581695 research-article2015
SJP0010.1177/1403494815581695S. Munson et al.Scandinavian Journal of Public Health
Scandinavian Journal of Public Health, 2015; 43: 657–666
Original Article
Clinical burden of pneumonia, meningitis and septicemia in Norway 2 years after 7-valent pneumococcal conjugate vaccine introduction
SAMANTHA MUNSON1, MIREIA RALUY-CALLADO2, DIMITRA LAMBRELLI2, RADEK WASIAK2, DANIEL ERIKSSON2 & SHARON GRAY1 1Pfizer
Inc., USA, and 2Evidera, UK
Abstract Aims: This population-based, retrospective study quantified the rates of all-cause and pneumococcal pneumonia, meningitis and septicemia in Norway from 2008 to 2009 and determined the proportions of cases caused by pneumococcal vaccine serotypes. Methods: Data on patients with all-cause and pneumococcal pneumonia, meningitis and septicemia were obtained from the Norwegian Patient Registry, which collects hospitalization data from all Norwegian public hospitals based on International Classification of Diseases codes. Norwegian Patient Registry case records linked to the Norwegian Surveillance System for Communicable Diseases provided serotype data for invasive pneumococcal disease in patients with microbiological cultures. Results: In 2008 and 2009, hospitalization rates were relatively stable for all-cause pneumonia (5.28 and 5.35, respectively, per 1000), meningitis (10.70 and 9.67, respectively, per 100,000), and septicemia (from 171.81 to 161.46 per 100,000). In contrast, rates decreased for International Classification of Diseases-10 diagnosed pneumococcal pneumonia (from 13.66 to 10.52 per 100,000), although these cases may be under-reported because of inclusion in all-cause pneumonia. Rates also decreased in diagnosed pneumococcal meningitis (from 1.60 to 1.19 per 100,000) and diagnosed pneumococcal septicemia (from 9.08 to 7.94 per 100,000). Diagnosed pneumococcal disease rates were highest in younger children and older adults, peaking at ⩾60 years old. Pneumococcal pneumonia, meningitis and septicemia caused by serotypes included in the 7-valent pneumococcal conjugate vaccine decreased substantially during the study period, with corresponding serotype replacement by non-7-valent pneumococcal conjugate vaccine serotypes. Conclusions: From 2008 to 2009, International Classification of Diseases-10 diagnosed pneumococcal pneumonia, meningitis and septicemia decreased in most age groups but remained greatest among subjects aged 0–1 and ⩾60 years. Key Words: Norway, pneumonia, meningitis, septicemia, epidemiology, pneumococcal
Introduction Invasive disease caused by Streptococcus pneumoniae results in substantial morbidity and mortality [1]. In 2001, invasive pneumococcal disease incidence in Norway was approximately 19–20 per 100,000 person-years [2]. Adults aged >65 years and children 65 years and children aged 0–2 years (46.6 and 18.6 per 100,000 population, respectively) [2]. In 2006, Norway began vaccinating children with two infant doses and one toddler dose of 7-valent pneumococcal conjugate vaccine (PCV7; serotypes 4,
6B, 9V, 14, 18C, 19F and 23F) [3]. By 2008, 95% of children aged >3 months had received one or more doses [4]. This study assessed the clinical burden of allcause and pneumococcal pneumonia, meningitis and septicemia in Norway starting 2 years after PCV7 introduction at a time when information from two national databases could be combined to generate a unique dataset. Data were analyzed for diagnostic information, corresponding pneumococcal serotype data and patient hospitalization outcomes from 2008 to 2009.
Correspondence: Sharon Gray, Pfizer Inc., 500 Arcola Road, Collegeville, PA, 19426, USA. E-mail: [email protected] (Accepted 19 March 2015) © 2015 the Nordic Societies of Public Health DOI: 10.1177/1403494815581695
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
658 S. Munson et al. Table I. ICD-10 Codes for pneumonia, meningitis, and septicemia. ICD-10 Pneumonia Meningitis Septicemia
J12; J13;a J14; J15:16; J17; J18; J10; J11 G00;a G0; G02; G03; A17.0; A20.3; A32.1; A39.0; A87; B00.3; B01.0; B02.1; B05.1; B26.1; B37.5; B38.4 A40a; A41; A02.1; A20.7; A22.7; A24.1; A26.7; A32.7; A42.7; B00.7; B37.7; O75.3; O85; P36
ICD-10: International Classification of Diseases, version 10. aPneumococcal-specific.
Objectives The primary objective of the study was to quantify rates of pneumonia, meningitis and septicemia in Norway. Secondary objectives were to quantify rates of International Classification of Diseases, version 10 (ICD-10)–diagnosed pneumococcal pneumonia, meningitis and septicemia, in-hospital mortality rates (IHMRs) of all-cause and diagnosed pneumococcal pneumonia, meningitis and septicemia and proportions of cases of pneumococcal pneumonia, meningitis and septicemia caused by pneumococcal vaccine serotypes. Methods Study design In this population-based retrospective study, data from January 2008 to December 2009 were obtained from the Norwegian Patient Registry (NPR) and the Norwegian Surveillance System for Communicable Diseases (MSIS). Disease rates and mortality data from NPR were evaluated for all-cause and ICD-10– diagnosed pneumococcal pneumonia, meningitis and septicemia; serotype data were evaluated for patients who could be matched in NPR and MSIS databases. Ethics The study protocol was reviewed and approved by the regional committees for medical and health research ethics with application reference number 2011/525: ‘A Retrospective Study to Assess Clinical Burden of All-cause Infectious Pneumonia, Meningitis and Septicemia in Norway’. Cases in MSIS and NPR were matched using a unique patient identifier, available for NPR data only after 2008, and anonymized. This review of anonymized data did not require informed consent. Data sources Hospitalization data from all public hospitals in Norway are collected by the Norwegian Health Directorate in the NPR, a national electronic
database comprising a population registry using ICD-10 codes [5]. NPR includes data on patient mortality based on discharge status. Pneumococcal diseases in NPR were diagnosed by the attending physician before laboratory confirmation. Not all ICD-10–diagnosed pneumococcal diseases in NPR were confirmed by microbiological culture, but these cases were coded based on clinical presentation and available laboratory data. Diagnosed pneumococcal diseases may have been under-reported due to classification as ‘all-cause’, rather than laboratory confirmed. MSIS, the Norwegian national surveillance system for infectious diseases, receives reports of all cases of notifiable diseases and pathogens (see Table I in Supplementary Material), including serotype identification for invasive pneumococcal disease cases. NPR and MSIS may differ from databases in other countries whose healthcare data are collected or classified in other ways. Study population All patients diagnosed with pneumonia, meningitis or septicemia from January 2008 to December 2009 based on ICD-10 codes (Table I) were included. Additional analyses of the subgroup of patients in both NPR and MSIS generated associations between pneumococcal serotype data and hospital outcomes. Linked cases were evaluated by year and month of ICD-10 diagnosis. Low sample sizes in linked cases precluded case stratification by age. Pneumococcal disease patient subgroup categorization was based on whether the identified serotype was included in PCV7, PCV13 or the 23-valent pneumococcal polysaccharide vaccine (PPSV23). Study variables Evaluation of the number and proportion of patients affected and IHMRs was based on age, sex and study year for each diagnosis of interest. NPR data included primary diagnosis, date of diagnosis and occurrence of selected comorbidities. Linked NPR-MSIS data included diagnosis, year of diagnosis and S. pneumoniae serotype.
