Changing diagnosis coding routines may confound the results of longitudinal childhood pneumonia studies

The pneumococcal conjugate vaccine had a sustained effect on Swedish children 8 years after its introduction

AIM

The pneumococcal conjugate vaccine, which covered seven serotypes of Streptococcus pneumoniae (PCV7), was introduced in Stockholm, Sweden, in 2007. It was replaced by a 13-valent vaccine (PCV13) in 2011. We previously reported a decreased incidence of pneumonia and sinusitis among young children 4 years after the introduction of the PCV7. This study followed the incidence of pneumonia, sinusitis, mastoiditis and meningitis for four more years.

METHODS

We studied validated hospital registry data covering children up to 17 years of age, who were hospitalised in the Stockholm region from 2003 to 2016, when the child population peaked at 485 687. All 11 115 cases diagnosed with pneumonia, coded as bacterial pneumonia, sinusitis, mastoiditis, bacterial meningitis or empyema, were identified. The controls had viral pneumonia or pyelonephritis.

RESULTS

The incidence rates for children under 2 years of age hospitalised for sinusitis, mastoiditis and meningitis decreased significantly by 61%-79% during the eight-year post-vaccination period. Hospitalisations for bacterial pneumonia decreased by 19%-25% in the same age group. These changes were probably due to both the vaccines and changes in diagnosis routines.

CONCLUSION

The effect of vaccination on children under 2 years of age was sustained 8 years after the introduction of the pneumococcal conjugate vaccines.

Pneumonia in Infancy and Risk for Asthma: The Role of Familial Confounding and Pneumococcal Vaccination

BACKGROUND

Studies have reported an increased risk for asthma following lower respiratory tract infections, but few studies have specifically assessed this risk in children diagnosed with pneumonia in infancy. Furthermore, it is not fully understood whether this association is indicative of a causal relationship or if certain children have a predisposition for both diseases.

RESEARCH QUESTION

Are children diagnosed with pneumonia in infancy at increased risk for asthma, and what is the role of familial confounding and pneumococcal conjugate vaccine immunization on the association?

STUDY DESIGN AND METHODS

This study was a nationwide register-based cohort analysis of > 900,000 Swedish children to assess the association between pneumonia in infancy and prevalent asthma at 4 years. A secondary aim was to assess if the association has changed after the introduction of nationwide pneumococcal conjugate vaccine (PCV) immunization as this has led to a shift in pneumonia etiology. The study controlled for important confounders, including shared environmental and familial confounding, by using sibling analyses.

RESULTS

There was a strong association between pneumonia diagnosis in infancy and prevalent asthma at 4 years (adjusted OR, 3.38; 95% CI, 3.26-3.51), as well as in the full sibling analyses (adjusted OR, 2.81; 95% CI, 2.58-3.06). The risk for asthma following pneumonia diagnosis in infancy was slightly higher for those born in the PCV period compared with the pre-PCV period (adjusted OR, 3.80 [95% CI, 3.41-4.24] vs 3.28 [95% CI, 3.15-3.42]) when the proportion of viral pneumonia etiology was also higher (14.5% vs 10.7%, respectively) and the overall asthma prevalence was lower (5.3% vs 6.6%).

INTERPRETATION

Children diagnosed with pneumonia in infancy have a highly increased risk for prevalent asthma at 4 years, which might have implications for future asthma preventive measures and needs to be considered when assessing the morbidity that can be attributed to pneumonia.

Rates and risk factors of severe respiratory syncytial virus infection in 2008-2016 compared with 1986-1998

AIM

Since the introduction in 1979 of rapid testing using immunofluorescence, we have collected information about children hospitalised for confirmed respiratory syncytial virus (RSV) infection in the northern Stockholm area. We here report hospitalisation rates, risk factors and complications in 2008-2016 compared with 1986-1998.

METHODS

Microbiological laboratory reports and retrospective chart review. Comparison of the two periods was complicated by changing testing routines, with a more sensitive method and increased testing of older children in the late period.

RESULTS

In infants, there was an 12.3% increase in the population-based rate of hospital admission for RSV infection from 12.2 to 13.7/1000. Including all children <5 years, there was a 48% increase from 2.7 to 4.0/1000. The median length of stay remained unchanged at 3 days. The need of intensive care decreased in healthy infants but remained high in older children with comorbidity.

