1will.oa

Comparison of Three Fluid Solutions for Resuscitation Bridget A. Wills, M.R.C.P., Nguyen M. Dung, M.D., Ha T. Loan, M.D., Dong T.H. Tam, M.D., Tran T.N. Thuy, M.D., Le T.T. Minh, M.D., Tran V. Diet, M.D., Nguyen T. Hao, M.D., Nguyen V. Chau, M.D., Kasia Stepniewska, Ph.D., Nicholas J. White, F.R.C.P., and Jeremy J. Farrar, F.R.C.P.
b a c k g r o u n d
Dengue shock syndrome is characterized by severe vascular leakage and disordered he- From the Oxford University Clinical Re- mostasis and progresses to death in 1 to 5 percent of cases. Although volume replace- eases, Ho Chi Minh City, Vietnam, and the ment is recognized as the critical therapeutic intervention, World Health Organization Centre for Clinical Vaccinology and Tropi-management guidelines remain empirical rather than evidence-based.
cal Medicine, Oxford University, Oxford,United Kingdom (B.A.W., K.S., N.J.W., J.J.F.); and the Hospital for Tropical Dis-eases, Ho Chi Minh City, Vietnam (N.M.D., We performed a double-blind, randomized comparison of three fluids for initial resus- H.T.L., D.T.H.T., T.T.N.T., L.T.T.M., T.V.D., citation of Vietnamese children with dengue shock syndrome. We randomly assigned N.T.H., N.V.C.). Address reprint requests 383 children with moderately severe shock to receive Ringer’s lactate, 6 percent dext- to Dr. Wills at Oxford University Clinical Research Unit, Hospital for Tropical Dis- ran 70 (a colloid), or 6 percent hydroxyethyl starch (a colloid) and 129 children with se- eases, 190 Ben Ham Tu, Quan 5, Ho Chi vere shock to receive one of the colloids. The primary outcome measure was require- Minh City, Vietnam, or at bridgetw@hcm.
ment for rescue colloid at any time after administration of the study fluid.
Only one patient died (<0.2 percent mortality). The primary outcome measure — re- Copyright 2005 Massachusetts Medical Society.
quirement for rescue colloid — was similar for the different fluids in the two severitygroups. The relative risk of requirement for rescue colloid was 1.08 (95 percent con-fidence interval, 0.78 to 1.47; P=0.65) among children with moderate shock who re-ceived Ringer’s lactate as compared with either of the colloid solutions, 1.13 (95 per-cent confidence interval, 0.74 to 1.74; P=0.59) among children who received dextranas compared with starch in the group with severe shock, and 0.88 (95 percent confi-dence interval, 0.66 to 1.17; P=0.38) among children who received dextran as comparedwith starch in the combined analysis. Although treatment with Ringer’s lactate result-ed in less rapid improvement in the hematocrit and a marginally longer time to initialrecovery than did treatment with either of the colloid solutions, there were no differ-ences in all other measures of treatment response. Only minor differences in efficacywere detected between the two colloids, but significantly more recipients of dextranthan of starch had adverse reactions. Bleeding manifestations, coagulation derange-ments, and severity of fluid overload were similar for all fluid-treatment groups.
c o n c l u s i o n s
Initial resuscitation with Ringer’s lactate is indicated for children with moderately severedengue shock syndrome. Dextran 70 and 6 percent hydroxyethyl starch perform similar-ly in children with severe shock, but given the adverse reactions associated with the useof dextran, starch may be preferable for this group.
Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine This was followed by a one-year study of 230 most serious manifestation of dengue children with dengue shock syndrome who received dhemorrhagic fever, a relatively new dis- treatment with similar fluids. The study focused on ease entity that has spread progressively through- sustained as well as immediate treatment effects.13out Asia and South America since its first appear- The study showed no clear advantages among theance in Bangkok, Thailand, in the 1950s.1 Dengue fluids in terms of the time to recover from the ini-hemorrhagic fever is characterized by systemic vas- tial episode of shock, the time to achieve final car-cular leakage and disordered hemostasis and may diovascular stability, or the requirement for rescuedevelop after infection with any of four dengue viral colloid. The most significant factor determiningserotypes.2,3 Between 250,000 and 500,000 cases of clinical response was the pulse pressure (i.e., thedengue hemorrhagic fever, mainly in children, are difference between the systolic and diastolic pres-reported to the World Health Organization (WHO) sures) at presentation with shock. Dengue is un-annually, with mortality rates of 1 to 5 percent usual in that a slow leak occurs over several days,among patients with shock.1,4 The pathophysio- permitting compensatory mechanisms to operate.