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Clinical burden of pneumonia, meningitis and septicemia 659 Statistical analysis Data were analyzed per calendar year by ICD-10 diagnosis, age group, sex, comorbidities and month (NPR only). NPR data were used to evaluate disease rates and IHMRs for all-cause and diagnosed pneumococcal pneumonia, meningitis and septicemia. Linked NPR-MSIS data were used to evaluate serotype incidence for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia. Categorical variables were described in terms of the number and proportion at each level of the variable, with 95% confidence intervals (CI) for proportions. Continuous variables were summarized in terms of the mean, SD and range of values, with 95% CI for means. Analyses were performed with SAS software (version 9; SAS Institute, Cary, NC, USA). Computation of disease rates and confidence intervals. Annual disease rates, with 95% CI, were calculated as the number of cases divided by the total numbers in the population at risk, multiplied by 1000 for all-cause pneumonia and by 100,000 for all-cause meningitis and septicemia and for diagnosed pneumococcal pneumonia, meningitis and septicemia. Disease cases. Cases were all new episodes with an ICD-10 confirmed diagnosis of interest as the primary diagnosis of hospital admission. Patients rehospitalized with the same diagnosis within 30 days were regarded as experiencing a progression of the previous episode and not considered a second case. Subsequent hospitalizations (after 30 days) for the same patient and ICD-10 diagnosis were considered new episodes. At-risk population. The total population of Norway was the denominator, overall and for subgroups (age, sex, year and month) [6]. Computation of mortality rates and confidence intervals. IHMRs with 95% CI were calculated for the diseases of interest. Hospital mortality rates were estimated as the number of hospitalized patients with a specific disease who died while hospitalized divided by the total number of hospitalized patients with the specific disease, multiplied by 100. Results Patient characteristics The number and proportion of cases by ICD-10 diagnosis, year, age and sex for NPR data are presented in Table II. NPR-based disease rates and
IHMRs by year for the six diseases of interest are shown in Table III. All-cause disease Pneumonia. Hospitalization rates for all-cause pneumonia were similar in 2008 and 2009 (Table III). The highest and lowest rates occurred in the winter and summer months, respectively. Rates were highest in children aged 0–4 years (6.36 and 6.78 per 1000 in 2008 and 2009, respectively) and adults aged ⩾65 years (22.49 and 21.41 per 1000, respectively) and increased sharply with age after age 50 years (Figure 1(a)). The most common comorbidity was chronic obstructive pulmonary disease (COPD), in 20.5% and 19.4% of patients in 2008 and 2009, respectively (see Tables III and IV in Supplementary Material). Comorbidities were less common among children compared with older adults; the most common comorbidity among children aged 0–4 years was central nervous system disorders (2.4% and 2.5% in 2008 and 2009, respectively). IHMRs were similar in 2008 and 2009 (Table III), but no clear patterns emerged in mortality by month. IHMRs were higher in males than females in 2008 (8.40% (95% CI, 7.93–8.90) vs. 6.71% (95% CI, 6.25–7.19)) and 2009 (7.76% (95% CI, 7.29–8.24) vs. 6.15% (95% CI, 5.71 vs 6.61)). IHMRs increased with age, ranging from 0.16% and 0.23% in 2008 and 2009, respectively, in children aged 0–4 years to 10.27% and 9.65%, respectively, in adults aged ⩾65 years. Meningitis. From 2008 to 2009, hospitalization rates for all-cause meningitis declined slightly (Table III); rate by month varied, with no clear trend. Rates were highest in children aged 0–4 years (24.51 and 13.40 per 100,000 in 2008 and 2009, respectively) and were relatively stable (ranging from 8.74 to 10.82 per 100,000) across the other age groups (Figure 1(b)). The most common comorbidity was central nervous system disorders, in 25.1% and 28.4% of patients in 2008 and 2009, respectively. IHMRs were low during 2008–2009 (Table III), although 11 hospitalized patients died of meningitis in each study year. Septicemia. The all-cause septicemia hospitalization rate declined slightly from 171.81 (95% CI, 168.10– 175.58) per 100,000 in 2008 to 161.46 (95% CI, 157.89–165.10) per 100,000 in 2009 (Table III) without apparent monthly trends. Rates were slightly higher in males than females in both 2008 (178.84 (95% CI, 173.49–184.31) vs. 164.84 (95% CI, 159.72–170.08) per 100,000) and 2009 (169.93 (95% CI, 164.76–175.23) vs. 153.02 (95% CI,
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
NPR: Norwegian Patient Registry.
