1- Faculty of Medicine, SEGi University, Kota Damansara, Selangor, Malaysia
2- Clinical Research Center, Sibu Hospital, Ministry of Health Malaysia, Sibu, Sarawak, Malaysia
3- Department of Paediatrics, Kapit Hospital, Ministry of Health Malaysia, Kapit, Sarawak, Malaysia
4- Duke Global Health Institute, Duke University, Durham, North Carolina, USA
5- Division of Infectious Diseases, Duke University School of Medicine, Durham, North Carolina, USA
6- Emerging Infectious Disease Program, Duke-NUS Medical School, Singapore
7- Global Health Center, Duke Kunshan University, Kunshan, China
8- Department of Paediatrics, Sibu Hospital, Ministry of Health Malaysia, Sibu, Sarawak, Malaysia
Background: Lower
respiratory tract infections (LRTI) have a great impact upon young children
globally, with high mortality rates. We studied the disease burden of
virus-associated LRTI among hospitalized young children in Central Sarawak.
Methods: This is a cross-sectional, observational study of under-five children recruited between June 2017 and May 2018 in Sibu and Kapit Hospitals. We reviewed medical records to compare the disease burden among children with nasopharyngeal swab that were positive by molecular assays for respiratory syncytial virus (RSV), Adenoviruses (AdV), Coronaviruses, Enteroviruses, Influenza or Para influenza viruses.
Results: A total of 196 children with LRTI (mean age: 1.34 +/- 1.08 years) were identified. RSV was the commonest virus detected (54.1%), followed by AdV (24.0%). Compared to the others, AdV-LRTI had a statistically significant longer length of stay (10.1 days vs 7.0 days), duration for intravenous hydration (33.8 hours vs. 17.0 hours), as well as nebulization using saline (128.4 hours vs. 66.3 hours) and salbutamol (151.5 hours vs. 111.32 hours). More children with AdV infection were treated with steroids (23.4%, OR= 7.3, 95% CI: 2.52, 21.01), magnesium sulphate (46.8%, OR= 2.7, 95% CI: 1.35, 5.27), high flow nasal cannula (48.9%, OR= 2.3, 95% CI: 1.17, 4.48), being intubated (10.6%, OR= 5.8, 95% CI: 1.33, 25.25) and received antibiotics (93.6%, OR= 4.0, 95% CI: 1.17, 13.77) than children with other etiologies.
Conclusion: Among under-five children with LRTI, RSV was the most prevalent viral etiological cause, but AdV infection was associated with a higher disease burden by several metrics.
Keywords: Disease burden, hospitalization, lowers respiratory tract infection, pediatrics, respiratory viral infection
INTRODUCTION
For more than a decade, lower respiratory tract infections (LRTI) have
been the deadliest communicable diseases worldwide with more than four million
deaths annually. [1] Children under five years of age are disproportionately
affected; in 2017, LRTI caused 808,694 deaths worldwide in under-five children.
[2] In Malaysia, LRTI caused 3.8% of all deaths among the under-five population
in 2016. [3]
Etiology and clinical presentation of LRTI varies among individuals
cross the globe. Viruses such as Respiratory Syncytial Virus (RSV), Influenza
virus (Flu), Parainfluenza virus (PIV) and Adenovirus (AdV) are frequent causes
of LRTI and responsible for the majority of LRTI-related hospitalizations.[4,
5] Among these viruses, RSV is the most prevalent respiratory pathogen detected
in young children with peak activity during the end of the year (September to
December) in the northern hemisphere.[6] Children hospitalized with LRTI often
present with cough, fever, and rapid breathing, and consequently often have
reduced energy as well as reduction in fluids intake and feeding. Treatment for
LRTI ranges from symptomatic support and invasive ventilation to antimicrobial
therapy. However, in the absence of the diagnostic and prognostic tools to
differentiate between viral and bacterial LRTI, antimicrobials are often
inappropriately and/or empirically used and mismanaged. [7, 8] Thus, it is
important to understand what pathogens are circulating among patients and
disease burden they cause.