logical mechanisms underlying the vascular leak- Before the development of overt cardiovascular col-age and coagulopathy are poorly understood, and lapse, the diastolic pressure rises to meet the sys-no specific treatment is available.
tolic pressure, and the pulse pressure narrows; thus Prompt restoration of the volume of circulating the pulse pressure is a marker of the severity of vas- plasma is the cornerstone of therapy for dengue cular leakage. According to WHO guidelines, a pa-shock syndrome. WHO management guidelines, tient with dengue hemorrhagic fever is consideredfirst proposed in 1975, recommend replacement to have dengue shock syndrome once the pulseof plasma losses with crystalloid solutions initially, pressure narrows to 20 mm Hg or less. Post hocfollowed by boluses of colloid for patients with re- analysis of the one-year study suggested that earlycurrent or refractory shock.5 Although the recom- treatment with a colloid solution hastened recoverymendations were initially invaluable in focusing in children with a pulse pressure of 10 mm Hg orattention on the need for volume replacement, they less at presentation (i.e., the group with profoundhave not been updated to any appreciable degree cardiovascular compromise).13since 1975.6 During this time there has been con- Many synthetic colloid preparations are now siderable debate in the medical literature regarding available, each with unique physicochemical char-the use of crystalloids versus colloids for volume acteristics that determine their likely efficacy andreplacement in critically ill patients.7-9 Theoretically, side-effect profiles.11 However, if colloid resuscita-colloid solutions offer advantages in patients with tion does confer a true benefit in children withincreased vascular permeability,10 although in clin- dengue shock syndrome, it is not clear which prep-ical practice a clear benefit has not been demonstrat- aration would be most appropriate. Pathophysio-ed. Conversely, most colloid solutions have adverse logical studies indicate that there is preferentialeffects on hemostasis,11 an important consider- leakage of relatively small plasma proteins (e.g., al-ation in patients with dengue.
bumin) as compared with larger molecules (e.g., Only two randomized and blinded trials have IgG),14 which implies that resuscitation with col- investigated the effect of different crystalloid and loid preparations of larger molecular weights maycolloid fluid regimens on the outcome from dengue offer therapeutic advantages.
shock syndrome. The first, a pilot study in which In an attempt to optimize management and ad- 50 children received one of four fluids (6 percent dress some of these unresolved issues, we conduct-dextran 70, 3 percent gelatin, Ringer’s lactate, or ed a trial comparing three resuscitation fluids, a0.9 percent saline) over a two-hour period for ini- crystalloid and two synthetic colloids, for primarytial resuscitation, showed that there were signifi- resuscitation of children with dengue shock syn-cantly greater improvements in cardiac index, he- drome. Since our pediatric intensive care unit con-matocrit, and blood pressure at the end of the sistently achieves mortality rates of less than 1 per-infusion of the study fluid among children who had cent for dengue shock syndrome, many thousandsreceived a colloid solution than among those who of children would need to be enrolled to examinehad received a crystalloid solution.12 All the chil- effects on mortality. Instead, we used the estab-dren recovered, but differences in sustained effects lished WHO format of primary resuscitation withamong the four fluids were not examined.
an initial infusion of a specific volume of fluid fol- Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e lowed by rescue colloid treatment as necessary, and Treatment allocation was determined in advance we tested the hypothesis that there is no difference with the use of computer-generated random num-in the requirement for treatment with rescue col- bers. To ensure prerandomization concealment asloid after initial resuscitation with Ringer’s lac- well as blinding, treatment packs comprising threetate, 6 percent dextran 70, or 6 percent hydroxy- 500-ml bottles of study fluid, sealed inside speciallyethyl starch (each at 25 ml per kilogram of body prepared cardboard containers and identified onlyweight) among Vietnamese children with dengue by a study number, were supplied to the ward forshock syndrome.
each patient. Treatment packs for at least the nextfive patients in each severity group were kept on theward at any time. Both the computer-generated ran- domization process and preparation of the treat- s t u d y d e s i g n
ment packs were carried out by independent re- The trial was a single-center, randomized, double- search staff not involved in clinical care. A sealedblind comparison of an isotonic crystalloid solution envelope containing the identity of the study fluid(Ringer’s lactate) and two isotonic colloid solutions was attached to the study file for each child in case(6 percent dextran 70 [dextran] and 6 percent hy- of emergency.
