Authors: Ahuva Averin, Jeffrey Vietri, Adriano Arguedas Mohs, Sarah J. Willis, Alexander Lonshteyn, Derek Weycker
Categories: Research Article, Pneumococcal vaccines, pneumonia, healthcare disparities, vaccination, immunization, Streptococcus pneumoniae
Source: AJPM Focus
Authors: Ahuva Averin, Jeffrey Vietri, Adriano Arguedas Mohs, Sarah J. Willis, Alexander Lonshteyn, Derek Weycker
Updated recommendations for adult pneumococcal vaccination in the U.S. (publication January 27, 2022) incorporated 2 new vaccines (15- and 20-valent pneumococcal conjugate vaccines), removed 13-valent pneumococcal conjugate vaccine, and called for pneumococcal conjugate vaccine use among immunocompetent adults aged 19–64 years with certain medical conditions. This study assessed uptake of recommendations and disparities in uptake across subgroups of adults.
A retrospective cohort design and data from Optum’s deidentified Clinformatics Data Mart Database were employed. Study population comprised all adults aged ≥65 years and adults aged 19–64 years with ≥1 chronic (at-risk) or immunocompromising (high-risk) condition. Vaccine uptake (including 23-valent pneumococcal polysaccharide vaccine) was estimated using the Kaplan–Meier method.
During 21-month follow-up period, 13.2% of adults (n=6.8 million) received pneumococcal vaccine, mostly 20-valent pneumococcal conjugate vaccine (9.6%). By age/risk conditions, 20-valent pneumococcal conjugate vaccine uptake was highest among adults aged 65–66 years (23.8%) and at-risk/high-risk adults aged 60–64 years (12.1%) and lowest among at-risk/high-risk adults aged 19–49 years (4.7%). By immunization history, 20-valent pneumococcal conjugate vaccine uptake was highest among adults with a history of 23-valent pneumococcal polysaccharide vaccine uptake only (15.1%) or 13-valent pneumococcal conjugate vaccine uptake only (10.6%) and lowest among those without prior pneumococcal vaccination (8.7%) or with a history of 13-valent pneumococcal conjugate vaccine + 23-valent pneumococcal polysaccharide vaccine uptake (7.9%).
Fewer than ∼1 in 7 U.S. adults received 20-valent pneumococcal conjugate vaccine in the first 21 months after the updated recommendations. Uptake was lower among at-risk/high-risk adults aged <60 years, adults aged ≥75 years, and adults without prior pneumococcal vaccination. Routine evaluation of vaccination status by providers and additional strategies to increase uptake of recommend vaccines are warranted.
Worldwide, lower respiratory tract illnesses, such as pneumonia, are a leading cause of death owing to communicable diseases.^1^ Pneumococcal disease—including invasive pneumococcal disease and nonbacteremic pneumonia caused by Streptococcus pneumoniae—continues to be a source of significant morbidity, mortality, and economic costs among adults, especially older adults and those with underlying medical conditions.2, 3, 4, 5, 6
In the U.S., recommendations for adult pneumococcal vaccination have been updated several times by the Advisory Committee on Immunization Practices (ACIP) of the Centers for Disease Control and Prevention (CDC). In 1984, the 23-valent pneumococcal polysaccharide vaccine (PPSV23) was first recommended for adults aged ≥65 years and those with chronic illnesses.^7^ In 2012, the 13-valent pneumococcal conjugate vaccine (PCV13) was recommended for immunocompromised adults, and in 2014, the recommendation was extended to all adults aged ≥65 years.^8^^,^^9^ In 2019, the recommendation for PCV13 among immunocompetent adults was revised to shared clinical decision making.^10^ Throughout these updates, the recommendation for PPSV23 persisted.
