TRANSFUSION PRACTICE

Prospective observational study of low thresholds for platelet transfusion in adult dengue patients
Laurent Thomas, Stphane Kaidomar, Brigitte Kerob-Bauchet, Victor Moravie, Yannick Brouste, Jean Philippe King, Sarah Schmitt, Franois Besnier, Sylvie Abel, Hossein Mehdaoui, Yves Plumelle, Fatiha Najioullah, Christiane Fonteau, Pascale Richard, Raymond Csaire, and Andr Cabi

BACKGROUND: The practice of platelet (PLT) transfusions has been adopted into the standard clinical practice in many dengue-endemic countries. Because many patients were found to have received unnecessary PLT transfusions, the development of guidelines for the management of dengue patients with thrombocytopenia has become a necessity. STUDY DESIGN AND METHODS: An emergency departmentbased prospective observational study was conducted in Martinique during a dengue outbreak in adult patients presenting with an acute febrile illness. Patients with severe bleeding and/or who underwent invasive intensive care procedures or emergency surgery were given PLT transfusion to achieve PLT counts of more than 50 109/L. PLT transfusion was also considered for patients with PLT counts of less than 5 109/L and for those with associated risk factors and PLT counts of less than 20 109/L. RESULTS: A total of 350 patients were admitted with conrmed dengue infections. Most of them had secondary serotype-2 infections. PLT counts of less than 50 109/L were recorded in 165 patients (47.1%). PLT transfusion was administered to 9 patients with thrombocytopenia. The indications included severe bleeding (5 cases), invasive procedures (3 cases), emergency surgery (1 case), and/or associated risk factors (2 cases). The median time duration from the onset of fever to PLT transfusion was 6 days (range, 4-10 days). The median amount of PLTs transfused was 3.66 1011 (range, 2.8 1011-13.2 1011). The median PLT yield was +12.4% (range, -3.9% to +67.1%). Three patients died. All other patients recovered during the second week after the onset of fever. CONCLUSION: A restrictive strategy for PLT transfusion based on clinical features and low PLT count thresholds proved to be feasible and safe for adult dengue patients.
_02132 1400..1411

engue is the most common vector-borne viral disease in the world, primarily prevalent in the tropical and subtropical regions, with an estimated annual incidence of 50 to 100 million cases.1 Dengue viruses belong to the Flaviviridae family and are classied into four antigenically distinct serotypes (from DENV-1 to DENV-4). Although the majority of the infections are asymptomatic, all the four DENV serotypes cause a spectrum of diseases ranging from ulike illness (dengue fever [DF]) to life-threatening dengue hemorrhagic fever (DHF) or dengue shock syndrome (DSS).2 The principal clinical feature that differentiates DHF from classical DF is an increased vascular permeability that results in serous effusions, hemoconcentration, hypovolemia, and in the most severe cases

ABBREVIATIONS: APTT = activated partial thromboplastin time; DF = dengue fever; DHF = dengue hemorrhagic fever; DSS = dengue shock syndrome; ED = emergency department; ER = emergency room. From the Centre Hospitalier Universitaire de Fort-de-France, Martinique (FWI); Etablissement Franais du Sang de la Martinique; and Centre dInvestigation Clinique et dEpidmiologie Clinique Antilles-Guyane (CIC-EC 802 INSERM). Address reprints requests to: Laurent Thomas, Service des Urgences, Hpital Pierre Zobda Quitman, Centre Hospitalier Universitaire, F-97200 Fort-de-France, Martinique; e-mail: laurent.thomas@chu-fortdefrance.fr. This study was supported by the Centre Hospitalier Universitaire de Fort-de-France, Martinique. Presented in abstract forms at the Second International Conference on Dengue and Dengue Haemorrhagic Fever, Phuket, Thailand, October 15-17, 2008; and at the 8me Congrs National dHmovigilance et de Scurit Transfusionnelle, Perpignan, France, November 19-21, 2008. Received for publication October 2, 2008; revision received December 26, 2008; and accepted January 10, 2009. doi: 10.1111/j.1537-2995.2009.02132.x TRANSFUSION 2009;49:1400-1411.

1400 TRANSFUSION

Volume 49, July 2009

PLT TRANSFUSION IN DENGUE PATIENTS

irreversible shock.3-5 However, the precise mechanism(s) by which dengue viruses cause DHF is not well understood. A correlation has been observed between the occurrence of DHF and secondary infection with a heterologous serotype, leading to the model of antibody-dependent enhancement,6 a phenomenon that is augmented in conditions where multiple serotypes circulate.3-7 A moderate thrombocytopenia is almost always observed in dengue patients with overt illness.2-5 Platelet (PLT) counts of less than 100 109/L develop between Days 3 and 7 of the disease and return to a normal level during the second week. A marked thrombocytopenia (PLT counts of less than 50 109/L) usually indicates clinical severity and is associated with an increased incidence of bleeding and with the development of DHF and DSS.8,9 However, the management of dengue patients with severe thrombocytopenia lacks standardized guidelines. Although there are few evidences that indicate that PLT transfusion improves the clinical outcome of dengue patients, the practice of PLT transfusions has been adopted into the standard clinical practice in many dengue-endemic countries.10-13 The practice of inappropriate PLT transfusion has been highlighted, particularly during dengue outbreaks during which 20% to 37% of the transfused patients were found to have received unnecessary PLT transfusion.11,14-16 Given the high number of acutely thrombocytopenic patients who may arrive in hospitals during an epidemic of DF, the absence of a protocol specifying instructions relating to PLT transfusions may result in unsuitable prescriptions and a shortage of resources, in addition to unnecessary costs. Moreover, because the PLT transfusions are potentially associated with acute and long-term complications,17 the development of guidelines for the management of dengue patients with severe thrombocytopenia has become a necessity. The Caribbean island of Martinique is an overseas department of France and has approximately 400,000 inhabitants mostly of African ancestry. Several dengue epidemics occurred in this island during the past decade. These epidemics were associated with DENV-2 and DENV-4 in 1995, DENV-1 in 1997, DENV-3 in 2001, and DENV-2 and DENV-4 in 2005.18 During these epidemics some patients exhibited severe thrombocytopenia without overt signs of plasma leakage or any other signs and symptoms of severe illness.19 An internal audit suggested that some of these patients were given unnecessary PLT transfusion. Therefore, a restrictive strategy for PLT transfusion based on clinical features and low PLT count thresholds was proposed by the emergency medicine specialists in coordination with medical staff from the transfusion medicine, infectious and tropical diseases, and hematology departments of the University Hospital. Our study was designed to evaluate the pertinence of this strategy in a dengue-epidemic setting. It was performed

during the last dengue outbreak observed in Martinique in 2007.20

MATERIALS AND METHODS

Study population
The study was conducted in the emergency department (ED) for adults in the University Hospital of Fort-deFrance, the capital city of Martinique. All the patients who were 15 years of age or older with a history of acute febrile illness admitted within 8 days of the onset of fever were enrolled prospectively between August 1, 2007, and January 31, 2008. Each patients informed consent was obtained. The inclusion criteria were acute dengue infection, conrmed either by reverse-transcriptase polymerase chain reaction (RT-PCR) or by detectable antidengue immunoglobulin M (IgM).

