Receipt of respiratory vaccines among patients with heart failure in a multicenter health system registry

26

Heart failure affects people of all ages and is a leading cause of death for both men 27 and women in most racial and ethnic groups in the United States. Infections are common causes of 28 hospitalizations in heart failure, with respiratory infections as the most frequent diagnosis. 29 Vaccinations provide significant protection against preventable respiratory infections. Despite being 30 an easily accessible intervention, prior studies suggest vaccines are underused in patients with heart 31 failure. 32

An observational study of 7341 adults with heart failure was conducted using data from 33 an integrated, multicenter, academic health system in Southern California from 2019 to 2022. 34 Logistic regression models were used to determine the rates of influenza, pneumococcal, and 35 COVID-19 vaccination among a population of patients with heart failure (heart failure preserved 36 ejection fraction [HFpEF], heart failure mildly reduced ejection fraction [HFmrEF], heart failure 37 reduced ejection fraction [HFrEF], and heart failure unspecified ejection fraction [HFuEF]) and 38 identify whether heart failure phenotype is associated with vaccination status. 39

Vaccination rates varied between influenza, pneumococcal, and COVID-19 vaccines. Of 40 the three respiratory vaccines, 54.5% of patients had received an influenza vaccine, 74.7% had 41 received a pneumococcal vaccine, and 81.3% had received a COVID-19 vaccine. There were no 42 sex-based differences by vaccination status. Patients with HFpEF and HFmrEF had the highest 43 vaccination levels in all three vaccine groups. In adjusted models, patients with HFpEF had higher 44 odds of being vaccinated for influenza (aOR=1.34, 95% CI=1.19-1.53), pneumococcal (aOR=1.28, 45 95% CI=1.10-1.48), and COVID-19 (aOR=1.25, 95% CI=1.07-1.47) compared to HFuEF patients. 46 Patients with HFrEF had lower odds of being vaccinated for pneumococcal (aOR=0.81, 95% 47 CI=0.70-0.93) than patients with HFuEF. 48

Patients with HFrEF had the lowest levels of respiratory vaccination compared to 49 other specified heart failure categories. Interventions are needed to increase vaccination education 50 and offerings, especially to patients with HFrEF.  51 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 1

52 Heart failure is associated with considerable morbidity, mortality, and cost. Infections are 53 common causes of hospitalizations in heart failure, with respiratory infections as the most frequent 54 diagnosis [1, 2]. Respiratory infections can cause significant disease and poor outcomes in patients 55 with heart failure. Influenza, pneumococcal, and COVID-19 vaccines can help protect against 56 significant morbidity or mortality in vulnerable heart failure patients and reduce the incidence of 57 common respiratory infections [3-6]. Multiple studies have shown the benefits of vaccination for 58 preventing and managing heart disease [3, 7-12]. Despite being an easily accessible intervention, 59 vaccinations are underutilized and understudied in patients with heart failure. 60 Patients with heart failure are at a higher risk of severe complications from respiratory 61 illnesses [2, 13]. Influenza, pneumonia, and COVID-19 infections can trigger arrhythmias, acute 62 coronary syndromes, and acute exacerbations of underlying heart failure [14]. Vaccinations provide 63 the best protection against preventable respiratory disease [3]. A large, recent trial conducted in 64 low- and middle-income countries confirmed that influenza vaccination could reduce the risk of 65 hospitalization and pneumonia in patients with symptomatic heart failure [15, 16]. However, prior 66 studies have found low levels of vaccination in the heart failure population. In a study of 313761 67 patients hospitalized at centers participating in the Get With The Guidelines-Heart Failure (GWTG-68 HF) registry from 2012 to 2017 [8], nearly 1 in 3 patients hospitalized with heart failure were not 69 vaccinated for influenza or pneumococcal, and vaccination rates did not improve across the five-70 year study period. The hospitals with higher vaccination rates performed well in other heart failure 71 quality of care measures [7, 8]. 