More recently, ACIP recommended that adults aged ≥65 years who have not previously received a pneumococcal conjugate vaccine (PCV) or whose previous vaccination history is unknown should receive 1 dose of a PCV, either 20-valent PCV (PCV20) or 15-valent PCV (PCV15).^11^ When PCV15 is used, it should be followed by a dose of PPSV23. Adults aged <65 years who have underlying medical conditions that increase their risk of pneumococcal disease were likewise recommended to receive either PCV20 alone or a sequence of PCV15 and PPSV23. These recommendations were published in the Morbidity and Mortality Weekly Report on January 27, 2022.^11^
Those recommendations targeted a much wider population for immunization with PCVs.^12^ Previous recommendations did not include PCV13 for immunocompetent adults aged <65 years, though adults with selected medical conditions such as chronic lung disease, chronic heart failure, or diabetes have long been recommended to receive PPSV23.^7^^,^^8^^,^^10^^,^^13^ The role of PPSV23 in adult pneumococcal vaccination was further eroded in an update published in September 2023, which recommended the use of either a single dose of PCV20 or ≥1 dose of PPSV23 for those who have initiated their vaccine series with PCV13 but have not received all recommended doses of PPSV23 and shared clinical decision making for use of a supplemental PCV20 dose among adults aged ≥65 years who have completed their recommended vaccine series with both PCV13 and PPSV23.^14^
Published studies have reported that adult pneumococcal vaccine uptake was low among target populations after issuance of previous updates to recommendations and that adherence to recommendations varied considerably across subgroups of adults defined on age and risk conditions.15, 16, 17, 18 Accordingly, a new study was undertaken to better understand the uptake of pneumococcal vaccines among U.S. adults since posting of the 2022 U.S. CDC-ACIP vaccine recommendations and disparities in uptake across subgroups of adults.
A retrospective observational cohort design and data from a large integrated U.S. healthcare claims repository—the Optum deidentified Clinformatics Data Mart Database (CDM)—spanning January 2015 through October 2023 were employed. The Optum CDM comprises medical (i.e., facility and professional service) claims, outpatient pharmacy claims, and enrollment information from a large U.S. private health insurer covering >15.0 million geographically diverse members annually, including enrollees, their spouses, and their dependents. Adults aged ≥65 years who elected to enroll in a Medicare Advantage Plan—and thus receive their healthcare coverage through a private health plan—are also included in the data source population.
Data available from each medical claim include dates and places of service, diagnoses, procedures performed/services rendered, and quantity of services (professional-service claims only). Data available for each outpatient pharmacy claim include the drug dispensed, dispensing date, quantity dispensed, and number of days supplied. Enrollment information includes dates of healthcare coverage and selected demographic characteristics. All healthcare claims are verified, adjudicated, adjusted, and deidentified by Optum prior to inclusion in the database. Variables for race and SES were derived by the data vendor using a separate marketing database matched to health plan members as well as algorithms that employ enhanced geocoding and predictive modeling combined with U.S. Census data; as a result, some data on race and SES in the Optum CDM may be imputed for some enrollees and missing/unknown for others.^19^
The study population comprised all adults aged ≥65 years and adults aged 19–64 years with ≥1 chronic medical (i.e., at-risk) condition or immunocompromising (i.e., high-risk) condition as of January 27, 2022. Adults without health plan enrollment on January 27, 2022 (i.e., index date) and during the 1-year preindex period and those with evidence of PCV20 or PCV15 prior to January 27, 2022 were excluded from the study population.
Age was calculated by subtracting birth year from year of index date (i.e., 2022). At-risk and high-risk conditions were identified on the basis of corresponding diagnosis, procedure, and drug codes recorded on medical claims and outpatient pharmacy claims during the 1-year preindex period (available from authors upon request). At-risk conditions included alcoholism, asthma, diabetes, chronic heart disease, chronic liver disease, chronic lung disease, and tobacco use; high-risk conditions included cochlear implant, congenital immunodeficiency, HIV, malignancy, organ transplant, renal failure (chronic), nephrotic syndrome, and white blood cell diseases.^14^ Adults with evidence of both at-risk and high-risk conditions were classified in the latter subgroup.
Use of pneumococcal vaccine (PCV13, PCV15, PCV20, and PPSV23) was ascertained from January 27, 2022 through October 31, 2023 and was identified through medical (ambulatory) claims with corresponding Current Procedural Terminology codes (PCV13: 90670; PCV15: 90671; PCV20: 90677; and PPSV23: 90732) and outpatient pharmacy claims with corresponding National Drug Codes (Appendix, available online).
Baseline characteristics of the study population included age, sex (male, female), race, geographic region of residence, risk conditions (low risk, at risk, high risk), pneumococcal vaccination history, and household income. Age and risk conditions were defined as described earlier (Study Population section); low-risk included adults without at-risk/high-risk conditions. Geographic region and household income were defined on the basis of the most proximate data relative to January 27, 2022. Pneumococcal vaccination history was defined on the basis of all available codes recorded on medical (ambulatory) claims and outpatient pharmacy claims between January 1, 2015, and January 26, 2022.
Observed data only were used in measuring study variables. Demographic information was available for nearly all enrollees. All other variables were defined on the basis of the presence of specific data (e.g., diagnosis, procedure, drug codes) in the study database; the absence of such data was assumed to indicate the absence of the characteristic/event captured by the variable.
Baseline characteristics of the study population, overall as well as by age, were summarized descriptively. Incidence proportions (and corresponding 95% CIs) for use of pneumococcal vaccine—overall and by type—were estimated using the Kaplan–Meier method, considering all adults qualifying for analyses. Use of PCV20 among subgroups defined on age and risk conditions, pneumococcal vaccination history, race, and household income was similarly estimated. Age- and risk-specific subgroup definitions were initially aligned with ACIP recommendations (i.e., at-risk/high-risk adults aged 19–64 years, all adults aged ≥65 years) and were subsequently refined on the basis of expected differences in vaccine uptake (i.e., at-risk/high-risk adults aged 19–49 years, 50–59 years, and 60–64 years; all adults aged 65–66 years, 67–74 years, and ≥75 years) consistent with published literature.15, 16, 17, 18^,^^20^
A Cox proportional hazards model was employed to evaluate the association between baseline characteristics and receipt of PCV20 within a multivariable framework. For these analyses, each age-specific subgroup from 19 years to 64 years was stratified into at risk versus high risk, and each age-specific subgroup from 65 years to ≥75 years was stratified into low risk, at risk, and high risk, respectively. Household income was excluded from the multivariable model owing to correlation with race. In all time to analyses, people who did not receive pneumococcal vaccine were censored at the end of follow-up, defined as the date of health plan disenrollment or the end of the study period, whichever occurred first.
A total of 14.9 million adults were aged ≥19 years and enrolled in a health plan contributing data to the Optum deidentified CDM on January 27, 2022. Among these adults, 11.2 million had ≥1 year of continuous health plan enrollment prior to January 27, 2022. After excluding adults with evidence of PCV20 or PCV15 prior to January 27, 2022 (<0.01%) and adults aged 19–64 years without evidence of at-risk/high-risk conditions, the study population totaled 6.8 million adults.
Most (84%) adults in the study population were aged ≥65 years, and 56% were female (Appendix Table 2, available online). The majority (70%) of adults were White; 10% were Black, 10% were Hispanic, and 4% were Asian (unknown, 7%). By risk conditions, 38% of the study population was designated as low risk, 39% were at risk, and 22% were high risk. More than one third (38%) of adults had evidence of pneumococcal vaccination prior to January 27, 2022 (PCV13 only, 16%; PPSV23 only, 11%; and PCV13 and PPSV23, 11%). Household income was <50,000–100,000 for 22% (unknown, 7%).
