Spatio-temporal clusters and patterns of spread of dengue, chikungunya, and Zika in Colombia.

BACKGROUND: Colombia has one of the highest burdens of arboviruses in South America. The country was in a state of hyperendemicity between 2014 and 2016, with co-circulation of several Aedes-borne viruses, including a syndemic of dengue, chikungunya, and Zika in 2015. METHODOLOGY/PRINCIPAL
FINDINGS: We analyzed the cases of dengue, chikungunya, and Zika notified in Colombia from January 2014 to December 2018 by municipality and week. The trajectory and velocity of spread was studied using trend surface analysis, and spatio-temporal high-risk clusters for each disease in separate and for the three diseases simultaneously (multivariate) were identified using Kulldorff's scan statistics. During the study period, there were 366,628, 77,345 and 74,793 cases of dengue, chikungunya, and Zika, respectively, in Colombia. The spread patterns for chikungunya and Zika were similar, although Zika's spread was accelerated. Both chikungunya and Zika mainly spread from the regions on the Atlantic coast and the south-west to the rest of the country. We identified 21, 16, and 13 spatio-temporal clusters of dengue, chikungunya and Zika, respectively, and, from the multivariate analysis, 20 spatio-temporal clusters, among which 7 were simultaneous for the three diseases. For all disease-specific analyses and the multivariate analysis, the most-likely cluster was identified in the south-western region of Colombia, including the Valle del Cauca department.
CONCLUSIONS/SIGNIFICANCE: The results further our understanding of emerging Aedes-borne diseases in Colombia by providing useful evidence on their potential site of entry and spread trajectory within the country, and identifying spatio-temporal disease-specific and multivariate high-risk clusters of dengue, chikungunya, and Zika, information that can be used to target interventions.

Introduction

Mosquito-borne diseases are thriving and expanding globally despite decades of large-scale control efforts [1]. Aedes mosquitoes are responsible for transmitting arboviral diseases such as dengue, chikungunya, and Zika. Aedes aegypti and Aedes albopictus are opportunist mosquitoes adapted to urban environments for which poor quality housing and sanitation management are key determinants for the sustained propagation of arboviral diseases [2,3]. Dengue virus (DENV) is endemic in more than 100 countries with estimates ranging from 105 to 390 million infections each year [4,5]. Similarly, chikungunya virus (CHIKV) has been identified in 112 different countries, has become endemic in several countries, and is prone to explosive epidemics in areas with no prior immunity [6,7]. As with chikungunya, epidemics of dengue also occur, usually coinciding with increased Aedes’ mosquito abundance (after a rainy season) in populations with no immunity to one of the four dengue virus types [8]. The introduction of Zika into the Americas likely occurred in late 2014, on the heels of chikungunya emergence in 2013 [9,10]. Local transmission of Zika was confirmed in 86 countries, and although Zika’s Public Health Emergency of International Concern (PHEIC) is over, sporadic new cases continue to be detected, indicating the potential for Zika to become endemic in previously Zika-naïve, Aedes-endemic countries [11-13].

Historically, Colombia has been significantly burdened with dengue, chikungunya, and Zika infections, with the Aedes aegypti mosquito being widely distributed throughout the country at elevations below 2,000 meters [14,15]. The severity of outcomes of arboviral diseases range from asymptomatic infections, mild febrile illnesses, to severe infections that can be fatal and or produce chronic sequelae, including persistent fatigue and myalgia, debilitating joint pain, and Guillain-Barré syndrome [10,16-18]. Zika virus (ZIKV) exposure in fetuses has been causally related to microcephaly in neonates and other congenital malformations [19]. A significant portion of infected individuals remain asymptomatic and lifelong immunity can be developed for each one of the four dengue serotypes and for CHIKV and ZIKV [20-22]. There is no current antiviral treatment for arboviruses and until effective vaccines become broadly commercially available, vector control through environmental (e.g., habitat removal), chemical (e.g., larvicide), biological (e.g., larvae eating fish), or educational campaigns remains the primary prevention strategy in most endemic settings [2,3]. Epidemiological surveillance is central to the successful control and prevention of arboviral diseases, as it provides data to monitor trends and outbreaks and identifies areas for targeted risk communication, travel advisories, and entomological response.

It is reasonable to anticipate that dengue, chikungunya, and Zika epidemiology is temporally and spatially related, given that they are transmitted by the same Aedes species, can co-circulate within the same region, and the presence of symptomatic infections for one virus may depend on previous or concurrent infection with one of the other viruses [23-28]. Previous work examined the spatio-temporal dependencies from 2015 to 2016 between dengue and Zika for one Colombian city and department [26]. Two other studies applied scan statistics to identify space-time clusters of arboviral diseases in Colombia, one only for dengue and chikungunya [29], and the other considered only disease-specific clusters [30]. Therefore, there have not been cluster analyses examining the co-occurrence of all three arboviruses and describing the patterns of introduction including the speed and direction of the spread of chikungunya and Zika in Colombia. We address this knowledge gap, providing meaningful insight into the shared and unique patterns of spatio-temporal disease risk in Colombia. Our study aimed to identify at-risk municipalities and time periods for dengue, chikungunya, and Zika in Colombia through disease-specific and multivariate cluster analyses and through estimating the direction and speed of chikungunya and Zika introduction in Colombia.

Methods

This is an ecological study in which notified cases of dengue, chikungunya, and Zika from January 2014 to December 2018 for the entire country were extracted from the national surveillance system. Our units of analysis were municipality for space and week for time. Two main different methods were applied: front wave velocity analysis and scan statistics.

