Vaccine: X 13 (2023) 100264 Contents lists available at ScienceDirect Vaccine: X journal homepage: www.elsevier.com/locate/jvacx Evaluating clinical effectiveness and impact of anti-pneumococcal vaccination in adults after universal childhood PCV13 implementation in Catalonia, 2017–2018 Angel Vila-Córcoles a,b, Olga Ochoa-Gondar a, Cinta de Diego-Cabanes a,⇑, Eva M. Satué-Gracia a,b, Verónica Torras-Vives a, M. José Forcadell-Peris a, Domingo Ribas-Seguí a, Angel Vila-Rovira a, Clara Rodríguez-Casado c a Primary Health Care Service ‘‘Camp de Tarragona”, Institut Català de la Salut, Tarragona, Spain b Unitat de Suport a la Recerca of Tarragona, Institut Universitari d’Investigació en Atenció Primària Jordi Gol (IDIAP Jordi Gol), Tarragona, Spain c Information System for the Improvement of Research in Primary Care (SIDIAP), Primary Care Research Institute Jordi Gol, Universitat Autonoma de Barcelona, Barcelona, Spain a r t i c l e i n f o a b s t r a c t Article history: Background: At present, because of indirect effects derived from routine childhood immunisation, clinical Received 1 June 2022 benefits vaccinating adults with the 23-valent pneumococcal polysaccharide vaccine (PPsV23) and/or the Received in revised form 18 January 2023 13-valent pneumococcal conjugate vaccine (PCV13) are uncertain. This study investigated clinical effec- Accepted 18 January 2023 tiveness for both PPsV23/PCV13 in preventing pneumonia among Catalonian adults during an earlier 2- Available online 21 January 2023 year period post-PCV13 free (publicly funded) approval for infants. Methods: We conducted a Population-based cohort study involving 2,059,645 adults  50 years in Keywords: Catalonia, Spain, who were followed between 01/01/2017–31/12/2018. Primary outcomes were hospital- Adults Effectiveness isation from pneumococcal pneumonia (PP) or all-cause pneumonia (ACP) and main explanatory variable Pneumococcal vaccination was PCV13/PPsV23 vaccination status. Cox regression models were used to estimate vaccination effec- Pneumococcal conjugate vaccine tiveness adjusted by age/sex and underlying-risk conditions. Pneumococcal polysaccharide vaccine Results: Cohort members were followed for 3,958,528 person-years (32,328 PCV13-vaccinated, 1,532,186 Pneumonia PPsV23-vaccinated), observing 3592 PP (131 in PCV13-vaccinated vs 2476 in PPsV23-vaccinated) and 24,136 ACP (876 in PCV13-vaccinated vs 17,550 in PPsV23-vaccinated). Incidence rates (per 100,000 person-years) were 90.7 for PP (394.2 in PCV13-vaccinated vs 161.6 in PPsV23-vaccinated) and 609.7 for ACP (2636.3 in PCV13-vaccinated vs 1145.4 in PPsV23-vaccinated). The PCV13 was associated with an increased risk of PP (hazard ratio [HR]: 1.24; 95% CI: 1.00–1.52; p = 0.046) and ACP (HR: 1.38; 95% CI: 1.28–1.49; p < 0.001) whereas the PPsV23 did not alter the risk of PP (HR: 1.07; 95% CI: 0.98–1.18; p = 0.153) and slightly increased the risk of ACP (HR: 1.14; 95% CI: 1.10–1.18; p < 0.001). In supplemen- tary analyses focused on at-risk individuals (i.e., elderly persons, immunocompromissing and other chronic illnesses) protective effects of vaccination did not emerge either. Conclusions: Data does not support clinical benefits from pneumococcal vaccination (nor PCV13 neither PPsV23) against pneumonia among Catalonian middle-aged and older adults in the current era of univer- sal PCV13 childhood immunisation in our setting. New extended valency PCVs are greatly needed. Ó 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Introduction major public health problem worldwide. Pneumococcal disease has a bimodal distribution, with a high burden among children To date, despite successive formulations of anti-pneumococcal under 5 years and adults over 50 years. In 2017, as well as 9,600 vaccines for adults and children during the past decades, infections deaths by pneumococcal meningitis, pneumococcus caused caused by Streptococcus pneumoniae, mainly invasive pneumococ- approximately 659,000 low-respiratory infection deaths among cal disease (IPD) and pneumococcal pneumonia (PP) remain a people over 50 around the world.[1]. The existence of more than 90 pneumococcal serotypes (differ- ⇑ Corresponding author at: Institut Catalá de la Salut, Rambla Nova 124, D, 1°A, ing in their immunogenicity, virulence and epidemiological distri- 43001 Tarragona, Spain. bution in distinct geographical areas),[2] together with the E-mail address: mcdiego.tgn.ics@gencat.cat (C. de Diego-Cabanes). observed phenomenon of serotype replacement after the introduc- https://doi.org/10.1016/j.jvacx.2023.100264 2590-1362/Ó 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 tion of conjugate vaccines, has largely complicated the develop- from the PHCC, death, or until the end of the two-year follow-up ment of a fully effective anti-pneumococcal vaccination strategy. (31/12/2018). The study was approved by the ethical committee [1,3]. of the Institution (ethic committee IDIAP Jordi Gol, file 20/065- At present, apart from childhood immunisation, two anti- PCV) and was conducted in accordance with the general principles pneumococcal vaccines have been available for use in adults: the for observational studies.[20] Considering population-based and 23-valent pneumococcal polysaccharide vaccine (PPSV23, classi- non-interventional design, an individual consent for all study par- cally recommended for high-risk individuals and elderly people ticipants (n = 2,059,645) was exempt. Data analyses were anon- since the 1990s) and the 13-valent protein-polysaccharide conju- ymised and risk of identification was null. gate vaccine (PCV13, initially marketed to replace PCV7 for child- hood immunisation in 2010 and licensed for use in adults since Data sources 2012).[3] A major advantage for the PCV13 would be its theoretical higher immunogenicity, but, it has a lower serotype coverage than To establish baseline characteristics (vaccinations, comorbidi- the PPSV23.[4]. ties and underlying risk conditions) of the cohort at study start, Currently, pneumococcal vaccination (using distinct PPSV23/ as well as to identify vaccinations after study started, we used PCV13 schedules) for at-risk adults and elderly people is recom- the information system for the development of research in primary mended in many countries.[1,3,5] However, diferent studies and care of Catalonia (SIDIAP),[21] which compiles administrative data meta-analyses estimating clinical effectiveness of PPSV23 and and clinical information contained in the ICS’s electronic PHCC PCV13 in adults in recent years have revealed heterogeneous medical records system and is usually used for epidemiological results,[6–12] and the cost-effectiveness of vaccinating adults with studies in the region. Quality criteria for clinical data of the SIDIAP PCV13 and/or PPSV23 remains unclear (especially if we consider research database has been reported elsewhere.[22]. possible indirect effects derived from routine pediatric PCV13 To identify study events (hospitalisations from pneumococcal use).