Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Effectiveness of the 23-Valent Pneumococcal Polysaccharide Vaccine (PPV23) against Pneumococcal Disease in the Elderly: Systematic Review and Meta-Analysis

Effectiveness of the 23-Valent Pneumococcal Polysaccharide Vaccine (PPV23) against Pneumococcal Disease in the Elderly: Systematic Review and Meta-Analysis

  • Gerhard Falkenhorst, 
  • Cornelius Remschmidt, 
  • Thomas Harder, 
  • Eva Hummers-Pradier, 
  • Ole Wichmann, 
  • Christian Bogdan



Routine vaccination of elderly people against pneumococcal diseases is recommended in many countries. National guidelines differ, recommending either the 23-valent polysaccharide vaccine (PPV23), the 13-valent conjugate vaccine (PCV13) or both. Considering the ongoing debate on the effectiveness of PPV23, we performed a systematic literature review and meta-analysis of the vaccine efficacy/effectiveness (VE) of PPV23 against invasive pneumococcal disease (IPD) and pneumococcal pneumonia in adults aged ≥60 years living in industrialized countries.


We searched for pertinent clinical trials and observational studies in databases MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, and Cochrane Database of Systematic Reviews. We assessed the risk of bias of individual studies using the Cochrane Risk of Bias tool for randomized controlled trials and the Newcastle-Ottawa Scale for observational studies. We rated the overall quality of the evidence by GRADE criteria. We performed meta-analyses of studies grouped by outcome and study design using random-effects models. We applied a sensitivity analysis excluding studies with high risk of bias.


We identified 17 eligible studies. Pooled VE against IPD (by any serotype) was 73% (95%CI: 10–92%) in four clinical trials, 45% (95%CI: 15–65%) in three cohort studies, and 59% (95%CI: 35–74%) in three case-control studies. After excluding studies with high risk of bias, pooled VE against pneumococcal pneumonia (by any serotype) was 64% (95%CI: 35–80%) in two clinical trials and 48% (95%CI: 25–63%) in two cohort studies. Higher VE estimates in trials (follow-up ~2.5 years) than in observational studies (follow-up ~5 years) may indicate waning protection. Unlike previous meta-analyses, we excluded two trials with high risk of bias regarding the outcome pneumococcal pneumonia, because diagnosis was based on serologic methods with insufficient specificity.


Our meta-analysis revealed significant VE of PPV23 against both IPD and pneumococcal pneumonia by any serotype in the elderly, comparable to the efficacy of PCV13 against vaccine-serotype disease in a recent clinical trial in elderly people. Due to its broader serotype coverage and the decrease of PCV13 serotypes among adults resulting from routine infant immunization with PCV13, PPV23 continues to play an important role for protecting adults against IPD and pneumococcal pneumonia.


Community-acquired pneumonia (CAP) is a major cause of hospital admissions and death in the elderly, with Streptococcus pneumoniae (pneumococcus) being the most frequently detected pathogen causing an estimated 20–30% of CAP cases [1, 2]. Invasive pneumococcal disease (IPD), in the elderly mostly presenting as pneumonia with bacteremia, is the most severe form of pneumococcal infections. Case fatality can exceed 20% in elderly patients [3]. More than 90 serotypes of S. pneumoniae can be distinguished on the basis of the antigen structure of the capsular polysaccharide.

Two pneumococcal vaccines are currently licensed for adults: one containing polysaccharides from 23 pneumococcal serotypes (PPV23), the other containing protein-conjugated polysaccharides from 13 serotypes (PCV13). Most industrialized countries recommend universal pneumococcal vaccination for the elderly, but there is considerable debate about the best vaccination strategy [4]. The choice of vaccine is primarily determined by the efficacy (i.e. the protective effect assessed in randomized controlled trials (RCTs)) or effectiveness (i.e. the protective effect assessed in observational studies) of the two vaccines against pneumococcal pneumonia (PP) and IPD, as well as by the prevalence of the pneumococcal serotypes contained in the respective vaccine among cases. In addition, cost-effectiveness aspects may be considered.

The pivotal RCTs leading to the license of the first commercial pneumococcal vaccine, a 14-valent plain polysaccharide vaccine (PPV14), were conducted in the 1970s among gold miners in South Africa, a population group with a high incidence of PP [5, 6]. In 1983, PPV14 was replaced by PPV23. Since then, its efficacy/effectiveness in the elderly has been investigated in several RCTs and observational studies.

We performed a systematic review and meta-analyses of RCTs and observational studies investigating the efficacy of PPV23 against the specific outcomes PP and IPD in people aged ≥60 years living in industrialized countries.

Since we started our review, three other systematic reviews and meta-analyses of PPV23 efficacy/effectiveness have been published in the beginning of 2016 [710]. Prior to these publications, a Cochrane review from 2013 [11] presented the most up-to-date meta-analysis. Remarkably, these four reviews have come to divergent conclusions regarding clinical effectiveness of PPV23. We scrutinized these reviews and discovered that they have ignored a major methodological flaw in two large efficacy trials of PPV23, likely resulting in in an underestimation of the efficacy of PPV23 against PP. We believe that our work not only helps to resolve the discrepancies between previously published meta-analyses, but also highlights the importance of a meticulous appraisal of the risk of bias of published VE studies.


We systematically assessed the evidence on the efficacy/effectiveness of PPV23 against clinical endpoints in the elderly, employing the following steps:

  1. We reviewed all studies that were assessed in the most comprehensive systematic review published so far, including studies that ultimately did not meet the inclusion criteria for the meta-analysis [11].
  2. We then updated the literature search of that review, and meta-analyzed all relevant studies, excluding studies with a high risk of bias in a sensitivity analysis.
  3. Finally, we compared results of our meta-analysis with those of the other recently published reviews.

We followed the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement [12]. Our review protocol is available as appendix S1 Text.

Eligibility criteria

According to our predefined PICOS criteria (Population, Intervention, Comparator, Outcome, Study design, see Table 1), eligible studies had to be an original report on the efficacy or effectiveness of PPV23 in individuals aged 60 years and older. The control group had to have received placebo or no vaccine. We considered publications in which the specific clinical outcomes IPD or PP (or both) were assessed. We included clinical trials and observational studies, using the term vaccine efficacy for data from clinical trials, vaccine effectiveness for data from observational studies, and the abbreviation VE for either one or both, depending on context. Observational studies were only included if they reported VE estimates that were adjusted at least for age and comorbidities. No restrictions were made regarding publication language, and publication status. We excluded immunogenicity studies, studies with older PPV formulations containing more antigen per serotype (e.g. PPV14 with 50μg compared to 25μg in PPV23), and studies conducted in developing countries.

Updated literature search

We used the review from the Cochrane Collaboration [11] as starting point and conducted an update literature search for subsequently published studies in the databases MEDLINE, EMBASE, and Cochrane Central Register of Controlled Trials from 01.01.2011 to 02.07.2015 with an adapted search strategy (S1 Table). Two reviewers (GF and CR) independently screened titles, abstracts and full text articles. In addition, reference lists of all identified studies and reviews were reviewed for additional studies. In case of discordances regarding literature screening process, data extraction, and quality assessment a final decision was made by consensus or resolved by a third reviewer (TH). We updated the literature search on 15.07.2016 and did not find additional studies.

Data extraction

From each eligible study, two independent reviewers (GF and CR) extracted the following information using standardized forms: authors, publication year, study design, country, study population, number of participants, duration of follow-up, person-years of follow-up, reported outcomes, reported effect measure (RR; adjusted HR or OR), and funding. The extraction forms were pilot-tested with the first identified study of each study type and the field "person-years of follow-up" was added. The corresponding author of one study was contacted to clarify discrepancies in published data.

Assessment of risk of bias and quality of the body of evidence

We used the Cochrane Risk of Bias tool [13] to assess risk of bias in randomized controlled trials (RCTs) and the Newcastle-Ottawa Scale for observational studies [14]. For each study, risk of bias by outcome was independently assessed by two reviewers (GF and CR) and expressed as considered judgment as either “low”, “high” or “unclear”. We judged the overall quality of the body of evidence using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) Working Group [15] criteria. In GRADE, bodies of evidence from RCTs are a priori regarded as "high" quality evidence, whereas those from observational studies start as "low" quality evidence. Defined criteria are applied to up- or downgrade the quality of evidence which is finally expressed as “high”, “moderate”, “low” or “very low” [15].

Statistical analysis

Extracted data were aggregated in tables. Risk ratios (RR), adjusted odds ratios (aOR) and corresponding 95% confidence intervals (95% CIs) of clinical endpoints in the PPV23-vaccinated group and the control group were either directly extracted from the publications or calculated using person-years of follow-up as denominator. Vaccine efficacy/effectiveness (VE) was calculated as (1–RR)x100 or (1-aOR)x100, respectively. If in observational studies data for various periods of time since vaccination were reported, we used the data for a period of 5 years.

Meta-analysis using a random-effects model was performed by study design if data on a given outcome were available from more than one study. We used the software Review Manager (RevMan, version 5.2, Cochrane Collaboration), which offers two options for the statistical analysis of random-effects models: the Mantel-Haenszel method and the inverse variance method. In our case, both methods produced identical results. Between-study variation was estimated by comparing each study’s result with a Mantel-Haenszel fixed-effect meta-analysis result. For comparisons with zero events in any cell, the software automatically adds 0.5 to all cells. I-squared was used to quantify the extent of heterogeneity.

In the primary analysis, all eligible studies were included. According to the recommendations of the Cochrane Collaboration [16], we conducted sensitivity analyses including only studies with a low risk of bias. Testing for publication bias was not done since study numbers for each outcome were too small. The results of the GRADE evidence rating were recorded in GRADE evidence profiles using the GRADEpro software [17].


This systematic review was financed by the authors’ respective institutional budgets without extramural funding.


Selection of studies

Overall, 4 clinical trials (3 RCTs and one pseudo-randomized trial) and 13 observational studies were included (Fig 1). Of those, 7 studies derived from the review of the Cochrane Collaboration [11] and further 10 studies were identified with the updated literature search in electronic databases. Screening of the reference lists of included studies did not reveal any additional eligible studies.

Characteristics of included studies

The three RCTs [1820] were conducted between 1991 and 2009 in Sweden, Spain, and Japan and included 596 to 1006 participants (Table 2). The pseudo-randomized trial [21] was conducted in Finland and included almost 27,000 participants.

Table 2. Characteristics of studies included in the systematic review of PPV23 efficacy/effectiveness.

Five register-based cohort studies were conducted between 1998 and 2011 in Spain [22, 23], US [24, 25], and in Taiwan [24, 26] including 34,000 to 458,000 person-years of follow-up (Table 2).

Three case-control studies were conducted between 2001 and 2010 in Spain [27, 28], and Israel [29]. A variation of the case-control design, the so-called Broome method, was used in four studies from the UK [30, 31], Canada [32], and Spain [33], in total covering IPD surveillance data from 1995 to 2012. Across all case-control studies, 4320 episodes of IPD were included. With the Broome method [34], VE against IPD caused by vaccine serotypes (VT) is estimated by comparing vaccine uptake in patients with VT-IPD ("cases") and patients with non-VT-IPD ("controls"). As both groups consist of cases of disease, this study design is also known as "case-case study".

With a similar approach, one multi-country study [35] analyzed cases of PP, using pneumonia cases of other or unknown etiology as controls.

Reported outcomes

VE against the outcome IPD was reported in all but one [35] study. The clinical trials, cohort studies, and classical case-control studies reported all-serotype IPD (i.e., IPD caused by any pneumococcal serotype). Two of the case-control studies also reported VT-IPD. The 4 studies using the Broome method reported estimates of VE against VT-IPD.

VE against all-serotype PP was assessed in 4 trials, in 2 cohort studies [22, 23], and in 2 case-control studies [28, 35]. In these studies, PP was diagnosed by a range of methods, including a urine-antigen test which does not allow differentiation of pneumococcal serotypes. Therefore, VE against PP caused by vaccine serotypes could not be calculated.

Risk of bias assessment

Clinical trials.

For the outcome IPD, we rated the risk of bias as low for all clinical trials. In the pseudo-randomized trial by Honkanen et al. [21] group allocation was based on participants' year of birth (odd vs. even), and participants were not blinded as to their vaccination status. Moreover, they were offered to switch groups, which only 4.5% of participants did, however. It appears very unlikely that these methodological shortcomings decisively altered the chance of being diagnosed with IPD during the follow-up period.

For the outcome PP, we judged the studies by Örtqvist et al. and Honkanen et al. [20] [21] to have a high risk of bias: In these studies diagnosis of PP was made on the basis of detection of serum antibodies against pneumolysin using poorly validated in-house ELISA methods [36, 37]. These assays were later shown to have poor specificity, thus biasing the observed VE in a vaccine trial towards no effect (see Discussion for details).

Observational studies.

10 of 13 observational studies were judged to have low risk of bias [2224, 2729, 3133]. The remaining three studies were judged to bear a high risk of bias for the following reasons: In the study by Hechter et al. [25] participants were men who were voluntarily participating in a broader longitudinal study on men’s health (high risk of selection bias); in the study by Wiemken et al. [35] vaccination status of participants was not sufficiently validated (high risk of differential misclassification bias). In the study by Tsai et al. [26], VE against all-cause mortality was implausibly high at 93%, suggesting an over-estimation of VE (also against other outcomes) due to healthy vaccinee bias [38].

Vaccine efficacy/effectiveness

Outcome IPD.

Pooled analysis of all included clinical trials showed a VE of 73% (95% CI: 10–92%, I2 = 0%) against IPD with any serotype (Fig 2).

Fig 2. Forest plots of meta-analyses of randomized controlled trials, outcomes IPD and pneumococcal pneumonia.

IPD = invasive pneumococcal disease

PP = pneumococcal pneumonia

RCT = randomized controlled trial.

In cohort studies, pooled VE against IPD (any serotype) including all studies was 58% (95% CI: 38–72%, I2 = 11%), but decreased to 45% (95% CI: 15–65%, I2 = 0%) when studies with high risk of bias [25, 26] were excluded (Fig 3).

Fig 3. Forest plots of meta-analyses of observational studies, outcome IPD.

IPD = invasive pneumococcal disease

VT-IPD = vaccine-serotype IPD.

In case-control studies, pooled VE was 59% (95% CI: 35–74%, I2 = 60%) against IPD (any serotype). Heterogeneity is due to the lower VE observed in the study by Leventer-Roberts et al.. This study was conducted several years later than the other two studies, at a time when the proportion of vaccine-preventable serotypes among all IPD cases had probably already declined due to herd protection resulting from universal pneumococcal vaccination of infants. Effectiveness against vaccine-type IPD was only reported in the two older case-control studies with a pooled estimate of 73% (95% CI: 56–84%, I2 = 0%). Pooled analysis of case-case studies revealed VE of 37% (95% CI: 27–45%, I2 = 0%) against VT-IPD (Fig 3).

Outcome pneumococcal pneumonia (PP).

Pooled analysis of all included clinical trials showed a VE of 25% (95% CI: -62-65%) against PP (any serotype) with marked heterogeneity (I2 = 78%). After exclusion of studies with high risk of bias [20, 21], VE increased to 64% (95% CI: 35–80%) without heterogeneity (Fig 2).

Pooled analysis of the cohort studies showed a VE of 48% (95% CI: 25–63%, I2 = 0%) against PP. Of the remaining observational studies, only one case-control study [28] (VE 53%, 95% CI: 33–68%) and one case-case study with high risk-of-bias [35] (VE 37%, 95% CI: 12–55%) reported on that outcome (Fig 4).

Fig 4. Forest plots of meta-analyses of observational studies, outcome pneumococcal pneumonia.

