Effectiveness of pneumococcal vaccination in reducing hospitalization and mortality among the elderly: A systematic review and meta-analysis

ABSTRACT

This meta-analysis assesses pneumococcal vaccine effectiveness in reducing hospitalization and mortality among the elderly. The studies retrieved from PubMed, Scopus, Embase, and Cochrane through March 2025. Of 3,002 identified studies, 35 met inclusion criteria (30 cohort, 3 RCTs, 2 case–control), encompassing 1.65 million vaccinated and 2.62 million unvaccinated individuals. Pneumonia-related hospitalization significantly declined in subgroups receiving mixed pneumococcal polysaccharide vaccine (PPV) and conjugate vaccine (PCV-13), notably among individuals aged >75 and those with chronic conditions. All-cause hospitalization dropped (OR: 0.94; 95% CI: 0.91–0.97), especially in PPV-23 recipients, Asian populations, and those with comorbidities. Pneumonia mortality decreased across age groups >60, and all-cause mortality declined among those aged 60–75. No significant impact was observed on emergency visit rates. Although findings favor vaccine effectiveness, high heterogeneity warrants further comparative trials to confirm outcomes.

Introduction

Pneumococcal pneumonia remains a leading cause of morbidity and mortality worldwide, particularly among older adults.¹ Streptococcus pneumoniae, the causative agent, is responsible, and disproportionately affects individuals aged 65 years or older due to age-related immunocompromised state and other comorbidities.² Globally, pneumococcal pneumonia significantly contributes to high hospitalization rates in elderly populations.¹ This often leads to complications such as respiratory failure, sepsis, and prolonged hospital stays, thereby imposing a considerable burden on healthcare systems.^3–5 The diagnostic challenges and associated medical needs in severe cases further increases this burden.⁶

The economic and clinical consequences of hospitalization for pneumococcal pneumonia are substantial.^7,8 Older adults who are hospitalized for this condition face increased risks of long-term disability, readmission, and death.^9–11 Additionally, the direct medical costs associated with inpatient care, as well as the indirect costs related to loss of independence and productivity, highlight the importance of effective preventive strategies.^12–14

Pneumococcal vaccines, including both conjugated (PCVs) and polysaccharide (PPVs), have shown promise in reducing the incidence of invasive pneumococcal disease and pneumonia in adults.^15–17 Several studies have suggested that these vaccines may significantly lower hospitalization rates and mortality associated with pneumococcal infections in the elderly.^18–20 Additionally, the vaccination programs and their implementation appear to be cost-effective and would yield substantial reductions in pneumococcal disease and associated burden.^20–22

While existing meta-analyses have evaluated the broader effectiveness of pneumococcal vaccines, including both polysaccharide and conjugate formulations, their focus has often been limited to the general population or general outcomes, like vaccine efficacy or invasive pneumococcal disease prevention.^23–27 Notably, these meta-analyses did not specifically investigate the impact of pneumococcal vaccines on hospitalization or all-cause mortality among the elderly as these are few major outcomes that contribute to the higher clinical and economic burden. To address this knowledge gap, the present meta-analysis aims to evaluate the effect of pneumococcal vaccination on hospitalization rates and mortality in elderly populations. By synthesizing current evidence, this review seeks to inform clinical and policy decisions regarding the optimal use of vaccination to reduce the burden of pneumococcal disease in older adults.

Materials and methods

This study followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines²⁸ to report this systematic review, and the protocol is registered in PROSPERO (CRD42025637010).

Literature search

A comprehensive search of electronic databases was conducted in PubMed, Scopus, the Cochrane Library, and Embase from inception to March 2025. Additionally, ClinicalTrials.gov and ProQuest Dissertations & Theses Global were searched for any additional resources. The bibliographic search of relevant articles was performed to locate any relevant citations. The search strategy comprised all entry terms, Medical Subject Heading (MeSH) terms, and all keywords that belong to two domains: 1) pneumococcal vaccine AND 2) elderly OR geriatric, which were combined using appropriate Boolean operators. A detailed search strategy in each database is provided in Supplementary file S1.

Study selection criteria

For inclusion in this review, we considered the following criteria; the studies should (i) include the elderly population (typically aged >60 years) irrespective of their gender and ethnicity, however, the definition of elderly was based on individual studies as per the author’s discretion; (ii) the administration of pneumococcal vaccine at any stage of life as a prevention or treatment were considered as an intervention; (iii) should have a comparator arm such as control, no vaccine, or different vaccine; (iv) should report any of the outcome such as hospitalization (pneumococcal associated or any cause) or mortality (pneumococcal associated or any cause); and (v) followed a study design of randomized (RCTs)/non-randomized control trials, or interventional or observational studies with a comparator arm published in the English language. We excluded studies that included a mixed population of elderly, mixed vaccination with any other vaccine, single-arm studies, reviews, case studies, cross-sectional studies, protocols, or guidelines.

