Effectiveness of pneumococcal vaccination in adults with common immune-mediated inflammatory diseases in the UK: a case-control study

Abstract

Background: People with immune-mediated inflammatory disease are at increased risk of pneumococcal pneumonia. The effectiveness of pneumococcal vaccination in people with immune-mediated inflammatory diseases has not been evaluated. We investigated the effectiveness of pneumococcal vaccination in preventing morbidity and mortality associated with pneumonia in patients with immune-mediated inflammatory diseases.

Methods: In this matched case-control study, we used primary-care electronic health record data from the Clinical Practice Research Datalink Gold database in the UK, with linked hospitalisation and mortality data. Adults with incident common immune-mediated inflammatory diseases diagnosed between April 1, 1997, and Dec 31, 2019, were followed up from the first diagnosis date to the occurrence of an outcome or date of last follow-up. Cases (ie, those with an outcome of interest) were age-matched and sex-matched to up to ten contemporaneous controls by use of incidence density sampling. Outcomes were hospitalisation due to pneumonia, death due to pneumonia, or primary-care consultation for lower respiratory tract infection requiring antibiotics. We defined hospital admission for pneumonia using hospital discharge diagnoses, death due to pneumonia using death certification data, and lower respiratory tract infection as present when primary-care consultation and antibiotic prescription occurred on the same date. We used multivariable, unconditional, logistical regression and constructed three models to examine the association between pneumococcal vaccination as an exposure and each of the three outcomes.

Findings: The first nested case-control analysis included 12 360 patients (7326 [59·3%] women and 5034 [40·7%] men): 1884 (15·2%) who were hospitalised due to pneumonia and 10 476 (84·8%) who were not admitted to hospital due to pneumonia. The second analysis included 5321 patients (3112 [58·5%] women and 2209 [41·5%] men): 781 (14·7%) who died due to pneumonia and 4540 (85·3%) who were alive on the index date. The third analysis included 54 530 patients (33 605 [61·6%] women and 20 925 [38·4%] men): 10 549 (19·3%) with lower respiratory tract infection treated with antibiotics and 43 981 (80·7%) without infection. In the multivariable analysis, pneumococcal vaccination was negatively associated with hospitalisation due to pneumonia (adjusted odds ratio 0·70 [95% CI 0·60-0·81]), death due to pneumonia (0·60 [0·48-0·76]), and lower respiratory tract infection treated with antibiotics (0·76 [0·72-0·80]).

Interpretation: Pneumococcal vaccination is associated with protection against hospitalisation and death due to pneumonia in patients with immune-mediated inflammatory diseases, without apparent residual confounding. However, residual unmeasured confounding cannot be fully excluded in observational research, which includes nested case-control studies. These findings should also be corroborated with data from other countries, given that this study used UK-based data.

Funding: National Institute for Health and Care Research.