Ethics statement

This study was approved by the Science and Health Research Ethics Committee (Comité d’éthique de la recherche en sciences et en santé—CERSES) of the University of Montreal, approval number CERSES-19-018-D. Informed consent was not required as only secondary data was used and the data were analyzed anonymously.

Study context and data

Colombia is located at the northern tip of South America with nearly 50.4 million inhabitants. The Andes mountains dominate its topography and there are 11 distinctive geographic and climatic conditions within the country. Colombia comprises 1,123 municipalities and 33 administrative states called departments that are economically diverse and range from urban to rural to natural landscapes. Dengue became a notifiable disease in Colombia in 1978 whereas chikungunya was first detected in Colombia in July 2014 and local transmission of Zika was confirmed in the country in October 2015 [31-34].

In Colombia, surveillance is the responsibility of each department’s Secretariat of Health although there is a national surveillance program administered by the National Institute of Health of Colombia that receives weekly reports from all health facilities that provide services to probable and confirmed cases of dengue, chikungunya, and Zika. The electronic platform, SIVIGILA (Sistema Nacional de Vigilancia en Salud Pública, http://portalsivigila.ins.gov.co/), provides publicly available aggregate data on chikungunya, dengue, and Zika cases at the municipal and departmental levels [35]. The aggregate data include probable and confirmed cases. Probable cases are those that present clinical signs within the case definition of each disease, following the official protocols of the National Institute of Health of Colombia [31-33]. In Colombia, probable cases can be confirmed either by laboratory (based upon a positive result from antigen, antibody, or virus detection and/or isolation) or by clinical and epidemiological criteria. Complete case definitions for probable and confirmed cases for each disease are available in the S1 Appendix. Population data was obtained from the National Administrative Department of Statistics of Colombia (DANE, https://www.dane.gov.co/).

Front wave velocity

A trend surface analysis was performed to estimate the front wave velocity for chikungunya and Zika, the two Aedes-diseases that emerged in Colombia during the study period. Dengue was not considered for this analysis as the disease was already endemic in the country. Trend surface analysis has been used to examine diffusion processes in two dimensions: time and space, using polynomial regression. A continuous surface is estimated with the order of the model capturing the general direction and speed of the emerging or front wave of an infectious disease [36-38].

For this analysis, we separately identified the first notification of a chikungunya and Zika case for each municipality and then used the centroids of the municipalities, calculated in meters using QGIS software [39]. The response variable was time in weeks from the first chikungunya or Zika case notified for each centroid (X and Y coordinates) in a given municipality. The continuous surface of time to notification was estimated by regressing it against the X and Y coordinates in meters. Parameters were estimated using least squares regression, and if a simple 2-D plane through the points is insufficient to model the data, high-order polynomials are often used to capture local scale trends [40]. The final models (one for each disease) were selected based on the Akaike Information Criterion (AIC), Bayesian information criterion (BIC) and expert judgment (S2 Appendix).

The rate of change was obtained by taking the partial derivatives with respect to X and Y, for the final linear models, shown below as order five polynomial (1).

Eqs (2) and (3) provide expressions for a slope vector at a given location (X,Y). The vectors can be converted to express the magnitude and direction of rate of change in km per week by finding the inner product of the vector, where magnitude ||xy|| = √(x + y) and the direction θ = tan(y/x). The rate we were primarily interested in was velocity (in km per week), which was obtained by inverting the final magnitude of the slope.

Scan statistics

To identify spatio-temporal clusters we used Kulldorff’s scan statistics [41]. We used a discrete Poisson model approach including the total number of cases per municipality offset by the population at-risk (municipal population). Cluster detection was performed to identify statistically significant space-time high-risk clusters of each disease separately and then performed for the three diseases simultaneously, using the multivariate scan analysis.

Kulldorff’s scan statistics determine the presence of clusters using a cylinder that moves across space and time under predefined spatial and temporal scanning windows. The clusters are identified by observing a higher risk within the cylinder compared to the risk outside of the cluster. The relative risk (RR) is calculated as follows: where c and C are the number of observed cases in the cylinder and the total number of observed cases in the study area, respectively, and E[c] is the expected number of cases inside the cylinder, calculated as: being P the total population in the study area and p the population within the cluster [41].

Clusters were ordered based on the likelihood ratio; clusters with the higher maximum likelihoods were more-likely, i.e. with stronger evidence of fitting the definition of a cluster. The likelihood function for the Poisson model is proportional to: where I(c>E[c]) is set to 1 when there are more cases than expected, and 0 otherwise. For the multivariate scan analysis, the likelihood ratio of the cluster is the sum of the likelihood ratio for each disease calculated using Eq 6 [41,42].

Each space-time cluster had its own start and end dates, which were used to calculate the duration in weeks of the cluster, the number of accumulated cases observed in this period, the population within the cluster, and its relative risk.

The spatial scanning window was based on the centroid of each municipality, with the maximum size set to 150 km of radius and 20% of the total population at risk. Colombia’s population is highly concentrated in a few municipalities, with only 5 municipalities accounting for nearly 30% of the country’s population. Considering only the maximum population at-risk would result in clusters with very different sizes, and therefore, we also considered the maximum radius of the cylinder. To define the values, we experimented with different combinations of maximum sizes for the radius and for the population at-risk (S3 Appendix). We chose the combination that resulted in clusters that were balanced in terms of number and size: not so large as to include very distinct areas and/or low-risk municipalities, nor too small as to be too numerous and include only one municipality. The temporal scanning window was set from 2 up to 26 weeks. Clusters were restricted to having at least 100 cumulative cases over the temporal scanning window. For each model, 999 Monte Carlo simulations were performed to assess the statistical significance of observed clusters and only the clusters with p-value less than 0.05 were reported.