Additional age groups (years) 0–4 1,769 (7.8) 1,921 (8.2) 20–49 2,526 (11.2) 3,245 (13.8) 50–64 3,472 (15.3) 3,497 (14.9) 65+ 14,040 (62.0) 13,741 (58.4) Gender Male 11,829 (52.2) 12,173 (51.7) Female 10,811 (47.8) 11,354 (48.3)
732 (3.1) 1,189 (5.1) 1,123 (4.8) 1,004 (4.3) 989 (4.2) 1,252 (5.3) 1,871 (8.0) 3,237 (13.8) 4,336 (18.4) 7,794 (33.1) 33 (6.7) 103 (20.8) 99 (20.0) 247 (49.8) 243 (49.0) 253 (51.0)
294 (46.6) 337 (53.4)
9 (1.8) 24 (4.8) 14 (2.8) 21 (4.2) 44 (8.9) 38 (7.7) 53 (10.7) 93 (18.8) 92 (18.5) 108 (21.8)
26 (4.1) 96 (15.2) 136 (21.6) 352 (55.8)
9 (1.4) 17 (2.7) 21 (3.3) 20 (3.2) 38 (6.0) 38 (6.0) 78 (12.4) 112 (17.7) 149 (23.6) 149 (23.6)
2009 (N=496)
2008 (N=631)
2008 (N=22,640)
2009 (N=23,527)
Pneumococcal pneumonia
Pneumonia
Age groups (years) 0–1 654 (2.9) 2–4 1,115 (4.9) 5–19 833 (3.7) 20–29 542 (2.4) 30–39 917 (4.1) 40–49 1,067 (4.7) 50–59 1,856 (8.2) 60–69 3,249 (14.4) 70–79 4,419 (19.5) 80+ 7,988 (35.3)
Number of patients (%)
231 (49.9) 232 (50.1)
58 (12.5) 179 (38.7) 71 (15.3) 71 (15.3)
40 (8.6) 18 (3.9) 84 (18.1) 49 (10.6) 82 (17.7) 48 (10.4) 45 (9.7) 57 (12.3) 29 (6.3) 11 (2.4)
2008 (N=463)
Meningitis
178 (44.7) 220 (55.3)
36 (9.0) 156 (39.2) 82 (20.6) 50 (12.6)
21 (5.3) 15 (3.8) 74 (18.6) 54 (13.6) 44 (11.1) 58 (14.6) 50 (12.6) 48 (12.1) 21 (5.3) 13 (3.3)
2009 (N=398)
Table II. Number and proportion of patients by diagnosis, year, age and gender (NPR database).
35 (50.7) 34 (49.3)
10 (14.5) 15 (21.7) 20 (29.0) 20 (29.0)
7 (10.1) 3 (4.3) 4 (5.8) 1 (1.4) 5 (7.2) 9 (13.0) 9 (13.0) 18 (26.1) 8 (11.6) 5 (7.2)
2008 (N=69)
19 (37.3) 32 (62.7)
4 (7.8) 10 (19.6) 21 (41.2) 12 (23.5)
3 (5.9) 1 (2.0) 4 (7.8) 2 (3.9) 3 (5.9) 5 (9.8) 12 (23.5) 13 (25.5) 5 (9.8) 3 (5.9)
2009 (N=51)
Pneumococcal meningitis
4,048 (51.7) 3,780 (48.3)
779 (10.0) 1,208 (15.4) 1,265 (16.2) 4,392 (56.1)
721 (9.2) 58 (0.7) 184 (2.4) 339 (4.3) 440 (5.6) 429 (5.5) 673 (8.6) 1,151 (14.7) 1,528 (19.5) 2,305 (29.4)
2008 (N=7,828)
Septicemia
3,875 (52.2) 3,548 (47.8)
758 (10.2) 1,145 (15.4) 1,159 (15.6) 4,181 (56.3)
705 (9.5) 53 (0.7) 180 (2.4) 320 (4.3) 402 (5.4) 423 (5.7) 637 (8.6) 1,101 (14.8) 1,391 (18.7) 2,211 (29.8)
2009 (N=7,423)
195 (46.4) 225 (53.6)
6 (1.4) 85 (20.2) 91 (21.7) 230 (54.8)
4 (1.0) 2 (0.5) 8 (1.9) 11 (2.6) 34 (8.1) 40 (9.5) 53 (12.6) 80 (19.0) 89 (21.2) 99 (23.6)
2008 (N=420)
187 (49.6) 190 (50.4)
9 (2.4) 63 (16.7) 92 (24.4) 203 (53.8)
7 (1.9) 2 (0.5) 10 (2.7) 9 (2.4) 24 (6.4) 30 (8.0) 46 (12.2) 85 (22.5) 77 (20.4) 87 (23.1)
2009 (N=377)
Pneumococcal septicemia
660 S. Munson et al.
IHMR, % (95% CI)
6.99 (6.67–7.33) 2.98 (1.50–5.27) 13.42 (12.68–14.18) 3.38 (1.94–5.42) 5.88 (1.23–16.24) 9.30 (6.60–12.65)
Incidenceb (95% CI)
5.35 (5.29–5.42) 9.67 (8.81–10.59) 161.46 (157.89–165.10) 10.52 (9.62–11.48) 1.19 (0.90–1.54) 7.94 (7.16–8.78) CI: confidence interval; NPR: Norwegian Patient Registry; IHMR: in-hospital mortality rate. aAt-risk population for 2008 was 4,737,171. bRate per 1000 population for pneumonia and per 100,000 for meningitis, septicemia and pneumococcal diseases. cAt-risk population for 2009 was 4,799,252.
25,692 464 7749 505 57 381 7.61 (7.28–7.95) 2.40 (1.20–4.25) 14.63 (13.88–15.41) 4.65 (3.11–6.65) 6.06 (1.68–14.80) 12.35 (9.36–15.88) 5.28 (5.21–5.34) 10.70 (9.79–11.68) 171.81 (168.10–175.58) 13.66 (12.63–14.75) 1.60 (1.26–2.01) 9.08 (8.24–9.98) 24,997 507 8139 647 76 430 Pneumonia Meningitis Septicemia Diagnosed pneumococcal pneumonia Diagnosed pneumococcal meningitis Diagnosed pneumococcal septicemia
IHMR, % (95% CI) Number of casesa
Incidenceb (95% CI)
Number of casesc
148.12–158.05) per 100,000). The highest rates were observed in adults aged ⩾65 years (656.56 and 616.76 per 100,000 in 2008 and 2009, respectively) and children aged 0–4 years (275.01 and 264.02 per 100,000, respectively) (Figure 1(c)). IHMRs were slightly higher in 2008 than 2009 (Table III). The most common comorbidities were cancer and malignancies, reported in 18.7% (in 2008) and 18.3% (in 2009) of patients. IHMRs were consistently higher in males (15.04% and 13.76% in 2008 and 2009, respectively) than in females (14.19% and 13.02%), and tended to increase with age in both years. IHMRs ranged from 0.12%–1.13% in children aged 0–4 years to 20.26%– 22.12% in adults aged ⩾65 years. Pneumococcal diseases
2009 2008
Table III. Rates of hospitalization and IHMRs in 2008 and 2009 for all-cause pneumonia, meningitis and septicemia and for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia (NPR Database).