Management of patients
Molecular and serologic identications had no impact on patient management, as the assays were carried out over several days. Clinical data were computerized at the bedside via a questionnaire tailored for acute febrile illnesses. Routine biochemical and hematologic tests along with blood and urine cultures were performed at the time of admission. In addition, a 10-mL serum sample was stored in aliquots at -70C for remote immunologic and virologic studies. Other diagnostic procedures (ultrasonography, computed tomography scanning, and/or magnetic resonance imaging) were performed at the discretion of the attending emergency physicians. Patients with overt febrile illness were administered with crystalloid infusion based on 50 mL/kg/day. Admission into the hospital was decided by the attending physicians on the basis of clinical severity, biologic abnormalities, and/or comorbidities. A daily clinical observation, repeated biochemical and hematologic tests, and ultrasonography were carried out on all the hospitalized patients. Nonhospitalized patients were discharged after complete evaluation within 24 hours after admission in the ED and were followed up at home by a general practitioner or in the outpatient infectious diseases clinic. The follow-up data were obtained from the medical records and/or through personal phone calls after discharge.

Case denition
The day of the onset of fever reported by the patients was recorded and dened as Day 1 of illness. The diseases were classied retrospectively based on the clinical and biologic data recorded during the follow-up. Patients who developed hemoconcentration (20% increase in hematocrit [Hct]), total protein of less than 50 g/L, and/or other
Volume 49, July 2009 TRANSFUSION 1401

THOMAS ET AL.

signs of plasma leakage as indicated by pleural effusion or ascites observed in ultrasounds or radiology were included in the DHF/DSS-like group. Consistent with the strict application of the World Health Organization (WHO) criteria,2 many patients diagnosed with DHF or DSS also exhibited hemorrhagic manifestations and PLT counts of less than 100 109/L. However, patients for whom the strict criteria were not applicable (i.e., either bleeding or thrombocytopenia was not present) were included in the incomplete DHF/DSS group. All the other dengue patients without any signs of plasma leakage were included in the DF-like group. Patients who presented only acute febrile illness or minor external bleeding (positive tourniquet test, petechiae, spontaneously resolving epistaxis, gum bleeding, or menorrhagia) were diagnosed as having uncomplicated DF. Patients with DF and PLT counts of less than 50 109/L without any other signs of severe illness were diagnosed as having DF with isolated thrombocytopenia. Patients with DF and gastrointestinal bleeding, metrorrhagia, and/or hematuria were considered to exhibit DF with internal bleeding. Other DF patients having at least one of the following featuresencephalopathy, symptomatic postural hypotension, dehydration, serum electrolyte or acid-base disorders, hepatitis (10-fold increase in aminotransaminases), rhabdomyolysis (20-fold increase in creatine kinase), or elevated cardiac enzymes (troponin-1c higher than 1 mg/L)were diagnosed as having DF with unusual manifestations. A positive anti-dengue immunoglobulin G (IgG) capture test on serum collected within 7 days of the onset of fever was considered to indicate a secondary infection.21,22 Sera negative, both by IgG capture and IgG indirect enzyme-linked immunosorbent assay (ELISA), were considered to indicate a primary infection.

Decision algorithm for PLT transfusion

According to the general recommendation of the French sanitary security agency,24 PLT transfusion is considered only in patients with PLT counts of less than 50 109/L (Fig. 1). Patients with severe bleeding (hemorrhage sufcient to cause a decrease in Hb level of 30 g/L or a decrease in blood pressure or bleeding into a vital organ such as brain, lungs, or pericardium) and/or those who underwent invasive intensive care procedures (arterial or central venous catheterization, tracheal intubation) or emergency surgery were given PLT transfusion to achieve PLT counts of at least 50 109/L. According to the guidelines and other earlier studies,17,24-28 PLT transfusion was also considered in stable patients with minor bleeding and PLT counts no higher than 5 109/L. In other cases without the tendency of spontaneous bleeding, PLT transfusion was considered for unstable patients or those with associated risk factors (uncontrolled arterial hypertension, recent stroke, head trauma or surgery, continuation of an anticoagulant treatment, existing hemorrhagic diathesis). These patients were given PLT transfusion to achieve PLT counts higher than 20 109/L. This protocol was approved by the College of Emergency Physicians and by the expert committee of the Cellule Inter Rgionale dEpidmiologie of Martinique.

PLT concentrates and PLT yield

PLT concentrates were provided by the Etablissement Franais du Sang of Martinique. Either apheresis or pooled buffy coatderived leukoreduced PLT concentrates, stored for up to 5 days, were used depending on the availability. Each concentrate was photochemically treated with amotosalen-HCl and contained a mean 3.7 1011 PLTs suspended in 35% plasma or 65% InterSol. Posttransfusion PLT counts were sampled within 2 hours after the end of the transfusion of each PLT concentrate. The PLT yield was calculated according to the following formula:24

Laboratory methods
Blood chemistry tests, blood cell and PLTs counts, and Hct and hemoglobin (Hb) levels were carried out on automated analyzers. Low PLT counts were veried manually from a separate specimen collected on citrate to avoid false thrombocytopenia. Activated partial thromboplastin time (APTT) and prothrombin and brinogen levels were measured using a coagulometer and commercially available reagents. Methods for the detection of dengue virusspecic antibodies and identication of viral RNA have already been reported.19 The dengue antibodies were detected using the IgM capture, IgG capture, and IgM indirect ELISA kits according to the manufacturers recommendations (Panbio, Sinnamon Park, QLD, Australia). Heminested RT-PCR was carried out with DENV serotypespecic primers, as indicated by Lanciotti and colleagues.23 Calibration of the laboratory methods was performed by internal and external quality control procedures.
1402 TRANSFUSION Volume 49, July 2009

PLT yield (%) = [PLT count after transfusion ( 109 L ) PLT count before transfusion ( 109 L ) ] body weight ( kg ) 0.075 number of PLTs transfused ( 1011 ).

