Abstract
Aim
The burden of respiratory disease is great among children. This study aimed to examine the temporal relationship between hospitalisation for respiratory syncytial virus (RSV) and bacterial pneumonia.
Methods
A Swedish population‐based cohort was created by combining data from the Swedish Medical Birth Register, the National Inpatient Register, the Cause of Death Register, the Total Population Register, and the Longitudinal Integration Database for Health Insurance and Labour Market Studies. Children born between 1998 and 2015 were included and followed for 2 years. We examined the temporal relationship between RSV hospitalisation and bacterial pneumonia using piecewise exponential models.
Results
The final cohort comprised 1 641 747 children, 48.5% were females. There were 23 632 RSV and 4722 bacterial pneumonia hospitalisations, with mean age of 137.8 and 424.2 days, respectively. RSV hospitalisation was associated with bacterial pneumonia with an adjusted incidence rate ratio (aIRR) of 3.18. The risk was highest in the first month after RSV hospitalisation, aIRR 11.19. The risk of bacterial pneumonia was elevated for 4 months after RSV hospitalisation and before RSV hospitalisation.
Conclusion
We found an increased risk for bacterial pneumonia hospitalisation in children hospitalised for RSV both before and after RSV hospitalisation, indicating a bidirectional relationship.
Keywords: bacterial pneumonia, children, epidemiology, respiratory syncytial virus, Streptococcus pneumoniae
Abbreviations
- aIRR
adjusted incidence rate ratio
- CI
confidence interval
- ICD‐10
International Classification of Disease, Tenth Revision
- PCV
pneumococcal conjugate vaccine
- RSV
respiratory syncytial virus
Key notes.
In this national cohort study, we aimed to examine the temporal relationship between hospitalisation for respiratory syncytial virus (RSV) and bacterial pneumonia in children under 2 years of age.
We found an increased risk of severe bacterial pneumonia before and after being hospitalised for RSV, indicating a bidirectional relationship.
The risk was highest in the month after RSV hospitalisation but was elevated for 4 months after RSV hospitalisation.
1. INTRODUCTION
The respiratory syncytial virus (RSV) is the most common viral agent that causes lower respiratory tract infections in infants and young children. Nearly all children experience at least one RSV infection during their first years of life, even though most are mild. 1 , 2 Globally, RSV accounts for approximately 33 million cases of lower respiratory tract infections, leading to 3.6 million admissions to hospitals and 26 300 hospital deaths in children under the age of 5 years. 2 Several preventive strategies such as new monoclonal antibodies and maternal vaccination are under implementation. 3
A relationship between RSV and bacterial pneumonia has been estabilsed. 4 , 5 Studies indicated that the risk of all‐cause pneumonia increased the following months after an RSV infection. 6 , 7 RSV infections and pneumococcal pneumonia exhibited similar seasonal patterns. 8 The social restrictions imposed during the COVID‐19 pandemic disrupted the RSV transmission, and the incidence of pneumonia decreased. 9 A randomised, placebo‐controlled trial conducted in South Africa before the COVID‐19 pandemic showed that vaccination against Streptococcus pneumoniae reduced the incidence of viral pneumonia, indicating a bidirectional relationship. 10
To our knowledge, no study has examined the duration of the association between bacterial pneumonia and RSV. Understanding this duration is important for estimating the burden of RSV infections and for cost‐effectiveness analyses of preventive measures against both RSV and bacterial pneumonia. The aim of this study was to investigate the relationship between RSV infection and bacterial pneumonia in children under 2 years of age and to examine the duration of this association.
2. METHODS
2.1. Study population
We conducted a register‐based cohort study of children born in Sweden between 1 January 1998, and 31 December 2015, with follow‐up data until 31 December 2016. A dataset was created by using the unique personal identification number assigned to all Swedish residents. This combined data from the Medical Birth Register, the National Inpatient Register, the Longitudinal Integration Database for Health Insurance and Labour Market Studies, the Cause of Death Register, and the Total Population Register. The Medical Birth Register contains information on prenatal, delivery and neonatal care, and covers 98%–99% of all births. 11 The National Inpatient Register contains information on all hospitalisations, and includes the principal diagnosis according to International Classification of Disease, Tenth Revision (ICD‐10) codes. 12 The Longitudinal Integration Database for Health Insurance and Labour Market Studies includes data on education. 13 The Cause of Death Register contains information on mortality. 14 The Total Population Register contains information on migration and county of residence. 15 The data were pseudo‐anonymised with a key kept at the Swedish National Board of Health and Welfare.
2.2. Severe bacterial pneumonia
The main outcome of the analyses was severe bacterial pneumonia, defined as a hospital admission with a principal diagnosis registered in any of the following ICD‐10 codes: J13, J14, J15, J16.0 or J85.1. Any readmissions with a pneumonia‐related ICD‐10 codes within 28 days were excluded.
