Spatiotemporal Dynamics and Forecasting of Severe Malaria Incidence Among Pregnant Women in the Democratic Republic of Congo (2020-2024): A Retrospective Observational Study

Spatiotemporal Dynamics and Forecasting of Severe Malaria Incidence Among Pregnant Women in the Democratic Republic of Congo (2020-2024): A Retrospective Observational Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Spatiotemporal Dynamics and Forecasting of Severe Malaria Incidence Among Pregnant Women in the Democratic Republic of Congo (2020-2024): A Retrospective Observational Study Aymar Akilimali, Jones Onesime, Abdisalam Hassan MUSE, Mukhtar Abdi HASSAN, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8707200/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Malaria remains a major public health concern in the Democratic Republic of Congo (DRC). This retrospective observational study examined the spatiotemporal variation in severe malaria incidence among pregnant women across DRC provinces from 2020 to 2024 and projected future trends to 2026. Methods Monthly data on severe malaria cases were obtained from the District Health Information Software 2 (DHIS2) managed by the DRC Ministry of Health. Data from 26 provinces were cleaned, harmonized, and analyzed using descriptive statistics, temporal trend visualizations, and spatial autocorrelation methods. Five forecasting models Seasonal Autoregressive Integrated Moving Average (SARIMA), Exponential Smoothing State Space (ETS), Trigonometric Box-Cox ARMA Trend Seasonal (TBATS), Autoregressive Neural Network (ARNN_NNAR), and Autoregressive Fractionally Integrated Moving Average (ARFIMA) were applied to predict future incidence. Model accuracy was assessed using Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Symmetric Mean Absolute Percentage Error (sMAPE), and Mean Absolute Scaled Error (MASE), with sMAPE used to identify the best-performing model for each province. Results Considerable spatial and temporal heterogeneity was observed. Kinshasa, Haut-Katanga, and Kasaï-Central reported persistently high incidence, while Nord-Ubangi and Mongala showed the lowest. Seasonal peaks occurred mainly between May and December. The ETS, ARNN_NNAR, and ARFIMA models demonstrated superior accuracy across different provinces, reflecting varied epidemic patterns. Forecasts for 2026 indicated persistent high-incidence clusters in western and central provinces, particularly Kongo-Central and Kwango. Conclusion The study underscores significant spatial disparities and rising trends in severe malaria among pregnant women in the DRC. The findings provide critical evidence to guide geographically targeted, seasonally timed interventions and inform policy to strengthen malaria prevention and control. Democratic Republic of Congo Severe Malaria Malaria Incidence Pregnant Women Retrospective Observational Study Spatiotemporal Dynamics and Forecasting Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1. INTRODUCTION Malaria remains one of the leading global public health challenges, particularly in low- and middle-income countries. The World Health Organization (WHO) ( 1 ) reported an estimated 249 million malaria cases and 608 000 deaths in 2022, with over 94% of these occurring in sub-Saharan Africa, where climatic and socioeconomic conditions facilitate persistent Plasmodium transmission. Within this context, the Democratic Republic of Congo (DRC) is among the six countries accounting for more than half of all global malaria cases, representing approximately 12% of the worldwide burden ( 1 ). Malaria’s impact extends beyond morbidity and mortality to considerable economic losses, further straining fragile health systems and impeding progress towards SDG 3, which seeks to end the epidemics of malaria and other communicable diseases by 2030. Pregnant women remain one of the most vulnerable groups to malaria infection. Physiological immunosuppression during pregnancy increases susceptibility, and infection often leads to maternal anaemia, low birth weight, preterm delivery, stillbirth, and elevated maternal and neonatal mortality ( 1 , 2 ). In 2022, an estimated 35.4 million pregnancies were at risk of malaria infection across the African region, with approximately 12.7 million (36%) directly exposed ( 3 ). In the DRC, the prevalence of malaria infection among pregnant women has been reported as high as 63.3%, with maternal anaemia affecting 56.8% of those infected ( 4 ). Such outcomes underscore the continuing public health significance of malaria in pregnancy and its intersection with maternal and child health priorities. Several factors aggravate the situation in the DRC. Limited access to quality antenatal care, inconsistent uptake of intermittent preventive treatment in pregnancy (IPTp), sub-optimal distribution of insecticide-treated nets (ITNs), and increasing insecticide resistance constrain the effectiveness of control strategies ( 5 ). Environmental variables such as rainfall, temperature, and humidity shape the abundance and distribution of Anopheles vectors, while sociocultural and economic disparities influence health-seeking behaviour and access to preventive measures. The country’s vast geography and ecological diversity lead to pronounced regional heterogeneity in malaria incidence, complicating national surveillance and response efforts ( 6 , 7 ). Despite growing evidence on malaria epidemiology, there is a paucity of research examining the spatiotemporal dynamics and forecasting of severe malaria among pregnant women in the DRC. Existing studies are often limited to aggregate national estimates, lacking fine-scale temporal and provincial analyses necessary for precision public health planning. This knowledge gap impedes the ability of health authorities to identify high-risk provinces, anticipate seasonal peaks, and target interventions effectively. Against this backdrop, the present study aims to analyse the spatiotemporal variation in severe malaria incidence among pregnant women across DRC provinces from 2020 to 2024 and to forecast future trends up to 2026. Findings from this study will provide actionable evidence to strengthen geographically targeted and seasonally aligned malaria interventions, enhance health-system responsiveness, and support national progress towards achieving SDG 3. 2. METHODOLOGY This study employed a retrospective observational design to analyse the spatiotemporal variation and forecast the incidence of severe malaria among pregnant women in the Democratic Republic of Congo. The methodology involved several stages: data collection and preprocessing, descriptive analytics, spatiotemporal analysis, time series modeling and forecasting, and model evaluation. 2.1. Data Sources Monthly incidence data for "A 1.5 Paludisme grave chez la femme enceinte" (Severe Malaria in Pregnant Women) was obtained from the District Health Information Software 2 (DHIS2) platform, managed by the DRC Ministry of Health, for the years 2020 to 2024. These data were provided in five separate CSV files, each corresponding to a specific year. For spatial analysis and visualization, a shapefile (cod_admbnda_adm1_rgc_itos_20190911.shp) representing the administrative boundaries of DRC provinces was utilized. 2.2. Data Collection and Preprocessing The initial step involved loading and merging the five annual CSV files into a single comprehensive dataset. This process required pivoting the monthly columns into a long format, creating a Month_Year column and a Severe_Malaria_Cases_PW column. Date conversion was a critical preprocessing step. The Month_Year column, containing French month names (e.g., "Janvier 2020"), was parsed to extract the month name and year. A mapping was created to convert French month names to their English abbreviations, allowing for the creation of a standardized Date column (representing the first day of each month). Rows with unparseable dates were filtered out. Province name mapping was performed to ensure consistency between the CSV data and the shapefile. A custom function was developed to standardize province names from the CSV (e.g., "bu Bas Uele Province") to match the ADM1_FR attribute in the shapefile (e.g., "Bas-Uele"). This involved removing two-letter codes and " Province" suffixes, and standardizing diacritics and hyphens (e.g., "Kasai" to "Kasaï", "Maindombe" to "Maï-Ndombe"). A verification step was included to identify and report any mismatches between the mapped CSV province names and the shapefile province names. Provinces not found in the shapefile were excluded from the analysis to maintain spatial integrity. Missing values in the Severe_Malaria_Cases_PW variable were addressed by first converting the column to a numeric type, coercing non-numeric entries to NA. Commas within numeric strings (e.g., "1,875") were removed prior to conversion. A complete historical date range was generated for all provinces, and any missing monthly entries were filled using linear interpolation (zoo::na.approx). Remaining NA values, typically at the beginning or end of a time series where interpolation was not possible, were replaced with zero, assuming no reported cases implied zero incidence. 2.3. Descriptive Analytics Descriptive statistics were computed to summarize the overall incidence of severe malaria among pregnant women in the DRC. This included minimum, maximum, mean, median, and standard deviation of cases across the entire study period (2020–2024). Monthly average cases for the entire DRC were calculated to observe general trends. Furthermore, provinces with the highest and lowest incidence in the latest available month (December 2024) were identified to highlight current hotspots and areas of lower burden. 2.4. Spatiotemporal Analysis Spatial distribution of severe malaria incidence was visualized through choropleth maps. An overall average map was generated by calculating the mean incidence for each province across the entire 2020–2024 period. Additionally, annual average maps were created for each year from 2020 to 2024 to illustrate yearly spatial shifts. A consistent color scale was applied across all maps to facilitate comparison. Temporal trends were analyzed using various visualizations. A time-series plot of the overall DRC average monthly incidence was generated to show national-level patterns. Individual time-series plots for each province, faceted for easy comparison, depicted provincial-level trends. Box plots were used to visualize seasonality (monthly distribution of cases across all years) and annual trends (yearly distribution of cases across all provinces). Inter-provincial correlations were assessed using a correlation heatmap. The data was pivoted to a wide format, with each province as a column, and the Pearson correlation coefficient was calculated between all pairs of provinces. Provinces with zero standard deviation in their case counts were excluded from this analysis to avoid undefined correlations. 2.5. Time Series Modeling and Forecasting For forecasting future incidence, five distinct time series models were applied to the monthly provincial data to capture both linear and non-linear temporal trends. The Seasonal Autoregressive Integrated Moving Average (SARIMA) model extended the conventional ARIMA framework to incorporate seasonal components, with the auto. Arima function from the forecast package automatically determining optimal model parameters. The Exponential Smoothing State Space (ETS) model, also from the forecast package, utilized exponential smoothing principles, assigning progressively smaller weights to older observations to account for evolving trends. The TBATS (Trigonometric, Box-Cox, ARMA, Trend, Seasonal) model was employed to manage complex seasonal patterns, non-linear trends, and Box-Cox transformations. Additionally, the Autoregressive Neural Network (ARNN_NNAR) model, implemented through the nnetar function, used a feed-forward neural network structure that incorporated lagged values of the series as predictors, enabling the modelling of non-linear temporal dependencies. Finally, the Autoregressive Fractionally Integrated Moving Average (ARFIMA) model was applied using the arfima function to accommodate fractional differencing and long-memory characteristics typical of persistent epidemiological data. The ANN_MLP model from the nnfor package was initially considered; however, execution errors occurred across all provinces due to incompatibility with the forecast: forecast method, which required a time-series (ts) object. Consequently, the ANN_MLP results were excluded from comparative analysis. 2.6. Model Evaluation Each model’s performance was evaluated through a systematic validation procedure. For every province, the dataset from January 2020 to December 2023 served as the training set, while data from January to December 2024 were retained for testing and validation. The models were trained using the historical data and subsequently generated forecasts for the validation period to assess predictive accuracy. Model performance was measured using multiple statistical indicators: the Root Mean Squared Error (RMSE), which quantified the average magnitude of prediction errors; the Mean Absolute Error (MAE), which captured the mean absolute deviation without regard to direction; the Mean Absolute Percentage Error (MAPE), which expressed accuracy as a percentage; the Symmetric Mean Absolute Percentage Error (sMAPE), which corrected MAPE’s sensitivity to scale variations; and the Mean Absolute Scaled Error (MASE), a scale-independent metric suitable for comparing accuracy across distinct time series. Among these, sMAPE was the principal criterion for identifying the best-performing model for each province, owing to its robustness and interpretability. Once the optimal model for each province was determined, it was re-trained on the complete dataset covering January 2020 to December 2024, and forecasts were subsequently generated for the future period spanning January 2025 to December 2026. 2.7. Spatial Autocorrelation Analysis of Forecasts To detect potential spatial clusters of severe malaria incidence during the forecasted period, a spatial autocorrelation analysis was conducted using the average forecasted cases for 2026. A queen contiguity spatial weights matrix was first developed, in which provinces sharing a common boundary or a single point were considered neighbors. The matrix was row-standardized to ensure equal weighting of neighboring relationships, while provinces without contiguous borders were excluded from the analysis. The Local Moran’s I statistic was then computed to identify spatial clusters and outliers, quantifying the degree to which each province’s incidence was similar to that of its neighbors. The results included the Moran’s I statistic (Ii), its corresponding Z-score, and p-value, enabling the identification of statistically significant clusters. In addition, the Getis–Ord Gi* statistic was applied to detect high-incidence clusters (hotspots) and low-incidence clusters (cold spots), where a high positive Gi* indicated a hotspot and a low negative Gi* signified a cold spot. The outputs of these spatial analyses were visualized through a set of maps depicting the overall forecasted averages, Local Moran’s I value, Z-scores, p-values, Getis–Ord Gi* values, and their significance levels, thereby illustrating the spatial distribution and intensity of predicted malaria risk across provinces. 2.8. Software All data manipulation, analysis, and visualisation were performed using R statistical software (version 4.5.0) with various packages, including tidyverse for data handling, lubridate for date operations, zoo for time series manipulation, ggplot2 and patchwork for visualisation, sf and spdep for spatial analysis, forecast for time series modelling, Metrics for accuracy assessment, and corrplot for correlation visualisation. 