PP = pneumococcal pneumonia.

Quality of the evidence

The quality of the evidence for both outcomes (IPD, PP) was assessed as moderate (the second highest level in the GRADE system) on the basis of data from RCTs (incl. the pseudo-randomized trial [21]). Reasons for downgrading by one level were wide CIs (GRADE criterion imprecision) in the case of IPD. For the outcome PP, the quality was downgraded because evidence is mainly based on one trial done in very old and frail nursing home residents [19] in whom the VE may be different (probably lower) than in the general population aged ≥60 years (GRADE criterion indirectness) (see S2 Table).

Comparison of this review with recently published meta-analyses

As shown in Table 3, previous systematic reviews of PPV23 efficacy/effectiveness have used different inclusion criteria for study selection and different outcomes. This partially explains why they reached divergent conclusions regarding clinical effectiveness of PPV23, in particular against pneumonia. Kraicer-Melamed et al. [8] used residence in nursing homes as an exclusion criterion, thereby excluding the Japanese RCT by Maruyama et al. [19]. Similarly, Schiffner-Rohe et al. [10] excluded that RCT in their stratified analysis. None of the review authors discusses the validity of serologic tests for the diagnosis of PP, as used in the trials by Örtqvist et al. [20] and Honkanen et al. [21].

Table 3. Overview of recent meta-analyses of PPV23 efficacy/effectiveness.


Our systematic literature review and meta-analysis revealed that PPV23 is effective against both IPD and PP (caused by any pneumococcal serotype) in the elderly. The point estimates of vaccine efficacy—derived from the meta-analysis of RCTs with low risk of bias—were 73% against IPD and 64% against PP (25% when including studies with high risk of bias). These estimates are supported by results from observational studies with low risk of bias. The pooled vaccine effectiveness against IPD (any serotype) in the first 5 years after vaccination was 45% in cohort and 59% in case-control studies, against PP it was 48% and 53%, respectively. These somewhat lower estimates may indicate waning of protection over the years, as the follow-up in the two RCTs of high quality lasted only 2.3 and 2.7 years, respectively, but on average 5 years in observational studies.

The question whether or not plain polysaccharide vaccines such as PPV23 can protect against PP is the subject of controversial discussions [4, 39, 40]. Historically, efficacy against PP has been clearly demonstrated in the 1970s in RCTs among workers in the gold mines in South Africa. In that population of young men, efficacy of 6- and 12-valent PPV against PP was 76% and 92%, respectively [6]. These results cannot be readily applied to the vaccination of an elderly population with the current 23-valent vaccine, but they provide a proof of principle. As further detailed below, efficacy/effectiveness of PPV23 against PP has been demonstrated in different settings and with different study designs, strongly suggesting that it is a real effect despite methodological limitations of individual studies.

Comparison with previous systematic reviews

Regarding efficacy against IPD, our results are in accordance with previous meta-analyses addressing this outcome [8, 11]. Regarding the outcome PP, the pooled VE of our meta-analysis of clinical trials with a low risk of bias was similar to that reported by Moberley et al. [11]. Contrarily, the latest meta-analyses [710] found no statistically significant VE against PP. Their estimates were driven by the trials by Örtqvist et al. [20] and Honkanen et al. [21], see Table 3.

We judged the trials by Örtqvist et al. and Honkanen et al. to have a high risk of bias regarding VE against the outcome PP, because diagnosis of PP was made by detection of antibodies against pneumolysin, a cholesterol-dependent cytotoxin produced by almost all strains of S. pneumoniae. In both trials, pneumolysin antibodies in serum (Ply-serum) and in circulating immune complexes (Ply-IC) were measured at the National Public Health Institute in Finland, using poorly validated in-house ELISA methods [36, 37]. These assays have not been used for the diagnosis of PP in any published study by other groups, nor have they become part of clinical routine. The main problem is their lack of specificity, which biases the observed effect in a VE study towards the null [39, 41, 42]. In the original publication by Jalonen et al. [36], specificity of the Ply-serum assay is not reported. In their validation study of the Ply-IC assay, Leinonen et al. [37] observed that geometric mean antibody titers measured with the Ply-serum ELISA were higher in healthy controls than in pneumonia patients, raising serious doubt about the specificity of the Ply-serum ELISA. For the Ply-IC assay, they report a specificity of 83%. However, as their healthy comparison group was much younger than their pneumonia patients, the validity of that specificity estimate is dubious.

The authors (including Leinonen) of a later validation study of the Ply-IC ELISA concluded that sensitivity and specificity of the assay were "insufficient for the performance of analytical epidemiological investigations or vaccine efficacy studies" [41]. A validation study of the Ply-IC ELISA by an independent group came to a similar result [42]. Moreover, that group showed that detection of antibodies to pneumolysin does not allow to differentiate between infection and mere colonization. Specificity of pneumolysin serology for the diagnosis of pneumococcal infection is further compromised by the fact that in addition to S. pneumoniae pneumolysin is expressed by other alpha-hemolytic streptococci such as S. viridans [43].

Another important difference between our and previous reviews is the in- or exclusion of the RCT by Maruyama et al. [19]. Kraicer-Melamed et al. [8] and Schiffner-Rohe et al. [10] excluded this trial, arguing that the study population of nursing home residents was not representative of the general elderly population. However, the same authors included the trial by Örtqvist et al. [20], which was carried out in patients who had recently been treated in hospital for pneumonia. The representativeness of these patients for the general elderly population is equally questionable.