Clinical burden of pneumonia, meningitis and septicemia 661
Pneumococcal pneumonia. From 2008–2009, the hospitalization rate for ICD-10–diagnosed pneumococcal pneumonia declined (Table III), although these cases may be under-reported because of possible inclusion in the all-cause pneumonia group, as the pathogen is usually unknown. Hospitalization was most frequent in the winter months. Hospitalization rates were highest in the youngest and oldest age groups (Figure 1(d)), rising sharply after age 50 years, and peaking in adults aged ⩾65 years (50.8 and 35.6 per 100,000 in 2008 and 2009, respectively). The most frequent comorbidity was COPD, occurring in 20.8% (in 2008) and 15.7% (in 2009) of patients, and often in older adults (22.7%–27.6% of adults aged ⩾65 years). IHMRs were 4.65% in 2008 and 3.38% in 2009 (Table III), without distinct monthly patterns. All deaths except one occurred in patients aged ⩾50 years (one adult aged 40–49 years died in 2009). Pneumococcal meningitis. Hospitalization for ICD10–diagnosed pneumococcal meningitis was low in both study years (Table III). Due to the small number of cases (78 in 2008 and 57 in 2009), variation by calendar month or by sex is difficult to interpret. The highest rates were in children aged 0–4 years and adults aged >50 years (Figure 1(e)). The most common comorbidity was central nervous system disorders, in 17.4% and 23.5% of patients in 2008 and 2009, respectively. IHMRs were 6.06% in 2008 and 5.88% in 2009 (Table III) based on four deaths in 2008 and three in 2009. Pneumococcal septicemia. Hospitalization rates for ICD-10–diagnosed pneumococcal septicemia were 9.08 per 100,000 in 2008 and 7.94 per 100,000 in
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
662 S. Munson et al.
Figure 1. Incidence rates of hospitalized all-cause and pneumococcal disease per year by age group (NPR database). (a) All-cause pneumonia; (b) all-cause meningitis; (c) all-cause septicemia; (d) pneumococcal pneumonia; (e) pneumococcal meningitis; (f) pneumococcal septicemia.
2009 (Table III). A total of 420 people were diagnosed with pneumococcal septicemia in 2008 and 377 in 2009, representing 5.4% and 5.1% of all cases
of septicemia, respectively. Rates were lower in the summer months and increased with age in adults (Figure 1(f)). Adults aged ⩾65 years had the highest
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Clinical burden of pneumonia, meningitis and septicemia 663 rates at 33.46 and 28.94 per 100,000 in 2008 and 2009, respectively. The most common comorbidity was chronic renal disease, in 15.0% and 16.4% of patients in 2008 and 2009, respectively. IHMRs were 12.35% in 2008 and 9.30% in 2009 (Table III). All but three deaths occurred in adults aged ⩾50 years; one adult each in the 30–39 years and 40–49 years age groups died in 2008, and one child in the 0–1 year age group died in 2009. The highest IHMRs occurred in adults aged 70–79 years (12.22% and 15.19% in 2008 and 2009, respectively) and in those aged >80 years (24.00% and 13.79%). Pneumococcal serotypes. The proportion of cases of ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia caused by specific pneumococcal serotypes was evaluated using matched NPR-MSIS data (Table IV, and see Table II in Supplementary Material). The proportion of pneumococcal disease cases caused by PCV7 serotypes decreased during 2008 and 2009. During this period, the proportion of cases caused by the six additional PCV13 serotypes increased in pneumococcal pneumonia patients, but decreased among pneumococcal meningitis and pneumococcal septicemia patients (Table IV). From 2008 to 2009, the proportion of pneumococcal disease cases caused by non-PCV13 serotypes increased. The most common PCV7 serotypes causing ICD10–diagnosed pneumococcal pneumonia in 2008 and 2009 were serotypes 4 and 14; the most common of the six additional PCV13 serotypes were serotypes 1 and 7F and the most common non-PCV13 serotypes were serotypes 8, 9N, 22F and 33F (Table IV). Notably, among the non-PCV13 serotypes, serotype 22F showed a substantial increase over 2008–2009. The most common PCV7 serotype causing diagnosed pneumococcal meningitis in 2008 and 2009 was serotype 4; in both years, the most common of the six additional PCV13 serotypes were serotypes 6A and 7F. The most common of the non-PCV13 serotypes were serotypes 9N and 10A, the latter having shown a substantial increase over 2008–2009 (Table IV). For diagnosed pneumococcal septicemia, similar proportions of cases were caused by PCV7 serotypes in 2008 and 2009; the most common of the six additional PCV13 serotypes were serotypes 6A and 7F and the most common of the non-PCV13 serotypes was serotype 22F, which increased substantially from 2008 to 2009 (Table IV). Discussion This population-based retrospective study analyzed data from two matched databases that provide a
national picture of hospitalizations and the clinical burden of disease associated with all-cause and pneumococcal pneumonia, meningitis and septicemia in Norway in the years 2008–2009. During this period, hospitalization rates for allcause pneumonia, meningitis and septicemia were similar, whereas rates declined for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia. Hospitalization rates were generally highest for younger children and older adults; rates increased with age in adults, but this pattern varied among diseases. These age-related trends in pneumococcal diseases were similar to those in other studies from Europe and the USA [7,8]. A recent Danish study of ICD-10–diagnosed all-cause pneumonia also concluded that the clinical burden of pneumonia is substantial. Disease rates for children aged
SJP0010.1177/1403494815581695S. Munson et al.Scandinavian Journal of Public Health
Scandinavian Journal of Public Health, 2015; 43: 657–666
Original Article
Clinical burden of pneumonia, meningitis and septicemia in Norway 2 years after 7-valent pneumococcal conjugate vaccine introduction
SAMANTHA MUNSON1, MIREIA RALUY-CALLADO2, DIMITRA LAMBRELLI2, RADEK WASIAK2, DANIEL ERIKSSON2 & SHARON GRAY1 1Pfizer
Inc., USA, and 2Evidera, UK