Study selection and data extraction

The records from the above-mentioned databases were retrieved to Microsoft Excel and the duplicates were removed. After the elimination of duplicates, titles and abstracts were screened to check for the satisfaction of the outlined eligibility criteria. The full text of included studies was then screened for eligibility using the same criteria. First author’s last name and year of publication were used to identify the studies. A pre-defined data extraction grid in Excel was used for data extraction. The following information was extracted from the included studies: characteristics of the studies (author, year, country and location of study, study design, and study duration), characteristics of participants (age, gender, sample size, etc.), type of vaccination, and outcomes. Two independent authors were involved in the study selection and data extraction. Any disagreement during the selection process and data extraction was resolved through discussion with a third author.

Risk of bias (quality) assessment

Two review authors independently assessed the methodological quality of all the included studies. The quality of RCTs was assessed using the Cochrane Risk of bias Assessment tool.²⁹ The Newcastle-Ottawa scale was used for the observational studies.³⁰ Disagreements were resolved by a discussion between the investigators involved or by a third reviewer.

Data synthesis

The data from the studies were summarized in a narrative manner and the quantitative data were pooled through a meta-analysis. Review Manager Software (RevMan, version 5.3 for Windows; The Cochrane Collaboration, Oxford, UK) was used for conducting the meta-analysis. The statistical heterogeneity of data was assessed using the I² statistic and Cochrane P value. The random-effects model was used to perform the meta-analysis where there was a high heterogeneity (I² > 50; P < .01), whereas a fixed-model method was adopted when there was no significant heterogeneity (I² < 50; P > .01).²⁹

Subgroup and sensitivity analysis

A subgroup analysis was planned to explore the sources of heterogeneity based on the study design, vaccine types, age category, duration of follow-up, geographic location, and comorbidities. A sensitivity analysis was conducted by excluding the studies with the lowest weight and/or sample size with less than 25%iles. A funnel plot was generated for visual inspection, and the Egger’s test was used to statistically confirm the publication bias.²⁹

Results

This study selection process

The literature search yielded a total of 3001 records and an additional study found from bibliographic search. After the removal of duplicates (n = 1206), and exclusion of 1300 initial records, 496 full texts were screened. Of these, 461 were excluded for appropriate reasons. Not having an outcome of interest (n = 216), elderly population (n = 90), pneumococcal vaccine (n = 77), being a non-comparative study (n = 42), non-English (n = 13), review (n = 16), or a mixed intervention (n = 7) were the reasons for exclusion. Finally, 35 publications^31–38 were included in this systematic review and meta-analysis. The PRISMA flow diagram of the study selection process is provided in Figure 1. The list of excluded studies is provided in Supplementary file S2.

Figure 1.

The PRISMA diagram for study selection process.

Study characteristics

A total of 35 studies published during 1997 to 2024 were included in this systematic review. These studies were from Spain (n = 9), Taiwan (n = 7), Japan (n = 4), the Unites States (n = 4), South Korea (n = 2), Canada (n = 2), Sweden (n = 2), Austria (n = 1), China (n = 1), Finland (n = 1), Germany (n = 1), and the United Kingdom (n = 1). The majority (n = 30; 85.7%) were cohort studies, followed by RCTs (n = 3; 8.6%), and case–control studies (n = 2; 5.7%). The studies reported the effects of 23‑valent pneumococcal polysaccharide vaccine (PPV-23), 13‑valent pneumococcal conjugate vaccine (PCV-13), 13‑valent pneumococcal polysaccharide vaccine PPV-13, and both PPV-13 and PCV-13. These studies reported a total of 1,648,919 patients to be vaccinated and 2,616,711 to be non-vaccinated with a median (interquartile range) of 1487 (511–12584), and 4581 (682–35702), respectively. All studies considered the age as equal to or above 65, except the study by Koivula et al.³³ which considered 60 years to specify the older population. The male–female proportion varied across the studies. Across the studies, “cases”³⁹ were defined as patients who received any of these vaccines and “controls” were those who were not vaccinated. Detailed study characteristics are provided in Table 1 and patient characteristics in Table 2. The case definition in each study is provided in Supplementary file S3.

Table 1.

Characteristics of included studies.