CONCLUSION

Considering the changed diagnosis routines, we believe that the rate of hospital admission of infants for RSV infection was unchanged throughout the observed years. The increased rates of older children with confirmed RSV likely resulted from increased testing of children with risk factors for a complicated course.

Impact of the 10-valent pneumococcal conjugate vaccine on hospital admissions in children under three years of age in Iceland

INTRODUCTION

Pneumococcus is an important respiratory pathogen. The 10-valent pneumococcal vaccine (PHiD-CV) was introduced into the Icelandic vaccination programme in 2011. The aim was to estimate the impact of PHiD-CV on paediatric hospitalisations for respiratory tract infections and invasive disease.

METHODS

The 2005-2015 birth-cohorts were followed until three years of age and hospitalisations were recorded for invasive pneumococcal disease (IPD), meningitis, sepsis, pneumonia and otitis media. Hospitalisations for upper- and lower respiratory tract infections (URTI, LRTI) were used as comparators. The 2005-2010 birth-cohorts were defined as vaccine non-eligible cohorts (VNEC) and 2011-2015 birth-cohorts as vaccine eligible cohorts (VEC). Incidence rates (IR) were estimated for diagnoses, birth-cohorts and age groups, and incidence rate ratios (IRR) between VNEC and VEC were calculated assuming Poisson variance. Cox regression was used to estimate the hazard ratio (HR) of hospitalisation between VNEC and VEC.

RESULTS

51,264 children were followed for 142,315 person-years, accumulating 1,703 hospitalisations for the respective study diagnoses. Hospitalisations for pneumonia decreased by 20% (HR 0.80, 95%CI:0.67-0.95) despite a 32% increase in admissions for LRTI (HR 1.32, 95%CI:1.14-1.53). Hospital admissions for culture-confirmed IPD decreased by 93% (HR 0.07, 95%CI:0.01-0.50) and no hospitalisations for IPD with vaccine-type pneumococci were observed in the VEC. Hospitalisations for meningitis and sepsis did not change. A decrease in hospital admissions for otitis media was observed, but did not coincide with PHiD-CV introduction.

CONCLUSION

Following the introduction of PHiD-CV in Iceland, hospitalisations for pneumonia and culture confirmed IPD decreased. Admissions for other LRTIs and URTIs increased during this period.

The changing epidemiology of community-acquired pneumonia: nationwide register-based study in Sweden

BACKGROUND

There is limited evidence on the impact of pneumococcal conjugate vaccine childhood immunization programmes (PCV-CIP) on community-acquired pneumonia (CAP) in individuals with underlying diseases.

METHODS

A nationwide cohort study using Swedish health registers to assess the incidence of hospitalization with all-cause (AC-CAP) and pneumococcal or lobar (PL-CAP) CAP between 2005 and 2015, in relation to PCV-CIP introduction in 2007-09.

RESULTS

In total, 303 691 episodes of AC-CAP occurred, of which 14 225 were PL-CAP. Comparing before (2005-06) with after (2014-15) PCV-CIP, relative incidence reductions were 36% (95% Confidence Interval 32-40), 20% (14-25) and 16% (11-22) of AC-CAP for age groups < 2, 2-4 and 5-17 years, respectively, with similar reductions in young children with and without comorbidities. The reductions were more pronounced for PL-CAP. In the age groups 40-64, 65-74, 75-84 and ≥85 years there were relative increases of 11% (8-14), 18% (15-22), 15% (12-17) and 30% (27-34) of AC-CAP, respectively, but these increases were attenuated after adjustment for admittance practices using four control conditions. In adults with comorbidities, there was an increase in incidence of AC-CAP, and PL-CAP, in contrast to adults without reported underlying diseases where the incidence was stable or diminished for some age groups. Over the study period, there was an increased proportion of pneumonia patients with underlying diseases in all ages.

CONCLUSION

This emphasizes that direct preventive interventions should be targeted towards individuals with underlying diseases. Future studies should investigate reasons for the observed increased risk in adults with comorbidities, for example due to pneumococcal nonvaccine serotypes, or other pathogens, preferentially affecting subjects with underlying diseases.