Uptake of any pneumococcal vaccine—based on the Kaplan–Meier method—was 13.2% (95% CI=13.1, 13.2) during the 21-month follow-up period (Figure 1). Cumulative uptake was highest for PPSV23 during the first 6 months of follow-up, though uptake of PCV20 relative to PPSV23 increased steadily and was approximately equal by May 2022 and higher thereafter. By the end of the follow-up period, use of PCV20 was highest (9.6%; 95% CI=9.6, 9.6), with lower uptake of PPSV23 (3.0%; 95% CI=3.0, 3.1), PCV13 (0.8%; 95% CI=0.8, 0.8), and PCV15 (0.2%; 95% CI=0.2, 0.3). By age/risk conditions, uptake of PCV20 was highest among adults aged 65–66 years (23.8%; 95% CI=23.5, 24.1), followed by at-risk/high-risk adults aged 60–64 years (12.1%; 95% CI=12.0, 12.3) and adults aged 67–74 years (11.6%; 95% CI=11.6, 11.6); uptake was lower among at-risk/high-risk adults aged 50–59 years (8.6%; 95% CI=8.5, 8.7) and adults aged ≥75 years (7.7%; 95% CI=7.7, 7.7) and lowest among at-risk/high-risk adults aged 19–49 years (4.7%; 95% CI=4.6, 4.8) (Figure 2). By history of immunization, PCV20 uptake was highest among adults who previously received PPSV23 only (15.1%; 95% CI=15.0, 15.2) or PCV13 only (10.6%; 95% CI=10.6, 10.7); uptake was lower among those who did not have evidence of prior pneumococcal vaccination (8.7%; 95% CI=8.7, 8.7) and those who previously received PCV13+PPSV23 (7.9%; 95% CI=7.8, 7.9) (Figure 3).Figure 1Cumulative uptake of pneumococcal vaccine, overall and by vaccine type*.The asterisk () denotes that uptake of vaccines was found to be statistically different across vaccine types during the last month of the study period (i.e., October 2023).PCV13, 13-valent pneumococcal conjugate vaccine; PCV15, 15-valent pneumococcal conjugate vaccine; PCV20, 20-valent pneumococcal conjugate vaccine; PPSV23, 23-valent pneumococcal polysaccharide vaccine.Figure 1Figure 2Cumulative uptake of PCV20, by age and comorbidity profile.The asterisk () denotes that uptake of PCV20 was found to be statistically different across subgroups defined on age and risk during the last month of the study period (i.e., October 2023).AR, at-risk; HR, high-risk; PCV20, 20-valent pneumococcal conjugate vaccine; y, year.Figure 2Figure 3Cumulative uptake of PCV20, by immunization historyThe asterisk (*) denotes that uptake of PCV20 was found to be statistically different across subgroups defined on immunization history during the last month of the study period (i.e., October 2023).PCV13, 13-valent pneumococcal conjugate vaccine; PCV20, 20-valent pneumococcal conjugate vaccine; PPSV23, 23-valent pneumococcal polysaccharide vaccine.Figure 3
Findings from adjusted analyses using a multivariable Cox proportional hazards model were largely comparable with those from unadjusted analyses using the Kaplan–Meier method (Table 1). Compared with at-risk adults aged 19–49 years, hazard ratios (95% CI) for PCV20 uptake were highest for adults aged 65–66 years (low 5.1, 95% CI=4.9, 5.2; at 6.2, 95% CI=6.0, 6.4; high 6.0, 95% CI=5.7, 6.2) and lowest for high-risk adults aged 19–49 years (1.0; 95% CI=1.0, 1.1). For adults aged <65 years, hazard ratios were similar for at-risk and high-risk adults within each age group; for adults aged ≥65 years, hazard ratios were lower for those in the low-risk subgroup. By immunization history, hazard ratios were lowest for adults who previously received PCV13+PPSV23 (0.84; 95% CI=0.83, 0.85) than for those who were vaccine naive (ref) and highest for those who previously received PPSV23 only (1.6; 95% CI=1.6, 1.7) or PCV13 only (1.3; 95% CI=1.2, 1.3). Results from analyses considering subgroups defined on race and household income are set forth in the Appendix (available online).Table 1Multivariable Cox Proportional Hazards Model for PCV20 UptakeTable 1Patient demographic/historyHazard ratios (95% CI) for PCV20 uptakeAge group (years) and risk conditions19–49, at-risk (ref)— 19–49, high-risk1.04 (0.99, 1.08) 50–59, at-risk1.87 (1.83, 1.91) 50–59, high-risk1.74 (1.68, 1.79) 60–64, at-risk2.60 (2.55, 2.67) 60–64, high-risk2.53 (2.46, 2.6) 65–66, low-risk5.07 (4.93, 5.21) 65–66, at-risk6.18 (6.01, 