Author, year	Country of participants	Study design	Type of Vaccine	No. of vaccinated & unvaccinated	Follow-up period/study duration
Maruyama T 2010	Japan	RCT	PPV-23	502 & 504	2.27 years
Rose N, 2021	Germany	Cohort	13- PCV and PPV	1,136 & 72867	2 years
Song JY 2018	South Korea	Cohort	PPV-23	871 & 1248	30 days
Koivula I 1997	Finland	RCT	14- PV	1364 & 1473	3 years
Kim S 2024	South Korea	Cohort	PV	1609 & 2906	NR
Li C 2015	United States	Cohort	PPV-23	1347 & 2010	1 year
Vila-Corcoles A 2012	Spain	Cohort	PPV-23	8981 & 18223	1 year
Murata F 2024	Japan	Cohort	PPV-23	55509 & 55509	3 years
Lee MS 2024	Taiwan	Cohort	PPV-23	1355 & 1355	9 years
Streeter AJ 2003	United Kingdom	Cohort 1	PPV-23	25870 & 29087	1 year
		Cohort 2	PPV-23	30028 & 42625
		Cohort 3	PPV-23	104969 & 92225
Hsiao A 2022	United States	Cohort	PCV-13	366931 & 311305	3 years
Naito T 2020	Japan	Cohort	PPV-23	310 & 1045	4.5 years
Vila-Corcoles 2019	Spain	Cohort	PCV-13	3617 & 1004937	2 years
			PPV-23	692016 & 316538
Chung-Yi Li 2021	Taiwan	Cohort	PPV-23	2188 & 2188	7 years
Chiou WY 2019	Taiwan	Cohort	PPV-23	2622 & 7866	2 years
Chiou WY 2018	Taiwan	Cohort	PPV-23	377 & 754	2 years
Payeras A 2017	Spain	Cohort	PV	7 & 236	5 years
Kuo SC 2016	Taiwan	Cohort	PPV-23	33395 & 33395	1 year
Tsai YH 2015	Taiwan	Cohort	PPV-23	229181 & 229181	1 year
Chiou WY 2015	Taiwan	Cohort	PPV-23	157 & 628	4 years
Ochoa-Gondar O 2014	Spain	Cohort	PPV-23	8981 & 12044	3 years
Ochoa-Gondar O 2013	Spain	Cohort	PPV-23	8981 & 18223	3 years
Domínguez A 2013	Spain	Case–control	PPV-23	175 & 700	3 years
Neupane B 2010	Canada	Cohort 1	PV	366 & 286	30 days
		Cohort 2	PV	175 & 144	90 days
Hung I.F.N 2010	China	Cohort	PPV-23	1875 & 25393	64 weeks
Kawakami K 2010	Japan	RCT	PPV-23	394 & 392	2 years
Ochoa-Gondar O 2008	Spain	Cohort	PPV-23	701 & 597	3 years
Johnstone J 2007	Canada	Cohort 1	PPV-23	750 & 2570	2 years
		Cohort 2	PPV-23	10 & 85	2 years
Vila-Corcoles A 2006	Spain	Cohort	PPV-23	4986 & 6255	3 years
Vila-Corcoles A 2005	Spain	Cohort	PPV-23	4986 & 6255	1 year
Christenson B 2003	Sweden	Cohort	PPV-23	23249 & 134045	1 year
Wagner C 2002	Austria	Case–control	PPV-23	514 & 562	2.3 years
Jackson LA 2003	United States	Cohort	PPV-23	26,313 & 21052	3 years
Christenson B 2001	Sweden	Cohort	PPV-23	841 & 159385	6 months
Nichol KL 1999	United States	Cohort	PPV-23	1280 & 618	2 years

NR: Not reported PV: Pneumococcal vaccine; PCV: Pneumococcal conjugated vaccine; PPV: Pneumococcal polysaccharide vaccine.

Table 2.

Characteristics of included participants.