Currently, there is a paucity of data regarding the disease burden of
viral causes of pneumonia among children less than five years of age in Sarawak,
on the northern central coast of Borneo. Metrics for disease burden include
duration of various presenting illness and requirement of different inpatient
treatments, as well as duration of hospitalization, and etiological causes of
LRTI in under-five children. In this study, we aimed to determine the viral
etiology and disease burden of LRTI among these children in Sarawak, as well as
disease burden differences between children infected by different viruses. We
also studied the differences of disease burden for children with a single viral
infection versus co-infections.
METHODS
Study Design
We conducted a cross-sectional case record review study using the
secondary data of a year-long cross-sectional surveillance study.
Setting
The initial study was conducted between June 2017 and May 2018 at Sibu
and Kapit Hospitals for viral etiological causes.[9]
Participants
Definition of the LRTI was described in details in the previous study
publication.[9] A medical officer would have evaluated eligible subjects for
inclusion and exclusion criteria, including confirmation by chest radiography
within 72 hours of hospitalization. We recruited all the children under five
years of age with a complete medical record who had a nasopharyngeal swab that
was positive by real-time polymerase chain reaction (PCR) or real-time reverse
transcription PCR for RSV, Flu, PIV, AdV, EV and/or CoV. The laboratory assays
used to examine the nasopharyngeal swabs for viral agents were described in the
previous study publication. [9]
Variables and Data
Collection
We reviewed the demographic and laboratory data as well as medical
records of all children with viral infections. We studied the disease burden,
including the length of stay (LOS), duration of symptoms (fever, activity and
feeding status) as well as the duration of treatment [intravenous (IV) fluids
rehydration, ventilation support, nebulization and antibiotics usage].
Statistical Analysis
We used Statistical Package for Social Sciences software (SPSS) version
22 to analyze the data. Demographic and laboratory data were examined using
descriptive statistics. We compared the categorical variables by using
chi-square or Fisher’s exact tests as appropriate. We used an independent
t-test to compare the mean between groups. The significance level was set at
0.05 and 95% confidence intervals (CIs) were calculated.
The initial study was approved by the Medical Research and Ethics
Committee, Ministry of Health Malaysia (NMRR-12-16-10787). Written informed
consent was obtained from all parents or caregivers; and the study was
conducted according to the Declaration of Helsinki. We report according to the
standard set by the Strengthening the Reporting of Observational Studies in
Epidemiology (STROBE) Statement.[10]
RESULTS
During the study period, 385 children under five years of age provided a
nasopharyngeal swab specimen, of which, 222 (57.7%) were positive for one or
more viruses. Among the children who were positive for viral infection, 196
(88.3%) had complete medical records and were recruited into the secondary
study, 115 (58.7%) of whom were hospitalized at Sibu Hospital and 81 (41.3%) at
Kapit Hospital. Table 1 summarizes the basic demographic data of the children,
as well as the length of stay (LOS), duration of fever, days being less active
and days of reduced feeding. The mean LOS for the cohort was 7.7 (SD: 4.20)
days, with 13 (6.6%) of them stayed for more than 2 weeks (max = 21 days).
There was no death during the admission among the children studied in this
cohort, and all of them were discharged well.