2.9. Ethical Consideration Data used in this study were extracted from the national DHIS2 database of the Democratic Republic of Congo, with authorisation from the Ministry of Health. The data were aggregated and anonymised secondary data and did not contain any personal identifiers. Therefore, ethical approval was not required 3. RESULTS 3.1. Descriptive Statistics The analysis of severe malaria cases among pregnant women in the DRC from 2020 to 2024 revealed significant variations across provinces and over time, see Table 1 . The overall descriptive statistics for the entire region indicated a wide range of reported cases. The minimum monthly average across all provinces was 0, while the maximum reached 6,348 cases. The mean monthly incidence across all provinces and months was approximately 776.5 cases (SD = 997.1), with a median of 420 cases, suggesting a right-skewed distribution where a few provinces experience very high incidence. Table 1 Overall Descriptive Statistics for Severe Malaria Cases in Pregnant Women in DRC (2020–2024) Statistic Value Min Cases 0 Max Cases 6,348 Mean Cases 776.5 Median Cases 420 SD Cases 997.1 Note. Data source: DHIS2, DRC Ministry of Health. In December 2024, the latest month in the historical data, Kinshasa Province reported the highest number of cases (5,915), followed by Haut-Katanga (2,068), Kasaï-Central (2,066), and Haut-Lomami (2,216). Conversely, Tshuapa (291), Nord-Ubangi (256), and Bas-Uele (473) consistently recorded some of the lowest incidences. These highlights persistent disparities in the burden of severe malaria across the DRC. 3.2. Temporal Trends Temporal analysis revealed distinct patterns in severe malaria incidence. The overall monthly average for the DRC (Fig. 1 ) showed fluctuations but an increasing trend from early 2020, with a noticeable peak in late 2021 and a more pronounced upward trajectory from mid-2023 into 2024. This suggests a growing challenge in controlling severe malaria among pregnant women at the national level. This figure illustrates the national average monthly incidence of severe malaria among pregnant women. A general upward trend is observed, particularly from 2023 onwards, indicating an increasing burden over the study period. Data source: DHIS2, DRC Ministry of Health. Individual provincial trends (Fig. 2 ) exhibited considerable heterogeneity. While some provinces like Kinshasa and Maniema showed significant increases, others like Tshuapa and Nord-Ubangi maintained relatively low and stable levels. Provinces such as Haut-Lomami and Sud-Kivu displayed more volatile patterns with sharp increases and decreases. This faceted plot reveals diverse temporal patterns across provinces. Kinshasa and Maniema show clear increasing trends, while others like Tshuapa exhibit more stable, lower incidence. The variability underscores the need for province-specific interventions. Data source: DHIS2, DRC Ministry of Health. Seasonality analysis (Fig. 3 ) indicated that severe malaria incidence tends to be higher during certain months. Across all years, the months of May to December generally showed higher median cases compared to January to April, with peaks often observed in the latter half of the year. This seasonal pattern is consistent with typical malaria transmission cycles influenced by rainfall and temperature. The boxplot illustrates a clear seasonal pattern, with higher median cases observed from May to December, suggesting increased malaria transmission during these months. This information is crucial for timing seasonal interventions. Data source: DHIS2, DRC Ministry of Health. Annual trends (Fig. 4 ) further highlighted the evolving situation. The median incidence across all provinces showed a gradual increase from 2020 to 2024, with 2024 recording the highest median and maximum values. This reinforces the observation of an escalating burden of severe malaria among pregnant women in the DRC. This boxplot shows an increasing trend in severe malaria cases among pregnant women from 2020 to 2024, both in terms of median and maximum reported cases, indicating a worsening situation over the study period. Data source: DHIS2, DRC Ministry of Health. 3.3. Spatial Distribution The spatial distribution of severe malaria incidence revealed persistent high-burden areas. The overall average map (Fig. 5 ) clearly identified Kinshasa, Haut-Katanga, Kasaï-Central, and Maniema as provinces with consistently high average cases over the 2020–2024 period. Conversely, Nord-Ubangi, Bas-Uele, and Tshuapa generally exhibited lower average incidence. This map visually represents the average severe malaria cases per province over the entire study period. Darker shades indicate higher average incidence, clearly showing Kinshasa, Haut-Katanga, and Kasaï-Central as high-burden provinces. Data source: DHIS2, DRC Ministry of Health. The annual maps (Fig. 6 ) demonstrated that while the high-burden provinces remained largely consistent, there were subtle shifts in intensity and spread over the years. Some provinces showed increasing intensity of cases over time, reinforcing the temporal trends observed earlier. These annual maps show the evolution of severe malaria incidence across provinces. While high-burden areas generally persist, some provinces exhibit increasing intensity over the years, indicating dynamic spatial patterns. Data source: DHIS2, DRC Ministry of Health. 3.4. Inter-Provincial Correlation The correlation heatmap (Fig. 7 ) illustrated the degree of linear relationship in severe malaria incidence between different provinces. Strong positive correlations were observed among geographically proximate provinces, suggesting that factors influencing malaria transmission often operate at a regional level. For example, provinces within the Kasaï region (Kasaï, Kasaï-Central, Kasaï-Oriental) showed high positive correlations. Similarly, provinces in the eastern part of the DRC (e.g., Nord-Kivu, Sud-Kivu, Maniema) also exhibited strong positive correlations. Some negative correlations were also present, though less pronounced, indicating inverse trends in incidence between certain distant provinces. This heatmap displays the correlation coefficients between severe malaria incidence in different provinces. Darker blue indicates strong positive correlation, while darker red indicates strong negative correlation. High positive correlations are evident among neighboring provinces, suggesting shared epidemiological drivers. Data source: DHIS2, DRC Ministry of Health. 3.5. Forecasting Model Performance Five time series models (SARIMA, ETS, TBATS, ARNN_NNAR, ARFIMA) were evaluated for their ability to forecast severe malaria incidence. The ANN_MLP model was excluded due to consistent execution errors across all provinces. The accuracy metrics (RMSE, MAE, MAPE, sMAPE, MASE) were calculated for the validation period (last 12 months of 2024). Table 2 Summary of Model Accuracy Metrics (sMAPE) by Province Province SARIMA sMAPE ETS sMAPE TBATS sMAPE ARNN_NNAR sMAPE ARFIMA sMAPE Bas-Uele 0.386 0.373 0.377 0.386 0.471 Equateur 0.219 0.214 0.204 0.251 0.322 Haut-Katanga 0.200 0.157 0.156 0.202 0.126 Haut-Lomami 0.187 0.269 0.268 0.254 0.145 Haut-Uele 0.148 0.153 0.166 0.082 0.256 Ituri 0.077 0.069 0.088 0.091 0.090 Kasaï 0.213 0.200 0.202 0.263 0.185 Kasaï-Central 0.281 0.323 0.324 0.232 0.254 Kasaï-Oriental 0.169 0.165 0.184 0.374 0.175 Kinshasa 0.643 0.470 0.772 0.344 0.522 Kongo-Central 0.231 0.235 0.271 0.204 0.275 Kwango 0.472 0.331 0.337 0.311 0.421 Kwilu 0.345 0.282 0.282 0.271 0.346 Lomami 0.144 0.142 0.375 0.144 0.129 Lualaba 0.103 0.093 0.088 0.103 0.107 Maniema 0.149 0.108 0.112 0.181 0.085 Maï-Ndombe 0.429 0.429 0.431 0.434 0.429 Mongala 0.493 0.381 0.336 0.471 0.592 Nord-Kivu 0.264 0.256 0.301 0.302 0.405 Nord-Ubangi 0.295 0.337 0.367 0.294 0.378 Sankuru 0.093 0.110 0.155 0.139 0.213 Sud-Kivu 0.227 0.127 0.203 0.194 0.236 Sud-Ubangi 0.336 0.362 0.385 0.506 0.462 Tanganyika 0.285 0.136 0.133 0.274 0.604 Tshopo 0.096 0.113 0.108 0.104 0.174 Tshuapa 0.336 0.355 0.354 0.401 0.353 Note. Bold values indicate the best performing model (lowest sMAPE) for each province. The results in Table 2 indicate that no single model consistently outperformed others across all provinces. ETS, ARNN_NNAR, and ARFIMA models frequently emerged as the best performers, suggesting the diverse nature of time series patterns across different provinces. For instance, ETS was the most accurate for Bas-Uele, Ituri, Kasaï-Oriental, Nord-Kivu, and Sud-Kivu. ARNN_NNAR showed superior performance for Haut-Uele, Kasaï-Central, Kinshasa, Kongo-Central, Kwango, Kwilu, and Nord-Ubangi. ARFIMA was optimal for Haut-Katanga, Haut-Lomami, Kasaï, and Maniema. SARIMA and TBATS also performed best in several provinces, such as Sankuru, Sud-Ubangi, Tshopo, Tshuapa (SARIMA), and Equateur, Lualaba, Mongala, Tanganyika (TBATS). This highlights the importance of selecting models tailored to the specific characteristics of each provincial time series. Table 3 Best Model (by sMAPE) per Province Province Best Model sMAPE Bas-Uele ETS 0.373 Equateur TBATS 0.204 Haut-Katanga ARFIMA 0.126 Haut-Lomami ARFIMA 0.145 Haut-Uele ARNN_NNAR 0.082 Ituri ETS 0.069 Kasaï ARFIMA 0.185 Kasaï-Central ARNN_NNAR 0.232 Kasaï-Oriental ETS 0.165 Kinshasa ARNN_NNAR 0.344 Kongo-Central ARNN_NNAR 0.204 Kwango ARNN_NNAR 0.311 Kwilu ARNN_NNAR 0.271 Lomami ARFIMA 0.129 Lualaba TBATS 0.088 Maniema ARFIMA 0.085 Maï-Ndombe SARIMA 0.429 Mongala TBATS 0.336 Nord-Kivu ETS 0.256 Nord-Ubangi ARNN_NNAR 0.294 Sankuru SARIMA 0.093 Sud-Kivu ETS 0.127 Sud-Ubangi SARIMA 0.336 Tanganyika TBATS 0.133 Tshopo SARIMA 0.096 Tshuapa SARIMA 0.336 Note. sMAPE = Symmetric Mean Absolute Percentage Error. The heatmaps for sMAPE, RMSE, and MAE (Figs. 8 , 9 , 10 ) visually confirm the varied performance. Provinces with lower sMAPE values (darker purple/blue) indicate better model fit, while higher values (yellow) suggest poorer performance, see Table 3 . Kinshasa, for example, shows higher error metrics across all models, reflecting the greater volatility and magnitude of cases in this densely populated province. This heatmap visualizes the sMAPE values for each model across all provinces. Darker shades indicate lower sMAPE (better accuracy). It clearly shows that model performance varies significantly by province, with no single model being universally superior. Data source: DHIS2, DRC Ministry of Health. Similar to the sMAPE heatmap, this figure displays RMSE values. It reinforces the observation that provinces with higher case counts, like Kinshasa, tend to have higher absolute error metrics, regardless of the model. Data source: DHIS2, DRC Ministry of Health. MAE heatmap provides another perspective on absolute error. The patterns are consistent with RMSE and sMAPE, indicating that model accuracy is highly dependent on the specific provincial context. Data source: DHIS2, DRC Ministry of Health. 3.6. Forecasted Trends and Spatial Autocorrelation (2025–2026) The combined forecasts for 2025–2026, using the best-performing model for each province, are presented in Fig. 11 . These forecasts generally project a continuation of historical trends. Provinces that showed increasing trends in the historical period are forecasted to continue this upward trajectory, while those with stable or lower incidence are expected to maintain similar patterns. This suggests that without significant intervention changes, the burden of severe malaria among pregnant women is likely to persist or increase in many areas. This figure shows historical data (black) and future forecasts (red) for each province. The dashed line marks the transition from historical data to forecasts. Many provinces are projected to continue their historical trends, with some high-burden areas showing sustained or increasing incidence. Data source: DHIS2, DRC Ministry of Health. The spatial autocorrelation analysis of the average forecasted severe malaria cases for 2026 (Fig. 12 ) provided crucial insights into future high-risk areas. Overall Forecast (Avg 2026): The map of overall forecasted average cases for 2026 (Panel 1) largely mirrors the historical spatial distribution, with Kinshasa, Kasaï-Central, and Haut-Katanga projected to remain high-incidence areas. Local Moran's I and Z-score: The Local Moran's I map (Panel 2) and its Z-score map (Panel 3) identified significant spatial clustering. The Z-score map, in particular, highlighted areas of positive spatial autocorrelation (red) in the western and central parts of the DRC, indicating clusters of high incidences. Conversely, some areas showed negative spatial autocorrelation (blue), suggesting spatial outliers or areas with values dissimilar to their neighbors. P-value (Moran's I): Panel 4, the p-value map for Local Moran's I, showed significant clustering (p < = 0.05, dark green) predominantly in the western provinces, including Kongo-Central and parts of Kasaï, confirming the statistical significance of the observed spatial patterns. Getis-Ord Gi* and P-value (Gi*): The Getis-Ord Gi* map (Panel 5) and its p-value map (Panel 6) further refined the identification of hot and cold spots. Significant hot spots (high Gi* with p < = 0.05) were identified in the western and central provinces, particularly Kongo-Central, Kwango, and parts of Kasaï-Central. These areas are projected to experience a statistically significant clustering of high severe malaria incidence in 2026. No significant cold spots were identified at the 0.05 level, but some areas showed lower Gi* values, indicating a tendency towards lower incidence. This figure presents a multi-panel view of the spatial autocorrelation analysis for 2026 forecasts. Panel 1 shows the forecasted average cases. Panels 2–4 (Local Moran's I) and 5–6 (Getis-Ord Gi) identify significant hot spots (clusters of high incidences) in the western and central provinces, particularly Kongo-Central and Kwango, indicating areas requiring prioritized intervention. Data source: DHIS2, DRC Ministry of Health. 