Statistical analysis
Data were analyzed using a statistical program (StatView 4.5, Abacus Concepts, Inc., Berkeley, CA). A descriptive analysis was performed by calculating the frequencies of the clinical features. The categories were compared using the chi-square analysis or Fishers exact test. Continuous variables were expressed as median values and ranges. Comparisons of the continuous variables between the groups were performed using nonparametric tests. A p value of 0.05 or less was considered to indicate a signicant difference. A multivariate logistic-regression analysis

PLT TRANSFUSION IN DENGUE PATIENTS

Yes

Dengue patient with PLT count <50 109/L

No

No PLT transfusion

Severe bleeding or invasive procedureA

No

Minor bleeding

No

PLT count <20 109/L

No

Yes

Yes

No PLT transfusion; monitor platelet count.

Yes PLT count <5 109/L No Unstable patient or associated risk factorB No

Monitor vital functions and coagulation profile; transfuse PLT concentrates and, if needed, the RBCs, to achieve platelet count >50 109/L and Hb >70 g/L; rectify blood coagulation defects.

Yes

Yes Transfuse PLT concentrates to achieve PLT count >20 10 9/L

Fig. 1. Decision algorithm for PLT transfusion in dengue patients. AInvasive procedure includes intensive care procedures and emergency surgery. BAssociated risk factor includes severe hypertension, recent stroke head trauma or surgery, continuation of anticoagulant treatment, or existing hemorrhagic diathesis.

was performed to assess whether independent association could be demonstrated between the variables of clinical interest recorded in the emergency room (ER) and the subsequent development of the principal clinical features of severe dengue infections.

Dengue classication
The nal classication of illnesses is shown in Table 1. The diagnosis of plasma leakage was based on hemoconcentration (46 cases), radiology and/or ultrasounds (19 cases), and/or severe hypoproteinemia (6 cases). Patients diagnosed with uncomplicated DF were signicantly younger than those with other illnesses (median, 29 years [range, 15-73 years] vs. 42 years [range, 15-88 years]; p < 0.001 by Mann-Whitney test). Male-to-female ratio was signicantly higher in the DHF/DSS-like group than in the DF-like group (49/22 = 2.23 vs. 117/162 = 0.72; p < 0.001 by Fishers exact test).

RESULTS
Patient demographics
During the outbreak, 636 adult patients were admitted in the ED for an acute febrile illness. Of these, 350 were diagnosed with conrmed dengue infections. The median age of the patients was 37 years (range, 15-88 years) and the male-to-female ratio was 0.9. The DENV-2 was detected in 297 patients (84.9%) and DENV-1 in 5 patients (1.4%); 48 other patients (13.7%) were diagnosed only by using a positive IgM determination. Among the 270 patients in whom serology could be interpreted, 227 demonstrated a secondary infection (84.1%).

Clinical characteristics
Bleeding episodes were documented in 81 patients (Table 2). Minor external bleeding was recorded in 67 patients (19.1%) and/or internal bleeding in 28 (8%). Among the patients who developed gastrointestinal
Volume 49, July 2009 TRANSFUSION 1403

THOMAS ET AL.

* Number of patients for whom serology could be interpreted. Lowest PLT counts recorded during follow-up. Indicated by 20% increase in Hct, total proteinemia of less than 50 g/L, and/or signs of pleural effusion or ascites. WHO criteria include history of fever, signs of plasma leakage, hemorrhagic manifestations, and PLT counts of less than 100 109/L (DHF) and signs and symptoms of shock (DSS). || Missing hemorrhagic manifestations and/or thrombocytopenia. PLT count of less than 50 109/L. ** DF with one of the following: encephalopathy, hepatitis, rhabdomyolysis, myocarditis, symptomatic postural hypotension, or metabolic abnormalities. M/F ratio = male-to-female ratio.

bleeding, endoscopic examination reported a MalloryWeiss syndrome (1 patient), a duodenal ulcer (1 patient), an acute hemorrhagic gastritis (2 patients), a colonic diverticulosis (2 patients), and of unknown origin in 3 patients. Other comorbidities were recorded in 13 patients. Pregnancy was reported in 4 women (all in the second trimester). Approximately 101 patients were hospitalized for more than 1 day (28.9%). This represents 28.7% of the patients who were hospitalized in Martinique for a conrmed dengue infection during this outbreak.20 The median duration of illness at presentation was 4 days (range, 1-8 days) in the nonhospitalized and hospitalized patients. In the hospitalized patients, the median hospital stay was 4 days (range, 2-39 days).

Bleeding episodes (%)

PLT counts, 109/L, median [range]

17 [3-56] 12 [2-48] 19 [3-45]

TABLE 1. Spectrum of diseases and characteristics of the study population

Secondary infections, % (n)*

140 26 21 55

[50-331] [8-49] [8-96] [8-243]

12.9 8.9 100 12.8

100 100 0

96.4 (n = 28) 75 (n = 4) 87.5 (n = 24)

(n = 116) (n = 53) (n = 7) (n = 38)

PLT counts
PLT counts of less than 100 109/L were recorded in 181 patients at the time of admission in the ER (51.7%). Furthermore, the prevalence of PLT counts no higher than 50 109/L was 34% at the time of admission, 47.1% during follow-up, and 85.2% in the hospitalized patients. The lowest PLT counts were recorded between Days 4 and 8 of the illness (Fig. 2). The distribution of the main clinical data according to the lowest PLT counts recorded during the follow-up is shown in Table 3. Minor external bleeding events were recorded more frequently in the patients with thrombocytopenia. In contrast, no correlation was observed between the lowest PLT counts recorded and the gastrointestinal and/or uterine bleeding in female patients. Among the 165 patients with PLT counts of less than 50 109/L, 67 (40.6%) demonstrated neither evidence of plasma leakage nor other clinical signs of severity. Among these 67 patients, 16 patients had PLT counts no higher than 20 109. When compared with patients of the DF-like group, patients of the DHF/DSS-like group demonstrated signicantly lower PLT count nadirs during follow-up (median, 18 109/L [range, 2-56 109/L] vs. median, 91 109/L [range 4-331 109/L]; p < 0.001 by Mann-Whitney test). Patients with secondary infections were characterized by lower PLT counts and higher liver enzymes and lymphocyte count levels at the time of admission and had lower PLT count nadirs during follow-up than patients diagnosed with primary infections (median, 41.5 109/L [range, 2-271 109/L] vs. 117 109/L [range, 6-294 109/L]; p < 0.001 by MannWhitney test).

DENV-2, %

87.9 85.7 93.5 2.67 6 1.58 42 [22-73] 60 [16-75] 50 [16-82]

M/F ratio

Age, years, median [range]

Plasma leakage (DHF/DSS-like), n = 71 DHF (strict WHO criteria), n = 33 DSS (strict WHO criteria), n = 7 Incomplete DHF/DSS,|| n = 31 No plasma leakage (DF-like), n = 279 Uncomplicated DF, n = 155 DF with isolated thrombocytopenia, n = 67 DF with internal bleeding, n = 9 DF with unusual manifestations,** n = 48

29 44 41 38

[15-73] [16-81] [18-76] [15-88]

0.57 0.97 0.13 1.18

89 70.1 77.8 85.4

70.7 96.2 100 78.9

Predictive factors of severe illnesses

Because age, sex, body temperature, PLT counts, APTT, liver enzymes levels, and other clinical or biologic abnormalities were found to be signicantly different among the patient groups, a multivariate regression analysis was performed to assess and demonstrate the possible

1404 TRANSFUSION

Case denition

Volume 49, July 2009

PLT TRANSFUSION IN DENGUE PATIENTS

TABLE 2. Bleeding episodes reported in 81 of 350 dengue patients admitted in the ED

Bleeding category No bleeding Minor bleeding Bleeding site Number of patients 269 19 49 9 16 5 3 1 4 1 1 Number of PLT transfusions* 2 1 Number of RBC transfusions

Skin (petechiae, ecchymosis) Mucosal (epistaxis, gum bleeding) Menorrhagia Metrorrhagia Gastrointestinal bleeding Spontaneous hematuria Catheter-associated hematuria Severe bleeding Gastrointestinal bleeding Metrorrhagia Catheter-associated femoral trauma * PLT transfusion was given according to a decision-tree algorithm (Fig. 1). Refused transfusion.

4 1

2 1

associated with the presence of minor external bleeding, lower PLT counts, and higher Hb levels.

Patients who underwent transfusion

PLT transfusion was administered to 9 of 165 patients with PLT counts of less than 50 109/L (5.5%; Tables 2 and 4). In addition 3 patients were administered red blood cells (RBCs) and 2 among them were also rendered fresh-frozen plasma (FFP). The proportions of patients who underwent PLT transfusion were 1 of 5 (20%) in patients with PLT count of less than 5 109/L, 5 of 64 (7.8%) in patients with PLT counts between 5 109 and 20 109/L, and 3 of 96 (3.1%) in patients with PLT count from 20 109 to 50 109/L. The proportions of patients who were given PLT transfusion were 6 of 71 (8.5%) in patients diagnosed with DHF/DSS-like illness and 3 of 124 (2.4%) in those diagnosed with other severe forms of DF (p = 0.076 by Fishers exact test). The median time duration from the onset of fever to PLT transfusion was 6 days (range, 4-10 days). The median amount of PLTs transfused was 3.66 1011 (range, 2.8 1011-13.2 1011) and was 0.35 1011 per 7 kg body weight (range, 0.29 10111.26 1011). The median PLT yield was +12.4% (range, -3.9% to +67.1%). PLT transfusion resulted in a poor PLT yield in 2 patients (Table 4, Cases 3 and 5). However, in both patients, the progressive recovery started on the day after transfusion. Among the 286 other febrile patients who were not eventually diagnosed with dengue, 8 patients with Hb levels of less than 70 g/L were given RBCs (inammatory anemia in 3, malignant diseases in 3, leptospirosis in 1, septic shock in 1), and 2 among them were also rendered PLTs (leukemia in 1, leptospirosis in 1).

Fig. 2. PLT counts recorded in 350 patients with conrmed dengue infections and plotted against the time elapsed between the onset of fever (Day 1) and sampling, including 143 patients who had varying measurements until recovery. Box plots show median values (horizontal line in the box), 25%-75% interquartile range (lower-upper limits of the box), and 90% range of data (additional bars); n = number of measurements.

independent association between these covariables recorded at the time of admission in the ER and the subsequent occurrence of the three principal clinical manifestations of severe dengue infections: severe thrombocytopenia, internal bleeding, and plasma leakage. The analysis showed that PLT count nadirs of less than 50 109/L were independently associated with older age, lower body temperature, and PLT counts no higher than 100 109/L at the time of admission. The occurrence of internal bleeding was independently associated with the presence of minor external bleeding at presentation, lower prothrombin and Hb levels, and elevation of creatininemia whatever the PLT counts recorded at the time of admission. Furthermore, the development of plasmaleakage syndrome was found to be independently

Patients outcome
In most of the patients in whom the follow-up of PLT counts could be monitored until recovery (n = 143), the
Volume 49, July 2009 TRANSFUSION 1405

THOMAS ET AL.

TABLE 3. Clinical and biologic features of the study subjects at entry and nal severity of illnesses according to the lowest PLT count recorded during follow-up
PLT counts nadir, 109/L [150-500]* Less than 50, n = 165 50 and more, n = 185 (range, 2-49) (range, 50-331) 1.26 44 (15-86) 37.2 (36-41) 81 (40-133) 43 (13-92) 118 (71.5) 49 (29.7) 7 (4.2) 10/71 (14.1) 147 (17-3203) 93 (10-1467) 195 (25-22400) 900 (194-23280) 12 (0-152) 81 (28-838) 72 (50-94) 41 (24-56) 143 (89-204) 2.08 (0.3-8.27) 1.17 (0.19-5.36) 95 (56-100) 45 (31-63) 3.06 (1.1-5.3) 133 (80.6) 120/131 (91.6) 95 (57.6) 70 (42.4) 86 (52.1) 0.66 30 (15-73) 38.5 (36-40.9) 90 (59-156) 47 (18-100) 111 (60) 18 (9.7) 2 (1.1) 7/105 (6.7) 38 (13-488) 32 (8-370) 139 (27-1533) 476 (61-4720) 13 (0-270) 78 (31-252) 76 (62-87) 39 (29-50) 135 (102-177) 2.11 (0.25-8.7) 0.69 (0.32-3.47) 87 (58-100) 39 (27-64) 3.2 (1.7-6.2) 164 (88.6) 101/139 (72.7) 184 (99.5) 1 (0.5) 17 (9.2)

