Temporal trends in dengue incidence and mortality across surveillance case classifications: a 13-year analysis from Davao City, Philippines

Abstract

Introduction:

Dengue remains a major public health concern in the Philippines, where routine surveillance systems classify reported cases as suspected, probable or confirmed. Although these classifications are widely used to monitor disease burden, their influence on the interpretation of long-term epidemiological trends and mortality patterns remains unclear. This study examined the temporal trends in dengue incidence and mortality across surveillance case classifications in Davao City, Philippines, over a 13-year period.

Methods:

A retrospective observational study was conducted using dengue surveillance data collected from the Philippine Integrated Disease Surveillance and Response (PIDSR) system from January 2012 to December 2024. Reported dengue cases were classified as suspected, probable, or confirmed, according to the national surveillance definitions. Annual dengue incidence and mortality were calculated per 100,000 population, and temporal trends were assessed using the Mann–Kendall trend test.

Results:

A total of 81,075 dengue cases were reported during the study period, including 59,021 suspected, 20,559 probable, and 1,495 confirmed cases. Confirmed dengue incidence exhibited a significant increasing trend over time (p = 0.003), whereas probable (p = 0.855) and suspected dengue incidences (p = 0.059) showed no statistically significant temporal trends. Increasing trends in confirmed dengue incidence were consistently observed across congressional districts, age groups, and sexes. Dengue mortality incidence showed no significant temporal trends in suspected (p = 0.089), probable (p = 0.242), or confirmed cases (p = 0.282). Most deaths recorded among reported dengue notifications occurred among probable and suspected cases, and the highest cumulative case fatality rate was observed among probable cases (0.89%).

Conclusion:

Surveillance classifications captured distinct epidemiological patterns of dengue incidence and mortality. In low- and middle-income country (LMIC) settings, where surveillance systems operate under varying levels of diagnostic certainty, considering probable and suspected case classifications may provide additional insights into surveillance trends and mortality patterns beyond those obtained from laboratory-confirmed cases.

1 Introduction

Dengue is a mosquito-borne viral disease that remains a major public health challenge in tropical and subtropical regions worldwide. Clinical manifestations range from self-limiting febrile illness to severe disease characterized by plasma leakage, hemorrhage, organ dysfunction, and death (1, 2). In recent decades, the dengue burden has substantially increased, driven by urbanization, population growth, climate variability, and expanding mosquito habitats (3). In 2024 alone, more than 13 million dengue cases and over 8,700 dengue-related deaths were reported globally, highlighting the continuing public health significance of the disease (4).

The Philippines is among the countries most affected by dengue in Southeast Asia, where cyclical outbreaks continue to place considerable pressure on the healthcare and public health systems (5, 6). Previous studies have identified meteorological, demographic, and environmental factors associated with dengue transmission (7, 8); however, the interpretation of long-term dengue trends remains challenging because surveillance data are influenced by changes in case definitions, reporting practices, and laboratory confirmation.

Davao City, one of the largest and most rapidly urbanizing cities in the Philippines, experiences sustained dengue transmission throughout the year, with periodic increases during the rainy season. In the Philippines, dengue is a nationally notifiable disease, and cases that meet the national surveillance definitions are required to be reported through the Philippine Integrated Disease Surveillance and Response (PIDSR) system. In Davao City, the PIDSR receives notifications from registered public and private healthcare facilities and classifies them as suspected, probable, or confirmed according to established clinical, epidemiological, and laboratory criteria (9). These surveillance classifications play an important role in disease monitoring and reporting; however, their temporal patterns and relationship with mortality have not been comprehensively examined over extended surveillance periods.

These surveillance classifications represent different levels of diagnostic certainty and may capture different dimensions of the dengue burden. While laboratory-confirmed cases provide the most specific evidence of dengue infection, probable and suspected cases frequently constitute a substantial proportion of notifications reported through routine surveillance systems, particularly in low- and middle-income countries (LMICs), where access to confirmatory testing may vary across healthcare facilities. Consequently, temporal patterns derived from different surveillance classifications may yield different interpretations of the disease burden, transmission dynamics, and mortality. Understanding these differences is important for interpreting surveillance data and informing public health decision-making in dengue-endemic settings.