6.36) 65–66, high-risk5.95 (5.73, 6.17) 67–74, low-risk2.23 (2.19, 2.28) 67–74, at-risk2.84 (2.78, 2.9) 67–74, high-risk2.72 (2.66, 2.78) ≥75, low-risk1.26 (1.23, 1.29) ≥75, at-risk1.99 (1.95, 2.03) ≥75, high-risk1.96 (1.92, 2.00)Race White (ref)— Black1.07 (1.06, 1.07) Hispanic1.04 (1.03, 1.05) Asian1.18 (1.17, 1.2) Unknown1.03 (1.02, 1.04)Geographic region Northeast (ref)— South1.00 (1.00, 1.01) Midwest1.08 (1.07, 1.09) West0.83 (0.82, 0.84)History of pneumococcal vaccination PCV13 only1.25 (1.24, 1.26) PPSV23 only1.64 (1.63, 1.66) PCV13 and PPSV230.84 (0.83, 0.85) None (ref)—PCV13, 13-valent pneumococcal conjugate vaccine; PCV20, 20-valent pneumococcal conjugate vaccine; PPSV23, 23-valent pneumococcal polysaccharide vaccine.
In a retrospective study using healthcare claims for a population of 6.8 million U.S. adults, we evaluated uptake of pneumococcal vaccine after issuance of new recommendations by the U.S. ACIP on January 27, 2022. Study findings indicate that among adults for whom pneumococcal vaccination is recommended, fewer than 1 in 7 received any pneumococcal vaccine, fewer than 1 in 10 received PCV20, and fewer than 1 in 500 received PCV15 during the first 21 months after issuance of the updated guidance. Although cumulative PCV20 uptake was highest by the end of the follow-up period, in the months soon after release of the updated recommendations, PPSV23 uptake exceeded that of PCV20, indicating a lag in provider/patient adherence with the new guidelines. Study findings also indicate that vaccine uptake was considerably lower among certain subgroups, including eligible younger adults (i.e., those aged <65 years) with at-risk/high-risk conditions, especially those aged 19–49 years; adults aged ≥75 years; and adults without a history of pneumococcal vaccination.
Observed low levels of pneumococcal vaccine uptake among eligible U.S. adults persist despite major U.S. public health goals focusing on increasing their use.^21^ Increasing pneumococcal vaccine uptake is especially important in light of the recent coronavirus disease 2019 (COVID-19) pandemic that disrupted routine immunization services and highlighted the vulnerability of adults to respiratory infections. Although pneumococcal disease decreased during the pandemic, the incidence has since resurged and with it the potential for vaccination to improve health outcomes as well as reduce the economic burden on the healthcare system.^22^ Accordingly, new and/or better uptake strategies involving educational campaigns, improving vaccine access and affordability, strengthening the provider recommendation, and integrating vaccination with other preventive services are needed.23, 24, 25
To the best of the authors’ knowledge, this study is the first evaluation of adult pneumococcal vaccine uptake after issuance of the updated recommendations by the U.S. CDC-ACIP on January 27, 2022. We note that notwithstanding the differences in study populations and follow-up duration, the findings described in this paper indicating low uptake of pneumococcal vaccine among eligible adults after release of the 2022 recommendations are largely consistent with results from other studies that evaluated uptake after the release of earlier recommendations.
In a 2017 evaluation by Black et al.,^18^ less than one third of Medicare beneficiaries received PCV13 during the ∼2-year period after the 2014 U.S. CDC-ACIP recommendation for routine use among persons aged ≥65 years. In a 2022 study by Vietri and colleagues,^17^ <15% of high-risk adults aged 19–64 years received PCV13 during the 4-year period after the 2012 recommendation for its use in this target population. In a 2022 publication by Morga et al.,^15^ only 17% of adults aged ≥65 years and 5% of high-risk adults aged 19–64 years received PCV13 and PPSV23 (per recommendation) during the 6 and 4 years, respectively, after the corresponding recommendations. Finally, in the study by Ostropolets and colleagues,^16^ use of PCV13 or PPSV23 among at-risk/high-risk adults aged 19–64 years was reported to increase by only 6% at Year 1 of follow-up and by 21% at Year 5 of follow-up.
