Dataset Information

Estimating dengue under-reporting in Puerto Rico using a multiplier model.

ABSTRACT: Dengue is a mosquito-borne viral illness that causes a variety of health outcomes, from a mild acute febrile illness to potentially fatal severe dengue. Between 2005 and 2010, the annual number of suspected dengue cases reported to the Passive Dengue Surveillance System (PDSS) in Puerto Rico ranged from 2,346 in 2006 to 22,496 in 2010. Like other passive surveillance systems, PDSS is subject to under-reporting. To estimate the degree of under-reporting in Puerto Rico, we built separate inpatient and outpatient probability-based multiplier models, using data from two different surveillance systems-PDSS and the enhanced dengue surveillance system (EDSS). We adjusted reported cases to account for sensitivity of diagnostic tests, specimens with indeterminate results, and differences between PDSS and EDSS in numbers of reported dengue cases. In addition, for outpatients, we adjusted for the fact that less than 100% of medical providers submit diagnostic specimens from suspected cases. We estimated that a multiplication factor of between 5 (for 2010 data) to 9 (for 2006 data) must be used to correct for the under-reporting of the number of laboratory-positive dengue inpatients. Multiplication factors of between 21 (for 2010 data) to 115 (for 2008 data) must be used to correct for the under-reporting of laboratory-positive dengue outpatients. We also estimated that, after correcting for underreporting, the mean annual rate, for 2005-2010, of medically attended dengue in Puerto Rico to be between 2.1 (for dengue inpatients) to 7.8 (for dengue outpatients) per 1,000 population. These estimated rates compare to the reported rates of 0.4 (dengue outpatients) to 0.1 (dengue inpatients) per 1,000 population. The multipliers, while subject to limitations, will help public health officials correct for underreporting of dengue cases, and thus better evaluate the cost-and-benefits of possible interventions.

SUBMITTER: Shankar MB

PROVIDER: S-EPMC6095627 | biostudies-literature | 2018 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Estimating dengue under-reporting in Puerto Rico using a multiplier model.

Shankar Manjunath B MB Rodríguez-Acosta Rosa L RL Sharp Tyler M TM Tomashek Kay M KM Margolis Harold S HS Meltzer Martin I MI

PLoS neglected tropical diseases 20180806 8

Dengue is a mosquito-borne viral illness that causes a variety of health outcomes, from a mild acute febrile illness to potentially fatal severe dengue. Between 2005 and 2010, the annual number of suspected dengue cases reported to the Passive Dengue Surveillance System (PDSS) in Puerto Rico ranged from 2,346 in 2006 to 22,496 in 2010. Like other passive surveillance systems, PDSS is subject to under-reporting. To estimate the degree of under-reporting in Puerto Rico, we built separate inpatient ...[more]

PMID: 30080848

Similar Datasets

Project description:BackgroundThe incidence and severity of dengue in Latin America has increased substantially in recent decades and data from Puerto Rico suggests an increase in severe cases. Successful clinical management of severe dengue requires early recognition and supportive care.MethodsFatal cases were identified among suspected dengue cases reported to two disease surveillance systems and from death certificates. To be included, fatal cases had to have specimen submitted for dengue diagnostic testing including nucleic acid amplification for dengue virus (DENV) in serum or tissue, immunohistochemical testing of tissue, and immunoassay detection of anti-DENV IgM from serum. Medical records from laboratory-positive dengue fatal case-patients were reviewed to identify possible determinants for death.ResultsAmong 10,576 reported dengue cases, 40 suspect fatal cases were identified, of which 11 were laboratory-positive, 14 were laboratory-negative, and 15 laboratory-indeterminate. The median age of laboratory-positive case-patients was 26 years (range 5 months to 78 years), including five children aged < 15 years; 7 sought medical care at least once prior to hospital admission, 9 were admitted to hospital and 2 died upon arrival. The nine hospitalized case-patients stayed a mean of 15 hours (range: 3-48 hours) in the emergency department (ED) before inpatient admission. Five of the nine case-patients received intravenous methylprednisolone and four received non-isotonic saline while in shock. Eight case-patients died in the hospital; five had their terminal event on the inpatient ward and six died during a weekend. Dengue was listed on the death certificate in only 5 instances.ConclusionsDuring a dengue epidemic in an endemic area, none of the 11 laboratory-positive case-patients who died were managed according to current WHO Guidelines. Management issues identified in this case-series included failure to recognize warning signs for severe dengue and shock, prolonged ED stays, and infrequent patient monitoring.

Project description:BackgroundPredictive models can serve as early warning systems and can be used to forecast future risk of various infectious diseases. Conventionally, regression and time series models are used to forecast dengue incidence, using dengue surveillance (e.g., case counts) and weather data. However, these models may be limited in terms of model assumptions and the number of predictors that can be included. Machine learning (ML) methods are designed to work with a large number of predictors and thus offer an appealing alternative. Here, we compared the performance of ML algorithms with that of regression models in predicting dengue cases and outbreaks from 4 to up to 12 weeks in advance. Many countries lack sufficient health surveillance infrastructure, as such we evaluated the contribution of dengue surveillance and weather data on the predictive power of these models.MethodsWe developed ML, regression, and time series models to forecast weekly dengue case counts and outbreaks in Iquitos, Peru; San Juan, Puerto Rico; and Singapore from 1990-2016. Forecasts were generated using available weekly dengue surveillance, and weather data. We evaluated the agreement between model forecasts and actual dengue observations using Mean Absolute Error and Matthew's Correlation Coefficient (MCC).ResultsFor near term predictions of weekly case counts and when using surveillance data, ML models had 21% and 33% less error than regression and time series models respectively. However, using weather data only, ML models did not demonstrate a practical advantage. When forecasting weekly dengue outbreaks 12 weeks in advance, ML models achieved a maximum MCC of 0.61.ConclusionsOur results identified 2 scenarios when ML models are advantageous over regression model: 1) predicting dengue weekly case counts 4 weeks ahead when dengue surveillance data are available and 2) predicting weekly dengue outbreaks 12 weeks ahead when dengue surveillance data are unavailable. Given the advantages of ML models, dengue early warning systems may be improved by the inclusion of these models.

Dataset Information

Estimating dengue under-reporting in Puerto Rico using a multiplier model.

Publications

Estimating dengue under-reporting in Puerto Rico using a multiplier model.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets