You are on page 1of 7

Failure of High-Dose Methylprednisolone in Established Dengue Shock

Syndrome: A Placebo-Controlled, Double-Blind Study

Sompon Tassniyom, MD*; Sirijitt Vasanawathana, MDX; Aroon Chirawatkul, MSc; and
Suntharee Rojanasuphot, BScII

ABSTRACT. Objective. Steroids are widely used in Infection with mosquito-borne dengue viruses
Thailand and other dengue-endemic countries to treat usually results in a mild, nonspecific febrile episode
severe dengue shock syndrome. This study was designed often accompanied by a skin rash. In a small propor-
to determine whether a single high dose of methyipred-
tion of cases, the infection causes an increase in vas-
nisolone will reduce mortality in children with dengue
shock syndrome who did not respond to simple fluid and cular permeability which leads to a bleeding diathe-
plasma replacement therapy. sis or disseminated bleeding, in which case the
Methods. A prospective, randomized, doubled-blind, condition is referred to as dengue hemorrhagic fever
controlled trial was conducted in two hospitals in Khon (DHF). The high prevalence of dengue infection in
Kaen Thailand during June to September in 1987 and endemic areas such as Southeast Asia causes DHF to
1988. Sixty-three children with severe dengue shock syn- be a leading cause of morbidity and mortality, par-
drome were randomized into two groups; the first group
ticularly in children aged 5 to 10 years. Since becom-
received a single dose of methylprednisolone (30 mgfkg)
and the second group received placebo.
ing endemic in Thailand in the 1960s, large outbreaks
Results. There was no significant difference in mor- of DHF have occurred at intervals of 1 to 2 years. The
tality between the two groups (P = .63). The mortality disease has also occurred sporadically in the Amen-
rate was 12.5% (4/32) in the steroid group and 12.9% (4/31) cas, including a large outbreak in Cuba in 1981.
in the group that received placebo. The sequelae at 2 Based on the experience in Southeast Asia, it is not
weeks among treatment and control survivors were not unreasonable to expect recurring outbreaks in the
significantly different These two groups were compara-
Western hemisphere.2
ble in age, sex, severity of illness, and duration of shock
In approximately 20% to 30% of DHF cases, shock
at the outset of the study. The two treatment groups were
similar in subsequent hospital course as determined by develops, creating a syndrome called dengue shock
maximum and minimum hematocrit level and bleeding syndrome (DSS). At the present time no drug is avail-
severity The numbers of patients in each group who had able to treat DHF specifically. In patients who do not
liver failure and evidence of disseminated intravascular have shock, treatment is supportive and includes re-
clotting defect were also comparable. Complications such hydration. But for patients in whom shock develops,
as occurrence of fever after shock, pneumonia, convul- steroids have long been used together with intrave-
sion, cardiac arrest, pulmonary hemorrhage, and positive
nous saline and plasma or plasma expanders (SPP).3
hemoculture were not significantly different in the treat-
ment and control groups. In addition to their well-known immunosuppres-
Conclusions. A single high dose of methylpredniso- sive effect, it is thought that steroids stabilize capil-
lone does not reduce mortality in severe dengue shock lary permeability and reduce accumulation of meta-
syndrome which does not respond to conventional criti- bolic products of shock. Based on this biologically
cal care. Pediatrics 199392:111-115; dengue hemorrhagic plausible hypothesis, pediatricians have tried to do-
f evei dengue shock syndrome, steroids, randomization,
termine whether of use in DSS.
steroids are
double-blind method.
Steroid efficacy in 055 has been investigated in
five studies published in the English language
ABBREVIATIONS. DHI dengue hemorrhagic fever; DSS, dengue literature.37#{176} These studies vary in the severity of
shock syndrome; SPP, saline and plasma or plasma expanders;
WHO, World Health Organization; ELISA, enzyme-linked immu-
illness in the patients selected, the steroid dosage
nosorbent assay. used, and the strength of the design and methodol-
ogy In one study, a high-dose steroid in a single
bolus was administered to patients with severe
From the *[)p,.flent of Pediatrics and Clinical Epidemiology Unit, Fac- shock, and there was a dramatic reduction in mor-
ulty of Medicine, Khon Kaen University; Department of Pediatrics, Khon tality rat&#{176};however, in general, most other studies
Kaen PrOVincial Hospital, Ministry of Public Health; Department of Bio-
have failed
to find an important effect on mortality.
statistics and Demography, Faculty of Public Health, Khon Kaen University;
and IVirology Division, Department of Medical Science, Ministry of Public The combined evidence at present is unconvincing
Health, Bangkok, Thailand. partly because of inherent methodologic weaknesses
Received for publication May 5, 1992; accepted Feb 22, 1993. in the design of the earlier studies (eg, non-blinding).
Reprint requests to(S.T.) Dept ofPediatrics and ClinicalEpideiiiology Unit,
Thus it was our objective to reinvestigate the efficacy
Faculty of Medicine, Thon Kaen University, Khon Kaen, Thailand 40002.
PEDIATRK3 (ISSN 0031 4005). Copyright C 1993 by the American Aced- of a high-dose steroid by applying standard clinical
emy of Pediatrics. trial methods.3

PEDIATRICS Vol. 92 No. I July 1993 111

Downloaded from by guest on October 25, 2016
METHODS volume, and vial container. All other investigators and nursing
staff remained blind with respect to study treatment until the trial
Definitions was completed.
Shock was defined in accordance with the World Health Orga- Study drug (either steroid or placebo) was allocated to individ-
nization (WHO) guidelines for the diagnosis, treatment, and con- ual patients according to a prescribed randomized schedule. The
trol of DHF. The WHO definition of shock is rapid and weak randomization scheme was blocked in groups of four patients and
pulse with narrowing
of the pulse pressure (20 mm Hg or less) or stratified by whether the patient originated from the two partici-
hypotension, with cold, clammy skin and restlessness. pating hospitals or was referred from an outlying center.
Hypotension was not defined quantitatively in the guideline. Supportive Care. All patients received the same supportive care.
This study defined hypotension as systolic blood pressure below Intravenous fluids were given at a rate adjusted to clinical status
80 mm Hg as applied in previous studies.74 as described in the WHO guideline. Maintenance fluid therapy
A patient with severe
DSS was defined as one in whom shock followed a regimen specified in a standard textbook of pediat-
developed despite the use of standard SPP therapy as applied in rica.2 Whole blood was given if there was a fall in hematocrit.
Futrakul and coworkersstudy.#{176} The concept of severe 055 is not Platelet concentrate was given if bleeding was uncontrolled. The
part of the WHO guideline but was necessary in this study to hematocrit was monitored every 2 to 4 hours depending on the
determine a clinical point at which randomization to steroids severity of shock and bleeding. Blood pressure, pulse, and respi-
would be appropriate. ration rate were monitored either half-hourly or hourly. Central
Consciousness level follows Plum and Posners classification.5 venous pressure monitoring was initiated and a urine catheter was
A patient without a normal state of consciousness was defined as inserted if the patients condition deteriorated. Some patients re-
one with any state of consciousness below normal alert state. ceived continuous blood pressure monitoring through an arterial
Liver failure was defined as a condition in which severe im- line.
pairment of hepatic function is associated with progressive mental Patients who needed a central venous pressure line were
changes.67 If a patient had only increased levels of liver enzymes moved to an intensive care unit. If patients had respiratory insuf-
without impairment of other functions, the patient did not meet ficiency an endotracheal tube was inserted and a respirator was
this definition. applied. Two types of respirators were used: a Bird (pressure)
Temperature rising after defervescence was defined as any tern- respirator and a Barnett (volume) respirator.
perature rise above 37.5#{176}C
measured orally after initial tempera- Antibiotics. Attending clinicians were free to make their own
ture reduction from the shock state. decision whether or not to give antibiotics. The indication for
antibiotic use was fever rising again after shock, resulting presum-
Study Design ably from contamination following instrument insertion.
A randomized, placebo-controlled, double-blind design was
used in this trial. Patients from the Khon Kaen University Hospital Follow-up
and Khon Kaen Provincial Hospital were considered for recruit- If a patient survived, he/she was called back
at 10 to 14 days
ment. To be eligible, patients had to meet the WHO clinical criteria post randomization for general physical and a blood
for DHF, they had to be younger than 15 years of ages and they sample for complete blood cell count, platelet count, and conva-
had to either have shock that had not responded to SPP therapy at lescent antibody level. Parents of the patients received payment
one of the two participating hospitals or had to have been referred for transportation, but even so 10 patients did not return to the
to one of these hospitals (usually already in advanced shock) from hospital for their appointment. In these cases, the following day
an outlying medical center. Patients meeting these inclusion crite- our team visited their home so that every case received follow-up.
na were subsequently excluded if they were found to have other The primary outcome for this trial was mortality from any
comorbid conditions, were already receiving steroid treatment, or cause. We expected mortality in the nonsteroid group to be high.
failed to provide an informed consent. One earlier study#{176}reported 100% mortality in comparable pa-
tients, but we used a more conservative figure of 90% for sample-
Investigations size calculations. Assuming that the use of steroids would halve
For patients who met eligibility criteria, a history was taken and this rate, a total sample size of 26 would yield 80% power for a
a physical examination performed. In every case, blood was one-tailed 5% test. We increased this to a planned recruitment of 38
drawn for complete blood cell count, platelet count, viral isolation,
to cater for some noncompliance and to achieve the required num-
and antibody testing, and urine was collected for analysis. In some her within the 85% of patients expected to be seropositive.