All analyses were performed using R (v. 4.1.2) [43] and we applied SaTScan (v. 9.6, https://www.satscan.org/) [44] using the package rsatscan (v. 0.3.9200) [45]. The front wave analysis was performed with codes based on the package outbreakvelocity (v. 0.1) [46]. The R codes and the data used in the analyses are available at https://github.com/laispfreitas/Colombia_DZC_satscan [47]. Maps were depicted using ggplot2 (v. 3.3.5) [48] and colorspace (v. 2.0) [49] packages in R, or QGIS (v. 3.22.3) [39].

Results

During the study period, from January 2014 to December 2018, there were 366,628, 77,345 and 74,793 cases of dengue, chikungunya, and Zika, respectively, captured by the national surveillance system. Dengue cases occurred throughout this period and peaked in 2016. Chikungunya notifications rapidly increased in September 2014 and lasted for approximately a year followed by a smaller wave of cases in December 2015 to August 2016 (Fig 1). Zika notifications began increasing in October 2015, with a peak number of cases in February 2016, coinciding with a peak of dengue cases and the second wave of chikungunya. The Zika epidemic in Colombia lasted for approximately a year, following which sporadic cases were continually detected along with sporadic chikungunya cases. In 2017, there was a sharp reduction in the number of cases of the three diseases, by 84.1% compared to the previous year. Cases of dengue increased again in 2018, but chikungunya and Zika counts remained low.

Fig 1

Number of notified dengue, chikungunya, and Zika cases by week of first symptoms, Colombia, 2014–2018.

The first cases of chikungunya observed in the surveillance data, at epidemiological week 23/2014, were from the municipalities of Tuluá and Barranquilla, with Tuluá being a smaller city inland in the Valle del Cauca department and over 700 kilometers away from Barranquilla, the capital city of Atlántico department, on the Atlantic northern coast. After the initial notifications, chikungunya was detected in 774 other municipalities during the first wave of cases, which lasted 14 months, following a southern pattern of dispersal from Barranquilla and a radiating pattern of dispersal from Tuluá (Fig 2A). Since the first notification of chikungunya until December 2018, it was identified in 875 different municipalities in Colombia with an average speed of 27 km/week, ranging from 1 to 397 km/week.

Fig 2

Chikungunya (A) and Zika (B) spread across Colombia, 2014–2018. The angle of the arrowhead represents the direction of spread. Yellow arrowheads represent the first cases observed in the data. Base layer source: GADM ( https://www.diva-gis.org/gdata.

The first cases of Zika observed from the surveillance data, at epidemiological week 32/2015, were from four municipalities from three different departments: San Andrés (from the islands of San Adrés and Providencia, which are off the Atlantic coast of Nicaragua), Cali (capital city of Valle del Cauca department), Cúcuta and El Zulia (both from the Norte de Santander department, in the northern border with Venezuela). During the period of the Zika epidemic, it was detected in 747 municipalities following a southern pattern of dispersal from the northern Atlantic coast, a eastern pattern towards the border of Venezuela and also a northern pattern in the western part of the country (Fig 2B). Since its initial introduction and until December 2018, it was identified in 768 different municipalities with an average speed of 79 km/week, ranging from 2 to 747 km/week.

We identified 21, 16 and 13 spatio-temporal clusters of dengue, chikungunya and Zika, respectively (Tables 1 and S1–S3). When we consider the median values for each disease, Zika case clusters included more municipalities, were shorter in duration, and had higher relative risks compared to chikungunya and dengue.

Table 1

Summary of the characteristics of the clusters for dengue, chikungunya and Zika cases, Colombia, 2014–2018.

	Dengue	Chikungunya	Zika
N° of clusters	21	16	13
	Median (IQR)
N° of municipalities	5 (2–34)	23 (9–61)	25 (4–106)
Duration in weeks	25 (21–26)	19.5 (16–24.25)	14 (11–21)
Relative risk	5.80 (3.34–8.75)	12.73 (7.68–21.53)	21.43 (7.89–30.81)
Population inside the cluster	203,979 (68,778–1,234,977)	975,063 (117,505–2,018,560)	940,273 (140,082–3,375,936)
N° of observed cases	639 (204–2976)	1207 (322–1962)	933 (348–4663)

IQR = Interquartile Range

The spatial distribution of the detected clusters was similar for all diseases (Fig 3A–3C). The most likely clusters for each disease were identified in the same region, in the south-west, including the Valle del Cauca department and its capital Cali. The most likely clusters were also the largest clusters in terms of population (S1 Fig). In 2014, dengue and chikungunya clusters were detected, while in 2015 and 2016, clusters of the three diseases were detected (Figs 3D–3F and 4). In 2017, no clusters were detected and in 2018, only dengue clusters were identified (Fig 4). The first chikungunya cluster was detected in northern Colombia, on the Atlantic coast, while the Zika clusters were first identified in the islands of San Andrés and Providencia, followed by a cluster on the Atlantic coast (S2 Fig). Generally, dengue clusters lasted longer than chikungunya and Zika (Figs 3G–3I and 4) and had smaller relative risks (Fig 3J–3L). Among the clusters, higher relative risks were observed for chikungunya and Zika, particularly in the south-western part (Pacific region) of Colombia (Fig 3J–3L).

