The research team, led by Joon Young Song of Korea University College of Medicine together with colleagues from Seoul National University, IQVIA and Merck, set out to quantify a gap that has long frustrated epidemiologists in the region. Reliable information on how pneumococcal disease affects adults in South Korea, and on whether vaccination programs have altered that picture, has been limited. To fill it, the investigators turned to one of the most powerful tools available in modern health research: a nationwide claims database covering virtually the entire population. They identified every hospitalization episode among adults aged 18 and older coded as invasive pneumococcal disease or pneumonia between 1 January 2010 and 31 December 2019, using the Korean version of the International Classification of Diseases to flag cases.

The study window was deliberately split around the 2014 vaccine rollout. Hospitalizations from 2010 to 2013 served as the pre-vaccination baseline, while those from 2015 to 2019 captured the post-introduction era, with the transition year excluded to allow the program to take hold. Rather than simply comparing average rates between the two periods, the researchers applied interrupted time series analysis, a statistical technique that separates an abrupt change in the level of a trend from a gradual change in its slope. Results were expressed as incidence rate ratios, with a value above 1 signaling an increase and a confidence interval crossing 1 indicating statistical uncertainty.

Over the full ten years, the database yielded 1,682 hospitalization episodes of invasive pneumococcal disease and 8,635 episodes of pneumonia among adults. Those raw counts translate into overall incidence rates of 0.4 episodes per 100,000 person-years for invasive disease and 2.1 per 100,000 person-years for pneumonia. Behind those averages lay a stark mortality gradient: 14.9 percent of adults hospitalized with invasive pneumococcal disease died, compared with 3.0 percent of those admitted with pneumonia. The burden was consistently concentrated among people aged 65 and older, the group in which pneumococcal infection most often turns fatal.

The interrupted time series results told two very different stories. For invasive pneumococcal disease, neither the level nor the slope of incidence changed significantly after the childhood program began. The level ratio was 1.26 with a 95 percent confidence interval of 0.99 to 1.59, just shy of conventional significance, and the slope ratio of 0.97, with an interval of 0.93 to 1.01, offered no evidence of a declining trend. In plain terms, the feared invasive infections, which include bacteremia and meningitis, neither dropped as herd-effect optimists might have hoped nor surged in any statistically convincing way. The adult burden simply persisted.

Pneumonia told a more surprising tale. The level of pneumonia incidence in the post-vaccination period was significantly higher than before, with an incidence rate ratio of 1.92 and a confidence interval of 1.51 to 2.42, meaning the post-program level was nearly double the baseline. The slope, however, was statistically flat, at 1.02 with an interval of 1.00 to 1.04, indicating that rates were not climbing year over year but had shifted upward as a step. The economic dimension followed the same pattern: the average total direct cost per pneumonia episode rose from 3.1 million Korean Won before the program to 4.1 million Won after it, an increase that ripples through a national health system when multiplied across thousands of admissions.

Why would a vaccine that protects children appear to coincide with more adult pneumonia? The authors are careful to avoid overinterpreting the association, and they point to several plausible explanations. Increased surveillance and heightened diagnostic awareness, particularly around respiratory illness, may have driven more hospitalizations to be coded and counted as pneumonia in the later period, inflating apparent incidence without a true rise in disease. Changes in coding practice, an aging population and shifting healthcare-seeking behavior could all contribute. It is also worth noting that pneumonia is a clinically broad category: many pneumonias in adults are not caused by pneumococcus at all, so a vaccine targeting one bacterium would not be expected to reverse trends driven by viruses and other pathogens.

The stability of invasive pneumococcal disease is equally instructive. Indirect, or herd, protection from childhood conjugate vaccination has been documented in several high-income countries, where declines in vaccine-type invasive disease among adults followed pediatric immunization. But the magnitude of that effect depends on how much of adult carriage and disease is caused by the strains covered by the vaccine, on how quickly non-vaccine strains fill the vacated ecological niche, and on adult vaccination coverage with the older polysaccharide vaccine, PPSV23, which South Korea has also deployed in older adults. The new study cannot disentangle these forces, but its results suggest that in the South Korean setting, the indirect benefit to adults from PCV13 in children was not strong enough to bend the invasive disease curve within five years.

The study comes with strengths and caveats that readers should weigh. Its great advantage is scale: a population-based database spanning a full decade, capturing hospitalizations across the entire country rather than a sample of sentinel hospitals. Its limitations are those inherent to claims data. Case identification relied on diagnostic codes rather than bacterial culture or polymerase chain reaction confirmation, so some episodes coded as pneumonia may not have been pneumococcal, and some true pneumococcal pneumonias may have been missed. The study was funded by Merck Sharp & Dohme, a subsidiary of Merck & Co., with study activities contracted to IQVIA, and several authors are employees or paid consultants of those companies, relationships disclosed in the publication. The descriptive design also means the analysis shows temporal association, not causation.

For policymakers, the message is less a verdict on PCV13 than a reminder that childhood immunization is not a complete answer for adult respiratory disease. The authors conclude that the incidence of invasive pneumococcal disease in South Korean adults remained stable while the incidence and economic burden of pneumonia increased after the vaccine entered the childhood program, and that vaccine use alongside factors such as increased surveillance may explain the observed results. As newer, higher-valency conjugate vaccines reach adult immunization schedules worldwide, studies like this one underscore the need for direct adult vaccination strategies, robust surveillance that can distinguish pneumococcal from non-pneumococcal pneumonia, and continued economic evaluation of who bears the cost when prevention falls short. The decade of Korean data suggests that protecting the oldest adults may require protecting them directly.

Subject of Research: Impact of childhood PCV13 vaccination on hospitalized adult pneumococcal disease in South Korea

Article Title: Association between introduction of 13-valent pneumococcal conjugate vaccine and burden of hospitalized adult pneumococcal disease in South Korea – a population-based retrospective database study

Article References: Song, J. Y., Choi, E. H., Choe, Y. J., Yeo, S.-H., Patel, D., Lim, C. T. K., Lee, Y. E., Yamada, E., & Sukarom, I. (2026). Association between introduction of 13-valent pneumococcal conjugate vaccine and burden of hospitalized adult pneumococcal disease in South Korea – a population-based retrospective database study. BMC Public Health. https://doi.org/10.1186/s12889-026-29058-6

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29058-6

Keywords: pneumococcal disease, PCV13, pneumonia, vaccination, herd immunity, South Korea, epidemiology, invasive pneumococcal disease, healthcare costs, public health, National Health Information Database, interrupted time series