Pneumonia hospitalisations in Scotland following the introduction of pneumococcal conjugate vaccination in young children

Background: Scotland introduced PCV7 and PCV13 immunisation in young children in 2006 and 2010 respectively. One recent study from the United States reported a decrease in hospitalisation rates for all-cause pneumonia most notably in adults older than 75 years of age following PCV7 introduction in the US child population. We aimed to examine the effect of PCV7 and PCV13 on hospitalisation rates for all-cause pneumonia across all age groups in Scotland. Methods: We linked hospital records and death certification datasets for the entire Scottish population for the period 2000 to 2012. We included all cases where the primary / secondary diagnosis was pneumonia. Differences in hospital admission rates for pneumonia by age group were calculated using the difference in average annual rates for each period. Results: We estimated that all-cause pneumonia hospitalisation rates in children <2 years decreased by about 30 % in the post-PCV-13 period compared with the pre-PCV period. However, in adults aged 75-84 years and >= 85 years, all-cause pneumonia hospitalisation rates increased by 63 and 46 % respectively in the post-PCV 13 period compared to the pre-PCV period. This resulted in an additional 7000 hospitalisations across all age groups in Scotland in 2012 about half of which were in adults > 75 years. At the same time, the median length of hospital stay decreased by a third in children < 2 years and by about 20 % in adults > 75 years in the post-PCV13 period compared to the pre-PCV period. Additionally, there was an 11 % reduction in deaths due to all-cause pneumonia, and 30 % reduction in pneumococcal hospitalisations across all age groups in the post-PCV13 period compared with pre-PCV period. Discussion: The modest and sustained decline in the rates of hospitalisation for all-cause pneumonia in children and the reduction in proportion of pneumonia hospitalisations in children coded as pneumococcal disease in the post-PCV period should alleviate concerns that pneumococcal serotype replacement may have resulted in an increased pneumonia burden in this age group. The indirect impact of child PCV immunisation in those not vaccinated (in terms of reduction in all-cause pneumonia hospitalisations in the elderly) has not been seen in Scotland. Our results are likely to be confounded by changes in clinical coding and healthcare practices over the same period. Conclusions: Our results illustrate that health care planners cannot, with confidence, predict indirect PCV vaccine impacts on hospitalisations. IPD surveillance across all age groups is needed to assess the indirect effects of PCV in the community.