Author, year	Age in intervention group	Age in control group	Male % in intervention group	Male % in control group
Maruyama T 2010	84.7 ± 7.7*	84.8 ± 7.6*	22.1	21.6
Rose N, 2021	69.62 ± 7.83*	72.38 ± 8.61*	43.84	42.22
Song JY 2018	76.0 ± 6.7*	NR	51.5	NR
Koivula I 1997	≥60	≥60	38.2	35.6
Kim S 2024	76.0 (71–82) **	78 (72–83) *	56.4	53
Li C 2015	77 (71–81) **	77 (72–82)**	97.8	97.7
Vila-Corcoles A 2012	72 ± 7.8*	71.5 ± 8.9*	45.2	44.3
Murata F 2024	≥65	≥65	37.1	37.1
Lee MS 2024	80.8 ± 4.4*	80.9 ± 4.6*	55	55
Streeter AJ 2003	Cohort 1: 84.5 ± 4.0* Cohort 2: 79.3 ± 4.3* Cohort 3: 71.6 ± 5.4*	Cohort 1: 85.9 ± 4.8* Cohort 2: 82.3 ± 5.8* Cohort 3: 75.1 ± 8.1*	Cohort 1: 36.7 Cohort 2: 40.4 Cohort 3: 44.7	Cohort 1: 28.5 Cohort 3: 40.3 Cohort 2: 31.8
Hsiao A 2022	NR	NR	43.4	43.5
Naito T 2020	77.4 ± 7.4*	77.5 ± 7.6*	54.8	54.5
Vila-Corcoles 2019 A	≥65	≥65	NR	NR
Chung-Yi Li 2021	80.88 ± 4.19*	80.97 ± 4.37*	100	100
Chiou WY 2019	79.7 ± 3.9*	79.8 ± 4.2*	59.1	59.1
Chiou WY 2018	82.69 ± 4.08*	82.54 ± 4.20*	63.9	61.4
Payeras A 2017	≥65	≥65	NR	NR
Kuo SC 2016	80.5 ± 4.7*	80.6 ± 5.4*	52.1	52.1
Tsai YH 2015	81.69 ± 4.21*	81.67 ± 4.29*	52.07	52.16
Chiou WY 2015	80.2 ± 4.3*	80.2 ± 4.3*	76.4	76.4
Ochoa-Gondar O 2014	72 ± 7.8*	71.5 ± 8.9*	45.2	45.3
Ochoa-Gondar O 2013	NR	NR	45.2	44
Domínguez A 2013	≥65	≥65	NR	NR
Neupane B 2010	NR	NR	NR	NR
Hung I.F.N 2010	75(71–80) *	75 (70–80) *	45	47
Kawakami K 2010	78.5 ± 7.3*	77.7 ± 7.2*	38.10	32.30
Ochoa-Gondar O 2008	76.5 ± 6.9*	74.2 ± 6.7*	73	75
Johnstone J 2007	≥65	≥65	Cohort 1: 49 Cohort 2: 20	Cohort 1: 54 Cohort 2: 61
Vila-Corcoles A 2006	≥65	≥65	43.6	43.4
Vila-Corcoles A 2005	≥65	≥65	43.6	43.4
Christenson B 2003	≥65	≥65	NR	NR
Wagner C 2002	82.2 ± 10.0*	82.6 ± 10.1*	NR	NR
Jackson LA 2003	74.5**	74.7**	41.8	41.5
Christenson B 2001	NR	NR	NR	NR
Nichol KL 1999	NR	74.9 ± 6.7*	55.2	45.6

*Mean^{; **}Median; NR: Not reported.

Methodological quality of included studies

Among three included RCTs, two studies^33,39 appeared to have a low risk of bias. The RCT by Koivula et al. was a single blind RCT, whereas the study by Kawakami et al.³⁹ had a moderate risk due to unclear information on allocation concealment and open-label trial design. The results are provided in Supplementary file S4A. The observational studies were well conducted either retrospectively or prospectively from the high-quality data from the healthcare database, population database or the preventive program projects that confirm the low bias in the selection of participants. All analyses were well-performed, accounting for the potential confounding factors with appropriate statistical analysis. Hence, all studies were graded to be of good quality with a score of 8 or 9 out of 9. A detailed score card is presented in Supplementary file S4B.

Effectiveness of pneumococcal vaccination on hospitalization

Pneumonia-associated hospitalization

The meta-analysis of 23 studies with 24 cohorts indicates that pneumococcal vaccination was not associated with a significant reduction in pneumonia-associated hospitalization (OR: 0.83; 95% CI: 0.60–1.15; P = .27; I²: 100%) compared to those who are not vaccinated (Figure 2). A similar result was observed in subgroup analysis based on the patients administered with PPV-23 (OR: 0.76; 95% CI: 0.56–1.03; p = .07; I²: 100%; 21 studies), or PCV-13 (OR: 2.88; 95% CI: 0.41–20; p = .29; I²: 100%; 2 studies). However, there was a significant reduction among the patients who received mixed PPV or PCV-13 (OR: 0.49; 95% CI: 0.35–0.69; p < .0001) compared to those who were not vaccinated³¹ as provided in Supplementary file S5A. Similarly, no significant reduction was found in the pooled estimates of observational studies (OR: 0.85; 95% CI: 0.61–1.18; P = .34; 22 studies), whereas the RCT by Maruyama et al.,³⁹ reported a significant reduction in pneumococcal hospitalization (OR: 0.54; 95% CI: 0.37–0.79; p = .001) (Supplementary file S5B). Interestingly, the subgroup analysis based on the duration of follow-up indicates that there was a significant reduction in pneumonia-associated hospitalization at 1-year (OR: 0.63; 95% CI: 0.42–0.95; p = .03) and 1–5 years (OR: 0.76; 95% CI: 0.61–0.94; p = .010) follow-up. But this was not significant after 5 years (p = .60) as provided in Supplementary file S5C. Also, there was a significant reduction among those who were aged >75 years (OR: 0.71; 95% CI: 0.54–0.94; p = .02), but not those who were aged 60–75 years (Supplementary file S5D). However, no significant differences were observed with studies from European, Asian, and North American populations (Supplementary file S5E). The elderly population who had a chronic disease (OR: 0.84; 95% CI: 0.73–0.96; p = .01) were observed to have significantly lower rates of pneumonia-associated hospitalization (Supplementary file S5F).