Fig 3

Space-time clusters ranked by likelihood ratio* (A-C), year of start date (D-F), duration in weeks (G-H) and relative risk (J-L) for dengue (1 column), chikungunya (2 column) and Zika (3 column), Colombia, 2014–2018. * The first cluster is the most likely cluster, i.e., with the maximum likelihood ratio. Base layer source: GADM (https://gadm.org/), available at https://www.diva-gis.org/gdata.

Fig 4

Temporal distribution of cases inside each cluster of dengue (A), chikungunya (B) and Zika (C), Colombia, 2014–2018. Orange bands represent the time period at which the cluster was detected. Clusters are ranked by likelihood ratio, being the first cluster the most likely cluster, i.e., with the maximum likelihood ratio.

From the multivariate analysis, 20 spatio-temporal clusters were identified (S4 Table). Among those, 7 were significant for the three diseases, 5 for dengue and chikungunya, 2 for dengue and Zika, 4 clusters for dengue only and 2 for chikungunya only. The median number of municipalities forming a cluster was 14.50 (Interquartile range, IQR, 2.00–46.25). Clusters’ median duration was of 16.50 weeks (IQR 14.50–25.25) and median population was of 305,307 inhabitants (IQR 5,127–8,580,330). The most likely cluster for the multivariate scan analysis was significant for dengue, chikungunya, and Zika (Fig 5A and 5B), and was also detected in the south-west of Colombia, in the Valle del Cauca department. All clusters significant for all three diseases were detected in 2015 (Fig 5C). In general, these clusters lasted longer than other clusters (Fig 5D) and presented higher relative risks for Zika compared to dengue and chikungunya (Fig 5E–5G).

Fig 5

Multivariate space-time clusters of dengue, chikungunya and Zika ranked by likelihood ratio* (A), diseases (B), year of start date (C), duration in weeks (D) and relative risk for each disease (E-G), Colombia, 2014–2018. * The first cluster is the most likely cluster, i.e., with the maximum likelihood ratio. Base layer source: GADM (https://gadm.org/), available at https://www.diva-gis.org/gdata.

Discussion

Colombia was in a state of hyperendemicity between 2014 and 2016, with co-circulation of several arboviruses, including a syndemic of dengue, chikungunya, and Zika in 2015. Under this scenario, we explored the introduction patterns of chikungunya and Zika, and identified disease-specific and multivariate space-time clusters. This paper presents novel results on the co-occurrence of all three diseases in Colombia using multivariate cluster analysis. Our results indicate that the northern Atlantic coast was likely the place of emergence of new Aedes-borne diseases in Colombia, while the south-western region concentrated higher disease burden.

Overall, the geographical pattern of spread within Colombia was very similar for both chikungunya and Zika, with general southern dispersion from the Atlantic coast, although Zika spread more quickly. Our trend surface analysis and cluster evaluation identified the northern Atlantic coast, Caribbean region, for early clusters of both chikungunya and Zika, and therefore a potential portal of entry of arboviruses in the country. This region has important cargo ports on the Atlantic coast through which a good proportion of the country’s imports enter. Also, there are well-known tourist sites in the region, such as Cartagena, Santa Marta and San Andrés. It is also close to the Venezuelan border, where there is an important and uncontrolled flow of people given the country’s serious economic crisis. The environmental and socioeconomic conditions in this border region facilitate the transmission of infectious diseases including vector-borne diseases [50-52]. If a new pathogen was introduced in the Colombian Atlantic coastline region, its transmission and dispersion would likely be facilitated given these conditions [34]. Zika’s accelerated spread is consistent with findings from another study using gravity models. The authors estimated that the spread of Zika was twice as fast as that of chikungunya, taking on average 15.7 weeks to invade half of the affected cities compared to 34.4 weeks for chikungunya [53]. We also observed that, compared to chikungunya, Zika’s clusters were shorter and had higher relative risks, indicating a more explosive epidemic. This is possibly a result of an increased demand of people seeking medical care as a consequence of being more concerned about Zika due to the association of the virus with congenital malformations and of neurological complications. Another possible explanation is of Zika virus being more transmissible than chikungunya, supported by evidence that Ae. aegypti is more efficient for transmitting the Zika virus than the chikungunya virus, even if co-infected [54,55].

For the disease-specific analyses and the multivariate analysis, the most likely cluster, i.e. the one with the strongest evidence of fitting the definition of a cluster, was consistently identified in the south-western region of Colombia, including the Valle del Cauca department and its capital Cali. This region is known for the presence of Aedes-borne diseases, as it has been described in other studies using the SIVIGILA data [29,30,34,56,57]. Between 2013 and 2016, Valle del Cauca was the department most affected by arboviral diseases in Colombia, accounting for 24.2% of the total disability-adjusted life years (DALYs) of the country and its capital city, Cali, contributed 13% of all arbovirus cases reported during the study period [56]. This region presents a favorable environment and suitable climate for the Aedes mosquitoes, with altitude less than 2000 meters above sea level, warm temperatures above 20°C, relative humidity below 80%, and rainy seasons that are essential to the life cycle of the mosquitoes [34,58,59]. Cali, where most of the cases in the region concentrate, has a warm and dry climate, with two rainy seasons (from March to May and from September to December). The average annual total precipitation in Cali is 1483 mm, the average maximum temperature is around 30°C, and the relative humidity of the air is on average between 70 and 76% [60]. Other factors such as environmental, cultural (e.g. water storage inside the houses) and social and material deprivation, in addition to a high population mobility due to trade with bordering countries and other cities have also favored the introduction and dissemination of arboviruses in the south-western region of Colombia [34,61-63].