Table 1. Demographic data, clinical symptoms and treatment of children under five years of age with lower respiratory tract infection
|
Frequency (%) n=196 |
Mean (SD) |
Age (years) |
- |
1.3 (1.08) |
Gender Male Female |
105 (53.6) 91 (46.4) |
- - |
LOS (days) |
- |
7.7 (4.20) |
Fever (days) |
- |
4.3 (15.46) |
Less active (days) |
- |
2.8 (3.31) |
Reduced feeding (days) |
- |
3.6 (3.70) |
IV Hydration (hours) |
109 (55.6) |
21.1 (31.03) |
NPO2 (hours) |
179 (91.3) |
98.0 (122.41) |
Nebulization (hours) Salbutamol Saline Ipratropium bromide |
165 (84.2) 136 (69.4) 61 (31.1) |
121.1 (99.62) 81.3 (90.46) 17.6 (43.53) |
Steroids No Yes |
179 (91.3) 17 (8.7) |
- |
IV MgSO4 No Yes |
137 (69.9) 59 (30.1) |
- |
SC Terbutaline No Yes |
160 (81.6) 36 (18.4) |
- |
HFNC No Yes |
129 (65.8) 67 (34.2) |
- |
Intubation No Yes |
188 (95.9) 8 (4.1) |
- |
Antibiotics (types) 0 1-3 4-6 |
35 (17.9) 151 (77.0) 10 (5.1) |
- |
SD= standard
deviation, LOS= length of stay, IV= intravenous, NPO2= nasal prong oxygen,
MgSO4= magnesium sulphate, SC= subcutaneous, HFNC= high flow nasal cannula
The analysis of treatments (Table 1) shows that children with LRTI
required a significant number of treatments with IV hydration, nasal prong
oxygen supply, salbutamol, saline and ipratropium bromide nebulization with the
mean durations of 21.1 (±31.03), 98.0 (±122.41), 121.1 (± 99.62), 81.3 (±
90.46) and 17.6 (± 43.53) hours respectively. While there were 17 children
(8.7%) received steroids, 30.1% (n = 59) received IV magnesium sulphate (MgSO4)
and 18.4% (n = 36) received subcutaneous terbutaline treatment. One third (n =
67, 34.2%) of children received high flow nasal cannula oxygen (HFNC)
treatment. Eight children (4.1%) were admitted to pediatric intensive care unit
and were ventilated. The majority of children (n = 151, 77.0%) received one to
three types of antibiotics during hospitalization, while 5.1% (n = 10) of
children were treated with four to six types of antibiotics during the hospital
stay.
Table 2. Comparison of disease burden between Adenovirus (AdV) and
non-Adenovirus (non-AdV) infection
|
Mean (SD) |
Mean Difference (95% CI) |
t statistic (df) |
P value* |
|
AdV |
Non-AdV |
||||
Age (years) |
1.0 (0.67) |
1.4 (1.16) |
-0.4 (-0.66, -0.12) |
-2.83 (136.15) |
0.005 |
LOS (days) |
10.1 (4.12) |
7.0 (3.94) |
3.2 (1.85, 4.48) |
4.74 (194) |
<0.001 |
Fever (days) |
7.4 (5.05) |
6.2 (3.58) |
1.2 (-0.37, 2.84) |
1.54 (59.87) |
0.129 |
Less active (days) |
3.5 (4.10) |
2.6 (3.02) |
0.9 (-0.44, 2.23) |
1.35 (57.63) |
0.183 |
Reduced feeding (days) |
3.9 (4.57) |
3.5 (3.40) |
0.3 (-1.13, 1.82) |
0.47 (59.48) |
0.643 |
IV Hydration (hours) |
33.8 (40.42) |
17.0 (26.26) |
16.8 (4.18, 29.32) |
2.67 (58.92) |
0.010 |
Nebulization (hours) Saline Salbutamol Ipratropium
bromide |
128.4 (97.69) 151.5 (95.61) 24.02 (44.27) |
66.3 (82.89) 111.32 (99.23) 15.62 (43.26) |
62.1 (30.56, 93.62) 40.1 (7.56, 72.70) 8.4 (-6.21, 23.02) |
3.93 (68.47) 2.43 (190) 1.134 (190) |
<0.001 0.016 0.258 |
NPO2 (days) |
119.7 (78.23) |
91.2 (132.83) |
28.4 (-12.30, 69.17) |
1.38 (190) |
0.170 |
|
Frequency (%) |
OR (95% CI) |
c2 statistic (df) |
P value† |
|
AdV
(n=47) |
Non-AdV
(n=149) |
||||
Steroids |
11 (23.4) |
6 (4.0) |
7.3 (2.52, 21.01) |
- |
<0.001‡ |
IV MgSO4 |
22
(46.8) |
37
(24.8) |
2.7
(1.35, 5.27) |
7.19
(1) |
0.007 |
SC Terbutaline |
10 (21.3) |
26 (17.4) |
1.3 (0.57, 2.89) |
0.140 (1) |
0.708 |
HFNC |
23
(48.9) |
44
(29.5) |
2.3
(1.17, 4.48) |
5.15
(1) |
0.023 |
Intubation |
5 (10.6) |
3 (2.0) |
5.8 (1.33, 25.25) |
- |
0.020‡ |
Antibiotics |