Abstract Aims: This population-based, retrospective study quantified the rates of all-cause and pneumococcal pneumonia, meningitis and septicemia in Norway from 2008 to 2009 and determined the proportions of cases caused by pneumococcal vaccine serotypes. Methods: Data on patients with all-cause and pneumococcal pneumonia, meningitis and septicemia were obtained from the Norwegian Patient Registry, which collects hospitalization data from all Norwegian public hospitals based on International Classification of Diseases codes. Norwegian Patient Registry case records linked to the Norwegian Surveillance System for Communicable Diseases provided serotype data for invasive pneumococcal disease in patients with microbiological cultures. Results: In 2008 and 2009, hospitalization rates were relatively stable for all-cause pneumonia (5.28 and 5.35, respectively, per 1000), meningitis (10.70 and 9.67, respectively, per 100,000), and septicemia (from 171.81 to 161.46 per 100,000). In contrast, rates decreased for International Classification of Diseases-10 diagnosed pneumococcal pneumonia (from 13.66 to 10.52 per 100,000), although these cases may be under-reported because of inclusion in all-cause pneumonia. Rates also decreased in diagnosed pneumococcal meningitis (from 1.60 to 1.19 per 100,000) and diagnosed pneumococcal septicemia (from 9.08 to 7.94 per 100,000). Diagnosed pneumococcal disease rates were highest in younger children and older adults, peaking at ⩾60 years old. Pneumococcal pneumonia, meningitis and septicemia caused by serotypes included in the 7-valent pneumococcal conjugate vaccine decreased substantially during the study period, with corresponding serotype replacement by non-7-valent pneumococcal conjugate vaccine serotypes. Conclusions: From 2008 to 2009, International Classification of Diseases-10 diagnosed pneumococcal pneumonia, meningitis and septicemia decreased in most age groups but remained greatest among subjects aged 0–1 and ⩾60 years. Key Words: Norway, pneumonia, meningitis, septicemia, epidemiology, pneumococcal
Introduction Invasive disease caused by Streptococcus pneumoniae results in substantial morbidity and mortality [1]. In 2001, invasive pneumococcal disease incidence in Norway was approximately 19–20 per 100,000 person-years [2]. Adults aged >65 years and children 65 years and children aged 0–2 years (46.6 and 18.6 per 100,000 population, respectively) [2]. In 2006, Norway began vaccinating children with two infant doses and one toddler dose of 7-valent pneumococcal conjugate vaccine (PCV7; serotypes 4,
6B, 9V, 14, 18C, 19F and 23F) [3]. By 2008, 95% of children aged >3 months had received one or more doses [4]. This study assessed the clinical burden of allcause and pneumococcal pneumonia, meningitis and septicemia in Norway starting 2 years after PCV7 introduction at a time when information from two national databases could be combined to generate a unique dataset. Data were analyzed for diagnostic information, corresponding pneumococcal serotype data and patient hospitalization outcomes from 2008 to 2009.
Correspondence: Sharon Gray, Pfizer Inc., 500 Arcola Road, Collegeville, PA, 19426, USA. E-mail: [email protected] (Accepted 19 March 2015) © 2015 the Nordic Societies of Public Health DOI: 10.1177/1403494815581695
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
658 S. Munson et al. Table I. ICD-10 Codes for pneumonia, meningitis, and septicemia. ICD-10 Pneumonia Meningitis Septicemia
J12; J13;a J14; J15:16; J17; J18; J10; J11 G00;a G0; G02; G03; A17.0; A20.3; A32.1; A39.0; A87; B00.3; B01.0; B02.1; B05.1; B26.1; B37.5; B38.4 A40a; A41; A02.1; A20.7; A22.7; A24.1; A26.7; A32.7; A42.7; B00.7; B37.7; O75.3; O85; P36
ICD-10: International Classification of Diseases, version 10. aPneumococcal-specific.
Objectives The primary objective of the study was to quantify rates of pneumonia, meningitis and septicemia in Norway. Secondary objectives were to quantify rates of International Classification of Diseases, version 10 (ICD-10)–diagnosed pneumococcal pneumonia, meningitis and septicemia, in-hospital mortality rates (IHMRs) of all-cause and diagnosed pneumococcal pneumonia, meningitis and septicemia and proportions of cases of pneumococcal pneumonia, meningitis and septicemia caused by pneumococcal vaccine serotypes. Methods Study design In this population-based retrospective study, data from January 2008 to December 2009 were obtained from the Norwegian Patient Registry (NPR) and the Norwegian Surveillance System for Communicable Diseases (MSIS). Disease rates and mortality data from NPR were evaluated for all-cause and ICD-10– diagnosed pneumococcal pneumonia, meningitis and septicemia; serotype data were evaluated for patients who could be matched in NPR and MSIS databases. Ethics The study protocol was reviewed and approved by the regional committees for medical and health research ethics with application reference number 2011/525: ‘A Retrospective Study to Assess Clinical Burden of All-cause Infectious Pneumonia, Meningitis and Septicemia in Norway’. Cases in MSIS and NPR were matched using a unique patient identifier, available for NPR data only after 2008, and anonymized. This review of anonymized data did not require informed consent. Data sources Hospitalization data from all public hospitals in Norway are collected by the Norwegian Health Directorate in the NPR, a national electronic
database comprising a population registry using ICD-10 codes [5]. NPR includes data on patient mortality based on discharge status. Pneumococcal diseases in NPR were diagnosed by the attending physician before laboratory confirmation. Not all ICD-10–diagnosed pneumococcal diseases in NPR were confirmed by microbiological culture, but these cases were coded based on clinical presentation and available laboratory data. Diagnosed pneumococcal diseases may have been under-reported due to classification as ‘all-cause’, rather than laboratory confirmed. MSIS, the Norwegian national surveillance system for infectious diseases, receives reports of all cases of notifiable diseases and pathogens (see Table I in Supplementary Material), including serotype identification for invasive pneumococcal disease cases. NPR and MSIS may differ from databases in other countries whose healthcare data are collected or classified in other ways. Study population All patients diagnosed with pneumonia, meningitis or septicemia from January 2008 to December 2009 based on ICD-10 codes (Table I) were included. Additional analyses of the subgroup of patients in both NPR and MSIS generated associations between pneumococcal serotype data and hospital outcomes. Linked cases were evaluated by year and month of ICD-10 diagnosis. Low sample sizes in linked cases precluded case stratification by age. Pneumococcal disease patient subgroup categorization was based on whether the identified serotype was included in PCV7, PCV13 or the 23-valent pneumococcal polysaccharide vaccine (PPSV23). Study variables Evaluation of the number and proportion of patients affected and IHMRs was based on age, sex and study year for each diagnosis of interest. NPR data included primary diagnosis, date of diagnosis and occurrence of selected comorbidities. Linked NPR-MSIS data included diagnosis, year of diagnosis and S. pneumoniae serotype.
