Statistical analysis of indigenous and imported cases of dengue fever in Kaohsiung City between 2013 and September 2018
Statistics for the total number of imported cases of dengue fever in Kaohsiung City
A total of 239 imported cases of dengue (32, 44, 61, 37, 34, and 31 cases in 2013, 2014, 2015, 2016, 2017, and 2018, respectively) were identified in Kaohsiung City during the examination period (Table 1), including 80 women (33.5%) and 159 men (66.5%). The majority of cases involved individuals arriving from Southeast Asian countries: 48 from the Philippines (20.0%), 47 from Indonesia (19.6%), 33 from Vietnam (13.8%), 30 from Malaysia (12.6%), 24 from Thailand (10.0%), and 13 from Singapore (5.4%). Additional cases were also imported from the Indian subcontinent (i.e., India, Bangladesh, Sri Lanka, and the Maldives) and the South Pacific region (Tuvalu). Supplementary Figure 1 shows the top and bottom 5 countries from which dengue was imported, and the distance from these countries to Kaohsiung.
Further analyses of the 239 imported cases of dengue revealed that 142 of the infected individuals (59.4%) were Taiwanese nationals who had traveled abroad. However, foreign workers and tourists entering the country accounted for a further 15.5% and 8.8% of cases, respectively (Table 2). Analysis of patient occupation is presented in Table 3, with home management, business, and studying being the most commonly cited occupations (together accounting for 46.4% of cases). Figure 1 shows fluctuations in the monthly reporting rates of imported cases of dengue in Kaohsiung City (Figure 1).
Predominant serotypes of annual indigenous cases of dengue fever in Kaohsiung City
Between 2013 and September 2018, there were 35,148 indigenous cases of dengue fever in Kaohsiung City, with 70, 14,999, 19,723, 342, 3, and 11 cases reported in 2013, 2014, 2015, 2016, 2017, and 2018 (up to the end of September), respectively. Primary cases of dengue fever were concentrated in the 24 consecutive months between May 2014 and April 2016 (Figure 2). During this period, there were 35,062 indigenous cases of dengue fever in Kaohsiung City (99.8%). The major epidemic serotype of dengue fever in 2014 was serotype 1, while that from 2015 to April 2016 was serotype 2.
Concealment of cases for more than 3 days
Further analysis of the 239 confirmed imported cases revealed that 37 infected individuals did not stay within local communities after contracting dengue. This included 5/32 cases that occurred in 2013 (15.6%); 3/44 cases in 2014 (6.8%); 1/61 cases in 2015 (1.7%); 10/37 cases in 2016 (27.0%); 12/34 cases in 2017 (35.3%); and 6/31 cases in 2018 (19.4%) (Table 4).
We also identified a significant number of cases whose concealment period was >3 days (139/239, 58.16%). Of these, we found 24 such cases in 2013 (75.0% of cases that year), 32 in 2014 (72.7%), 39 in 2015 (63.9%), 22 in 2016 (59.5%), 14 in 2017 (41.2%), and 8 in 2018 (19.4%) (Figure 3).
Statistical analysis of the number of indigenous and imported cases of dengue fever in Kaohsiung City between 2013 and September 2018
The 35,148 confirmed cases of indigenous dengue fever reported in Kaohsiung City between 2013 and September of 2018, along with the 239 imported cases, were sorted according to the month of identification (n = 69). A significant positive correlation (paired sample correlation coefficient, r = 0.407, p = 0.001) was found between the numbers of imported and indigenous cases reported each month (Table 5).
However, actual clinical surveillance reports were delayed by approximately 1–2 months during local epidemics of dengue fever; furthermore, a concealment period ≥2 days was common. Therefore, instead of investigating correlations between reported cases in the same month, we tested for relationships between the number of imported cases in a single month and the number of indigenous cases reported 2 months later, also accounting for differences in the concealment period. When accounting for concealment periods >2 days, we found significant positive correlations between the number of imported cases and the number of indigenous cases reported 2 months later (Pearson’s correlation coefficient, r = 0.459, p = 0.0001), >3 days (r = 0.394, p = 0.001), and >4 days (r = 0.387, p = 0.001) (Table 6).