Clinical and biologic data Male-to-female ratio Age, years Body temperature, C Heart rate, beats/min Pulse pressure, mmHg Gastrointestinal signs, n (%) Minor external bleeding, n (%) Gastrointestinal bleeding, n (%) Metrorrhagia, n (% female) AST, U/L [4-37] ALT, U/L [4-50] CK, U/L [20-170] LDH, U/L [219-439] CRP, mg/L [0-10] Creatininemia, mmol/L [30-115] Proteinemia, g/L [60-80] Hct, % [36-44] Hemoglobinemia, g/L [130-170] Polymorphonuclear, 109/L [1-7.5] Lymphocytes, 109/L [1-4] Prothrombin level, % normal [80-100] APTT, sec [24-40] Fibrinogen, g/L [2-4] DENV-2, n (%) Secondary infection, n (%) DF-like, n (%) DHF/DSS-like, n (%) Hospital stay >1 day, n (%)

p Value 0.004 <0.001 <0.001 <0.001 0.063 0.017 <0.001 0.09 0.122 <0.001 <0.001 0.019 <0.001 0.786 0.423 <0.003 <0.001 <0.001 0.260 <0.001 <0.001 <0.001 0.033 0.150 <0.001 <0.001 <0.001 <0.001

* [Normal range]. Continuous data are expressed by median value (range). Categorical and continuous data were compared using Fishers exact test and Mann-Whitney test, respectively. Clinical severity recorded during follow-up. ALT = alanine aminotransferase; AST = aspartate aminotransferase; CK = creatine kinase; CRP = C-reactive protein; LDH = lactate dehydrogenase.

PLT counts were found to increase before the eighth day after the onset of fever (median, 7 days; range, 4-12 days) and subsequently return to normal level during the second week (Fig. 2). A 30-year-old female patient with a history of uterine broma and chronic anemia, diagnosed with a secondary DENV-2 infection, presented 5 days after the onset of fever with intense metrorrhagia and a PLT count of 27 109/L. She refused transfusion. Cessation of bleeding was observed 2 days later together with the progressive normalization of PLT counts. In other female patients with thrombocytopenia who presented with uterine bleeding as the only hemorrhagic manifestation, bleeding was never sufcient to warrant consideration for PLT transfusion. In all the patients who manifested external or internal bleeding and subsequently recovered, cessation of bleeding was observed before the ninth day after the onset of fever. However, three patients with severe thrombocytopenia died (i.e., 1.8% of the patients with PLT counts of less than 50 109/L), including two of the nine transfused patients and one nontransfused patient. None of the
1406 TRANSFUSION Volume 49, July 2009

deaths was related to hemorrhage. Death was the consequence of irreversible fulminating hepatitis in one patient (Table 4, Case 1) and refractory pulmonary failure after aspiration pneumonia in the other patient (Table 4, Case 2). Another 65-year-old patient diagnosed with a primary DENV-2 infection manifested seizures and developed diffuse brain edema, transtentorial herniation, and refractory shock. Six days after the onset of fever, the PLT count was 3 109/L. Because the patient was diagnosed with brain death, he was not rendered PLT transfusion. Furthermore, postmortem pathologic examination did not reveal any internal hemorrhagic complication. All the other patients recovered.

DISCUSSION
This prospective observational study was conducted in 2007 during a DENV-2 outbreak in Martinique. Most of the patients presented with secondary infections. This explains the large proportion of patients developing severe thrombocytopenia.2-5 Although the last outbreak

TABLE 4. Characteristics of dengue patients who underwent PLT transfusion

PLT count,* 109/L 4 Invasive ICU procedure, internal bleeding 2.4 Indication of PLT transfusion Units of blood products administered Number of PLTs transfused (1011) PLT yield (%) -2.3

Patients main characteristics and outcome Case 1: male, 63 years, BW 75 kg, secondary DENV-1 infection, DHF, fulminating hepatitis, gastrointestinal bleeding, coagulopathy, died 3 10 25 9 Invasive ICU procedure 1 APC (D5) 1 APC (D6) 1 RDPC (D7) 1 APC (D7) 5 RBC 8 FFP 1 APC (D10) 3.1 3.6 4.1 4.6

+67.1 +10.9 NA +19.6

Internal bleeding

1 APC (D6)

-3.9

Case 2: male, 59 years, BW 80 kg, DENV-2 infection, incomplete DSS, status epilepticus, aspiration pneumonia, died Case 3: female, 42 years, BW 70 kg, secondary DENV-2 infection, DF with hemorrhagic complications, metrorrhagia, colonic bleeding, recovered Case 4: male, 60 years, BW 80 kg, secondary DENV-2 infection, DSS, multiple organ failure, major bleeding from iatrogenic femoral trauma (hemodialysis catheter), recovered 24 46 7 8 15 Associated risk factor Internal bleeding Invasive ICU procedure, emergency vascular surgery 1 APC (D8) 1 APC (D8) 11 RBC 6 FFP 1 APC (D5) 1 APC (D5) 1 APC (D6)

3.2 4.8 3.71 2.45 5.9

+16.9 +27.5 +2.8 -2.1 +12.5

48

Internal bleeding

1 APC + 1 RDPC (D5) 7 RBC

8.7

+9

20 9

Internal bleeding Associated risk factor

1 APC (D6) 1 APC (D6)

3 2.8

+15.3 +12.4

Case 5: male, 56 years, BW 70 kg, DENV-2 infection, DHF, colonic bleeding, recovered Case 6: female, 31 years, BW 58 kg, primary dengue infection, DF with unusual manifestations, postural hypotension (syncope), head trauma, periorbital ecchymosis, menorrhagia, recovered Case 7: male, 63 years, BW 80 kg, secondary DENV-2 infection, DSS, gastric ulcer hemorrhage, ecchymosis, recovered Case 8: male, 37 years, BW 68 kg, secondary DENV-2 infection, DHF, lower gastrointestinal bleeding, recovered Case 9: male, 83 years, BW 66 kg, DENV-2 infection, DF with unusual manifestations, postural hypotension (syncope), coumadin treatment for atrial brillation, no bleeding, recovered

Volume 49, July 2009

PLT TRANSFUSION IN DENGUE PATIENTS

* Recorded at the time of the onset of PLT transfusion. PLT yield = [PLT count after transfusion (109/L) - PLT count before transfusion (109/L)] BW 0.075/number of PLTs transfused (1011). APC = apheresis PLT concentrate; BW = body weight; D# = number of days from the onset of fever to PLT transfusion; ICU = intensive care unit; NA = not available; RDPC = random donor PLT concentrate.