Dengue, chikungunya, and Zika simultaneous clusters were identified around the cities of Cali, Ibagué and Neiva, and in western Colombia near the border with Ecuador. Cali, Ibagué and Neiva were part of the most likely clusters for all three diseases separately and simultaneously, and accounted for almost 20% of all cases of dengue, chikungunya, and Zika in the country during the study period. Cali alone contributed 13% of all cases. The western region near the border with Ecuador was previously classified as high-risk for dengue [57]. This region comprises an area with a large proportion of rural population, with limited public, educational, and health services. There is also a very active armed conflict in the region, generated by drug trafficking, further exacerbating health vulnerabilities. Therefore, the important presence of arboviruses observed in this region can be explained by its social and health inequities, in addition to a favorable climate for the presence of the Aedes mosquito [34].

Dengue and Zika clusters were mostly observed between the September of 2015 and the September of 2016. It is possible that during this period the transmission of Aedes-borne diseases was being influenced by the El Niño Southern Oscillation (ENSO). The El Niño is a climate cycle that occurs in the equatorial Pacific Ocean affecting weather patterns, and has been identified as one of the factors increasing dengue transmission in Colombia [64]. No clusters were identified in 2017, when a sharp decrease in the number of arboviral diseases reported compared to previous years was observed. This decrease is likely due to immunity development (i.e. decrease in the number of susceptible individuals to the circulating viruses), which is typical after large outbreaks [65]. In fact, we observed very small counts of chikungunya and Zika in 2017 and 2018, while dengue cases began increasing in 2018. Dengue has four circulating serotypes, and is expected to generate cyclic outbreaks every two to three years [66]. After our study period, Colombia was affected by another dengue epidemic in 2019, with numbers of cases exceeding those of 2016 [67].

The main limitations of this study are related to the quality and timeliness of the surveillance data. We used official case counts that are from a passive surveillance system, meaning our study population included only patients who sought health care. Underreporting is a known limitation when working with surveillance data and is also an important challenge with Colombia’s surveillance system [68,69]. During a syndemic of arboviral diseases that share similar symptoms such as dengue, chikungunya, and Zika, misclassification likely occurred as only a small proportion of cases are laboratory confirmed (27.2%, 4.4% and 2.7% of cases of dengue, chikungunya, and Zika, respectively, in Colombia in 2017), although differential diagnosis algorithms are used [69-72]. Earlier introductions of chikungunya may not have been captured by the surveillance system. In the case of Zika, its introduction was expected after the Brazilian epidemic, however, given the mild and generic nature of symptoms and the high proportion of asymptomatic persons, some cases may not have been captured by the surveillance system, especially in non Aedes-endemic areas [20]. Sporadic geographically dispersed cases were recorded in various parts of Colombia, which increased the uncertainty associated with the front wave analysis. These cases, such as those in southeastern Colombia, increased uncertainty in direction and speed estimates, which are also related to edge effects. Edge effects occurred along the boundary of the study area, which in this study were constructed by using fewer data points and are therefore less stable. One limitation of scan statistics is that the method uses circular scans to detect clusters. This could result in low-risk municipalities being considered as part of a cluster if surrounded by high risk municipalities. We counterbalanced this by restricting the clusters’ size and verifying the relative risk of the municipalities forming clusters (S3 Fig). Finally, we also acknowledge that Aedes-borne diseases are complex systems, and that there are many other levels necessary to explain disease presence and distribution, including but not limited to environmental conditions, climate, connectivity of municipalities, human mobility, and socioeconomic and demographic characteristics.

In this study, we applied two methods to examine the emergence and spatio-temporal clusters of Aedes-borne diseases in Colombia. The detection of simultaneous clusters using the multivariate scan statistics analysis highlights the key hotspots areas for dengue, chikungunya, and Zika, and that should be prioritized for interventions to reduce the burden of these diseases. Additionally, these areas may also be at higher risk for other emerging Aedes-borne diseases. It is important that the surveillance in these locations is strengthened to be able to early detect the circulating viruses as well as unusual increase in the number of cases. Both methods are simple and provide helpful insight into the trajectory of arboviral diseases in Colombia region, which can be used to inform targeted interventions, such as enhanced surveillance activities and prevention activities. The methods have the potential to be applied for other emerging infectious diseases or variants of concern for COVID-19. It is worth mentioning that the Aedes aegypti mosquitoes are increasing their geographic range due to the rising global mean temperature, one of the consequences of climate change, which means regions historically not suitable for these mosquitoes are at risk of future outbreaks [73,74]. The methods described here could also be useful in these settings.

Population inside clusters of dengue, chikungunya and Zika, Colombia, 2014–2018.

Base layer source: GADM (https://gadm.org/), available at https://www.diva-gis.org/gdata.

(PNG)

Click here for additional data file.

Week of start date of dengue, chikungunya and Zika clusters, Colombia, 2014–2018.

Base layer source: GADM (https://gadm.org/), available at https://www.diva-gis.org/gdata.

(PNG)

Click here for additional data file.

Relative risk for dengue, chikungunya and Zika by municipality inside a cluster, Colombia, 2014–2018.

Base layer source: GADM (https://gadm.org/), available at https://www.diva-gis.org/gdata.

