Our unique forecasting techniques combine in-house data with client information to deliver unrivalled forecasts for anywhere in the world. Contact us Email: offshoresales meteogroup.
Founded in , the company has an unrivalled reputation for quality and innovation in weather forecasting. Millions discover their favorite reads on issuu every month. Give your content the digital home it deserves. Get it to any device in seconds. Nowcasting Pro6. Publish for free today. Go explore. However, we find that many health facilities in endemic areas opt to send their reports directly to Georgetown or through alternative routes, due to issues in road access and other challenges, often resulting in delayed or missed reports personal communication.
Furthermore, while the data is intended to be documented in the system in near real-time, with summary reports sent from regions to the central office via USBs, physical copies of individual case registers from facilities are sometimes delivered to the central office, leading to additional bottlenecks in reporting personal communication. Together, these challenges contribute to the number of cases recorded in the central database at the end of the month potentially failing to reflect the true burden of disease at a given time, with cases typically reported more than a month after they occur See Fig 1B.
These issues are not only common for malaria control and surveillance programs, but for many other epidemiological reporting systems, hindering the ability of public health programs to implement appropriate and optimal measures for control [ 3 , 12 — 16 ]. Modeling efforts to translate available reporting data to more meaningful and accurate measures of real-time incidence could help address some of these challenges.
Map of malaria endemic regions and region 4; B. Boxplot reporting the proportion of annually aggregated cases from — reported by the end of the month for each region, with colors corresponding to the regions depicted in the map. Map created using the R package ssplot. The most widely used are Bayesian estimation and regularized regression, which take advantage of the fact that the timing of delays may be relatively predictable. The Bayesian approach, broadly considered the paragon approach, encapsulates both delay trends given prior known cases and epidemiological trends to estimate disease occurrence and mortality, with several recent extensions implemented to better uncover the true magnitude of COVID cases and deaths [ 14 , 15 , 17 — 23 ].
Extensions of these methods, proposed by Bastos et al. Relatedly, Reich et al. Penalized regression methods, such as the ARGO, Net, and ARGONet models proposed by Santillana, Lu, and others have also been used to incorporate a diversity of case reporting and alternative data sources, to learn from previous time trends in delays and transmission, and the spatial dependencies in these trends, to predict rates of influenza-like illness [ 24 ].
However, this class of models is only well-suited for contexts in which these varied digital sources are readily available and easily integrated into routine model training, testing, and validation efforts [ 24 ].
To address challenges in reporting delays for malaria surveillance in Guyana, we developed nowcasting approaches to estimate monthly cases of malaria in each endemic region, using data from the central VCS database from to to develop and test our model.
We show that proximity to high-risk vulnerable populations, namely mining sites and Amerindian populations, is associated with reporting delays and that delays are relatively consistent over time. Importantly, our predictions provide extensive improvements in surveillance capacity for remote areas and are presently being used by VCS to help characterize prevailing case counts across regions in the face of added challenges due to COVID, to help facilitate the allocation of malaria prevention and control interventions and overall planning initiatives centered on elimination.
Historically, malaria has affected groups that primarily live and work in Regions 1, 7, 8, and 9, otherwise known as hinterland regions because they are sparsely populated, forested areas that are difficult to navigate. These characteristics contribute to the ongoing challenge of time lags in reporting of malaria cases which threaten the delivery of an appropriate public health response.
The nowcasting tool generates estimates of current malaria burden at the sub-national level based on real-time reporting and a contextual understanding of the malaria epi-trends over the past years. Thus, we illustrate how nowcasting methods offer a general, tractable approach for improving decision-making for malaria control programs in countries that have significant reporting delays.
Delays are defined as the days elapsing from when a patient is registered as a case at a health facility and when the patient is documented at the central VCS database. We excluded the 5. While we acknowledge that the inclusion of patients with recurrent infections may hinder a mechanistic analysis of potential transmission patterns in each region, the primary goal of this analysis is to learn from prior clinical reporting trends to arrive at more informed estimates of current case loads, which include all patients, in the face of reporting delays.
Monthly total precipitation data was extracted for each region by year [ 26 ]. We defined NDCs, rather than regions, as our spatial unit of analysis to ensure adequate statistical power, and aggregated median delays across years by locality for each NDC.
A shapefile of mapped mines was drawn from the U. Geological Survey Mineral resources data system [ 29 ]. For each region, we additionally estimated a the Pearson correlation between median reporting delays and regional connectivity in , and b the Pearson correlation and cross-correlations between monthly total precipitation levels and median delays from — We focused our analysis on case register data from region 4, where many malaria patients come for treatment but there is no local transmission, and from malaria-endemic regions 1, 7, 8, and 9 see Fig 1A for map.
We further used the Getis-ord statistic, which compares the sum of characteristics in each neighborhood to their overall mean [ 32 , 33 ], to identify local clusters of NDCs with higher reporting delays. For all spatial analyses, we omitted the 2. All significance maps for local autocorrelation tests that we provide report FDR-adjusted p-values. We used the following two approaches to estimate revised i. In all of the proposed approaches, our real-time malaria case count estimations were produced via a data assimilation process, in that only information that would have been available at the time of estimation was used to train our models.
Subsequent estimates were produced by dynamically training the models below, such that as more information became available the training set consisted of a new set of month long observations. This approach was chosen as a better way to capture the potential time-evolving nature of reporting delays.
For all models, we centered the covariates to ensure standardization of inputs. For each region i, we used an elastic net penalized regression, i. Note that the number of known cases for a given month t increases with the month of separation, as more information accumulates over time. For example, as we would expect to have more information about cases occurring in the prior month than cases occurring in the current month, would exceed.
A matrix of case counts from each of the twelve months prior to and including time t the most up to date case count that were known by the end of time t was used to dynamically train and test the data imputation model in order to predict the number of revised cases for t. We computed a 3-fold cross validation within each twelve month window in order to identify the tuning parameters that resulted in the lowest mean square error and used these parameters for the subsequent unseen monthly prediction.
In other words, training of our models was conducted using strictly data that would have been available at the time of prediction. We experimented with different number of folds 3, 5, or 10 for cross-validation to explore whether this choice would have an impact on the quality of our predictions and found that these additional choices led to qualitatively similar results, so we report the results of the most cost-efficient approach in the manuscript.
We also implemented an elastic net regression which additionally takes as inputs the predicted number of cases in month t for all other regions j from the previous data imputation models. Both network models consider the location where malaria activity is estimated as a node in a network potentially influenced by malaria activity happening in other potentially neighboring locations nodes.
The second network model NM2 takes the following form 3 , where y j t denotes the predicted case counts for region j at time t, estimated from the data imputation model for region j 1 3. Please refer to S1 Diagram for a schematic visualizing the distinct and shared inputs of the two network models. A matrix of case counts from each of the twelve months prior to and including t that were known by the end of t for region i, case counts from each of the twelve months prior to t that were known by the end of t for all other regions j, and total precipitation levels for region i in month t, was used to dynamically train and test NM1 in order to predict the number of revised cases for t for region i.
A matrix of case counts from each of the twelve months prior to and including t that were known by the end of t for region i, case counts from each of the twelve months prior to t that were known by the end of t for all other regions j, DIM-predicted case counts in t for all other regions j, and total precipitation levels for region i in t was used to dynamically train and test NM2 in order to predict the number of revised cases for t for region i.
We replicate the same procedure as in the data imputation model to test and train the two network models. We assigned the mean across models as the point estimate for the first three months of observations. Subsequently, following an approach outlined in Poirier et al [ 35 ] citing work by Yang et al [ 36 ] to capture changing uncertainty at different points in time, we generated the lower and upper limits of the confidence intervals by subtracting or adding the RMSE associated to a moving window of the last 24 errors, respectively, to the upcoming point estimate [ 35 , 36 ].
Lastly, we compared our model to a modeling approach detailed by Bastos et al [ 17 ]. We chose to compare to the Bastos et al.
For simplicity, we used all priors and hyperparameters as defined in their open access code provided in their manuscript [ 17 ]. See S1 Diagram for a diagram outlining key steps in this data processing to model implementation process. All data pre-processing, a-spatial and nowcasting analyses were conducted in R version 4. All spatial analyses were conducted in ArcMap version Fig 1B illustrates how these delays occurred in different regions see S2 Fig for full empirical density functions and S1 Fig for annual confirmed cases and median delays for each region , and highlights the considerable heterogeneity in the extent of reporting delays observed between them.
Due to the pervasiveness of mobile populations, particularly in mining regions, population size estimates and consequently, measures of incidence, are neither meaningful nor tractably quantifiable, so we chose not to report population standardized case counts. For all other regions, we found no significant cross-correlations within a meaningful range in lags. Significant clusters of NDCs marked by higher delays were most concentrated in regions 1 and 7, while clusters of NDCs marked by lower delays were most concentrated in regions 4 and 6 S3 and S4 Figs.
Both clustering maps signal NDCs within region 1 marked by higher reporting delays and a relatively increased presence of Amerindian settlements and mines. Note that NDCs showing the inverse of this relationship, i. Map of mining sites; C. Map of Amerindian areas. For visual purposes, we show only high-high clusters of areas reporting at higher delays with a greater density of mines or Amerindian settlements.
Full bivariate cluster and significance maps can be found in S5 Fig. We found weak evidence for overall synchronicity in delay distributions across regions, with an estimated mean cross-correlation between the regional monthly delay distributions equal to 0. Overall, these minimally correlated delay distributions may reflect varying seasonality in transmission between regions. These findings support the use of data imputation models, which rely on previous trends in delays in a given region to inform ongoing predictions.
DIM model predictions coincided with the two highest peaks in true cases and generally reflected trends throughout time, despite substantially underestimating a moderate peak in late and generally overshooting revised cases in late Fig 3B. Given that the number of cases known by the end of each month in region 4 closely approximated the number of revised cases, this approach may be of limited practical utility. Due to the high accuracy of region 4 DIM, we chose not to implement network models for this region.
Solid red lines indicate the number of cases estimated from the data imputation model, dashed red lines indicate the number of cases known by the end of the month and solid blue lines indicate the true number of eventually reported cases. In contrast, the DIM models for regions 1, 7 and 8 significantly improved on revised case counts for each of these regions.
However, region 1 predictions broadly approximated converged case trends, although still of an attenuated magnitude, from onwards, and model predictions for region 7 still generally reflected the most recent upward trend in cases from , with the exception of late , improving on month-end reporting substantially Fig 3A and 3C. Finally, the DIM model for region 9 failed to converge identify a stable set of parameter values due to data sparsity, largely driven by the exclusion of cases with implausible dates of documentation, as defined previously.
Of note, the data imputation model estimates for all regions, particularly for region 1 and 7, were marked by wide confidence intervals from —, a consequence of the poor model performance in the years just preceding. However, it is important to highlight that we observe consistently sharper confidence intervals, and thus an increased level of certainty in our estimates, for cases occurring in the more recent past, i.
Finally, for all regions, a slight majority of monthly converged case counts fall within our estimated confidence intervals, testifying the effectiveness of our instituted method for prospectively estimating confidence intervals.
Both network models exhibited improvements for all regions Fig 4A—4D. However, both models demonstrated continued underestimation of case counts from to NM1 predictions for region 1 reported only marginally greater accuracy than the corresponding predictions from NM2. Model predictions from late onwards better mirrored the magnitude and timing of current trends Fig 4E and 4F. Unlike regions 1 and 7, model predictions were most improved for NM2.
The data imputation and network models significantly improved over low known case counts for regions 1, 7, and 8, most strikingly for regions 1 and 8, with error rates reduced by a factor of 1.
In all cases, when relying on predictions from the best performing models, the median fraction of cases captured within a month for all regions from to now ranges from 0. Teplice Kunovice okr. Letovice okr. Liberec Libina okr.
Svitavy Litovel okr. Most Loket okr. Sokolov Lom okr. Most Lomnice nad Popelkou okr. Louny Lovosice okr. Brno-venkov Mohelnice okr. Most Mosty u Jablunkova okr. Karlovy Vary Nepomuk okr. Netolice okr. Prachatice Nivnice okr. Olomouc Opava okr. Teplice Oslavany okr. Brno-venkov Ostrava okr.
Pardubice Paskov okr. Jihlava Poruba okr. Praha Praha 1 okr. Praha Praha 10 okr. Praha Praha 11 okr. Praha Praha 12 okr. Praha Praha 13 okr. Praha Praha 14 okr.
Praha Praha 15 okr. Praha Praha 2 okr. Praha Praha 3 okr. Praha Praha 4 okr. Praha Praha 5 okr.
0コメント