Severe dengue in adults admitted to intensive care units on réunion island: clinical spectrum, outcomes, and predictors of mortality Authors: Charles Vidal, Quentin Soupin-Coulin, Willelm Doucet, Marjolaine Devineau, Zoé Blanc, Romane Crouzet, Anne-Cecile Morvan, Margot Caron, Julien Jabot, Nicolas Allou Published: 2025 ABSTRACT AbstractBackgroundSince 2017, Réunion Island has experienced successive dengue epidemics involving the co-circulation of DENV-1 and DENV-2 serotypes, leading to an increased number of severe cases. This study aimed to describe the clinical characteristics outcomes, and predictors of in-hospital mortality among adults admitted to intensive care units (ICUs) for severe dengue.MethodsWe conducted a multicenter retrospective study including all adults admitted to the five ICUs of Réunion Island between January 2017 and December 2022 with laboratory-confirmed severe dengue (SD), as defined by U.S. Centers for Disease Control and Prevention criteria. Clinical, biological, and outcome data were collected from electronic records. Significant variables in univariate analysis (p < 0.05) were entered into a multivariate logistic regression model to identify independent predictors of mortality.ResultsAmong 72.434 confirmed dengue cases, 2.817 required hospitalization and 137 SD were admitted to ICUs (1.8 per 1,000 infections). The median age was 63 years (IQR 50–70), and 62% were male. Circulatory failure occurred in 62% of cases including 12% with distributive shock and 9% with myocarditis; six patients required veno-arterial extra corporeal membrane oxygenation (VA-ECMO). Hepatic and renal failures were respectively observed in 50% and 59% of patients and neurological presentations occurred in 13 patients (9%). Co-infection were present in 33% of cases and fifty-two patients (38%) were admitted with an initial diagnosis other than dengue. Despite thrombocytopenia in 78% of cases, hemorrhagic events were uncommon. The in-hospital mortality rate was 25%. Catecholamine use (OR 4.71 [1.58–14.07], p = 0.005) and respiratory rate at admission (OR 1.10 [1.03–1.17], p = 0.004) were independent predictors of mortality.ConclusionsSD on Réunion Island exhibited a wide clinical spectrum in ICUs and was associated with substantial mortality, primarily due to multiple organ failure rather than hemorrhagic complications. Co-infections were common, and comorbidities likely contributed to unfavorable outcomes. Early identification and optimal management of shock and organ dysfunction remain essential to improving survival during future dengue outbreaks.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13054-025-05779-9.Keywords: Severe dengue, Mortality, Intensive care units, Shock, Indian ocean islands Introduction Dengue fever is a mosquito-borne arboviral disease caused by the dengue virus (DENV), a member of the Flavivirus genus, and transmitted primarily by Aedes aegypti and Aedes albopictus mosquitoes. Four distinct pathogenic serotypes (DENV-1 to DENV-4) have been well characterized. The disease is endemic in over 100 countries, particularly in tropical and subtropical regions such as Southeast Asia, Latin America, and sub-Saharan Africa. According to current global estimates, dengue affects approximately 400 million individuals each year, including about 100 million symptomatic infections and an estimated 20,000 deaths [1]. Among symptomatic dengue cases, approximately 5% progress to severe or critical forms requiring hospitalization. Hospitalization rates vary widely across endemic regions from about 2% to 10% depending on circulating serotypes, host immune status, and access to healthcare [1, 2].The ongoing expansion of Aedes vectors, driven by globalization, urbanization, and climate change, is projected to place nearly 3.9 billion people at risk of dengue virus exposure worldwide [3, 4]. The clinical presentation of dengue fever is highly variable and often nonspecific. Approximately 75% of infections remain asymptomatic. When symptoms occur, dengue usually manifests as a self-limiting febrile illness characterized by abrupt onset of high fever, headache, retro-orbital pain, myalgia, arthralgia, and anorexia, frequently associated with transient thrombocytopenia. Cutaneous manifestations such as a morbilliform rash, petechiae, or ecchymoses may also appear. The disease typically progresses through three phases: a febrile phase, a critical phase, and a recovery phase. The febrile phase lasts 2 to 7 days and is generally benign. The critical phase begins around the time of defervescence and is characterized by warning signs such as persistent vomiting, severe abdominal pain, mucosal bleeding, lethargy, or clinical fluid accumulation, which may herald plasma leakage, shock, or organ dysfunction. The recovery phase is marked by gradual reabsorption of extravascular fluid and clinical improvement [5, 6]. Severe