This is the author’s version of a work that was submitted/accepted for publication in the following source:
Xu, Zhiwei, Hu, Wenbiao, & Tong, Shilu (2014) The geographical codistribution and socio-ecological drivers of childhood pneumonia and diarrhoea in Queensland, Australia. Epidemiology and Infection. (In Press)
This file was downloaded from: http://eprints.qut.edu.au/74391/
c Copyright 2014 Cambridge University Press
Notice: Changes introduced as a result of publishing processes such as
copy-editing and formatting may not be reflected in this document. For a
definitive version of this work, please refer to the published source:
http://dx.doi.org/10.1017/S095026881400171X
Epidemiology and Infection
The geographic co-distribution and socio-ecological drivers
of childhood pneumonia and diarrhea in Queensland,
Australia
r
Fo
Journal:
Manuscript ID:
Manuscript Type:
Date Submitted by the Author:
Original Manuscript
n/a
Xu, Zhiwei; Queensland University of Technology, School of Public Health
and Social Work
Hu, Wenbiao
Tong, Shilu
vi
Keyword:
HYG-OM-5365-Dec-13.R1
Re
Complete List of Authors:
Epidemiology and Infection
Diarrhoea, Pneumonia
ew
ly
On
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 1 of 23
1
The geographic co-distribution and socio-ecological drivers of
2
childhood pneumonia and diarrhea in Queensland, Australia
3
Zhiwei Xu, Wenbiao Hu, Shilu Tong*
4
5
Affiliations for all authors:
6
School of Public Health and Social Work & Institute of Health and Biomedical Innovation,
7
Queensland University of Technology, Brisbane, Australia.
8
*Corresponding author:
9
Dr. Shilu Tong, School of Public Health and Social Work & Institute of Health and
r
Fo
vi
Re
10
Biomedical Innovation, Queensland University of Technology, Kelvin Grove, Brisbane, Qld.
11
4059, Australia. Telephone: 61-7-3138 9745; Fax: 61-7-3138 3369; Email:
12
[email protected]
13
Running head:
14
Childhood pneumonia and diarrhea in Queensland
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
15
16
17
18
19
20
1
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
Summary
2
This study aimed to explore the spatiotemporal patterns, geographic co-distribution, and
3
socio-ecological drivers of childhood pneumonia and diarrhea in Queensland. A Bayesian
4
conditional autoregressive model was used to quantify the impacts of socio-ecological factors
5
on both childhood pneumonia and diarrhea at a postal area level. A distinct seasonality of
6
childhood pneumonia and diarrhea was found. Childhood pneumonia and diarrhea mainly
7
distributed in northwest of Queensland. Mount Isa was the high-risk cluster where childhood
8
pneumonia and diarrhea co-distributed. Emergency department visits (EDVs) for pneumonia
9
increased by 3% per 10-mm increase in monthly average rainfall, in wet seasons. In
r
Fo
10
comparison, a 10-mm increase in monthly average rainfall may increase 4% of EDVs for
11
diarrhea. Monthly average temperature was negatively associated with EDVs for childhood
12
diarrhea, in wet seasons. Low socioeconomic index for areas (SEIFA) was associated with
13
high EDVs for childhood pneumonia. Future pneumonia and diarrhea prevention and control
14
measures in Queensland should focus more on Mount Isa.
15
Keywords: Climate change, pneumonia, diarrhea, geographic co-distribution
ew
vi
Re
On
16
ly
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 2 of 23
17
18
19
20
21
22
2
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 3 of 23
1
INTRODUCTION
2
Pneumonia and diarrhea are the leading causes of child mortality [1]. In 2011, two million
3
children died before reaching their fifth birthday because of pneumonia and diarrhea
4
worldwide [2]. Though declining trend in mortality from pneumonia and diarrhea have been
5
witnessed in some industrialized countries, pneumonia and diarrhea are still an important
6
source of morbidity in these regions [3]. For example, in Western Pacific region, the total
7
episodes of pneumonia and diarrhea in children under five years of age were 256.3 million
8
and 12.2 million in 2010 [2].
9
Pneumonia and diarrhea are largely preventable, and hence it is essential to identify the risk
r
Fo
Re
10
factors and take targeted preventive measures [4]. Existing studies have confirmed some
11
individual-level biological factors for pneumonia and diarrhea, such as underweight, stunting
12
and zinc deficiency [2]. Some of these poverty-related risk factors, such as suboptimum
13
breastfeeding, under-nutrition and zinc deficiency, are shared by pneumonia and diarrhea,
14
and these overlapping risk factors may result in the geographic co-distribution of pneumonia
15
and diarrhea [5].