2.3. Severe RSV infection
Severe RSV infection was defined as hospitalisation with a principal diagnosis of RSV infection. RSV hospitalisations were identified using any of the following ICD‐10 codes: J12.1, J20.5, J21.0 or B97.4. In cases where a child had multiple RSV hospitalisations, only the first event was included in the analyses.
2.4. Alternative outcome
We assumed that pyelonephritis was unrelated to RSV infections and used it as an alternative outcome. Pyelonephritis was defined as a hospital admission with a principal diagnosis of ICD‐10 code N10.9.
2.5. Covariates
Information on pregnancy and birth characteristics was retrieved from the Medical Birth Register. Sex was categorised as male or female. Gestational age was categorised as extremely preterm (22 + 0–27 + 6 weeks/days), very preterm (28–31 weeks), moderately preterm (32–36 weeks), term (37–41 weeks) and post‐term (≥42 weeks). Small for gestational age was defined as a birth weight below the 10th percentile for gestational age. Large for gestational age was defined as a birth weight above the 90th percentile for gestational age. Congenital malformation was defined as abnormalities recorded during the first 28 days of life (ICD‐10 codes Q00–Q99). Additionally, we obtained information on bronchopulmonary dysplasia (ICD‐10 code P27.1), maternal age at delivery, maternal smoking during pregnancy, maternal county of residence, and year of birth. A pneumococcal conjugate vaccine (PCV) was included in the Swedish national immunisation program in 2009, although the first County Council started PCV vaccination in 2007. During the study period, different County Councils tendered different PCV vaccines during different time periods. PCV coverage for children born in Sweden is very high, at more than 95%. 16 PCV status was estimated by combining information on the County Council and the date of birth and county of residence, and categorised as pre‐PCV, PCV7, PCV10 and PCV13. Information on maternal education level was obtained from the Longitudinal Integration Database for Health Insurance and Labour Market Studies and categorised as secondary school or less (≤9 years), upper secondary school (10–12 years), short postsecondary education (13–14 years), and long post‐secondary education (≥15 years).
2.6. Statistical analysis
We estimated the incidence rates with 95% confidence intervals (CI) of bacterial pneumonia hospitalisations as the number of hospital admissions per 100 000 person‐years at risk for children under 2 years of age, first for all children and then by categories relative to time between RSV hospitalisation and bacterial pneumonia hospitalisation. The time categories were defined as follows: The time before RSV hospitalisation was divided into two time periods: more than 1 month before, and 1 month before. Time after RSV hospitalisation was separated into monthly periods up to 5 months, and finally a period more than 5 months after RSV hospitalisation. We also included no RSV hospitalisation, and we had in total nine different categories of bacterial pneumonia to analyse in relation to RSV hospitalisations. We classified 1 month as a one‐twelfth of a year.
We examined associations between severe RSV infections and hospitalisations for bacterial pneumonia using piece‐wise exponential regression models applying robust standard errors. Results were presented as crude and adjusted incidence rate ratios with 95% CI. All analyses were restricted to observations with complete data.
The first step was to examine the association between severe RSV infection and hospitalisation for bacterial pneumonia in the first 2 years of life. We performed crude and adjusted analyses. The adjusted analysis was controlled for the potential confounding effect of pregnancy characteristics, birth characteristics, maternal education, PCV status, seasonality defined as month of birth, county of residence, and time trends defined as year of birth.
The second step was to examine the temporal relationship between severe RSV infection and hospitalisation for bacterial pneumonia. The dataset was split at time‐relative to RSV hospitalisation, and age in months. Time‐relative to RSV hospitalisation and age in months were included as time‐dependent covariates. We performed crude and adjusted analyses. The crude analysis was controlled for age in months. The adjusted analysis was additionally controlled for pregnancy characteristics, birth characteristics, maternal education, PCV status, seasonality, county of residence, and time trends.
Finally, we compared the association between RSV infection and bacterial pneumonia with the association between RSV infection and pyelonephritis, in the same way as described above.
Statistical analyses were performed using Stata Statistical Software: Release V.17 (College Station, Texas, USA).
2.7. Ethics
This study was approved by the Regional Ethics Board in Umeå (numbers 2012‐265‐31M and 2017‐399‐32M), and by the Swedish Ethical Review Authority (number 2021‐06337‐02).
3. RESULTS
The national cohort comprised 1 828 430 children born in Sweden between 1 January 1998, and 31 December 2015. We excluded 186 683 children with missing data in covariates, leaving 1 641 747 children for this study. All children were followed from birth until 2 years of age, 31 December 2016, or censoring due to death (n = 2777) or emigration (n = 8194). In total, the study included 3 220 487 person‐years of follow‐up time.