Our meta-analysis of the efficacy against PP rests on only two RCTs with a low risk of bias and is dominated by the larger Japanese study by Maruyama et al. [19]. This study was undertaken in a population of very old, frail nursing home residents with an unusually high incidence of PP (32 per 1000 person-years in the placebo group). This study population is certainly not representative for the entire Japanese population aged 60 years and older. However, there is no biological reason to assume that the vaccine will be less effective in elderly people living outside nursing homes in Japan or other industrialized countries, as these vaccine recipients will be on average younger and have fewer comorbidities than nursing home residents. Furthermore, two register-based cohort studies among the resident elderly population in Tarragona, Spain, also showed a statistically significant VE against PP among persons vaccinated within the last 5 years [22, 23].

Another limitation for the interpretation of our data is the wide confidence intervals around the pooled VE estimates, leaving some uncertainty about the degree of protection. Also, the available data is insufficient to precisely determine the duration of protection afforded by PPV23.

Choosing the right vaccine

Most industrialized countries recommend routine pneumococcal vaccination for the elderly. In the USA, the Advisory Committee on Immunization Practices (ACIP) recommends sequential vaccination with PCV13 followed by PPSV23 [44], whereas in the UK the Joint Committee on Vaccination and Immunisation (JCVI) recommends PPV23 only [45]. In Europe, some countries recommend sequential vaccination, others the use of PPV23 or PCV13 only, yet others (e. g. France, The Netherlands) do not advocate routine vaccination of healthy elderly at all (

PCV13 was originally developed for young children whose immature immune system lacking splenic marginal zone B cells and circulating IgM+ memory B cells does not respond well to plain polysaccharide antigens during the first 2 years of life [46]. In 2011, PCV13 was also licensed for use in adults on the basis of immunogenicity studies. Its efficacy against clinical endpoints in immunocompetent elderly was subsequently examined in a single randomized placebo-controlled trial in the Netherlands (CAPITA trial [47]). In the modified intention-to-treat analysis of that trial, efficacy of PCV13 against IPD and PP caused by vaccine serotypes was 76% (95% CI: 47–90%) and 38% (95% CI: 14–55%), respectively. These estimates are similar to our pooled VE estimates of PPV23 efficacy against IPD and PP by any serotype. VE of PCV13 against IPD and PP by any serotype was lower, reaching only 49% (95% CI: 21–67%) and 22% (95% CI: 2–39%), respectively [47]. However, for two reasons the CAPITA trial might overestimate the VE of PCV13 in the general elderly population: (i) Persons with immunocompromising conditions and those residing in nursing homes were not eligible; therefore, the CAPITA study population was in better health and possibly mounted a better immune response to the vaccine than the overall elderly population. (ii) The trial was conducted in 2008–2012, before the introduction of PCV13 for infant vaccination in the Netherlands. Hence, VE against IPD and PP by any serotype was observed at a time, when the proportion of PCV13 serotypes among cases of all ages was still high. In countries using PCV13 for infant immunization, a marked reduction of IPD cases by PCV13 serotypes has been seen in all age groups due to herd protection, reducing the potential benefit of PCV13 for the elderly [4851]. For example in Germany, the proportion of PCV13 serotypes among IPD cases in ≥60 year old patients dropped from ~60% in the 2010/2011 season to ~30% in the 2015/2016 season, when still ~70% of cases were caused by serotypes included in PPV23 (

Data on the serotype distribution among cases of non-bacteremic PP is scarce, because often no isolate is available for serotyping. Serotype-specific assays for the detection of pneumococcal antigens in urine have only recently been developed and are so far limited to the 13 serotypes contained in PCV13 [52, 53]. A study of non-bacteremic PP cases in adults (median age 71 years) in Nottingham/England has shown a 30% reduction of the proportion of PCV13 serotypes within 3 years of the switch from PCV7 to PCV13 in the infant immunization program [54]. In a similar study in Germany, 79% of bacteremic PP and 62% of non-bacteremic PP cases were caused by PCV13 serotypes in the period 2007–2011, i. e. 3 years before and 2 years after the switch from PCV7 to PCV13 [55]. If anything, it appears that PCV13 serotypes are less prevalent among non-bacteremic PP than among bacteremic PP cases. Data covering a more recent time period is highly desirable to judge the further impact of infant immunization with PCV13 on serotype distribution in adult non-bacteremic PP.


Our systematic review and meta-analysis indicates that PPV23 is effective against IPD and pneumococcal pneumonia in the elderly. In view of its broader serotype coverage compared to PCV13, PPV23 should be recommended for routine vaccination of the elderly. Sequential vaccination with PCV13 followed by PPV23 may be justified in countries where a large proportion of pneumococcal disease in the elderly is caused by PCV13 serotypes.

Regarding future research, an RCT directly comparing the efficacy of different vaccination strategies (PPSV23 only, PCV13 only, and sequential vaccination) on clinical endpoints is highly desirable. In addition, more data on the duration of protection by either vaccine as well as data on the optimal age for vaccinating elderly people would be useful.

Author Contributions

  1. Conceptualization: GF CR TH EHP OW CB.
  2. Formal analysis: GF CR TH.
  3. Investigation: GF CR TH EHP CB.
  4. Methodology: GF CR TH EHP OW CB.
  5. Project administration: GF.
  6. Resources: OW.
  7. Supervision: CB.
  8. Validation: GF CR EHP CB.
  9. Visualization: GF CR.
  10. Writing – original draft: GF CR.
  11. Writing – review & editing: TH EHP OW CB.


  1. 1. Rozenbaum MH, Pechlivanoglou P, van der Werf TS, Lo-Ten-Foe JR, Postma MJ, Hak E. The role of Streptococcus pneumoniae in community-acquired pneumonia among adults in Europe: a meta-analysis. European journal of clinical microbiology & infectious diseases. 2013;32(3):305–16.
  2. 2. Klapdor B, Ewig S, Pletz MW, Rohde G, Schutte H, Schaberg T, et al. Community-acquired pneumonia in younger patients is an entity on its own. The European respiratory journal. 2012;39(5):1156–61. pmid:22088967
  3. 3. Harboe ZB, Thomsen RW, Riis A, Valentiner-Branth P, Christensen JJ, Lambertsen L, et al. Pneumococcal serotypes and mortality following invasive pneumococcal disease: a population-based cohort study. PLoS Med. 2009;6(5):e1000081. pmid:19468297
  4. 4. Fedson DS. Preventing non bacteremic pneumococcal pneumonia in older adults: historical background and considerations for choosing between PCV13 and PPV23. Hum Vaccin Immunother. 2014;10(5):1322–30. pmid:24732438
  5. 5. Austrian R, Douglas RM, Schiffman G, Coetzee AM, Koornhof HJ, Hayden-Smith S, et al. Prevention of pneumococcal pneumonia by vaccination. Transactions of the Association of American Physicians. 1976;89:184–94. pmid:14433
  6. 6. Smit P, Oberholzer D, Hayden-Smith S, Koornhof HJ, Hilleman MR. Protective efficacy of pneumococcal polysaccharide vaccines. JAMA. 1977;238(24):2613–6. pmid:21973
  7. 7. Diao WQ, Shen N, Yu PX, Liu BB, He B. Efficacy of 23-valent pneumococcal polysaccharide vaccine in preventing community-acquired pneumonia among immunocompetent adults: A systematic review and meta-analysis of randomized trials. Vaccine. 2016;34(13):1496–503. pmid:26899376
  8. 8. Kraicer-Melamed H, O'Donnell S, Quach C. The effectiveness of pneumococcal polysaccharide vaccine 23 (PPV23) in the general population of 50 years of age and older: A systematic review and meta-analysis. Vaccine. 2016;34(13):1540–50. pmid:26899372
  9. 9. Kraicer-Melamed H, O'Donnell S, Quach C. Corrigendum to "The effectiveness of pneumococcal polysaccharide vaccine 23 (PPV23) in the general population of 50years of age and older: A systematic review and meta-analysis" [Vaccine 34 (2016) 1540–1550]. Vaccine. 2016;34(34):4083–4. pmid:27329185
  10. 10. Schiffner-Rohe J, Witt A, Hemmerling J, von Eiff C, Leverkus FW. Efficacy of PPV23 in Preventing Pneumococcal Pneumonia in Adults at Increased Risk—A Systematic Review and Meta-Analysis. PloS One. 2016;11(1):e0146338. pmid:26761816
  11. 11. Moberley S, Holden J, Tatham DP, Andrews RM. Vaccines for preventing pneumococcal infection in adults. The Cochrane database of systematic reviews. 2013;1:Cd000422.
  12. 12. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. pmid:19621072
  13. 13. Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. pmid:22008217
  14. 14. Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses [10.07.2015].
  15. 15. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables. Journal of clinical epidemiology. 2011;64(4):383–94. pmid:21195583
  16. 16. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0: The Cochrane Collaboration; 2011.
  17. 17. Guyatt GH, Thorlund K, Oxman AD, Walter SD, Patrick D, Furukawa TA, et al. GRADE guidelines: 13. Preparing summary of findings tables and evidence profiles-continuous outcomes. Journal of clinical epidemiology. 2013;66(2):173–83. pmid:23116689
  18. 18. Alfageme I, Vazquez R, Reyes N, Munoz J, Fernandez A, Hernandez M, et al. Clinical efficacy of anti-pneumococcal vaccination in patients with COPD. Thorax. 2006;61(3):189–95. pmid:16227328
  19. 19. Maruyama T, Taguchi O, Niederman MS, Morser J, Kobayashi H, Kobayashi T, et al. Efficacy of 23-valent pneumococcal vaccine in preventing pneumonia and improving survival in nursing home residents: double blind, randomised and placebo controlled trial. BMJ. 2010;340:c1004. pmid:20211953
  20. 20. Örtqvist A, Hedlund J, Burman LA, Elbel E, Hofer M, Leinonen M, et al. Randomised trial of 23-valent pneumococcal capsular polysaccharide vaccine in prevention of pneumonia in middle-aged and elderly people. Swedish Pneumococcal Vaccination Study Group. Lancet. 1998;351(9100):399–403. pmid:9482293
  21. 21. Honkanen PO, Keistinen T, Miettinen L, Herva E, Sankilampi U, Laara E, et al. Incremental effectiveness of pneumococcal vaccine on simultaneously administered influenza vaccine in preventing pneumonia and pneumococcal pneumonia among persons aged 65 years or older. Vaccine. 1999;17(20–21):2493–500. pmid:10418894
  22. 22. Ochoa-Gondar O, Vila-Corcoles A, Rodriguez-Blanco T, Gomez-Bertomeu F, Figuerola-Massana E, Raga-Luria X, et al. Effectiveness of the 23-valent pneumococcal polysaccharide vaccine against community-acquired pneumonia in the general population aged >/ = 60 years: 3 years of follow-up in the CAPAMIS study. Clin Infect Dis. 2014;58(7):909–17. pmid:24532544
  23. 23. Vila-Corcoles A, Ochoa-Gondar O, Hospital I, Ansa X, Vilanova A, Rodriguez T, et al. Protective effects of the 23-valent pneumococcal polysaccharide vaccine in the elderly population: the EVAN-65 study. Clin Infect Dis. 2006;43(7):860–8. pmid:16941367
  24. 24. Jackson LA, Neuzil KM, Yu O, Benson P, Barlow WE, Adams AL, et al. Effectiveness of pneumococcal polysaccharide vaccine in older adults. NEJM. 2003;348(18):1747–55. pmid:12724480
  25. 25. Hechter RC, Chao C, Jacobsen SJ, Slezak JM, Quinn VP, Van Den Eeden SK, et al. Clinical effectiveness of pneumococcal polysaccharide vaccine in men: California Men's Health Study. Vaccine. 2012;30(38):5625–30. pmid:22789510
  26. 26. Tsai YH, Hsieh MJ, Chang CJ, Wen YW, Hu HC, Chao YN, et al. The 23-valent pneumococcal polysaccharide vaccine is effective in elderly adults over 75 years old-Taiwan's PPV vaccination program. Vaccine. 2015;33(25):2897–902. pmid:25936662
  27. 27. Dominguez A, Salleras L, Fedson DS, Izquierdo C, Ruiz L, Ciruela P, et al. Effectiveness of pneumococcal vaccination for elderly people in Catalonia, Spain: a case-control study. Clin Infect Dis. 2005;40(9):1250–7. pmid:15825026
  28. 28. Vila-Corcoles A, Salsench E, Rodriguez-Blanco T, Ochoa-Gondar O, de Diego C, Valdivieso A, et al. Clinical effectiveness of 23-valent pneumococcal polysaccharide vaccine against pneumonia in middle-aged and older adults: a matched case-control study. Vaccine. 2009;27(10):1504–10. pmid:19171174
  29. 29. Leventer-Roberts M, Feldman BS, Brufman I, Cohen-Stavi CJ, Hoshen M, Balicer RD. Effectiveness of 23-valent pneumococcal polysaccharide vaccine against invasive disease and hospital-treated pneumonia among people aged >/ = 65 years: a retrospective case-control study. Clin Infect Dis. 2015;60(10):1472–80. pmid:25669354
  30. 30. Andrews NJ, Waight PA, George RC, Slack MP, Miller E. Impact and effectiveness of 23-valent pneumococcal polysaccharide vaccine against invasive pneumococcal disease in the elderly in England and Wales. Vaccine. 2012;30(48):6802–8. pmid:23000122
  31. 31. Wright LB, Hughes GJ, Chapman KE, Gorton R, Wilson D. Effectiveness of the 23-valent pneumococcal polysaccharide vaccine against invasive pneumococcal disease in people aged 65 years and over in the North East of England, April 2006–July 2012. Trials in Vaccinology. 2013;2(0):45–8.
  32. 32. Rudnick W, Liu Z, Shigayeva A, Low DE, Green K, Plevneshi A, et al. Pneumococcal vaccination programs and the burden of invasive pneumococcal disease in Ontario, Canada, 1995–2011. Vaccine. 2013;31(49):5863–71. pmid:24099873
  33. 33. Gutierrez Rodriguez MA, Ordobas Gavin MA, Garcia-Comas L, Sanz Moreno JC, Cordoba Deorador E, Lasheras Carbajo MD, et al. Effectiveness of 23-valent pneumococcal polysaccharide vaccine in adults aged 60 years and over in the Region of Madrid, Spain, 2008–2011. Eurosurveillance. 2014;19(40):20922. pmid:25323079
  34. 34. Broome CV, Facklam RR, Fraser DW. Pneumococcal disease after pneumococcal vaccination: an alternative method to estimate the efficacy of pneumococcal vaccine. NEJM. 1980;303(10):549–52. pmid:6995835
  35. 35. Wiemken TL, Carrico RM, Klein SL, Jonsson CB, Peyrani P, Kelley RR, et al. The effectiveness of the polysaccharide pneumococcal vaccine for the prevention of hospitalizations due to Streptococcus pneumoniae community-acquired pneumonia in the elderly differs between the sexes: results from the Community-Acquired Pneumonia Organization (CAPO) international cohort study. Vaccine. 2014;32(19):2198–203. pmid:24613522
  36. 36. Jalonen E, Paton JC, Koskela M, Kerttula Y, Leinonen M. Measurement of antibody responses to pneumolysin—a promising method for the presumptive aetiological diagnosis of pneumococcal pneumonia. The Journal of infection. 1989;19(2):127–34. pmid:2809235
  37. 37. Leinonen M, Syrjälä H, Jalonen E, Kujala P, Herva E. Demonstration of pneumolysin antibodies in circulating immune complexes—a new diagnostic method for pneumococcal pneumonia. Serodiagnosis and Immunotherapy in Infectious Disease. 1990;4(6):451–8.
  38. 38. Nelson JC, Jackson ML, Weiss NS, Jackson LA. New strategies are needed to improve the accuracy of influenza vaccine effectiveness estimates among seniors. Journal of clinical epidemiology. 2009;62(7):687–94. pmid:19124221
  39. 39. Jackson LA. Pneumococcal polysaccharide vaccines. In: Plotkin S, Orenstein W, Offit P, editors. Vaccines. 6th ed: Elsevier Saunders; 2013.
  40. 40. Hollingsworth R, Isturiz R. The stubborn persistence of adult pneumococcal pneumonia as a public health problem. Hum Vaccin Immunother. 2014;10(5):1319–21. pmid:24553362
  41. 41. Scott JA, Hall AJ, Leinonen M. Validation of immune-complex enzyme immunoassays for diagnosis of pneumococcal pneumonia among adults in Kenya. Clinical and diagnostic laboratory immunology. 2000;7(1):64–7. pmid:10618279
  42. 42. Musher DM, Mediwala R, Phan HM, Chen G, Baughn RE. Nonspecificity of assaying for IgG antibody to pneumolysin in circulating immune complexes as a means to diagnose pneumococcal pneumonia. Clin Infect Dis. 2001;32(4):534–8. pmid:11181114
  43. 43. Kaijalainen T, Saukkoriipi A, Bloigu A, Herva E, Leinonen M. Real-time pneumolysin polymerase chain reaction with melting curve analysis differentiates pneumococcus from other alpha-hemolytic streptococci. Diagn Microbiol Infect Dis. 2005;53(4):293–9. pmid:16360553
  44. 44. Tomczyk S, Bennett NM, Stoecker C, Gierke R, Moore MR, Whitney CG, et al. Use of 13-valent pneumococcal conjugate vaccine and 23-valent pneumococcal polysaccharide vaccine among adults aged >/ = 65 years: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Morbidity and mortality weekly report. 2014;63(37):822–5. pmid:25233284
  45. 45. Joint Committee on Vaccination and Immunisation (JCVI). Interim JCVI statement on adult pneumococcal vaccination in the UK. November 2015.
  46. 46. Iyer AS, Ohtola JA, Westerink MA. Age-related immune response to pneumococcal polysaccharide vaccination: lessons for the clinic. Expert review of vaccines. 2015;14(1):85–97. pmid:25269650
  47. 47. Bonten MJ, Huijts SM, Bolkenbaas M, Webber C, Patterson S, Gault S, et al. Polysaccharide Conjugate Vaccine against Pneumococcal Pneumonia in Adults. NEJM. 2015;372(12):1114–25. pmid:25785969
  48. 48. Harboe ZB, Dalby T, Weinberger DM, Benfield T, Molbak K, Slotved HC, et al. Impact of 13-valent pneumococcal conjugate vaccination in invasive pneumococcal disease incidence and mortality. Clin Infect Dis. 2014;59(8):1066–73. pmid:25034421
  49. 49. Moore MR, Link-Gelles R, Schaffner W, Lynfield R, Lexau C, Bennett NM, et al. Effect of use of 13-valent pneumococcal conjugate vaccine in children on invasive pneumococcal disease in children and adults in the USA: analysis of multisite, population-based surveillance. The Lancet Infectious Diseases. 2015;15(3):301–9. pmid:25656600
  50. 50. van der Linden M, Falkenhorst G, Perniciaro S, Imöhl M. Effects of infant pneumococcal conjugate vaccination on serotype distribution in invasive pneumococcal disease (IPD) among children and adults in Germany. PloS One. 2015;10(7):e0131494. pmid:26132078
  51. 51. Waight PA, Andrews NJ, Ladhani SN, Sheppard CL, Slack MPE, Miller E. Effect of the 13-valent pneumococcal conjugate vaccine on invasive pneumococcal disease in England and Wales 4 years after its introduction: an observational cohort study. The Lancet Infectious Diseases. 2015(0).
  52. 52. Pride MW, Huijts SM, Wu K, Souza V, Passador S, Tinder C, et al. Validation of an immunodiagnostic assay for detection of 13 Streptococcus pneumoniae serotype-specific polysaccharides in human urine. Clinical and vaccine immunology. 2012;19(8):1131–41. pmid:22675155
  53. 53. Sheppard CL, Harrison TG, Smith MD, George RC. Development of a sensitive, multiplexed immunoassay using xMAP beads for detection of serotype-specific streptococcus pneumoniae antigen in urine samples. Journal of medical microbiology. 2011;60(Pt 1):49–55. pmid:20864547
  54. 54. Rodrigo C, Bewick T, Sheppard C, Greenwood S, McKeever TM, Trotter CL, et al. Impact of infant 13-valent pneumococcal conjugate vaccine on serotypes in adult pneumonia. The European respiratory journal. 2015;45(6):1632–41. pmid:25792633
  55. 55. Pletz MW, Ewig S, Rohde G, Schuette H, Rupp J, Welte T, et al. Impact of pneumococcal vaccination in children on serotype distribution in adult community-acquired pneumonia using the serotype-specific multiplex urinary antigen detection assay. Vaccine. 2016;34(20):2342–8. pmid:27016653