44
(93.6) |
117
(78.5) |
4.0
(1.17, 13.77) |
4.57
(1) |
0.033 |
SD= standard deviation, LOS= length of stay, IV= intravenous, NPO2=
nasal prong oxygen
IV= intravenous, MgSO4= magnesium sulphate, SC= subcutaneous, HFNC= high flow
nasal cannula
*Independent t-test, equal variance assumed
†Yates’ chi-square test
‡Fisher exact test
Among all respiratory viruses identified, RSV was the most common virus
detected with the prevalence of 54.1% (RSV A = 74, 37.8%; RSV B = 32, 16.3%),
followed by AdV (n = 47, 24.0%), Flu virus (Flu A = 38, 19.4%; Flu B = 5,
2.6%), EV (n = 20, 10.2%), PIV (PIV1 = 4, 2.0%; PIV2 = 1, 0.5%) and coronavirus
(n = 3, 1.5%). However, Flu virus was more prevalent than AdV in Kapit Hospital
(n = 15, 18.5% vs 12, 14.8%) compared to Sibu Hospital (n = 28, 24.4% vs n =
35, 30.4%). (Figure 1)
When we compared the burden of disease in children infected with the
different viruses, children with RSV infection had shorter period of fever (1.6
days, 95% CI: 0.49, 2.80, t statistic (df) = -2.82 (150.18), p = 0.005), but
otherwise no statistical difference in symptoms or required treatment compared
to those with infection
Figure 1. Prevalence of respiratory virus detections among children
under five years of age with lower respiratory tract infections. RSV=
respiratory syncytial virus, AdV= adenovirus, Flu= influenza, EV= enterovirus,
PIV= parainfluenza virus, CoV= coronavirus.
caused by non-RSV viruses (Supplementary Table 1). By comparing AdV and
non-AdV infection (Table 2), children with AdV infection were significantly
younger (0.4 years, 95% CI: 0.12, 0.66, t statistic (df) = -2.83 (136.15), p =
0.005) and more likely to have longer LOS (3.2 days, 95% CI: 1.85, 4.48, t
statistic (df) = 4.74 (194.00), p < 0.001), duration of IV hydration (16.8
hours, 95% CI: 4.18, 29.32, t statistic (df) = 2.67 (58.92), p = 0.010) as well
as duration for nebulization using saline (62.1 hours, 95% CI: 30.56, 93.62, t
statistic (df) = 3.93 (68.47), p < 0.001) and salbutamol (40.1 hours, 95% CI:
7.56, 72.70, t statistic (df) = 2.43 (190.00), p = 0.016). It was also clear
that more children with AdV infection were treated with steroid (n=11, 23.4%,
OR: 7.3, 95%CI: 2.52, 21.01, p < 0.001), MgSO4 (n=22, 46.8%, OR = 2.7, 95%
CI: 1.35, 5.27, 2 = 7.19, p = 0.007), HFNC (n=23, 48.9%, OR = 2.3, 95% CI:
1.17, 4.48, 2 = 5.15, p = 0.023), being intubated (n=5, 10.6%, OR = 5.8, 95%
CI: 1.33, 25.25, p = 0.020) and treated with antibiotics (n=44, 93.6%, OR =
4.0, 2 = 4.57, p = 0.033).
Children with influenza infection were more likely to be older (0.4
years, 95% CI: 0.07, 0.80, t statistic (df) = 2.37 (194.00), p = 0.019) and
have a prolonged fever (2.2 days, 95% CI: 0.83, 3.54, t statistic (df) = 3.17
(190.00), p = 0.002), while requiring shorter duration of saline (36.9 hours,
95% CI: 12.70, 61.09, t statistic (df) = -3.02 (102.76), p = 0.003) and
ipratropium bromide nebulization (13.3 hours, 95% CI: 2.83, 23.73, t statistic
(df) = -2.50 (138.76), p = 0.013) compared to those with non-influenza
infection (Table 3). Children with flu infection were also negatively
associated with MgSO4 treatment (n=6, 14.3%, OR = 0.3, 95% CI: 0.13, 0.80, 2 =
5.44, p = 0.020). Children with EV infection had a shorter duration of fever
(1.9 days, 95% CI: 0.01, 3.80, t statistic (df) = -1.98 (190.00), p = 0.049)
and reduced feeding (2.2 days, 95% CI: 1.03, 3.43, t statistic (df) = -3.79
(32.11), p = 0.001). There was no significant difference between EV and non-EV
infection in term of duration of treatment (Supplementary Table 2).
Table 3. Comparison of disease burden between Influenza (Flu) and non-Influenza
(non-Flu) infection
|
Mean (SD) |
Mean
Difference (95% CI) |
t statistic (df) |
P value* |
|
Flu |
Non-Flu |
||||
Age (years) |
1.7 (1.38) |
1.2 (0.96) |
0.4 (0.07, 0.80) |
2.37 (194) |
0.019 |
LOS (days) |
6.8 (4.29) |
8.0 (4.15) |
-1.1 (-2.57, 0.30) |
-1.56 (194) |
0.121 |
Fever (days) |
8.2 (3.42) |
6.0 (4.03) |
2.2 (0.83, 3.54) |
3.17 (190) |
0.002 |
Less active (days) |
3.2 (3.05) |
2.7 (3.38) |
0.6 (-0.60,1.70) |
0.94 (190) |
0.346 |
Reduced feeding (days) |
4.5 (3.22) |
3.4 (3.79) |
1.2 (-0.09, 2.44) |
1.83 (192) |
0.068 |
IV Hydration (hours) |
21.1 (32.47) |
21.1 (25.28) |
0 (-10.82, 10.76) |
-0.01 (192) |
0.996 |
Nebulization (hours) Saline Salbutamol Ipratropium bromide |
89.3 (95.67) 95.4 (110.60) 7.21 (23.28) |
52.4 (61.00) 127.7 (95.91) 20.5 (47.33) |
-36.9 (-61.09, -12.70) -32.4 (-67.39, 2.68) -13.3 (-23.73, -2.83) |
-3.02 (102.76) -1.82 (190) -2.50 (138.76) |
0.003 0.070 0.013 |
NPO2 (days) |
95.45 (191.95) |
98.8 (95.23) |
-3.3 (-45.56, 38.96) |
-0.15 (190) |
0.878 |
|
Frequency (%) |
OR (95% CI) |
c2 statistic (df) |
P value† |
|
Flu (n=42) |
Non-Flu (n=154) |
||||
Steroids |
1 (2.4) |
16 (10.4) |
0.2 (0.03, 1.63) |
- |
0.128‡ |
IV MgSO4 |
6 (14.3) |
53 (34.4) |
0.3 (0.13, 0.80) |
5.44 (1) |
0.020 |
SC Terbutaline |
3 (7.1) |
33 (21.4) |
0.3 (0.08, 0.97) |
3.59 (1) |
0.058 |
HFNC |
9 (21.4) |
58 (37.7) |
0.45 (0.20, 1.01) |
3.18 (1) |
0.075 |
Intubation |
1 (2.4) |
7 (4.5) |
0.5 (0.06, 4.28) |
- |
1.000‡ |
Antibiotic |
37 (88.1) |
124 (80.5) |
1.8 (0.65, 4.94) |
0.83 (1) |
0.363 |
SD= standard deviation, LOS= length of stay, IV= intravenous, NPO2=
nasal prong oxygen,
MgSO4= magnesium sulphate, HFNC= high flow nasal cannula
*Independent t-test, equal variance assumed
†Yates’ chi-square test
‡Fisher exact test
Of the total 196 children with LRTI, 34 (17.3%) had respiratory viral
co-infection, of which one of these children was co-infected with three types
of respiratory viruses (Table 4). Among all the respiratory viruses we
identified, PIV1 had the highest co-infection rate although the number were
small (n=3, 75.0%, OR = 15.6, 95%CI: 1.57, 154.72, p = 0.017). The other
viruses that had significant co-infection rate were AdV (n = 23, 48.9%, OR =
12.0, 95%CI: 5.20, 27.82, 2 = 40.18, p < 0.001); EV (n = 9, 45.0%, OR =
4.9, 95%CI: 1.86, 13.13, p = 0.002); and RSV B (n = 12, 37.5%, OR = 3.9, 95%CI:
1.66, 9.01, 2 = 9.22, p = 0.002). . There was no significant difference in the
symptoms and treatments provided, when we compared the disease burden between
children with co-infections versus single infection (Supplementary Table 3).