Figure 2.

Effectiveness of vaccines on pneumonia-associated hospitalization.

All-cause hospitalization

The meta-analysis of five studies indicates that pneumococcal vaccination significantly reduced all-cause hospitalization by 6% (OR: 0.94; 95% CI: 0.91–0.97; P < .0001; I2:0%) compared to the non-vaccinated cohort (Figure 3). A similar significant benefit was observed in subgroup analysis based on the patients administered with PPV-23 (p < .00001), 1-year follow-up (P = .00002), older elderly who are aged more than 75 years (P = .0001), Asian population (P = .0001), and those with a chronic disease (P = .0002). But the effect was not significant (P > .05) with vaccines other than PPV-23, follow-up of 1–5 years, follow-up of >5 years, elderly aged <75, Europeans, North Americans, and those with a non-chronic disease (Supplementary file S5G-S5K).

Figure 3.

Effectiveness of vaccines on all-cause hospitalization.

Emergency visits

Only the study by Kuo et al.⁴⁰ reported the emergency visit outcome. The study reported a non-significant effect of pneumococcal vaccination on emergency visits (OR: 0.99; 95% CI: 0.95–1.03) compared to the non-vaccinated group.

Effectiveness of pneumococcal vaccination on mortality

Pneumonia-associated mortality

The meta-analysis of 15 studies with 16 cohorts indicates that pneumococcal vaccination was not associated with a significant reduction in pneumonia-associated mortality (OR: 0.69; 95% CI: 0.35–1.39; P = .30; I²: 99%) compared to the non-vaccinated group (Figure 4). However, there was a significant reduction in pneumonia-associated mortality in studies that reported mixed PCV-13 or PPV-13 (P = .01), intermediate follow-up of 1–5 years (P < .00001), younger elderly (aged 60–75; P = .004), and older elderly (aged >75 years; P = .02). Whereas, there was no significant (P > .05) benefit among those who were administered with PPV-23 alone, PCV-13, PPV-14, had <1-year follow-up, >5-year follow-up, were European or Asian, or with chronic and non-chronic comorbidities. Also, there were no significant benefits in studies which were RCT and observational in design. The results are provided in Supplementary file S5L – S5Q.

Figure 4.

Effectiveness of vaccines on pneumonia-associated mortality.

All-cause mortality

The meta-analysis of 19 studies with 21 cohorts indicates that pneumococcal vaccination was not associated with a significant reduction in all-cause mortality (OR: 0.71; 95% CI: 0.25–1.98; P = .51; I²: 100%) compared to those who are not vaccinated (Figure 5). However, there was a significant reduction in pneumonia-associated mortality in studies that reported mixed PCV-13 or PPV-13 (P < .0001), and younger elderly (aged 60–75; P = .005). Whereas, there was no significant (P > .05) benefit among those who were administered with PPV-23 alone, PCV-13; had <1-year, 1–5-year, and >5-year follow-up; aged >75 years; were European, Asian, and North American; had chronic and non-chronic comorbidities. The results are provided in Supplementary file S5R – S5V.

Figure 5.

Effectiveness of vaccines on all-cause mortality.

Exploration of heterogeneity

The subgroup analysis based on the type of vaccines, study design, duration of follow-up, age, geographical location, and comorbidities indicates that there was no difference in heterogeneity (I²: > 75%) on pneumonia-associated hospitalization, all-cause hospitalization, pneumonia-associated mortality, and all-cause mortality (Supplementary file S5A – and S5V). Hence, these factors are not contributing to the heterogeneity. However, the RCT subgroup might have contributed to heterogeneity (I2: 0%) to the pneumonia-associated death mortality (Supplementary file S5M). The studies with non-chronic population (I²: 18%), Asian population (I²: 0%), European population (I²: 68%), older elderly (I²: 0%), long-term follow-up, and PPV-23 might have contributed to all-cause hospitalization (Supplementary file S5G – S5K). Table 3 summarizes the findings from the meta-analyses and subgroup analyses.

Table 3.

The summary of meta-analyses.