droxyethyl starch 200/0.5 [starch]) for emergency After the infusion of the study fluid, the children resuscitation of children with dengue shock syn- received a standard schedule of Ringer’s lactate thatdrome. The children were stratified according to involved a reduction at specific time intervals topulse pressure at admission, a marker of the sever- maintenance levels after eight hours. Pulse, bloodity of the vascular leak.13 No children in the group pressure, and peripheral perfusion were monitoredwith severe shock received a crystalloid because of at least hourly until they were stable for a minimumconcerns about the potential development of criti- of 24 hours, and then every 4 hours until discharge.
cal fluid overload without access to advanced respi- The capillary hematocrit was measured at baseline,ratory support. The study took place in the pediatric 2 and 6 hours after study entry, and then approxi-intensive care unit at the Hospital for Tropical Dis- mately every 12 hours or in the event of cardiovas-eases in Ho Chi Minh City, Vietnam. The ethics and cular deterioration. Additional citrated plasmascience committee of the hospital approved the samples for coagulation screening were obtainedprotocol.
on study days 2 and 4, together with a second se-rum sample for serologic testing for dengue infec- s t u d y p o p u l a t i o n a n d c l i n i c a l m e t h o d s
tion at discharge. An ultrasound scan of the chest Children 2 to 15 years of age presenting directly to and abdomen was carried out on study day 3 by onethe hospital with clinical dengue shock syndrome of two trained observers with the use of a standard-were eligible for enrollment provided a parent or ized protocol to measure the depth of any pleuralguardian gave written informed consent. WHO effusions and assess the severity of ascites.
guidelines were used for the diagnosis of dengue Patients whose cardiovascular status did not im- shock syndrome.6 At study entry we recorded dem- prove after administration of the study fluid (i.e.,ographic data, history, and examination findings, those who had further narrowing or no response inand we obtained citrated plasma samples for coag- pulse pressure, together with persisting or worsen-ulation screening and serum samples for the diag- ing peripheral shutdown, a rising hematocrit, ornosis of dengue. Patients were enrolled in one of both) received infusions of 5 to 10 ml per kilogramtwo groups according to the pulse pressure at ad- of rescue colloid (usually dextran) at the discretionmission. Children with shock of moderate severity of the clinician. Similarly, if after an initial favorable(pulse pressure, >10 and ≤20 mm Hg) constituted response, the pulse pressure subsequently nar-group 1 and were randomly assigned to receive rowed again to 20 mm Hg or less with peripheralRinger’s lactate, dextran, or starch. Group 2 con- vasoconstriction, a rising hematocrit, or both, res-sisted of those with severe shock (pulse pressure, cue colloid could be given. It was not possible to fix≤10 mm Hg); these children were randomly as- absolute criteria for the use of rescue colloid, but thesigned to receive either dextran or starch. Each same core group of clinicians was responsible forchild received 15 ml per kilogram of body weight of patient care throughout the study, and the generalthe allocated fluid over a one-hour period, followed policy of the pediatric intensive care unit for inter-by 10 ml per kilogram over the second hour.
vention after initial resuscitation is conservative.
Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine Patients received inotropes, blood transfusions, of an earlier study, in which approximately 30 per-diuretics, and other therapy at the discretion of the cent of children with dengue shock syndrome re-treating clinician.
quired rescue colloid.13 Parallel recruitment to thestudy comparing use of the two colloids in children l a b o r a t o r y p r o c e d u r e s
with severe shock was expected at a ratio of approx- A diagnosis of dengue infection was made with the imately three patients with moderately severe shockuse of Dengue Duo IgM capture and IgG capture to one patient with severe shock.
enzyme-linked immunosorbent assay kits (PanBio) A statistician who was not involved in the de- on paired serum samples. Coagulation screening sign or execution of the study performed all analy-was performed with the use of kits obtained from ses with the use of Stata (version 8.0) or StatsDi-Diagnostica Stago; tests included those for pro- rect statistical software. A preplanned interimthrombin time, activated partial-thromboplastin analysis was carried out approximately halfwaytime, and fibrinogen level and a semiquantitative through the study, and the results were reviewedassay for fibrin-degradation products. Only results by the data and safety monitoring committee. Af-of samples separated within 12 hours of venipunc- ter a series of adverse reactions, a second analysis,ture and without visible hemolysis or clot forma- focusing on safety, was carried out after 440 chil-tion were included in the analysis.