A few limitations and potential biases associated with the use of healthcare claims data for this study deserve mention. Use of operational algorithms to classify adults on the basis of their risk conditions undoubtedly resulted in some misclassification (i.e., false positives and false negatives). However, we note that these algorithms have been employed in several previously published studies.^5^^,^^6^^,^26, 27, 28 Because age was defined as of January 27, 2022, on the basis of birth year, it is likely that some persons were classified in a younger age group (e.g., 60–64 years) but received pneumococcal vaccine at an age that would qualify them for inclusion in an older age group (e.g., 65–66 years); a similar bias is noted for the classification of adults by risk conditions. Time-dependent variables for age and risk conditions were not employed because the study data source lacks birth dates (and thus precise age on a specific date is unknown) and owing to uncertainty regarding dates of onset/diagnosis for at-risk/high-risk conditions. However, we believe that given the relatively short follow-up period, the impact of these limitations on misclassification of adults by age and risk conditions was minimal.
Ascertainment of pneumococcal vaccine history is limited to available healthcare claims information during the approximate 7-year period preceding the updated recommendations. Across age-/risk-specific subgroups, mean duration of health plan enrollment during this ∼7-year period ranged from 3.5 to 5.4 years. Accordingly, vaccine histories may be misclassified for some adults, such as those who received vaccine while enrolled in other health plans or prior to 2015. Although person-level characteristics associated with vaccine uptake were evaluated within a multivariable framework, results from these analyses may be confounded if other (e.g., unobserved) characteristics not considered in the regression model were associated with 1 or more of the independent variables and vaccine uptake. It was implicitly assumed that differential follow-up was noninformative; the veracity of this assumption is unknown. Variables for race and household income in the Optum CDM are based on a mix of individual-level and ecologic-level data; to the best of the authors’ knowledge, the methodology for designating race and household incomes has not been evaluated against a gold standard, and thus the accuracy of these variables is uncertain.^19^
Because the composition of the study population is not reflective of the overall U.S. population by age and risk conditions and because adults in the study population may be more compliant with vaccination recommendations than others, some estimates of vaccine uptake may be upwardly biased. For example, 84% of vaccine-eligible adults in the study population were aged ≥65 years, whereas in the U.S. population, the corresponding value is substantially lower.^5^^,^^29^ Persons who are uninsured and those with other types of health insurance coverage are not represented in the study database. Accordingly, caution should be used when generalizing study results to other populations and settings. Likewise, the adult pneumococcal recommendation has been revised since the conduct of the study to both expand the choice of pneumococcal vaccines and extend the recommendation to all adults aged ≥50 years, potentially further altering uptake across age and risk groups.^30^
Finally, this study describes delayed implementation of the recommendation but cannot explain it with the available data. Vaccine choice shortly after recommendation could be driven by a variety of factors, such as use of in-stock vaccine, delays in receipt of the newer vaccines, lack of awareness of the new recommendation, and concerns over insurance coverage. Future research assessing the drivers of delayed implementation of vaccine recommendations may be useful in developing interventions to facilitate adherence to new recommendations.
Fewer than 1 in 7 U.S. adults aged ≥19 years in this study population received any pneumococcal vaccine—mostly PVC20—in the first 21 months after publication of the updated ACIP recommendations, and PCV20 uptake was especially low in the months soon after issuance of the guidelines. Uptake was highest among those near the age of 65 years and considerably lower in both younger and older adults; uptake was particularly low among adults aged 19–49 years and those without a history of pneumococcal vaccination. Routine evaluation of vaccination status by providers and additional strategies to increase the uptake of current recommendations are warranted.