16
Prior studies also have reported that climate may play an important role in the transmission of
17
pneumonia and diarrhea, highlighting that high temperature [6, 7] and rainfall [8, 9] may
18
trigger pneumonia and diarrhea. As Intergovernmental Panel on Climate Change projected,
19
the Earth’s surface average temperature will increase, and there will be more intense rainy
20
seasons in Asia, Africa, and the Pacific [10]. As climate change continues, the burden of
21
pneumonia and diarrhea in these regions may increase, though there are still regional
22
differences and contrasting effects of climate on pneumonia and diarrhea due to different
23
aetiological agents.
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
3
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
Australia shoulders a considerable burden of childhood pneumonia and diarrhea [11, 12]. It is
2
urgently needed to reveal the spatiotemporal patterns of childhood pneumonia and diarrhea in
3
Australia. This study explored the spatiotemporal patterns, geographic co-distribution and
4
socio-ecological determinants of childhood pneumonia and diarrhea in Queensland, Australia.
5
6
MATERIALS AND METHODS
7
Data collection
8
Queensland is located in the northeast of Australia. Its mean temperature of summer is 25 °C
9
and mean temperature of winter is 15 °C. There is significant variation in mean annual
r
Fo
Re
10
rainfall across Queensland, varying from less than 150 mm/year in the southwest to 4000
11
mm/year in the northern coast [13]. Data on emergency department visits (EDVs) by
12
postcode from January 1st 2007 through 31st December 2011 in Queensland were obtained
13
from Queensland Health. The anonymised EDV data were classified according to the
14
International Classification of Disease, 10th version (ICD–code10). In this study, we included
15
EDVs with the principle cause coded as pneumonia (ICD–10 codes: J12–J18) and diarrheal
16
disease of any cause (ICD–10 codes: A00–A03, A04, A05, A06.0–A06.3, A06.9, A07.0–
17
A07.2, A07.9, A08–A09) among children aged 0–14 years. Ethical approval was obtained
18
from the Human Research Ethics Committee of Queensland University of Technology
19
(Australia) prior to the data collection. Patient information was de-identified and thus no
20
written informed consent was obtained. Data on weather (including temperature and rainfall)
21
were supplied by the worldclimate website, including interpolated monthly mean temperature
22
and monthly rainfall. Temperature and rainfall values for each postal area during the study
23
period were extracted using the ArcMap software package (ESRI Inc., Redlands, CA, USA).
24
Data for the same period for each postcode on the socioeconomic index for areas (SEIFA)
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 4 of 23
4
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 5 of 23
1
and population size, were obtained from the Australian Bureau of Statistics [13]. SEIFA is a
2
product developed by the Australian Bureau of Statistics that ranks areas in Australia
3
according to relative socio-economic advantage and disadvantage. Lower values of SEIFA
4
indicate lower socioeconomic status.
5
Statistical analysis
6
We plotted the decomposed daily distributions of EDVs for childhood pneumonia and
7
diarrhea using time-series approach. The change of EDVs for childhood pneumonia and
8
diarrhea from 2008-2009 to 2010-2011 was calculated using the following equation:
9
Mc = ( EDVi 2010−2011 − EDVi 2008−2009 ) / populationi
r
Fo
Re
10
Where Mc represents the morbidity change, EDVi 2010−2011 represents the EDVs for childhood
11
pneumonia (diarrhea) for postal area i during 2010-2011, EDVi 2008−2009 represents the EDVs
12
for childhood pneumonia (diarrhea) for postal area i during 2008-2009, and populationi
13
refers to the population for postal area i.
14
Bayesian conditional autoregressive (CAR) model [14] was used to estimate the relative risk
15
of EDVs for childhood pneumonia and diarrhea in postcode districts with a range of
16
independent variables, including temperature, rainfall and SEIFA.
17
Oi ~ Poisson( µi )
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
log( µi ) = log( Ei ) + α + β1 ( temi ) + β 2 (rainfalli ) + β 3 (SEIFA i ) + u i +v i
18
where Oi is the observed counts of pneumonia (diarrhea) from the ith postcode (i=1…424),
19
Ei , the expected number of cases in postal area i, is an offset to control for population, α is
20
the intercept, β1 is the coefficient for temperature, β2 is the coefficient for rainfall, β3 is the
5
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
coefficient for SEIFA, ui is a spatially structured random effect with mean zero and variance
2
σu2, and vi is a spatially unstructured random effect with mean zero and variance σv2.