dengue (SD), affecting approximately 5% of symptomatic cases, is defined by the presence of at least one of the following: severe plasma leakage, severe bleeding, or significant organ dysfunction commonly involving the central nervous system, liver, kidney or heart [5–7]. Severe cases predominantly occur during secondary infections involving a different serotype driven by antibody-dependent enhancement (ADE) and are estimated to affect around 500,000 individuals annually, with a reported in-hospital mortality rate reaching up to 30% [8–10]. Réunion Island, a French overseas territory located in the Indian Ocean, is home to an estimated 870,000 inhabitants. The island reported its first dengue outbreak in 1977–1978, primarily involving DENV-2, which affected approximately 30% of the population [11, 12]. Over the following decades, dengue transmission remained sporadic. However, since 2017, the island has experienced a marked resurgence of dengue activity, with increasing reports of severe cases linked to the co-circulation of DENV-1 and DENV-2 serotypes [12–14]. Recent epidemics on Réunion Island have revealed a wide spectrum of severe dengue manifestations, including hepatic failure, myocarditis, severe hemorrhagic syndromes, and central nervous system involvement [15–19]. The aim of this study was to describe clinical characteristics and outcomes of adult patients with Severe Dengue (SD) admitted to intensive care units on Réunion Island and to investigate the risk factors of in-hospital mortality in this population. Materials and methods Study sample selection This retrospective study included all adult patients diagnosed with SD and admitted to one of the five adult intensive care units (ICUs) on Réunion Island—Centre Hospitalier Universitaire (CHU) Félix Guyon, CHU Saint-Pierre, and Centre Hospitalier Ouest Réunion—between January 2017 and December 2022. The total number of confirmed cases and hospitalizations were obtained from the Regional Health Agency of Réunion Island (ARS Réunion). ICU cases were identified using administrative coding systems from each participating hospital. Definition of severe dengue and inclusion/exclusion criteria Patients were eligible for inclusion if they met the following criteria [1, 7, 12, 13, 19, 20]: Laboratory-confirmed dengue virus infection, diagnosed using at least one of the following methods: Detection of dengue viral RNA by reverse transcription polymerase chain reaction (RT-PCR) in a blood sample, detection of the non-structural protein 1 (NS1) antigen using enzyme-linked immunosorbent assay (ELISA) or rapid diagnostic tests (RDTs) and/or Evidence of a humoral immune response characterized by either seroconversion or a four-fold increase in total antibody titers, confirmed by one of the following methods: IgM/IgG ELISA, RDTs, hemagglutination inhibition assay, or plaque reduction neutralization test. Laboratory-confirmed dengue virus infection, diagnosed using at least one of the following methods: Detection of dengue viral RNA by reverse transcription polymerase chain reaction (RT-PCR) in a blood sample, detection of the non-structural protein 1 (NS1) antigen using enzyme-linked immunosorbent assay (ELISA) or rapid diagnostic tests (RDTs) and/or Evidence of a humoral immune response characterized by either seroconversion or a four-fold increase in total antibody titers, confirmed by one of the following methods: IgM/IgG ELISA, RDTs, hemagglutination inhibition assay, or plaque reduction neutralization test. Presence of at least two clinical symptoms within 21 days prior to ICU admission, including: fever, headache, skin rash, myalgia, arthralgia, abdominal pain, bleeding, and/or thrombocytopenia (platelet count < 150 G/L). Presence of at least two clinical symptoms within 21 days prior to ICU admission, including: fever, headache, skin rash, myalgia, arthralgia, abdominal pain, bleeding, and/or thrombocytopenia (platelet count < 150 G/L). Severe dengue requiring intensive care unit (ICU) admission for organ support as defined by the Surviving Sepsis Campaign guidelines. Severe dengue was defined according to the U.S. Centers for Disease Control and Prevention (CDC) 2015 case definition (CDC, National Notifiable Diseases Surveillance System, 2015), which includes the presence of at least one of the following: Severe plasma leakage leading to hypovolemic shock or severe respiratory distress or major hemorrhagic complications requiring medical or surgical intervention, including fluid resuscitation and blood transfusion or impaired level of consciousness and/or a diagnosis of encephalopathy, encephalitis, or meningitis or clinical presentation consistent with myocarditis. Severe dengue requiring intensive care unit (ICU) admission for organ support as defined by the Surviving Sepsis Campaign guidelines. Severe dengue was defined according to the U.S. Centers for Disease