TRANSFUSION 1407

THOMAS ET AL.

reported was characterized by an increase in both severe cases and hospital stay,20 no increase in mortality was observed in Martinique during the following epidemics. Data recorded in Martinique over the past years outline that life-threatening bleeding only occurred in those dengue patients with severe gastrointestinal hemorrhage or in those who underwent emergency surgery or who were at risk for intracranial hemorrhage; however, death mainly resulted from a major visceral organ failure (such as liver, heart, brain, or lung). It should be underlined that these observations were made in adult patients. Signicant clinical differences between pediatric and adult dengue cases have been demonstrated.9 During a large outbreak in Cuba in 1981, the death risk in children was found to be 15-fold higher than in adults.29 In many countries in Southeast Asia as well as in the Americas, the DSS with concomitant gastrointestinal bleeding remains the leading cause of dengue-related deaths among children.8,9 In Martinique, this has not been observed so far. During the last outbreak reported, the proportion of patients diagnosed with uncomplicated DF (44.4%) was similar to the proportions reported by different teams in the Caribbean basin and Latin America.30-32 The differences observed in the distribution of severe clinical forms of dengue infections reported in these countries and in Martinique are mainly owing to the lack of reproducibility among the different severity scoring systems used (e.g., DHF and DSS as dened by the WHO system,2 DHF/ DSS-like,19 DF with signs associated with shock,30 DF with hemorrhagic manifestations,30,31 intermediate DF/DHF,32 provisional diagnosis of DHF,33 and other unusual manifestations19,34,35) and the considerable overlap among all these clinical forms. Hemorrhagic complications may occur in DHF as well as in DF patients. In this respect, the recent reviews have underlined the need to revise and standardize the dengue classication system and redene a threshold for thrombocytopenia.34-36 In addition, information about patients ethnicity should be reported. The lower rate of DHF/DSS in the Americas than in Asia is well established.37 A dengue resistance gene in individuals of African ancestry could explain the differences in clinical severity observed in Cuba and in Haiti between African Americans and Caucasians.38,39 Apart from patients diagnosed with classical clinical forms of dengue infections, the study population included a group of patients who developed signs of plasma leakage without hemorrhagic manifestations. This incomplete DHF/DSS group included 43.7% of the patients diagnosed with a plasma-leakage syndrome. Another important nding of this study is that 40.6% of patients with severe thrombocytopenia demonstrated no other clinical signs of severity (included in the DF with isolated thrombocytopenia group). Follow-up of these patients revealed that none developed signicant bleeding or showed evidence of plasma leakage. However, most of these patients
1408 TRANSFUSION Volume 49, July 2009

exhibited three- to vefold increased plasma liver enzymes levels, increased APTT, and normal prothrombin and brinogen levels. All these patients were discharged within 24 hours after admission in the ED and recovered within a few days. Although it is well documented that the most severe cases of DF occur more often in patients with severe thrombocytopenia,2-5,8-9,40 our ndings raise the question on the signicance of thrombocytopenia as a specic marker of the severe disease in adult dengue patients in Martinique. These clinical features might be dependent on the genetic backgrounds. The degree of thrombocytopenia, liver enzymes, or vascular permeability observed during primary and secondary dengue infections might be quite different among Caucasians, Africans, and Asians. The Creole population of Martinique is the result of mixing European, African, and Indian ancestry over centuries. Given this specic ethnic makeup, it is difcult to identify ethnic subclasses and to study this feature in Martinique. The lack of correlation between thrombocytopenia and the occurrence of severe bleeding in dengue patients has been previously reported,11,16,41 suggesting the distinct physiopathologic processes. However, to date, the exact mechanisms of thrombocytopenia and bleeding in dengue infections remain poorly understood. The mechanisms of thrombocytopenia include early bone marrow suppression and increased peripheral immune-mediated destruction of PLTs.5,7 Bleeding could result from the combined effects of vascular alteration, thrombocytopenia, and coagulation disorders.40,42,43 However, the principal clinical feature that differentiates DHF from uncomplicated DF is not bleeding but the increased vascular permeability.3-5 This might be the consequence of a very complex phenomenon involving the production of cytokines and other cellular factors induced by the virulence and/or the complexes between DENV and heterologous dengue antibodies.7 The resulting plasma-leakage syndrome is characterized by hypovolemia, hemoconcentration indicated by elevation of Hct, and a marked decrease in plasma proteins levels. Internal bleeding especially at the gastrointestinal tract may be difcult to recognize in the presence of hemoconcentration. In our study, multivariate analysis identied increased Hb levels and low PLT counts as signicantly associated with plasma leakage whereas normal to low Hb levels were signicantly associated with the occurrence of internal bleeding whatever the PLT count levels. Another study reported similar ndings in children with DSS and concurrent gastrointestinal bleeding.41 Owing to the lack of precise guidelines, the proportion of dengue patients receiving PLT transfusion is subjected to variation in the decision among different physicians and the differences in the level of care provided by the attending hospitals.10-16 The general recommendation according to which a PLT transfusion can only be

PLT TRANSFUSION IN DENGUE PATIENTS

given when the number of PLTs falls below 50 109/L must be applied to patients suffering from DF. An initial distinction must be drawn between transfusions for curative purposes and prophylactic transfusions.24,27 In Martinique, PLT transfusion was essentially reserved for patients suffering from DF who had severe internal bleeding or who had to undergo surgery or other invasive procedures. Given the condition of these patients it was often also necessary to give a simultaneous transfusion of RBCs and plasma. In addition to these instructions for curative purposes, a prophylactic PLT transfusion could also be given to highly thrombocytopenic patients with a high risk of cerebral bleeding (traumatic brain injury or recent strokes, chronic high blood pressure not controlled by treatment, etc.). The practice of PLT transfusion in treating dengue patients has several major limitations, such as the lack of availability of PLT compounds owing to the restriction of the donor population or eligibility during an epidemic, the potential hazards of PLT transfusion, and the potential refractoriness to transfusion. During the outbreak reported, the availability of PLT compounds was quite a challenge for the Regional Blood Donation Center, with a global increase of 31.3% in PLT use (monthly use [in 0.5 1011 PLTs] was 376 in the pre-epidemic period vs. 494 during the outbreak), leading to an increase in PLT apheresis procedures of 23.3% (mean, 43 per month to 53 per month during the outbreak) and buffy coatpooled compounds production. The ratio of postdonation information leading to product destruction (mainly febrile illness) increased from 3.22 per 1000 to 4.03 per 1000. The potential hazards of PLT transfusion include febrile and allergic reactions, infections, transmission of viral diseases (limited in this study by the viroattenuation of pathogens by amotosalen), alloimmunization to PLTs, hemolysis, hypotensive reactions, graft-versus-host reactions, and transfusion-related acute lung injury.17 It has also been reported that the duration of survival of the transfused PLTs in dengue patients was markedly shortened.44 However, refractoriness to PLT transfusion in dengue patients has been poorly investigated in the clinical settings.11,16 Our study observed a mean PLT yield of +12.4%, measured within 2 hours after the PLT transfusion. As the PLT transfusions were mostly performed on Day 6 of the illness, it was difcult to differentiate the effects of transfusion from the spontaneous recovery from thrombocytopenia. Another limitation is that there is no proof that prophylactic PLT transfusions can benet patients suffering from DF.45 Until controlled clinical trials have been carried out, the most recent expert recommendations must be followed.24,27,46 In summary, during a DENV-2 epidemic in Martinique, 84.1% of adult patients admitted in the ED were diagnosed with secondary infections. The PLT counts of less than 50 109/L were recorded in 47.1% of the