72 Guideline‐recommended therapies are vital in improving outcomes and preventing recurrent 73 cardiovascular events [17, 18]. Due to a lack of randomized clinical trials, heart failure guidelines 74 from the American Heart Association and American College of Cardiology Foundation 75 (AHA/ACCF), Heart Failure Society of America (HFSA), and European Society of Cardiology 76 (ESC) do not cover vaccination recommendations in detail. The U.S. Preventive Services Task 77 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 2 Force (USPSTF) and Advisory Committee on Immunization Practices (ACIP) remain vague on 78 specific vaccine recommendations but make general recommendations for chronic heart, lung, liver, 79 diabetes, and smoking [5]. Nonetheless, there is consensus in the scientific community to provide 80 influenza, pneumococcal, and COVID-19 vaccines to all patients with heart failure in the absence of 81 contraindications [19, 20]. Given limited evidence on respiratory vaccines among patients with 82 heart failure, especially for the newer COVID-19 vaccine, the aim of this study was to investigate 83 levels of influenza, pneumococcal, and COVID-19 vaccination among a heart failure patient 84 population and whether heart failure phenotype was associated with vaccination status. 85

86 Design and Data 87 This observational study used data from an integrated, multicenter academic health system 88 from 2019 to 2022. The database included all known adult patients with heart failure diagnoses on 89 their Electronic Health Record (EHR) problem list. The problem list is updated by clinicians who 90 are members of the patient’s care team in ambulatory and inpatient settings. The registry includes 91 data about patient demographics, clinical and social characteristics, treatment background, and 92 clinician type. The UCLA Institutional Review Board provided a determination of exempt status for 93 this study as an analysis of existing de-identified data. 94 Sample and Setting 95 The study cohort included 7341 adults over 18 years of age with heart failure. The health 96 system in Southern California is comprised of four hospitals and over 250 ambulatory care clinics 97 with more than 2.5 million patient visits and 100000 hospital admissions annually. All individuals 98 listed in the dataset were eligible for inclusion in the study sample, and the entire cohort was used 99 for the analysis. 100 Outcome Variable: Vaccination Status 101 Vaccination status data were populated into the patient registry from the patient’s 102 immunization record in the EHR. Vaccination status data included within-health system and 103 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 3 external data sources. External immunization records were populated from the California 104 Immunization Registry (CAIR2) and Epic Care Everywhere (Epic Systems Corporation, Inc. 105 Verona, WI). The CAIR2 is a secure, confidential, statewide computerized immunization 106 information system for California residents. Epic Care Everywhere allows health systems to share 107 medical records with other health systems using the same EHR. Patients were considered 108 vaccinated for each vaccine type when the following conditions were met: 109 1. Influenza vaccine: documentation of immunization type (42 total) or professional billing 110 charge (20 total) in the EHR during the influenza vaccination period from August of the 111 previous year to March of the current year. 112 2. Pneumococcal vaccine: documentation of one dose of pneumococcal polysaccharide vaccine 113 23-valent (Pneumovax) or pneumococcal conjugate vaccine 13-valent (Prevnar 13) (no time 114 specified). The pneumococcal conjugate vaccine 20-valent (Prevnar 20) was not captured in 115 the EHR during the study period; it was added two months later. 116 3. COVID-19 vaccine: documentation of two doses for the mRNA manufacturers (Pfizer-117 BioNTech, Moderna), one dose for Johnson and Johnson, or one dose for unspecified (no 118 time specified). 