The key background characteristics of the study population are presented in Table 1 and further characteristics are available in the online supplement Table S1. During the study period, 23 632 children were hospitalised at least once for RSV. There were 4722 cases of bacterial pneumonia, and 13 146 cases of pyelonephritis. The mean age at first hospitalisation for RSV was 137.8 days (95% CI 135.9–139.6) and for bacterial pneumonia 424.2 days (95% CI 418.5–429.9).
TABLE 1.
Key background characteristics of the study population. a
| Included | Excluded b | |||
|---|---|---|---|---|
| RSV | Non‐RSV | All | ||
| n = 23 632 | n = 1 618 115 | n = 1 641 747 | n = 186 683 | |
| % (n) | % (n) | % (n) | % (n) | |
| Sex | ||||
| Female | 42.6 (10078) | 48.6 (786762) | 48.5 (796840) | 48.7 (90974) |
| Male | 57.4 (13554) | 51.4 (831353) | 51.5 (844907) | 51.3 (95702) |
| Gestation age c | ||||
| Extremely preterm | 0.5 (129) | 0.2 (2811) | 0.2 (2940) | 1.4 (2615) |
| Very preterm | 1.6 (384) | 0.4 (6172) | 0.4 (6556) | 2.4 (4472) |
| Moderately preterm | 7.9 (1871) | 4 (63933) | 4 (65804) | 14 (26149) |
| Normal | 85.4 (20170) | 88.3 (1428247) | 88.2 (1448417) | 77.2 (144080) |
| Post‐term | 4.6 (1078) | 7.2 (116952) | 7.2 (118030) | 4.5 (8392) |
| Small for gestational age | ||||
| No | 97.1 (22957) | 97.8 (1583006) | 97.8 (1605963) | 66.8 (124647) |
| Yes | 2.9 (675) | 2.2 (35109) | 2.2 (35784) | 2.1 (3879) |
| Large for gestational age | ||||
| No | 95.6 (22587) | 96.3 (1559009) | 96.3 (1581596) | 66.4 (124047) |
| Yes | 4.4 (1045) | 3.7 (59106) | 3.7 (60151) | 2.4 (4479) |
| Congenital malformation d | ||||
| No | 94.6 (22365) | 96.6 (1562944) | 96.6 (1585309) | 95.8 (178888) |
| Yes | 5.4 (1267) | 3.4 (55171) | 3.4 (56438) | 4.2 (7795) |
| Bronchopulmonary dysplasia | ||||
| No | 99.7 (23563) | 99.9 (1617024) | 99.9 (1640587) | 99.6 (185906) |
| Yes | 0.3 (69) | 0.1 (1091) | 0.1 (1160) | 0.4 (777) |
| PCV | ||||
| Pre PCV | 48.4 (11429) | 53.9 (871639) | 53.8 (883068) | 61.7 (115098) |
| PCV7 | 10.9 (2569) | 8.7 (141296) | 8.8 (143865) | 7.8 (14515) |
| PCV10 | 13.5 (3199) | 15.8 (255407) | 15.8 (258606) | 14.4 (26885) |
| PCV13 | 27.2 (6435) | 21.6 (349773) | 21.7 (356208) | 16.2 (30185) |
| Maternal age | ||||
| Mean (SD) | 30.4 (5.0) | 30.2 (5.1) | 30.2 (5.1) | 30.0 (5.7) |
| Maternal smoking | ||||
| No | 89.6 (21163) | 92.1 (1489845) | 92 (1511008) | 45.2 (84309) |
| Yes | 10.4 (2469) | 7.9 (128270) | 8 (130739) | 4.5 (8341) |
| Parity | ||||
| Mean (SD) | 2.2 (1.0) | 2.0 (1.0) | 1.8 (1.0) | 2.0 (1.2) |
| Maternal education | ||||
| ≥15 | 31.6 (7470) | 32.8 (529956) | 32.7 (537426) | 26.1 (48791) |
| 13–14 | 12.6 (2968) | 13.6 (219388) | 13.5 (222356) | 11.3 (21050) |
| 10–12 | 42 (9914) | 42.1 (681953) | 42.1 (691867) | 31.5 (58736) |
| ≤9 | 13.9 (3280) | 11.5 (186818) | 11.6 (190098) | 8.2 (15271) |
| Year of birth | ||||
| Mean (SD) | 2007.5 (4.8) | 2007.1 (5.1) | 2007.1 (5.1) | 2006.2 (5.2) |
Note: Values are percentages (numbers) unless stated otherwise.
Abbreviations: PCV, pneumococcal conjugate vaccine; RSV, respiratory syncytial virus; SD, standard deviation.
Additional background characteristics on county of residence and month of birth are included in online supplement, Table S1.
Children with missing data (10%) were excluded. Data on variables were available in 50%–100% of the excluded children.
Gestational age was categorised as extremely premature (22–27 weeks), very premature (28–31 weeks), moderately premature (32–36 weeks), term (37–41 weeks) and post‐term (≥42 weeks).
International Classification of Diseases, 10th revision codes: Q00–Q99.
The incidence rate of bacterial pneumonia was 146.7 (95% CI 142.6–150.9) per 100 000 person‐years. The monthly incidence rates of bacterial pneumonia before and after RSV hospitalisation are shown in Table 2. The incidence rates were highest the month before and after RSV hospitalisation.
TABLE 2.
Incidence rates of bacterial pneumonia hospitalisation per 100 000 PY at risk by time relative to RSV hospitalisation.
| Time from RSV hospitalisation | PY | Events | IR (95% CI) |
|---|---|---|---|
| <−1 month | 7070 | 45 | 636.5 (475.2–852.5) |
| −1 month | 1840 | 24 | 1302.5 (873.1–1943.3) |
| RSV 1 month | 1950 | 26 | 1331.1 (906.3–1955) |
| RSV 2 months | 1940 | 13 | 669.4 (388.7–1152.8) |
| RSV 3 months | 1950 | 13 | 665.3 (386.3–1145.8) |
| RSV 4 months | 1920 | 9 | 469 (244–901.4) |
| RSV 5 months | 1900 | 5 | 262.6 (109.3–630.8) |
| RSV >5 months | 28 130 | 125 | 444.4 (372.9–529.6) |
| No RSV | 3 173 770 | 4462 | 140.7 (136.6–144.8) |
| All | 3 220 487 | 4722 | 146.7 (142.6–150.9) |
Note: IRs estimated as the number of inpatient hospital admissions per 100 000 PY at risk.
Abbreviations: CI, confidence interval; IR, incidence rate; PY, person‐years; RSV, respiratory syncytial virus.
Associations between RSV hospitalisation, perinatal factors, and hospitalisation due to bacterial pneumonia are presented in Table 3 and additional data are found in the online supplement Table S2. RSV hospitalisation was associated with bacterial pneumonia adjusted incidence rate ratio (aIRR) 3.18 (95% CI 2.77–3.66). Prematurity, congenital malformation, parity, small for gestational age, and male sex were also associated with an increased risk for bacterial pneumonia in the crude and adjusted analysis. The pneumococcal conjugate vaccine and maternal age were associated with a decreased risk for bacterial pneumonia in the adjusted analysis. Pyelonephritis was weakly associated with RSV hospitalisation aIRR 1.36 (95% CI 1.18–1.56). Regression results are included in online supplement Table S3.
TABLE 3.
Associations between RSV infection, perinatal and sociodemographic factors, and hospitalisation for bacterial pneumonia.
| Crude IRR (95% CI) | Adjusted IRR a (95% CI) | |
|---|---|---|
| RSV b | ||
| Yes | 3.96 (3.46–4.53) | 3.18 (2.77–3.66) |
| No | 1 ref | 1 ref |
| Sex | ||
| Female | 1 ref | 1 ref |
| Male | 1.23 (1.16–1.31) | 1.20 (1.13–1.28) |
| Gestation age c | ||
| Extremely preterm | 6.92 (5.06–9.47) | 4.46 (3.02–6.59) |
| Very preterm | 5.21 (4.16–6.51) | 3.74 (2.98–4.69) |
| Moderately preterm | 1.71 (1.51–1.93) | 1.56 (1.38–1.77) |
| Normal | 1 ref | 1 ref |
| Post‐term | 0.82 (0.72–0.93) | 0.85 (0.75–0.96) |
| Small for gestational age | ||
| Yes | 1.92 (1.64–2.26) | 1.52 (1.28–1.79) |
| No | 1 ref | 1 ref |
| Large for gestational age | ||
| Yes | 1.13 (0.97–1.32) | 1.07 (0.92–1.25) |
| No | 1 ref | 1 ref |
| Congenital malformation d | ||
| Yes | 2.7 (2.42–3.01) | 2.43 (2.18–2.72) |
| No | 1 ref | 1 ref |
| Bronchopulmonary dysplasia | ||
| Yes | 8.64 (5.51–13.56) | 1.52 (0.88–2.63) |
| No | 1 ref | 1 ref |
| PCV | ||
| Pre PCV | 1 ref | 1 ref |
| PCV7 | 1.06 (0.96–1.18) | 0.88 (0.77–0.99) |
| PCV10 | 0.97 (0.89–1.05) | 0.71 (0.61–0.82) |
| PCV13 | 0.77 (0.71–0.84) | 0.64 (0.55–0.74) |
| Maternal age | 1.00 (0.99–1.00) | 0.98 (0.97–0.99) |
| Maternal smoking | ||
| Yes | 1.12 (1.01–1.24) | 0.95 (0.85–1.06) |
| No | 1 ref | 1 ref |
| Parity | 1.20 (1.18–1.23) | 1.23 (1.20–1.26) |
| Maternal education in years | ||
| ≥15 | 1 ref | 1 ref |
| 13–14 | 1.04 (0.95–1.15) | 1.01 (0.91–1.11) |
| 10–12 | 1.12 (1.04–1.20) | 1.01 (0.94–1.09) |
| ≤9 | 1.39 (1.26–1.53) | 1.10 (0.99–1.23) |
| Year of birth | 0.99 (0.99–1.00) | 1.02 (1.01–1.04) |
Abbreviations: CI, confidence interval; IRR, incidence rate ratio; PCV, pneumococcal conjugate vaccine; RSV, respiratory syncytial virus.
The adjusted analysis was adjusted for all factors in the table and county of residence and month of birth. Regression results of county of residence and month of birth are available in online supplement.
RSV hospitalisation due to RSV during first 2 years of life.
Gestational age is categorised as extremely premature (22–27 weeks), very premature (28–31 weeks), moderately premature (32–36 weeks), term (37–41 weeks) and post‐term (≥42 weeks).
International Classification of Diseases, 10th revision codes: Q00–Q99.
Associations between time to RSV hospitalisation and bacterial pneumonia, and between RSV hospitalisation and pyelonephritis are shown in Figure 1. The risk of bacterial pneumonia was elevated more than 1 month before RSV hospitalisation (aIRR 4.68 95% CI 3.43–6.40). The highest risks were found in the month before RSV hospitalisation (aIRR 10.65, 95% CI 7.08–16.01) and in the first month after RSV hospitalisation (aIRR 11.19, 95% CI 7.56–16.58). Thereafter, the risk of bacterial pneumonia decreased continuously until the fifth month after RSV hospitalisation. In the fifth month after RSV hospitalisation, there were only five bacterial pneumonia admissions and no significant difference (aIRR 1.80 95% CI 0.75–4.32) in relation to no RSV hospitalisation. However, the risk of bacterial pneumonia was still higher more than 5 months after RSV hospitalisation compared to those with no RSV hospitalisation (aIRR 2.17 95% CI 1.80–2.61). We found no temporal relationship between time to RSV hospitalisation and pyelonephritis. Regression results are published in the online supplement Tables S4 and S5.
FIGURE 1.
Temporal relationship between RSV hospitalisation and hospitalisation for bacterial pneumonia and pyelonephritis. Time was divided into nine different categories: more than 1 month before RSV hospitalisation, 1 month before RSV hospitalisation, monthly periods related to RSV hospitalisation up to 5 months after RSV hospitalisation, more than 5 months after RSV hospitalisation, and no RSV hospitalisation. Crude (blue circles) and adjusted analyses (red squares) of the associations between time to RSV hospitalisation and hospitalisation for bacterial pneumonia and pyelonephritis in children under 2 years of age. Incidence rate ratios are presented on a logarithmic scale. Vertical lines represent the 95% CIs around the point estimates. Crude analysis was controlled for age in months. The adjusted analysis was controlled for pregnancy characteristics, birth characteristics, maternal education, PCV‐status, seasonality, county of residence, time trends, and age in months.
4. DISCUSSION
This national cohort study found an increased risk for bacterial pneumonia hospitalisation in children hospitalised for RSV. The risk of bacterial pneumonia after RSV hospitalisation was highest in the first month. After that, it decreased successively until the fifth month. Interestingly, the risk of bacterial pneumonia hospitalisation was also elevated before RSV hospitalisation, indicating a bidirectional relationship. There was a small increased risk for pyelonephritis in children hospitalised for RSV but no temporal association, this may be attributed to unmeasured or residual confounding. Our study contributes to establishing this association and shows that the increased risk of bacterial pneumonia is present several months after RSV infection.
An increased risk of bacterial pneumonia after RSV infection was expected, because other studies had showed an association between RSV infection and pneumococcal disease, including community‐acquired alveolar Pneumonia. 4 , 5 , 9 , 17 In addition one study showed that hospitalisation due to invasive pneumococcal disease was correlated with a previous RSV infection, and the correlation between invasive pneumococcal disease and RSV was strong up to 4 weeks, compared to only 2 weeks for invasive pneumococcal disease and influenza and human metapneumoviruses. 18 The risk of bacterial infection after viral infections has been attributed to barrier damage, loss of tight junctions, cilia dysfunction, and affected mucociliary clearance, leading to obstruction of the small airways. RSV can interfere with the repairing of the epithelium by producing factors that lead to fibrosis rather than repair. 5 RSV infection enhanced bacterial adhesion of S. pneumoniae, 5 , 19 and also of Haemophilus influenzae. 20 The innate immune system was also affected by RSV by reduced production of reactive oxygen species used by macrophages and neutrophils leading to reduced bacterial clearance. 5 The adaptive immune system was affected by disturbed cytokines and pro‐inflammatory factors. 5 These mechanisms may explain the association we found between RSV and bacterial pneumonia.
A growing body of evidence has emerged regarding the role of the microbiota of the airways and lower respiratory tract infections. 4 , 21 , 22 Dysbiosis in the microbiota led to susceptibility to infection. 22 Examples of hypotheses are that RSV may trigger an outgrowth of a pathogen already presents in the airway microbiota, that RSV infection may promote novel bacteria into the microbiota, cause co‐shedding of disease with droplets containing both RSV and bacteria, and that colonisation of certain bacteria make the child more susceptible to severe RSV infection. 23 , 24 For instance, in the Orchid birth cohort, an outgrowth of S. pneumoniae or Moraxella catarrhalis was associated with RSV infection. 23 In a cotton rat model the viral load of RSV increased when the rats were inoculated with certain strains of pneumococci before RSV. 25 In a prospective observational study from the Netherlands, an increase in viral load was seen in patients with an abundance of H. influenzae in the nasopharynx. 26 Streptococcus pneumoniae in nasopharyngeal swaps was associated with an increased need for mechanical ventilation for infants hospitalised for RSV disease. Co‐detection of S. pneumoniae or H. influenzae and RSV had more severe disease. 4 Consequently, disturbances in the airway microbiota may increase the risk of bacterial pneumonia before and after RSV infection.
The bidirectional relationship between bacterial pneumonia and RSV has been studied between RSV and S. pneumoniae. In a randomised controlled trial, a decrease in RSV was seen in HIV‐negative children who were given PCV. 10 This was confirmed in several observational studies after PCV implementation. 8 , 27 In Israel, an increase in RSV activity was associated with an increased incidence of radiology‐confirmed alveolar pneumonia. 28 Furthermore, restrictions implemented to reduce the transmission of COVID‐19 reduced the incidence rate of invasive pneumococcal diseases, influenza, and RSV. 29 In the autumn of 2021 after the lifting of COVID‐19 restrictions, there was an unusually large outbreak of RSV in Quebec, followed by a rise in invasive pneumococcal disease in children under 5 years. 30 No large difference in the pneumococcal carriage during the COVID‐19 pandemic was seen indicating that both the presence of pneumococcal serotypes and the effect of a viral infection are part of the development of invasive pneumococcal disease. 9 , 29 In a study from Israel, RSV was attributable to 49% of community‐acquired alveolar pneumonia, 21% of community‐acquired other than alveolar pneumonia, and 18% of pneumococcal bacteremic pneumonia. 9 In our study only a small part of bacterial pneumonia had a prior RSV hospitalisation, but since RSV infection is common among young children a larger part of bacterial pneumonia may be attributed to RSV disease that did not require hospitalisation. Our results together with findings from other studies, suggest that the burden of disease might be lessened both by implementing preventive measures against RSV and vaccination against pneumococcal disease.
4.1. Strengths and limitations
A strength of this study was that we used nationwide registers with complete data on 89.8% of the children. However, our study had several limitations. First, it relied on register data, infections could not be confirmed by laboratory findings or radiographic images, and we were not able to examine viral load, dysregulated immune responses, and interactions with other pathogens or commensals in the airway microbiota. Second, we were unable to adjust for several potential confounders including air pollution, weather factors, breastfeeding, antibiotic exposure, and genetic vulnerability. This may explain the increased risk of bacterial pneumonia that remained more than 5 months after RSV hospitalisation; it may also explain, at least partly, the increased risk of bacterial pneumonia before RSV hospitalisation. However, the temporal relationship between RSV hospitalisation and bacterial pneumonia may not be explained by these confounders. Third, we have not been able to perform interaction analysis due to the low number of cases of bacterial pneumonia in the months after RSV hospitalisation, only 66 cases occurred during the first 4 months after RSV hospitalisation. Fourth, our study included only children born before 2016, and we were unable to examine the association between RSV infections and bacterial pneumonia during and after the COVID‐19 pandemic. Finally, we have not been able to examine the temporal relationship between less severe RSV disease and bacterial pneumonia.
5. CONCLUSION
We found an increased risk for bacterial pneumonia hospitalisation in children hospitalised for RSV. The risk was highest in the month after RSV hospitalisation and decreased successively until the fifth month. Additionally, we found that the risk of bacterial pneumonia increased before RSV, indicating a bidirectional relationship. We suggest that the impact of bacterial pneumonia is evaluated when implementing new preventive strategies against RSV disease.
AUTHOR CONTRIBUTIONS
Karin Strandell: Writing – original draft; visualization. Samuel Videholm: Conceptualization; methodology; visualization; formal analysis; writing – review and editing; funding acquisition. Andreas Tornevi: Writing – review and editing; validation. Maria Björmsjö: Writing – review and editing; funding acquisition. Sven Arne Silfverdal: Conceptualization; methodology; supervision; writing – review and editing; funding acquisition.
FUNDING INFORMATION
This study was supported in part by a research grant from Investigator‐Initiated Studies Program of MSD, Sweden (to SV). The opinions expressed in this paper are those of the authors and do not necessarily represent those of MSD. The study was also supported in part by a research grant from Oskarsfonden (to KS and SV) and by grants from the Swedish state under the agreement between the Swedish government and the county councils, the ALF‐agreement (to SAS and MB).
CONFLICT OF INTEREST STATEMENT
SV and MB have received funding to their institution from MSD during the conduct of the study and serving on the advisory board for pneumococcal vaccines from MSD outside the submitted work. SAS has been involved in vaccine trials for MSD, and has participated in Ad board meetings on hexavalent, influenza, meningococcal, pneumococcal vaccines, paid to his Institution or personally but outside this work.
Supporting information
Table S1.
Table S2.
Table S3.
Table S4.
Table S5.
Strandell K, Videholm S, Tornevi A, Björmsjö M, Silfverdal SA. Increased risk of bacterial pneumonia before and after respiratory syncytial virus infection in young children. Acta Paediatr. 2025;114:83–91. 10.1111/apa.17405
REFERENCES
- 1. Andeweg SP, Schepp RM, van de Kassteele J, Mollema L, Berbers GAM, van Boven M. Population‐based serology reveals risk factors for RSV infection in children younger than 5 years. Sci Rep. 2021;11(1):8953. doi: 10.1038/s41598-021-88524-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Li Y, Wang X, Blau DM, et al. Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in children younger than 5 years in 2019: a systematic analysis. Lancet. 2022;399(10340):2047‐2064. doi: 10.1016/s0140-6736(22)00478-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Messina A, Germano C, Avellis V, et al. New strategies for the prevention of respiratory syncytial virus (RSV). Early Hum Dev. 2022;174:105666. doi: 10.1016/j.earlhumdev.2022.105666 [DOI] [PubMed] [Google Scholar]
- 4. Pacheco GA, Gálvez NMS, Soto JA, Andrade CA, Kalergis AM. Bacterial and viral coinfections with the human respiratory syncytial virus. Microorganisms. 2021;9(6):1293. doi: 10.3390/microorganisms9061293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Oliva J, Terrier O. Viral and bacterial co‐infections in the lungs: dangerous liaisons. Viruses. 2021;13(9):1725. doi: 10.3390/v13091725 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Abreo A, Wu P, Donovan BM, et al. Infant respiratory syncytial virus bronchiolitis and subsequent risk of pneumonia, otitis media, and antibiotic utilization. Clin Infect Dis. 2020;71(1):211‐214. doi: 10.1093/cid/ciz1033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Weber MW, Milligan P, Giadom B, et al. Respiratory illness after severe respiratory syncytial virus disease in infancy in The Gambia. J Pediatr. 1999;135(6):683‐688. doi: 10.1016/s0022-3476(99)70085-5 [DOI] [PubMed] [Google Scholar]
- 8. Weinberger DM, Klugman KP, Steiner CA, Simonsen L, Viboud C. Association between respiratory syncytial virus activity and pneumococcal disease in infants: a time series analysis of US hospitalization data. PLoS Med. 2015;12(1):e1001776. doi: 10.1371/journal.pmed.1001776 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Dagan R, van der Beek BA, Ben‐Shimol S, et al. The COVID‐19 pandemic as an opportunity for unravelling the causative association between respiratory viruses and pneumococcus‐associated disease in young children: a prospective study. EBioMedicine. 2023;90:104493. doi: 10.1016/j.ebiom.2023.104493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Madhi SA, Klugman KP. A role for Streptococcus pneumoniae in virus‐associated pneumonia. Nat Med. 2004;10(8):811‐813. doi: 10.1038/nm1077 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. The National Board of Health and Welfare in Sweden . The Swedish Medical Birth Register – A summary of content and quality. 2003.
- 12. Ludvigsson JF, Andersson E, Ekbom A, et al. External review and validation of the Swedish national inpatient register. BMC Public Health. 2011;11:450. doi: 10.1186/1471-2458-11-450 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ludvigsson JF, Svedberg P, Olén O, Bruze G, Neovius M. The longitudinal integrated database for health insurance and labour market studies (LISA) and its use in medical research. Eur J Epidemiol. 2019;34(4):423‐437. doi: 10.1007/s10654-019-00511-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Brooke HL, Talbäck M, Hörnblad J, et al. The Swedish cause of death register. Eur J Epidemiol. 2017;32(9):765‐773. doi: 10.1007/s10654-017-0316-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ludvigsson JF, Almqvist C, Bonamy AK, et al. Registers of the Swedish total population and their use in medical research. Eur J Epidemiol. 2016;31(2):125‐136. doi: 10.1007/s10654-016-0117-y [DOI] [PubMed] [Google Scholar]
- 16. Lindstrand A, Galanis I, Darenberg J, et al. Unaltered pneumococcal carriage prevalence due to expansion of non‐vaccine types of low invasive potential 8years after vaccine introduction in Stockholm, Sweden. Vaccine. 2016;34(38):4565‐4571. doi: 10.1016/j.vaccine.2016.07.031 [DOI] [PubMed] [Google Scholar]
- 17. Ben‐Shimol S, Greenberg D, Hazan G, Shemer‐Avni Y, Givon‐Lavi N, Dagan R. Seasonality of both bacteremic and nonbacteremic pneumonia coincides with viral lower respiratory tract infections in early childhood, in contrast to nonpneumonia invasive pneumococcal disease, in the pre‐pneumococcal conjugate vaccine era. Clin Infect Dis. 2015;60(9):1384‐1387. doi: 10.1093/cid/civ023 [DOI] [PubMed] [Google Scholar]
- 18. Ampofo K, Bender J, Sheng X, et al. Seasonal invasive pneumococcal disease in children: role of preceding respiratory viral infection. Pediatrics. 2008;122(2):229‐237. doi: 10.1542/peds.2007-3192 [DOI] [PubMed] [Google Scholar]
- 19. Kimaro Mlacha SZ, Peret TC, Kumar N, et al. Transcriptional adaptation of pneumococci and human pharyngeal cells in the presence of a virus infection. BMC Genomics. 2013;14:378. doi: 10.1186/1471-2164-14-378 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Novotny LA, Bakaletz LO. Intercellular adhesion molecule 1 serves as a primary cognate receptor for the type IV pilus of nontypeable Haemophilus influenzae . Cell Microbiol. 2016;18(8):1043‐1055. doi: 10.1111/cmi.12575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Diaz‐Diaz A, Garcia‐Maurino C, Jordan‐Villegas A, Naples J, Ramilo O, Mejias A. Viral bacterial interactions in children: impact on clinical outcomes. Pediatr Infect Dis J. 2019;38(6S Suppl 1):S14‐S19. doi: 10.1097/inf.0000000000002319 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Natalini JG, Singh S, Segal LN. The dynamic lung microbiome in health and disease. Nat Rev Microbiol. 2023;21(4):222‐235. doi: 10.1038/s41579-022-00821-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Brealey JC, Young PR, Sloots TP, et al. Bacterial colonization dynamics associated with respiratory syncytial virus during early childhood. Pediatr Pulmonol. 2020;55(5):1237‐1245. doi: 10.1002/ppul.24715 [DOI] [PubMed] [Google Scholar]
- 24. Cebey‐López M, Herberg J, Pardo‐Seco J, et al. Does viral co‐infection influence the severity of acute respiratory infection in children? PLoS One. 2016;11(4):e0152481. doi: 10.1371/journal.pone.0152481 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Nguyen DT, Louwen R, Elberse K, et al. Streptococcus pneumoniae enhances human respiratory syncytial virus infection in vitro and in vivo. PLoS One. 2015;10(5):e0127098. doi: 10.1371/journal.pone.0127098 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Ederveen THA, Ferwerda G, Ahout IM, et al. Haemophilus is overrepresented in the nasopharynx of infants hospitalized with RSV infection and associated with increased viral load and enhanced mucosal CXCL8 responses. Microbiome. 2018;6(1):10. doi: 10.1186/s40168-017-0395-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Fathima P, Blyth CC, Lehmann D, et al. The impact of pneumococcal vaccination on bacterial and viral pneumonia in Western Australian children: record linkage cohort study of 469589 births, 1996–2012. Clin Infect Dis. 2018;66(7):1075‐1085. doi: 10.1093/cid/cix923 [DOI] [PubMed] [Google Scholar]
- 28. Weinberger DM, Givon‐Lavi N, Shemer‐Avni Y, et al. Influence of pneumococcal vaccines and respiratory syncytial virus on alveolar pneumonia, Israel. Emerg Infect Dis. 2013;19(7):1084‐1091. doi: 10.3201/eid1907.121625 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Rybak A, Levy C, Angoulvant F, et al. Association of nonpharmaceutical interventions during the COVID‐19 pandemic with invasive pneumococcal disease, pneumococcal carriage, and respiratory viral infections among children in France. JAMA Netw Open. 2022;5(6):e2218959. doi: 10.1001/jamanetworkopen.2022.18959 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ouldali N, Deceuninck G, Lefebvre B, et al. Increase of invasive pneumococcal disease in children temporally associated with RSV outbreak in Quebec: a time‐series analysis. Lancet Reg Health Am. 2023;19:100448. doi: 10.1016/j.lana.2023.100448 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1.
Table S2.
Table S3.
Table S4.
Table S5.