patients. However, 40.6% of these acutely thrombocytopenic patients demonstrated no clinical sign of severe illness. A restrictive strategy for PLT transfusion based on clinical features and low PLT count thresholds reduced the practice of PLT transfusion to 5.5% of the adult dengue patients with severe thrombocytopenia and proved to be feasible and safe.

ACKNOWLEDGMENTS We thank the medical and nursing staff of the Emergency and General Medicine Departments and the Intensive Care Units of Fort-de-France University Hospital for their care toward the patients.

CONFLICT OF INTEREST The authors declare no competing nancial interests.

REFERENCES
1. Gibbons RV, Vaughn DW. Dengue: an escalating problem. BMJ 2002;324:1563-6. 2. WHO. Dengue hemorrhagic fever: diagnosis, treatment, prevention and control. Geneva: WHO; 1997. 3. Gubler DJ. Dengue and dengue hemorrhagic fever. Clin Microbiol Rev 1998;11:480-96. 4. Guzman MG, Kouri G. Dengue: an update. Lancet Infect Dis 2002;2:33-42. 5. Halstead SB. Dengue. Lancet 2007;370:1644-52. 6. Kliks SC, Nisalak A, Brandt WE, Wahl L, Burke DS. Antibody-dependent enhancement of dengue virus growth in human monocytes as a risk factor for dengue hemorrhagic fever. Am J Trop Med Hyg 1989;40:444-51. 7. Green S, Rothman A. Immunopathological mechanisms in dengue and dengue hemorrhagic fever. Curr Opin Infect Dis 2006;19:429-36. 8. Phuong CX, Nhan NT, Kneen R, Thuy PT, van Thien C, Nga NT, Thuy TT, Solomon T, Stepniewska K, Wills B, and the Dong Nai Study Group. Clinical diagnosis and assessment of severity of conrmed dengue infections in Vietnamese children: is the World Health Organization classication system helpful? Am J Trop Med Hyg 2004;70:172-9. 9. Hammond SN, Balmaseda A, Perez L, Tellez Y, Saboro SI, Mercado JC, Videa E, Rodriguez Y, Prez MA, Cuadra R, Solano S, Rocha J, Idiaquez W, Gonzalez A, Harris E. Differences in dengue severity in infants, children, and adults in a 3-year hospital-based study in Nicaragua. Am J Trop Med Hyg 2005;73:1063-70. 10. Chuansumrit A, Phimolthares V, Tardtong P, TapaneyaOlarn C, Tapaneya-Olarn W, Kowsathit P, Chantarojsiri T. Transfusion requirements in patients with dengue hemorrhagic fever. Southeast Asian J Trop Med Public Health 2000;31:10-4. Volume 49, July 2009 TRANSFUSION 1409

THOMAS ET AL.

11. Chairulfatah A, Setiabudi D, Agoes R, Colebunders R. Thrombocytopenia and platelet transfusions in dengue haemorrhagic fever and dengue shock syndrome. Dengue Bull 2003;27:141-3. 12. Makroo RN, Raina V, Kumar P, Kanth K. Role of platelet transfusion in the management of dengue patients in a tertiary care hospital. Asian J Transf Sci 2007;1:4-7. 13. Chandralekha, Pratyush G, Anjan T. The north Indian dengue outbreak 2006: a retrospective analysis of intensive care unit admissions in a tertiary care hospital. Trans R Soc Trop Med Hyg 2008;102:143-7. 14. Jamal R, Hoe TS, Ong LC, Ong LC, Afah I, Khuzaiah R, Doraisamy G. A clinical audit on the practice of platelet transfusions at a tertiary paediatric referral centre. Malays J Pathol 1998;20:35-40. 15. Kumar ND, Tomar V, Singh B, Kela K. Platelet transfusion practice during dengue fever epidemic. Indian J Pathol Microbiol 2000;43:55-60. 16. Chaudhary R, Khetan D, Sinha S, Sinha P, Sonker A, Pandey P, Das SS, Agarwal P, Ray V. Transfusion support to dengue patients in a hospital based blood transfusion service in north India. Transfus Apher Sci 2006;35:239-44. 17. Stroncek DF, Rebulla P. Platelet transfusions. Lancet 2007; 370:427-38. 18. Cellule Interrgionale dEpidmiologie (CIRE) de la Martinique. Situation pidmiologique de la dengue dans les Antilles franaises. Mto: Bulletin dAlerte et de Surveillance Antilles Guyane (BASAG). [cited November, 2006.] Available from: http://www.invs.sante.fr/publications/ basag/basag2006-7.pdf 19. Thomas L, Verlaeten O, Cabi A, Kaidomar S, Moravie V, Martial J, Najioullah F, Plumelle Y, Fonteau C, Dussart P, Csaire R. Inuence of the dengue serotype, previous dengue infection and plasma viral load on clinical presentation and outcome during a dengue-2 and dengue-4 co-epidemic. Am J Trop Med Hyg 2008;78:990-8. 20. Cellule Interrgionale dEpidmiologie (CIRE) de la Martinique. Bilan de lpidmie 2007-2008 en Martinique. Mto: Bulletin dAlerte et de Surveillance Antilles Guyane (BASAG). [cited April, 2008.] Available from: http:// www.invs.sante.fr/publications/basag/Basag2008-4.pdf 21. Vaughn DW, Nisalak A, Solomon T, Kalayanarooj S, Nguyen MD, Keen R, Cuzzubbo A, Devine PL. Rapid serologic diagnosis of dengue virus infection using a commercial capture ELISA that distinguishes primary and secondary dengue infection. Am J Trop Med Hyg 1999;60: 693-8. 22. Vazquez S, Hafner G, Ruiz D, Calzada N, Guzman MG. Evaluation of immunoglobulin M and G capture enzymelinked immunosorbent assay Panbio kits for diagnostic dengue infections. J Clin Virol 2007;39:194-8. 23. Lanciotti RS, Calisher CH, Gubler DJ, Chang GJ, Vorndam AV. Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptase-polymerase chain reaction. J Clin Microbiol 1992;30:545-51.