Although previous studies have described dengue epidemiology in the Philippines, including temporal trends, disease burden, and mortality patterns (10–12), recent surveillance-based investigations have characterized dengue trends using routinely collected surveillance data (8, 13). However, these studies primarily focused on describing surveillance trends and disease burden, with limited discussion on how different surveillance case classifications may influence the interpretation of long-term temporal trends and mortality. As suspected, probable, and confirmed cases represent different levels of diagnostic certainty, examining these surveillance classifications separately may provide additional insights into the interpretation of routine dengue surveillance data in endemic urban settings.

In this study, we analyzed 13 years (2012–2024) of dengue surveillance data from Davao City, comprising 81,075 notifications reported in the PIDSR system. We aimed to characterize temporal trends in dengue incidence across surveillance case classifications, describe their demographic and geographic distributions, and examine mortality patterns within the surveillance system. By examining long-term surveillance trends across multiple surveillance classifications, this study provides insights into the interpretation of dengue burden and mortality within routine surveillance systems in LMIC settings.

2 Methods

2.1 Study design

This retrospective observational study analyzed dengue surveillance data collected from Davao City, Philippines, between January 1, 2012, and December 31, 2024. This study aimed to characterize temporal trends in dengue notifications and mortality across surveillance case classifications (suspected, probable, and confirmed) and to describe demographic and geographic patterns within the surveillance system.

2.2 Setting

Davao City is a highly urbanized metropolitan area in the southeastern Philippines, with an estimated population of 1.9 million in 2024. The city is administratively divided into three congressional districts comprising 11 administrative districts. Dengue transmission occurs year-round, with increased activity typically observed during the rainy season (June–October). As a nationally notifiable disease in the Philippines, dengue is routinely reported through the Philippine Integrated Disease Surveillance and Response (PIDSR) system. In Davao City, surveillance is coordinated by the City Epidemiology and Surveillance Unit (CESU), which receives dengue notifications from all registered Disease Reporting Units (DRUs), including government and private hospitals, rural health units, city health centers, and other accredited public and private healthcare facilities. Healthcare providers are required to report dengue cases that meet the national surveillance case definitions to the local epidemiology unit, where reports are validated before being incorporated into the PIDSR database and forwarded through the national surveillance system. Consequently, the PIDSR database serves as the official citywide surveillance database for routine dengue monitoring and public health reports.

2.3 Dengue surveillance case definitions

Surveillance data were classified using a three-tier case classification system for dengue surveillance.

• Suspected case. A previously healthy person with acute febrile illness of 2–7 days duration with clinical signs and symptoms of dengue.

• Probable case. A suspected case with a positive dengue IgM antibody test result.

• Confirmed case. A suspected case with positive results for viral culture isolation, polymerase chain reaction (PCR), or dengue NS1 antigen test.

2.4 Data source and study population

The PIDSR database initially contained 81,192 dengue notifications recorded by the City Epidemiology and Surveillance Unit (CESU) between January 1, 2012, and December 31, 2024. During data cleaning, 117 records with substantial missing information (i.e., missing data for more than two key study variables) were excluded, resulting in a final analytical dataset of 81,075 dengue notifications. As PIDSR is the official routine surveillance system that consolidates dengue notifications from registered public and private DRUs throughout Davao City, the final dataset represents routine citywide dengue surveillance across the local healthcare system. Records with missing hospital admission information alone were retained in the analysis and recoded as “No Record” to distinguish missing values from reported admission categories.

2.5 Variables

The primary variables included surveillance case classification (suspected, probable, or confirmed), age group (<15 or ≥15 years), sex (male or female), congressional district of residence, hospital admission status, and clinical outcome (survived or died). Dengue case classification followed the standard surveillance definitions established by the DOH. Age was categorized according to the grouping used in routine PIDSR dengue surveillance.

2.6 Potential sources of bias

Potential sources of bias included underreporting, delayed reporting, and misclassification, which are inherent to routine passive surveillance systems. As dengue notifications were obtained from routinely collected surveillance data, reporting practices may have varied across facilities and over time. The use of standardized PIDSR case definitions and reporting procedures helped promote consistency across the reporting units.

2.7 Study size

All eligible dengue notifications reported during the study period were included, comprising 81,075 notifications. Because the analysis used complete surveillance data, no formal sample size calculations were performed.