[13–15]. and all-cause pneumonia) that occurred in the study population In Catalonia, a region in Northeastern Spain with 7.6 million throughout the study period, we used the national surveillance people, the PPSV23 has been recommended (and publicly funded) system for hospital discharge data (CMBD, Conjunto Mínimo for all persons over 60 years (with or without risk conditions) and Básico de Datos), which is maintained by the Spanish Ministry of individuals under 60 years with at-risk conditions since 1999. The Health and covers an estimated 99.5% of the total Spanish popula- PCV13 (available for use in adults since 2012) is recommended for tion (covered in the National Health Care System by compulsory the same adult target population, although it is publicly funded health insurance).[23] In the present study we used CMBD hospital only for high-risk individuals (basically immunocompromised discharge codes, coded according to the International Classification patients). In children, a publicly funded vaccination programme of Diseases, 10th Revision (ICD-10), reported during 2017–2018 offering a free PCV13 for all infants  2 years began in late 2016; from 68 Catalonian hospitals. before this date, the PCV7/PCV13 had been also used in children (without public funding, except for immunocompromised chil- Outcomes dren).[16] PPSV23 and PCV13 effectiveness among adults was eval- uated by our research team in a prior study performed during Pneumococcal pneumonia (ICD-10 code J13), pneumonia by 2015–2016.[10] The present study is aimed to update information other microorganisms (codes J11.0, J12, J14-J17) and pneumonia about the clinical effectiveness of adult vaccination after free by unidentified/unspecified microorganism (code J18) were PCV13 pediatric approval. Concretely, we assessed PPSV23 and defined on the basis of hospital discharge ICD-10 codes (any listed PCV13 effectiveness in preventing hospitalised pneumonia (pneu- position) reported by the CMBD in hospitalisations occurring mococcal and all-cause), death from pneumonia and death from among cohort members from 01/01/2017 to 31/12/2018. Bac- any cause among Catalonian people over 50 years-old throughout teremic PP was considered in those cases with ICD-10 code J13 the 2017–2018 period (early 2-year period post-PCV13 free imple- plus A40.3 or B95.3. Death from any cause was considered accord- mentation in infants). ing to administrative data (vital status), which is periodically updated in the SIDIAP data base. Death from pneumonia (case- fatality) was considered when the patient died within 30-day after Methods pneumonia diagnosis. Design, setting and study population Vaccination status This is a closed population-based retrospective cohort study PPSV23 and PCV13 vaccination status were determined accord- involving 2,059,645 Catalonian middle-aged and older adults fol- ing to data registered in the PHCCs’ electronic clinical records lowed between 01/01/2017–31/12/2018. Cohort members were which contain specially designated fields for pneumococcal and all persons 50 years or older (birth date before 01/01/1967) who influenza vaccinations (virtually all of them are administered at were assigned to the 274 Primary Health Care Centers (PHCCs) the PHCCs in the Spanish Health System). At baseline, cohort mem- managed by the Catalonian Health Institute (ICS, Institut Catala bers were classified as PCV13 and/or PPSV23 vaccinated if they had de la Salut) on January 1, 2017 (date of study start). received at least one dose of the vaccine before the study started. In the study region (Catalonia, Spain) there are 358 PCCs, of Throughout the study period, PCV13/PPSV23 vaccination status which 274 (76.5%) are managed by the ICS and 84 are managed was a time-varying condition since some individuals received the by other providers. The analysed cohort (n = 2,059,645 vaccine after the study started. Subjects were considered to be vac- persons  50 years) represented a 72.6% of the total 2,838,002 Cat- cinated 14 days after vaccine administration. A subject was consid- alonian inhabitants in this age strata according to census data in ered as unvaccinated when a vaccination was not recorded. January 2017.[17] At the start of the study, antipneumococcal vac- cination uptakes among at-risk/older adults were approximately Covariates 60% for the PPSV23 and 1% for the PCV13 (48% in children).[18,19]. For this report, cohort members were followed since the begin- Baseline covariates were age, sex, influenza vaccination in prior ning of the study until the occurrence of any event, disenrollment autumn, history of hospitalisation for pneumococcal disease or 2 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 pneumonia in the previous 2 years, history of chronic respiratory and, therefore, the analyses do not include multiple events per per- disease, chronic heart disease, chronic liver disease, diabetes mel- son. To compare baseline characteristics of study subjects accord- litus, current smoking, alcoholism, and immunological situation. ing to their PPSV23 and PCV13 vaccination status we used Chi- Immunocompromise was defined by the presence of any one of squared or Fisher’s test as appropriate. Cox regression models for the following: asplenia, immunodeficiency/HIV-infection, severe time-varying covariables were used to estimate hazard ratios chronic renal disease, bone marrow transplantation, cancer (solid (HRs) and evaluate the association between having received organ or haematological neoplasia) and/or immunosuppressive PPSV23/PCV13 and the time of the first outcome during the study medication. Definitions used to identify comorbidities/underlying period. The PPSV23 and the PCV13 vaccination status were consid- conditions were listed in an appendix elsewhere.[10]. ered time-varying conditions (e.g., subjects vaccinated after study started), whereas the other covariables were defined at study entry. All above mentioned covariables (i.e., age, sex, vaccinations’ Statistical analyses history, and presence of comorbidities/underlying risk conditions) were initially considered potential candidates for the calculation of Incidence rates (IRs) were calculated as person-years, consider- multivariable Cox models. The method to select a subset of covari- ing that the numerator was the number of events and the denom- ates to include in the final models was the purposeful selection. inator was the sum of the persons-time contributed to each cohort [24] The proportional hazard assumptions were assessed by adding member during the study period. Only a first episode of hospitali- the covariate by log-time interactions to the model. Both PCV13/ sation from pneumonia during the study period was considered Table 1 Baseline characteristics of 2,059,645 cohort members according to their PCV13 and PPSV23 vaccination status before the study started (31/12/2016). PCV13 statusa PPSV23 statusb VACCINATED UNVACCINATED VACCINATED UNVACCINATED N=13,917 N=2,045,728 N=798,548 N=1,261,097 n (%) n (%) n (%) n (%) Age 50-59 years 2720 (19.5) 738,216 (36.1) 31,953 (4.0) 708,983 (56.2) 60-69 years 3814 (27.4) 571,322 (27.9) 201,392 (25.2) 373,744 (29.6) 70-79 years 3994 (28.7) 409,285 (20.0) 297,072 (37.2) 116,207 (9.2) 80 years 3389 (24.4) 326,905 (16.0) 268,131 (33.6) 62,163 (4.9) Sex men 7643 (54.9) 943,368 (46.1) 356,149 (44.6) 594,862 (47.2) women 6274 (45.1) 1,102,360 (53.9) 442,399 (55.4) 666,235 (52.8) Hospitalization history (previous 2 yrs) Invasive pneumococcal disease 50 (0.4) 682 (0.0) 478 (0.1) 254 (0.0) Pneumococcal disease 186 (1.3) 2740 (0.1) 2055 (0.3) 871 (0.1) Pneumococcal