(PNG)

Click here for additional data file.

Case definitions for dengue, chikungunya and Zika.

(PDF)

Click here for additional data file.

Tested models for the front wave velocity analysis.

(PDF)

Click here for additional data file.

SaTScan parameters setting.

(PDF)

Click here for additional data file.

Space-time clusters of dengue cases, Colombia, 2014–2018.

(PDF)

Click here for additional data file.

Space-time clusters of chikungunya cases, Colombia, 2014–2018.

(PDF)

Click here for additional data file.

Space-time clusters of Zika cases, Colombia, 2015–2018.

(PDF)

Click here for additional data file.

Space-time clusters of dengue, chikungunya and Zika cases detected using multivariate scan statistics, Colombia, 2014–2018.

(PDF)

Click here for additional data file.

18 Apr 2022

Dear Dr Freitas,

Thank you very much for submitting your manuscript "Spatio-temporal clusters and patterns of spread of dengue, chikungunya, and Zika in Colombia" for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Alberto Novaes Ramos Jr

Associate Editor

PLOS Neglected Tropical Diseases

Victor S. Santos

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: Freitas et al. describe the epidemiology of the triple epidemic of Zika, chikungunya, and dengue in Colombia. To do this they identify at-risk regions using spatiotemporal scan statistics, and estimate the speed of spread using trend surface analysis. The methods are clearly described and are appropriate to address the study's objectives. The study population is clearly described (although I was a little unclear on the definition of a case - see below), and I have no concerns about ethical requirements as the study used publicly available data. The specific concerns relating to the methodology that I would like to be addressed are:

- L107-113: I find the definition of a case a little confusing, please could you re-write this section. In particular it is unclear what category a positive antibody test would be in, as it is included in both definitions.

- Related to this, please could you describe how you used probable vs confirmed cases and what the breakdown of these were (i.e. the proportion in each group).

- L128+: For the trend surface analysis, it would be useful to see the other models tested, perhaps as a supplementary table. In this table it would be useful to include the model tested, the BIC of each model, and the average and range of the estimated velocities. Including this would help the reader interpret these results by being able to compare across models.

- Please could you include a GitHub link to the code used to run the analyses.

Reviewer #2: In SatScan the authors performed a retrospective analysis of spatiotemporal clusters in Colombia. It would be interesting to see the spatial variations in the temporal trend clusters. The spatial variation in the temporal trends identifies the clusters with growth rates (or decreases) in and out of these clusters, and evaluates whether the difference between them is statistically significant.

Reviewer #3: The objectives of the study are clearly articulated and ordered in a logical manner. The study design is appropriate to address the stated objectives. The population is clearly described. I have minor concerns about the methods.

Reviewer #4: Objectives are clearly articulated, however, there is not a clear proposed hypothesis related to the transmission type of disease, seasonality, or geographical regions as initial point of the transmission. The authors can propose a hypothesis (null or alternative) for the temporal trends (seasonality or extreme events and outbreaks), or spatially related to socio-ecological drivers of the transmission (urban determinants of health, human mobility). See:

Filho, A.S.N., Murari, T.B., Ferreira, P. et al. A spatio-temporal analysis of dengue spread in a Brazilian dry climate region. Sci Rep 11, 11892 (2021). https://doi.org/10.1038/s41598-021-91306-z

Chen et al. 2019. Int J Environ Res Public Health. 2019 Jul; 16(14): 2486. Published online 2019 Jul 12. doi: 10.3390/ijerph16142486

Yes, the study design seems appropriate to explore the spatio-temporal of high-risk cluster of the diseases either separately or concurrent.

- The population of municipalities, as unit of analysis, is described as well the number of cases by epidemiological weeks. However, what is/are the hypotheses that want to be tested?

-Yes, municipalities jurisdiction allows a adequate analysis of the whole country.

However, the authors need to state clear the hypothesis.

- Statistical analysis: I am not an expert in spatial statistics, so I suggest a reviewer with this expertise. However, I wonder about confounding factors such as seasonality in the different regions of Colombia. Are different patterns of rainy and hot temperatures. Considering that dengue, zika, Chikungunya are climate sensitive diseases, something to explore is how the seasonality may be a factor (or not) on the patterns of disease transmission (week number).

-Are there concerns about ethical or regulatory requirements being met?

No, the study met all the requirements of ethics standards.

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: The analysis matches the analysis plan and the results are clearly and sufficiently presented. I only have two comments:

- Could you label individually the importation locations which you mentioned in the text, so that a reader can find them.

- Figs 3 and 4 captions, I assume you rank all clusters by their likelihood (not just the first)? Perhaps better to write clusters are ranked by likelihood than the first cluster is most likely if so. Or if not, it would be better to rank them this way I think.

Reviewer #2: The results and figures are clearly and completely presented.

Reviewer #3: The results follow naturally from the analysis plan and are presented clearly. The figures are best in the .tif format included and seem of sufficient quality.

Reviewer #4: Yes, the results are linked to the objective of exploring the spatio-temporal clusters for the three diseases in Colombia. The discussion on the geographical sites for high-risk clusters is interesting and supported by the data and references. Yet, is there any additional explanation about the cities connectivity, urbanization trends, human mobility on the period of analysis?

Line 293: Could be more specific about what are the suitable climate conditions for the diseases.

-Are the results clearly and completely presented?