Nevertheless, children with co-infection were more likely to be intubated
during hospitalization (11.8%, OR = 5.3, 95% CI: 1.25, 22.23, p = 0.032).
Table 4. Distribution of respiratory viral co-infection among children
under five years of age with lower respiratory tract infection and odd ratio
for having co-infection
Viral Type |
Co-infection vs Single Infection |
||||
Frequency
(%) |
OR
(95%CI) |
c2 statistic (df) |
P
value† |
||
Co-infection
(n=34) |
Single
infection (n=162) |
||||
AdV |
23
(67.6) |
24
(14.8) |
12.0
(5.20, 27.82) |
40.18
(1) |
<0.001 |
RSV
A |
13
(38.2) |
61
(37.7) |
1.0
(0.48, 2.19) |
0.00
(1) |
1.000 |
RSV
B |
12
(35.3) |
20
(12.3) |
3.9
(1.66, 9.01) |
9.22
(1) |
0.002 |
EV |
9
(26.5) |
11
(6.8) |
4.9
(1.86, 13.13) |
- |
0.002‡ |
Flu
A |
6
(17.6) |
32
(19.8) |
0.9
(0.33, 2.28) |
0.00
(1) |
0.965 |
PIV
1 |
3
(8.8) |
1
(0.6) |
15.6
(1.57, 154.72) |
- |
0.017‡ |
Flu
B |
2
(5.9) |
3
(1.9) |
3.3
(0.53, 20.63) |
- |
0.208‡ |
PIV
3 |
1
(2.9) |
6
(3.7) |
0.8
(0.09, 6.76) |
- |
1.000‡ |
CoV |
0
(0.0) |
3
(1.9) |
- |
- |
1.000‡ |
PIV
2 |
0
(0.0) |
1
(0.6) |
- |
- |
1.000‡ |
RSV= respiratory syncytial virus, AdV= adenovirus, Flu= influenza virus,
EV= enterovirus, PIV= parainfluenza virus, CoV= coronavirus
†Yates’ chi-square test
‡Fisher exact test
DISCUSSION
Our study showed that RSV accounted more than half of the viral causes
of LRTI among the children under five years of age in Sibu and Kapit Hospitals,
followed by about one-fifth each by AdV and Flu virus. These findings are
consistent with the existing literature; RSV is understood to be the primary
pathogen that causes LRTI in under-five children globally, accounting for 3.2
million hospitalizations and 59,600 in-hospital mortality in 2015.[11]
In this study, LRTI with different viral etiologies was found to result
in different disease burdens among under-five children in terms of clinical
presentation and treatment requirements. These differences may be due to
different pathogenesis of the viruses and the immune responses they induce.[12]
Our study showed that children with AdV were 4.8 months younger whereas those
with Flu virus were 4.8 months older. These findings are consistent with
studies from other countries. [13, 14, 15]
Our findings were in agreement with other studies, which showed that
children with AdV had LOS of about 10 days and had more severe disease burden,
including the needs for nebulization, oxygen therapy and intubation.[13]
Nevertheless, other studies have suggested children with AdV are less likely to
develop lower respiratory illness.[14] The discrepancy between those reports
and our findings could be explained by distinct AdV serotypes among the
children. It is known that several AdV serotypes such as serotypes 3 and 7 were
involved in outbreak and caused severe disease burden. [14, 15] Our team had
successfully sequenced 25 AdV-positive specimens from these under-five
population and 56% of them were detected as AdV type 7.[16] This suggested that
children with AdV type 7 infection might result in more severe burden of
disease than other viral infections analyzed in our study. A study in Malaysia
on AdV also revealed that AdV type 7 was the most common serotype circulating
in children and that severe respiratory illness was associated with prolonged
LOS.[17]
Several EV strains are responsible for LRTI with different clinical
severity.[18] Although EV is self-limiting and associated with less severe
respiratory illness,[18] one of the strain, EV-D68 was reported to cause asthma
exacerbation and severe respiratory illness.[ 19, 20] In this current study,
there were about 10% of children infected with EV virus, and of the positive EV
specimens successfully sequenced, there were three EV-71 specimens and one
EV-D68 detected and one coxsackievirus B5.[16] However, they did not seem to
have much difference in terms of disease burden compared to those without EV
virus, may be due to its relatively smaller number of children with EV.