pneumonia 160 (1.1) 2321 (0.1) 1755 (0.2) 726 (0.1) All cause pneumonia 746 (5.4) 16,021 (0.8) 12,294 (1.5) 4473 (0.4) Vaccination history PPSV23 (at any time) 11,291 (81.1) 787,257 (38.5) - - PCV13 (in previous 5 yrs) - - 11,291 (1.4) 2626 (0.2) Flu vaccine (in prior autumn) 10,649 (76.5) 641,972 (31.4) 548,638 (68.7) 103,983 (8.2) Risk level High risk 6779 (48.7) 196,668 (9.6) 122,606 (15.4) 80,841 (6.4) At-risk 4763 (34.2) 796,229 (38.9) 374,802 (46.9) 426,190 (33.8) Healthy 2375 (17.1) 1,052,831 (51.5) 301,140 (37.7) 754,066 (59.8) Chronic respiratory disease 3798 (27.3) 196,898 (9.6) 134,005 (16.8) 66,691 (5.3) Chronic heart disease 3349 (24.1) 216,333 (10.6) 156,673 (19.6) 63,009 (5.0) Diabetes mellitus 4245 (30.5) 347,042 (17.0) 242,761 (30.4) 108,526 (8.6) Chronic liver disease 986 (7.1) 39,937 (2.0) 19,208 (2.4) 21,715 (1.7) Chronic renal disease 1612 (11.6) 20,924 (1.0) 17,657 (2.2) 4879 (0.4) Alcoholism 505 (3.6) 64,513 (3.2) 23,047 (2.9) 41,971 (3.3) Current smoking 1700 (12.2) 342,840 (16.8) 69,091 (8.7) 275,449 (21.8) Asplenia 92 (0.7) 270 (0.0) 190 (0.0) 172 (0.0) Primary immunodeficiency 124 (0.9) 674 (0.0) 459 (0.1) 339 (0.0) HIV infection 820 (5.9) 3066 (0.1) 1950 (0.2) 1936 (0.2) Bone marrow transplantation 1558 (11.2) 4597 (0.2) 3893 (0.5) 2262 (0.2) Haematological neoplasia 395 (2.8) 6582 (0.3) 4051 (0.5) 2926 (0.2) Solid neoplasia 1884 (13.5) 109,491 (5.4) 62,946 (7.9) 48,429 (3.8) Immunosuppressive treatment 3308 (23.8) 78,997 (3.9) 50,964 (6.4) 31,341 (2.5) Immunocompromisec 6779 (48.7) 196,668 (9.6) 122,606 (15.4) 80,841 (6.4) NOTE: PCV13, 13 valent pneumococcal conjugate vaccine; PPSV23, 23 valent pneumococcal polysaccharide vaccine. a Comparing PCV13 vaccinated vs unvaccinated, p-values (chi-squared test) were always statistically significant at p < 0.001, except for alcoholism (p = 0.142). b Comparing PPSV23 vaccinated vs unvaccinated, p-values (chi-squared test) were always statistically significant at p < 0.001. c Immunocompromise was a composite variable defined by the presence of any one of the following: cancer (solid organ or haematological neoplasia), chronic severe nephropathy (nephrotic syndrome, renal failure, dialysis or transplantation), anatomical or functional asplenia, immunodeficiency (including AIDS), and long-term corti- costeroid therapy (20 mg/day of prednisone) or another immunosuppressive medication. 3 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 PPSV23, influenza vaccination status, history of prior pneumococ- nated and 33,228 were PCV13 vaccinated. Considering the PPSV23 cal disease and immunocompromising conditions were considered status, 798,548 individuals had received PPSV23 before study start as epidemiologically relevant covariates, being included in all the (contributing to the analyses with 1,500,833 person-years as final models. All models were compared by the partial likelihood PPSV23 vaccinated) and 41,745 individuals received PPSV23 later ratio test and Akaike information criterion. Besides the main anal- (contributing to the analyses with 51,392 person-years as PPSV23 ysis including the total study cohort, we performed supplementary unvaccinated and 31,353 person-years as PPSV23 vaccinated). As analyses by age subgroups, immunological status and four specific for the PCV13 status, 13,917 individuals had received PCV13 before at-risk conditions (patients with chronic respiratory disease, study start (contributing to the analyses with 25,565 person-years chronic heart disease, diabetes mellitus and smokers). All results as PCV13 vaccinated) and 7,769 individuals received PCV13 later were expressed with 95% confidence intervals (CIs). Statistical sig- (contributing to the analyses with 7,419 person-years as PCV13 nificance was set at p < 0.05 (two-tailed). The analyses were per- unvaccinated and 7,664 person-years as PCV13 vaccinated). The formed using IBM SPSS Statistics for Windows, version 24 (IBM vast majority (81.1%) of PCV13 vaccinated subjects had dual vacci- Corp., Armonk, N.Y., USA). nation (PCV13 + PPSV23). At baseline, mean age of cohort members was 66 years-old (s- Results tandard deviation: 11.4), 951,011 (46.2%) men and 1,108,634 (53.8%) women. Considering comorbidities/underlying risk condi- The 2,059,645 cohort members were observed for a total of tions, 351,287 (17.1%) cohort members had diabetes mellitus, 3,958,528 person-years, of which 1,532,186 were PPSV23 vacci- 344,540 (16.7%) were smokers, 219,682 (10.7%) had chronic heart Table 2 Incidence and risk of hospitalisation from pneumococcal and all-cause pneumonia in relation to PCV13 and PPSV23 vaccination status in the total study population (N = 2,059,645). Hospitalisation from pneumonia Parameter Pneumococcal Other microorganisms Unknown ethiology All-cause Bacteremic PP All PP Number of events: All people 472 3592 4145 15399 24,136 PCV13 vaccinated 23 131 193 552 876 PCV13 unvaccinated 449 3461 4952 14847 23,260 PPSV23 vaccinated 273 2476 3656 11418 17,550 PPSV23 unvaccinated 199 1116 1489 3981 6586 Overall IR 11.9 60.7 130.0 389.0 609.7 (95% CI) (10.8-13.1) (85.2-96.5) (122.1-138.3) (365.3-413.9) (572.5-648.7) PCV13-vaccinated IR 69.2 394.2 580.8 1661.2 2636.3 (95% CI) (43.9-103.8) (332.7-466.7) (504.1-668.5) (1531.6-1800.7) (2467.6-2815.6) PCV13-unvaccinated IR 11.4 88.2 126.2 378.2 592.6 (95% CI) (10.4-12.5) (82.8-93.8) (118.5-134.3) (355.1-402.4) (556.5-630.5) PPSV23-vaccinated IR 17.8 161.6 238.6 745.2 1145.4 (95% CI) (15.7-20.2) (151.7-171.9) (224.0-253.9) (699.7-792.9) (1075.5-1218.7) PPSV23-unvaccinated IR 8.2 46.0 61.4 164.1 271.4 (95% CI) (7.1-9.4) (43.2-48.9) (57.7-65.3) (154.1-174.6) (254.8-288.8) Unadjusted HR For PCV13 4.65 3.63 4.20 4.09 4.04 (95% CI) (2.90-7.45) (2.99-4.40) (3.61-4.89) (3.74-4.46) (3.76-4.34) p value <0.001 <0.001 <0.001 <0.001 <0.001 For PPSV23 2.02 3.40 3.75 4.40 4.09 (95% CI) (1.69-2.43) (3.16-3.64) (3.53-3.98) (4.25-4.57) (3.97-4.20) p value <0.001 <0.001 <0.001 <0.001 <0.001 Age and sex-adjusted HR For PCV13 3.60 2.55 2.80 2.70 2.71 (95% CI) (2.24-5.79) (2.10-3.10) (2.40-3.26) (2.47-2.95) (2.52-2.90) p value <0.001 <0.001 <0.001 <0.001 <0.001 For PPSV23 1.11 1.40 1.61 1.55 1.53 (95% CI) (0.89-1.39) (1.29-1.52) (1.50-1.73) (1.49-1.61) (1.48-1.58) p value 0.361 <0.001 <0.001 <0.001 <0.001 Multivariable-adjusted HR For PCV13 1.72 1.24 1.28 1.45 1.38 (95% CI) (1.03-2.86) (1.00-1.52) (1.09-1.51) (1.32-1.60) (1.28-1.49) p value 0.038 0.046 <0.001 <0.001 <0.001 For PPSV23 0.78 1.07 1.21 1.14 1.14 (95% CI) (0.60-1.01) (0.98-1.18) (1.12-1.31) (1.09-1.19) (1.10-1.18) p value 0.060 0.153 <0.001 <0.001 <0.001 NOTE: PP, pneumococcal pneumonia. IR, unadjusted incidence rate per 100,000 person-years. CI, confidence interval. PCV13, 13-valent pneumococcal conjugate vaccine. PPSV23, 23-valent pneumococcal polysaccharide vaccine. HRs (hazard ratios) are for vaccinated subjects as compared with unvaccinated subjects and were adjusted, where appropriate, for age, sex, history of pneumococcal disease or pneumonia during the previous 24 months, presence of chronic pulmonary/respiratory disease, chronic heart disease, diabetes, chronic liver disease, alcoholism, current smoking, asplenia, immunodeficiency, HIV infection, chronic renal disease, bone marrow transplantation, cancer, immunosuppressive therapy, history of pneumococcal polysaccharide vaccination at any time and receipt or non receipt of influenza vaccine in prior autumn. . 