2.8 Statistical analysis

Descriptive statistics were used to summarize dengue cases according to surveillance case classification, age group, sex, congressional district, hospital admission status, and clinical outcomes. Categorical variables are presented as frequencies and percentages.

Annual dengue incidence and mortality were calculated per 100,000 population using annual population estimates for Davao City. Temporal trends in dengue incidence and mortality were assessed using the Mann–Kendall trend test, a non-parametric method for detecting monotonic trends in time-series data. This approach was selected because it does not require assumptions of normality and is appropriate for evaluating temporal trends in epidemiological rates. Separate trend analyses were performed according to the surveillance case classification. Additional stratified analyses were conducted to evaluate trends in confirmed dengue incidence according to congressional district, age group, and sex.

To characterize mortality across surveillance case classifications, cumulative case fatality rates (CFRs) were calculated by dividing the total number of deaths recorded among reported dengue notifications by the total number of reported dengue cases within each surveillance classification and were expressed as percentages. Statistical significance was defined as a two-sided p-value of <0.05. All analyses were performed using R version 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria).

3 Results

3.1 Characteristics of dengue notifications by surveillance case classification

Between 2012 and 2024, 81,075 dengue notifications were reported in Davao City, including 59,021 suspected cases, 20,559 probable cases, and 1,495 confirmed cases (Table 1). Suspected cases accounted for most notifications throughout the surveillance period.

Across all surveillance case classifications, most notifications occurred among individuals younger than 15 years of age, representing 59.4%, 61.4%, and 62.7% of suspected, probable, and confirmed cases, respectively. Males comprised a slightly higher proportion of notifications than females across all classifications, accounting for 51.2%–52.2% of the reported notifications. The geographic distribution of the notifications varied across congressional districts. Congressional District I contributed the largest proportion of suspected (42.5%) and probable (43.6%) notifications, whereas confirmed notifications were more evenly distributed across the three congressional districts. Hospital admission was reported in more than 90% of notifications across all surveillance case classifications.

3.2 Temporal trends in dengue incidence by surveillance case classification

Citywide temporal trends in dengue incidence varied across surveillance case classifications (Figure 1; Supplementary Table S1). The confirmed dengue incidence increased significantly over the study period (Mann–Kendall test, p = 0.003). In contrast, the probable dengue incidence remained stable, with no significant temporal trend (p = 0.855). Suspected dengue incidence showed a decreasing trend over time, although this was not statistically significant (p = 0.059). Overall, the findings indicate distinct temporal patterns in the incidence of reported dengue cases across the surveillance case classifications during the study period.

Figure 1

3.3 District-specific trends in confirmed dengue incidence

Because confirmed dengue incidence was the only surveillance case classification that exhibited a significant citywide temporal trend, subsequent analyses focused on its geographic distribution across congressional districts (Figure 2; Supplementary Table S2). Significant increasing trends in confirmed dengue incidence were observed in all three congressional districts: Congressional District I (p = 0.005), Congressional District II (p = 0.016), and Congressional District III (p = 0.002). These findings indicate that the increasing citywide trend in confirmed dengue incidence was consistently observed in all congressional districts during the study period.

Figure 2

3.4 Age- and sex-specific trends in confirmed dengue incidence

Because the confirmed dengue incidence exhibited a significant citywide increasing trend, temporal patterns were further examined according to age group and sex (Figure 3; Supplementary Table S3). Significant increasing trends were observed among individuals younger than 15 years (p = 0.005) and those aged 15 years or older (p = 0.014). Similarly, the confirmed dengue incidence increased significantly among females (p = 0.005) and males (p = 0.003). The consistency of these trends across age and sex strata suggests that the observed increase in dengue incidence occurred broadly throughout the population rather than being restricted to specific demographic groups.

Figure 3

3.5 Temporal trends in dengue mortality incidence

No significant temporal trends were observed in dengue mortality incidence during the study period (Figure 4; Supplementary Table S4). When stratified by surveillance case classification, mortality incidence among suspected (p = 0.089), probable (p = 0.242), and confirmed (p = 0.282) cases showed no significant temporal trends in mortality. Although the mortality incidence among suspected cases demonstrated a decreasing tendency over time, the trend did not reach statistical significance. Despite the absence of significant temporal changes, the annual distribution of deaths recorded among reported dengue notifications differed across surveillance classifications, with probable and suspected cases accounting for the majority of deaths in most years, whereas confirmed cases consistently represented a smaller proportion of the reported dengue deaths.