Control and Prevention (CDC) 2015 case definition (CDC, National Notifiable Diseases Surveillance System, 2015), which includes the presence of at least one of the following: Severe plasma leakage leading to hypovolemic shock or severe respiratory distress or major hemorrhagic complications requiring medical or surgical intervention, including fluid resuscitation and blood transfusion or impaired level of consciousness and/or a diagnosis of encephalopathy, encephalitis, or meningitis or clinical presentation consistent with myocarditis. Exclusion criteria were: age under 18 years, patients with no criteria of Severe Dengue in ICU and missing data. The criteria used to define organ failures are detailed in Appendix 1 [21–24]. Study endpoints Data were collected using the patient data management system’s routine documentation (Crossway®, Cegedim Santé, Paris, France). The variables include demographic data (age, sex, comorbidities), clinical presentation at ICU admission (vital parameters, organ failures), laboratory findings (hematological and biochemical parameters), and outcome measures (use of organ support, length of stay, and in-hospital mortality). The primary endpoint was in-hospital mortality. Secondary endpoints included determining the incidence of SD cases, describing the various clinical forms of SD and identifying risk factors associated with mortality. This observational study followed STROBE guidelines. Statistical analyses Results were expressed as median [25th–75th percentile] for continuous variables and as number (percentage) for categorical variables. Continuous variables were compared using the Mann–Whitney U test or unpaired Student’s t-test. Categorical variables were compared using the Chi-square test or Fisher’s exact test. Risk factors for in-hospital mortality were assessed by comparing clinical and demographic characteristics between survivors and non-survivors using univariate analyses. Risk factors found to be predictive of in-hospital mortality in the bivariate analysis with p < 0.05 were entered into a multivariate logistic regression analysis using backward selection with a criterion of p < 0.05. Collinearity between in dependent factors was investigated. When identified, the most clinically relevant factor was chosen for use in the multivariate model [26, 27]. All statistical analyses were performed using SAS software (SAS Institute Inc., Cary, NC, USA). Ethics An informational letter was sent to each participant or their legal representative (Appendix 2). Formal consent was not required due to the retrospective and observational nature of the study. The protocol was reviewed by the French Ethics Committee for Infectious and Tropical Diseases (CER-MIT), and the study was declared to the French Data Protection Authority in accordance with international recommendations for observational studies (Ref CER-BDX 2025 − 137, Appendix 3). Results Descriptive data Over the study period, 72,434 laboratory-confirmed dengue cases were reported on Réunion Island, including 2,817 hospitalizations. A total of 148 potential cases of SD requiring ICU admission were initially identified, of which 137 were finally included in the analysis. This corresponds to a median of 1.8 severe cases per 1,000 confirmed dengue infections and 5.1% of hospitalized patients progressing to SD requiring ICU care. The annual incidence of severe dengue admitted to the ICU varied between 1.1 and 57.7 cases per million inhabitants, depending on the study period (Fig. 1 and Appendix 4). Dengue virus serotyping was performed in 35 patients, identifying 20 cases of serotype 1 and 15 of serotype 2. Most serotype 2 infections occurred before 2020, whereas all serotype 1 infection were reported from 2020 onward (Appendix 5). Confirmed Dengue and Severe Dengue requiring ICU admission in Reunion Island between 2017 and 2022 The baseline characteristics and initial symptoms of the patients are presented in Table 1 . The median age of hospitalized patients was 63 years [IQR: 50–70], and the majority were male (62%). The median body mass index (BMI) was 25.6 [22.4–29.6] kg/m². Chronic Hypertension was present in 69 patients (50%), diabetes in 50 patients (36%), heart failure in 25 patients (18%), immunosuppression in 19 patients (14%), and chronic respiratory failure in 16 patients (12%). Table 1Baseline characteristics of survivors and non-survivors in patients with severe dengue admitted in ICU in reunion Island between 2017–2022CharacteristicsAll patients(n = 137 )Non Survivors(n = 34)Survivors(n = 103) p Age. years [IQR]63 [50 ;70]65 [58 ;70]61 [47.5 ;70]0.25Male. n (%)85 (62)24 (71)61 (59)0.24BMI. kg/m2 [IQR]25.6 [22.4 ;29.6]24.2 [20.5 ;29.2]25.9 [23 ;29.6]0.12Comorbidities Chronic heart disease. n (%)25 (18)5 (15)20 (19)0.54 Chronic lung disease. n (%)16 (12)8 (24)8 (8) 0.026 Chronic renal failure. n (%)25 (18)8 (24)17 (17)0.36 Arterial