Outcomes & Subgroup	No. of studies	Odds Ratio (95% CI)	P-value	Heterogeneity (I2)
Pneumonia-associated hospitalization
Overall	23	0.83 (0.60–1.15)	.27	100%
PPV-23	21	0.76 (0.56–1.03)	.07	100%
PCV-13	2	2.88 (0.41–20)	.29	100%
Mixed PPV or PCV	1	0.49 (0.35–0.69)	<.0001*	NA
Observational studies	22	0.85 (0.61–1.18)	.34	100%
RCT	1	0.54 (0.37–0.79)	.001*	NA
≤1-year follow-up	7	0.63 (0.42–0.95)	.03*	97%
1–5 years follow-up	8	0.76 (0.61–0.94)	.010*	95%
>5 years follow-up	9	1.13 (0.72–1.78)	.60	100%
Aged >75 years	11	0.71 (0.54–0.94)	.02*	99%
Aged 60–75 years	7	0.71 (0.48–1.06)	.10	97%
Age not mentioned	6	1.38 (0.74–2.58)	.31	100%
European population	13	0.88 (0.57–1.36)	.55	100%
Asian population	9	0.73 (0.50–1.06)	.10	98%
North American population	2	1.13 (1–1.27)	.05	71%
Patients with chronic disease	8	0.84 (0.73–0.96)	.01	63%
Patients without a chronic disease	16	0.88 (0.60–1.15)	.27	100%
All-cause hospitalization
Overall	5	0.94 (0.91–0.97)	<.0001*	0%
PPV-23	21	0.94 (0.91–0.97)	<.0001*	6%
Unknown vaccine type	1	1.16 (0.64–2.10)	.62	NA
≤1-year follow-up	1	0.94 (0.91–0.97)	.0002	NA
1–5 years follow-up	1	1.01 (0.82–1.24)	.93	NA
>5 years follow-up	3	0.90 (0.77–1.05)	.17	31%
Aged >75 years	2	0.94 (0.91–0.97)	<.0001*	0%
Age not mentioned	3	0.96 (0.80–1.15)	.63	44%
European population	2	0.94 (0.77–1.13)	.51	68%
Asian population	2	0.94 (0.91–0.97)	.0001*	0%
North American population	1	1.16 (0.64–2.10)	.62	NA
Patients with chronic disease	1	0.94 (0.91–0.97)	.0002	NA
Patients without a chronic disease	4	0.94 (0.83–1.06)	.27	100%
Pneumonia-associated mortality
Overall	15	0.69 (0.35–1.39)	.30	99%
PPV-23	13	0.60 (0.28–1.32)	.20	99%
PCV-13	1	3.86 (2.81–5.30)	<.0001*	NA
PPV-14	1	0.90 (0.27–3.0)	.86	NA
Mixed PPV and PCV	1	0.57 (0.37–0.88)	.01*	NA
Observational studies	13	0.69 (0.32–1.50)	.35	100%
RCT	3	0.65 (0.4–1.04)	.07	0%
≤1-year follow-up	6	0.49 (0.13–1.89)	.30	98%
1–5 years follow-up	5	0.46 (0.35–0.60)	<.00001	90%
>5 years follow-up	5	1.45 (0.56–3.75	.44	98%
Aged 60–75 years	2	0.43 (0.27–0.68)	.0004	78%
Aged >75 years	7	0.41 (0.20–0.86)	<.0001*	99%
Age not mentioned	7	1.33 (0.60–2.95)	.48	97%
European population	12	0.86 (0.42–1.74)	.67	99%
Asian population	4	0.36 (0.08–1.64)	.19	97%
Patients with chronic disease	1	1.07 (0.45–2.54)	.88	NA
Patients without a chronic disease	14	0.67 (0.33–1.39)	.29	99%
All-cause mortality
Overall	19	0.71 (0.25–1.98)	.51	100%
PPV-23	18	0.66 (0.20–2.16)	.49	100%
PCV-13	2	1.60 (0.73–3.54)	.24	100%
Mixed PPV and PCV	1	0.48 (0.34–0.68)	<.0001*	NA
≤1-year follow-up	3	0.30 (0.06–1.61)	.16	100%
1–5 years follow-up	6	0.76 (0.53–1.09)	.13	95%
>5 years follow-up	13	0.86 (0.51–1.44)	.56	100%
Aged 60–75 years	4	0.64 (0.46–0.87)	.005*	93%
Aged >75 years	7	0.47 (0.13–1.69)	.25	100%
Age not mentioned	11	1.06 (0.60–1.89)	.83	100%
European population	9	1.08 (0.58–2.00)	.81	100%
Asian population	6	0.41 (0.13–1.35)	.14	100%
North American population	7	0.82 (0.67–1.00)	.05	92%
Patients with chronic disease	3	0.60 (0.25–1.44)	.25	96%
Patients without a chronic disease	19	0.73 (0.24–2.22)	.58	100%

*indicate the significant results.

Publication bias

The visual inspection of funnel plot analyses reports that there was obvious asymmetry on pneumonia-associated mortality (Supplementary file S6A) and all-cause mortality (Supplementary file S6B). There was a symmetrical funnel plot for pneumonia-associated hospitalization (Supplementary file S6C), which suggest no publication bias. Meanwhile, the Egger’s test reported a significant publication bias (P < .05).

