Dengue vaccine supplies under endemic and epidemic conditions in three dengue-endemic countries: Colombia, Thailand, and Vietnam

A machine learning model to assess potential misdiagnosed dengue hospitalization

Dengue, like other arboviruses with broad clinical spectra, can easily be misdiagnosed as other infectious diseases due to the overlap of signs and symptoms. During large outbreaks, severe dengue cases have the potential to overwhelm the health care system and understanding the burden of dengue hospitalizations is therefore important to better allocate medical care and public health resources. A machine learning model that used data from the Brazilian public healthcare system database and the National Institute of Meteorology (INMET) was developed to estimate potential misdiagnosed dengue hospitalizations in Brazil. The data was modeled into a hospitalization level linked dataset. Then, Random Forest, Logistic Regression and Support Vector Machine algorithms were assessed. The algorithms were trained by dividing the dataset in training/test set and performing a cross validation to select the best hyperparameters in each algorithm tested. The evaluation was done based on accuracy, precision, recall, F1 score, sensitivity, and specificity. The best model developed was Random Forest with an accuracy of 85% on the final reviewed test. This model shows that 3.4% (13,608) of all hospitalizations in the public healthcare system from 2014 to 2020 could have been dengue misdiagnosed as other diseases. The model was helpful in finding potentially misdiagnosed dengue and might be a useful tool to help public health decision makers in planning resource allocation.

The Threat of Climate Change on Tick-Borne Infections: Rising Trend of Infections and Geographic Distribution of Climate Risk Factors Associated With Ticks

Ticks transmit a wide range of pathogens. The spread of tick-borne infections is an emerging, yet often overlooked, threat in the context of climate change. The infections have rapidly increased over the past few years in South Korea despite no significant changes in socioeconomic circumstances. We investigated the impact of climate change on the surge of tick-borne infections and identified potential disease hot spots at a resolution of 5 km by 5 km. A composite index was constructed based on multiple climate and environmental indicators and compared with the observed tick-borne infections. The surge of tick-borne episodes corresponded to the rising trend of the index over time. High-risk areas identified by the index can be used to prioritize locations for disease prevention activities. Monitoring climate risk factors may provide an opportunity to predict the spread of the infections in advance.

Dengue models based on machine learning techniques: A systematic literature review

BACKGROUND

Dengue modeling is a research topic that has increased in recent years. Early prediction and decision-making are key factors to control dengue. This Systematic Literature Review (SLR) analyzes three modeling approaches of dengue: diagnostic, epidemic, intervention. These approaches require models of prediction, prescription and optimization. This SLR establishes the state-of-the-art in dengue modeling, using machine learning, in the last years.

METHODS

Several databases were selected to search the articles. The selection was made based on Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology. Sixty-four articles were obtained and analyzed to describe their strengths and limitations. Finally, challenges and opportunities for research on machine-learning for dengue modeling were identified.

RESULTS

Logistic regression was the most used modeling approach for the diagnosis of dengue (59.1%). The analysis of the epidemic approach showed that linear regression (17.4%) is the most used technique within the spatial analysis. Finally, the most used intervention modeling is General Linear Model with 70%.

CONCLUSIONS

We conclude that cause-effect models may improve diagnosis and understanding of dengue. Models that manage uncertainty can also be helpful, because of low data-quality in healthcare. Finally, decentralization of data, using federated learning, may decrease computational costs and allow model building without compromising data security.

Geographical distribution of risk factors for invasive non-typhoidal Salmonella at the subnational boundary level in sub-Saharan Africa

Background

Invasive non-typhoidal Salmonella (iNTS) is a growing health-concern in many parts of sub-Saharan Africa. iNTS is associated with fatal diseases such as HIV and malaria. Despite high case fatality rates, the disease has not been given much attention. The limited number of population-based surveillance studies hampers accurate estimation of global disease burden. Given the lack of available evidence on the disease, it is critical to identify high risk areas for future surveillance and to improve our understanding of iNTS endemicity.

Methods

Considering that population-based surveillance data were sparse, a composite index called the iNTS risk factor (iNRF) index was constructed based on risk factors that commonly exist across countries. Four risk factors associated with the prevalence of iNTS were considered: malaria, HIV, malnutrition, and safe water. The iNRF index was first generated based on the four risk factors which were collected within a 50 km radius of existing surveillance sites. Pearson product-moment correlation was used to test statistical associations between the iNRF index and the prevalence of iNTS observed in the surveillance sites. The index was then further estimated at the subnational boundary level across selected countries and used to identify high risk areas for iNTS.

Results

While the iNRF index in some countries was generally low (i.e. Rwanda) or high (i.e. Cote d’Ivoire), the risk-level of iNTS was variable not only by country but also within a country. At the provincial-level, the highest risk area was identified in Maniema, the Democratic Republic of Congo, whereas Dakar in Senegal was at the lowest risk.

Conclusions

The iNRF index can be a useful tool to understand the geographically varying risk-level of iNTS. Given that conducting a population-based surveillance study requires extensive human and financial resources, identifying high risk areas for iNTS prior to a study implementation can facilitate an appropriate site-selection process in the future.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-021-06198-1.

Sensitivity, uncertainty and identifiability analyses to define a dengue transmission model with real data of an endemic municipality of Colombia

Dengue disease is a major problem for public health surveillance entities in tropical and subtropical regions having a significant impact not only epidemiological but social and economical. There are many factors involved in the dengue transmission process. We can evaluate the importance of these factors through the formulation of mathematical models. However, the majority of the models presented in the literature tend to be overparameterized, with considerable uncertainty levels and excessively complex formulations. We aim to evaluate the structure, complexity, trustworthiness, and suitability of three models, for the transmission of dengue disease, through different strategies. To achieve this goal, we perform structural and practical identifiability, sensitivity and uncertainty analyses to these models. The results showed that the simplest model was the most appropriate and reliable when the only available information to fit them is the cumulative number of reported dengue cases in an endemic municipality of Colombia.

The threat of climate change to non-dengue-endemic countries: increasing risk of dengue transmission potential using climate and non-climate datasets

Background

Dengue is a major public health problem in the tropics and sub-tropics, but the disease is less known to non-dengue-endemic countries including in Northeast Asia. However, an unexpected dengue outbreak occurred in 2014 in Japan. Given that autochthonous (domestic) dengue cases had not been reported for the past 70 years in Japan, this outbreak was highly unusual and suggests that several environmental factors might have changed in a way that favors vector mosquitoes in the Northeast Asian region.

Methods

A Climate Risk Factor (CRF) index, as validated in previous work, was constructed using climate and non-climate factors. This CRF index was compared to the number of reported dengue cases in Tokyo, Japan where the outbreak was observed in 2014. In order to identify high-risk areas, the CRF index was further estimated at the 5 km by 5 km resolution and mapped for Japan and South Korea.

Results

The high-risk areas determined by the CRF index corresponded well to the provinces where a high number of autochthonous cases were reported during the outbreak in Japan. At the provincial-level, high-risk areas for dengue fever were the Eastern part of Tokyo and Kanakawa, the South-Eastern part of Saitama, and the North-Western part of Chiba. While a relatively small number of high-risk areas were identified in South Korea compared with Japan, the high-risk areas in South Korea include popular tourist destinations where international visitors have been increasing.

Conclusion

The recent dengue outbreak in Japan may signal that the two adjacent non-dengue-endemic countries are also exposed to the risk of temporal and sporadic behavior of dengue fever. It is critical to understand potential high-risk areas for future outbreaks and to set up appropriate prevention activities at the governmental-level.

A multi-country study of dengue vaccination strategies with Dengvaxia and a future vaccine candidate in three dengue-endemic countries: Vietnam, Thailand, and Colombia

BACKGROUND

The dengue vaccination era began when Dengvaxia (CYD-TDV) became available in 2016. In addition, several second-generation vaccine candidates are currently in phase 3 trials, suggesting that a broader availability of dengue vaccines may be possible in the near future. Advancing on the recent WHO-SAGE recommendations for the safe and effective use of CYD-TDV at the regional level on average, this study investigates the vaccination impacts and cost-effectiveness of CYD-TDV and of a hypothetical new vaccine candidate (NVC) in a country-specific manner for three endemic countries: Vietnam, Thailand, and Colombia.

METHODS

The vaccination impacts of CYD-TDV and NVC were derived by fitting the empirical seroprevalence rates of 9 year olds into an individual-based meta-population transmission model, previously used for the WHO-SAGE working group. The disability-adjusted life years were estimated by applying country-specific parametric values. The cost-effectiveness analyses of four intervention strategies in combination with routine and catch-up campaigns were compared for both vaccines to inform decision makers regarding the most suitable immunization program in each of the three countries.

RESULTS AND CONCLUSION

Both CYD-TDV and NVC could be cost-effective at the DALY threshold cost of $2000 depending upon vaccination costs. With CYD-TDV, targeting 9 year olds in routine vaccination programs and 10-29 year olds as a one-off catch-up campaign was the most cost-effective strategy in all three countries. With NVC, while the most cost-effective strategy was to vaccinate 9-29 and 9-18 year olds in Vietnam and Thailand respectively, vaccinating younger age cohorts between 1 and 5 years old in Colombia was more cost-effective than other strategies. Given that three countries will soon face decisions regarding whether and how to incorporate CYD-TDV or future dengue vaccines into their budget-constrained national immunization programs, the current study outcomes can be used to help decision makers understand the expected impacts and cost-effectiveness of such vaccines.

Feasibility and willingness to pay for dengue vaccine in the threat of dengue fever outbreaks in Vietnam

BACKGROUND

The escalation of dengue fever (DF) cases in recent years and the occurrence of a large-scale DF outbreak in 2017 underline the importance of dengue vaccines in Vietnam. Given the potential benefits of the dengue vaccines and the need for copayment by the private sector, this study aims to evaluate the willingness to pay (WTP) for the dengue vaccines in patients with DF in Northern Vietnam.

METHODS

A cross-sectional study was conducted on 330 in-and-out patients with DF admitted to the Bach Mai Hospital. We used the contingent valuation method to evaluate the WTP for dengue vaccines. Socioeconomic and clinical characteristics were also investigated. Multivariate interval and logistic regression models were used to estimate the average amount of WTP and identify the factors associated with the WTP.

RESULTS

Around 77.3% patients were willing to pay an average amount of US$ 67.4 (95% CI=57.4-77.4) for the vaccine. People of higher ages, those having health insurance, those traveling in the past 15 days or suffering from anxiety/depression were less likely to be willing to pay for the dengue vaccine. However, people having a longer duration of DF or having problems with mobility were positively associated with WTP for the dengue vaccine. Patients educated to more than high school levels (Coeff.=31.31; 95% CI=3.26-59.35), those in the richest quintile (Coeff.=62.76; 95% CI=25.40; 100.13), or those having a longer duration of the disease (Coeff.=6.18; 95% CI=0.72-11.63) were willing to pay a higher amount.

CONCLUSION

This study highlights a relatively high rate and amount of WTP for the dengue vaccine among patients with DF. Psychological counseling services as well as educational campaigns should be undertaken to improve the WTP for the vaccine. Moreover, government subsidies should be given to increase the coverage of the vaccine in the future, especially for the poor.