The explanation of the temporal cluster analysis considering the peaks of the VBD outbreaks, would need additional discussion: potential relation with extreme years such as El Nino (2015-2016), seasonality (weeks of the year), and interannual variability (2018) compared to other studies on South America region. See:

- Rachel Lowe, Anna M Stewart-Ibarra, Desislava Petrova, Markel García-Díez, Mercy J Borbor-Cordova, Raúl Mejía, Mary Regato, Xavier Rodó. Climate services for health: predicting the evolution of the 2016 dengue season in Machala, Ecuador. The Lancet Planetary Health, Volume 1, Issue 4, July 2017, Pages e142-e151

- Petrova D et al 2020 The 2018–2019 weak El Niño: predicting the risk of a dengue outbreak in Machala, Ecuador Int. J. Climatol. 41 3813–23

The maps of figure 3 are not easy to understand, a better explanation for:

- The clusters are for all the period (2014-2018), this makes difficult to understand the dynamics of the disease’s transmission which is clearly presented in the time series.

- Specially the relative’s risk for dengue, zika and chikungunya, individual or in co-occurrence, maybe would be better represented for various points on time (years and weeks) to visualize the temporal dynamics of the disease’s transmission.

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: The conclusions are generally supported by the results, and the limitations are well described. I have a few comments:

- In the limitations you touch upon the effect of isolated cases, but I wonder if you could also discuss the effect of repeated introductions and how this might affect your trend surface analysis in particular.

- L268: as your model is not predictive but descriptive, please say "was likely the place of emergence" here, rather than "is likely the place of emergence".

- L279: could you briefly describe what this other study did and what they found (one sentence)

- L280: "quicker" -> "shorter"

- L266: Please remove the sentence were you say "To our knowledge, this is the first study..." as it is not informative and hard to verify.

Reviewer #2: The conclusions are supported by the data presented.

Reviewer #3: The conclusions are mostly supported by the data presented and the public health implications are discussed. I think more could be done around the limitations.

Reviewer #4: Yes, partially. The temporal dynamics need to be better explained.

Need to expand the limitation of the study considering the following ideas:

Vector-borne diseases are complex systems which are driven by natural (climate, environment) and socio-ecological factors. The study just considered the population and number of cases of the diseases individually or in co-occurrence. However, there are many more variables that are relevant in the dynamics of the transmission. Environmental conditions, climate factors, vectors distribution, type of virus circulating, host mobility, and health and policy systems. Some of these limitations may be mentioned in this study.

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

Yes, with the limitation above mentioned.

-Is public health relevance addressed?

Yes, I suggest that considering issues of global health and climate change.

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

Reviewer #3: Line 24: (typo) the number of dengue cases reported here do not match the number reported in line 184.

Line 96: (typo) shouldn't this be 1,123 municipalities?

Line 160, 161: Missing indicator function in equation 6.

Line 184: Rewording this sentence so that there aren't 16 numbers in a row might improve readability.

Line 186-187: Split this sentence into two. One regarding dengue and one regarding CHIK.

Line 218: (typo) ", an e[a]stern pattern"

Line 268: (typo) "likely the place of emergenc[e] of new..."

Line 276-277: why is proximity to Venezuelan border a condition that would favor the introduction of new pathogens? I think additional clarification is necessary here.

Line 286: This sentence could use a bit of reworking for the sake of clarification. What does is mean for a cluster to be "most likely"?

Lines 298-307: I think this paragraph needs to be reworked or removed. The point of the paragraph is not coming through clearly. The three cities discussed are not near the border with Ecuador, yet half of the paragraph is about proximity to Ecuador. I can't quite see how these connect.

Reviewer #4: A better explanation on the figure 3 description, and improve the visualization of the risk maps along the time.

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: Other comment:

Abstract:

Please rewrite the third (Conclusions) section of the abstract. In particular, 1. remove the statement "To our knowledge, this is the first study...", and 2. instead of stating that this study furthers our knowledge, explain briefly how it furthers our knowledge / what knowledge it adds.

Reviewer #2: line 55: which species of Aedes is incriminated to transmit these arboviruses in Colombia?

Manuscript needs a full proofread. Text writing needs to be more fluid and easier-to-read. Maybe join similar sentences in a more organic passage without repeating the same words so much, and also create acronyms for dengue, Zika and chikungunya.

Can you clarify better the methodology applied to collect data? Were only confirmed cases considered? Or all those reported, even as suspects?

Regarding Kulldorff’s scan statistics, the authors used GINI index for optimization?

The first sentence of the discussion has already been written in the Introduction. I think the authors could start the discussion by mentioning the main findings of the study

line 284: please correct "Ae. aegypti"

In SatScan the authors performed a retrospective analysis of spatiotemporal clusters in Colombia. It would be interesting to see the spatial variations in the temporal trend clusters. The spatial variation in the temporal trends identifies the clusters with growth rates (or decreases) in and out of these clusters, and evaluates whether the difference between them is statistically significant.

The cases of these arboviruses are strictly related to the abundance of the Aedes vector. Would benefit from a short discussion about the climate (temperature, rainfall, humidity, vegetation cover) of the locations where the clusters were identified.

Is there any control method for Aedes in Colombia? e.g. some insecticide? Maybe discuss a little bit about how this could affect the distribution of cases over time

Reviewer #3: Thank you for the opportunity to review this work. This research contributes to the understanding of similarities and differences in the spatiotemporal spread of dengue, chikungunya, and Zika across Colombia. The better we can understand these dynamics, the better equipped we hopefully become to disrupt them in the future.

