Simulation models of dengue transmission in Funchal, Madeira Island: Influence of seasonality

Abstract

The recent emergence and established presence of Aedes aegypti in the Autonomous Region of Madeira, Portugal, was responsible for the first autochthonous outbreak of dengue in Europe. The island has not reported any dengue cases since the outbreak in 2012. However, there is a high risk that an introduction of the virus would result in another autochthonous outbreak given the presence of the vector and permissive environmental conditions. Understanding the dynamics of a potential epidemic is critical for targeted local control strategies. Here, we adopt a deterministic model for the transmission of dengue in Aedes aegypti mosquitoes. The model integrates empirical and mechanistic parameters for virus transmission, under seasonally varying temperatures for Funchal, Madeira Island. We examine the epidemic dynamics as triggered by the arrival date of an infectious individual; the influence of seasonal temperature mean and variation on the epidemic dynamics; and performed a sensitivity analysis on the following quantities of interest: the epidemic peak size, time to peak, and the final epidemic size. Our results demonstrate the potential for summer and autumn season transmission of dengue, with the arrival date significantly affecting the distribution of the timing and peak size of the epidemic. Late-summer arrivals were more likely to produce large epidemics within a short peak time. Epidemics within this favorable period had an average of 11% of the susceptible population infected at the peak, at an average peak time of 95 days. We also demonstrated that seasonal temperature variation dramatically affects the epidemic dynamics, with warmer starting temperatures producing large epidemics with a short peak time and vice versa. Overall, our quantities of interest were most sensitive to variance in the date of arrival, seasonal temperature, transmission rates, mortality rate, and the mosquito population; the magnitude of sensitivity differs across quantities. Our model could serve as a useful guide in the development of effective local control and mitigation strategies for dengue fever in Madeira Island.

A Systematic Review of Mathematical Models of Dengue Transmission and Vector Control: 2010–2020

“…Only three papers address environmental controls ( Figure 3 ), and two of the three are from the last two years of publications in this study, suggesting this may be a growing area of interest [ 4 ]. Most (28 out of 32) of the overall articles selected are deterministic modelling studies of dengue transmission [ 32 , 34 , 35 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 46 , 48 , 49 , 50 , 51 , 52 , 53 , 54 , 56 , 57 , 58 , 59 , 61 , 62 , 63 , 64 , 65 ]. Three of the thirty-two articles are stochastic modelling studies [ 33 , 47 , 60 ].…”

“…Ten (10) of the 32 articles investigated the chemical method for vector control [ 33 , 37 , 46 , 47 , 48 , 49 , 54 , 56 , 61 , 62 , 63 ]; nineteen (19) articles described the biological method of control [ 32 , 34 , 35 , 38 , 40 , 41 , 42 , 43 , 44 , 50 , 51 , 52 , 53 , 55 , 57 , 58 , 64 , 65 ]; while only three (3) of the articles partially discussed the environmental method for vector control [ 39 , 59 , 60 ]. More than half (60%) of the selected modelling studies were categorised under biological control methods.…”

“…Almost half (13 out of 32) of the selected studies performed sensitivity analysis [ 33 , 35 , 37 , 38 , 42 , 43 , 44 , 53 , 54 , 57 , 59 , 61 , 65 ]. Four of the studies [ 35 , 37 , 59 , 61 ] used the Latin hypercube sampling (LHS) and partial rank correlation coefficients (PRCC) and provided evidence that the most effective parameters in curtailing dengue viral spread include reducing the transmission probability per contact of either infectious mosquitoes with susceptible humans or susceptible mosquitoes with infectious humans; mosquito recruitment and death rates; and human recovery rates [ 37 , 61 ]. Other highly sensitive parameters include the epidemic size, date of arrival of the infectious human and mean annual temperature [ 59 ].…”

“…Four of the studies [ 35 , 37 , 59 , 61 ] used the Latin hypercube sampling (LHS) and partial rank correlation coefficients (PRCC) and provided evidence that the most effective parameters in curtailing dengue viral spread include reducing the transmission probability per contact of either infectious mosquitoes with susceptible humans or susceptible mosquitoes with infectious humans; mosquito recruitment and death rates; and human recovery rates [ 37 , 61 ]. Other highly sensitive parameters include the epidemic size, date of arrival of the infectious human and mean annual temperature [ 59 ]. One study used probabilistic sensitivity and threshold analyses and showed that if the cost of adult mosquito control was greater than 16 times that of larval control, all the adult vector control strategies were dominated [ 54 ].…”

“…The introduction of mathematical models to understand viral infection dynamics has long been helpful in the area of disease control [ 30 , 31 , 32 , 33 , 34 , 35 , 36 ]. Several models involving vector control of the transmission of different dengue serotypes have been formulated and analysed [ 30 , 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 , 61 , 62 , 63 , 64 , 65 ]. Some studies [ 35 , 36 , 38 , 41 , 42 , 44 , 52 ] investigated dengue transmission models capturing the different dengue serotypes together with their vectors.…”