Sensitivity analysis

The sensitivity analysis, which involved removing the study with lowest weight, did not change the outcomes, such as pneumonia-associated hospitalization (OR: 0.86; 95% CI: 0.62–1.19), all-cause hospitalization (OR: 0.94; 95% CI: 0.91–0.97), pneumonia-associated mortality (OR: 0.68; 95% CI: 0.33–1.40), and all-cause mortality (OR: 0.73; 95% CI: 0.26–2.08) by removing the studies by Christenson et al.⁴¹, Neupane et al.⁴², Koivula et al.³³, and Johnstone et al.⁴³ respectively. All sensitivity results are provided in Supplementary file S7.

Discussion

Pneumococcal vaccination has been a cornerstone of public health strategies aimed at reducing the burden of pneumococcal disease, particularly among older adults and individuals with chronic comorbidities.²⁷ Numerous observational studies and some randomized controlled trials have examined the effectiveness of pneumococcal vaccines in preventing hospitalization and reducing mortality. More recent studies and meta-analyses show mixed results.^23,24,27 These inconsistencies are often attributed to limitations in outcome ascertainment, unmeasured confounding, and the inclusion of deaths not directly attributable to pneumococcal disease. Despite the variability in findings, most studies agree that pneumococcal vaccination remains a valuable intervention, particularly when used in conjunction with influenza vaccination and other preventive measures.^44,45 Hence, this meta-analysis aimed to investigate the effect of pneumococcal vaccines on reducing hospitalization and mortality among the elderly population.

This metanalysis included a total of 35 studies published between 1997 and 2024 from diverse geographical locations. In terms of study design, the majority were cohort studies (n = 30; 85.7%), reflecting real-world observational evidence on pneumococcal vaccine effectiveness. A smaller number of RCTs (n = 3; 8.6%) and case–control studies (n = 2; 5.7%) were also included, offering additional insights with varying levels of methodological rigor. This distribution highlights a reliance on observational data in the current evidence base and this evidence requires caution due to the influence of indication and residual confounding, outcome misclassification, less rigorous methodology, temporal and immortal bias, and lack of randomization.^46–50 However, all studies appeared to have a good quality as per the Newcastle-Ottawa scale, scoring 8 to 9.

The pooled estimates found that pneumococcal vaccination was not significantly associated with a reduction in pneumonia-associated hospitalizations (OR: 0.83; P = .27). The substantial heterogeneity observed (I² = 100%) underscores the variability across studies and calls for cautious interpretation. Similar results were observed with the subgroups of individual vaccine types. This was similar to the findings of a large long-term cohort study that followed the patients for about 3.8 years. The study records that PPV was not associated with a reduced risk of death or hospitalization and better pneumococcal vaccination strategies are urgently needed.¹⁸ In contrast, a Canadian study reported that the publicly funded PCV immunization program is responsible for substantial reductions in pneumonia hospitalizations and related healthcare costs among the aged population.⁵⁰ A greater number of PPV studies in our analysis may have influenced our results. Additionally, factors such as heterogeneity in study population, vaccine types, temporal and regional variabilities may have affected our results. There are still investigations warranted to understand the direct and indirect effects of current and expanded valency pneumococcal vaccination in the US population, though it has been introduced as part of the routine vaccination program.¹⁹ Interestingly, there was a significant reduction of pneumonia hospitalization in the RCT subgroup. This highlights the need for more rigorous, high-quality RCT methodologies to estimate the actual benefits of vaccines in pneumonia hospitalization.

Interestingly, our findings indicated a significantly reduced all-cause hospitalization among those vaccinated with any pneumococcal vaccine (OR: 0.94; P < .0001) and PPV-23 (OR: 0.94; P = .07). These results were aligned with previous studies that reported a beneficial effect among various populations such as older children,¹⁹ the pneumonia population,^20,51 and the older population,⁵¹ and with community-acquired pneumonia (CAP).³⁴ This clearly indicates that pneumococcal vaccination is more effective against all-cause hospitalization and offers broader protection, better prevention of morbidity, and better public health benefits as opposed to specific pneumococcal benefits. This will also improve economic benefits for patients.⁵⁰ The significant reduction in pneumonia-associated and all-cause hospitalization, particularly among high-risk subgroups such as those aged >75 and individuals with chronic conditions, supports the targeted use of pneumococcal vaccines in elderly populations. These were aligned with previous studies demonstrating the effectiveness of pneumococcal vaccines in high-risk populations.^23,52