hypertension. n (%)69 (50)15 (44)54 (52)0.40 Diabetes. n (%)50 (36)13 (38)37 (36)0.81 Chronic alcoholic use. n (%)19 (14)6 (18)13 (13)0.57 Immunosuppression. n (%)19 (14)8 (24)11 (11)0.089 History of Dengue. n (%)5 (4)2 (6)3 (3)0.6Initial symptoms Fever > 38.5 °C. n (%)101 (74)20 (61)81 (79)0.039 Myalgia. n (%)77 (57)13 (39)64 (62)0.022 Headache. n (%)62 (46)9 (27)53 (51) 0.015 Abdominal pain. n (%)58 (43)13 (39)45 (44)0.66 Nausea/vomiting. n (%)46 (34)8 (24)38 (37)0.18 Arthralgia. n (%)13 (9)013 (13) 0.038 Retro-orbital pain. n (%)5 (4)05(5)0.34 Cutaneous rash. n (%)4 (3)04 (4)0.57 Jaundice. n (%)9 (7)3 (9)6 (6)0.69 Hepatosplenomegaly. n (%)2 (1)1 (3)1 (1)0.44Delai between 1st symptoms – diagnosis (days) [IQR]3 [1 ;5]3 [1 ;5]2 [1 ;4]0.31Diagnostic methods RDT. n (%)72 (54)16 (50)56 (55)0.63 RT-PCR. n (%)96 (72)24 (75)72 (71)0.63 IgM/IgG antibody. n (%)44 (34)11 (34)33 (32)0.83Serotype available. n (%)35 (26)12 (9)23 (17)0.11 Serotype 1. n (%)207 (35)13 (65)0.92 Serotype 2. n (%)155 (33)10 (66)0.92ICU : Intensive Care Unit. IQR = Interquartile range; n = Number; BMI = Body mass index; RDT = rapid diagnostic test; RT-PCR = reverse transcription polymerase chain reaction Baseline characteristics of survivors and non-survivors in patients with severe dengue admitted in ICU in reunion Island between 2017–2022 ICU : Intensive Care Unit. IQR = Interquartile range; n = Number; BMI = Body mass index; RDT = rapid diagnostic test; RT-PCR = reverse transcription polymerase chain reaction The most common initial symptoms before hospitalization were fever > 38.5 °C in 101 patients (74%), followed by myalgia in 77 patients (57%), headache in 62 patients (46%), abdominal pain in 56 patients (43%), and nausea/vomiting in 46 patients (34%). Regarding diagnostic confirmation, 96 patients (70%) had a positive RT-PCR, 29 patients (21%) tested positive using a rapid diagnostic test (RDT), and 12 patients (9%) had a positive serology. Admission in ICU Eighty-five patients (62%) presented with circulatory failure in the ICU, including 50 (36%) with hypovolemic shock, 17 (12%) with vasoplegic shock, 12 (9%) with a clinical picture of myocarditis, 6 (4%) with a hemorrhagic syndrome including 5 patients with gastrointestinal bleeding and one with postoperative bleeding. Forty-three patients (31%) showed signs of respiratory distress, and 81 patients (59%) had acute kidney injury. Thirteen patients (9%) presented with neurological manifestations of dengue, including 4 patients (3%) with confirmed meningoencephalitis, 3 (2%) patients for intracranial hemorrhage and 3 with acute polyradiculoneuropathy (PRN). Two patients presented acute disseminated encephalomyelitis (ADEM). Fifty-two patients were admitted with an initial diagnosis other than dengue, including 22 for exacerbations of preexisting chronic conditions. Co-infection were present in 33% of cases including 3 Covid-19 pneumonia. Antibiotic treatment was administered to 67% of the patients at admission. The patients’ physiological and biological parameters are presented respectively in Tables 2 and 3. Thrombocytopenia (< 150 G/L) were present in 104 patients (78%) at admission, with a median nadir of 40 [18–69] G/L occurring 5 [4–7] days after the onset of symptoms. Table 2Clinical presentation and organ dysfunction in ICU patients with severe dengue in Réunion Island between 2017 and 2022CharacteristicsAll patients(n = 137)Non Survivors’ group (n = 34)Survivors’ group(n = 103) p Scores SAPS II [IQR]41 [29;55]61 [49;77]35 [26;47] < 0.001 SOFA at admission [IQR]6 [4;9]9 [6;11]5 [3;8] < 0.001 Delay Between first symptoms – ICU4 [3;6]4 [2;6]4 [3;6]0.91Clinical presentation / organ failure in ICU admission Circulatory failure n (%)85 (62)24 (71)61 (59)0.24 Myocarditis / Cardiogenic Shock n (%)12 (9)5 (15)7 (7)0.17 Vasoplegic Shock, n (%)17 (12)8 (24)7 (7) 0.01 Non hemorragic hypovolemic Shock, n (%)50 (36)9 (26)43 (43)0.15 Hemorrhage (no intra-cranial), n (%)6 (4)2 (6)4 (4)0.64Respiratory distress, n (%)43 (31)16 (47)27 (26)0.023Hepatitis, n (%)68 (50)18 (52)50 (49)0.70 Mild (3-5UNL), n (%)20 (15)6 (18)14 (14)0.58 Moderate (5–10 UNL), n (%)12 (9)2 (6)10 (10)0.58 Severe (> 10 UNL), n (%)36 (26)10 (29)26 (25)0.65Acute Kidney Injury, n (%)81 (59)25 (73)56 (54)0.069 Kdigo 1, n (%)29 (21)8 (24)21 (20)0.7 Kdigo 2, n (%)18 (13)7 (21)11 (11)0.15 Kdigo 3, n (%)35 (24)10 (29)24 (23)0.47Neurological presentation, n (%)13 (9)5 (15)8 (8)0.31 Meningo-encephalitis, n (%)4 (3)1 (3)3 (3)1 Intra-cranial Hemorrhage n (%)3 (2)3 (9)0 0.014 Acute polyradiculoneuropathy, n (%)3 (2)03 (3)1Admitted in ICU for another disease, n (%)52 (38)16 (47)36 (35)0.23Exacerbation of previous complaints, n (%)22 (16)8 (24)14 (14)0.17Co-infection, n (%)43 (33)14 (42)29 (30)0,19 Pneumonia, n (%)23 (17)10 (29)13 (13)0.08 Urinary sepsis, n (%)8 (6)1(3)7 (7)0.67Antibiotic therapy at admission, n (%)85 (62)27 (82)58 (56) 0.01 Physiological