Figure 4

3.6 Cumulative case fatality rates by surveillance case classification

To characterize mortality across surveillance case classifications, cumulative case fatality rates (CFRs) were calculated for the entire study period (Table 2). Probable cases exhibited the highest CFR (0.85%), followed by confirmed (0.67%) and suspected cases (0.26%). Despite the higher CFR observed among probable and confirmed cases, most deaths recorded among reported dengue notifications occurred among suspected and probable cases because these classifications accounted for most of the reported dengue cases during the study period.

To further characterize mortality patterns, cumulative case fatality rates were stratified according to the hospital admission status (Table 3). Among the admitted patients, CFRs were highest for probable notifications (0.89%), followed by confirmed (0.72%) and suspected notifications (0.29%). Among non-admitted patients, CFRs were low across all surveillance case classifications, and no deaths were recorded among non-admitted confirmed notifications. Across all surveillance case classifications, most deaths occurred among the admitted patients.

4 Discussion

Surveillance case classifications are fundamental components of dengue monitoring systems and are widely used to estimate disease burden, characterize epidemiological trends and guide public health responses (10). In this 13-year surveillance analysis, we observed distinct temporal patterns across dengue surveillance case classifications in Davao City. Confirmed dengue incidence increased significantly over time, whereas probable and suspected dengue incidences remained relatively stable. In contrast, mortality incidence did not exhibit significant temporal trends across any of the surveillance classifications. Notably, most deaths recorded among reported dengue notifications occurred among probable and suspected cases, with the highest cumulative case fatality rate observed among probable cases. These findings indicate that different surveillance classifications may capture distinct dimensions of dengue burden and severity and may therefore yield different interpretations of long-term epidemiological patterns.

The increasing incidence of laboratory-confirmed dengue was consistently observed across congressional districts, age groups, and sexes, suggesting that the observed pattern was widespread rather than restricted to specific demographic or geographic subpopulations. Similar increases in laboratory-confirmed dengue incidence have been reported in other dengue-endemic settings and have been attributed to a combination of changing transmission dynamics (11), expanded availability of diagnostic testing, improvements in laboratory capacity (8, 12), and enhanced surveillance practices (13). Toward the end of the study period, the Department of Health updated the national dengue surveillance guidance to formally recognize positive NS1 antigen test results as a criterion for laboratory-confirmed dengue. Because this update occurred only in 2024 (14), its influence on the long-term temporal trends observed in this 13-year analysis is likely limited, although it may have contributed to the distribution of surveillance classifications in the final year of this study. Consequently, temporal increases in confirmed dengue incidence should be interpreted in the context of evolving surveillance systems and diagnostic practices. At the same time, the consistency of the observed trend across demographic and geographic strata suggests that the increase is unlikely to be explained solely by localized reporting practices and may reflect broader epidemiological changes within the population.

The temporal trends observed in this study should be interpreted in the context of the COVID-19 pandemic. Consistent with reports from other dengue-endemic countries, dengue notifications declined during the pandemic, likely reflecting the combined effects of movement restrictions, school closures, altered healthcare-seeking behavior, and disruptions to routine surveillance and healthcare services. The subsequent increase in dengue incidence following the relaxation of these public health measures may represent a post-pandemic rebound driven by the resumption of population mobility, renewed opportunities for dengue virus transmission, and improved healthcare utilization and case detection. Although these factors may have influenced the magnitude of annual dengue incidence, the increasing trend in laboratory-confirmed dengue observed over the 13-year study period remained evident beyond the immediate post-pandemic period, suggesting that the observed pattern is unlikely to be explained solely by the effects of the COVID-19 pandemic.

An important finding of this study was the substantial contribution of probable cases to dengue-related mortality. Despite the absence of definitive laboratory confirmation, probable cases exhibited the highest cumulative case fatality rate and accounted for a considerable proportion of all deaths recorded among reported dengue notifications. This observation may be particularly relevant in dengue-endemic settings, where access to confirmatory diagnostic testing varies across healthcare facilities (15, 16). In such settings, probable case classifications may capture clinically compatible dengue cases that would otherwise remain underrepresented in laboratory-confirmed datasets. Furthermore, more than 90% of notifications across all surveillance classifications were hospitalized, suggesting that patients with suspected, probable, and confirmed dengue received comparable levels of clinical attention within the healthcare system. Together, these findings indicate that surveillance classifications may differ in diagnostic certainty but do not necessarily correspond to differences in clinical management or the risk of severe outcomes.

