Global Antimicrobial Non-Susceptibility Patterns of Streptococcus pneumoniae: A Systematic Literature Review (2014–2023) Stratified by Age, Region, and Infection Type Authors: Mejbah Uddin Bhuiyan, John A Carter, Victoria Abbing-Karahagopian Published: 2026 ABSTRACT AbstractAntimicrobial non-susceptibility of Streptococcus pneumoniae is an important global health challenge. While pneumococcal vaccination has significantly lowered the burden of pneumococcal disease, including that caused by antibiotic-resistant strains, non-susceptible S. pneumoniae isolates are still important contributors to morbidity and mortality, complicating treatment. Empiric antibiotic selection for treating pneumococcal disease should consider the local epidemiology and non-susceptibility profile of isolates. To comprehensively characterize the distribution of non-susceptibility to commonly used antibiotics of S. pneumoniae isolates across different age groups, geographical regions and infection types, we conducted a systematic literature review in accordance with PRISMA guidelines. We searched PubMed, Medline, and Embase to retrieve studies published between 2014 and 2023 reporting the proportion of non-susceptible S. pneumoniae isolates detected in patients with pneumococcal disease. Data were analyzed overall and stratified by age groups (pediatric and adult), World Health Organization region, and infection type (any invasive pneumococcal disease, meningitis, bacteremic pneumonia, non-bacteremic pneumonia, bacteremia/septicemia, and acute otitis media). Study quality was assessed using the Newcastle-Ottawa Scale. Due to observed data heterogeneity, meta-analysis was not performed. Of the 1200 publications identified, 128 (113 full-text articles and 15 conference abstracts) were included in the review, comprising antimicrobial susceptibility testing outcomes for >250,000 isolates. Globally, the highest non-susceptibility rates were observed for sulfonamides (52%), macrolides (40%), and lincosamides (30%). Higher rates were reported in the Western Pacific region, with ≥91% of isolates non-susceptible to commonly used antibiotics (lincosamides, tetracyclines, macrolides, and sulfonamides); elevated non-susceptibility to tetracyclines and sulfonamides was also observed in the Southeast Asian region. Non-susceptibility proportions, in general, were higher in the pediatric population than in adults, and they varied markedly by pneumococcal infection type. Up-to-date, stratified information on the non-susceptibility profile of S. pneumoniae isolates causing different types of infection can orient local prescription guidance to use appropriate antibiotics, thereby improving the treatment outcomes and reducing the potential threat of antibiotic resistance. Introduction Pneumococcal disease comprises infections caused by the bacterium Streptococcus pneumoniae (pneumococcus), which include bacteremia, pneumonia, meningitis, sinusitis, and acute otitis media (AOM).1,2 While pneumococcal disease can occur in all age groups, some population groups, such as infants, young children, the elderly, and individuals with certain underlying medical conditions are at an increased risk.1,3 Out of >100 pneumococcal serotypes identified to date, 20 to 30 are responsible for most cases of pneumococcal disease worldwide.2,4 Antibiotics remain the primary choice of therapy for bacterial infections, including pneumococcal disease; however, antimicrobial-resistant pneumococcal strains have been increasingly reported, becoming a global health concern.5–7 S. pneumoniae strains resistant to various antibiotics contribute significantly to morbidity and mortality, complicate the treatment plan and increase healthcare costs.3,5–7 In 2024, the World Health Organization (WHO) included macrolide-resistant S. pneumoniae on its list of bacterial priority pathogens, which constitute the focus of surveillance and control of antibacterial resistance programs.8 In S. pneumoniae, key drivers of antimicrobial resistance include inappropriate and/or excessive use of antibiotics,9,10 as well as selective pressure from pneumococcal vaccination.11 While vaccination has substantially reduced the burden of pneumococcal disease caused by vaccine serotypes (including antimicrobial-resistant isolates), it has also led to the emergence of non-vaccine serotypes with distinct resistance profiles.12–17 These strains can act as a genetic reservoir for antimicrobial resistance due to the ability of pneumococci to undergo horizontal gene transfer and recombination.11 Genomic epidemiology approaches such as multilocus sequence typing and whole-genome sequencing have increasingly been used to observe the evolution and dissemination of antimicrobial resistance in S. pneumoniae. These methods allow for a precise tracking of resistant clones and provide important context for interpreting the observed resistance patterns. Pneumococcal resistance to antimicrobials is class-specific and arises primarily through drug target-site modifications (eg, mutations in the penicillin-binding protein or the ermB gene) or alteration of efflux systems (eg, mutations in the mefE or mefA genes).3 However, mapping of resistance patterns determined using phenotypic methods across various settings remains an essential tool for orienting clinical decision-making. Pneumococcal disease surveillance data demonstrated a considerable geographical variability in antimicrobial resistance, with increasing macrolide resistance reported in many regions, and a generally low (but rising) resistance