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Clinical burden of pneumonia, meningitis and septicemia 659 Statistical analysis Data were analyzed per calendar year by ICD-10 diagnosis, age group, sex, comorbidities and month (NPR only). NPR data were used to evaluate disease rates and IHMRs for all-cause and diagnosed pneumococcal pneumonia, meningitis and septicemia. Linked NPR-MSIS data were used to evaluate serotype incidence for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia. Categorical variables were described in terms of the number and proportion at each level of the variable, with 95% confidence intervals (CI) for proportions. Continuous variables were summarized in terms of the mean, SD and range of values, with 95% CI for means. Analyses were performed with SAS software (version 9; SAS Institute, Cary, NC, USA). Computation of disease rates and confidence intervals. Annual disease rates, with 95% CI, were calculated as the number of cases divided by the total numbers in the population at risk, multiplied by 1000 for all-cause pneumonia and by 100,000 for all-cause meningitis and septicemia and for diagnosed pneumococcal pneumonia, meningitis and septicemia. Disease cases. Cases were all new episodes with an ICD-10 confirmed diagnosis of interest as the primary diagnosis of hospital admission. Patients rehospitalized with the same diagnosis within 30 days were regarded as experiencing a progression of the previous episode and not considered a second case. Subsequent hospitalizations (after 30 days) for the same patient and ICD-10 diagnosis were considered new episodes. At-risk population. The total population of Norway was the denominator, overall and for subgroups (age, sex, year and month) [6]. Computation of mortality rates and confidence intervals. IHMRs with 95% CI were calculated for the diseases of interest. Hospital mortality rates were estimated as the number of hospitalized patients with a specific disease who died while hospitalized divided by the total number of hospitalized patients with the specific disease, multiplied by 100. Results Patient characteristics The number and proportion of cases by ICD-10 diagnosis, year, age and sex for NPR data are presented in Table II. NPR-based disease rates and
IHMRs by year for the six diseases of interest are shown in Table III. All-cause disease Pneumonia. Hospitalization rates for all-cause pneumonia were similar in 2008 and 2009 (Table III). The highest and lowest rates occurred in the winter and summer months, respectively. Rates were highest in children aged 0–4 years (6.36 and 6.78 per 1000 in 2008 and 2009, respectively) and adults aged ⩾65 years (22.49 and 21.41 per 1000, respectively) and increased sharply with age after age 50 years (Figure 1(a)). The most common comorbidity was chronic obstructive pulmonary disease (COPD), in 20.5% and 19.4% of patients in 2008 and 2009, respectively (see Tables III and IV in Supplementary Material). Comorbidities were less common among children compared with older adults; the most common comorbidity among children aged 0–4 years was central nervous system disorders (2.4% and 2.5% in 2008 and 2009, respectively). IHMRs were similar in 2008 and 2009 (Table III), but no clear patterns emerged in mortality by month. IHMRs were higher in males than females in 2008 (8.40% (95% CI, 7.93–8.90) vs. 6.71% (95% CI, 6.25–7.19)) and 2009 (7.76% (95% CI, 7.29–8.24) vs. 6.15% (95% CI, 5.71 vs 6.61)). IHMRs increased with age, ranging from 0.16% and 0.23% in 2008 and 2009, respectively, in children aged 0–4 years to 10.27% and 9.65%, respectively, in adults aged ⩾65 years. Meningitis. From 2008 to 2009, hospitalization rates for all-cause meningitis declined slightly (Table III); rate by month varied, with no clear trend. Rates were highest in children aged 0–4 years (24.51 and 13.40 per 100,000 in 2008 and 2009, respectively) and were relatively stable (ranging from 8.74 to 10.82 per 100,000) across the other age groups (Figure 1(b)). The most common comorbidity was central nervous system disorders, in 25.1% and 28.4% of patients in 2008 and 2009, respectively. IHMRs were low during 2008–2009 (Table III), although 11 hospitalized patients died of meningitis in each study year. Septicemia. The all-cause septicemia hospitalization rate declined slightly from 171.81 (95% CI, 168.10– 175.58) per 100,000 in 2008 to 161.46 (95% CI, 157.89–165.10) per 100,000 in 2009 (Table III) without apparent monthly trends. Rates were slightly higher in males than females in both 2008 (178.84 (95% CI, 173.49–184.31) vs. 164.84 (95% CI, 159.72–170.08) per 100,000) and 2009 (169.93 (95% CI, 164.76–175.23) vs. 153.02 (95% CI,
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
NPR: Norwegian Patient Registry.
