Epidemiological, Clinical, and Diagnostic Characteristics of a Large-Scale Upsurge of Dengue in the Rohingya Refugee Camps and Host Communities in Coxs Bazar, Bangladesh, 2021 to 2024: A Retrospective Study

Background: Dengue emerged as a new public health threat in the Rohingya refugee camps in Coxs Bazar, Bangladesh, in 2021 and expanded into large-scale upsurges in subsequent years. Evidence on dengue epidemiology and clinical presentation in protracted refugee settings remains limited, despite the need for stronger outbreak preparedness and case management in these contexts. Objectives: To describe the epidemiological, clinical, and diagnostic characteristics of the dengue upsurge among Rohingya refugees and surrounding host communities in Coxs Bazar, Bangladesh, and to identify predictors of inpatient admission and diagnostic positivity patterns. Methods: This retrospective observational study used anonymized surveillance data from the International Organization for Migration (IOM) dengue patient database. Rapid diagnostic test (RDT)-confirmed dengue cases identified across 36 IOM-supported health facilities in Ukhiya and Teknaf between 1 October 2021 and 31 December 2024 were included. Demographic, epidemiological, clinical, and laboratory variables were summarized using descriptive statistics. Weekly incidence was aggregated by epidemiological week and calendar year, and epidemic growth and decay phases were modelled using phase-specific Poisson regression. Multivariable logistic regression was used to identify predictors of inpatient admission and to examine associations between delay in presentation and RDT positivity patterns, adjusting for age and sex. Results: A total of 35,581 RDT-confirmed dengue cases were reported, of which 90.2% occurred among Rohingya refugees. The median age was 17 years (IQR 7-30), and 46.0% of cases were among children aged 0-14 years. Annual caseload increased from 1,011 in 2021 to 11,752 in 2022, 10,669 in 2023 and 12,149 in 2024, with seasonal peaks during the monsoon period and progressively later peaks and longer epidemic tails over time. Poisson models showed decreasing growth rates across years (r=0.449 in 2021 to r=0.091 in 2024) with increasing doubling times, while decay rates remained broadly comparable (halving time ~4.4-6.0 weeks). Overall, 8.0% of cases required inpatient admission, 1.3% were referred, and four deaths were reported (case fatality <0.1%). In multivariable analysis, inpatient admission was associated with older age ([&ge;]60 vs 0-14: aOR 2.31), delayed presentation (aOR 1.06 per day), refugee status (aOR 1.39), presence of any World Health Organization (WHO) warning sign (aOR 26.60), low systolic BP (aOR 2.84) and chronic co-morbidity (aOR 6.07). In addition, males had lower odds of admission than females (aOR 0.88). NS1 antigen alone was positive in 62.1% of cases, IgM alone in 33.6%, and dual positivity in 4.3%. Longer delay from symptom onset to presentation was strongly associated with IgM-only positivity compared with NS1-only positivity (adjusted models controlling for age and sex). Conclusion: Sustained dengue preparedness is required in Coxs Bazar, including strengthened surveillance, community-based early referral, targeted monitoring of high-risk groups, environmental vector control, and phase-appropriate use of NS1 and IgM/IgG diagnostics to reduce missed diagnoses and prevent progression to severe disease. These findings highlight the need for a policy shift from episodic outbreak response toward sustained dengue preparedness in humanitarian settings, including strengthened surveillance systems, integrated diagnostic strategies, community-based early referral, and coordinated vector control interventions.