Below are a handful of comments summarizing revisions or clarifications that I think would strengthen the impact of work. In particular, further clarification around the thresholds used for cluster detection *and/or* the inclusion of sensitivity analyses varying these thresholds will improve the confidence in the stability of the reported results.

Overall, this is informative research that can be even more impactful if it more directly communicates with the findings presented in the existing literature, particularly those constructed from the same (SIVIGILA) data source.

Methods:

- line 168. A maximum radius of 150 km is set for the space-time clusters. The description of why this value was chosen is quite vague and non-informative (lines 173-175). I would like to know more about what it means to balance clusters in terms of number and size. How were decisions made about what was "too large" or "too small"? Why is a cluster comprised of a single municipality a bad thing? Previous literature on space-time clusters using municipality data have allowed clusters to be comprised of 1 municipality. Were sensitivity analyses performed for the maximum radius? This information would be useful for future decision-making in future research. I think the evidence for setting the radius could be included in the supplemental material and would be of interest to other researchers.

- line 176. Why were clusters restricted to a minimum of 100 cumulative cases over the temporal scanning window? What effect does this have on the comparability of dengue versus CHIK and Zika clusters since there are so many fewer cases of CHIK and Zika overall? More discussion of the setting of this minimum threshold and its impact on the results would be informative.

Results:

- Line 204. In describing the pattern of dispersal from Barranquilla, the text says that there was a southern pattern yet the arrowhead points north.

- Table 1. Related to my comment for lines 176 and 168, I wonder how these cluster results are impacted by the thresholds set on maximum distance and minimum number of cumulative cases per cluster? I recognize the need to set thresholds, but without further information on the validity of the thresholds, it would be useful to know how conclusions change as the thresholds themselves change. It is interesting that the minimum observed number of cases is highest for the disease with the lowest total case count.

Discussion:

- Line 292. Authors note that clusters of Aedes-borne diseases in the south-western region of Colombia have been identified in previous studies. Why weren’t they also identified here? Their references (29, 30) analyze the same SIVIGILA data over a similar temporal window at the same level (municipality) with, in the case of reference 29, the same scan statistic. A further discussion of how this work matches and does not match the existing knowledge generated by this data would better situate this work in the context of the literature.

Reviewer #4: The study aims to develop new knowledge examining the co-occurrence of all three arboviruses and describing the patterns of introduction including the speed and direction of the spread of chikungunya and Zika in Colombia. The authors propose to apply a cluster analysis to determine unique patterns of spatiotemporal disease risk in Colombia: through disease-specific and multivariate cluster analyses and through estimating the direction and speed of chikungunya and Zika introduction in Colombia.

Method is promising, however, the inclusion of other factors related with the complex dynamics of the VBD need to be include for future studies. Also, a better visual representation of the dynamics of the cluster may be generated. Considerations of climate variability, change and extremes would need to be included in the limitations of the study, and for future discussions of direction and speed on the region. Global and environmental changes potentially would generate new risk areas for VBD, that would need a better understanding of the dynamics for developing early warning systems for epidemics.

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: No

Reviewer #4: Yes: Mercy J. Borbor-Cordova

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

20 Jun 2022

Submitted filename: response_comments PNTDs.pdf

Click here for additional data file.

12 Jul 2022

Dear Dr Freitas,

We are pleased to inform you that your manuscript 'Spatio-temporal clusters and patterns of spread of dengue, chikungunya, and Zika in Colombia' has been provisionally accepted for publication in PLOS Neglected Tropical Diseases.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Alberto Novaes Ramos Jr

Academic Editor

PLOS Neglected Tropical Diseases

Victor Santana Santos

Section Editor

PLOS Neglected Tropical Diseases

***********************************************************

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: (No Response)

Reviewer #3: I would like to thank the authors for their thoughtful feedback to the previous round of comments. The objectives of the study are clearly articulated and the changes they have made have improved the clarity of the methods, results, and discussion.

**********

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: (No Response)

Reviewer #3: (No Response)

**********

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #3: A minor point, but for the sake of completeness I would like to revisit the discussion around the indicator function in Equation 6. I acknowledge that the indicator function is left empty (i.e., "I()") in the Kulldorff SaTScanJ user guide (https://www.satscan.org/techdoc.html, "Likelihood Ratio Test"). This appears to be done strictly for user flexibility. For example, the user of the software would want to set "I(c < E[c])" to identify low rate clusters. In this work, the authors are strictly searching for high rate clusters (line 161, 177) and so the indicator function should be explicitly defined as "I(c > E[c])". This practice of explicitly defining the indicator notation is demonstrated in the two Kulldorff manuscript references included in this work (citations: 41, 42).

**********

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: The authors have addressed all of my prior concerns

Reviewer #3: (No Response)

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: No

17 Aug 2022

Dear Dr Freitas,

We are delighted to inform you that your manuscript, "Spatio-temporal clusters and patterns of spread of dengue, chikungunya, and Zika in Colombia," has been formally accepted for publication in PLOS Neglected Tropical Diseases.

We have now passed your article onto the PLOS Production Department who will complete the rest of the publication process. All authors will receive a confirmation email upon publication.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any scientific or type-setting errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript. Note: Proofs for Front Matter articles (Editorial, Viewpoint, Symposium, Review, etc...) are generated on a different schedule and may not be made available as quickly.

Soon after your final files are uploaded, the early version of your manuscript will be published online unless you opted out of this process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Shaden Kamhawi

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Paul Brindley

co-Editor-in-Chief

PLOS Neglected Tropical Diseases