to fluoroquinolones.14,18–20 The overall prevalence of multidrug-resistant S. pneumoniae isolates (ie, resistant to ≥3 antimicrobial classes) ranged from 15% in Europe to 36% in Asia.3 Resistance patterns also vary between invasive pneumococcal disease (IPD, such as bacteremia, bacteremic pneumonia, meningitis) and non-IPD (such as AOM, sinusitis, and non-bacteremic pneumonia).18,21–23 For a suspected pneumococcal meningitis, a combination therapy with vancomycin and a third-generation cephalosporin is frequently recommended,3,19 whereas beta-lactam antibiotics like amoxicillin or macrolides are commonly used for pneumococcal community-acquired pneumonia (CAP); however, these recommendations may not be optimal in settings with high circulation of non-susceptible strains.3,24 Interpretation of resistance patterns is further complicated by variability in study design, antimicrobial susceptibility testing methods, and breakpoint definitions in the various settings. The available evidence suggests that the choice of antibiotics must rely on the local epidemiology and non-susceptibility profiles of isolates causing pneumococcal disease. Previous systematic reviews have documented global and regional trends in resistance, reporting variable but often high prevalence of antibiotic non-susceptibility rates, depending on geographic region, antimicrobial class, and circulating pneumococcal isolates and/or serotypes.25–28 These reviews either focused only on the pediatric population, certain infection types, or specific antimicrobial classes, and therefore a comprehensive, contemporary comparison stratified simultaneously by age group, infection type, and WHO region was lacking. This systematic literature review, by integrating data from the published literature, covering all age groups and geographic regions, and applying consistent stratification, addresses this important knowledge gap and provides a more detailed understanding of how S. pneumoniae non-susceptibility varies across clinically relevant subpopulations and settings. As resistance patterns continue to evolve, such up-to-date, stratified data can be useful to inform and guide empiric antibiotic selection in treating pneumococcal disease. Materials and Methods This systematic literature review aimed to assess the proportion of S. pneumoniae isolates displaying antimicrobial non-susceptibility (including both intermediate resistant and resistant isolates) that were associated with symptomatic pneumococcal infection in pediatric and adult patients. The systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines.29 Search Strategy PubMed, Medline, and Embase were searched using a comprehensive multi-string search strategy to identify all relevant articles published between January 2014–December 2023 and conference abstracts published between 2021–2023 that reported antimicrobial susceptibility and non-susceptibility of S. pneumoniae. The search was performed using the ProQuest platform.30 The search strategy used to identify relevant literature is available in Supplementary Table 1. Eligibility Criteria for Study Selection The search output and screening of articles were managed using EndNote. We included epidemiological observational studies that reported proportions of susceptible and/or non-susceptible S. pneumoniae isolates detected in pediatric and adult patients with different pneumococcal infections, including IPD, ie any IPD without explicit mention of clinical condition (further labeled as “any IPD”), bacteremia, septicemia, meningitis, or bacteremic CAP, and non-IPD, such as non-bacteremic CAP or AOM. Articles published in other languages than English, asymptomatic nasopharyngeal carriage studies, interventional studies, case reports, narrative review articles, editorials, comments, and correspondence were excluded. To minimize the risk of selection bias, two reviewers independently screened the titles and abstracts of all identified articles based on pre-defined eligibility criteria. Any discrepancies between the reviewers were resolved by a third reviewer. In addition, an artificial intelligence (AI) platform, Nested Knowledge AI,31 was used as a shared environment for multiple researchers to conduct study selection, while offering functionalities such as record deduplication and the highlighting of key words, phrases, and concepts. Employing the AI platform added another layer of quality control on top of the work completed by the independent reviewers as described above. Data Extraction Key study characteristics were extracted from each eligible study using a pre-defined data extraction template. Extracted data included authors, publication year, study location(s), data collection timeline, study design, study population (size, age group), infection type from which S. pneumoniae isolates were detected, number of S. pneumoniae isolates detected and tested for antimicrobial susceptibility, name of the antimicrobial(s) tested, and outcome of antimicrobial testing, ie, number of S. pneumoniae isolates susceptible and non-susceptible to a specific antimicrobial. Antimicrobial susceptibility was defined as reported in the source publication, either based on the guidelines from the Clinical and Laboratory Standards Institute or the European Committee on Antimicrobial Susceptibility Testing criteria, or on assay data that allowed classification as susceptible or non-susceptible in the respective study setting. For studies that reported antimicrobial susceptibility outcomes for multiple years, data for each year