4. DISCUSSION This study provides one of the most comprehensive spatiotemporal analyses of severe malaria incidence among pregnant women in the Democratic Republic of Congo (DRC) between 2020 and 2024, with projections extending to 2026. The findings reveal significant spatial and temporal heterogeneity, an overall upward trend in disease burden, and clear evidence of spatial clustering, particularly in western and central provinces. These results underscore the persistence and, in some areas, intensification of malaria risk among pregnant women, highlighting the need for geographically targeted and seasonally timed interventions. The temporal analysis revealed a steadily increasing trend in severe malaria incidence among pregnant women from 2020 to 2024, with the most pronounced rise occurring between mid-2023 and 2024. This finding aligns with national malaria surveillance reports indicating stagnation or resurgence in malaria transmission in several high-burden African countries during the post-COVID-19 period ( 1 ). The increase may reflect multiple intersecting factors, including disruptions in health service delivery during the pandemic, reduced distribution of insecticide-treated nets (ITNs), interruptions in intermittent preventive treatment in pregnancy (IPTp), and declining efficacy of current vector control tools due to insecticide resistance ( 4 , 5 ). The seasonal patterns observed, higher incidence between May and December, are consistent with the climatic and ecological characteristics of the DRC. Malaria transmission in tropical regions such as the DRC is closely linked to rainfall and temperature, which influence mosquito breeding and parasite development cycles ( 3 ). The increased transmission during the rainy season underscores the need for enhanced surveillance and preventive measures timed to precede peak transmission months. Interventions such as intensified community sensitization, IPTp coverage expansion, and vector control campaigns should be strategically aligned with these seasonal cycles to maximize impact. Spatial analysis revealed persistent high-burden provinces, notably Kinshasa, Haut-Katanga, Kasaï-Central, and Maniema, throughout the study period. These findings are consistent with prior epidemiological assessments that identified similar provinces as endemic hotspots due to a combination of demographic density, ecological suitability, and limited access to effective prevention and treatment ( 6 , 7 ). For example, Kinshasa’s large urban population and peri-urban slums present favorable conditions for malaria transmission, despite being an area of relatively advanced healthcare infrastructure. Urban malaria transmission is often underestimated, yet evidence suggests that poor sanitation, informal settlements, and water storage practices can sustain local mosquito populations ( 8 ). In contrast, provinces such as Tshuapa, Nord-Ubangi, and Bas-Uele consistently exhibited lower incidence. These differences may reflect variations in ecological zones, vector species composition, and the success of local malaria control programs. However, low reported incidence might also indicate underdiagnosis or underreporting due to limited surveillance infrastructure in remote areas. Strengthening provincial data systems, particularly in low-resource settings, is therefore critical for obtaining an accurate national picture of malaria burden. The inter-provincial correlation analysis further demonstrated strong positive associations among neighboring provinces, especially within regional clusters such as the Kasaï and Kivu regions. This spatial coherence suggests that malaria drivers, such as climate patterns, cross-border population movement, and shared health system challenges, operate at a regional rather than strictly provincial level. Consequently, malaria control strategies should adopt a regional approach, fostering cross-provincial coordination and data sharing to address these shared determinants effectively. Forecasting results suggest that, without major programmatic changes, the upward trend in severe malaria incidence among pregnant women is likely to continue through 2026. Provinces already burdened by high incidence, such as Kinshasa, Kasaï-Central, and Haut-Katanga, are projected to maintain or even exacerbate their high transmission levels. The spatial autocorrelation and Getis-Ord Gi* analyses identified significant hot spots of projected high incidence in western and central provinces, particularly Kongo-Central, Kwango, and parts of Kasaï-Central. These findings have critical implications for prioritization within national malaria control efforts. Identifying these “hot spots” is particularly valuable for optimizing resource allocation in a resource-limited context like the DRC. Targeting interventions to statistically significant high-incidence clusters can improve cost-effectiveness and public health impact. Evidence from other malaria-endemic countries suggests that geographically focused interventions, such as reactive case detection, focal indoor residual spraying (IRS), and community-based IPTp distribution, are effective in reducing transmission within identified clusters ( 9 ). The absence of significant “cold spots” indicates that malaria remains widespread and endemic across the DRC, underscoring the need for sustained national-scale efforts alongside regional targeting. The comparative evaluation of forecasting models revealed that no single model performed best across all provinces, reflecting the complex, non-stationary nature of malaria incidence data. ETS, ARNN_NNAR, and ARFIMA models were most frequently identified as optimal, highlighting that both classical time series and machine learning-based approaches have context-specific strengths. ETS models, which capture trend and seasonal components adaptively, performed well in provinces with smoother seasonal variations such as Ituri and Sud-Kivu. In contrast, ARNN_NNAR models, which incorporate nonlinear relationships, were superior in provinces with more volatile or irregular patterns, such as Kinshasa and Kwango. The strong performance of ARFIMA models in multiple provinces suggests the presence of long-memory processes in malaria transmission dynamics, possibly linked to persistent climatic or behavioral factors influencing disease recurrence ( 10 , 11 ). These findings emphasize the importance of adopting a flexible, province-specific modeling strategy rather than applying a single forecasting framework nationwide. Future studies could explore hybrid or ensemble approaches that combine statistical and machine learning methods to further enhance predictive accuracy. Moreover, incorporating exogenous variables such as rainfall, temperature, and vector control coverage could refine model performance and improve the interpretability of forecasted trends. Implications for public health policy, practice and research The results underscore the urgent need for targeted and data-driven malaria policies in the Democratic Republic of Congo (DRC). The Ministry of Health, through the Programme National de Lutte contre le Paludisme (PNLP), should integrate spatiotemporal surveillance data into planning and resource allocation. Policies must prioritize high-burden provinces such as Kinshasa, Haut-Katanga, and Kasaï-Central for intensified prevention and control measures. Provincial health authorities, in collaboration with the National Institute for Biomedical Research (INRB) and the National Statistics Institute (INS), should establish decentralized monitoring systems that use real-time data for decision-making. International partners including the WHO, United Nations Children’s Fund (UNICEF), the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM), and the United States Agency for International Development (USAID) should support subnational stratification of interventions. These agencies can facilitate resource mobilization, technical capacity-building, and joint cross-border strategies to address regional transmission clusters. At the community level, engagement of local leaders, faith-based organizations, and women’s associations will be vital in sustaining preventive behavior and trust in malaria control initiatives. Operationally, the findings demand strengthening of preventive and clinical practices across the health system. The observed seasonal transmission peaks between May and December highlight the need for synchronized timing of insecticide-treated net (ITN) distribution, intermittent preventive treatment in pregnancy (IPTp), and community sensitization campaigns. The Directorate of Reproductive Health and provincial health directorates should work closely with district health management teams (DHMTs) to ensure complete IPTp coverage through antenatal care (ANC) platforms. Professional nursing and midwifery bodies, including the Ordre National des Infirmiers du Congo (ONIC), must play a central role in training frontline health workers on malaria-in-pregnancy management and surveillance reporting. Collaboration with non-governmental organizations (NGOs) such as Médecins Sans Frontières (MSF), Catholic Relief Services (CRS), and Malaria Consortium can enhance community-based case detection and health education in underserved areas. Private health facilities and pharmacies should also be integrated into national reporting and supply chains to reduce treatment delays and improve continuity of care. The study identifies significant gaps that future research should address. Research institutions such as the University of Kinshasa, University of Lubumbashi, and regional health research centres should collaborate with international partners like the London School of Hygiene and Tropical Medicine (LSHTM) and the Institut Pasteur to develop predictive models incorporating climatic, demographic, and socioeconomic determinants. The inclusion of environmental and meteorological agencies (e.g., Agence Nationale de Météorologie et de Télédétection par Satellite) is essential to integrate rainfall and temperature data into forecasting systems. Donor-funded initiatives such as the President’s Malaria Initiative (PMI) and the Global Technical Strategy for Malaria (GTS 2021–2030) can support operational research focusing on the cost-effectiveness of spatially targeted interventions. Engagement of civil society networks, maternal health advocates, and local universities in dissemination will strengthen knowledge translation and encourage uptake of evidence-based practices ( 12 ). Limitations and Future Research Directions While the present study provides valuable insights, several limitations warrant consideration. First, the analysis relies on routine surveillance data from the DHIS2 system, which, although comprehensive, may be subject to underreporting, data entry errors, or inconsistencies across provinces. The observed differences between low- and high-incidence provinces could partially reflect disparities in reporting completeness rather than true epidemiological differences. Strengthening data quality audits and integrating community-based reporting systems could enhance reliability in future analyses. Second, the study did not incorporate environmental or socioeconomic covariates, such as rainfall, temperature, population density, or poverty levels, which are known to influence malaria transmission dynamics ( 13 , 14 ). Including these variables in future spatiotemporal models could help elucidate causal pathways and improve the precision of forecasts. Third, while this study focused specifically on severe malaria cases, integrating data on uncomplicated malaria, treatment-seeking behavior, and vector control coverage would offer a more holistic understanding of malaria risk among pregnant women. Finally, although the forecasting models provided valuable projections up to 2026, these predictions assume the continuation of current intervention levels and transmission dynamics. Any major changes in vector control strategies, drug resistance patterns, or health system performance could substantially alter future outcomes. Continuous model updating and validation against real-time surveillance data will therefore be essential for maintaining predictive relevance ( 14 ). 5. CONCLUSION This study has successfully characterized the spatiotemporal variation of severe malaria incidence among pregnant women in the DRC from 2020 to 2024 and provided forecasts up to 2026. The findings reveal a persistent and increasing burden of severe malaria, with distinct geographical hotspots and clear seasonal patterns. Kinshasa, Haut-Katanga, and Kasaï-Central consistently emerge as high-incidence provinces, and spatial autocorrelation analysis of future forecasts identifies significant hot spots in western and central regions. The application of diverse time series models underscores the need for context-specific forecasting approaches. These insights are crucial for informing evidence-based public health policies, enabling the targeted allocation of resources, and optimizing the timing of interventions to protect pregnant women from the devastating effects of severe malaria in the DRC. Future research should integrate additional covariates and explore more advanced machine learning models to enhance predictive accuracy and deepen our understanding of malaria epidemiology. Abbreviations DRC Democratic Republic of Congo DHIS2 District Health Information Software 2 SARIMA Seasonal Autoregressive Integrated Moving Average ETS Exponential Smoothing State Space TBATS Trigonometric, Box-Cox, ARMA, Trend, Seasonal ARNN_NNAR Autoregressive Neural Network ARFIMA Autoregressive Fractionally Integrated Moving Average RMSE Root Mean Squared Error MAE Mean Absolute Error MAPE Mean Absolute Percentage Error sMAPE Symmetric Mean Absolute Percentage Error MASE Mean Absolute Scaled Error Declarations Acknowledgment s The authors would like to thank the direction of Medical Research Circle (MedReC) of Democratic Republic of the Congo for the realization of this present paper. Author contributions Conceptualization: Aymar AKILIMALI, Jones ONESIME and Michée Sanza KANDA , Data curation: Aymar AKILIMALI, Samson HANGI, Excellent RUGENDABANGA and Michée Sanza KANDA, Formal analysis: Aymar AKILIMALI, Abdisalam Hassan MUSE, Mukhtar Abdi HASSAN, Saralees NADARAJAH, Jones ONESIME and Innocent MUFUNGIZI, Investigation: Aymar AKILIMALI and Abel AIZEQUE, Methodology: Abdisalam Hassan MUSE, Isaac ISIKO, Mukhtar Abdi HASSAN, Samson HANGI, Saralees NADARAJAH, Project administration: Aymar AKILIMALI, Samson HANGI and Abel AIZEQUE, Resources: Michée Sanza KANDA, Isaac ISIKO, Jones ONESIME and Samson HANGI, Supervision: Amos Kipkorir LANGAT and Saralees NADARAJAH, Validation: Abel AIZEQUE, Innocent MUFUNGIZI, Elie KIHANDUKA, Christian TAGUE, Amidu ALHASSAN, Jones ONESIME and Isaac ISIKO, Visualization: Michée Sanza KANDA, Isaac ISIKO, Innocent MUFUNGIZI and Excellent RUGENDABANGA, Writing - original draft: Aymar AKILIMALI, Abdisalam Hassan MUSE, Abel AIZEQUE, Mukhtar Abdi HASSAN, Amidu ALHASSAN and Saralees NADARAJAH, Writing - review & editing ; All Authors , Final approval of manuscript ; All Authors. Ethics approval and consent to participate Data used in this study were extracted from the national DHIS2 database of Democratic Republic of Congo. The data were aggregated and anonymized secondary data and did not contain any personal identifiers. Therefore, ethical approval was not required. Consent for publication Not applicable. Funding The authors did not receive any financial support for this work. No funding has been received for the conduct of this study. Conflict of interest All authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability Not applicable. Supplementary information Not applicable. Use of artificial intelligence tools No artificial intelligence was used in generating the manuscript. Provenance and peer review Not commissioned, externally peer reviewed. Clinical trial number Not applicable. References World Health Organization. World Malaria Report 2023 . Geneva: WHO; 2023. Desai M, ter Kuile FO, Nosten F, McGready R, Asamoa K, Brabin B, et al. Epidemiology and burden of malaria in pregnancy. Lancet Infect Dis . 2007;7(2):93–104. Weiss DJ, Lucas TCD, Nguyen M, Nandi AK, Bisanzio D, Battle KE, et al. Mapping the global prevalence, incidence, and mortality of Plasmodium falciparum , 2000–2017: a spatial and temporal modelling study. Lancet . 2019;394(10195):322–331. Bhatt S, Weiss DJ, Mappin B, Dalrymple U, Cameron E, Bisanzio D, et al. Coverage and system efficiencies of insecticide-treated nets in Africa from 2000 to 2017. eLife . 2015;4:e09672. Gleave K, Lissenden N, Richardson M, Choi L, Ranson H. Association between insecticide resistance and malaria vector control effectiveness: a systematic review. Lancet Infect Dis . 2021;21(1):76–88. Ministry of Health, DRC. Annual Malaria Epidemiological Report 2023 . Kinshasa: PNLP; 2024. Bosco AB, Tchouassi DP, Gimonneau G, Sangbakembi-Ngounou C, Mavoko HM, Cohuet A, et al. Spatiotemporal heterogeneity of malaria transmission in Central Africa: implications for control. Parasit Vectors . 2022;15(1):472. Hay SI, Guerra CA, Tatem AJ, Noor AM, Snow RW. The global distribution and population at risk of malaria: past, present, and future. Lancet Infect Dis . 2004;4(6):327–336. Sturrock HJW, Hsiang MS, Cohen JM, Smith DL, Greenhouse B, Bousema T, et al. Targeting asymptomatic malaria infections: active surveillance in control and elimination. PLoS Med . 2013;10(6):e1001467. Box GE, Jenkins GM, Reinsel GC, Ljung GM. Time Series Analysis: Forecasting and Control . 5th ed. Hoboken: Wiley; 2015. UNICEF & WHO. DRC Malaria Programme Review 2023 . Geneva: UNICEF/WHO; 2023. World Health Organization. High burden to high impact: a targeted malaria response . Geneva: WHO; 2018. Tatem AJ, Smith DL. International population movements and regional Plasmodium falciparum malaria elimination strategies. Proc Natl Acad Sci USA . 2010;107(27):12222–12227. Gething PW, Patil AP, Smith DL, Guerra CA, Elyazar IR, Johnston GL, et al. Climate change and the global malaria burden: a spatial modeling analysis. Nature . 2010;465(7296):342–345. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8707200","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":585541561,"identity":"c4651d1c-db29-40dd-9584-b9ad87a4c774","order_by":0,"name":"Aymar Akilimali","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYPCCBB4+ZiD1wMAGSDI2HiBKCxtIS4JBGkhLA1FaGNgg1GEwF68W/hnZiZ8ratJk2Ni5Ez8kFJy3W9t+GGhLjU00Li0SN3I3S545lgN0GO9miQSD28nbziQCtRxLy23ApedG7gbJBrYKkJYNYC1mB4BaGBsO49QiD7TlZ8M/sJbNPxIMziWbnX+IX4vBjdxtko1tYIdtA9pywM7sBgFbDM+83WbZ2JcG1mKRYJCcYHYDaEsCHr/IHc/dfLPhW7I9P//ZzTc+/LGzNzuf/vDBhxob3N4XSEDlJ4JVJmCoQwL8B1D59vgUj4JRMApGwcgEAN5HYcaPYQ4bAAAAAElFTkSuQmCC","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":true,"prefix":"","firstName":"Aymar","middleName":"","lastName":"Akilimali","suffix":""},{"id":585541562,"identity":"907cf976-089f-450e-b588-7490b6a7c807","order_by":1,"name":"Jones Onesime","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Jones","middleName":"","lastName":"Onesime","suffix":""},{"id":585541563,"identity":"a842c67a-ac9e-4f38-84ee-2ffb0acecc58","order_by":2,"name":"Abdisalam Hassan MUSE","email":"","orcid":"","institution":"Amoud University","correspondingAuthor":false,"prefix":"","firstName":"Abdisalam","middleName":"Hassan","lastName":"MUSE","suffix":""},{"id":585541564,"identity":"de0a3e55-3417-45ec-8303-4ae5e6b120a8","order_by":3,"name":"Mukhtar Abdi HASSAN","email":"","orcid":"","institution":"University of Manchester","correspondingAuthor":false,"prefix":"","firstName":"Mukhtar","middleName":"Abdi","lastName":"HASSAN","suffix":""},{"id":585541565,"identity":"74e9cd8e-16ce-43ae-9416-2ce342b2f602","order_by":4,"name":"Saralees NADARAJAH","email":"","orcid":"","institution":"University of Manchester","correspondingAuthor":false,"prefix":"","firstName":"Saralees","middleName":"","lastName":"NADARAJAH","suffix":""},{"id":585541566,"identity":"ff2b63a7-0b12-410b-a80b-bd04130c10d0","order_by":5,"name":"Abel AIZEQUE","email":"","orcid":"","institution":"Catholic University of Mozambique","correspondingAuthor":false,"prefix":"","firstName":"Abel","middleName":"","lastName":"AIZEQUE","suffix":""},{"id":585541567,"identity":"ebe5e443-5011-4c17-89fc-f37cbb7889fb","order_by":6,"name":"Amidu ALHASSAN","email":"","orcid":"","institution":"University of Cape Coast","correspondingAuthor":false,"prefix":"","firstName":"Amidu","middleName":"","lastName":"ALHASSAN","suffix":""},{"id":585541568,"identity":"e651a604-8167-4f12-a83e-1829330417a4","order_by":7,"name":"Michée Sanza KANDA","email":"","orcid":"","institution":"Ministry of Public Health","correspondingAuthor":false,"prefix":"","firstName":"Michée","middleName":"Sanza","lastName":"KANDA","suffix":""},{"id":585541569,"identity":"54f437d3-19e4-4aca-9081-b99929f587ac","order_by":8,"name":"Isaac ISIKO","email":"","orcid":"","institution":"University of British Columbia","correspondingAuthor":false,"prefix":"","firstName":"Isaac","middleName":"","lastName":"ISIKO","suffix":""},{"id":585541571,"identity":"4446f012-d075-4797-889b-387fb6a7cd4e","order_by":9,"name":"Excellent RUGENDABANGA","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Excellent","middleName":"","lastName":"RUGENDABANGA","suffix":""},{"id":585541573,"identity":"3d4b8403-3f50-4e88-b3e6-b0cb0a3ba755","order_by":10,"name":"Innocent MUFUNGIZI","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Innocent","middleName":"","lastName":"MUFUNGIZI","suffix":""},{"id":585541575,"identity":"eb60e6f4-ff26-4827-9b95-ec328291fd82","order_by":11,"name":"Samson HANGI","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Samson","middleName":"","lastName":"HANGI","suffix":""},{"id":585541577,"identity":"c536a9ce-deaa-402a-84a1-2adac9f89883","order_by":12,"name":"Elie KIHANDUKA","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Elie","middleName":"","lastName":"KIHANDUKA","suffix":""},{"id":585541579,"identity":"242fa99f-f22a-49fb-ac30-55b3f0cfc0db","order_by":13,"name":"Christian TAGUE","email":"","orcid":"","institution":"Medical Research Circle (MedReC)","correspondingAuthor":false,"prefix":"","firstName":"Christian","middleName":"","lastName":"TAGUE","suffix":""},{"id":585541580,"identity":"93379161-15df-4a67-8f3e-5ba88b92b0a2","order_by":14,"name":"Antoine Osongo ONANGA","email":"","orcid":"","institution":"La Sapientia Catholic University","correspondingAuthor":false,"prefix":"","firstName":"Antoine","middleName":"Osongo","lastName":"ONANGA","suffix":""},{"id":585541581,"identity":"b0a62976-9552-4183-85c0-2bfa2bbb0c22","order_by":15,"name":"Amos Kipkorir LANGAT","email":"","orcid":"","institution":"Pan African University for Basic Sciences Technology and Innovation","correspondingAuthor":false,"prefix":"","firstName":"Amos","middleName":"Kipkorir","lastName":"LANGAT","suffix":""}],"badges":[],"createdAt":"2026-01-27 07:38:39","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8707200/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8707200/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102296784,"identity":"8f0a1668-4531-4ad0-88e3-dd1f06920ffa","added_by":"auto","created_at":"2026-02-10 10:21:41","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":44231,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMonthly Average Severe Malaria Cases in Pregnant Women in DRC (2020-2024)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/2335605b1eb0120b1f073689.jpg"},{"id":102297071,"identity":"edee5a90-96f2-492c-aada-a21001d2f16c","added_by":"auto","created_at":"2026-02-10 10:25:28","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":164381,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSevere Malaria Cases in Pregnant Women Trends in Individual DRC Provinces (2020-2024)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/62f3a38413c8f5a881e51d11.jpg"},{"id":102210323,"identity":"2c6282a3-78a8-4035-84b4-5eec969396a0","added_by":"auto","created_at":"2026-02-09 12:19:18","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":60593,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSevere Malaria Cases in Pregnant Women Distribution by Month in DRC (Seasonality)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/d609f6cdea933244bfecde0c.jpg"},{"id":102210325,"identity":"f4f51299-63b2-4075-b2d8-51dfda42a470","added_by":"auto","created_at":"2026-02-09 12:19:18","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":49299,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSevere Malaria Cases in Pregnant Women Distribution by Year in DRC\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/39a7e63c17aecb60f1347e2a.jpg"},{"id":102297149,"identity":"cec0fff7-f6eb-42f9-9479-947853a87cfa","added_by":"auto","created_at":"2026-02-10 10:26:08","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":77259,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverall Average Severe Malaria Cases in Pregnant Women in DRC (2020-2024)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/ea213e453082c949f91ff106.jpg"},{"id":102210327,"identity":"f1446e47-96c2-4bf0-bea3-7fcd21034543","added_by":"auto","created_at":"2026-02-09 12:19:18","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":110105,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnnual Average Severe Malaria Cases in Pregnant Women in DRC Provinces (2020-2024)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/e45cece3792f959f2ce1aa2a.jpg"},{"id":102210331,"identity":"1aad1f1c-e42f-4a32-8d18-fa2f67658df6","added_by":"auto","created_at":"2026-02-09 12:19:18","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":185079,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation Heatmap of Severe Malaria Cases in Pregnant Women in DRC Provinces\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/afd870d4aba5a99155a5832d.jpg"},{"id":102297134,"identity":"a0669c1f-7619-476d-84c3-09a7dbe0ddbe","added_by":"auto","created_at":"2026-02-10 10:26:01","extension":"jpg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":41098,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003esMAPE by Model and Province\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"8.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/39af7faa697241ddd5edf01b.jpg"},{"id":102296716,"identity":"75f11af9-bf40-4673-8360-dd9bba3f0c33","added_by":"auto","created_at":"2026-02-10 10:21:04","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":33704,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRMSE by Model and Province\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/efd64ebadfd1ca8e7f06b8f2.jpg"},{"id":102297043,"identity":"74011ce3-30a4-4d9f-836b-206c77127846","added_by":"auto","created_at":"2026-02-10 10:25:17","extension":"jpg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":35231,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMAE by Model and Province\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"10.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/0940e1be46c2562e47471b6b.jpg"},{"id":102210333,"identity":"0bfd2a24-e51d-445f-91c8-d2f189d7a271","added_by":"auto","created_at":"2026-02-09 12:19:19","extension":"jpg","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":184806,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSevere Malaria Cases in Pregnant Women Forecasts for DRC Provinces (2020-2026)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"11.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/81416b18958bd15be72c8f29.jpg"},{"id":102210328,"identity":"352a19f3-91ca-4746-8cad-1b7c9341cde1","added_by":"auto","created_at":"2026-02-09 12:19:18","extension":"jpg","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":107258,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpatial Autocorrelation Analysis of Forecasted Severe Malaria Cases in Pregnant Women in DRC Provinces (Avg 2026)\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"12.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/179ce71f04ad943d52c3100a.jpg"},{"id":102962537,"identity":"d8d6af34-19e0-4588-b421-841661a1b4bf","added_by":"auto","created_at":"2026-02-19 04:09:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2738649,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8707200/v1/53084b8b-e4c6-4cb1-b787-750409bc1a4b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Spatiotemporal Dynamics and Forecasting of Severe Malaria Incidence Among Pregnant Women in the Democratic Republic of Congo (2020-2024): A Retrospective Observational Study","fulltext":[{"header":"1. INTRODUCTION","content":"\u003cp\u003eMalaria remains one of the leading global public health challenges, particularly in low- and middle-income countries. The World Health Organization (WHO) (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) reported an estimated 249\u0026nbsp;million malaria cases and 608 000 deaths in 2022, with over 94% of these occurring in sub-Saharan Africa, where climatic and socioeconomic conditions facilitate persistent Plasmodium transmission. Within this context, the Democratic Republic of Congo (DRC) is among the six countries accounting for more than half of all global malaria cases, representing approximately 12% of the worldwide burden (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Malaria\u0026rsquo;s impact extends beyond morbidity and mortality to considerable economic losses, further straining fragile health systems and impeding progress towards SDG 3, which seeks to end the epidemics of malaria and other communicable diseases by 2030.\u003c/p\u003e \u003cp\u003ePregnant women remain one of the most vulnerable groups to malaria infection. Physiological immunosuppression during pregnancy increases susceptibility, and infection often leads to maternal anaemia, low birth weight, preterm delivery, stillbirth, and elevated maternal and neonatal mortality (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). In 2022, an estimated 35.4\u0026nbsp;million pregnancies were at risk of malaria infection across the African region, with approximately 12.7\u0026nbsp;million (36%) directly exposed (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). In the DRC, the prevalence of malaria infection among pregnant women has been reported as high as 63.3%, with maternal anaemia affecting 56.8% of those infected (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e). Such outcomes underscore the continuing public health significance of malaria in pregnancy and its intersection with maternal and child health priorities.\u003c/p\u003e \u003cp\u003eSeveral factors aggravate the situation in the DRC. Limited access to quality antenatal care, inconsistent uptake of intermittent preventive treatment in pregnancy (IPTp), sub-optimal distribution of insecticide-treated nets (ITNs), and increasing insecticide resistance constrain the effectiveness of control strategies (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e). Environmental variables such as rainfall, temperature, and humidity shape the abundance and distribution of Anopheles vectors, while sociocultural and economic disparities influence health-seeking behaviour and access to preventive measures. The country\u0026rsquo;s vast geography and ecological diversity lead to pronounced regional heterogeneity in malaria incidence, complicating national surveillance and response efforts (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite growing evidence on malaria epidemiology, there is a paucity of research examining the spatiotemporal dynamics and forecasting of severe malaria among pregnant women in the DRC. Existing studies are often limited to aggregate national estimates, lacking fine-scale temporal and provincial analyses necessary for precision public health planning. This knowledge gap impedes the ability of health authorities to identify high-risk provinces, anticipate seasonal peaks, and target interventions effectively. Against this backdrop, the present study aims to analyse the spatiotemporal variation in severe malaria incidence among pregnant women across DRC provinces from 2020 to 2024 and to forecast future trends up to 2026. Findings from this study will provide actionable evidence to strengthen geographically targeted and seasonally aligned malaria interventions, enhance health-system responsiveness, and support national progress towards achieving SDG 3.\u003c/p\u003e"},{"header":"2. METHODOLOGY","content":"\u003cp\u003eThis study employed a retrospective observational design to analyse the spatiotemporal variation and forecast the incidence of severe malaria among pregnant women in the Democratic Republic of Congo. The methodology involved several stages: data collection and preprocessing, descriptive analytics, spatiotemporal analysis, time series modeling and forecasting, and model evaluation.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1. Data Sources\u003c/h2\u003e \u003cp\u003eMonthly incidence data for \"A 1.5 Paludisme grave chez la femme enceinte\" (Severe Malaria in Pregnant Women) was obtained from the District Health Information Software 2 (DHIS2) platform, managed by the DRC Ministry of Health, for the years 2020 to 2024. These data were provided in five separate CSV files, each corresponding to a specific year. For spatial analysis and visualization, a shapefile (cod_admbnda_adm1_rgc_itos_20190911.shp) representing the administrative boundaries of DRC provinces was utilized.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2. Data Collection and Preprocessing\u003c/h2\u003e \u003cp\u003eThe initial step involved loading and merging the five annual CSV files into a single comprehensive dataset. This process required pivoting the monthly columns into a long format, creating a Month_Year column and a Severe_Malaria_Cases_PW column.\u003c/p\u003e \u003cp\u003eDate conversion was a critical preprocessing step. The Month_Year column, containing French month names (e.g., \"Janvier 2020\"), was parsed to extract the month name and year. A mapping was created to convert French month names to their English abbreviations, allowing for the creation of a standardized Date column (representing the first day of each month). Rows with unparseable dates were filtered out.\u003c/p\u003e \u003cp\u003eProvince name mapping was performed to ensure consistency between the CSV data and the shapefile. A custom function was developed to standardize province names from the CSV (e.g., \"bu Bas Uele Province\") to match the ADM1_FR attribute in the shapefile (e.g., \"Bas-Uele\"). This involved removing two-letter codes and \" Province\" suffixes, and standardizing diacritics and hyphens (e.g., \"Kasai\" to \"Kasa\u0026iuml;\", \"Maindombe\" to \"Ma\u0026iuml;-Ndombe\"). A verification step was included to identify and report any mismatches between the mapped CSV province names and the shapefile province names. Provinces not found in the shapefile were excluded from the analysis to maintain spatial integrity.\u003c/p\u003e \u003cp\u003eMissing values in the Severe_Malaria_Cases_PW variable were addressed by first converting the column to a numeric type, coercing non-numeric entries to NA. Commas within numeric strings (e.g., \"1,875\") were removed prior to conversion. A complete historical date range was generated for all provinces, and any missing monthly entries were filled using linear interpolation (zoo::na.approx). Remaining NA values, typically at the beginning or end of a time series where interpolation was not possible, were replaced with zero, assuming no reported cases implied zero incidence.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3. Descriptive Analytics\u003c/h2\u003e \u003cp\u003eDescriptive statistics were computed to summarize the overall incidence of severe malaria among pregnant women in the DRC. This included minimum, maximum, mean, median, and standard deviation of cases across the entire study period (2020\u0026ndash;2024). Monthly average cases for the entire DRC were calculated to observe general trends. Furthermore, provinces with the highest and lowest incidence in the latest available month (December 2024) were identified to highlight current hotspots and areas of lower burden.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4. Spatiotemporal Analysis\u003c/h2\u003e \u003cp\u003eSpatial distribution of severe malaria incidence was visualized through choropleth maps. An overall average map was generated by calculating the mean incidence for each province across the entire 2020\u0026ndash;2024 period. Additionally, annual average maps were created for each year from 2020 to 2024 to illustrate yearly spatial shifts. A consistent color scale was applied across all maps to facilitate comparison.\u003c/p\u003e \u003cp\u003eTemporal trends were analyzed using various visualizations. A time-series plot of the overall DRC average monthly incidence was generated to show national-level patterns. Individual time-series plots for each province, faceted for easy comparison, depicted provincial-level trends. Box plots were used to visualize seasonality (monthly distribution of cases across all years) and annual trends (yearly distribution of cases across all provinces).\u003c/p\u003e \u003cp\u003eInter-provincial correlations were assessed using a correlation heatmap. The data was pivoted to a wide format, with each province as a column, and the Pearson correlation coefficient was calculated between all pairs of provinces. Provinces with zero standard deviation in their case counts were excluded from this analysis to avoid undefined correlations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5. Time Series Modeling and Forecasting\u003c/h2\u003e \u003cp\u003eFor forecasting future incidence, five distinct time series models were applied to the monthly provincial data to capture both linear and non-linear temporal trends. The Seasonal Autoregressive Integrated Moving Average (SARIMA) model extended the conventional ARIMA framework to incorporate seasonal components, with the auto. Arima function from the forecast package automatically determining optimal model parameters. The Exponential Smoothing State Space (ETS) model, also from the forecast package, utilized exponential smoothing principles, assigning progressively smaller weights to older observations to account for evolving trends. The TBATS (Trigonometric, Box-Cox, ARMA, Trend, Seasonal) model was employed to manage complex seasonal patterns, non-linear trends, and Box-Cox transformations. Additionally, the Autoregressive Neural Network (ARNN_NNAR) model, implemented through the nnetar function, used a feed-forward neural network structure that incorporated lagged values of the series as predictors, enabling the modelling of non-linear temporal dependencies. Finally, the Autoregressive Fractionally Integrated Moving Average (ARFIMA) model was applied using the arfima function to accommodate fractional differencing and long-memory characteristics typical of persistent epidemiological data. The ANN_MLP model from the nnfor package was initially considered; however, execution errors occurred across all provinces due to incompatibility with the forecast: forecast method, which required a time-series (ts) object. Consequently, the ANN_MLP results were excluded from comparative analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6. Model Evaluation\u003c/h2\u003e \u003cp\u003eEach model\u0026rsquo;s performance was evaluated through a systematic validation procedure. For every province, the dataset from January 2020 to December 2023 served as the training set, while data from January to December 2024 were retained for testing and validation. The models were trained using the historical data and subsequently generated forecasts for the validation period to assess predictive accuracy. Model performance was measured using multiple statistical indicators: the Root Mean Squared Error (RMSE), which quantified the average magnitude of prediction errors; the Mean Absolute Error (MAE), which captured the mean absolute deviation without regard to direction; the Mean Absolute Percentage Error (MAPE), which expressed accuracy as a percentage; the Symmetric Mean Absolute Percentage Error (sMAPE), which corrected MAPE\u0026rsquo;s sensitivity to scale variations; and the Mean Absolute Scaled Error (MASE), a scale-independent metric suitable for comparing accuracy across distinct time series. Among these, sMAPE was the principal criterion for identifying the best-performing model for each province, owing to its robustness and interpretability. Once the optimal model for each province was determined, it was re-trained on the complete dataset covering January 2020 to December 2024, and forecasts were subsequently generated for the future period spanning January 2025 to December 2026.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7. Spatial Autocorrelation Analysis of Forecasts\u003c/h2\u003e \u003cp\u003eTo detect potential spatial clusters of severe malaria incidence during the forecasted period, a spatial autocorrelation analysis was conducted using the average forecasted cases for 2026. A queen contiguity spatial weights matrix was first developed, in which provinces sharing a common boundary or a single point were considered neighbors. The matrix was row-standardized to ensure equal weighting of neighboring relationships, while provinces without contiguous borders were excluded from the analysis. The Local Moran\u0026rsquo;s I statistic was then computed to identify spatial clusters and outliers, quantifying the degree to which each province\u0026rsquo;s incidence was similar to that of its neighbors. The results included the Moran\u0026rsquo;s I statistic (Ii), its corresponding Z-score, and p-value, enabling the identification of statistically significant clusters. In addition, the Getis\u0026ndash;Ord Gi* statistic was applied to detect high-incidence clusters (hotspots) and low-incidence clusters (cold spots), where a high positive Gi* indicated a hotspot and a low negative Gi* signified a cold spot. The outputs of these spatial analyses were visualized through a set of maps depicting the overall forecasted averages, Local Moran\u0026rsquo;s I value, Z-scores, p-values, Getis\u0026ndash;Ord Gi* values, and their significance levels, thereby illustrating the spatial distribution and intensity of predicted malaria risk across provinces.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8. Software\u003c/h2\u003e \u003cp\u003eAll data manipulation, analysis, and visualisation were performed using R statistical software (version 4.5.0) with various packages, including tidyverse for data handling, lubridate for date operations, zoo for time series manipulation, ggplot2 and patchwork for visualisation, sf and spdep for spatial analysis, forecast for time series modelling, Metrics for accuracy assessment, and corrplot for correlation visualisation.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.9. Ethical Consideration\u003c/h2\u003e \u003cp\u003eData used in this study were extracted from the national DHIS2 database of the Democratic Republic of Congo, with authorisation from the Ministry of Health. The data were aggregated and anonymised secondary data and did not contain any personal identifiers. Therefore, ethical approval was not required\u003c/p\u003e \u003c/div\u003e"},{"header":"3. RESULTS","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Descriptive Statistics\u003c/h2\u003e \u003cp\u003eThe analysis of severe malaria cases among pregnant women in the DRC from 2020 to 2024 revealed significant variations across provinces and over time, see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The overall descriptive statistics for the entire region indicated a wide range of reported cases. The minimum monthly average across all provinces was 0, while the maximum reached 6,348 cases. The mean monthly incidence across all provinces and months was approximately 776.5 cases (SD\u0026thinsp;=\u0026thinsp;997.1), with a median of 420 cases, suggesting a right-skewed distribution where a few provinces experience very high incidence.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eOverall Descriptive Statistics for Severe Malaria Cases in Pregnant Women in DRC (2020\u0026ndash;2024)\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStatistic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMin Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMax Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6,348\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e776.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMedian Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e420\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSD Cases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e997.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn December 2024, the latest month in the historical data, Kinshasa Province reported the highest number of cases (5,915), followed by Haut-Katanga (2,068), Kasa\u0026iuml;-Central (2,066), and Haut-Lomami (2,216). Conversely, Tshuapa (291), Nord-Ubangi (256), and Bas-Uele (473) consistently recorded some of the lowest incidences. These highlights persistent disparities in the burden of severe malaria across the DRC.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Temporal Trends\u003c/h2\u003e \u003cp\u003eTemporal analysis revealed distinct patterns in severe malaria incidence. The overall monthly average for the DRC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) showed fluctuations but an increasing trend from early 2020, with a noticeable peak in late 2021 and a more pronounced upward trajectory from mid-2023 into 2024. This suggests a growing challenge in controlling severe malaria among pregnant women at the national level.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis figure illustrates the national average monthly incidence of severe malaria among pregnant women. A general upward trend is observed, particularly from 2023 onwards, indicating an increasing burden over the study period. Data source: DHIS2, DRC Ministry of Health. Individual provincial trends (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) exhibited considerable heterogeneity. While some provinces like Kinshasa and Maniema showed significant increases, others like Tshuapa and Nord-Ubangi maintained relatively low and stable levels. Provinces such as Haut-Lomami and Sud-Kivu displayed more volatile patterns with sharp increases and decreases.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis faceted plot reveals diverse temporal patterns across provinces. Kinshasa and Maniema show clear increasing trends, while others like Tshuapa exhibit more stable, lower incidence. The variability underscores the need for province-specific interventions. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003eSeasonality analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) indicated that severe malaria incidence tends to be higher during certain months. Across all years, the months of May to December generally showed higher median cases compared to January to April, with peaks often observed in the latter half of the year. This seasonal pattern is consistent with typical malaria transmission cycles influenced by rainfall and temperature.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe boxplot illustrates a clear seasonal pattern, with higher median cases observed from May to December, suggesting increased malaria transmission during these months. This information is crucial for timing seasonal interventions. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003eAnnual trends (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) further highlighted the evolving situation. The median incidence across all provinces showed a gradual increase from 2020 to 2024, with 2024 recording the highest median and maximum values. This reinforces the observation of an escalating burden of severe malaria among pregnant women in the DRC.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis boxplot shows an increasing trend in severe malaria cases among pregnant women from 2020 to 2024, both in terms of median and maximum reported cases, indicating a worsening situation over the study period. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.3. Spatial Distribution\u003c/h2\u003e \u003cp\u003eThe spatial distribution of severe malaria incidence revealed persistent high-burden areas. The overall average map (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) clearly identified Kinshasa, Haut-Katanga, Kasa\u0026iuml;-Central, and Maniema as provinces with consistently high average cases over the 2020\u0026ndash;2024 period. Conversely, Nord-Ubangi, Bas-Uele, and Tshuapa generally exhibited lower average incidence.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis map visually represents the average severe malaria cases per province over the entire study period. Darker shades indicate higher average incidence, clearly showing Kinshasa, Haut-Katanga, and Kasa\u0026iuml;-Central as high-burden provinces. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003eThe annual maps (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e) demonstrated that while the high-burden provinces remained largely consistent, there were subtle shifts in intensity and spread over the years. Some provinces showed increasing intensity of cases over time, reinforcing the temporal trends observed earlier.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThese annual maps show the evolution of severe malaria incidence across provinces. While high-burden areas generally persist, some provinces exhibit increasing intensity over the years, indicating dynamic spatial patterns. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.4. Inter-Provincial Correlation\u003c/h2\u003e \u003cp\u003eThe correlation heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) illustrated the degree of linear relationship in severe malaria incidence between different provinces. Strong positive correlations were observed among geographically proximate provinces, suggesting that factors influencing malaria transmission often operate at a regional level. For example, provinces within the Kasa\u0026iuml; region (Kasa\u0026iuml;, Kasa\u0026iuml;-Central, Kasa\u0026iuml;-Oriental) showed high positive correlations. Similarly, provinces in the eastern part of the DRC (e.g., Nord-Kivu, Sud-Kivu, Maniema) also exhibited strong positive correlations. Some negative correlations were also present, though less pronounced, indicating inverse trends in incidence between certain distant provinces.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis heatmap displays the correlation coefficients between severe malaria incidence in different provinces. Darker blue indicates strong positive correlation, while darker red indicates strong negative correlation. High positive correlations are evident among neighboring provinces, suggesting shared epidemiological drivers. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.5. Forecasting Model Performance\u003c/h2\u003e \u003cp\u003eFive time series models (SARIMA, ETS, TBATS, ARNN_NNAR, ARFIMA) were evaluated for their ability to forecast severe malaria incidence. The ANN_MLP model was excluded due to consistent execution errors across all provinces. The accuracy metrics (RMSE, MAE, MAPE, sMAPE, MASE) were calculated for the validation period (last 12 months of 2024).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eSummary of Model Accuracy Metrics (sMAPE) by Province\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProvince\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA sMAPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eETS sMAPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTBATS sMAPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eARNN_NNAR sMAPE\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eARFIMA sMAPE\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBas-Uele\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.386\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0.373\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.377\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.386\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.471\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEquateur\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.219\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.214\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e0.204\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.251\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.322\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Katanga\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.157\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.202\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0.126\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Lomami\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.187\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.269\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.268\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.254\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0.145\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Uele\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.148\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.153\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.166\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.082\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.256\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIturi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.077\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0.069\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.091\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.090\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.213\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.202\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.263\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0.185\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;-Central\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.323\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.324\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.232\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.254\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;-Oriental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.169\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0.165\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.184\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.374\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.175\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKinshasa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.643\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.470\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.772\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.344\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.522\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKongo-Central\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.231\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.235\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.271\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.204\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.275\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKwango\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.472\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.331\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.337\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.311\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.421\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKwilu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.282\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.282\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.271\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.346\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLomami\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.144\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.142\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.375\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.144\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0.129\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLualaba\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.093\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e0.088\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.107\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eManiema\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.149\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.181\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0.085\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMa\u0026iuml;-Ndombe\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.429\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.429\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.431\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.434\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.429\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMongala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.493\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.381\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e0.336\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.471\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.592\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNord-Kivu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.264\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0.256\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.301\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.302\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.405\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNord-Ubangi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.295\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.337\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.367\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e0.294\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.378\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSankuru\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.093\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.155\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.139\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.213\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSud-Kivu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.227\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0.127\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.203\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.194\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.236\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSud-Ubangi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.336\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.385\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.506\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.462\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTanganyika\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.285\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.136\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e0.133\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.274\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.604\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTshopo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.096\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.113\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.104\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.174\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTshuapa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e0.336\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.355\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.354\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.401\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.353\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cem\u003eNote.\u003c/em\u003e Bold values indicate the best performing model (lowest sMAPE) for each province.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe results in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e indicate that no single model consistently outperformed others across all provinces. ETS, ARNN_NNAR, and ARFIMA models frequently emerged as the best performers, suggesting the diverse nature of time series patterns across different provinces. For instance, ETS was the most accurate for Bas-Uele, Ituri, Kasa\u0026iuml;-Oriental, Nord-Kivu, and Sud-Kivu. ARNN_NNAR showed superior performance for Haut-Uele, Kasa\u0026iuml;-Central, Kinshasa, Kongo-Central, Kwango, Kwilu, and Nord-Ubangi. ARFIMA was optimal for Haut-Katanga, Haut-Lomami, Kasa\u0026iuml;, and Maniema. SARIMA and TBATS also performed best in several provinces, such as Sankuru, Sud-Ubangi, Tshopo, Tshuapa (SARIMA), and Equateur, Lualaba, Mongala, Tanganyika (TBATS). This highlights the importance of selecting models tailored to the specific characteristics of each provincial time series.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cem\u003eBest Model (by sMAPE) per Province\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProvince\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBest Model\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003esMAPE\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBas-Uele\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eETS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.373\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEquateur\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTBATS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.204\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Katanga\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARFIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.126\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Lomami\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARFIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.145\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHaut-Uele\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.082\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIturi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eETS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.069\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARFIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.185\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;-Central\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.232\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKasa\u0026iuml;-Oriental\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eETS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.165\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKinshasa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.344\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKongo-Central\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.204\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKwango\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.311\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKwilu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.271\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLomami\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARFIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.129\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLualaba\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTBATS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.088\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eManiema\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARFIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.085\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMa\u0026iuml;-Ndombe\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.429\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMongala\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTBATS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.336\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNord-Kivu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eETS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.256\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNord-Ubangi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eARNN_NNAR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.294\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSankuru\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.093\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSud-Kivu\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eETS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.127\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSud-Ubangi\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.336\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTanganyika\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTBATS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.133\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTshopo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.096\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTshuapa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSARIMA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.336\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cem\u003eNote.\u003c/em\u003e sMAPE\u0026thinsp;=\u0026thinsp;Symmetric Mean Absolute Percentage Error.\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe heatmaps for sMAPE, RMSE, and MAE (Figs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, \u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e) visually confirm the varied performance. Provinces with lower sMAPE values (darker purple/blue) indicate better model fit, while higher values (yellow) suggest poorer performance, see Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. Kinshasa, for example, shows higher error metrics across all models, reflecting the greater volatility and magnitude of cases in this densely populated province.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis heatmap visualizes the sMAPE values for each model across all provinces. Darker shades indicate lower sMAPE (better accuracy). It clearly shows that model performance varies significantly by province, with no single model being universally superior. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSimilar to the sMAPE heatmap, this figure displays RMSE values. It reinforces the observation that provinces with higher case counts, like Kinshasa, tend to have higher absolute error metrics, regardless of the model. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMAE heatmap provides another perspective on absolute error. The patterns are consistent with RMSE and sMAPE, indicating that model accuracy is highly dependent on the specific provincial context. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e3.6. Forecasted Trends and Spatial Autocorrelation (2025\u0026ndash;2026)\u003c/h2\u003e \u003cp\u003eThe combined forecasts for 2025\u0026ndash;2026, using the best-performing model for each province, are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e. These forecasts generally project a continuation of historical trends. Provinces that showed increasing trends in the historical period are forecasted to continue this upward trajectory, while those with stable or lower incidence are expected to maintain similar patterns. This suggests that without significant intervention changes, the burden of severe malaria among pregnant women is likely to persist or increase in many areas.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis figure shows historical data (black) and future forecasts (red) for each province. The dashed line marks the transition from historical data to forecasts. Many provinces are projected to continue their historical trends, with some high-burden areas showing sustained or increasing incidence. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003cp\u003eThe spatial autocorrelation analysis of the average forecasted severe malaria cases for 2026 (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e) provided crucial insights into future high-risk areas.\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eOverall Forecast (Avg 2026): The map of overall forecasted average cases for 2026 (Panel 1) largely mirrors the historical spatial distribution, with Kinshasa, Kasa\u0026iuml;-Central, and Haut-Katanga projected to remain high-incidence areas.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLocal Moran's I and Z-score: The Local Moran's I map (Panel 2) and its Z-score map (Panel 3) identified significant spatial clustering. The Z-score map, in particular, highlighted areas of positive spatial autocorrelation (red) in the western and central parts of the DRC, indicating clusters of high incidences. Conversely, some areas showed negative spatial autocorrelation (blue), suggesting spatial outliers or areas with values dissimilar to their neighbors.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eP-value (Moran's I): Panel 4, the p-value map for Local Moran's I, showed significant clustering (p\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;0.05, dark green) predominantly in the western provinces, including Kongo-Central and parts of Kasa\u0026iuml;, confirming the statistical significance of the observed spatial patterns.\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eGetis-Ord Gi* and P-value (Gi*): The Getis-Ord Gi* map (Panel 5) and its p-value map (Panel 6) further refined the identification of hot and cold spots. Significant hot spots (high Gi* with p\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;0.05) were identified in the western and central provinces, particularly Kongo-Central, Kwango, and parts of Kasa\u0026iuml;-Central. These areas are projected to experience a statistically significant clustering of high severe malaria incidence in 2026. No significant cold spots were identified at the 0.05 level, but some areas showed lower Gi* values, indicating a tendency towards lower incidence.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThis figure presents a multi-panel view of the spatial autocorrelation analysis for 2026 forecasts. Panel 1 shows the forecasted average cases. Panels 2\u0026ndash;4 (Local Moran's I) and 5\u0026ndash;6 (Getis-Ord Gi) identify significant hot spots (clusters of high incidences) in the western and central provinces, particularly Kongo-Central and Kwango, indicating areas requiring prioritized intervention. Data source: DHIS2, DRC Ministry of Health.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. DISCUSSION","content":"\u003cp\u003eThis study provides one of the most comprehensive spatiotemporal analyses of severe malaria incidence among pregnant women in the Democratic Republic of Congo (DRC) between 2020 and 2024, with projections extending to 2026. The findings reveal significant spatial and temporal heterogeneity, an overall upward trend in disease burden, and clear evidence of spatial clustering, particularly in western and central provinces. These results underscore the persistence and, in some areas, intensification of malaria risk among pregnant women, highlighting the need for geographically targeted and seasonally timed interventions.\u003c/p\u003e \u003cp\u003eThe temporal analysis revealed a steadily increasing trend in severe malaria incidence among pregnant women from 2020 to 2024, with the most pronounced rise occurring between mid-2023 and 2024. This finding aligns with national malaria surveillance reports indicating stagnation or resurgence in malaria transmission in several high-burden African countries during the post-COVID-19 period (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). The increase may reflect multiple intersecting factors, including disruptions in health service delivery during the pandemic, reduced distribution of insecticide-treated nets (ITNs), interruptions in intermittent preventive treatment in pregnancy (IPTp), and declining efficacy of current vector control tools due to insecticide resistance (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe seasonal patterns observed, higher incidence between May and December, are consistent with the climatic and ecological characteristics of the DRC. Malaria transmission in tropical regions such as the DRC is closely linked to rainfall and temperature, which influence mosquito breeding and parasite development cycles (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). The increased transmission during the rainy season underscores the need for enhanced surveillance and preventive measures timed to precede peak transmission months. Interventions such as intensified community sensitization, IPTp coverage expansion, and vector control campaigns should be strategically aligned with these seasonal cycles to maximize impact.\u003c/p\u003e \u003cp\u003eSpatial analysis revealed persistent high-burden provinces, notably Kinshasa, Haut-Katanga, Kasa\u0026iuml;-Central, and Maniema, throughout the study period. These findings are consistent with prior epidemiological assessments that identified similar provinces as endemic hotspots due to a combination of demographic density, ecological suitability, and limited access to effective prevention and treatment (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). For example, Kinshasa\u0026rsquo;s large urban population and peri-urban slums present favorable conditions for malaria transmission, despite being an area of relatively advanced healthcare infrastructure. Urban malaria transmission is often underestimated, yet evidence suggests that poor sanitation, informal settlements, and water storage practices can sustain local mosquito populations (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). In contrast, provinces such as Tshuapa, Nord-Ubangi, and Bas-Uele consistently exhibited lower incidence. These differences may reflect variations in ecological zones, vector species composition, and the success of local malaria control programs. However, low reported incidence might also indicate underdiagnosis or underreporting due to limited surveillance infrastructure in remote areas. Strengthening provincial data systems, particularly in low-resource settings, is therefore critical for obtaining an accurate national picture of malaria burden.\u003c/p\u003e \u003cp\u003eThe inter-provincial correlation analysis further demonstrated strong positive associations among neighboring provinces, especially within regional clusters such as the Kasa\u0026iuml; and Kivu regions. This spatial coherence suggests that malaria drivers, such as climate patterns, cross-border population movement, and shared health system challenges, operate at a regional rather than strictly provincial level. Consequently, malaria control strategies should adopt a regional approach, fostering cross-provincial coordination and data sharing to address these shared determinants effectively.\u003c/p\u003e \u003cp\u003eForecasting results suggest that, without major programmatic changes, the upward trend in severe malaria incidence among pregnant women is likely to continue through 2026. Provinces already burdened by high incidence, such as Kinshasa, Kasa\u0026iuml;-Central, and Haut-Katanga, are projected to maintain or even exacerbate their high transmission levels. The spatial autocorrelation and Getis-Ord Gi* analyses identified significant hot spots of projected high incidence in western and central provinces, particularly Kongo-Central, Kwango, and parts of Kasa\u0026iuml;-Central. These findings have critical implications for prioritization within national malaria control efforts. Identifying these \u0026ldquo;hot spots\u0026rdquo; is particularly valuable for optimizing resource allocation in a resource-limited context like the DRC. Targeting interventions to statistically significant high-incidence clusters can improve cost-effectiveness and public health impact. Evidence from other malaria-endemic countries suggests that geographically focused interventions, such as reactive case detection, focal indoor residual spraying (IRS), and community-based IPTp distribution, are effective in reducing transmission within identified clusters (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e). The absence of significant \u0026ldquo;cold spots\u0026rdquo; indicates that malaria remains widespread and endemic across the DRC, underscoring the need for sustained national-scale efforts alongside regional targeting.\u003c/p\u003e \u003cp\u003eThe comparative evaluation of forecasting models revealed that no single model performed best across all provinces, reflecting the complex, non-stationary nature of malaria incidence data. ETS, ARNN_NNAR, and ARFIMA models were most frequently identified as optimal, highlighting that both classical time series and machine learning-based approaches have context-specific strengths. ETS models, which capture trend and seasonal components adaptively, performed well in provinces with smoother seasonal variations such as Ituri and Sud-Kivu. In contrast, ARNN_NNAR models, which incorporate nonlinear relationships, were superior in provinces with more volatile or irregular patterns, such as Kinshasa and Kwango. The strong performance of ARFIMA models in multiple provinces suggests the presence of long-memory processes in malaria transmission dynamics, possibly linked to persistent climatic or behavioral factors influencing disease recurrence (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThese findings emphasize the importance of adopting a flexible, province-specific modeling strategy rather than applying a single forecasting framework nationwide. Future studies could explore hybrid or ensemble approaches that combine statistical and machine learning methods to further enhance predictive accuracy. Moreover, incorporating exogenous variables such as rainfall, temperature, and vector control coverage could refine model performance and improve the interpretability of forecasted trends.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImplications for public health policy, practice and research\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe results underscore the urgent need for targeted and data-driven malaria policies in the Democratic Republic of Congo (DRC). The Ministry of Health, through the Programme National de Lutte contre le Paludisme (PNLP), should integrate spatiotemporal surveillance data into planning and resource allocation. Policies must prioritize high-burden provinces such as Kinshasa, Haut-Katanga, and Kasa\u0026iuml;-Central for intensified prevention and control measures. Provincial health authorities, in collaboration with the National Institute for Biomedical Research (INRB) and the National Statistics Institute (INS), should establish decentralized monitoring systems that use real-time data for decision-making. International partners including the WHO, United Nations Children\u0026rsquo;s Fund (UNICEF), the Global Fund to Fight AIDS, Tuberculosis and Malaria (GFATM), and the United States Agency for International Development (USAID) should support subnational stratification of interventions. These agencies can facilitate resource mobilization, technical capacity-building, and joint cross-border strategies to address regional transmission clusters. At the community level, engagement of local leaders, faith-based organizations, and women\u0026rsquo;s associations will be vital in sustaining preventive behavior and trust in malaria control initiatives.\u003c/p\u003e \u003cp\u003eOperationally, the findings demand strengthening of preventive and clinical practices across the health system. The observed seasonal transmission peaks between May and December highlight the need for synchronized timing of insecticide-treated net (ITN) distribution, intermittent preventive treatment in pregnancy (IPTp), and community sensitization campaigns. The Directorate of Reproductive Health and provincial health directorates should work closely with district health management teams (DHMTs) to ensure complete IPTp coverage through antenatal care (ANC) platforms. Professional nursing and midwifery bodies, including the Ordre National des Infirmiers du Congo (ONIC), must play a central role in training frontline health workers on malaria-in-pregnancy management and surveillance reporting. Collaboration with non-governmental organizations (NGOs) such as M\u0026eacute;decins Sans Fronti\u0026egrave;res (MSF), Catholic Relief Services (CRS), and Malaria Consortium can enhance community-based case detection and health education in underserved areas. Private health facilities and pharmacies should also be integrated into national reporting and supply chains to reduce treatment delays and improve continuity of care.\u003c/p\u003e \u003cp\u003eThe study identifies significant gaps that future research should address. Research institutions such as the University of Kinshasa, University of Lubumbashi, and regional health research centres should collaborate with international partners like the London School of Hygiene and Tropical Medicine (LSHTM) and the Institut Pasteur to develop predictive models incorporating climatic, demographic, and socioeconomic determinants. The inclusion of environmental and meteorological agencies (e.g., Agence Nationale de M\u0026eacute;t\u0026eacute;orologie et de T\u0026eacute;l\u0026eacute;d\u0026eacute;tection par Satellite) is essential to integrate rainfall and temperature data into forecasting systems. Donor-funded initiatives such as the President\u0026rsquo;s Malaria Initiative (PMI) and the Global Technical Strategy for Malaria (GTS 2021\u0026ndash;2030) can support operational research focusing on the cost-effectiveness of spatially targeted interventions. Engagement of civil society networks, maternal health advocates, and local universities in dissemination will strengthen knowledge translation and encourage uptake of evidence-based practices (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cb\u003eLimitations and Future Research Directions\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWhile the present study provides valuable insights, several limitations warrant consideration. First, the analysis relies on routine surveillance data from the DHIS2 system, which, although comprehensive, may be subject to underreporting, data entry errors, or inconsistencies across provinces. The observed differences between low- and high-incidence provinces could partially reflect disparities in reporting completeness rather than true epidemiological differences. Strengthening data quality audits and integrating community-based reporting systems could enhance reliability in future analyses. Second, the study did not incorporate environmental or socioeconomic covariates, such as rainfall, temperature, population density, or poverty levels, which are known to influence malaria transmission dynamics (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e). Including these variables in future spatiotemporal models could help elucidate causal pathways and improve the precision of forecasts. Third, while this study focused specifically on severe malaria cases, integrating data on uncomplicated malaria, treatment-seeking behavior, and vector control coverage would offer a more holistic understanding of malaria risk among pregnant women. Finally, although the forecasting models provided valuable projections up to 2026, these predictions assume the continuation of current intervention levels and transmission dynamics. Any major changes in vector control strategies, drug resistance patterns, or health system performance could substantially alter future outcomes. Continuous model updating and validation against real-time surveillance data will therefore be essential for maintaining predictive relevance (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e).\u003c/p\u003e"},{"header":"5. CONCLUSION","content":"\u003cp\u003eThis study has successfully characterized the spatiotemporal variation of severe malaria incidence among pregnant women in the DRC from 2020 to 2024 and provided forecasts up to 2026. The findings reveal a persistent and increasing burden of severe malaria, with distinct geographical hotspots and clear seasonal patterns. Kinshasa, Haut-Katanga, and Kasa\u0026iuml;-Central consistently emerge as high-incidence provinces, and spatial autocorrelation analysis of future forecasts identifies significant hot spots in western and central regions. The application of diverse time series models underscores the need for context-specific forecasting approaches. These insights are crucial for informing evidence-based public health policies, enabling the targeted allocation of resources, and optimizing the timing of interventions to protect pregnant women from the devastating effects of severe malaria in the DRC. Future research should integrate additional covariates and explore more advanced machine learning models to enhance predictive accuracy and deepen our understanding of malaria epidemiology.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"606\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDRC\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eDemocratic Republic of Congo\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDHIS2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eDistrict Health Information Software 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eSARIMA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eSeasonal Autoregressive Integrated Moving Average\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eETS\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eExponential Smoothing State Space\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTBATS\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eTrigonometric, Box-Cox, ARMA, Trend, Seasonal\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eARNN_NNAR\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eAutoregressive Neural Network\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eARFIMA\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eAutoregressive Fractionally Integrated Moving Average\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRMSE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eRoot Mean Squared Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMAE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eMean Absolute Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMAPE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eMean Absolute Percentage Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003esMAPE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eSymmetric Mean Absolute Percentage Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 120px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eMASE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 486px;\"\u003e\n \u003cp\u003eMean Absolute Scaled Error\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgment\u003c/strong\u003e\u003cstrong\u003es\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the direction of Medical Research Circle (MedReC) of Democratic Republic of the Congo for the realization of this present paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eConceptualization:\u003c/em\u003e\u003c/strong\u003e Aymar AKILIMALI, Jones ONESIME and Mich\u0026eacute;e Sanza KANDA\u003cstrong\u003e,\u003c/strong\u003e \u003cstrong\u003e\u003cem\u003eData curation:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAymar AKILIMALI, Samson HANGI, Excellent RUGENDABANGA and Mich\u0026eacute;e Sanza KANDA, \u003cstrong\u003e\u003cem\u003eFormal analysis:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAymar AKILIMALI, Abdisalam Hassan MUSE, Mukhtar Abdi HASSAN, Saralees NADARAJAH, Jones ONESIME and Innocent MUFUNGIZI, \u003cstrong\u003e\u003cem\u003eInvestigation:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAymar AKILIMALI and Abel AIZEQUE,\u003cstrong\u003e\u003cem\u003e\u0026nbsp;Methodology:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAbdisalam Hassan MUSE,\u0026nbsp;Isaac ISIKO,\u0026nbsp;Mukhtar Abdi HASSAN,\u0026nbsp;Samson HANGI,\u0026nbsp;Saralees NADARAJAH,\u003cstrong\u003e\u003cem\u003e\u0026nbsp;Project administration:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAymar AKILIMALI, Samson HANGI and Abel AIZEQUE, \u003cstrong\u003e\u003cem\u003eResources:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eMich\u0026eacute;e Sanza KANDA, Isaac ISIKO, Jones ONESIME and Samson HANGI, \u003cstrong\u003e\u003cem\u003eSupervision:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAmos Kipkorir LANGAT and Saralees NADARAJAH, \u003cstrong\u003e\u003cem\u003eValidation:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAbel AIZEQUE, Innocent MUFUNGIZI, Elie KIHANDUKA, Christian TAGUE, Amidu ALHASSAN, Jones ONESIME and Isaac ISIKO, \u003cstrong\u003e\u003cem\u003eVisualization:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eMich\u0026eacute;e Sanza KANDA,\u0026nbsp;Isaac ISIKO, Innocent MUFUNGIZI\u0026nbsp;and Excellent RUGENDABANGA,\u003cstrong\u003e\u003cem\u003e\u0026nbsp;Writing - original draft:\u003c/em\u003e\u003c/strong\u003e Aymar AKILIMALI, Abdisalam Hassan MUSE, Abel AIZEQUE, Mukhtar Abdi HASSAN, Amidu ALHASSAN and Saralees NADARAJAH, \u003cstrong\u003e\u003cem\u003eWriting - review \u0026amp; editing\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e;\u0026nbsp;\u003c/strong\u003eAll Authors\u003cstrong\u003e\u003cem\u003e, Final approval of manuscript\u003c/em\u003e\u003c/strong\u003e\u003cem\u003e;\u003c/em\u003e All Authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData used in this study were extracted from the national DHIS2 database of Democratic Republic of Congo. The data were aggregated and anonymized secondary data and did not contain any personal identifiers. Therefore, ethical approval was not required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors did not receive any financial support for this work. No funding has been received for the conduct of this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eUse of artificial intelligence tools\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo artificial intelligence was used in generating the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProvenance and peer review\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot commissioned, externally peer reviewed. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial number\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eWorld Health Organization. \u003cem\u003eWorld Malaria Report 2023\u003c/em\u003e. Geneva: WHO; 2023.\u003c/li\u003e\n\u003cli\u003eDesai M, ter Kuile FO, Nosten F, McGready R, Asamoa K, Brabin B, et al. Epidemiology and burden of malaria in pregnancy. \u003cem\u003eLancet Infect Dis\u003c/em\u003e. 2007;7(2):93\u0026ndash;104.\u003c/li\u003e\n\u003cli\u003eWeiss DJ, Lucas TCD, Nguyen M, Nandi AK, Bisanzio D, Battle KE, et al. Mapping the global prevalence, incidence, and mortality of \u003cem\u003ePlasmodium falciparum\u003c/em\u003e, 2000\u0026ndash;2017: a spatial and temporal modelling study. \u003cem\u003eLancet\u003c/em\u003e. 2019;394(10195):322\u0026ndash;331.\u003c/li\u003e\n\u003cli\u003eBhatt S, Weiss DJ, Mappin B, Dalrymple U, Cameron E, Bisanzio D, et al. Coverage and system efficiencies of insecticide-treated nets in Africa from 2000 to 2017. \u003cem\u003eeLife\u003c/em\u003e. 2015;4:e09672.\u003c/li\u003e\n\u003cli\u003eGleave K, Lissenden N, Richardson M, Choi L, Ranson H. Association between insecticide resistance and malaria vector control effectiveness: a systematic review. \u003cem\u003eLancet Infect Dis\u003c/em\u003e. 2021;21(1):76\u0026ndash;88.\u003c/li\u003e\n\u003cli\u003eMinistry of Health, DRC. \u003cem\u003eAnnual Malaria Epidemiological Report 2023\u003c/em\u003e. Kinshasa: PNLP; 2024.\u003c/li\u003e\n\u003cli\u003eBosco AB, Tchouassi DP, Gimonneau G, Sangbakembi-Ngounou C, Mavoko HM, Cohuet A, et al. Spatiotemporal heterogeneity of malaria transmission in Central Africa: implications for control. \u003cem\u003eParasit Vectors\u003c/em\u003e. 2022;15(1):472.\u003c/li\u003e\n\u003cli\u003eHay SI, Guerra CA, Tatem AJ, Noor AM, Snow RW. The global distribution and population at risk of malaria: past, present, and future. \u003cem\u003eLancet Infect Dis\u003c/em\u003e. 2004;4(6):327\u0026ndash;336.\u003c/li\u003e\n\u003cli\u003eSturrock HJW, Hsiang MS, Cohen JM, Smith DL, Greenhouse B, Bousema T, et al. Targeting asymptomatic malaria infections: active surveillance in control and elimination. \u003cem\u003ePLoS Med\u003c/em\u003e. 2013;10(6):e1001467.\u003c/li\u003e\n\u003cli\u003eBox GE, Jenkins GM, Reinsel GC, Ljung GM. \u003cem\u003eTime Series Analysis: Forecasting and Control\u003c/em\u003e. 5th ed. Hoboken: Wiley; 2015.\u003c/li\u003e\n\u003cli\u003eUNICEF \u0026amp; WHO. \u003cem\u003eDRC Malaria Programme Review 2023\u003c/em\u003e. Geneva: UNICEF/WHO; 2023.\u003c/li\u003e\n\u003cli\u003eWorld Health Organization. \u003cem\u003eHigh burden to high impact: a targeted malaria response\u003c/em\u003e. Geneva: WHO; 2018.\u003c/li\u003e\n\u003cli\u003eTatem AJ, Smith DL. International population movements and regional \u003cem\u003ePlasmodium falciparum\u003c/em\u003e malaria elimination strategies. \u003cem\u003eProc Natl Acad Sci USA\u003c/em\u003e. 2010;107(27):12222\u0026ndash;12227.\u003c/li\u003e\n\u003cli\u003eGething PW, Patil AP, Smith DL, Guerra CA, Elyazar IR, Johnston GL, et al. Climate change and the global malaria burden: a spatial modeling analysis. \u003cem\u003eNature\u003c/em\u003e. 2010;465(7296):342\u0026ndash;345.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Democratic Republic of Congo, Severe Malaria, Malaria Incidence, Pregnant Women, Retrospective Observational Study, Spatiotemporal Dynamics and Forecasting","lastPublishedDoi":"10.21203/rs.3.rs-8707200/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8707200/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eMalaria remains a major public health concern in the Democratic Republic of Congo (DRC). This retrospective observational study examined the spatiotemporal variation in severe malaria incidence among pregnant women across DRC provinces from 2020 to 2024 and projected future trends to 2026.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eMonthly data on severe malaria cases were obtained from the District Health Information Software 2 (DHIS2) managed by the DRC Ministry of Health. Data from 26 provinces were cleaned, harmonized, and analyzed using descriptive statistics, temporal trend visualizations, and spatial autocorrelation methods. Five forecasting models Seasonal Autoregressive Integrated Moving Average (SARIMA), Exponential Smoothing State Space (ETS), Trigonometric Box-Cox ARMA Trend Seasonal (TBATS), Autoregressive Neural Network (ARNN_NNAR), and Autoregressive Fractionally Integrated Moving Average (ARFIMA) were applied to predict future incidence. Model accuracy was assessed using Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Symmetric Mean Absolute Percentage Error (sMAPE), and Mean Absolute Scaled Error (MASE), with sMAPE used to identify the best-performing model for each province.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eConsiderable spatial and temporal heterogeneity was observed. Kinshasa, Haut-Katanga, and Kasa\u0026iuml;-Central reported persistently high incidence, while Nord-Ubangi and Mongala showed the lowest. Seasonal peaks occurred mainly between May and December. The ETS, ARNN_NNAR, and ARFIMA models demonstrated superior accuracy across different provinces, reflecting varied epidemic patterns. Forecasts for 2026 indicated persistent high-incidence clusters in western and central provinces, particularly Kongo-Central and Kwango.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThe study underscores significant spatial disparities and rising trends in severe malaria among pregnant women in the DRC. The findings provide critical evidence to guide geographically targeted, seasonally timed interventions and inform policy to strengthen malaria prevention and control.\u003c/p\u003e","manuscriptTitle":"Spatiotemporal Dynamics and Forecasting of Severe Malaria Incidence Among Pregnant Women in the Democratic Republic of Congo (2020-2024): A Retrospective Observational Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-09 12:19:01","doi":"10.21203/rs.3.rs-8707200/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"963dcb42-6c6e-4062-9b3e-b8902a4203a2","owner":[],"postedDate":"February 9th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-16T16:26:22+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-09 12:19:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8707200","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8707200","identity":"rs-8707200","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}