dren had been recruited; the committee recom-mended that the trial continue. All analyses were o u t c o m e m e a s u r e s
performed on an intention-to-treat basis. Patient The primary outcome measure was the requirement characteristics and treatment effects of the variousfor supplemental intervention with rescue colloid fluids were compared with the use of the chi-at any time after the infusion of the study fluid. The square or Fisher’s exact test for categorical varia-following secondary outcome measures were ex- bles and the Mann–Whitney or Kruskal–Wallis testamined: the time taken to achieve initial cardiovas- for continuous variables. Cardiovascular recoverycular stability (defined as the time in hours from times were compared with the use of the log-rankstudy entry until the pulse pressure reached and test, and the estimated probability of recovery iswas maintained at ≥25 mm Hg with a systolic pres- presented as Kaplan–Meier curves. Dextran wassure of ≥80 mm Hg for a minimum of two hours), compared with starch across the categories ofthe time taken to achieve sustained cardiovascular pulse pressure; comparisons were carried out withstability (defined as the time in hours from study the use of the Mantel–Haenszel test for categoricalentry to reach and maintain these cardiovascular outcomes, with conditional logistic regressionindexes indefinitely without further intervention), used to test fluid association with continuous out-the volumes of rescue colloid and total parenteral comes. Comparisons of event rates among the var-fluid required, the pattern of change in the hemato- ious fluid-treatment groups are presented as rela-crit, and the number of days in the hospital. A sin- tive risks, focusing on comparisons between thegle observer not involved in clinical management crystalloid and either of the colloid groups, or be-calculated all recovery times and fluid volumes. In tween the dextran and the starch groups. We usedaddition, the following four possible adverse effects Koopman’s method for the ratio of binomials toof the various fluids were investigated: clinical determine 95 percent confidence intervals.
bleeding; laboratory evidence of coagulopathy; theseverity of vascular leakage as assessed clinically, by ultrasonography, and by the requirement for di-uretic therapy; and the incidence of allergic-type The profile of the trial is presented in Figure 1. A to-reactions.
tal of 512 children were recruited into the study be-tween August 1999 and March 2004, and all received s t a t i s t i c a l a n a l y s i s
their designated study fluid. Of the 512 patients, A sample size of 360 patients (120 in each fluid 476 (93 percent) had confirmed dengue, were cor-group) was calculated for the main study to give 80 rectly enrolled and randomly assigned to receive apercent power to detect a 50 percent reduction in fluid, and received the assigned fluid to within 10the requirement for rescue colloid at a 5 percent percent of the intended volume of 25 ml per kilo-significance level, taking as baseline the findings gram over two hours for initial resuscitation. Treat- Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e 5 Refused consent 91 Because of logistic problems with supply of study fluids in Vietnam 33 Because of temporary suspension of study pending safety analysis Figure 1. Profile of the Trial.
ment allocation was unblinded in six patients (five differences in effects of fluid treatment
receiving dextran, and one receiving starch) after a There was no significant difference among the flu-
severe allergic-type reaction to the study fluid, in ids in terms of the overall proportion of children
order to permit decisions to be made about which requiring rescue colloid in either severity group
colloid to use for subsequent rescue therapy. All (Table 2). The relative risk of a requirement for res-
baseline characteristics were similar among the flu- cue colloid was 1.08 (95 percent confidence inter-
id-treatment groups for the 383 children with mod- val, 0.78 to 1.47; P=0.65) among children with
erately severe shock (group 1) and for the 129 chil- moderate shock who received Ringer’s lactate as
dren with severe shock (group 2) (Table 1). One compared with either of the colloid solutions, 1.13
child (a recipient of starch) died of profound shock (95 percent confidence interval, 0.74 to 1.74; P=
and gastrointestinal bleeding. The remaining study 0.59) among children who received dextran as com-
patients recovered fully. Outcome data reported here pared with starch in the severe shock group, and
are for all 512 children, except where indicated.
0.88 (95 confidence interval, 0.66 to 1.17; P=0.38) Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine Table 1. Baseline Characteristics of the Patients.*
Characteristic
Fluid treatment before study entry — no. (%) Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e Table 1. (Continued.)
Characteristic
Activated partial-thromboplastin time — sec¶** Fibrin-degradation products — no. (%)¶¿ * Patients in group 1 had a pulse pressure greater than 10 mm Hg and less than or equal to 20 mm Hg. Patients in group 2 had a pulse pres- sure of 10 mm Hg or less. Percentages may not sum to 100 because of rounding.
† Data are for those patients with recordable values for cardiovascular characteristics at admission. In group 1, data are for 119 patients receiving dextran, 117 receiving starch, and 119 receiving Ringer’s lactate. In group 2, data are for 39 patients receiving dextran and 34 receiving starch.
‡ Data are for those patients with recordable values for cardiovascular characteristics at admission. In group 1, data are for 125 patients receiv- ing dextran and 127 receiving Ringer’s lactate. In group 2, data are for 47 patients receiving dextran and 39 receiving starch.