3
We conducted an initial burn-in of 10,000 iterations that were subsequently discarded.
4
Convergence was evaluated through visual inspection of posterior density plots, history plots,
5
and autocorrelation of selected parameters, and it was reached within the first 10,000
6
iterations for the model. We conducted a subsequent set of 200,000 iterations for the accuracy.
7
Model selection was conducted by comparing the deviance information criterion (DIC) of
8
different models. In this study, we defined May to October as dry season, and January,
9
February, March, April, November and December as wet season.
r
Fo
10
Time-series analysis was conducted using the R statistical environment, version 2.15.3.
11
Visual maps were created using ArcGIS version 9.3 (ESRI Inc., Redlands, CA, USA). Spatial
12
cluster analysis was conducted using SatScan version 9.1, and Bayesian CAR model was
13
conducted using WinBugs software, version 1.4.3 (MRC Biostatistics Unit 2008).
ew
vi
Re
14
15
RESULTS
16
Summary statistics
17
Table 1 presents the summary statistics of EDVs for childhood pneumonia and diarrhea,
18
mean temperature, rainfall and SEIFA by postcode in Queensland. The average counts of
19
childhood pneumonia and diarrhea were 43.7 and 138.5, respectively, and the mean values of
20
mean temperature, rainfall and SEIFA were 20.1°C, 95.7 mm, and 976.6.
21
Spatial pattern
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 6 of 23
6
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 7 of 23
1
Figure 1 shows the spatial distribution of rates of EDVs for childhood pneumonia and
2
diarrhea, illustrating that EDVs for both pneumonia were the highest in central west,
3
northwest and far north of Queensland, and the EDVs for childhood diarrhea were the highest
4
in the northwest of Queensland (Mount Isa). Figure 2 illustrates the change in EDVs for
5
childhood pneumonia and diarrhea from years 2008-2009 to 2010-2011, indicating that EDVs
6
for pneumonia and diarrhea changed from northwest or southeast of Queensland in the past
7
couple of years.
8
Temporal pattern
9
Figure 3 shows the decomposed daily distributions of EDVs for childhood pneumonia and
10
diarrhea, showing a distinct seasonal trend for the two diseases, especially for pneumonia.
11
This figure indicates that EDVs for childhood pneumonia in Queensland were more likely to
12
occur in cold season. The particularly great number of EDVs for childhood pneumonia in
13
2009 is because of the 2009 pandemic H1N1 influenza.
14
Geographical co-distribution
15
The cluster results in Figure 4 reveal that EDVs for childhood pneumonia and diarrhea in
16
Queensland were co-distributed in Mount Isa.
17
Socio-ecological drivers
18
The effect of socio-ecological factors on EDVs for childhood pneumonia and diarrhea in dry
19
and wet seasons is reflected in Table 2. SEIFA played an important role in driving the
20
distribution of pneumonia, highlighting that more EDVs for pneumonia occurred in the
21
regions with low socioeconomic status. The relationship between rainfall and EDVs for
22
pneumonia was significant in wet seasons, with 3% (95% confidence interval (CI): 1% –5%)
23
increase in EDVs for pneumonia by 10-mm increase in monthly average rainfall. Rainfall
r
Fo
ew
vi
Re
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
7
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
was also significantly associated with EDVs for diarrhea in both dry and wet seasons, with 4%
2
(95% CI: 2% –7%) increase of EDVs for diarrhea for 10 mm increase in monthly average
3
rainfall. Mean temperature was negatively associated with EDVs for diarrhea in wet seasons,
4
but not in dry seasons.
5
Posterior estimated RRs of childhood pneumonia and diarrhea reveal that high-risk areas of
6
childhood pneumonia were located in northwest of Queensland, and high-risk areas of
7
childhood diarrhea were located in central west of Queensland (Figure 5). Estimated residual
8
variation after taking into account the socio-ecological factors indicate that high-incidence
9
postal areas for childhood pneumonia were located in far north and northwest of Queensland,
r
Fo
10
and high-incidence postal clusters for childhood diarrhea were located in Mount Isa (Figure
11
6).
12
ew
vi
Re
13
DISCUSSION
14
This study has yielded several notable findings. There was a strong seasonal trend in EDVs
15
for childhood pneumonia, with more cases occurring in the cold season. Children suffering
16
pneumonia and diarrhea who visited emergency departments in Queensland from 2007 to
17
2011 were mainly from central west, northwest and far north of Queensland. According to the
18
cluster analysis results, Mount Isa is the high risk area for both childhood pneumonia and
19
diarrhea. Interestingly, in recent years, Mount Isa has been experiencing a substantial
20
decrease in EDVs for childhood pneumonia and diarrhea, and EDVs for childhood
21
pneumonia and diarrhea were moving from west to southeast of Queensland. We found
22
SEIFA played a relatively more important role than climate in the driving the spatial
23
transmission of childhood pneumonia, while climate may be more essential in the spread of
24
childhood diarrhea. Only in wet seasons, rainfall was significantly associated with EDVs for
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 8 of 23
8
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 9 of 23
1
childhood pneumonia. Low temperature may significantly increase EDVs for childhood
2
diarrhea, also only in wet seasons.
3
Mount Isa city, a major lead, zinc and copper producer, is the largest emitter of sulphur
4
dioxide, lead and some other metals in Australia [15]. It has been convincingly documented
5
that the blood lead level of children in Mount Isa, especially for those aged 1–4 years, is way
6
higher than children from other regions of Australia [16], and the consequent life-long
7
negative health and intellectual impacts of lead exposure on children have also been
8
extensively reported [17, 18]. In this study, we found that pneumonia and diarrhea in children
9
were co-distributed in Mount Isa, highlighting that there might be some common risk factors
r
Fo
10
in this area. Exposure to air pollutants (e.g., sulphur dioxide) emitted by mining could
11
increase hospital admissions for childhood pneumonia [19]. Mining also had a significant
12
adverse effect on semi-arid freshwater aquatic system in Mount Isa [20]. The densities of
13
bacterial indicators in remnant pools throughout Leichhardt River have exceeded the
14
acceptable guidelines, which might expose children to greater risk of diarrhea. In this study,
15
we also found the risk areas for childhood pneumonia and diarrhea changed from northwest
16
to southeast of Queensland, and the EDVs for childhood pneumonia and diarrhea in Mount
17
Isa have been decreasing sharply (though still high) in recent years, indicating that protective
18
measures may have been taken to prevent children from continuously adverse impacts of
19
mining.
20
In this study, we found the average score of SEIFA was significantly associated with
21
childhood pneumonia, but not for diarrhea, implying that socioeconomic factors may play a
22
more prominent role in pneumonia than diarrhea in Queensland. This finding conflicts with
23
previous study conducted in China which found socioeconomic factors played a more
24
important part in driving the transmission of pneumonia than diarrhea [21]. The inconsistency
25
reveals that the patterns of risk factors for pneumonia and diarrhea in developed and
ew
vi
Re
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
9
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
developing countries may not be the same, suggesting that future preventive measures should
2
focus on the economic characteristics of specific regions. With regards to the mechanism of
3
social and economic impacts on childhood pneumonia, we found in the literature that most
4
risk factors for childhood pneumonia (e.g., underweight) are socioeconomic-related [2].
5
Children in the lower socioeconomic groups appear to be living in more crowded houses and
6
suffer under-nutrition more often than those with higher socioeconomic status, possibly
7
increasing their risk of getting pneumonia.
8
High rainfall was found to be significantly associated with more pneumonia and diarrhea,
9
especially in wet seasons. Two studies, so far, have formally explored the relationship
r
Fo
10
between rainfall and childhood pneumonia in Philippine [22] and the USA [23], both using
11
time-series approach, but did not find significant results. In rainy days, children are more
12
likely to spend time indoors, which may increase crowding and their exposure to biomass
13
fuel smoke, and decrease their sunlight exposure, possibly resulting in a higher risk of getting
14
pneumonia. The association between high rainfall and more EDVs for childhood diarrhea we
15
found in this study corresponds to the findings of previous studies in Brazil [24] and the USA
16
[25]. Increased rainfall may raise the risk of sewer overflow which render water supply
17
contamination [26]. Further, runoff of human excreta in soil and subsurface may increase,
18
and result in more concentrations of pathogens in surface water. Turbulences may be caused
19
by increased heavy rainfall, leading to sediment re-suspension and dispersing accumulated
20
pathogens. Apart from high rainfall, low temperature was also found associating with more
21
diarrheal cases in wet seasons in this study, while little evidence on the temperature-
22
pneumonia relation was found. Rotavirus has been reported as the most important cause of
23
acute and severe childhood diarrhea in industrialized countries [27-29], and low temperature
24
increases the survival and replication of rotavirus [30].