119 Exposure Variable: Heart Failure Category 120 Heart failure type was classified into four categories: heart failure with preserved ejection 121 fraction (HFpEF, ejection fraction [EF] ≥50%), heart failure with mildly reduced ejection fraction 122 (HFmrEF, EF 41-49%), heart failure with reduced ejection fraction (HFrEF, EF ≤40%), and heart 123 failure with unspecified ejection fraction (HFuEF, EF unspecified) [21, 22]. Over 2000 ICD-10-CM 124 codes/subsets are associated with a heart failure diagnosis, comprising approximately 3% of total 125 ICD-10 codes. In the health system heart failure registry, these various ICD-10-CM diagnosis codes 126 were populated from the EHR’s problem list into the registry and categorized into the four heart 127 failure groups. To improve the accuracy of the problem list, clinicians were notified through a Best 128 Practice Alert (BPA) in the EHR when a heart failure diagnosis may be erroneously missing from 129 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 4 the problem list based on documentation in the medical history, one or more billing codes, EF 130 value, or prescriptions of heart failure medications. Within the healthcare system, there was an 131 active push to refine generic heart failure diagnoses to more specific ones as problems became more 132 defined during the course of patient treatment. If multiple heart failure diagnoses were on the 133 problem list, the algorithm selected the diagnosis based on the lowest EF on file in the last three 134 years. Based on this logic, there was no overlap between the heart failure categories. 135 Covariates 136 Covariates were demographic, s ocial, and health characteristics that might influence the 137 relationship between exposures and outcomes. These included patient gender (male/female), age 138 (<65 years, ≥65 years), race/ethnicity (Asian, Black, Hispanic, White, Other), primary language 139 (English, non-English), Accountable Care Organization status (ACO), insurance type (Commercial, 140 Medicaid, Medicare, Managed Care, Other), and a Social Vulnerability Index (SVI) score. SVI 141 measured four domains of social vulnerability from United States Census data, including (1) 142 socioeconomic status, (2) household composition and disability, (3) minority status and language, 143 and (4) housing and transportation [23]. Higher scores denote higher social vulnerability. We 144 measured patient comorbidity characteristics, including the presence of chronic kidney disease 145 (CKD), diabetes mellitus, hypertension, and ischemic heart disease. Finally, we assessed whether 146 the patient was registered with a primary care physician, cardiologist, or both in the health system. 147 Statistical Analysis 148 RStudio was used for statistical analysis. Descriptive statistics and frequencies were used to 149 characterize the sample on all analytic variables. Chi square tests were used to compare sample 150 differences by patient and clinician characteristics and vaccination status. We estimated odds of 151 vaccination by heart failure type in three logistic regression models, first assessing the bivariate 152 relationship between the exposure and outcome and then adjusting for all covariates listed above. 153 For both unadjusted and adjusted models, we estimated separate models for each vaccination 154 outcome. 155 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 5

156 Patient and clinician characteristics of the heart failure population by vaccination status are 157 presented in Table 1. The sample was primarily male (55.2%, n=4049), aged 65 years or older 158 (74.5%, n=5467), self-identified as White (52.9%, n=3886), and insured by Medicare (66.1%, 159 n=4851). Thirty-one percent (n=2256) had an SVI score of medium to high. Sixty percent of 160 patients had a registered primary care physician (n=4430), while 52.6% had a registered 161 cardiologist (n=3863) and 36.4% had both clinician types (n=2675). Of the total sample, 37.7 % 162 (n=2764) were patients with HFpEF, 2.7% (n=198) were heart failure with HFmrEF, 28.9% 163 (n=2127) were patients with HFrEF, and 30.7% (n=2252) were patients with HFuEF (Figure 1). 164 Patient comorbidity characteristics included 30% CKD (n=2205), 32.1% diabetes mellitus 165 (n=2356), 74.8% hypertension (n=5494), and 21% ischemic heart disease (n=1545). 166 There were no sex-based differences seen in vaccination status. An estimated 54.5% of 167 patients (n=4001) received an influenza vaccine, 74.7% (n=5483) received a pneumococcal vaccine, 168 and 81.3% (N=5970) received a COVID vaccine (Figure 2). In bivariate tests, there were significant 169 differences in vaccination status by heart failure category for all three vaccines (p<0.001). In 170 multiple logistic regression models, patients with HFpEF (aOR=1.34, 95% CI=1.19-1.53) and 171 HFmrEF (aOR=1.40, 95% CI=1.01-1.95) had higher odds of being vaccinated for influenza 172 compared to patients with HFuEF. Likewise, patients with HFpEF had higher odds of being 173 vaccinated for pneumococcal (aOR=1.28, 95% CI=1.10-1.48), and patients with HFrEF had lower 174 odds of being vaccinated for pneumococcal compared to patients with HFuEF (aOR=0.80, 95% 175 CI=0.70-0.93). Patients with HFpEF had higher odds of being vaccinated for COVID-19 than those 176 with HFuEF (aOR=1.25, 95% CI=1.07-1.47) (Table 2). 177

178 There have been very few studies examining the receipt of common respiratory vaccines 179 concurrently among patients with heart failure, especially investigating the receipt of COVID-19 180 vaccines. In this study of adults with heart failure, we found that individuals with HFrEF had the 181 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 6 lowest rates of respiratory vaccination that could prevent infections and hospitalizations. Patients 182 with HFpEF had the highest odds of receiving all three vaccine types, adjusting for demographic 183 and social factors and patient comorbidities. Vaccination rates for the three vaccines were 184 comparable to the general public [20], but there were significant gaps in vaccination for patients 185 with HFrEF who could benefit most from this preventive intervention. 186 Of the three respiratory vaccines, the vaccination rates were the lowest for the influenza 187 vaccine (54.5%). The influenza vaccine has been available since 1945, but the public interest in 188 receiving this vaccine is low. A 2021 survey found that in the general U.S. public, only 56% 189 (N=631) of adults wanted to receive an influenza vaccine, and among those who were unsure, 39% 190 did not believe flu vaccines worked very well [24]. However, survey participants also believed that 191 healthcare providers were the primary and most trusted source of information about influenza and 192 influenza vaccination, suggesting that healthcare providers play an important role in vaccine uptake. 193 The Centers for Disease Control and Prevention recommends assessing the vaccination status of 194 patients and addressing misconceptions about vaccines at all clinical encounters [19]. Incorporating 195 tailored patient education on preventive practices like annual flu shots as part of patient self-196 management, especially for patients with more severe heart failure, may increase vaccine uptake. 197 Pneumococcal pneumonia infections are one of the leading causes of heart failure admission 198 [25, 26]. In this study, patients with HFrEF had the lowest levels of pneumococcal vaccination and 199 lower odds of vaccination than other heart failure types. Some possible reasons might be that these 200 patients were managed mainly by cardiologists focused on advanced therapies than closing 201 preventive care gaps. In addition, some cardiology clinics did not have the workflows to provide 202 respiratory vaccinations to patients during their clinic visits. Prior studies have found a high 203 incidence of pneumonia among patients with heart failure, even those with HFpEF [27]. Meta-204 analyses have also found that in the general adult population, pneumococcal vaccination is 205 associated with a decreased risk of cardiovascular events and myocardial infarction in all age 206 groups [4]. Since much of the clinical management of patients with HFpEF is primarily directed 207 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 7 toward treating associated conditions [17], vaccination in this population may produce more 208 favorable outcomes. 209 Underlying heart disease is associated with an increased risk for in-hospital death among 210 patients hospitalized with COVID-19 [28]. Of the three vaccines in this study, the vaccination rates 211 were the highest for the COVID-19 vaccine (81.3%). At the beginning of the COVID-19 pandemic, 212 there was a great deal of media attention to COVID-19 hospitalization and mortality rates for those 213 with high-risk medical conditions, and initially, COVID-19 vaccines became available earlier for 214 individuals with high-risk medical conditions, including heart failure [24]. Efforts to increase 215 COVID-19 vaccination in this population may explain why COVID-19 vaccination levels were high 216 in our sample. However, there were still significant vaccination differences among the four heart 217 failure categories, with those with HFrEF or HFuEF having the lowest levels of COVID-19 218 vaccination. As such, there is a need for targeted communication and outreach to patients with 219 HFrEF. 