parameters in ICU admission Heart Rate (beats per minute) [IQR]97 [75 ;115]105 [85 ;126]93 [72 ;111] 0.046 Systolic Arterial Pressure, mmHg [IQR]120 [97 ;137]114 [92 ;144]120 [100 ;137]0.38 Diastolic Arterial Pressure, mmHg [IQR]70 [58 ;80]63 [54;80]70 [59;80]0.069 Mean Arterial Pressure, mmHg [IQR]85 [72 ;97]74 [67 ;95]87 [74 ;97]0.052 Respiratory Rate (breath per minute) [IQR]25 [20 ;30]34 [28 ;40]21 [20 ;28] < 0.001 SpO2 (%) [IQR]98 [95 ;99]98 [93 ;100]98 [96 ;99]0.59 Glasgow score [IQR]15 [14 ;15]14 [10 ;15]15 [14 ;15]0.11ICU: Intensive Care Unit, IQR = Interquartile range- n = Number, SAPS : Simplified Acute Physiology Score, SOFA : Sequential Organ Failure Assessment, Kdigo = Kidney Disease: Improving Global outcomes Clinical presentation and organ dysfunction in ICU patients with severe dengue in Réunion Island between 2017 and 2022 ICU: Intensive Care Unit, IQR = Interquartile range- n = Number, SAPS : Simplified Acute Physiology Score, SOFA : Sequential Organ Failure Assessment, Kdigo = Kidney Disease: Improving Global outcomes Table 3Laboratory parameters of patients with severe dengue at ICU admission in reunion Island between 2017 and 2022CharacteristicsAll patients(n = 137)Non Survivors’ group(n = 34)Survivors’ group(n = 103) p Laboratory values Leukocytes G/L [IQR]6.5 [4.4;9.4]7.5 [5.6;12.0]6.2 [3.6;8.8] 0.042 Lymphocytes G/L [IQR]0.8 [0.5;1.5]0.7 [0.6;1.0]0.9 [0.5;1.5]0.21 Lymphopenia < 1 G/L n (%)67 (59)19 (70)48 (55)0.16 Hemoglobin, g/dL [IQR]13.8 [11.8;15.3]13.8 [120;152]13.8 [12.0;15.2]0.15 Platelet count, G/L [IQR]87 [24;144]81 [17;163]90 [33;142]0.92 Thrombopenia < 150 G/L, n (%)104 (78)9 (74)79 (77)0.81 Nadir Thrombopenia G/L [IQR]40 [18;69]32 [13;72]42 [20_72]0.051 Nadir Thrombopenia, days [IQR]5 [4;7]5 [4;7]6 [4;7]0.44 Prothrombin Time, % [IQR]83 [67;96]74 [59;84]86 [72;99]< 0.01 APTT ratio [IQR]1.28 [1.06;1.56]1.31[1.13;1.65]1.25 [1.05;1.45]0.098 Fibrinogen, g/L [IQR]3.5 [2.7;4.4]3.4 [2.0;5.0]3.5 [2.7;4.3]0.61 Natremia, mmol/L [IQR]135 [131;138]136 [131;139]135 [131;138]0.56 Kaliemia, mmol/L [IQR]4.1 [3.7;5.0]4.6 [4.0;5.2]4.0 [3.6;4.7] 0.036 Serum creatinine, µmol/L [IQR]149 [84;302]192 [103;308]146 [84;294]0.5 Uremia, mmol/L[IQR]9.3 [5.4;16.9]10.1 [6.6;16.8]8.5 [5.1;18.0]0.74 ALT UI/L, [IQR]70 [35;222]69 [38;453]70 [34;210]0.32 AST UI/L, [IQR]152 [72;536]180 [78;1052]147 [71;490]0.35 GGT UI/L, [IQR]81 [38;196]72 [39;164]85 [37;202]0.39 ALP, UI/L, [IQR]77 [61;118]74 [64;152]78 [64;110]0.25 Total bilirubinemia mg/L [IQR]0.64 [0.47 ;1.24]0.70 [0.52 ;1.64]0.58 [0.47 ;1.12]0.7 Lipasemia UI/L [IQR]65 [33;149]43 [37;99]78 [33;162]0.47 Serum Creatinine Kinase, UI/L[IQR]377 [164;1242]400 [148;2400]340 [169;1070]0.41 Serum Troponin, ng/L [IQR]54 [21;149]87 [29;361]54 [19;138]0.081 Reactive Protein C, mg/L [IQR]22 [8;61]42 [7;145]18 [8;41]0.054 Procalcitonine ng/mL [IQR]1.11 [0.34;6.47]1.08 [0.32;9.27]1.15 [0.39;5.64]0.7Blood Gases at ICU admission pH, [IQR]7.39 [7.26;7.45]7.31 [7.17;7.46]7.4 [7.27;7.45] 0.031 PaO2, mmHg [IQR]87 [24;144]84 [69;122]88 [72;102]0.53 PaCO2, mmHg [IQR]31 [26;39]30 [26;46]32 [26;39]0.78 HCO3− mmol/L [IGR]20 [14;25]21 [12;23]20 [15;25]0.39 Base Exces mmol/L [IGR]4.0 [;10.5;0.7]7.0 [;13.0; ;0.8]3.9 [9.3;1.5]0.14 Arterial lactate, mmol/L [IQR]2.1 [1.2;4.2]3.9 [1.8;7.9]1.8 [1.1;3.2] < 0.01 ICU: Intensive Care Unit, IQR = Interquartile range; n = Number, APTT ratio: Activated Partial Thromboplastin Time ratio, ALT : Alanine Aminotransferase, AST : Aspartate Aminotransferase, ALP Alkaline Phosphatase, GGT : Gamma-Glutamyl Transferase Laboratory parameters of patients with severe dengue at ICU admission in reunion Island between 2017 and 2022 ICU: Intensive Care Unit, IQR = Interquartile range; n = Number, APTT ratio: Activated Partial Thromboplastin Time ratio, ALT : Alanine Aminotransferase, AST : Aspartate Aminotransferase, ALP Alkaline Phosphatase, GGT : Gamma-Glutamyl Transferase Outcomes are presented in Table 4. Fifty-eight patients (42%) received catecholamines, and 50 (36%) required invasive mechanical ventilation. Six patients (4%) required VA-ECMO support, and one patient (1%) required VV-ECMO for Covid 19 co-infection. Twenty-nine patients (21%) underwent renal replacement therapy. Fifteen (11%) patients presented bleeding or hemorrhagic manifestations, blood transfusion was performed for 39 patients (29%), including 23 who received red blood cell concentrates and 31 who received platelet concentrates. The median length of stay in the ICU was 4 [2–7] days. The in-hospital mortality rate was 25%, with 19 patients dying from multi-organ failure syndrome, 6 from neurological causes, and 3 from hemorrhagic shock. Table 4Outcomes of patients with severe dengue in ICU in reunion Island between 2017 and 2022OutcomesAll patients(n = 137)Non Survivors’ group (n = 34)Survivors’ group(n = 103) p Requirement for organ support Catecholamines, n (%)58 (42)27 (79)31 (30) < 0.001 