The findings should also be interpreted considering the potential classification and verification biases inherent in routine surveillance systems. Surveillance case classifications depend on the availability and use of laboratory confirmation, which may vary across healthcare facilities and over time. Patients with more severe clinical presentations may have been more likely to undergo confirmatory diagnostic testing and consequently be classified as probable or confirmed cases, whereas patients with milder illness or limited access to laboratory testing may have remained classified as suspected cases. These factors may have influenced both the observed temporal trends and the differences in case fatality rates across surveillance case classifications and should, therefore, be considered when interpreting the findings.

These findings are particularly relevant to low- and middle-income countries (LMICs), where dengue surveillance systems frequently operate under constraints related to diagnostic availability and laboratory capacity (17, 18). Although laboratory-confirmed cases provide the most specific estimate of the dengue burden, analyses restricted exclusively to confirmed cases may underestimate severe outcomes recorded within routine surveillance systems. In the present study, probable and suspected cases accounted for most deaths recorded among reported dengue notifications, despite stable incidence trends over time. These observations suggest that surveillance classifications beyond laboratory-confirmed dengue continue to provide meaningful epidemiological information regarding the disease burden and severe outcomes in resource-constrained settings. These findings illustrate how dengue surveillance systems in LMICs balance diagnostic certainty with case detection and burden estimation.

Several limitations should be considered when interpreting the findings of this study. First, the study relied on routinely collected passive surveillance data, which may be subject to underreporting, reporting delays, and misclassifications. Second, detailed clinical information, including disease severity, healthcare utilization, treatment practices, diagnostic testing, and medical record reviews of fatal cases, were unavailable. Consequently, we were unable to clinically validate deaths recorded within the surveillance system, particularly among probable and suspected case classifications, or evaluate alternative diagnoses and the extent to which individual deaths were directly attributable to dengue infection. Future studies incorporating detailed clinical record reviews and diagnostic verifications would strengthen the interpretation of mortality patterns observed in routine surveillance data. Finally, because this study relied exclusively on routinely collected surveillance notifications without an external reference standard or independent case ascertainment, surveillance performance measures, such as sensitivity and positive predictive value, could not be evaluated. Future studies incorporating laboratory verification, active surveillance, and independent case validation are warranted to assess these important surveillance system attributes. Nevertheless, the inclusion of all reported dengue cases over a 13-year period provides a comprehensive assessment of dengue surveillance patterns in a large urban endemic setting. These findings may be relevant to other dengue-endemic LMICs, where surveillance systems incorporate multiple case classifications to monitor disease burden and inform public health decision-making.

5 Conclusion

Surveillance case classifications capture different epidemiological patterns in routine dengue surveillance systems. In LMIC settings, where surveillance systems operate under varying levels of diagnostic certainty, considering probable and suspected case classifications may provide additional insights into surveillance trends and mortality patterns beyond those obtained from laboratory-confirmed cases alone. These findings highlight the importance of interpreting dengue surveillance data in the context of surveillance case classifications and varying levels of diagnostic certainty.

References

• 1

  PourzangiabadiMNajafiHFallahAGoudarziAPouladiI. Dengue virus: Etiology, epidemiology, pathobiology, and developments in diagnosis and control – a comprehensive review. Infection Genet Evol. (2025) 127:105710. doi: 10.1016/j.meegid.2024.105710

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  YangZ-SBauaAHemdanMAssavalapsakulWWangW-HLinC-Yet al. Dengue virus infection: A systematic review of pathogenesis, diagnosis and management. J Infect Public Health. (2025) 18:102982. doi: 10.1016/j.jiph.2025.102982

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3

  SchaeferTJPandaPKWolfordRW. Dengue Fever. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing (2025). Bookshelf ID: NBK430732. Available online at: https://pubmed.ncbi.nlm.nih.gov/28613483/.

  • Google Scholar

• 4

  eClinicalMedicine. Dengue as a growing global health concern. EClinicalMedicine. (2024) 77:102975. doi: 10.1016/j.eclinm.2024.102975

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  CokerRHunterBRudgeJLiveraniMHanvoravongchaiP. Emerging infectious diseases in southeast Asia: Regional challenges to control. Lancet. (2011) 377:599–609. doi: 10.1016/S0140-6736(10)62004-1

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  AhmadLGillBSulaimanLMuhamadNSinghSTeeKet al. Molecular epidemiology of dengue in Southeast Asia (SEA): Protocol of systematic review and meta-analysis. BMJ Open. (2025) 15:e088890. doi: 10.1136/bmjopen-2024-088890

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7

  BravoLRoqueVBrettJDizonRL’AzouM. Epidemiology of dengue disease in the Philippines (2000–2011): A systematic literature review. PloS NeglTrop Dis. (2014) 8:e3027. doi: 10.1371/journal.pntd.0003027

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  OlanaKPopromNSiewchaisakulPPunyapornwithayaVThongprachumA. Spatial distribution analysis and comparative forecasting of dengue resurgence in the Philippines (2025–2027): A nationwide study. Transbound Emerg Dis. (2025) 2025:7480710. doi: 10.1155/tbed/7480710

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9

  GallardoFde los ReyesVSucalditoMLigon-ImperioLPeñasJRebatoNet al. An assessment of the case notification system 16 months after Typhoon Haiyan in Region 8, the Philippines. Western Pacific Surveillance Response J. (2015) 6:71–5. doi: 10.5365/wpsar.2015.6.2.HYN_002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  Runge-RanzingerSHorstickOMarxMKroegerA. What does dengue disease surveillance contribute to predicting and detecting outbreaks and describing trends? Trop Med Int Health. (2008) 13:1022–41. doi: 10.1111/j.1365-3156.2008.02112.x

  • CrossRef
  • Google Scholar

• 11

  AbbasiE. The impact of climate change on travel-related vector-borne diseases: A case study on dengue virus transmission. Travel Med Infect Dis. (2025) 65:102841. doi: 10.1016/j.tmaid.2025.102841

  • CrossRef
  • Google Scholar

• 12

  AngLWLimCTKMaSTayJCutterJ. Adjustment for baseline level of dengue cases due to increased testing in Singapore. Online J Public Health Inform. (2015) 7. doi: 10.5210/ojphi.v7i1.5770

  • CrossRef
  • Google Scholar

• 13

  ZonunmawiiCKumarMSBoopathiKJeyashreeK. Investigation of laboratory-confirmed dengue outbreak in Bethlehem Vengthlang and Chite Veng, Aizawl, Mizoram, 2022. Indian J Public Health. (2026) 70:4. doi: 10.4103/ijph.ijph_654_25

  • CrossRef
  • Google Scholar

• 14

  BeattyMEStoneAFitzsimonsDWHannaJNLamSKVongSet al. Best practices in dengue surveillance: A report from the Asia-Pacific and Americas Dengue Prevention Boards. PLoS Negl Trop Dis. (2010) 4. doi: 10.1371/journal.pntd.0000890

  • CrossRef
  • Google Scholar

• 15

  Department of Health. DM 2024-0319: Heightened surveillance for leptospirosis and dengue. Manila: Department of Health (2024).

  • Google Scholar

• 16

  HaunerAAroni-SierraJMerinoXVillaCTorresFLagatieOet al. Assessing the diagnostic performance of clinical, serological and molecular approaches to improve dengue case detection in the Peruvian Amazon. PLoS Negl Trop Dis. (2026) 20. doi: 10.1371/journal.pntd.0013984

  • CrossRef
  • Google Scholar

• 17

  RaafatNLoganathanSMukakaMBlacksellSDMaudeRJ. Diagnostic accuracy of the WHO clinical definitions for dengue and implications for surveillance: A systematic review and meta-analysis. PLoS Negl Trop Dis. (2021) 15. doi: 10.1371/journal.pntd.0009359

  • CrossRef
  • Google Scholar

• 18

  SanapVSahaySMartinez-VegaRAgnihotriS. Informational challenges of dengue surveillance and response: A case study from India. Front Public Health. (2026) 14:1809584. doi: 10.3389/fpubh.2026.1809584

  • CrossRef
  • Google Scholar