were entered separately. The specific antimicrobials tested in the included studies were categorized under the broader antimicrobial class to which the antimicrobial belongs to (eg, beta-lactam carbapenems, beta-lactam penicillins, beta-lactam cephalosporins, macrolides, or fluoroquinolones, Supplementary Table 2). Data on the number of susceptible, intermediate resistant, and resistant isolates (as reported in the eligible studies) were extracted. The non-susceptible isolates included both intermediate resistant and resistant isolates. Data were extracted by a single reviewer, and a confirmatory review of selected data was subsequently performed by a second reviewer. Any discrepancies between the two reviewers were resolved by a third senior reviewer. All extracted data were collated within tables in Microsoft Excel. Data unavailable in numerical form (eg, only presented in figures) were digitized using DigitizeIt,32 a program demonstrated to reliably and accurately extract data from graphs.33 Data Analysis All analyses were performed with Microsoft Excel and were descriptive in nature. The proportion of S. pneumoniae isolates non-susceptible to a specific antimicrobial class was calculated by dividing the total number of non-susceptible isolates to a given antimicrobial class, by the total number of isolates tested for antimicrobial susceptibility. For each clinical condition of interest, the proportion of S. pneumoniae isolates not susceptible to various antimicrobial classes was assessed; here, we report non-susceptibility data for the eight antimicrobial classes most commonly tested: beta-lactams (carbapenems, cephalosporins, and penicillins), fluoroquinolones, lincosamides, macrolides, sulfonamides, and tetracyclines. Data retrieved for other antimicrobial classes are included in Supplementary Tables 3–7. Within each clinical condition, the proportion of non-susceptible isolates associated with an infection was further stratified by age group (ie, pediatric [<18 years] and adult [≥18 years]), and by WHO regions (African region [AFR], region of the Americas [AMR], Southeast Asian region [SEAR], European region [EUR], Eastern Mediterranean region [EMR], and the Western Pacific region [WPR]). All results were reported as the percentage of non-susceptible isolates. Two-sided 95% confidence intervals (CI) for proportions were calculated using the Wald normal approximation; when this method produced bounds outside the 0–100% range (eg, due to small sample sizes), Wilson score 95% CIs were used instead. For clarity and readability of the text, the 95% CIs are presented in the Supplementary Tables. Due to the inherent heterogeneity in population selection, sampling methods, and antimicrobial testing assays, no meta-analysis was performed. Quality assessment for the studies included in the analysis was conducted using an adapted version of the Newcastle-Ottawa Scale for each study type,34 by using the following domains: Case definition, Representativeness (of exposed cohort), Selection of controls, Definition of controls, Comparability (cases vs controls), Ascertainment of exposure, Ascertainment for cases and controls, Assessment of outcomes, and Adequacy of follow-up. Each included study was assessed by one reviewer, and all quality scores were checked by a senior reviewer. Discrepancies were adjudicated by consensus. Results Search Outcome and Study Characteristics The ProQuest search strategy resulted in the identification of 1200 articles published between 2014 and 2023 that underwent title/abstract screening. A total of 235 articles were selected for full-text screening, of which 113 full-text articles were included in the review. In addition, 23 conference abstracts were identified from various sources, of which 15 were found eligible and were included in the review. Altogether, 128 publications reporting antimicrobial susceptibility and/or non-susceptibility of S. pneumoniae isolates were included in the review (Figure 1).35–162 The studies were conducted in 50 countries located across all six WHO regions; one study was conducted in multiple settings. All eligible full-text publications were rated as moderate-to-high quality. Figure 1PRISMA flow diagram.This PRISMA flow diagram details the screening process. In the identification phase, 1200 articles were identified from ProQuest without duplicates. In the screening phase, these 1200 articles underwent title and abstract screening, with 965 records excluded. Full-text screening involved 235 articles, of which, 122 articles were excluded for reasons such as outcome (77), population (33), publication type (1), full text not accessible (6) and non-English language (5). Additionally, 23 conference abstracts were identified, and 8 were excluded. In the inclusion phase, 128 records were selected for data extraction, comprising 113 articles and 15 conference abstracts.A PRISMA flow diagram showing article and abstract screening process.Abbreviation: PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses. PRISMA flow diagram. Antimicrobial Non-Susceptibility of S. pneumoniae Isolates Distribution of Non-Susceptible Pneumococcal Isolates The analysis included a total of 256,012 antimicrobial susceptibility testing outcomes of pneumococcal isolates identified from pediatric and adult patients with any of the pneumococcal infections of our interest. Globally, 52% of isolates (9606/18,375 isolates) were non-susceptible to sulfonamides, 40% (21,524/53,609) to macrolides, 30% (4163/13,773) to lincosamides, and 29% (5082/17,707) to tetracyclines. With regards to beta-lactam antibiotics, 31% (1000/3217) of isolates were non-susceptible to carbapenems, 26% (18,459/72,098) to penicillins, and 10% (3943/38,290) to cephalosporins (Figure 2). Figure 2Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in patients with any pneumococcal infection, by WHO region. Global N values are bold.The figure compares pneumococcal non-susceptibility rates for beta-lactam carbapenems, beta-lactam cephalosporins, beta-lactam penicillins, fluoroquinolones, lincosamides, macrolides, sulfonamides, and tetracyclines across six WHO regions (AFR, AMR, EMR, EUR, SEAR, and WPR) and globally. Non-susceptibility rates varied substantially by antibiotic classes and regions, with the highest rates generally observed for sulfonamides, tetracyclines, and macrolides, particularly in the Western Pacific and South-East Asia regions. In contrast, non-susceptibility to fluoroquinolones was generally low across regions. Sample sizes for each antibiotic class and region are indicated beside the bars.A horizontal bar graph showing proportion (%) of non-susceptible pneumococcal isolates by antibiotic classes and WHO regions.Abbreviations: AFR, African region; AMR, region of the Americas; EMR, Eastern Mediterranean region; EUR, European region; N, number of non-susceptible isolates; NR, not reported; SEAR, Southeast Asian region; WHO, World Health Organization; WPR, Western Pacific region. Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in patients with any pneumococcal infection, by WHO region. Global N values are bold. Among the WHO regions, high proportions of isolates non-susceptible to several antibiotics, such as lincosamides (91%, 1623/1787), tetracyclines (88%, 1638/1867), macrolides (88%, 5235/5931), and sulfonamides (73%, 1809/2486), were reported in WPR. In AFR, EMR, and EUR, approximately 40% of isolates were not susceptible to penicillins. Globally and across most regions, the non-susceptibility rate was very rare for glycopeptides (vancomycin) and oxazolidinones (linezolid). Distribution of Non-Susceptible Pneumococcal Isolates Associated with Specific Pneumococcal Infections, by Age Groups The proportion of pneumococcal isolates not susceptible to commonly prescribed antimicrobials identified in patients with specific pneumococcal infections was assessed separately in the pediatric and adult populations (Figure 3A and B). Figure 3Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in pediatric (3A) and adult (3B) patients with different pneumococcal infections. Error bars indicate 95% confidence intervals. Colored dots indicate that susceptibility data were not reported for a specific antimicrobial class. Where the non-susceptibility proportion is 0%, no bar is visible, and there is no colored dot.This figure shows a comparison of antimicrobial non-susceptibility among Streptococcus pneumoniae isolates causing different pneumococcal diseases in pediatric (panel A) and adult (panel B) populations. Non-susceptibility rates are shown for eight antibiotic classes (i.e, beta-lactam carbapenems, beta-lactam cephalosporins, beta-lactam penicillins, fluoroquinolones, lincosamides, macrolides, sulfonamides, and tetracyclines) across different infection types (i.e, invasive pneumococcal disease, meningitis, bacteremic pneumonia, non-bacteremic pneumonia, bacteremia/septicemia, and acute otitis media [only in the pediatric population]). Error bars represent 95% confidence intervals around the estimates. Non-susceptibility patterns differed by age groups, infection types, and antibiotic classes, with generally higher rates observed for lincosamides, macrolides, sulfonamides, and tetracyclines, while beta-lactam carbapenem and fluoroquinolone non-susceptibility remained comparatively low globally.A grouped bar graph showing the proportion (%) of non-susceptible Streptococcus pneumoniae isolates in pediatric and adult pneumococcal infections.Abbreviations: AOM, acute otitis media; CAP, community-acquired pneumonia; IPD, invasive pneumococcal disease. Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in pediatric (3A) and adult (3B) patients with different pneumococcal infections. Error bars indicate 95% confidence intervals. Colored dots indicate that susceptibility data were not reported for a specific antimicrobial class. Where the non-susceptibility proportion is 0%, no bar is visible, and there is no colored dot. A total of 79 publications reporting 109,268 antimicrobial susceptibility testing outcomes of pneumococcal isolates identified among pediatric patients were included in the analysis. In this age group, non-susceptibility to tetracycline was frequently observed: in 92% (259/281) of isolates from non-bacteremic CAP, 85% (824/967) from any IPD, 83% (189/228) from meningitis, and 70% (28/40) from septicemia. A substantial proportion (>19%) of isolates from AOM patients were non-susceptible to common antimicrobial classes, eg, carbapenems, cephalosporins, penicillins, lincosamides, macrolides, sulfonamides, and tetracyclines (Figure 3A). A total of 36 publications reporting 102,599 antimicrobial susceptibility testing outcomes of pneumococcal isolates among adult patients were included in the analysis. In adults, non-susceptibility to sulfonamides was commonly reported: in 69% (5472/7954) of isolates from any IPD, 63% (81/128) from bacteremic CAP, and 54% (123/228) from non-bacteremic CAP. In both pediatric and adult patients, the lowest proportion of non-susceptible isolates was observed to fluoroquinolones, ranging between 1–8% in pediatric and 2–4% in adult patients with different pneumococcal infections (Figure 3B). For specific pneumococcal infections, the proportion of isolates non-susceptible to a specific antimicrobial class varied between the pediatric and adult groups. Among meningitis patients, 55% (1604/2935) of isolates from pediatric and 29% (647/2260) from adult patients were non-susceptible to penicillins, and 24% (619/2547) and 10% (217/2187), respectively, were non-susceptible to cephalosporins. Among patients with any IPD, the proportion of isolates not susceptible to lincosamides was 55% (888/1607) in the pediatric group and 31% (280/903) in the adult group, while 30% (274/907) and 38% (250/658), respectively, were non-susceptible to carbapenems. Among patients with non-bacteremic CAP, non-susceptibility to macrolides was 92% (706/770) in the pediatric group and 44% (2964/6703) in the adult group. Within the pediatric group of patients, 14% (343/2527) of isolates from patients with any IPD were non-susceptible to cephalosporins, compared to 24% (619/2547) of isolates from patients with meningitis. A similar variation in the proportion of non-susceptible isolates across infection types was also observed for other antimicrobial classes. The details of pneumococcal isolates tested for antimicrobial non-susceptibility to different antimicrobial classes and the variation in the non-susceptibility proportion within each age group are presented in Supplementary Tables 3–8. Distribution of Non-Susceptible Pneumococcal Isolates Associated with Specific Pneumococcal Infections, by Regions The distribution of non-susceptible S. pneumoniae isolates from various infection sites in pediatric and adult patients was further stratified by regions, to identify the geographical areas with the highest antimicrobial non-susceptibility burden (Figure 4A and B). Figure 4Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in pediatric (4A) and adult (4B) patients with different pneumococcal infections, by region. Colored dots indicate that susceptibility data were not reported in a specific WHO region/globally. Where the non-susceptibility proportion is 0%, no bar is visible, and there is no colored dot.The figure shows a comparison of regional antimicrobial non-susceptibility patterns among Streptococcus pneumoniae isolates causing different pneumococcal disease manifestations in pediatric (panel A) and adult (panel B) populations. Non-susceptibility rates are presented for eight antibiotic classes (i.e, beta-lactam carbapenems, beta-lactam cephalosporins, beta-lactam penicillins, fluoroquinolones, lincosamides, macrolides, sulfonamides, and tetracyclines) across WHO regions (AFR, AMR, EMR, EUR, SEAR, and WPR) and globally for the following infection types: invasive pneumococcal disease, meningitis, bacteremia/septicemia, non-bacteremic pneumonia, bacteremic pneumonia, and acute otitis media (only in the pediatric population). Considerable variation in non-susceptibility was observed across regions, disease presentations, and antibiotic classes by age groups. Higher non-susceptibility rates were generally reported for macrolides, lincosamides, sulfinamides, and tetracyclines, particularly in the Western Pacific Region, whereas fluoroquinolones and beta-lactam carbapenems showed consistently lower non-susceptibility across most settings in both age groups across regions.A grouped horizontal bar graph showing proportion (%) of non-susceptible Streptococcus pneumoniae isolates by region in pediatric and adult pneumococcal infections.Abbreviations: AFR, African region; AMR, region of the Americas; AOM, acute otitis media; CAP, community-acquired pneumonia; EMR, Eastern Mediterranean region; EUR, European region; IPD, invasive pneumococcal disease; SEAR, Southeast Asian region; WHO, World Health Organization; WPR, Western Pacific region. Proportion of antimicrobial non-susceptible S. pneumoniae isolates identified in pediatric (4A) and adult (4B) patients with different pneumococcal infections, by region. Colored dots indicate that susceptibility data were not reported in a specific WHO region/globally. Where the non-susceptibility proportion is 0%, no bar is visible, and there is no colored dot. Any IPD A total of 19,021 antimicrobial susceptibility testing outcomes of pneumococcal isolates in pediatric and 42,800 outcomes in adult patients were analyzed (Supplementary Table 3). Among the two age groups, the highest proportion of non-susceptible strains was observed in WPR for all common antimicrobial classes, except for penicillins, sulfonamides, and fluoroquinolones in both pediatric and adult patients, and for tetracyclines in adult patients. Among the isolates from the pediatric age group in WPR, non-susceptibility to lincosamides was 97% (589/610), to macrolides 94% (911/969), and to carbapenems 36% (274/751). The highest proportion of isolates non-susceptible to penicillins was observed in EUR (52%, 17/33) and to fluoroquinolones in AMR (12%, 4/33); both these values were higher than the corresponding global proportions of non-susceptible strains in this age group (24% and 2%, respectively). The highest proportion of isolates non-susceptible to sulfonamides (85%, 307/361) was observed in SEAR (Figure 4A). The distribution of non-susceptible isolates in adults was very similar to that observed in the pediatric age group: the highest proportion of non-susceptibility to most antimicrobials was observed in WPR, where more than half of the isolates were non-susceptible to carbapenems. EUR had the highest proportion of strains not susceptible to penicillins (Figure 4B). Meningitis A total of 8623 antimicrobial susceptibility testing outcomes of pneumococcal isolates in pediatric and 4848 outcomes in adult patients were analyzed (Supplementary Table 4). The highest proportions of S. pneumoniae isolates not susceptible to common antimicrobials were observed in WPR, where in the pediatric group 94% (85/90) of isolates were not susceptible to lincosamides, and 82% (270/328) were not susceptible to macrolides (Figure 4A). Isolates not susceptible to penicillins were frequent in most regions, with proportions ranging between 28% (9/32) in AFR and 71% (941/1326) in WPR; non-susceptibility to penicillins was not assessed in SEAR and was not observed in EUR. In adults, the proportion of isolates not susceptible to penicillins ranged between 26% (138/530) in EUR and 75% (27/36) in WPR; no non-susceptible isolates were detected in AFR (Figure 4B). Isolates not susceptible to fluoroquinolones were mainly observed in AFR, where they reached 33% (1/3) in pediatric and 11% (1/9) in adult patients. Non-susceptibility to carbapenems was frequently observed in WPR, in 21% (46/216) of isolates from the pediatric group and 78% (7/9) of isolates from the adult group. However, as seen some of these proportions derived from studies with a small number of samples (susceptibility tested for <10 isolates). Bacteremic CAP The proportion of non-susceptible strains isolated from cases of bacteremic CAP came primarily from the adult population, in which 9647 antimicrobial susceptibility outcomes were analyzed (Supplementary Table 5). In WPR, the proportion of isolates not susceptible to penicillins (30%, 22/74) was 2–3 times higher compared to other regions and the global proportion (12%, 254/2150), while the proportion of isolates not susceptible to fluoroquinolones (25%, 37/148) was more than six times higher than the global proportion (4%, 49/1284) (Figure 4B). The highest proportion of isolates not susceptible to fluoroquinolones was observed in SEAR (30%, 6/20). The proportions of isolates non-susceptible to macrolides were >25% across regions, with increased frequency in WPR (87%, 138/158) and in SEAR (65%, 13/20). Non-Bacteremic CAP A total of 6522 antimicrobial susceptibility testing outcomes of pneumococcal isolates in pediatric and 40,388 outcomes in adult patients were analyzed (Supplementary Table 6). Across regions, the proportion of isolates not susceptible to different antimicrobials in the pediatric group were only reported in SEAR and WPR (Figure 4A). In WPR, the proportion of non-susceptible isolates was 98% (681/696) to macrolides, 94% (493/525) to lincosamides, 93% (256/275) to tetracyclines, and 86% (520/607) to sulfonamides. Nearly half of the isolates were not susceptible to penicillins in these two regions (49% [96/195] in SEAR and 47% [324/689] in WPR). Among adults, isolates not susceptible to macrolides were frequently observed in EUR (63%, 26/41) and AMR (43%, 2620/6065), with the highest proportion of non-susceptible isolates reported in WPR (85%, 246/290) (Figure 4B). Bacteremia and Septicemia A total of 1480 antimicrobial susceptibility testing outcomes of pneumococcal isolates in pediatric and 4916 outcomes in adult patients were analyzed (Supplementary Table 7). In the pediatric age group, non-susceptibility to macrolides was the most common finding across regions, ranging from 11% (12/107) in EUR to 70% (28/40) in WPR, followed by lincosamides, ranging from 5% (5/107) in EUR to 100% (2/2) in EMR, and penicillins, ranging from 8% (4/52) in EMR to 60% (3/5) in AFR (Figure 4A). However, for some instances, these proportions were calculated using studies with a small number of samples (susceptibility tested for <10 isolates). The proportion of non-susceptible isolates in adult patients with bacteremia was reported only in countries from AMR and EUR. In AMR, the proportion of isolates non-susceptible to sulfonamides (44%, 67/154) was nearly twice as high than the global non-susceptibility proportion (16%, 103/661) and six times higher than the proportion of non-susceptible strains in EUR (7%, 36/507). Furthermore, the proportions of isolates not susceptible to macrolides, lincosamides, and tetracyclines were also higher in AMR compared to EUR; however, the proportions of isolates not susceptible to penicillins and fluoroquinolone were higher in EUR (Figure 4B). No data on carbapenem non-susceptibility were retrieved for either age group in any of the WHO regions. AOM A total of 73,529 antimicrobial susceptibility testing outcomes of pneumococcal isolates in pediatric patients were analyzed (Supplementary Table 8). A high proportion of isolates not susceptible to most commonly used antimicrobials was reported in countries from WPR, ranging from 86% (182/211) for carbapenems to 99% (107/108) for lincosamides (Figure 4A). However, the highest proportion of isolates not susceptible to cephalosporins was reported in SEAR (37%, 20/54), while isolates not susceptible to penicillins were frequently reported in countries from AFR (45%, 96/213) and EUR (42%, 9457/22,478). Discussion This systematic literature review describes the distribution of antimicrobial non-susceptible pneumococcal isolates using data published between 2014 and 2023 from 128 studies conducted across all WHO regions. To our knowledge, this is the most comprehensive assessment of how non-susceptible S. pneumoniae isolates from adult and pediatric patients with different pneumococcal infections are distributed worldwide, during a period of widespread pneumococcal vaccination. According to this estimated global distribution, more than half of S. pneumoniae isolates causing any symptomatic pneumococcal infection were non-susceptible to sulfonamides, the highest percentage among different antimicrobial classes. Furthermore, 40% of isolates were non-susceptible to macrolides and nearly a third to tetracyclines. A review that assessed antimicrobial resistance of pneumococcal isolates from children before and after widespread pneumococcal conjugate vaccination showed a reduction in the prevalence of resistance to sulfonamides, macrolides, and tetracyclines following vaccination implementation.25 Despite this decrease, our estimates confirm that there is still a substantial burden of circulating S. pneumoniae isolates not susceptible to antimicrobials commonly used to treat pneumococcal infections. Globally and across WHO regions, we observed very low non-susceptibility to fluoroquinolones (3%), although antibiotics from this class were widely used during the study period;163 there was also no noticeable non-susceptibility to vancomycin and linezolid, however, these antimicrobial classes were not as frequently prescribed as other antimicrobials.163 Among the WHO regions, the proportion of non-susceptible isolates was highest in WPR, reaching ~90% for lincosamides, tetracyclines, and macrolides each, and nearly 40% of isolates in EUR were non-susceptible to penicillin. Taken together, our findings are largely consistent with a previously published review on antimicrobial resistance of pneumococcus isolates obtained from the pediatric population only, although the previous review included nasopharyngeal carriage data from healthy children in their assessment,25 whereas we focused on antimicrobial non-susceptibility data from the patient population only, limiting our capacity to directly compare the results. In our stratified analysis by population age group, we found that the non-susceptible proportion of isolates to commonly prescribed antimicrobials was generally higher in pediatric patients than in adults, overall as well as for any specific infection. Non-susceptibility to penicillins, macrolides, and tetracyclines was substantially higher in children than in adults. Several studies have highlighted young age, daycare attendance, and increased colonization with S. pneumoniae as risk factors for children developing an infection with penicillin- and macrolides-resistant pneumococcus.164–167 In addition, consumption of beta-lactam drugs and macrolides in the previous 1–6 months was strongly correlated with developing an infection with penicillin- and macrolide-resistant pneumococcal strains.165,166 WHO guidelines recommend using oral amoxicillin as the first-line therapy for pediatric pneumococcal infections, including pneumonia and AOM.168,169 In contrast, several studies report frequent use of macrolides to treat childhood pneumonia in clinical practice,170–172 indicating a discrepancy between WHO guideline recommendations and real-world prescription patterns.169 Considering the outcomes of our study, using either penicillins (as recommended by the WHO) or macrolides (as reported in clinical practice) to treat childhood pneumonia and AOM might be less effective in certain settings where resistance to these antibiotics is common, potentially leading to treatment failure. In addition, we observed a variation in the prevalence of non-susceptibility to specific antimicrobial classes within the pediatric age group across the different infection types. These findings highlight the importance of having an in-depth understanding of antimicrobial resistance patterns in various settings. Activities such as targeted antimicrobial resistance surveillance of commonly used empiric agents could inform local resistance data and support the development and strengthening of antimicrobial stewardship programs across all regions, especially where the burden of antimicrobial non-susceptibility is high. Such initiatives remain key to inform local resistance profile and guide the clinician to prescribe appropriate empiric therapy, limiting further resistance development. Our analysis demonstrated that the rates of non-susceptibility to common antimicrobials varied widely by pneumococcal infection type in pediatric and adult patients, with geographical differences. Inconsistent antibiotic prescribing patterns, variation in the case-management guidelines between countries and regions, and the widespread over-the-counter availability of antibiotics in many low- and middle-income settings are likely key factors influencing the observed geographical variation in the distribution of non-susceptible S. pneumoniae isolates. Zooming out, there are several closely linked, broader aspects, including differences in healthcare infrastructure, availability of standard etiological diagnostics and antibacterial testing, extent of antimicrobial stewardship programs, and pneumococcal vaccination coverage in a specific setting, which could influence the observed antibacterial non-susceptibility of S. pneumoniae. The highest burden of pneumococcal isolates non-susceptible to most commonly used antimicrobials was reported in WPR, in both the pediatric and adult populations, regardless of pneumococcal infection type. High rates of non-susceptibility to tetracyclines and sulfonamides were also observed across most infection types in both age groups in SEAR, which is geographically close and has similar socio-economic and