Norepinephrine, n(%)58 (42)28 (82)30 (29) < 0.001 Dobutamine, n (%)15 (11)7 (21)8 (8)0.73 Invasive Mechanical Ventilation, n (%)50 (36)32 (94)18 (17) < 0.001 Renal Remplacement Therapy, n (%)29 (21)15 (45)14 (14) < 0.001 VA; ECMO, n (%)6 (4)4 (12)2 (2) 0.033 VV; ECMO, n (%)1 (1)1 (3)00,25Bleeding or haemorrhagic manifestations, n (%)15 (11%)7 (21)8 (8)0.06Requirement of transfusion, n (%)39 (29)19 (59)20 (19) < 0.001 Packed Red Blood Cells, n (%)23 (17)10 (29)13 (13) 0.033 Platelet Concentrates, n (%)31 (23)15 (44)16 (16) 0.002 Duration ICU stay, days [IQR]4 [2;7]4 [1;10]4 [2;6]0.19 Hospital stay, days [IQR]8 [5;16]5 [2;12]9 [6;16]0.96In-hospital mortality n (%)34 (25)34 (100)0 (0) < 0.001 Causes of death Multi Organ Failure syndrome, n (%)-19 (56)-- Neurological causes, n (%)-6 (18)-- (Non intra; cranial) Hemorrhage, n (%)-3 (9)-- Other causes n (%)6 (18)ICU : Intensive care Unit, VA Veno-Arterial, VV : Veno-Venous, ECMO Extra Corporeal Membrane Oxygenation Outcomes of patients with severe dengue in ICU in reunion Island between 2017 and 2022 ICU : Intensive care Unit, VA Veno-Arterial, VV : Veno-Venous, ECMO Extra Corporeal Membrane Oxygenation In-hospital mortality risk factors Tables 1, 2, 3 and 4 compare the clinical, paraclinical characteristics, management strategies, and outcomes between survivors and non-survivors. Among the risk factors for in-hospital mortality, 5 variables were entered into a multivariate logistic regression: history of chronic lung disease, respiratory rate, catecholamine use, lactatemia and kaliemia at admission. Only the use for catecholamine support at admission (p = 0.005; OR 4.71 [1.58–14.07]) and respiratory rate (p = 0.005; OR 1.10 [1.03–1.17]) were identified as independent predictors of in-hospital mortality (Appendix 6). Discussion This multicenter retrospective study describes the clinical spectrum, outcomes, and mortality predictors of adults admitted to ICUs for severe dengue on Réunion Island between 2017 and 2022. Severe dengue accounted for 1.8 cases per 1,000 laboratory-confirmed infections and 5% of hospitalizations. The predominant clinical form was a hepato-nephritic syndrome with thrombocytopenia and, in more than half of cases, hemodynamic failure. Other presentations included neurological forms such as meningoencephalitis or acute polyradiculoneuropathy. Catecholamine use and elevated respiratory rate were identified as independent predictors of in-hospital mortality, with deaths mainly related to multiorgan failure rather than bleeding. Dengue activity on Réunion Island was characterized by the successive predominance of DENV-2 between 2017 and 2020, followed by DENV-1 from 2020 to 2022. The co-circulation of these two serotypes, documented in 2020–2021, has been previously linked to higher incidence of Dengue cases, and so, a higher risk of Severe Dengue [13]. SD is partly driven by antibody-dependent enhancement (ADE), in which non-neutralizing antibodies from a prior infection facilitate viral entry into immune cells expressing Fcγ receptors, leading to increased viral replication. This antibody-mediated enhancement triggers an exaggerated immune response and vascular permeability, thereby contributing to the pathogenesis of shock and severe dengue manifestations [25]. These dynamics are consistent with findings from other regions experiencing serotype shifts and secondary infections, particularly in the Indian Ocean basin and Southeast Asia [17, 26]. Thus, in a series of 220 patients in India, Dinkar et al. demonstrated that individuals with secondary dengue infections presented with more severe clinical forms, characterized by a higher frequency of hemorrhagic, hepatic, and capillary leak complications [26]. In the French West Indies, Camous et al. reported a higher incidence of acute liver failure among critically ill dengue patients related to the emergence of the new DENV-2 serotype [8]. The in-hospital mortality rate in our cohort reached 25%, a figure comparable to that reported in other ICU series of adult patients with severe dengue. Chen et al. observed a mortality rate of 23%, while Amâncio et al. reported 19% in a Brazilian population [10, 27]. This concordance reflects the broader epidemiological transition in dengue-endemic regions, where population aging and the accumulation of chronic comorbidities contribute to more severe disease courses [2, 9]. In their meta-analysis, Sangkaew et al. highlighted diabetes, hypertension, and cardiovascular or renal diseases as major risk factors for progression to severe dengue, which likely account for the poorer outcomes observed in critically ill patients [2]. Importantly, one-third of patients were initially admitted to the ICU for another acute condition or exacerbations of preexisting chronic diseases. This phenomenon mirrors observations from the 2005–2006 Chikungunya outbreak, caused by another arbovirus, during