§ Data are for those patients with recordable values for cardiovascular characteristics at admission. In group 1, data are for 125 patients receiv- ing dextran and 127 receiving Ringer’s lactate. In group 2, data are for 43 patients receiving dextran and 36 receiving starch.
¶ Data are for samples separated within 12 hours, with no hemolysis or visible clot formation.
¿ In group 1, data are for 46 patients receiving dextran, 36 receiving starch, and 44 receiving Ringer’s lactate. In group 2, data are for 22 patients receiving dextran and 21 receiving starch.
** In group 1, data are for 45 patients receiving dextran, 35 receiving starch, and 44 receiving Ringer’s lactate. In group 2, data are for 22 patients receiving dextran and 20 receiving starch.
among children who received dextran as compared recipients of starch than of dextran required res-with starch in the combined analysis. Children in cue colloid at this early stage in either severitygroup 1 who received Ringer’s lactate for primary group; the relative risk of a requirement for rescueresuscitation took longer to achieve initial cardio- colloid for the initial episode of shock was 0.34vascular stability than patients receiving either of (95 percent confidence interval, 0.07 to 1.71;the colloids (Fig. 2A), but the degree of compro- P=0.03) among starch recipients as comparedmise during this period was generally not sufficient with dextran recipients in the combined analysis.
to warrant intervention with rescue colloid, and the There was a corresponding minor advantage intime to final cardiovascular stability was not differ- initial cardiovascular recovery times in the recipi-ent among the fluid-treatment groups (Fig. 2B).
ents of starch as compared with the recipients of No child in group 1 who had received starch for dextran in group 2 (median, one hour vs. two primary resuscitation required rescue colloid to re- hours; P=0.03 by the log-rank test). However, thecover from this episode. Overall, significantly fewer numbers involved were small and the effect was Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine Table 2. Primary and Secondary Outcome Measures.*
Ringer’s Lactate
P Value†
Rescue colloid for initial resuscitation — no. (%) Rescue colloid required subsequently — no. (%) Percentage reduction in hematocrit at 2 hr‡ * In group 1, data are for 126 patients receiving dextran, 129 receiving starch, and 128 receiving Ringer’s lactate. In group 2, data are for 67 pa- tients receiving dextran and 62 receiving starch. In the groups combined, data are for 193 patients receiving dextran and 191 receiving starch. A dash denotes not applicable.
† The chi-square or Fisher’s exact test was used for categorical variables, and the Mann–Whitney or Kruskal–Wallis test for continuous variables. Comparisons between the children receiving dextran and starch across the two severity groups were carried out with the use of the Mantel–Haenszel test for categorical outcomes, with conditional logistic regression used to test fluid association with continuous outcomes. P values are for three-way comparisons in group 1 and for two-way comparisons in group 2 and the combined group analyses. For the combined analy-ses, tests for heterogeneity were not significant in each instance.
‡ In group 1, data are for 121 patients receiving dextran, 123 receiving starch, and 126 receiving Ringer’s lactate. In group 2, data are for 62 pa- tients receiving dextran and 60 receiving starch.
Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e not sustained. Thus, there was no difference in ei- ther severity group in the requirement for colloid subsequent to the initial episode of shock, in the volumes of rescue colloid or total parenteral fluid administered, in the final recovery times, or in the Marked differences were noted among the fluids in their effects on the capillary hematocrit. Twohours after study entry, the median reduction in the hematocrit from baseline was 9 percent (90 percent range, 1 to 19 percent) for children in group 1 whoreceived Ringer’s lactate as compared with 25 per- Proportion of Patients with Shock
cent (90 percent range, 10 to 35 percent) for those who received dextran and 22 percent (90 percent range, 7 to 31 percent) for those who received starch Time (hr)
(P<0.001) (Table 2). However, the subsequent in-crease in the hematocrit between two and six hours was significantly greater for the two colloids than for the crystalloid. The median increase in the he- matocrit during this period was 5 percent (90 per- cent range, ¡8 to 20 percent) for dextran and 5 per- cent (90 percent range, ¡10 to 21 percent) for starchas compared with 0 percent (90 percent range, ¡12 to 12 percent) for Ringer’s lactate in group 1 (P<0.001), and 8 percent (90 percent range, ¡6 to 22percent) and 5 percent (90 percent range, ¡9 to 21 percent) for dextran and starch, respectively, in group 2 (no statistical difference).