220 This study suggests an urgent need to reach patients with heart failure with respiratory 221 vaccines, especially the influenza vaccine, which had the lowest rates of uptake in our study. 222 Although education on the need for respiratory vaccines in clinical encounters can improve care for 223 some patients, a systems approach could be more effective at improving the health of the entire 224 population and reducing vaccine inequities. Clinicians should utilize both primary care and 225 cardiology encounters to assess the vaccination status of patients and address patient hesitancy and 226 vaccine side effects and long-term effects. We also recommend integration with external state 227 vaccine registries to ensure up-to-date and accurate capture of vaccination status, provide vaccines 228 within the visit, and offer referrals to retail pharmacy if needed. Clinicians may also benefit from 229 tracking population-level data at the provider and clinic levels to set organizational benchmarks for 230 vaccination, make data transparent, and provide incentives to increase vaccination rates. Future 231 research could evaluate patient and clinician factors that influence vaccine rates. The 2022 232 ACC/AHA/HFSA heart failure guidelines provide a Class 2a, B-NR recommendation that 233 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 8 vaccinating against respiratory illnesses is reasonable to reduce mortality [29]. In the future, it may 234 be beneficial for such clinical guidelines to make specific recommendations for respiratory 235 vaccines. 236

237 There are strengths and limitations to this study to consider in interpreting the findings. Our 238 study used a large, population-based cohort of adults with heart failure derived from a multicenter 239 integrated academic health system. Validity testing of the data was conducted by a team of primary 240 care and cardiology clinicians, quality specialists, and computer programmers and analysts during a 241 six-month period prior to data analysis. However, a clinician had to actively include heart failure on 242 the EHR’s problem list for the patient to be included, and thus it is possible that this dataset does 243 not capture all patients in the health system with heart failure. We had complete vaccine outcomes 244 data pulled from multiple sources, including external sources of CAIR2 and Epic Care Everywhere. 245 Our study did not capture patients who received Prevnar 20 for the pneumococcal vaccine, which 246 was added to the health system EHR after the time of the study. In addition, the COVID-19 boosters 247 were not counted for numerator compliance. At the time of the study, hospitalization and mortality 248 data were unavailable for the patient population. Therefore, additional analysis could not be done to 249 link vaccination status to outcomes data. In addition, vaccination status in relation to guideline-250 directed medical therapy compliance was not measured. These areas are of interest for future 251 studies. Despite these limitations, our findings can inform ways to implement population-level 252 interventions to improve vaccination outcomes among patients with heart failure. 253

254 Heart failure is a complex chronic disease to manage, but preventive care practices like 255 vaccinations can improve patient outcomes and reduce the risk of infections, hospitalization, and 256 mortality. This study found differences in respiratory vaccination rates by heart failure categories 257 and highlights care gaps that can be addressed through population-level strategies to increase 258 vaccination. Because patients with HFrEF who could potentially benefit most from respiratory 259 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 9 vaccines are least likely to receive them, it is important for individual clinicians as well as health 260 systems to develop strategies to improve use and overcome heart failure vaccine differences. 261 262 263 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 10 Acknowledgments: We want to acknowledge the UCLA Department of Medicine Quality and 264 Cardiology Division’s dedication and hard work in supporting quality improvement efforts for the 265 heart failure patient population. 266 Sources of Funding: No funding was received for this study. 267 Author Disclosures: Dr. Gregg Fonarow reports consulting for Abbott, Amgen, AstraZeneca, 268 Bayer, Boehringer Ingelheim, Cytokinetics, Eli Lilly, Johnson & Johnson, Medtronic, Merck, 269 Novartis, and Pfizer. Dr. Chidinma Chima-Melton reports consulting for AstraZeneca, Boehringer 270 Ingelheim, and Gilead. Both are unrelated to the contents of this article. 271 272 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 11