4 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 Table 3 Incidence and risk of death from pneumonia and death from any cause in relation to PCV13 and PPSV23 vaccination status in the total study population (N = 2,059,645). Death from pneumonia Death from Pneumococcal All-cause Parameter any cause Number of events: All people 272 2480 83,440 PCV13 vaccinated 8 66 1520 PCV13 unvaccinated 264 2414 81,920 PPSV23 vaccinated 211 1862 61,476 PPSV23 unvaccinated 61 618 21,964 Overall IR 6.9 62.6 2107.9 (95% CI) (6.1-7.8) (58.8-66.6) (1979.3-2242.8) PCV13-vaccinated IR 24.1 198.6 4574.5 (95% CI) (10.4-47.5) (155.9-252.2) (4295.5-4867.3) PCV13-unvaccinated IR 6.7 61.5 2087.0 (95% CI) (5.9-7.6) (57.7-65.4) (1959.7-2220.5) PPSV23-vaccinated IR 13.8 121.5 4012.3 (95% CI) (12.0-15.9) (114.1-129.3) (3767.6-4269.1) PPSV23-unvaccinated IR 2.5 25.5 905.2 (95% CI) (1.9-3.3) (23.5-27.6) (850.0-963.1) Unadjusted HR For PCV13 3.66 3.31 2.20 (95% CI) (1.81-7.40) (2.59-4.22) (2.10-2.32) p value <0.001 <0.001 <0.001 For PPSV23 5.47 4.77 4.43 (95% CI) (4.11-7.27) (4.35-5.22) (4.36-4.50) p value <0.001 <0.001 <0.001 Age and sex-adjusted HR For PCV13 2.39 2.24 1.58 (95% CI) (1.18-4.84) (1.75-2.86) (1.50-1.66) p value 0.015 <0.001 <0.001 For PPSV23 1.48 1.11 1.06 (95% CI) (1.08-2.01) (1.01-1.23) (1.04-1.08) p value 0.014 0.033 <0.001 Multivariable-adjusted HR For PCV13 1.43 1.24 1.03 (95% CI) (0.68-3.00) (0.96-1.60) (0.98-1.09) p value 0.344 0.105 0.275 For PPSV23 1.12 0.91 0.97 (95% CI) (0.78-1.60) (0.81-1.02) (0.95-0.99) p value 0.545 0.097 0.002 NOTE: IR, unadjusted incidence rate per 100,000 person-years. CI, confidence interval. PCV13, 13-valent pneumococcal conjugate vaccine. PPSV23, 23-valent pneumococcal polysaccharide vaccine. HRs (hazard ratios) are for vaccinated subjects as compared with unvaccinated subjects and were adjusted, where appropriate, for age, sex, history of pneumococcal disease or pneumonia during the previous 24 months, presence of chronic pulmonary/respiratory disease, chronic heart disease, diabetes, chronic liver disease, alcoholism, current smoking, asplenia, immunodeficiency, HIV infection, chronic renal disease, bone marrow transplantation, cancer, immunosuppressive therapy, history of pneumococcal polysaccharide vaccination at any time and receipt or non receipt of influenza vaccine in prior autumn. disease, 200,696 (9.7%) chronic respiratory disease, 118,352 (5.7%) in PCV13 vaccinated), 130.0 (95% CI: 122.1–138.3) for pneumonia cancer diagnosed in prior 5 years (6977 haematological neoplasia due to other microorganisms (238.6 in PPSV23 vaccinated and and 111,375 solid cancer), 65,018 (3.2%) alcoholism, 40,923 (2%) 580.8 in PCV13 vaccinated), 389.0 (95% CI: 365.3–413.9) for pneu- severe liver disease, 22,536 (1.1%) severe renal disease and 4684 monia due to unknown ethiology (745.2 in PPSV23 vaccinated and (0.2%) immunodeficiency/HIV-infection. 1661.2 in PCV13 vaccinated) and 609.7 (95% CI: 572.5–648.7) for Table 1 shows baseline characteristics of the 2,059,645 cohort all-cause pneumonia (1145.4 in PPSV23 vaccinated and 2636.3 in members according to their PCV13 and PPSV23 vaccination status PCV13 vaccinated). at study start. To summarize, vaccinated persons were older (espe- Table 2 shows number of events, IRs, unadjusted, age and sex- cially PPSV23 vaccinated), had more comorbidities/underlying adjusted and multivariable-adjusted risks of pneumococcal, other conditions (especially PCV13 vaccinated) and had greater influenza microorganisms, unknown ethiology and all-cause pneumonia in vaccine coverage than unvaccinated subjects. relation with PCV13 and PPSV23 vaccination status in the total During the two-year study period, 83,440 (4.1%) cohort mem- study cohort. In the unadjusted analyses, as well as in the age bers died and 51,175 (2.5%) moved or were lost subjects. A total and sex-adjusted analyses, both PCV13 and PPSV23 were associ- of 24,136 cohort members suffered a first episode of hospitalisa- ated with an increased risk for all analysed outcomes. In the mul- tion for all-cause pneumonia, of which 3592 were pneumococcal tivariable analyses, having received the PPSV23 did not pneumonia (472 bacteremic PP cases), 5145 were due to other significantly alter the risk of overall pneumococcal pneumonia microorganisms and 15,399 were due to unknown ethiology. (HR: 1.07; 95% CI: 0.98–1.18; p = 0.153) and slightly increased IRs (per 100,000 person-years) were 90.7 (95% CI: 85.2–96.5) for the risk of all-cause pneumonia (HR: 1.14; 95% CI: 1.10–1.18; pneumococcal pneumonia (161.6 in PPSV23 vaccinated and 394.2 p < 0.001). Considering the PCV13, it appeared significantly associ- 5 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 Table 4 Stratified analyses on PPSV23 vaccination effectiveness according to age subgroups, immunological situation and distinct at-risk conditions. Age Strata Immunological Status At-risk conditions 50-59 years 60 years Compromised Competent CRD CHD DM Smoking (N=740,936) (N=1,318,709) (N=203,447) (N=1,856,198) (N=200,696) (N=219,682) (N=351,287) (N=344,540) Bacteremic PP Number of events 98 374 126 346 150 105 133 123 PPSV23 multivariable HR 1.08 0.74 0.88 0.72 1.01 0.88 0.79 0.55 (95% CI) (0.52-2.26) (0.56-0.97) (0.54-1.43) (0.53-0.98) (0.63-1.62) (0.49-1.56) (0.50-1.27) (0.32-0.95) p value 0.841 0.031 0.609 0.035 0.981 0.652 0.336 0.031 All PP Number of events 471 3121 877 2715 1169 885 1051 604 PPSV23 multivariable HR 1.49 1.05 1.26 0.98 1.18 1.32 1.02 1.08 (95% CI) (1.09-2.03) (0.95-1.16) (1.04-1.52) (0.88-1.10) (0.98-1.42) (1.07-1.64) (0.85-1.22) (0.86-1.35) p value 0.012 0.375 0.019 0.747 0.078 0.011 0.830 0.500 All-cause pneumonia Number of events 2557 21,579 6265 17,871 7735 7000 7939 3558 PPSV23 multivariable HR 1.49 1.13 1.18 1.08 1.21 1.20 1.08 1.07 (95% CI) (1.30-1.69) (1.08-1.17) (1.09-1.27) (1.04-1.13) (1.13-1.31) (1.11-1.30) (1.01-1.15) (0.97-1.17) p value <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.031 0.177 Death from PP Number of events 18 254 71 201 66 82 80 35 PPSV23 multivariable HR 2.97 1.06 2.20 0.89 2.73 1.17 1.22 1.44 (95% CI) (0.73-12.05) (0.73-1.54) (1.00-4.87) (0.60-1.34) (1.02-7.35) (0.57-2.39) (0.59-2.51) (0.56-3.71) p value 0.128 0.757 0.051 0.586 0.047 0.666 0.597 0.445 Death from pneumonia Number of events 124 2356 730 1750 681 845 821 311 PPSV23 multivariable HR 1.37 0.91 0.94 0.86 0.86 1.05 0.90 0.81 (95% CI) (0.76-2.49) (0.81-1.02) (0.76-1.16) (0.76-0.99) (0.68-1.09) (0.85-1.30) (0.73-1.11) (0.60-1.10) p value 0.298 0.120 0.556 0.031 0.215 0.690 0.334 0.176 Death from any cause Number of events 5214 78,226 25,751 57,689 16,369 25,551 26,958 9045 PPSV23 multivariable HR 1.38 1.00 0.93 0.95 1.05 1.02 1.00 1.03 (95% CI) (1.25-1.52) (0.98-1.02) (0.90-0.96) (0.93-0.98) (1.00-1.10) (0.98-1.06) (0.97-1.04) (0.97-1.09) p value <0.001 0.980 <0.001 <0.001 0.055 0.363 0.956 0.371 Note: PP denotes pneumococcal pneumonia (any serotype). CRD: chronic respiratory disease; CHD: chronic heart disease; DM: diabetes mellitus. The HRs (hazard ratios) are for PPSV23 vaccinated subjects as compared with PPSV23 nonvaccinated and were adjusted, where appropriate, for age, sex, history of pneumococcal disease or pneumonia during