ew
vi
Re
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 10 of 23
10
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 11 of 23
1
This study has several strengths. To the best of our knowledge, this is the first study to
2
explore the geographic co-distribution of childhood pneumonia and diarrhea. An advanced
3
Bayesian spatial model was used to quantify the effect of socio-ecological factors on both
4
childhood pneumonia and diarrhea. The results from this study, especially the high risk areas
5
of pneumonia and diarrhea we identified, may have important implications for future control
6
and prevention for childhood pneumonia and diarrhea in Queensland. Two major limitations
7
should also be acknowledged. First, the disease data we collected from emergency
8
departments may underestimate the actual infected population because only children with
9
severe symptoms would go to emergency departments for treatment. Second, only aggravated
10
r
Fo
data were used, which may result in some measurement bias.
11
Re
12
CONCLUSIONS
13
Childhood pneumonia and diarrhea were predominantly distributed in northwest of
14
Queensland, and Mount Isa was the region where these two childhood diseases co-distributed.
15
In recent years, the high risk areas of these two childhood diseases have been changing from
16
northwest to southeast of Queensland. Low temperature and high rainfall were associated
17
with more childhood diarrhea cases, and low SEIFA was associated with more childhood
18
pneumonia cases. Future precautionary measures should be taken before the rainy seasons to
19
prevent children from the impact of pneumonia and diarrhea.
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
20
21
ACKNOWLEDGEMENTS
11
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
ZX was funded by a China Scholarship Council Postgraduate Scholarship and Queensland
2
University of Technology fee waiving scholarship; ST was supported by a National Health
3
and Medical Research Council Research Fellowship (#553043).
4
5
DECLARATION OF INTERESTS
6
None.
7
r
Fo
8
REFERENCES
9
(1)
Liu L, et al. Global, regional, and national causes of child mortality: an updated
10
systematic analysis for 2010 with time trends since 2000. The Lancet 2012; 379(9832): 2151-
11
2161.
12
(2)
13
Lancet 2013; 381(9875): 1405-1416.
14
(3)
15
Seminars in Pediatric Infectious Diseases 2004; 15(3): 155-168.
16
(4)
17
diarrhoea equitably: what works and at what cost? The Lancet 2013; 381(9875): 1417-1429.
18
(5)
19
among young children in low income settings. Journal of Global Health 2013; 3(1): 10402.
20
(6)
21
for diarrhoeal diseases in Peruvian children. The Lancet 2000; 355(9202): 442-450.
22
(7)
23
counties. International Journal of Public Health 2010; 55(2): 113-121.
24
(8)
25
pneumophila Pneumonia. PLoS ONE 2013; 8(4): e61036.
Re
Walker CLF, et al. Global burden of childhood pneumonia and diarrhoea. The
ew
vi
Podewils LJ, et al. Acute, infectious diarrhea among children in developing countries.
On
Bhutta ZA, et al. Interventions to address deaths from childhood pneumonia and
ly
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 12 of 23
Walker C, et al. Diarrhea as a risk factor for acute lower respiratory tract infections
Checkley W, et al. Effect of El Niño and ambient temperature on hospital admissions
Green R, et al. The effect of temperature on hospital admissions in nine California
Garcia-Vidal C, et al. Rainfall Is a Risk Factor for Sporadic Cases of Legionella
12
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 13 of 23
1
(9)
Hashizume M, et al. Association between climate variability and hospital visits for
2
non-cholera diarrhoea in Bangladesh: effects and vulnerable groups. International Journal of
3
Epidemiology 2007; 36(5): 1030-1037.
4
(10)
5
basis. Contribution of Working Group I to the Fifth Assessment Report of the
6
Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. 2013.
7
(11)
8
Ireland, and the United States. International Journal of Epidemiology 2005; 34(2): 454-460.
9
(12)
IPCC. Summary for policymakers. In: Climate change 2013: the physical science
Scallan E, et al. Prevalence of diarrhoea in the community in Australia, Canada,
r
Fo
Rudan I, et al. Epidemiology and etiology of childhood pneumonia in 2010:
10
estimates of incidence, severe morbidity, mortality, underlying risk factors and causative
11
pathogens for 192 countries. Journal of Global Health 2013; 3(1): 10401.
12
(13)
13
http://www.abs.gov.au/CDataOnline In, 2010.
14
(14)
15
spatial statistics. Annals of the Institute of Statistical Mathematics 1991; 43(1): 1-20.
16
(15)
17
facility-detail/criteria/year/2009/browse-type/Company/regbusiness-
18
name/MOUNT%2BISA%2BMINES%2BLTD/jurisdiction-facility/Q020MIM001. . In, 2010.
19
(16)
20
into the results of a blood-lead screening program of 1–4 year old children in Mount Isa,
21
Queensland. http://www.health.qld.gov.au/ph/documents/tphn/mtisa_leadrpt.asp. In, 2009.
22
(17)
23
function: an international pooled analysis. Environmental Health Perspectives 2005; 113(7):
24
894-899.
Re
Australian Bureau of Statistics. Census Data. Available
ew
vi
Besag J, York J, Mollié A. Bayesian image restoration, with two applications in
National Pollutant Inventory. http://www.npi.gov.au/npidata/action/load/individual-
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
Queensland Health. Mount Isa community lead screening program 2006–07: a report
Lanphear B, et al. Low-level environmental lead exposure and children's intellectual
13
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
(18)
Tong S, et al. Declining blood lead levels and changes in cognitive function during
2
childhood: The port pirie cohort study. Journal of American Medical Association 1998;
3
280(22): 1915-1919.
4
(19)
5
Respiratory and Critical Care Medicine 2005; 171(11): 1272-1278.
6
(20)
7
the example of Mount Isa, Queensland. Journal of Environmental Monitoring 2009; 11(5):
8
977-986.
9
(21)
Barnett AG, et al. Air Pollution and Child Respiratory Health. American Journal of
Taylor MP, et al. Mining and urban impacts on semi-arid freshwater aquatic systems:
r
Fo
Feng X, et al. Social, economic, political and health system and program
10
determinants of child mortality reduction in China between1990 and 2006: A systematic
11
analysis. Journal of Global Health 2012; 2(1): 10405.
12
(22)
13
time-series analyses. Epidemiology & Infection 2013; 141(06): 1328-1336.
14
(23)
15
hospitalization for viral pneumonia in females: California, 1983-1998. American Journal of
16
Public Health 2001; 91(8): 1200-1208.
17
(24)
18
rates of hospitalization for diarrhoea in Brazilian infants. Transactions of The Royal Society
19
of Tropical Medicine and Hygiene 2009; 103(5): 506-511.
20
(25)
21
visits for acute gastrointestinal illness. Environmental Health Perspectives 2010; 118(10):
22
1439-1443.
23
(26)
24
disease outbreaks in the United States, 1948-1994. American Journal of Public Health 2001;
25
91(8): 1194-1199.
Re
PAYNTER S, et al. Sunshine, rainfall, humidity and child pneumonia in the tropics:
ew
vi
Ebi K, et al. Association of normal weather periods and El Niño events with
On
Andrade IG, et al. Improved sanitation and income are associated with decreased
ly
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 14 of 23
Drayna P, et al. Association between rainfall and pediatric emergency department
Curriero F, et al. The association between extreme precipitation and waterborne
14
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 15 of 23
1
(27)
2
Children Less Than 5 Years of Age: United States, 1997 and 2000. Pediatrics 2006; 117(6):
3
1887-1892.
4
(28)
5
Study. Journal of Clinical Microbiology 2005; 43(8): 3636-3641.
6
(29)
7
Children in 2004. Journal of Infectious Diseases 2009; 200(Supplement 1): S9-S15.
8
(30)
9
hospitalizations for rotavirus diarrhoea in children under 5 years of age. Epidemiology &
10
Malek MA, et al. Diarrhea- and Rotavirus-Associated Hospitalizations Among
Olesen B, et al. Etiology of Diarrhea in Young Children in Denmark: a Case-Control
Parashar UD, et al. Global Mortality Associated with Rotavirus Disease among
D'SOUZA RM, HALL G, BECKER NG. Climatic factors associated with
r
Fo
Infection 2008; 136(01): 56-64.
11
vi
Re
12
Figure legends
13
Figure 1. The spatial distribution of EDVs for childhood pneumonia and diarrhea in
14
Queensland, from 2007 to 2011.