273 1. Fonarow, G.C., et al., Factors identified as precipitating hospital admissions for heart 274 failure and clinical outcomes: findings from OPTIMIZE-HF. Arch Intern Med, 2008. 275 168(8): p. 847-54. 276 2. Alon, D., et al., Predictors and outcomes of infection-related hospital admissions of heart 277 failure patients. PLoS One, 2013. 8(8): p. e72476. 278 3. Girerd, N., et al., Vaccination for Respiratory Infections in Patients with Heart Failure. J 279 Clin Med, 2021. 10(19). 280 4. Marra, F., et al., The protective effect of pneumococcal vaccination on cardiovascular 281 disease in adults: A systematic review and meta-analysis. Int J Infect Dis, 2020. 99: p. 204-282 213. 283 5. Bhugra, P., et al., Determinants of Influenza Vaccine Uptake in Patients With 284 Cardiovascular Disease and Strategies for Improvement. J Am Heart Assoc, 2021. 10(15): 285 p. e019671. 286 6. Gotsman, I., et al., Influenza Vaccination and Outcome in Heart Failure. Am J Cardiol, 287 2020. 128: p. 134-139. 288 7. Bhatt, A.S., et al., Can Vaccinations Improve Heart Failure Outcomes?: Contemporary 289 Data and Future Directions. JACC Heart Fail, 2017. 5(3): p. 194-203. 290 8. Bhatt, A.S., et al., Vaccination Trends in Patients With Heart Failure: Insights From Get 291 With The Guidelines-Heart Failure. JACC Heart Fail, 2018. 6(10): p. 844-855. 292 9. Fountoulaki, K., et al., Beneficial Effects of Vaccination on Cardiovascular Events: 293 Myocardial Infarction, Stroke, Heart Failure. Cardiology, 2018. 141(2): p. 98-106. 294 10. Rodrigues, B.S., et al., Influenza vaccination in patients with heart failure: a systematic 295 review and meta-analysis of observational studies. Heart, 2020. 106(5): p. 350-357. 296 11. Vardeny, O., et al., Influenza Vaccination in Patients With Chronic Heart Failure: The 297 PARADIGM-HF Trial. JACC Heart Fail, 2016. 4(2): p. 152-158. 298 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 12 12. Ciszewski, A., Cardioprotective effect of influenza and pneumococcal vaccination in 299 patients with cardiovascular diseases. Vaccine, 2018. 36(2): p. 202-206. 300 13. Drozd, M., et al., Infection-Related Hospitalization in Heart Failure With Reduced Ejection 301 Fraction: A Prospective Observational Cohort Study. Circ Heart Fail, 2020. 13(5): p. 302 e006746. 303 14. Rojulpote, C., et al., COVID-19 and the Heart. Colomb Med (Cali), 2020. 51(2): p. e4320. 304 15. Kumbhani, D.J. Influenza Vaccine to Prevent Adverse Vascular Events - IVVE. 2022 [cited 305 2022 September 10, 2022]; Available from: https://www.acc.org/latest-in-306 cardiology/clinical-trials/2022/04/02/15/50/ivve. 307 16. Loeb, M., et al., Randomized controlled trial of influenza vaccine in patients with heart 308 failure to reduce adverse vascular events (IVVE): Rationale and design. Am Heart J, 2019. 309 212: p. 36-44. 310 17. Yancy, C.W., et al., 2013 ACCF/AHA guideline for the management of heart failure: 311 executive summary: a report of the American College of Cardiology Foundation/American 312 Heart Association Task Force on practice guidelines. Circulation, 2013. 128(16): p. 1810-313 52. 314 18. Yancy, C.W., et al., 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA 315 Guideline for the Management of Heart Failure: A Report of the American College of 316 Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the 317 Heart Failure Society of America. Circulation, 2017. 136(6): p. e137-e161. 318 19. Centers for Disease Control and Prevention. Strategies for Increasing Adult Vaccination 319 Rates. 2021; Available from: https://www.cdc.gov/vaccines/hcp/adults/for-320 practice/increasing-vacc-rates.html. 321 20. Centers for Disease Control and Prevention. Vaccination Coverage among Adults in the 322 United States, National Health Interview Survey. 2022; Available from: 323 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 13 https://www.cdc.gov/vaccines/imz-managers/coverage/adultvaxview/pubs-324 resources/vaccination-coverage-adults-2019-2020.html. 325 21. Ponikowski, P., et al., 2016 ESC Guidelines for the diagnosis and treatment of acute and 326 chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic 327 heart failure of the European Society of Cardiology (ESC)Developed with the special 328 contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J, 2016. 37(27): 329 p. 2129-2200. 330 22. Savarese, G., et al., Heart failure with mid-range or mildly reduced ejection fraction. Nat 331 Rev Cardiol, 2022. 19(2): p. 100-116. 332 23. Lewis, B.E.F.E.W.G.E.J.H.J.L.H.a.B., A Social Vulnerability Index for Disaster 333 Management. Journal of Homeland Security and Emergency Management, 2011. 8(1). 334 24. National Foundations for Infectious Diseases. 2021 National Survey: Attitudes about 335 Influenza, Pneumococcal Disease, and COVID-19. 2021; Available from: 336 https://www.nfid.org/2021-national-survey-attitudes-about-influenza-pneumococcal-337 disease-and-covid-19/. 338 25. Mor, A., et al., Chronic heart failure and risk of hospitalization with pneumonia: a 339 population-based study. Eur J Intern Med, 2013. 24(4): p. 349-53. 340 26. Thomsen, R.W., et al., The impact of pre-existing heart failure on pneumonia prognosis: 341 population-based cohort study. J Gen Intern Med, 2008. 23(9): p. 1407-13. 342 27. Shen, L., et al., Incidence and Outcomes of Pneumonia in Patients With Heart Failure. J 343 Am Coll Cardiol, 2021. 77(16): p. 1961-1973. 344 28. Mehra, M.R., et al., Cardiovascular Disease, Drug Therapy, and Mortality in Covid-19. N 345 Engl J Med, 2020. 382(25): p. e102. 346 29. Heidenreich, P.A., et al., 2022 AHA/ACC/HFSA Guideline for the Management of Heart 347 Failure: A Report of the American College of Cardiology/American Heart Association Joint 348 Committee on Clinical Practice Guidelines. Circulation, 2022. 145(18): p. e895-e1032. 349 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 14 Figure 1: Heart failure patient population by heart failure category 350 HFpEF, 37.7% HFmrEF, 2.7% HFrEF, 28.9% HFuEF, 30.7% Heart Failure Patient Population by Heart Failure Category 351 Legend: Proportion of adult patients with each of four heart failure types in a sample of 7341 352 patients with heart failure derived from a health system Heart Failure Registry in Southern 353 California from 2019 to 2022. HFpEF= heart failure preserved ejection fraction; HFmrEF= heart 354 failure mildly reduced ejection fraction; HFrEF= heart failure reduced ejection fraction; HFuEF= 355 heart failure unspecified ejection fraction. 356 357 358 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 15 Figure 2: Vaccination rates for influenza, pneumococcal, and COVID-19 vaccines by heart failure 359 category 360 361 Legend: Percentage of adult patients who received respiratory vaccines (influenza, pneumococcal, 362 COVID-19) by heart failure type. The sample includes 7341 adult patients with heart failure derived 363 from a health system Heart Failure Registry in Southern California from 2019 to 2022. HFpEF= 364 heart failure preserved ejection fraction; HFmrEF= heart failure mildly reduced ejection fraction; 365 HFrEF= heart failure reduced ejection fraction; HFuEF= heart failure unspecified ejection fraction. 366 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 17 Table 1. Patient and clinician characteristics by influenza, pneumococcal, and COVID-19 vaccination status 367 368 Total Population (n=7341) Influenza Vaccine (n=4001) P Value Pneumococcal Vaccine (n=5483) P Value COVID-19 Vaccine (n=5970) P Value n (%) n (%) n (%) n (%) Heart Failure Category <0.001 <0.001 <0.001 HFpEF (EF ≥50%) 2764 (37.7) 1749 (63.3) 2288 (82.8) 2382 (86. 2) HFmrEF (EF 41-49%) 198 (2.7) 129 (65.2) 139 (70.2) 175 (88.4) HFrEF (EF ≤40%) 2127 (28.9) 1075 (50.5) 1450 (68.2) 1683 (79. 