the previous 24 months, presence of chronic pulmonary/respiratory disease, chronic heart disease, diabetes, chronic liver disease, alcoholism, current smoking, asplenia, immunodeficiency, HIV infection, chronic renal disease, bone marrow transplantation, cancer, immunosuppressive therapy, history of pneumococcal polysaccharide vaccination at any time and receipt or non receipt of influenza vaccine in prior autumn. CI denotes confidence interval. ated with an increased risk for both pneumococcal pneumonia population subgroups (with HRs ranging between 1.07 and 1.49 for (HR: 1.24; 95% CI: 1.00–1.52; p = 0.046) and all-cause pneumonia the PPSV23 and 1.11–1.41 for the PCV13). (HR: 1.38; 95% CI: 1.28–1.49; p < 0.001). In pneumococcal pneumonia, crude case-fatality rate was 7.6% Discussion [272/3592] overall, 6.1% [8/131] in PCV13 vaccinated and 8.5% [211/2476] in PPSV23 vaccinated (p = 0,343). With all-cause pneu- At present, after PCV13 introduction for infants, benefits from monia, crude case-fatality rate was 10.3% [2480/24,147] overall, using PCV13 (and also PPSV23) in adults are uncertain.[13–15] 7.5% [66/876] in PCV13 vaccinated and 10.6% [1852/17,550] in The present study assessed clinical effectiveness and public health PPSV23 vaccinated (p = 0.004). Considering all-cause death, mor- impact of PCV13/PPSV23 vaccinations in the general adult popula- tality rate was 2042 deaths per 100,000 person-years in the total tion over 50 years in Catalonia, Spain, throughout 2017–2018 study cohort, 4795 per 100,000 in PCV13 vaccinated and 3725 (early two-year period after free PCV13 approval for infants). Data per 100,000 in PPSV23 vaccinated. After multivariable- provided here updates data reported during 2015–2016 (before adjustements, pneumococcal vaccination (neither PCV13 nor free PCV13 approval).[10]. PPSV23) did not significantly alter the risk of death from pneumo- As its main findings, pneumococcal vaccination has not nia (pneumococcal and/or all-cause) while PPSV23 was associated emerged effective (neither PCV13 nor PPSV23) in preventing hospi- with a little reduction risk of all-cause death (HR: 0.97; 95% CI: talised pneumonia (pneumococcal or all-cause) and death from 0.95–0.99; p = 0.002) (Table 3). pneumonia in the total study cohort. In stratified analyses, the Supplementary analyses focused on age subgroups (under/over PPSV23 was associated with a marginally significant reduction risk 60 years), immunocompetent/immunocompromised subjects and of IPD (bacteremic PP) in the age subgroup over 60 years (where specific at-risk comorbidity subgroups are shown in Table 4 (for vaccine is recommended) and in immunocompetent persons. How- the PPSV23) and Table 5 (for the PCV13). The PPSV23 was associ- ever, pneumococcal vaccination did not emerge effective (neither ated with a marginally significant reduction risk of bacteremic PPSV23 nor PCV13) against all pneumococcal pneumonia and/or pneumococcal pneumonia in the age subgroup >=60 years (HR: all-cause pneumonia in specific at-risk population subgroups 0.74; 95% CI: 0.56–0.97; p = 0.031) and immunocompetent persons where vaccination is also recommended (i.e, immunocompromised (HR: 0.72; 95% CI: 0.53–0.98; p = 0.035), but no benefits emerged subjects, chronic respiratory disease, heart disease or diabetes). against all pneumococcal pneumonia and/or all-cause pneumonia. Considering pneumococcal pneumonia, our results are essen- Considering the PCV13, clinical benefits did not emerge either. tially similar to data observed during the 2015–2016 period and both PCV13 and PPSV23 were associated with an increased supports that vaccinating adults has not public health impact in multivariable-adjusted risk of all-cause pneumonia in all analysed reducing hospitalised pneumococcal pneumonia (all serotypes) in 6 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 Table 5 Stratified analyses on PCV13 vaccination effectiveness according to age subgroups, immunological situation and distinct at-risk conditions. Age Strata Immunological Status At-risk conditions 50-59 years 60 years Compromised Competent CRD CHD DM Smoking (N=740,936) (N=1,318,709) (N=203,447) (N=1,856,198) (N=200,696) (N=219,682) (N=351,287) (N=344,540) Bacteremic PP Number of events 98 374 126 346 150 105 133 123 PCV13 multivariable HR 0.73 1.99 1.40 2.06 1.77 1.37 2.16 NA (95% CI) (0.16-3.39) (1.16-3.39) (0.71-2.75) (0.97-4.40) (0.87-3.62) (0.53-3.59) (0.98-4.77) (-) p value 0.687 0.012 0.326 0.061 0.117 0.518 0.056 - All PP Number of events 471 3121 877 2715 1169 885 1051 604 PCV13 multivariable HR 1.20 1.22 1.09 1.39 1.12 1.09 1.08 0.75 (95% CI) (0.68-2.12) (0.97-1.52) (0.83-1.45) (1.03-1.88) (0.83-1.51) (0.74-1.58) (0.75-1.54) (0.40-1.41) p value 0.528 0.084 0.537 0.031 0.474 0.672 0.688 0.366 All-cause pneumonia Number of events 2557 21,579 6265 17,871 7735 7000 7939 3558 PCV13 multivariable HR 1.11 1.40 1.30 1.41 1.32 1.33 1.30 1.16 (95% CI) (0.86-1.43) (1.29-1.51) (1.18-1.44) (1.26-1.58) (1.18-1.47) (1.17-1.51) (1.15-1.47) (0.92-1.46) p value 0.415 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.212 Death from PP Number of events 18 254 71 201 66 82 80 35 PCV13 multivariable HR 1.93 1.28 1.57 0.83 0.95 2.13 0.35 2.66 (95% CI) (0.26-14.57) (0.56-2.95) (0.63-3.89) (0.21-3.38) (0.27-3.33) (0.76-5.92) (0.05-2.70) (0.49-14.34) p value 0.523 0.561 0.334 0.799 0.933 0.149 0.312 0.255 Death from any cause Number of events 124 2356 730 1750 681 845 821 311 PCV13 multivariable HR 0.97 1.25 1.36 0.97 1.35 1.17 1.01 1.60 (95% CI) (0.34-2.78) (0.95-1.62) (0.99-1.87) (0.63-1.52) (0.92-1.98) (0.79-1.73) (0.66-1.55) (0.78-3.27) p value 0.956 0.107 0.061 0.908 0.121 0.437 0.979 0.201 Death from any cause Number of events 5214 78,226 25,751 57,689 16,369 25,551 26,958 9045 PCV13 multivariable HR 1.00 1.03 0.93 1.10 0.98 0.99 1.00 0.91 (95% CI) (0.81-1.23) (0.97-1.09) (0.86-0.99) (1.02-1.19) (0.89-1.07) (0.91-1.08) (0.92-1.08) (0.76-1.08) p value 0.999 0.314 0.031 0.018 0.637 0.864 0.933 0.283 Note: PP denotes pneumococcal pneumonia (any serotype). CRD: chronic respiratory disease; CHD: chronic heart disease; DM: diabetes mellitus. The HRs (hazard ratios) are for PCV13 vaccinated subjects as compared with PCV13 nonvaccinated and were adjusted, where appropriate, for age, sex, history of pneumococcal disease or pneumonia during the previous 24 months, presence of chronic pulmonary/respiratory disease, chronic heart disease, diabetes, chronic liver disease, alcoholism, current smoking, asplenia, immunodeficiency, HIV infection, chronic renal disease, bone marrow transplantation, cancer, immunosuppressive therapy, history of pneumococcal polysaccharide vaccination at any time and receipt or non receipt of influenza vaccine in prior autumn. NA denotes non available HR (zero cases in vaccinated).CI denotes confidence interval. the current era of universal PCV13 infants’ vaccination in our set- effectiveness estimates (multivariable HRs) because it could sim- ting. In addition, data alerts about a posible increasing risk of all- ply reflect that the immunocompromised PCV13 vaccinated popu- cause pneumonia (non-vaccine pneumococcal serotypes, other lation had an enormously increased a priori risk of pneumococcal microorganisms and/or unknown ethiology) among vaccinated disease outcomes, and that PCV13 could not overcome this excess subjects. risk. In this cohort study, crude IRs of hospitalisation from all-type A lower risk for death from any cause was observed in certain pneumonia were largely higher in vaccinated than in nonvacci- analysis, such as the fully adjusted overall cohort and the immuno- nated persons, reflecting the baseline excess risk of vaccinated sub- competent adults (also oserved for death from all cause pneumo- jects who were older and had more comorbidities/risk conditions nia). This data could simply reflect baseline risk differences than unvaccinated (especially for the PCV13). We try to resolve between vaccinated and unvaccinated subjects, But It could also these baseline differences between vaccinated and unvaccinated be related with the fact that pneumococcal pneumonia increases subjects by multivariable analysis but, even after age/sex and the risk for subsequent cardiac events and all-cause mortality.