which Lemant et al. showed that 42% of ICU admissions were related to concomitant acute infectious diseases or exacerbations of pre-existing complaints [28]. Additionally, one third of patients had documented co-infections, pneumonia or urinary sepsis and a majority of patients received antibiotics at ICU admission. These findings are consistent with other studies. Chen et al. reported that two-thirds of patients admitted to the ICU for severe dengue had a co-infection, mainly involving the respiratory or urinary tract, and See et al. demonstrated that such co-infections were associated with increased mortality [10, 30]. The clinical spectrum of severe dengue was remarkably broad. Before ICU admission, fever and myalgia were the most frequently reported symptoms, consistent with the classic presentation of dengue [5–7]. Only a few patients in our study reported a prior history of dengue infection during the initial assessment. However, this figure is likely underestimated, as seroprevalence studies suggest that nearly 30% of the population of Reunion Island had already been exposed to dengue prior to 2017 [11, 12]. Abdominal pain, often described in severe disease, was observed in 43% of patients [9, 10, 31]. Interestingly, arthralgia and headache were more commonly observed among survivors than non-survivors, which contrasts with data from Latin America reported by Paraná et al., where these two symptoms were more frequently associated with severe forms of the disease [9]. At ICU admission, hepatic and renal dysfunctions were among the most frequent organ failures observed. Acute kidney injury appeared to be multifactorial, resulting from a combination of hypovolemia, rhabdomyolysis, and possible immune-mediated mechanisms [5, 32]. Hepatic impairment likely reflected both direct viral cytopathic effects and systemic hypoperfusion. Alvarez et al. reported that hepatic involvement was more frequent during secondary dengue infections, consistent with the findings of Camous et al. [8, 17]. Similarly, Padyana et al. found that the coexistence of acute renal and hepatic dysfunction was associated with increased mortality in critically ill dengue patients [31]. Circulatory failure was the most frequent form of organ dysfunction observed in our cohort, reflecting the complex and multifactorial nature of severe dengue. Most patients exhibited hypovolemic or distributive shock related to plasma leakage and dehydration from anorexia and vomiting, while a smaller proportion developed vasoplegic shock, consistent with the classical dengue shock syndrome [5, 6]. Myocardial involvement, although relatively uncommon, represents a severe and often underrecognized manifestation of dengue infection that may progress to refractory cardiogenic shock [16]. Reported prevalence varies widely across studies, ranging from isolated cases to over two-thirds of patients. In their recent meta-analysis, Sandeep et al. estimated that dengue myocarditis could affect up to one in five hospitalized patients [33]. In our cohort six patients required veno-arterial extracorporeal membrane oxygenation (VA-ECMO) for circulatory support. These observations highlight the importance of early hemodynamic assessment and echocardiographic evaluation in patients with severe dengue and circulatory instability. Thrombocytopenia was a very common finding at ICU admission, yet significant bleeding events remained rare. Only a few patients required admission for active hemorrhage, most often of gastrointestinal origin and hemorrhagic complications during the ICU stay were infrequent. Fatal bleeding was exceptional, and deaths were predominantly linked to multiorgan failure rather than hemorrhage. In the cohorts reported by Padyana et al. and Chen et al., nearly one-third of patients presented with bleeding manifestations at ICU admission, but no deaths due to hemorrhagic shock were observed. In a study investigating the causes of death related to dengue in a cohort of 60 fatal cases, Lee et al. identified only one death attributable to hemorrhagic shock secondary to gastrointestinal bleeding [34]. However, some patients may develop intracranial hemorrhage, which was fatal in our cohort. Xu et al. reported that this complication occurred in approximately 1% to 2% of patients hospitalized with dengue infection [35]. Interestingly, several studies have reported a weak correlation between thrombocytopenia and bleeding, and the limited efficacy of prophylactic platelet concentrate transfusion [36, 37]. Neurological involvement, beyond intracranial hemorrhagic complications, was also a noteworthy feature in our cohort. Several patients exhibited dengue-associated meningoencephalitis, while others developed acute polyradiculoneuropathy requiring ventilatory support. A few cases of acute disseminated encephalomyelitis (ADEM) were also identified after multidisciplinary