We detected viral co-infection in 17% of the specimens collected from
children in our study, of which, PIV 1 as well as AdV, EV and RSV B were the
most common viruses with such co-infection. Our study found that children with
co-infections were more likely to be intubated compared to those with a single
infection. However, the clinical severity of disease among children with
co-infection was still controversial with some findings that suggested to have
increased clinical severity while some did not due to different study designs,
seasonality and pathogens covered in the studies. [ 21, 22]
These findings, along with the seasonal distribution of the viruses, age
distribution and social background, provide important information for
clinicians to understand which pathogens may be causing respiratory infections
among the under-five population, thus improving their clinical practice. By
knowing which pathogens are associated with the infection, clinicians can both
focus on symptomatic treatment and target the specific pathogen. For example,
children with influenza virus infection can be treated by oseltamivir or
zanamivir. These antiviral agents can reduce the duration of illness in
children.[23] Currently, there are several antiviral therapies for RSV under
research.[24] Detection of viral infection of LRTI can also help to reduce the
misuse of antibiotics. Furthermore, understanding which pathogens are
predominant and causing a more severe disease burden is important for the
development of antiviral agents, vaccines and prophylaxis treatments.
This was the first of this kind of epidemiological study for LRTI from
Sarawak, and it reflected a true clinical practice of children with LRTI,
showing the accurate disease burden. Nevertheless, the study has a few
limitations. Study subjects were collected through convenience sampling and
hence may not represent the general LRTI population; we believe children with
more serious conditions were recruited during the study period. The actual
disease burden for the viruses in overall LRTI may be lower, which likely would
have been reflected had we included children in the study with milder LRTI
disease manifestation. We also did not study the bacteriological causes of
these LRTI, although during the recruitment, if the children were to have known
bacteriological causes, they would not have been recruited. Because the NP were
only processed in batches due to the study nature, and sometimes weeks later,
the clinicians did not have the results in hands most of the time while the
children were in the wards. We also could not be completely sure that molecular
evidence of viruses in the NP swab were truly the cause of the LRTI morbidity.
Therefore, a further study to compare the swab-negative LRTI cases with the
swab-positive cases is required. Especially in young children, prolonged
shedders and the issue of difficulty getting a good sample may skew the
positive isolation rates either way. Other factors that might have influenced
on the prevalence of viruses and / or disease burdens include flu vaccine
uptake, bacterial-viral co-infection and co-morbidities such as prematurity or
chronic lung / heart diseases. This information were not available to our
study, although in Sarawak, the flu vaccine uptake among children population is
generally very low.
CONCLUSION
This study compared metrics for disease burden for several viruses
associated with LTRI in Central Sarawak. In this study, LTRI hospitalizations
due to AdV infection among children under five years of age were more severe
compared to LRTI hospitalizations due to other causes, although RSV was the
commonest cause among all hospitalizations. Having the diagnostic capability to
determine the cause of LRTI is critically important in developing effective
programs to prevent and treat infections.
ACKNOWLEDGEMENT
We thank the Director-General of Health, Malaysia for his approval to
publish the findings in this study. This work was conducted with support from
the Duke University, the Duke Global Health Institute, Clinical Research Centre
Sibu Hospital and SEGi University Sibu Clinical Campus. We thank the pediatric
doctors in Sibu Hospital and Kapit Hospital for enrolling the patients and for
gathering essential clinical data.
REFERENCES: