3.1. Overview of overseas imported dengue fever cases in 13 provinces
3.1.1. Sources of the imported cases
From 2005 to 2019, a total of 11407 imported cases of dengue fever were reported in the 13 provinces in China. Among these cases, 11045 were imported from the Asian region, accounting for 96.83%, particularly from Southeast Asia which accounted for 94.50%. In addition, Africa, Oceania, South America and North America are also the sources of imported cases, accounting for 1.60%, 0.51%, 0.25% and 0.10%, respectively. The source of the remaining 0.72% of imported cases wasn’t identified.
3.1.2. Distribution areas and epidemic trend of imported cases
All the municipalities had imported dengue fever cases in the studied 13 provinces from 2005 to 2019 with a few exceptions for the northwestern part of Yunnan province, the northwestern part of Sichuan province, Shennongjia forest region of Hubei province and Rizhao city of Shandong province and a few areas in Chongqing (Fig. 1). The areas in western Yunnan Province bordering Myanmar had the largest number of overseas imported cases, reaching 2,485 cases, accounting for 21.78% of the total overseas imported cases in all the 13 provinces. In general, the overseas imported cases are more likely to be observed in the border areas of Yunnan Province, the southeast coastal areas and the capital areas with developed economies and convenient transportation (Fig. 2).
It can be seen from Table 1 that overseas imported dengue fever cases in most of the provinces began to increase since 2013 and accounted for 94.91% of the total cases, so here we mainly described the prevalence of the imported dengue fever cases from 2013 to 2019. During this period, Yunnan province and Guangdong province reported a large proportion of imported cases, accounting for 62.08% in total. The prevalence curve of these cases from 2013 to 2019, in provinces with more imported cases, showed a clear seasonal trend with most import cases happened from July to November (Fig. 3).
Table 1. The proportion of overseas imported dengue fever cases in provinces where dengue fever is locally outbreak in mainland China, 2005-2019.
Area
|
2005-2012
|
2013-2019
|
2005-2019
|
overseas imported dengue fever cases
|
Percentage(%)
|
overseas imported dengue fever cases
|
Percentage(%)
|
overseas imported dengue fever cases
|
Percentage(%)
|
Yunnan
|
177
|
30.46
|
4483
|
41.41
|
4660
|
40.85
|
Guangdong
|
173
|
29.78
|
2249
|
20.77
|
2422
|
21.23
|
Fujian
|
95
|
16.35
|
1009
|
9.32
|
1104
|
9.68
|
Zhejiang
|
40
|
6.88
|
952
|
8.79
|
992
|
8.70
|
Sichuan
|
15
|
2.58
|
433
|
4.00
|
448
|
3.93
|
Hunan
|
28
|
4.82
|
412
|
3.81
|
440
|
3.86
|
Henan
|
4
|
0.69
|
257
|
2.37
|
261
|
2.29
|
Hubei
|
13
|
2.24
|
240
|
2.22
|
253
|
2.22
|
Chongqing
|
4
|
0.69
|
236
|
2.18
|
240
|
2.10
|
Jiangxi
|
9
|
1.55
|
192
|
1.77
|
201
|
1.76
|
Shandong
|
7
|
1.20
|
149
|
1.38
|
156
|
1.37
|
Guangxi
|
10
|
1.72
|
138
|
1.27
|
148
|
1.30
|
Hainan
|
6
|
1.04
|
76
|
0.71
|
82
|
0.71
|
Total
|
581
|
100.00
|
10826
|
100.00
|
11407
|
100.00
|
5.09%
|
94.91%
|
100.00%
|
3.2. Population distribution characteristics of overseas imported dengue fever cases in 13 provinces
3.2.1. Gender characteristics
As shown in Fig. 4, the number of males was higher than females for the imported cases from 2005 to 2019, with a male-to-female ratio = 1.97:1. The ratio in Yunnan province and Hainan province was relatively low, which was below the overall male-to-female ratio in 13 provinces, 1.28:1 and 1.83:1 respectively. The highest ratio was observed in Shandong Province, about 6:1.
3.2.2. Age characteristics
As shown in Fig. 5, the age distribution of the imported dengue fever cases in 13 provinces exhibited similar patterns. The young and middle-aged people aged 21–50 accounted for a large proportion (76.61%) of the imported cases. Specifically, imported cases in Yunnan province and Guangxi province were concentrated in the 21–30 age group. In Fujian Province, Henan Province, Chongqing City and Hainan Province the number of imported cases also was the largest in the 21–30 age group, and there were the second-largest imported cases in the 31–40 age group. The imported cases in other provinces were more concentrated in the 31–40 year-old age group, followed by the 21–30 year-old age group.
3.2.3. Occupational characteristics
Among the 13 provinces, the top ten occupations for the imported cases were farmers, businessman, housework or unemployed, workers, students, cadre, migrant laborer, scattered children, retiree, and catering industries, accounting for 85.59% of the total number of imported cases. Those above occupations showed different proportions slightly in different provinces. Farmers accounted for the highest proportion of imported cases in Yunnan, Hunan, Henan, Jiangxi and Shandong provinces. In contrast, the proportion of commercial services were highest in Guangdong, Zhejiang and Hainan provinces. Most imported cases in Fujian Province and Chongqing City were domestic and unemployed people. In Sichuan Province, Hubei Province, and Guangxi Province, in addition to farmers, the proportion of housework and unemployed among imported cases was also very high.(Fig. 6).
3.3. Temporal-spatial distribution characteristics of overseas imported dengue fever cases in 13 provinces
3.3.1. Global autocorrelation statistics
Using ArcGIS, we estimated the global Moran's I index as 0.110708 and the z-score as 3.595916 (P < 0.05) for all the imported cases from 2005 to 2012. The results suggested that the spatial autocorrelation of these cases during this period was significantly positive. Getis-Ord General G value was estimated as 0.000001 and z score as 2.526695(P < 0.05), suggesting a tendency for clustering of the overseas imported dengue fever cases from 2005 to 2012.
From 2013 to 2019, the global Moran's I index was 0.083838 and the z-score was 3.642993(P < 0.05). The results showed that the spatial autocorrelation was significantly positive for the imported cases during this period. Getis-Ord General G value was 0.000001 and z score was 1.756175(P < 0.1), suggesting a tendency for clustering of dengue fever cases imported from 2013 to 2019.
3.3.2. Local autocorrelation analysis
Anselin Local Moran's I analysis of the imported cases from 2005 to 2012 revealed high-high clusters at the border area of of Yunnan province, southeast of Guangdong province, and two coastal areas in Fujian province. In contrast, the low-low clusters were detected at most areas of Henan province and Shandong province, as well as one area or two areas in Guangxi province, Jiangxi province, Sichuan province, and Hubei province respectively. Sichuan province and Shandong province had one region exhibited high-low clusters respectively. Two border areas in Yunnan province, and the other 6 cities in the northern part of Guangdong Province were identified as low-high clusters (Fig. 7. a).
From 2013 to 2019, the number of provinces with local spatial autocorrelation areas decreased. four border areas in Yunnan province, and a region in the southeast of Guangdong province were high-high clusters. The central part of Guangxi Province, the central part of Hubei Province, Henan Province and Shandong Province were mostly low-low clusters. The capital of Hubei province was a high-low cluster. Two border areas in Yunnan Province and a city in eastern Guangdong province were low-high clusters(Fig. 7.b).
Getis-Ord Gi* analysis showed that imported cases from 2005 to 2012 had formed hot spots in some areas in Yunnan, Guangdong, and Fujian provinces including the border areas of Yunnan province, most parts of Guangdong province and the coastal areas of Fujian province(Fig. 7.c). However, the hot spots of the imported cases during 2013–2019 were only identified in Yunnan province(Fig. 7.d).
3.3.3. Temporal-spatial distribution characteristics of overseas dengue fever cases
We used the maximum spatial cluster size of 50% of the at-risk population to scan the clustering areas in the two time periods of 2005–2012 and 2013–2019. The results showed that the Dehong Dai and Jingpo Autonomous Prefecture in Yunnan Province was the most likely cluster for both of the two time periods(P < 0.01). The regional scope of the secondary cluster was significantly larger from 2013 to 2019. During 2005–2012, the secondary clusters were identified in Hainan Province, eastern Yunnan Province, Guangxi Province, Guangdong Province, central and southern Hunan Province, southern Jiangxi Province, and southern Fujian Province (P < 0.01). During 2013–2019, the secondary cluster expanded westward and northward, reaching the areas around Chongqing (P < 0.01) (Fig. 8, Table 2).
Table 2. Temporal-spatial scan results of overseas imported dengue fever, 2005-2019.
period
|
Clusters
|
Aggregation time
|
Observed cases
|
Expected cases
|
Relative risk
|
LLR
|
P-value
|
2005-2012
|
I
|
2008/8/1-2008/11/30
|
46
|
0.038
|
1321.43
|
282.688853
|
<0.01
|
II
|
2010/7/1-2012/12/31
|
223
|
66.37
|
4.83
|
140.393001
|
<0.01
|
2013-2019
|
I
|
2014/8/1-2017/12/31
|
1986
|
8.62
|
282.05
|
9019.769506
|
<0.01
|
II
|
2019/5/1-2019/11/30
|
3266
|
463.81
|
9.65
|
3991.103028
|
<0.01
|