evaluation. These findings highlight the neurotropic potential of dengue virus and the broad polymorphism of its clinical manifestations. The coexistence of both central and peripheral nervous system involvement suggests that multiple pathogenic mechanisms, including direct viral invasion and post-infectious immune responses, are involved [38, 39]. In multivariate analysis, catecholamine use and respiratory rate at admission were identified as the two independent predictors of in-hospital mortality. Numerous risk factors for mortality have been identified, including extreme age, the presence of comorbidities, severe thrombocytopenia at admission, abdominal pain, and fluid accumulation [2, 9, 19]. Paraná et al. also reported that the presence of respiratory signs was associated with increased mortality [9]. Pulmonary complications are less common, and dyspnea may occur due to pleural effusion, acute respiratory distress syndrome, pulmonary hemorrhage, pneumonia, or shock [40]. In our cohort, respiratory distress and tachypnea may also reflect associated pulmonary co-infections or decompensation of underlying chronic lung disease [5, 9, 29, 30]. Future management of severe dengue should focus on early recognition of organ dysfunction and tailored hemodynamic support. Rapid triage protocols integrating simple bedside parameters, such as respiratory rate and hemodynamic status, could facilitate early identification of high-risk patients before progression to multiorgan failure [19]. This study has several limitations. Its retrospective design may have led to incomplete data collection and heterogeneity in patient management between ICUs. Virologic and immunological data were lacking, preventing formal assessment of the link between serotype, prior immunity, and disease severity. Co-infections may have led to underestimation of severe dengue incidence. Indeed, to investigate risk factors of in-hospital mortality, our statistical data was based on a small cohort resulting a limited choice of variables to perform in the multivariable analysis. Conclusion Severe dengue on Réunion Island exhibited a wide clinical spectrum in ICUs. Despite widespread thrombocytopenia, hemorrhagic events were uncommon, while neurological, renal and hepatic complications were frequent. Co-infections were common, and comorbidities likely contributed to unfavorable outcomes. Severe Dengue in ICU was associated with substantial mortality, primarily due to multiple organ failure rather than hemorrhagic complications. Catecholamine requirement and increased respiratory rate were independent predictors of death, reflecting the severity of systemic involvement. Early identification and optimal management of shock and organ dysfunction remain essential to improving survival during future dengue outbreaks. Supplementary Information Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Supplementary Material 6 Acknowledgements The authors thank all the intensive care unit teams. Abbreviations Antibody-Dependent Enhancement Acute Disseminated EncephaloMyelitis Body Mass Index Centre Hospitalier Universitaire Intensive Care Units Inter Quartile Range PolyRadiculoNeuropathy Rapid Diagnostic Test Reverse Transcription Polymerase Chain Reaction Severe Dengue Veno Arterial Extra Corporeal Membrane Oxygenation Veno Venous Extra Corporeal Membrane Oxygenation Author contributions Charles Vidal: Writing original draft, Supervision, Investigation, Conceptualization. Quentin Soupin Coulin: Investigation, Data curation. Willelm Doucet and Marjolaine Devineau: Investigation. Zoé Blanc Data curation. Romane Crouzet, Anne-Cécile Morvan, Margot Caron: Investigation. Julien Jabot: Validation, Visualization. Nicolas Allou: Validation, Methodology, Conceptualization. Funding This work did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Formal consent was not required due to the retrospective and observational nature of the study. In accordance with current French law, informed written consent for demographic, physiologic and hospital outcome data analyses was not obtained because this observational study did not modify existing diagnostic or therapeutic strategies. Nonetheless, patients and/or relatives were informed about the anonymous data collection and told that they could decline inclusion. The protocol was reviewed by the French Ethics Committee for Infectious and Tropical Diseases (CER-MIT), and the study was declared to the French Data Protection Authority in accordance with international recommendations for observational studies (Ref CER-BDX 2025 − 137, Appendix 2). Consent for publication A newsletter was sent to each participant or their legal representative in accordance with the French Data Protection Act of January 6, 1978 (Appendix 1). Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Supplementary Material 6 Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.