15
Figure 2. The change of EDVs for childhood pneumonia and diarrhea in Queensland, from
16
2008-2009 to 2010-2011.
17
Figure 3. The daily distribution of EDVs for childhood pneumonia and diarrhea in
18
Queensland, from 2007 to 2011.
19
Figure 4. The spatial clusters of EDVs for childhood pneumonia and diarrhea in Queensland,
20
from 2007 to 2011.
21
Figure 5. Relative risk of childhood pneumonia and diarrhea from spatial CAR model.
22
Figure 6. Spatial random effects for EDVs for childhood pneumonia and diarrhea
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
15
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
1
2
Table 1. Summary statistics for EDVs for childhood pneumonia and diarrhea, monthly mean
temperature and rainfall, and SEIFA by postcode in Queensland, Australia, during 2007-2011
Variables
Mean
SD
Min
Max
Pneumonia (cases)
43.7
79.5
0
739
Diarrhea (cases)
135.8
247.7
0
1750
r
Fo
Mean temperature (°C)
20.1
2.1
13.9
26.8
Rainfall (mm)
95.7
42.0
19.7
318.0
74.3
589.0
1147.0
SEIFA
3
6
ew
5
976.6
vi
4
Re
7
On
8
9
ly
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 16 of 23
10
11
12
13
14
15
16
17
16
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 17 of 23
1
2
Table 2. Bayesian spatial CAR models of socioecological drivers of childhood pneumonia
and diarrhea in Queensland, Australia.
Variable
Poster mean ± SD
Monte Carlo error
RR(95% CI)
Intercept
-0.774 ± 0.122
<0.01
0.460 (0.365 – 0.590)
Temperature
-0.056 ± 0.040
<0.01
0.945 (0.876 – 1.024)
Rainfall
0.003 ± 0.001
<0.01
1.003 (0.999 – 1.005)
SEIFA
-0.001 ± 0.004
<0.01
0.998 (0.997 – 0.999)*
-0.724 ± 0.128
<0.01
0.486 (0.376 – 0.619)
-0.023 ± 0.041
<0.01
0.977 (0.901 – 1.058)
0.003 ± 0.001
<0.01
1.003 (1.001 – 1.005)*
-0.002 ± 0.001
<0.01
0.998 (0.997 – 0.999)*
Model 1: Pnuemonia (dry season)
r
Fo
Model 2 : Pnuemonia (wet season)
Intercept
Temperature
Rainfall
Model 1: Diarrhea (dry season)
ew
vi
SEIFA
Re
Intercept
-1.268 ± 0.126
<0.01
0.281 (0.220 – 0.362)
Temperature
-0.059 ± 0.043
<0.01
0.942 (0.887 – 1.028)
Rainfall
0.004 ± 0.001
<0.01
1.004 (1.002 – 1.007)*
SEIFA
0.001 ± 0.001
<0.01
1.000 (0.999 – 1.002)
ly
Model 2 : Diarrhea (wet season)
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
Intercept
-1.130 ± 0.124
<0.01
0.322 (0.254 – 0.410)
Temperature
-0.139 ± 0.040
<0.01
0.870 (0.805 – 0.943)*
Rainfall
0.004 ± 0.001
<0.01
1.004 (1.002 – 1.007)*
SEIFA
-0.001 ± 0.001
<0.01
0.999 (0.998 – 1.000)
3
4
5
17
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
r
Fo
vi
Re
296x185mm (96 x 96 DPI)
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 18 of 23
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 19 of 23
r
Fo
Re
330x199mm (96 x 96 DPI)
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
r
Fo
Re
315x188mm (96 x 96 DPI)
ew
vi
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 20 of 23
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 21 of 23
r
Fo
vi
Re
304x196mm (96 x 96 DPI)
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Epidemiology and Infection
r
Fo
vi
Re
296x187mm (96 x 96 DPI)
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 22 of 23
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK
Page 23 of 23
r
Fo
vi
Re
277x193mm (96 x 96 DPI)
ew
ly
On
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Epidemiology and Infection
London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK

"Sdy! IⅡ.THANH PHAN THAM D丁,DIA DЁM TttP HUAN

Who helped NEC find its perfect partner?

Additional topic for course

Quebec Grandparents Association

Sat 14 May - The Oxford and Cambridge Musical Club