healthcare system characteristics as WPR. In these regions, limited antimicrobial stewardship and the availability of over-the-counter antibiotics could contribute to increased selective pressure and subsequent emergence of antimicrobial non-susceptibility.173,174 In contrast, in EUR, where countries generally have stronger antimicrobial stewardship policies and higher adherence to antibiotic therapy guidelines, the proportion of non-susceptible pneumococcal isolates was generally low. In addition to antibiotic use patterns, factors like the timing of pneumococcal conjugate vaccine introduction and vaccine uptake can also influence regional resistance profiles of circulating S. pneumoniae strains, as high vaccination coverage may reduce disease burden and antibiotic use.175–177 Understanding the association between clinical infection type and antimicrobial non-susceptibility in the local context represents a critical consideration for antimicrobial recommendation guidelines. Improved knowledge of antimicrobial susceptibility data could not only potentially decrease the risk of treatment failure but also to a reduced consumption of antimicrobials with suboptimal efficacy in that setting, thus indirectly limiting the development of antibiotic resistance. Taken together, these findings highlight the importance of strengthening antimicrobial stewardship programs (especially in countries with high burden of non-susceptibility), improving diagnostics, establishing targeted surveillance, ensuring appropriate guideline implementation, and maintaining high pneumococcal vaccination coverage, which have been found effective in reducing overall pneumococcal disease burden, as key interventions to mitigate antimicrobial resistance. The strengths of our study are represented by the large number (>250,000) of antimicrobial susceptibility testing outcomes analyzed and in the stratification of these outcomes by clinical infection type, WHO region, and age group (pediatric and adult patients). However, our study has several key limitations. The substantial heterogeneity across the included studies in terms of study population, antibiotic practices, disease epidemiology, and laboratory methodology limited the feasibility of performing a meta-analysis, as pooled estimates would have likely been biased and challenging to interpret. We did not report the proportion of multidrug-resistant S. pneumoniae isolates as most eligible publications did not report these data stratified by infection and antibiotic type. This lack of detail on multidrug resistance estimates limited our ability to generate and interpret additional clinically relevant estimates on pneumococcal isolates resistant to multiple antibiotics in a specific setting. Similarly, our non-susceptibility prevalence estimates could be influenced by studies that reported antimicrobial outcomes of <10 pneumococcal isolates for a specific antimicrobial class since such small sample sizes could result in disproportionately high or low estimates that are not generalizable to other settings; although only a small number of studies reported results based on small sample sizes, and this was only done for a few antibiotics, those estimates of non-susceptibility proportions should be interpreted with caution. Furthermore, data extraction was conducted by a single reviewer with robust secondary verification, and some data were obtained by digitizing graphical data; while both approaches may introduce minor errors, their impact on overall findings is likely minimal. Future studies aiming to assess antimicrobial susceptibility of S. pneumoniae should provide more detailed characterization of resistance profiles, including multidrug resistance, preferably collected from larger patient populations with different clinical conditions, to ensure robustness and generalizability of the outcomes. Furthermore, incorporating mechanistic insights into antimicrobial resistance in S. pneumoniae besides epidemiological patterns could increase the relevance of the findings for clinical and policy decision-making. Conclusions Non-susceptibility of pneumococcal isolates towards various antimicrobials remains an important global health challenge. Our analysis showed that many of the S. pneumoniae isolates obtained from patients with symptomatic pneumococcal disease had decreased susceptibility to commonly prescribed antimicrobials. Notable geographical variation was observed, with the highest burden of non-susceptible isolates identified in WPR and SEAR. In general, the prevalence of non-susceptible pneumococcal isolates was higher in pediatric than in adult patients, and the high proportion of isolates not susceptible to penicillin and macrolides among pediatric patients is concerning from both clinical and public health perspectives. While our review did not include multidrug resistance data, these findings still underscore the importance of continued monitoring of antimicrobial susceptibility patterns and highlight the need to better understand age- and region-specific drivers of resistance. These results do not necessarily reflect the clinical outcomes and/or treatment effectiveness, but the observed patterns of non-susceptibility emphasize the importance of establishing and strengthening ongoing antimicrobial surveillance. Targeted antimicrobial surveillance can contribute to the evidence supporting antimicrobial stewardship initiatives, guiding informed decision-making on empiric therapy and promoting efforts to address the global antimicrobial resistance challenge.