Additional age groups (years) 0–4 1,769 (7.8) 1,921 (8.2) 20–49 2,526 (11.2) 3,245 (13.8) 50–64 3,472 (15.3) 3,497 (14.9) 65+ 14,040 (62.0) 13,741 (58.4) Gender Male 11,829 (52.2) 12,173 (51.7) Female 10,811 (47.8) 11,354 (48.3)
732 (3.1) 1,189 (5.1) 1,123 (4.8) 1,004 (4.3) 989 (4.2) 1,252 (5.3) 1,871 (8.0) 3,237 (13.8) 4,336 (18.4) 7,794 (33.1) 33 (6.7) 103 (20.8) 99 (20.0) 247 (49.8) 243 (49.0) 253 (51.0)
294 (46.6) 337 (53.4)
9 (1.8) 24 (4.8) 14 (2.8) 21 (4.2) 44 (8.9) 38 (7.7) 53 (10.7) 93 (18.8) 92 (18.5) 108 (21.8)
26 (4.1) 96 (15.2) 136 (21.6) 352 (55.8)
9 (1.4) 17 (2.7) 21 (3.3) 20 (3.2) 38 (6.0) 38 (6.0) 78 (12.4) 112 (17.7) 149 (23.6) 149 (23.6)
2009 (N=496)
2008 (N=631)
2008 (N=22,640)
2009 (N=23,527)
Pneumococcal pneumonia
Pneumonia
Age groups (years) 0–1 654 (2.9) 2–4 1,115 (4.9) 5–19 833 (3.7) 20–29 542 (2.4) 30–39 917 (4.1) 40–49 1,067 (4.7) 50–59 1,856 (8.2) 60–69 3,249 (14.4) 70–79 4,419 (19.5) 80+ 7,988 (35.3)
Number of patients (%)
231 (49.9) 232 (50.1)
58 (12.5) 179 (38.7) 71 (15.3) 71 (15.3)
40 (8.6) 18 (3.9) 84 (18.1) 49 (10.6) 82 (17.7) 48 (10.4) 45 (9.7) 57 (12.3) 29 (6.3) 11 (2.4)
2008 (N=463)
Meningitis
178 (44.7) 220 (55.3)
36 (9.0) 156 (39.2) 82 (20.6) 50 (12.6)
21 (5.3) 15 (3.8) 74 (18.6) 54 (13.6) 44 (11.1) 58 (14.6) 50 (12.6) 48 (12.1) 21 (5.3) 13 (3.3)
2009 (N=398)
Table II. Number and proportion of patients by diagnosis, year, age and gender (NPR database).
35 (50.7) 34 (49.3)
10 (14.5) 15 (21.7) 20 (29.0) 20 (29.0)
7 (10.1) 3 (4.3) 4 (5.8) 1 (1.4) 5 (7.2) 9 (13.0) 9 (13.0) 18 (26.1) 8 (11.6) 5 (7.2)
2008 (N=69)
19 (37.3) 32 (62.7)
4 (7.8) 10 (19.6) 21 (41.2) 12 (23.5)
3 (5.9) 1 (2.0) 4 (7.8) 2 (3.9) 3 (5.9) 5 (9.8) 12 (23.5) 13 (25.5) 5 (9.8) 3 (5.9)
2009 (N=51)
Pneumococcal meningitis
4,048 (51.7) 3,780 (48.3)
779 (10.0) 1,208 (15.4) 1,265 (16.2) 4,392 (56.1)
721 (9.2) 58 (0.7) 184 (2.4) 339 (4.3) 440 (5.6) 429 (5.5) 673 (8.6) 1,151 (14.7) 1,528 (19.5) 2,305 (29.4)
2008 (N=7,828)
Septicemia
3,875 (52.2) 3,548 (47.8)
758 (10.2) 1,145 (15.4) 1,159 (15.6) 4,181 (56.3)
705 (9.5) 53 (0.7) 180 (2.4) 320 (4.3) 402 (5.4) 423 (5.7) 637 (8.6) 1,101 (14.8) 1,391 (18.7) 2,211 (29.8)
2009 (N=7,423)
195 (46.4) 225 (53.6)
6 (1.4) 85 (20.2) 91 (21.7) 230 (54.8)
4 (1.0) 2 (0.5) 8 (1.9) 11 (2.6) 34 (8.1) 40 (9.5) 53 (12.6) 80 (19.0) 89 (21.2) 99 (23.6)
2008 (N=420)
187 (49.6) 190 (50.4)
9 (2.4) 63 (16.7) 92 (24.4) 203 (53.8)
7 (1.9) 2 (0.5) 10 (2.7) 9 (2.4) 24 (6.4) 30 (8.0) 46 (12.2) 85 (22.5) 77 (20.4) 87 (23.1)
2009 (N=377)
Pneumococcal septicemia
660 S. Munson et al.
IHMR, % (95% CI)
6.99 (6.67–7.33) 2.98 (1.50–5.27) 13.42 (12.68–14.18) 3.38 (1.94–5.42) 5.88 (1.23–16.24) 9.30 (6.60–12.65)
Incidenceb (95% CI)
5.35 (5.29–5.42) 9.67 (8.81–10.59) 161.46 (157.89–165.10) 10.52 (9.62–11.48) 1.19 (0.90–1.54) 7.94 (7.16–8.78) CI: confidence interval; NPR: Norwegian Patient Registry; IHMR: in-hospital mortality rate. aAt-risk population for 2008 was 4,737,171. bRate per 1000 population for pneumonia and per 100,000 for meningitis, septicemia and pneumococcal diseases. cAt-risk population for 2009 was 4,799,252.
25,692 464 7749 505 57 381 7.61 (7.28–7.95) 2.40 (1.20–4.25) 14.63 (13.88–15.41) 4.65 (3.11–6.65) 6.06 (1.68–14.80) 12.35 (9.36–15.88) 5.28 (5.21–5.34) 10.70 (9.79–11.68) 171.81 (168.10–175.58) 13.66 (12.63–14.75) 1.60 (1.26–2.01) 9.08 (8.24–9.98) 24,997 507 8139 647 76 430 Pneumonia Meningitis Septicemia Diagnosed pneumococcal pneumonia Diagnosed pneumococcal meningitis Diagnosed pneumococcal septicemia
IHMR, % (95% CI) Number of casesa
Incidenceb (95% CI)
Number of casesc
148.12–158.05) per 100,000). The highest rates were observed in adults aged ⩾65 years (656.56 and 616.76 per 100,000 in 2008 and 2009, respectively) and children aged 0–4 years (275.01 and 264.02 per 100,000, respectively) (Figure 1(c)). IHMRs were slightly higher in 2008 than 2009 (Table III). The most common comorbidities were cancer and malignancies, reported in 18.7% (in 2008) and 18.3% (in 2009) of patients. IHMRs were consistently higher in males (15.04% and 13.76% in 2008 and 2009, respectively) than in females (14.19% and 13.02%), and tended to increase with age in both years. IHMRs ranged from 0.12%–1.13% in children aged 0–4 years to 20.26%– 22.12% in adults aged ⩾65 years. Pneumococcal diseases
2009 2008
Table III. Rates of hospitalization and IHMRs in 2008 and 2009 for all-cause pneumonia, meningitis and septicemia and for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia (NPR Database).
Clinical burden of pneumonia, meningitis and septicemia 661
Pneumococcal pneumonia. From 2008–2009, the hospitalization rate for ICD-10–diagnosed pneumococcal pneumonia declined (Table III), although these cases may be under-reported because of possible inclusion in the all-cause pneumonia group, as the pathogen is usually unknown. Hospitalization was most frequent in the winter months. Hospitalization rates were highest in the youngest and oldest age groups (Figure 1(d)), rising sharply after age 50 years, and peaking in adults aged ⩾65 years (50.8 and 35.6 per 100,000 in 2008 and 2009, respectively). The most frequent comorbidity was COPD, occurring in 20.8% (in 2008) and 15.7% (in 2009) of patients, and often in older adults (22.7%–27.6% of adults aged ⩾65 years). IHMRs were 4.65% in 2008 and 3.38% in 2009 (Table III), without distinct monthly patterns. All deaths except one occurred in patients aged ⩾50 years (one adult aged 40–49 years died in 2009). Pneumococcal meningitis. Hospitalization for ICD10–diagnosed pneumococcal meningitis was low in both study years (Table III). Due to the small number of cases (78 in 2008 and 57 in 2009), variation by calendar month or by sex is difficult to interpret. The highest rates were in children aged 0–4 years and adults aged >50 years (Figure 1(e)). The most common comorbidity was central nervous system disorders, in 17.4% and 23.5% of patients in 2008 and 2009, respectively. IHMRs were 6.06% in 2008 and 5.88% in 2009 (Table III) based on four deaths in 2008 and three in 2009. Pneumococcal septicemia. Hospitalization rates for ICD-10–diagnosed pneumococcal septicemia were 9.08 per 100,000 in 2008 and 7.94 per 100,000 in
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
662 S. Munson et al.
Figure 1. Incidence rates of hospitalized all-cause and pneumococcal disease per year by age group (NPR database). (a) All-cause pneumonia; (b) all-cause meningitis; (c) all-cause septicemia; (d) pneumococcal pneumonia; (e) pneumococcal meningitis; (f) pneumococcal septicemia.
2009 (Table III). A total of 420 people were diagnosed with pneumococcal septicemia in 2008 and 377 in 2009, representing 5.4% and 5.1% of all cases
of septicemia, respectively. Rates were lower in the summer months and increased with age in adults (Figure 1(f)). Adults aged ⩾65 years had the highest
Downloaded from sjp.sagepub.com at CHINESE UNIV HONG KONG LIB on December 10, 2015
Clinical burden of pneumonia, meningitis and septicemia 663 rates at 33.46 and 28.94 per 100,000 in 2008 and 2009, respectively. The most common comorbidity was chronic renal disease, in 15.0% and 16.4% of patients in 2008 and 2009, respectively. IHMRs were 12.35% in 2008 and 9.30% in 2009 (Table III). All but three deaths occurred in adults aged ⩾50 years; one adult each in the 30–39 years and 40–49 years age groups died in 2008, and one child in the 0–1 year age group died in 2009. The highest IHMRs occurred in adults aged 70–79 years (12.22% and 15.19% in 2008 and 2009, respectively) and in those aged >80 years (24.00% and 13.79%). Pneumococcal serotypes. The proportion of cases of ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia caused by specific pneumococcal serotypes was evaluated using matched NPR-MSIS data (Table IV, and see Table II in Supplementary Material). The proportion of pneumococcal disease cases caused by PCV7 serotypes decreased during 2008 and 2009. During this period, the proportion of cases caused by the six additional PCV13 serotypes increased in pneumococcal pneumonia patients, but decreased among pneumococcal meningitis and pneumococcal septicemia patients (Table IV). From 2008 to 2009, the proportion of pneumococcal disease cases caused by non-PCV13 serotypes increased. The most common PCV7 serotypes causing ICD10–diagnosed pneumococcal pneumonia in 2008 and 2009 were serotypes 4 and 14; the most common of the six additional PCV13 serotypes were serotypes 1 and 7F and the most common non-PCV13 serotypes were serotypes 8, 9N, 22F and 33F (Table IV). Notably, among the non-PCV13 serotypes, serotype 22F showed a substantial increase over 2008–2009. The most common PCV7 serotype causing diagnosed pneumococcal meningitis in 2008 and 2009 was serotype 4; in both years, the most common of the six additional PCV13 serotypes were serotypes 6A and 7F. The most common of the non-PCV13 serotypes were serotypes 9N and 10A, the latter having shown a substantial increase over 2008–2009 (Table IV). For diagnosed pneumococcal septicemia, similar proportions of cases were caused by PCV7 serotypes in 2008 and 2009; the most common of the six additional PCV13 serotypes were serotypes 6A and 7F and the most common of the non-PCV13 serotypes was serotype 22F, which increased substantially from 2008 to 2009 (Table IV). Discussion This population-based retrospective study analyzed data from two matched databases that provide a
national picture of hospitalizations and the clinical burden of disease associated with all-cause and pneumococcal pneumonia, meningitis and septicemia in Norway in the years 2008–2009. During this period, hospitalization rates for allcause pneumonia, meningitis and septicemia were similar, whereas rates declined for ICD-10–diagnosed pneumococcal pneumonia, meningitis and septicemia. Hospitalization rates were generally highest for younger children and older adults; rates increased with age in adults, but this pattern varied among diseases. These age-related trends in pneumococcal diseases were similar to those in other studies from Europe and the USA [7,8]. A recent Danish study of ICD-10–diagnosed all-cause pneumonia also concluded that the clinical burden of pneumonia is substantial. Disease rates for children aged