1) HFuEF (EF unspecified) 2252 (30.7) 1048 (46.5) 1606 (71.3) 1 730 (76.8) Gender 0.182 0.201 0.889 Female 3292 (44.8) 1823 (55.4) 2483 (75.4) 2680 (81.4) Male 4049 (55.2) 2178 (53.8) 3000 (74.1) 3290 (81.3) Age <0.001 <0.001 <0.001 <65 1874 (25.5) 755 (40.3) 1010 (53.9) 1365 (72.8) ≥65 5467 (74.5) 3246 (59.4) 4473 (81.8) 4605 (84.2) Race/Ethnicity <0.001 <0.001 <0.001 Asian 535 (7.3) 316 (59.1) 425 (79.4) 451 (84.3) Black 758 (10.3) 363 (47.9) 554 (73.1) 589 (77.7) Hispanic 996 (13.6) 521 (52.3) 754 (75.7) 791 (79.4) White 3886 (52.9) 2283 (58.7) 3023 (77.8) 3263 (84.0) Other 1166 (15.9) 518 (44.4) 727 (62.3) 876 (75.1) Language <0.001 0.404 <0.001 English 6532 (89.0) 3630 (55.6) 4889 (74.8) 5359 (82.0) Non-English 809 (11.0) 371 (45.9) 594 (73.4) 611 (75.5) ACO Status <0.001 <0.001 <0.001 Member 1656 (22.6) 1197 (72.3) 1464 (88.4) 1529 (92.3) Non-Member 5685 (77.4) 2804 (49.3) 4019 (70.7) 4441 (78.1) Insurance <0.001 <0.001 <0.001 Commercial 1168 (15.9) 503 (43.1) 615 (52.7) 925 (79.2) Medicaid 460 (6.3) 151 (32.8) 250 (54.3) 284 (61.7) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 18 Medicare 4851 (66.1) 2793 (57.6) 3918 (80.8) 4027 (83.0) Managed Care 671 (9.1) 511 (76.2) 580 (86.4) 624 (93.0) Other 191 (2.6) 43 (22.5) 120 (62.8) 110 (57.6) SVI <0.001 <0.001 90 381 (5.2) 163 (42.8) 267 (70.1) 262 (68.8) Medium risk 50 to 90 1875 (25.5) 880 (46.9) 1311 (69.9) 1437 (76.6) Low risk <50 4826 (65.8) 2866 (59.4) 3732 (77.3) 4110 (85.2) Risk unknown 259 (3.5) 92 (35.5) 173 (66.8) 161 (62.2) Comorbidities <0.001 <0.001 <0.001 CKD 2205 (30.0) 1559 (70.7) 1899 (86.1) 2014 (91.3) Diabetes mellitus 2356 (32.1) 1376 (58.4) 1936 (82.2) 1963 (8 3.3) Hypertension 5494 (74.8) 3204 (58.3) 4350 (79.2) 4581 (83.4) Ischemic disease 1545 (21.0) 975 (63.1) 1253 (81.1) 1338 (86. 6) Clinician Type <0.001 <0.001 <0.001 Primary Care 4430 (60.3) 2989 (67.5) 3668 (82.8) 3992 (90.1) Cardiology 3863 (52.6) 2501 (64.7) 2987 (77.3) 3417 (88.5) Both 2675 (36.4) 1953 (73.0) 2243 (83.9) 2459 (91.9) Other 1723 (23.5) 464 (26.9) 1071 (62.1) 1090 (63.3) Notes. HFpEF= heart failure preserved ejection fraction; HFmrEF= heart failure mildly reduced ejection fraction; HFrEF= heart failure reduced ejection fraction; HFuEF= 369 heart failure unspecified ejection fraction; ACO= accountable care organization; SVI= social vulnerability index; CKD= chronic kidney disease; COVID-19= coronavirus 370 disease 2019. Race/Ethnicity “Other” category included American Indian/Alaskan Native, Native Hawaiian/Pacific Islander, and Unknown. Insurance ‘Other” category 371 included International Payor, Package Billing, Self-Pay, and Worker’s Compensation. ACO plans included CMS Medicare Shared Savings Plan, Anthem Blue Cross PPO 372 Enhanced Care Coordination, Cigna Collaborative Accountable Care, and Health Net Blue & Gold ACO. In California, not all addresses were mapped to the Census Tract 373 due to their geographical location; therefore, there were some SVI “Risk unknown” values. Clinician type “Other” category included patients who had no primary care or 374 cardiology clinicians registered in the health system. P value is significant at the 0.01 level. 375 376 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint 19 Table 2. Odds of vaccination by heart failure category 377 Predictors Influenza Vaccine Pne umococcal Vaccine COVID-19 Vacc ine OR (95% CI) aOR (95% CI) OR (95% CI) aOR (95% CI) OR (95% CI) a OR (95% CI) Heart Failure Category (reference: HFuEF [EF unspecified]) HFpEF (EF ≥50) 1.98 (1.77-2.22) 1.34 (1.19-1.53) 1.93 (1.69-2.2 1) 1.28 (1.10-1.48) 1.88 (1.63-2.18) 1.25 (1.07-1.47) HFmrEF (EF 41-49%) 2.15 (1.59-2.92) 1.40 (1.01-1.95) 0.95 (0.69 -1.31) 0.76 (0.54-1.09) 2.30 (1.50-3.68) 1.34 (0.86-2.20) HFrEF (EF ≤40%) 1.17 (1.04-1.32) 1.00 (0.87-1.14) 0.86 (0.76-0. 98) 0.81 (0.70-0.93) 1.14 (0.99-1.32) 0.94 (0.81-1.10) Notes. Three unadjusted and adjusted multiple logistic regression models estimating odds of vaccination for influenza, pneumococcal, and COVID-19 vaccines by heart 378 failure category. Adjusted models included covariates for gender, age, race/ethnicity, insurance type, whether the patient was a member of an Accountable Care 379 Organization (ACO), primary language, Social Vulnerability Index (SVI) score, clinician type, and number of comorbidities. In this analysis, N=7341 represents patients 380 with heart failure from 2019 to 2022 in the health system Heart Failure Registry. HFpEF= heart failure preserved ejection fraction; HFmrEF= heart failure mildly reduced 381 ejection fraction; HFrEF= heart failure reduced ejection fraction; HFuEF= heart failure unspecified ejection fraction; EF= Ejection Fraction; OR= odds ratio; CI= confidence 382 interval. 383 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 8, 2023. ; https://doi.org/10.1101/2023.09.05.23295101doi: medRxiv preprint