[28]. underlying risk condition’s adjustments, both PCV13 and PPSV23 In this concern, there is insufficient evidence to draw conclu- remained significantly associated with an increased risk of all- sions on the impact of PCV13 or PPSV23 on mortality among older cause pneumonia (both in the total study cohort as well as in strat- adults.[1]. ified subgroups). During the past decades, likely related to differences in vaccina- Opposite other recent studies in USA elderly people,[25,26] tion practices and smoking, the role of pneumococcus as a causa- population benefits of PCV13 vaccination against all-type pneumo- tive microorganism in pneumonia cases in developed countries nia have not emerged in the present study. We note that no adjust- has clearly declined (representing currently less than 10–15% of ment method fully resolves confounding by indication in overall pneumonia cases in North America and Europe).[29,30] observational studies.[27] Regarding PCV13 (with very small vac- Therefore, the potential value of pneumococcal vaccination in cine coverage in our cohort since it was only funded for immuno- reducing all-cause pneumonia has also decreased, and it would compromised patients), caution is needed to interpret vaccines’ be relatively low at present. On this concern, pneumococcal 7 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 pneumonia represented 14.9% (3592/24,136) of overall pneumonia recommended in aged > 65 years) and all PCV13 vaccinated had cases in the present study, supporting that the role of pneumococ- received the vaccine in previous 5 years (within 2012–2016). cus as causative pathogen of pneumonia has also declined in recent We note that definition criteria of pneumococcal pneumonia years in our setting (where pneumococcus represented near 40% of vary between distinct studies, but we also note that the use of community-acquired pneumonia cases at the beginning of the pre- ICD diagnoses codes to define pneumococcal pneumonia, despite sent century).[31]. recognised limitations,[39] has been commonly used in many In the present study, data about pneumococcal serotypes was studies evaluating this issue. On this concern, we underline that not available (since it is not reflected in the Spanish CMBD system) all participating Catalonian hospitals basically apply similar diag- and, as consequent limitation, this study is not able to assess vac- noses checklist and treatment for patients with a clinical suspicion cination effectiveness against vaccine-type pneumococcal pneu- of pneumonia (which is established on the basis of an acute respi- monia (which would be the most specific outcome evaluating ratory illness, with evidence of a new infiltrate in a chest radio- vaccines efficacy). It must be noted that the value of PCV13/PPSV23 graph), being blood/sputum cultures and urinary antigen testing vaccinating adults against all pneumococcal pneumonia (any sero- used as conventional diagnostic workup performed according to type) has also decreased in recent years because indirect effects the attending physician.[40] As real world data study, vaccinations from routine PCV13 pediatric use.[32–37] In Spain, data from the in this study were more frequently prescribed for persons with National Surveillance Microbiology Laboratory of Pneumococcus comorbidities/underlying risk conditions and, as above noted, has revealed that only a 29.7% of overall IPD cases (mostly bac- may exists residual confounding in vaccines’ effectiveness estima- teremic pneumococcal pneumonia) were caused by PCV13 sero- tions despite multivariable adjustements. We recognise these lim- types during the 2017–2018 period.[33] In Catalonia, during the itations but, opposite to vaccine’s efficacy that must be assessed by same period, PCV13 and PPSV23 serotypes represented 30.4% and trials with controlled conditions, vaccination’s effectiveness must 70.5%, respectively, of overall IPD cases (they were 40.4% and be assessed by observational studies conducted in the real-world 73.6%during 2014–2015).[34]. practice conditions, as in the present study, where vaccines are In European PCV13 sites, several years after vaccine introduc- not homogeneously used. Caution is needed in extrapolating data tion, the incidence of IPD caused by PCV13 serotypes declined sub- to other geographical settings with distinct epidemiological condi- stantially in older adults (as indirect effect from childhood PCV13 tions (i.e., different vaccines’ use, vaccination uptakes among chil- vaccination) while the incidence of non-PCV13 serotypes increased dren and adults, prevalence of circulating serotypes, etc). 63%, resulting in a non-significant 9% reduction (95% CI: 4% to In conclusion, in this real world data study involving 2,059,645 19%) of IPD caused by all serotypes.[35] In the United States, where Catalonian adults  50 years followed throughout 2017–2018, PCV13/PPSV23 uptakes among children and older adults are rela- apart of a protective effect against IPD, clinical benefits of adults tively high, only 4.6% of overall pneumonia cases hospitalised vaccination in preventing hospitalised all-type pneumococcal between October 2013 and September 2016 in people aged pneumonia or all-cause pneumonia have not emerged, neither 18 years or more were caused by PCV13 serotypes (3.8–5.3% in for the PCV13 nor for the PPSV23. Our data supports that the cur- patients aged 18–64 years depending on their risk status and rent anti-pneumococcal vaccination strategy is insufficient to 4.2% in people aged 65 years or older).[36]. reduce pneumonia burden in at-risk and older adults in our setting. The new generation of higher-valent PCVs (PCV15, PCV20 and At present, when PCV13 childhood immunisation is working future PCV24) have the potential to reduce a substantial proportion and indirect effects have occurred (by reducing circulating PCV13 of pneumococcal disease cases among all age groups. However, strains in the population), the potential value vaccinating adults considering the repeated serotype-replacement phenomenon with the PCV13 (but also with the PPSV23 which shares twelve observed after each PCV introduction, a new technology pneumo- common serotypes) has clearly decreased. Vaccinating adults using coccal vaccine (with complete protection regardless of serotype) PCV13/PPSV23 could still provide clinical benefits in those settings remains desirable. where vaccine-type disease burden remains relatively high, but Major strengths in this study are the largest size of the study changes in adults vaccination strategies are necessary in settings cohort (which included more than 2 million adults over 50 years where universal infants vaccination is working and, therefore, cir- and represented almost 73% of the total Catalonian inhabitants in culating vaccine-type serotypes are low.