24. Agence Franaise de Scurit Sanitaire des Produits de Sant (AFSSAPS). Transfusion de plaquettes: produits, indications, recommandations. Afssaps: AFSSAPS. [cited June, 2003.] Available from: http://www.afssaps.fr/var/ afssaps_site/storage/original/application/ 4d0c07b9ebaac5b2f096d7f332b71500.pdf 25. Sagmeister M, Oec L, Gmr G. A restrictive platelet transfusion policy allowing long-term support of outpatients with severe aplastic anemia. Blood 1999;93:3124-6. 26. Drews RE, Weinberger SE. Thrombocytopenic disorders in critically ill patients. Am J Respir Crit Care Med 2000;162: 347-51. 27. Slichter SJ. Relationship between platelet count and bleeding risk in thrombocytopenic patients. Transfus Med Rev 2004;18:153-67. 28. Arnold DM, Crowther MA, Cook RJ, Sigouin C, Heddle NM, Molnar L, Cook DJ. Utilization of platelet transfusions in the intensive care unit: indications, transfusion triggers, and platelet count responses. Transfusion 2006;46: 1286-91. 29. Guzmn MG, Kouri G, Bravo J, Valdes L, Vazquez S, Halstead SB. Effect of age on outcome of secondary dengue 2 infections. Int J Infect Dis 2002;6:118-24. 30. Harris E, Videa E, Prez L, Sandoval E, Tllez Y, Prez ML, Cuadra R, Rocha J, Idiaquez W, Alonso RE, Delgado MA, Campo LA, Acevedo F, Gonzalez A, Amador JJ, Balmaseda A. Clinical, epidemiologic, and virologic features of dengue in the 1998 epidemic in Nicaragua. Am J Trop Med Hyg 2000;63:5-11. 31. Lorono-Pino MA, Farfn-Ale JA, Zapata-Peraza AL, Rosado Paredes EP, Flores-Flores LF, Garca-Rejn JE, Daz FJ, Blitvich BJ, Andrade-Narvez M, Jimnez-Ros E, Blair CD, Olson KE, Black W 4th, Beaty BJ. Introduction of the American/Asian genotype of dengue 2 virus into the Yucatan state of Mexico. Am J Trop Med Hyg 2004;71: 485-92. 32. Guilarde AO, Turchi MD, Siqueira JB, Feres VC, Rocha B, Levi JE, Souza VA, Boas LS, Pannuti CS, Martelli CM. Dengue and dengue hemorrhagic fever among adults: clinical outcomes related to viremia, serotypes, and antibody response. J Infect Dis 2008;197:817-24. Ramos MM, Argello DF, Luxemburger C, Quiones L, Muoz JL, Beatty M, Lang J, Tomashek KM. Epidemiological and clinical observations on patients with dengue in Puerto Rico: results from the rst year of enhanced surveillanceJune 2005-May 2006. Am J Trop Med Hyg 2008;79:123-7. Rigau-Perez JG. Severe dengue: the need for new case denitions. Lancet Infect Dis 2006;6:297-302. Deen JL, Harris E, Wills B, Balmaseda A, Hammond SN, Rocha C, Dung NM, Hung NT, Hien TT, Farrar JJ. The WHO dengue classication and case denitions: time for a reassessment. Lancet 2006;368:170-3. Farrar J, Focks D, Gubler D, Barrera R, Guzman MG, Simmons C, Kalayanarooj S, Lum L, McCall PJ, Lloyd L,

33.

34. 35.

36.

1410 TRANSFUSION

Volume 49, July 2009

PLT TRANSFUSION IN DENGUE PATIENTS

Horstick O, Dayal-Drager R, Nathan MB, Kroeger A; WHO/ TDR Dengue Scientic Working Group. Towards a global dengue research agenda. Trop Med Intern Health 2007;12: 695-9. 37. Halstead SB. Dengue in the Americas and Southeast Asia: do they differ? Rev Panam Salud Publica 2006;20:407-15. 38. Guzman MG, Kouri GP, Bravo J, Soler M, Vazquez S, Morier L. Dengue hemorrhagic fever in Cuba, 1981: a retrospective seroepidemiologic study. Am J Trop Med Hyg 1990;42:179-84. 39. Halstead SB, Streit TG, Lafontant JG, Putvatana R, Russell K, Sun W, Kanesa-Thasan N, Hayes CG, Watts DM. Haiti: absence of dengue hemorrhagic fever despite hyperendemic dengue virus transmission. Am J Trop Med Hyg 2001;65:180-3. 40. Krishnamurti C, Kalayanarooj S, Cutting MA, Peat RA, Rothwell SW, Reid TJ, Green S, Nisalak A, Endy TP, Vaughn DW, Nimmannitya S, Innis BL. Mechanisms of hemorrhage in dengue without circulatory collapse. Am J Trop Med Hyg 2001;65:840-7. 41. Lum LC, Goh AY, Chan PW, El-Amin AL, Lam SK. Risk

factors for hemorrhage in severe dengue infections. J Pediatr 2002;140:629-31. 42. Mitrakul C, Poshyachinda M, Futrakul P, Sangkawibha N, Ahandrik S. Hemostatic and platelet kinetic studies in dengue hemorrhagic fever. Am J Trop Med Hyg 1977;26: 975-84. 43. Wills BA, Oragui EE, Stephens AC, Daramola OA, Dung NM, Loan HT, Chau NV, Chambers M, Stepniewska K, Farrar JJ, Levin M. Coagulation abnormalities in dengue hemorrhagic fever: serial investigations in 167 Vietnamese children with dengue shock syndrome. Clin Infect Dis 2002;35:277-85. 44. Isarangkura P, Tuchinda S. The behaviour of transfused platelets in dengue hemorrhagic fever. Southeast Asian J Trop Med Public Health 1993;24 Suppl 1:222-4. 45. Lum LC, Latif ME, Goh AY, Chan PW, Lam SK. Preventive transfusion in dengue shock syndromeis it necessary? J Pediatr 2003;143:682-4. 46. Slichter SJ. Evidence-based platelet transfusion guidelines. Hematol Am Soc Hematol Educ Progr 2007;2007:172-8.

Volume 49, July 2009

TRANSFUSION 1411