Our findings indicated that there was no significant difference in pneumonia-associated mortality (P = .30) and all-cause mortality (P = .51), among those who were or were not vaccinated. These findings align with previous literature that has reported mixed or inconclusive evidence regarding the mortality benefits of pneumococcal vaccination in adult populations, which warrants future research.^53,54 The CDC provides updated, expanded age-based recommendations to improve pneumococcal disease prevention in adults aged 50–64 years, particularly among demographic groups experiencing higher disease rates. However, this needs future evaluation to understand its effectiveness.⁵⁵ This was in contrast to the results of Kim et al. who reported that prior pneumococcal vaccination was associated with improved in-hospital mortality and 30-day mortality among elderly hospitalized CAP patients.³⁴ However, this study included only CAP patients. Despite observed reductions in pneumonia-related and all-cause mortality with mixed PPV and PCV-13, the lack of effect on emergency visits and inconsistencies across subgroups highlight the need for further high-quality studies to clarify vaccine impact.¹⁹ Although mixed PPV and PCV-13 vaccination showed mortality benefits, the inconsistent impact across outcomes and subgroups reflects findings from earlier research, which also emphasized variability in vaccine effectiveness and the need for further targeted studies.^56,57 The immunological differences between PPV23 and PCV13 are important when considering vaccine effectiveness in older adults. PPV23 elicits a T-cell – independent response, which does not generate immunological memory and offers limited mucosal protection, potentially reducing its efficacy in this population. In contrast, PCV13, as a conjugate vaccine, induces a T-cell – dependent response that may confer longer-lasting immunity.^52–54 However, our meta-analysis included only two studies evaluating PCV13, resulting in a wide confidence interval, and thus no definitive conclusions can be drawn regarding its effectiveness based on current data. This results also might had affected by the lower sample size, which warrant future adequately sampled studies.

Despite providing good evidence on the effectiveness of pneumococcal vaccines in the elderly population, this meta-analysis has some limitations. First, a smaller number of RCTs were included, and a greater number of observational studies might have contributed to the obvious study bias. Hence, we need more well-conducted, adequately powered large RCTs to decide the effectiveness. Secondly, there was a smaller number of studies from lower-middle-income countries, thus the generalizability of these findings in this region of the world is limited. Hence, more research is needed from the lower-middle-income countries on this topic. Thirdly, the restriction to English language studies was a limitation. However, the comprehensive search in multiple databases and additional resources helped us to accumulate as much as evidence as possible on this topic. Additionally, not restricting to specific disease, condition, or study design among the elderly population helped us to receive the maximum amount of evidence on this topic. Although the funnel plot for pneumonia-associated hospitalization appeared symmetrical, Egger’s test revealed significant publication bias (P < .05). This discrepancy highlights the limitations of visual inspection alone and emphasizes the importance of combining graphical and statistical methods in bias assessment. Then, the differences in diagnostic criteria across studies including reliance on non‑specific ICD codes, low‑sensitivity sputum culture results, or lack of microbiological confirmation may have inflated heterogeneity and influenced outcome measurement; this limitation should be considered when interpreting the pooled estimates. Finally, we could not perform a subgroup analysis based on the immunocompromised patient status due to the limited availability of the data in the included studies.

Conclusions

Pneumococcal vaccination was associated with reduced all‑cause hospitalization and showed benefits in specific subgroups (e.g., older adults over 75 years, those receiving mixed PPV/PCV regimens and follow‑up of 1–5 years). However, the evidence for pneumonia‑specific hospitalization and mortality remains inconsistent because of heterogeneity among predominantly observational studies. Large, high‑quality randomized trials are needed to confirm the effectiveness of pneumococcal vaccination in the elderly.

Biography

Adeeb Bulkhi is an associate professor of medicine and a consultant physician at Umm Al-Qura University and chairman of the department of medicine at International Medical Center, specializing in allergy, immunology, and respiratory diseases. He holds a master’s degree in Clinical Research and Statistical Analysis, complementing his extensive clinical and academic expertise. His primary research interests include respiratory infections, severe asthma, chronic urticaria, chronic rhinosinusitis, and biologic therapies. Dr. Bulkhi is experienced in conducting systematic reviews and meta-analyses, contributing significantly to evidence-based clinical guidelines and improved patient outcomes. He actively participates in international collaborations, guideline committees, and medical education, regularly presenting at global conferences to advance clinical excellence in Allergy and Immunology.

Funding Statement

The authors reported there is no funding associated with the work featured in this article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Abbreviations

PCVs

Pneumococcal Conjugated Vaccines

PPVs

Polysaccharide (Vaccines)/Pneumococcal Polysaccharide Vaccines

IRB

Th Institutional Review Board

PRISMA

Preferred Reporting Items for Systematic reviews and Meta-Analyses

MeSH

Medical Subject Heading

RCTs

Randomized Controlled Trials

CAP

Community-Acquired Pneumonia

OR

Odds Ratio

RevMan

Review Manager Software

CDC

Centers for Disease Control and Prevention

Availability of data, code, and other materials

All the data and materials associated with this work are presented in the manuscript and the supplementary file. Any additional data can be made upon appropriate request to the corresponding author.

Ethical approval

All authors take full responsibility for the accuracy of all aspects of this manuscript.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21645515.2025.2561315