE, LH, DG, CS, JR, MG, and MP contributed to the acquisition, analysis, or interpretation of data for the work. All authors contributed substantially to drafting the work or revising it critically for important intellectual content. All authors provided final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Acknowledgments

The authors thank Brian Samsell, PhD, of RTI Health Solutions for medical writing assistance.

Disclosure statement

LH, DG, CS, JR, MG, and MP are full-time employees of RTI Health Solutions, an independent nonprofit research organization, which was retained by MSD to conduct the research which is the subject of this manuscript. Their compensation is unconnected to the studies on which they work. MG was a full-time employee of RTI Health Solutions at the time this study was conducted. ALE and AB are current employees of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA, who may own stock and/or stock options in Merck & Co., Inc., Rahway, NJ, USA.

Data availability statement

The data that support the findings of the BRFSS analysis are openly available in the BRFSS Annual Survey Data at https://www.cdc.gov/brfss/annual_data/annual_data.htm. All data generated or analyzed within the TLR are included in this article and its supplementary information files.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2024.2381283

Additional information

Funding

This work was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA (MSD). RTI Health Solutions, an independent nonprofit research organization, received funding under a research contract with MSD to conduct this study and provide publication support in the form of manuscript writing, styling, and submission. Authors from MSD were involved in the study as disclosed in the Contributors subsection. The content is the sole responsibility of the authors and does not necessarily represent the official views of MSD or RTI Health Solutions.

Notes

[a] ACIP, Advisory Committee on Immunization Practices; ACOG, American College of Obstetricians and Gynecologists; BRFSS, Behavioral Risk Factor Surveillance System; CDC, Centers for Disease Control and Prevention; CI, Confidence interval; COPD, chronic obstructive pulmonary disease; EMR, electronic medical record; HPV, human papillomavirus; HZV, herpes zoster virus; IIS, immunization information systems; NR, Not reported; OR, Odds ratio; PCV, pneumococcal conjugate vaccine; PPSV, pneumococcal polysaccharide vaccine; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; SOP, standing order protocol; Td, tetanus and diphtheria; Tdap, tetanus, diphtheria, and acellular pertussis; TLR, targeted literature review; UCLA, University of California, Los Angeles; US, United States; VCR, vaccination coverage rate.