Proportion of Patients with Shock
p o s s i b l e c o m p l i c a t i o n s o f f l u i d t r e a t m e n t
There were no significant differences in any adverse Time (hr)
effects of the various fluid treatments, except in the Figure 2. Kaplan–Meier Curves for Time from Study Entry to Initial (Panel A)
incidence of allergic-type reactions (Table 3). Fif- and Sustained (Panel B) Cardiovascular Stability among Children in Group 1,
teen of 193 patients (8 percent) receiving dextran According to the Resuscitation Fluid Received.
had severe reactions (transient high fever and rigors Of 383 patients in group 1, 126 received dextran, 129 starch, and 128 Ringer’s without cardiorespiratory compromise) that oc- curred within six hours of commencing the studyfluid. Bacterial cultures and tests for endotoxin con-tamination were negative, but ongoing investiga- athy at study entry in all groups of patients, as ex-tions suggest contamination of certain batches with pected, that worsened slightly on study day 2 buta nonendotoxin pyrogen (Dr. S. Poole, National In- improved again by study day 4 (Table 4). The onlystitute for Biological Standards and Control, Lon- significant difference was between the absolutedon: personal communication). An urticarial rash values for the partial-thromboplastin time in thewithout fever developed in one patient in the starch children in group 2 who received dextran or starchgroup at the end of the infusion. All patients re- on day 2. However, all the coagulation variablessponded to symptomatic treatment alone.
were slightly worse before study entry in the children There were no differences among the fluid- in group 2 who received starch, and the day 2 re- treatment groups in the development of new bleed- sults actually represented an improvement froming manifestations, clinical fluid overload, objective day 1. All other comparisons showed no differencesmeasures of the overall severity of vascular leakage, among the fluid-treatment groups, including com-or the use of furosemide (Table 3). The serial coag- parisons of the percentage changes in each varia-ulation-screening tests revealed a mild coagulop- ble on the different days (data not shown).
Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine Table 3. Possible Adverse Effects of Fluid Treatment.*
Adverse Effect
New bleeding after study entry — no. (%) Depth of right pleural effusion — cm† Severe allergic-type reactions after infusion — no (%)‡ * Percentages may not sum to 100 because of rounding.
† Data are for children for whom an ultrasound scan was performed 48 to 72 hours after study entry (i.e., in group 1, 78 patients receiving dex- tran, 82 receiving starch, and 77 receiving Ringer’s lactate; and in group 2, 41 receiving dextran and 36 receiving starch). To assess the severity of leakage, the depth of fluid (in centimeters) in the right hemithorax posterior to the liver was measured with the patient lying flat and the probe placed at the costal margin in the mid-clavicular line.
‡P<0.001 for group 1 and P=0.03 for group 2 by Fisher’s exact test. All other statistical comparisons were not significant.
tensive care units, has advanced considerably the general debate about crystalloids versus colloids.
Although volume replacement is accepted as the Our trial, focusing on a single disease entity formainstay of treatment for children with dengue which fluid resuscitation is the essential and usual-shock syndrome, the two previous studies that in- ly sole intervention required, indicates that even investigated the efficacy of different fluids in this sit- patients with vascular leak, isotonic crystalloid so-uation were not adequately powered with respect lutions are as effective as colloid solutions for theto a clinically relevant outcome to permit definitive majority of patients.
comparisons between management with crystalloid More patients with severe shock at presentation and colloid solutions. This study, with requirement required rescue colloid than did patients with mod-for rescue colloid as the outcome indicator, estab- erately severe shock, but only minor differences inlishes that the cheapest and safest choice, Ring- efficacy were detected between the colloids in eitherer’s lactate, is as effective as either of the colloids severity group. Although minor benefits were seenfor initial resuscitation of children with moderate- among the recipients of starch during the initial re-ly severe shock. The recent publication of the Sa- suscitation, the relatively large molecular size ofline versus Albumin Fluid Evaluation (SAFE) Study starch as compared with dextran did not result infindings,15 indicating that albumin and normal sa- prolonged intravascular persistence or a more sus-line were equally effective for fluid resuscitation in tained volume-expanding effect. We found no ev-a large, heterogeneous population of patients in in- idence of adverse effects of either colloid on the Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e Table 4. Results of Serial Coagulation-Screening Tests According to Day of Shock.*
Variable
* The day of study entry is day 1. Data are for samples separated within 12 hours, with no hemolysis or visible clot formation. Dashes denote not applicable. Percentages may not sum to 100 because of rounding.
† In group 2 on day 2, the activated partial-thromboplastin time was significantly prolonged in the children who received starch as compared with those who received dextran (P<0.001 by the Kruskal–Wallis test). All other comparisons for the absolute values and for the percentage changes in the variables at each time point showed no differences among the fluids.
intrinsic coagulopathy or clinical bleeding mani- fect that is followed by a rebound increase in vascu-festations or on the severity of fluid overload.
The serial hematocrit measurements reflect a Current theories of microvascular ultrafiltration combination of the effects of fluid treatment and support the basic Starling principle of a balancedongoing vascular leak. The hematocrit data indicate equilibrium between opposing oncotic and hydro-that the two colloids exert a dramatic immediate ef- static pressures but postulate that the glycocalyx, Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. The new england journal of medicine rather than the endothelial cells themselves, is the mise, and yet most did well with Ringer’s lactatemajor regulator of fluid flow.16,17 There is good ev- alone.
idence that plasma proteins, particularly albumin, Unlike the SAFE trial, the present trial did not adsorb to positive residues in the glycocalyx layer examine mortality as an outcome; our primary out-and restrict ultrafiltration.18-20 Albumin is probably come was an intervention based on the treatingwashed out of this layer during dengue infections clinician’s subjective assessment of need. However,but may be replaced temporarily by the synthetic col- our study took place in a single ward staffed by theloids, which are known to permeate the glycocalyx at same core group of doctors throughout and studydifferent rates, depending on molecular size.21 In treatment was concealed and blinded so that dif-this way, colloids may briefly alter the selective per- ferences in the threshold for intervention are likelymeability of the endothelial barrier, reducing out- to have been distributed evenly across the groupsward flux and permitting the low hydrostatic pres- of children receiving the various fluids. Althoughsure of the capillaries to rise until the colloid there is no proven relationship to mortality, colloidmolecules themselves are washed out and the net rescue is an integral part of the WHO managementStarling forces again favor leakage but from a higher guidelines for resuscitation of dengue shock syn-baseline hydrostatic pressure. In contrast, crystal- drome, and a substantial requirement for rescueloid solutions equilibrate rapidly throughout the in- colloid is considered by physicians in endemic ar-travascular and interstitial fluid spaces and appear eas to be a poor prognostic marker.
to have no effect on the function of the endothelial It is likely that the excellent overall outcome barrier. The effects of colloids are transient, how- (one death among 641 children with dengue shockever, and despite the early rebound in the hemat- syndrome treated during the trial period) reflectsocrit seen in the children receiving colloids, we meticulous overall medical and nursing care asfound no difference between the different fluids much as the specific treatments used. Hourly ob-in the overall severity of fluid overload when it was servations, immediate access to ward-based hema-assessed 48 to 72 hours after the study infusion.
tocrit measurements, and a conservative interven- During the study, an excess of febrile responses tion policy ensure that patients receive volumes of occurred in the recipients of dextran. Dextrans are intravenous fluid that are titrated carefully to re-produced by a process involving bacterial degrada- quirements, thus providing sufficient fluid to main-tion, and despite purification, residual pyrogens tain vital functions during the period of systemicmay be sufficient to induce febrile responses. In the leakage without overfilling the intravascular space.
present study, febrile responses were associated Respiratory compromise secondary to fluid over-with particular batches of fluids. The overall fre- load is a major contributor to mortality in settingsquency and importance of this adverse effect in the with poor resources, few personnel, and limitedmanagement of dengue remain to be determined.
equipment. In general, in the Southeast Asian re- In addition to its relevance to dengue, major gion, mortality rates of 1 to 5 percent persist, and strengths of this large, randomized trial in the efforts to improve management must continue. Forgeneral debate about crystalloids versus colloids are ethical reasons, we did not address the issue of thethe uniform nature of the underlying disease pro- use of crystalloids for patients with profound or re-cess and the fact that fluid resuscitation is the single current shock, two situations in which colloid so-most important therapeutic intervention. Subgroup lutions are thought to be beneficial, despite a lackanalysis of the SAFE trial suggested a treatment of good supporting evidence. Further studies areeffect favoring albumin in patients with severe sep- needed that focus on these high-risk groups. Thesis of mixed underlying causes,15 many of whom uniformly good outcome in children with shock ofare likely to have had an associated vascular-leak moderate severity who received the crystalloid insyndrome. In our study, however, there was no this study may help to provide reassurance for futureclear benefit to the use of a colloid in children with studies. In addition, work to better define the patho-moderately severe shock due to vascular-leak syn- physiological mechanisms underlying the vascular-drome. Although the pathophysiological mecha- leak process will be useful to inform future studies.
nism underlying the vascular leak associated with In conclusion, most children with dengue shock severe sepsis may well be different from that asso- syndrome respond well to judicious treatment withciated with the dengue virus, this group of chil- isotonic crystalloid solutions. Early interventiondren showed considerable cardiovascular compro- with colloid solutions is not indicated. The fluid Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved. f l u i d r e s u s c i t a t i o n f o r d e n g u e s h o c k s y n d r o m e regimen of Ringer’s lactate at 25 ml per kilogram Supported by the Wellcome Trust. Dr. Wills is a Wellcome Trust over a period of two hours is now supported by Career Development Fellow, Dr. Farrar is a Wellcome Trust Senior Fellow, and Professor White is a Wellcome Trust Principal Fellow.
strong prospective evidence and should be recom- We are indebted to the directors of the Hospital for Tropical Dis- mended for children with moderately severe shock. eases for their encouragement, to the medical and nursing staff ofFor those with severe shock, the situation is less the pediatric intensive care unit for their care of the patients, to the staff of the Oxford University Clinical Research Unit for laboratory clear-cut, and clinicians must continue to rely on and clerical support, to B. Braun Medical Industries in Vietnam forpersonal experience, familiarity with particular donating all the parenteral fluids and providing the containers forproducts, local availability, and cost. Minor advan- blinding, to Dr. Stephen Poole (principal scientist, National Insti- tute for Biological Standards and Control, London) for investigat- tages in initial recovery were shown with starch, ing possible contamination of dextran fluid batches, and to theand significantly more adverse reactions were as- data and safety monitoring committee (Dr. Rose McGready at thesociated with dextran, so if the use of a colloid is Shoklo Malaria Research Unit, Mae Sot, Thailand, and Dr. Delia Bethell at the Mahidol University–Oxford Tropical Medicine Re- considered necessary, starch may be the preferred search Progamme, Bangkok, Thailand, in addition to Professoroption.
r e f e r e n c e s
Crystalloids vs. colloids in fluid resuscita- 15. The SAFE Study Investigators. A com-
and dengue haemorrhagic fever. In: Gubler parison of albumin and saline for fluid re- suscitation in the intensive care unit. N Engl orrhagic fever. Wallingford, England: CAB 16. Starling EH. On the absorption of fluids
fluid resuscitation in critically ill patients.
from the connective tissue spaces. J Physiol ated with dengue infection. I. Clinical and physiologic manifestations of dengue hem- 17. Michel CC, Curry FE. Microvascular per-
orrhagic fever in Thailand, 1964. J Pediatr 10. Haupt MT, Kaufman BS, Carlson RW.
meability. Physiol Rev 1999;79:703-61.
Fluid resuscitation in patients with increased 18. Haraldsson B, Rippe B. Orosomucoid
vascular permeability. Crit Care Clin 1992;8: as one of the serum components contribut- ing to normal capillary permselectivity in rat 11. Griffel MI, Kaufman BS. Pharmacology
skeletal muscle. Acta Physiol Scand 1987; of colloids and crystalloids. Crit Care Clin 19. Schneeberger EE, Lynch RD, Neary BA.
12. Dung NM, Day NP, Tam DT, et al. Fluid
Interaction of native and chemically modi- fied albumin with pulmonary microvascular ment, surveillance, prevention and control four intravenous-fluid regimens. Clin Infect 20. Huxley VH, Curry FE. Differential ac-
13. Ngo NT, Cao XT, Kneen R, et al. Acute
solute permeability. Am J Physiol 1991;260: treatment, prevention and control. 2nd ed.
Geneva: World Health Organization, 1997.
4 intravenous fluid regimens in the first 21. Vink H, Duling BR. Capillary endothe-
Schierhout G, Roberts I. Fluid resuscita- hour. Clin Infect Dis 2001;32:204-13.
lial surface layer selectively reduces plasma tion with colloid or crystalloid solutions in 14. Wills BA, Oragui EE, Dung NM, et al.
critically ill patients: a systematic review of Size and charge characteristics of the protein Heart Circ Physiol 2000;278:H285-H289.
randomised trials. BMJ 1998;316:961-4.
leak in dengue shock syndrome. J Infect Dis Copyright 2005 Massachusetts Medical Society. Choi PT, Yip G, Quinonez LG, Cook DJ.
Downloaded from www.nejm.org on January 9, 2007 . Copyright 2005 Massachusetts Medical Society. All rights reserved.

Source: http://www.spi.org.py/docs/NEJM-Comparison_of_Three_Fluid_Solutions_for_Resuscitation_in_Dengue_Shock_Syndrome.pdf