Abstract
BACKGROUND: There is an ongoing need to assess the impact of pneumococcal conjugate vaccines (PCVs) to guide the use of these potentially valuable but under-utilized vaccines against pneumonia, which is one of the most common causes of post-neonatal mortality. METHODS: We conducted a systematic review of the literature on PCV10 and PCV13 impact on all-cause, radiologically confirmed and severe pneumonia hospitalisation rates as well as all-cause and pneumonia-specific mortality rates. We included studies that were published from 2003 onwards, had a post-licensure observational study design, and reported on any of our defined outcomes in children aged between 0-9 years. We derived incidence rates (IRs), incidence rate ratios (IRRs) or percent differences (%). We assessed all studies for risk of bias using the Effective Public Health Practice Project (EPHPP) quality assessment tool. RESULTS: We identified a total of 1885 studies and included 43 comparing one or more of the following hospitalised outcomes of interest: all-cause pneumonia (n = 27), severe pneumonia (n = 6), all-cause empyema (n = 8), radiologically confirmed pneumonia (n = 8), pneumococcal pneumonia (n = 7), and pneumonia mortality (n = 10). No studies evaluated all-cause mortality. Studies were conducted in all WHO regions except South East Asia Region (SEAR) and low- or middle-income countries (LMICs) in the Western Pacific Region (WPR). Among children <5 years old, PCV impact ranged from 7% to 60% for all-cause pneumonia hospitalisation, 8% to 90% for severe pneumonia hospitalisation, 12% to 79% for radiologically confirmed pneumonia, and 45% to 85% for pneumococcal confirmed pneumonia. For pneumonia-related mortality, impact was found in three studies and ranged from 10% to 78%. No obvious differences were found in vaccine impact between PCV10 and PCV13. One study found a 17% reduction in all-cause pneumonia among children aged 5-9 years, while another found a reduction of 81% among those aged 5-17 years. A third study found a 57% reduction in all-cause empyema among children 5-14 years of age. CONCLUSION: We found clear evidence of declines in hospitalisation rates due to all-cause, severe, radiologically confirmed, and bacteraemic pneumococcal pneumonia in children aged <5 years, supporting ongoing use of PCV10 and PCV13. However, there were few studies from countries with the highest <5-year mortality and no studies from SEAR and LMICs in the WPR. Standardising methods of future PCV impact studies is recommended.