[1] In this way, CDC has this age stratum), as well as the use of multivariable survival anal- recently approved new vaccinations’ recommendations criteria ysis methods to estimate accurately PCV13/PPSV23 effectiveness for at-risk and older adults, supporting the use of the new genera- adjusted by major known underlying risk conditions. To our tion of PCVs (PCV15 or better PCV20) in all age groups.[37]. knowledge, this is the largest cohort study evaluating pneumococ- Author contributions cal vaccination effectiveness in adults performed to date. Impor- AVC and OOG conceptualized and designed the study; AVC, CDC tantly, the study provides scarce population-based incidence data and ESG wrote and edited the manuscript; CDC, VTV, MFP and DRS and it has been able to assess vaccination effectiveness against assessed outcomes; CRC obtained data; ESG and AVR did statistical public health relevant outcomes (such as hospitalised all-type analyses; AVC coordinated the study. All authors have read and pneumococcal pneumonia, all-cause pneumonia and death from agreed to the final version of the manuscript. pneumonia) in distinct at-risk population subgroups of interest (e.g., chronic respiratory or cardiac disease, diabetics and smokers) Data availability where PPSV23/PCV13 effectiveness data is relatively rare in the lit- erature.[38]. Interested authors might obtain SIDIAP data (previous ethics and Major limitations in this study are related with its observational scientific approval by the ethics and clinical research committee nature; mainly, the non-randomised vaccination, the small PCV13 of the Primary Care Research Institute Jordi Gol (IDIAP Jordi Gol)). coverage and the absence of serotype data. We assumed that hos- pital discharge coding was correct, but a validation diagnosis was Declaration of Competing Interest not feasible considering study design and enormous size of the cohort. We did not consider time since vaccination in the analyses The authors declare that they have no known competing finan- because of the vast majority of PPSV23 vaccinated had received the cial interests or personal relationships that could have appeared vaccine more than 5 years ago (since revaccination is not routinely to influence the work reported in this paper. 8 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 Acknowledgements [17] Generalitat de Catalunya. Population on 1 January. By age group. Catalunya: Statistical Institute of Catalonia. Annual indicators. [accessed 2022 May 11]. https://www.idescat.cat/indicadors/?id=anuals&n=10329&col=1&lang=en This work was supported by a grant from the ‘‘Fondo de Inves- [18] Vila-Córcoles A, Ochoa-Gondar O, de Diego C, Satué E, Vila-Rovira A, Aragón M. tigación Sanitaria” of the ”Instituto de Salud Carlos III” (call 2020) Pneumococcal vaccination coverages by age, sex and specific underlying risk conditions among middle-aged and older adults in Catalonia, Spain, 2017. Euro for the ‘‘Acción Estratégica en Salud 2020/2030 (code file Surveill 2019;24(29):1800446. https://doi.org/10.2807/1560-7917. PI20/01223), cofinanced by the European Union through the ES.2019.24.29.1800446. ‘‘Fondo Europeo de Desarrollo Regional” (FEDER). [19] González R, Armadans L, Martínez X, Moraga F, Campins M.Pneumococcal vaccination coverage in at-risk children in Catalonia [in Spanish]. Enferm Infecc Microbiol Clin. 2015;33(9):597-602. Spanish. doi: 10.1016/j. eimc.2015.01.003. References [20] World Medical Association (WMA). Declaration of Helsinki – Ethical Principles for Medical Research Involving Human Subjects. Amended by 64th WMA General Assembly, Fortaleza, Brazil, October 2013 [accessed 2022 May 11]. [1] World Health Organization. (2021). Considerations for pneumococcal [21] Information system for the development of research in primary care (SIDIAP vaccination in older adults. Weekly Epidemiological Record, 96 (23), 217 - data base). Catalonia: SIDIAP; 2022 [accessed 2022 May 11]. 228. World Health Organization. [accessed 2022, Mar 11]. https://apps.who. [22] García-Gil Mdel M, Hermosilla E, Prieto-Alhambra D, Fina F, Rosell M, Ramos R, int/iris/handle/10665/341722 et al. Construction and validation of a scoring system for the selection of high- [2] Hausdorff WP, Feikin DR, Klugman KP. Epidemiological differences among quality data in a Spanish population primary care database (SIDIAP). Inform pneumococcal serotypes. Lancet Infect Dis 2005;5(2):83–93. https://doi.org/ Prim Care 2011;19(3):135–45. https://doi.org/10.14236/jhi.v19i3.806. 10.1016/S1473-3099(05)01280-6. [23] Ministerio de Sanidad. Conjunto Mínimo Básico de Datos de Hospitalización [3] Matanock A, Lee G, Gierke R, Kobayashi M, Leidner A, Pilishvili T. Use of 13- (CMBD-H). Madrid: Ministerio de Sanidad. Portal estadístico. Área de Valent Pneumococcal Conjugate Vaccine and 23-Valent Pneumococcal inteligencia de gestión [accessed 2022 May 11]. https://pestadistico. Polysaccharide Vaccine Among Adults Aged 65 Years: Updated inteligenciadegestion.mscbs.es/publicoSNS/N/rae-cmbd/cmbd-h. Recommendations of the Advisory Committee on Immunization Practices. [24] Hosmer DW, Lemeshow S. Applied Survival Analysis. Regression Modeling of MMWR Morb Mortal Wkly Rep 2019;68(46):1069–75. https://doi.org/ Time to Event Data. New York: John Wiley & Sons; 1999. 10.15585/mmwr.mm6846a5. Erratum. In: MMWR Morb Mortal Wkly Rep. [25] Lewnard J.A., Bruxvoort K.J., Fischer H., Hong V.X., Grant L.R., Jódar L., et al. 2020 Jan 03;68(5152):1195. Effectiveness of 13-Valent Pneumococcal Conjugate Vaccine Against Medically [4] Fedson DS. Preventing non bacteremic pneumococcal pneumonia in older Attended Lower Respiratory Tract Infection and Pneumonia Among Older adults: historical background and considerations for choosing between PCV13 Adults. Clin Infect Dis 2022 Sep 14;75(5):832–841. doi: 10.1093/cid/ciab1051 and PPV23. Hum Vaccin Immunother 2014;10(5):1322–30. https://doi.org/ [26] Hsiao A., Hansen J., Timbol J., Lewis N., Isturiz R., Alexander-Parrish R., et al. 10.4161/hv.28797. Incidence and Estimated Vaccine Effectiveness Against Hospitalizations for [5] Bonnave C, Mertens D, Peetermans W, Cobbaert K, Ghesquiere B, Deschodt M, All-Cause Pneumonia Among Older US Adults Who Were Vaccinated and Not et al. Adult vaccination for pneumococcal disease: a comparison of the national Vaccinated With 13-Valent Pneumococcal Conjugate Vaccine. JAMA Netw guidelines in Europe. Eur J Clin Microbiol Infect Dis 2019;38(4):785–91. Open 2022 Mar 1;5(3):e221111. doi: 10.1001/jamanetworkopen.2022.1111 https://doi.org/10.1007/s10096-019-03485-3. [27] Bosco JLF, Silliman RA, Thwin SS, Geiger AM, Buist DSM, Prout MN, et al. A [6] Bonten MJ, Huijts SM, Bolkenbaas M, Webber C, Patterson S, Gault S, et al. most stubborn bias: no adjustment method fully resolves confounding by Polysaccharide conjugate vaccine against pneumococcal pneumonia in adults. indication in observational studies. J Clin Epidemiol 2010;63:64–74. https:// N Engl J Med 2015;372(12):1114–25. https://doi.org/10.1056/ doi.org/10.1016/j.jclinepi.2009.03.001. NEJMoa1408544. [28] Marques Antunes M, Duarte GS, Brito D, et al. Pneumococcal vaccination in [7] McLaughlin JM, Jiang Q, Isturiz RE, Sings HL, Swerdlow DL, Gessner BD, et al. adults at very high risk or with established cardiovascular disease: systematic Effectiveness of 13-Valent Pneumococcal Conjugate Vaccine Against review and meta-analysis. Eur Heart J Qual Care Clin Outcomes 2021;7 Hospitalization for Community-Acquired Pneumonia in Older US Adults: A (1):97–106. https://doi.org/10.1093/ehjqcco/qcaa030. Test-Negative Design. Clin Infect Dis 2018;67(10):1498–506. https://doi.org/ [29] Musher DM, Abers MS, Bartlett JG. Evolving Understanding of the Causes of 10.1093/cid/ciy312. Pneumonia in Adults, With Special Attention to the Role of Pneumococcus. Clin [8] Prato R, Fortunato F, Cappelli MG, Chironna M, Martinelli D. Effectiveness of Infect Dis 2017;65(10):1736–44. https://doi.org/10.1093/cid/cix549. the 13-valent pneumococcal conjugate vaccine against adult pneumonia in [30] Navarro-Torné A, Montuori EA, Kossyvaki V, Méndez C. Burden of Italy: a case-control study in a 2-year prospective cohort. BMJ Open 2018;8 pneumococcal disease among adults in Southern Europe (Spain, Portugal, (3):. https://doi.org/10.1136/bmjopen-2017-019034e019034. Italy, and Greece): a systematic review and meta-analysis. Hum Vaccin [9] Kim JH, Chun BC, Song JY, Kim HY, Bae IG, Kim DM, et al. Direct effectiveness of Immunother 2021;17(10):3670–86. https://doi.org/10.1080/ pneumococcal polysaccharide vaccine against invasive pneumococcal disease 21645515.2021.1923348. and non-bacteremic pneumococcal pneumonia in elderly population in the era [31] Zalacain R, Torres A, Celis R, Blanquer J, Aspa J, Esteban L, et al. Pneumonia in of pneumococcal conjugate vaccine: A case-control study. Vaccine 2019;37 the elderly working group, Area de Tuberculosis e Infecciones Respiratorias. (21):2797–804. https://doi.org/10.1016/j.vaccine.2019.04.017. Community-acquired pneumonia in the elderly: Spanish multicentre study. [10] Vila-Corcoles A, Hospital I, Ochoa-Gondar O, Satue E, de Diego C, Vila-Rovira A, Eur Respir J 2003;21(2):294–302. https://doi.org/10.1183/ et al. Clinical effectiveness of 13-valent and 23-valent pneumococcal 09031936.03.00064102. vaccination in middle-aged and older adults: The EPIVAC cohort study, [32] Marimon JM, Ardanuy C. Epidemiology of pneumococcal diseases in Spain 2015–2016. Vaccine 2020;38(5):1170–80. https://doi.org/10.1016/ after the introduction of pneumococcal conjugate vaccines. Enferm Infecc j.vaccine.2019.11.012. Microbiol Clin (Engl Ed). 2021;39(3):142-150. English, Spanish. doi: 10.1016/j. [11] Berild JD, Winje BA, Vestrheim DF, Slotved HC, Valentiner-Branth P, Roth A, eimc.2020.02.016. et al. A Systematic Review of Studies Published between 2016 and 2019 on the [33] Carmona R, Cano R, Arias P. Resultados de la vigilancia Epidemiológica de las Effectiveness and Efficacy of Pneumococcal Vaccination on Pneumonia and enfermedades transmisibles. Informe anual. Años 2017-2018. Madrid: Centro Invasive Pneumococcal Disease in an Elderly Population. Pathogens 2020;9 Nacional de Epidemiología. Instituto de Salud Carlos III; 2020 [accessed 2022 (4):259. https://doi.org/10.3390/pathogens9040259. May 11]. https://www.isciii.es/QueHacemos/Servicios/ [12] Niederman MS, Folaranmi T, Buchwald UK, Musey L, Cripps AW, Johnson KD. VigilanciaSaludPublicaRENAVE/EnfermedadesTransmisibles/Documents/ Efficacy and effectiveness of a 23-valent polysaccharide vaccine against INFORMES/INFORMES%20RENAVE/RENAVE_Informe_anual__2017-2018.pdf. invasive and noninvasive pneumococcal disease and related outcomes: a [34] Ciruela P, Broner S, Izquierdo C, Nogueras MM, Hernández S, Muñoz-Almagro review of available evidence. Expert Rev Vaccines 2021;20(3):243–56. https:// C, Pallarés R, Jané M. Epidemiología de la enfermedad neumocócica invasiva en doi.org/10.1080/14760584.2021.1880328. Cataluña. Informe 2017-2018. Barcelona: Agencia de Salud Pública de [13] Leidner AJ, Murthy N, Chesson HW, Biggerstaff M, Stoecker C, Harris AM, et al. Cataluña. Departamento de Salud; 2020 [accessed 2022 May 11]. Cost-effectiveness of adult vaccinations: A systematic review. Vaccine 2019;37 https://canalsalut.gencat.cat/web/.content/_Professionals/ (2):226–34. https://doi.org/10.1016/j.vaccine.2018.11.056. Vigilancia_epidemiologica/documents/arxius/malaltia-pneumoccocica- [14] Treskova M, Scholz SM, Kuhlmann A. Cost Effectiveness of Elderly invasiva-informe-2017-2018_final-es.pdf. Pneumococcal Vaccination in Presence of Higher-Valent Pneumococcal [35] Hanquet G, Krizova P, Valentiner-Branth P, Ladhani SN, Nuorti JP, Lepoutre A, Conjugate Childhood Vaccination: Systematic Literature Review with Focus et al. SpIDnet/I-MOVE+ Pneumo Group. Effect of childhood pneumococcal on Methods and Assumptions. Pharmacoeconomics 2019;37(9):1093–127. conjugate vaccination on invasive disease in older adults of 10 European https://doi.org/10.1007/s40273-019-00805-5. countries: implications for adult vaccination. Thorax 2019;74(5):473–82. [15] Shao Y, Stoecker C. Cost-effectiveness of pneumococcal vaccines among adults https://doi.org/10.1136/thoraxjnl-2018-211767. over 50 years old in low- and middle-income countries: a systematic review. [36] Isturiz RE, Ramirez J, Self WH, Grijalva CG, Counselman FL, Volturo G, et al. Expert Rev Vaccines 2020;19(12):1141–51. https://doi.org/10.1080/ Pneumococcal epidemiology among us adults hospitalized for community- 14760584.2020.1874929. acquired pneumonia. Vaccine 2019;37(25):3352–61. https://doi.org/10.1016/ [16] Generalitat de Catalunya. Calendari de vacunacions sistematiques 2016. j.vaccine.2019.04.087. Barcelona (SP): Generalitat de Catalunya. Departament de Salut; 2017 [37] CDC. Pneumococcal Vaccination: Summary of Who and When to Vaccinate. [accessed 2022 Mar 25]. https://canalsalut.gencat.cat/web/ Centers for Disease Control and Prevention. CDC; 2022 Jan 24 [accessed 2022 .content/contingut_responsiu/salutAZ/ V/vacunacions/documents/calendari_vacunacions.pdf. 9 A. Vila-Córcoles, O. Ochoa-Gondar, C. de Diego-Cabanes et al. Vaccine: X 13 (2023) 100264 May 11]. https://www.cdc.gov/vaccines/vpd/pneumo/hcp/who-when-to- pneumonia for incidence and vaccine efficacy studies. Am J Epidemiol vaccinate.html. 1999;149(3):282–9. https://doi.org/10.1093/oxfordjournals.aje.a009804. [38] Schiffner-Rohe J, Witt A, Hemmerling J, von Eiff C, Leverkus FW. Efficacy of [40] Torres A, Barberán J, Falguera M, Menéndez R, Molina J, Olaechea P, Rodríguez PPV23 in Preventing Pneumococcal Pneumonia in Adults at Increased Risk–A A; Grupo de la Guía Multidisciplinar para el Manejo de la Neumonía Adquirida Systematic Review and Meta-Analysis. PLoS One 2016;11(1):. https://doi.org/ en la Comunidad. Multidisciplinary guidelines for the management of 10.1371/journal.pone.0146338e0146338. community-acquired pneumonia [in Spanish]. Med Clin (Barc). 2013;140 [39] Guevara RE, Butler JC, Marston BJ, Plouffe JF, File Jr TM, Breiman RF. Accuracy (5):223.e1-223.e19. Spanish. doi: 10.1016/j.medcli.2012.09.034. of ICD-9-CM codes in detecting community-acquired pneumococcal 10