cases, blood was drawn for serum glucose, electrolyte, blood urea
nitrogen, creatinine, and liver function determinations and coag- Statistical Analysis
ulation studies if indicated. Data were edited by the principal investigator. The data were
Blood specimens for complete blood cell count and platelet double-entered into an IBM-compatible computer by using the Epi
count were sent to the university laboratory for quality control. Info proamfl and checked for entry error. All analyses were
Blood specimens for viral isolation were collected by standard conducted using the SPSS PC+ statistical package. The case-
methods.8 The collected blood was centrifuged immediately to fatality rate and other dichotomous measures were compared by
separate cells from plasma and then kept frozen (-70#{176}C).These x2 test. If the minimum expected frequency requirements for the x
blood specimens were transported in a liquid nitrogen tank to the test were not met, the Fishers Exact Test was used instead. For
Medical Science Department in Bangkok. Blood for dengue anti- continuous data, a Students unpaired t test was used.
body tests were put on two or three pieces of standard filter paper.
These samples were kept at room temperature and then sent to the RESULTS
Medical Science Department and to the US Armed Forces Research
Fairly early in the trial it became apparent that the
Institute of Medical Science in Bangkok for confirmatory antibody
testing. The antibody tests used were hemagglutination inhibi- overall mortality rate was substantially lower than
tion9 and enzyme-linked immunosorbent assay (ELISA).2#{176}All that assumed in the sample-size calculation. Follow-
specimens were identified only by code number. ing the 1987 dengue season, a decision was made to
continue recruitment during the next available peak
period, which turned out to be the following year, as
Study Drugs. Placebo was 5% dextrose in normal saline solu-
a result, a total of 63 subjects were enrolled. There
lion. It was prepared by the hospital pharmaceutical unit. Each
vial contained 8 mL, which is equal to the amount of active drug.
were no patients admitted with the diagnosis of 055
Methylprednisolone sodium succinate (Solu-Medrol, Upjohn) who did not have
a laboratory diagnosis of recent
was given in a single bolus dose of 30 mg/kg. Upon recruitment dengue infection. The decision to consider recruit-
of an eligible case, one of the investigators (A.C.), who was not ment closed at this point was made independently
involved with subsequent patient care and evaluation, prepared
from observed treatment differences and based pri-
the study medication by mixing the appropriate powder with
solvent. The liquid that resulted from active powders and the marily on power considerations and the expectation
liquid of placebo were indistinguishable with respect to color, that another major dengue epidemic would not occur

112 DENGUE SHOCK SYNDROME Downloaded from by guest on October 25, 2016
for at least 2 years. All surviving cases were con- TABLE 2. Clinical Course of All the Patients*
firmed as dengue infection from paired sera tested MP Placebo
for hemagglutination inhibition and ELISA antibod- Group Group
ies. Diagnosis of eight fatal cases was established (n=32) (n=31)
from a single blood specimen by the 1gM-capture Hematocrit, % (mean SEM)
ELISA test. In one fatal case dengue type 3 virus was Maximum 48 5.5 47 6.4
isolated. Minimum 295.6 304.9
No. (%) needing transfusion 11 (34.4) 8(25.8)
Comparability of the Study Groups No. (%) with encephalopathy 15 (46.9) 11 (35.5)
No. (%) with liver failure 8 (25) 6(19.4)
Thirty-two patients were randomized to the meth- No. (%) with prolonged PT and PT! 11 (34.4) 8(25.8)
ylprednisolone group and 31 to the placebo group. Duration of hospitalization, d
The two groups were comparable at the time of entry All cases

into the study (Table 1). The exact duration of shock n 32 31

Mean SEM 7.3 1.20 6.2 0.71
was unknown in 21 patients (10 methylprednisolone, Range 2-38 1-21
11 placebo) and only an estimated duration was Only survivors
available. The duration of shock was known accu- n 28 27
rately if it developed while the patient was in the Mean SEM 7.6 1.32 6.7 0.76
hospital. For children admitted with shock, duration Range 3-38 4-21
Only death
was estimated based on the occurrence of symptoms
n 4 4
as reported by the parents. Shock duration is thus Mean SEM 5.2 2.59 2.5 0.65
summarized separately according to the certainty of Range 2-13 1-4
the history. The steroid group had slightly more se- * There was no significant difference between the two groups (P>
vere grade shock patients, but the mean durations of .2 for all parameters). Abbreviations: MP, methylprednisolone; VF,
shock were quite similar. prothrombin time; PTT, partial thromboplastin time.

Response to Methyiprednisolone
Duration of Hospitalization. Duration of hospital-
Case Fatality. Four
of the 32 patients who received
ization of the two groups was compared. Each treat-
methylpredmsolone and 4 of the 31 patients who
ment group was subdivided into dead and survivor
received placebo died. The 12.5% case-fatality rate in
subgroups (Table 2). There was no significant differ-
the methylprednisolone group was not significantly
ence between the two groups in total or in the groups
different from the 12.9% case-fatality rate in the pla-
of surviving patients.
cebo group (P = .63). The 95% confidence interval for
the difference in mortality rates between steroid and Complications During Treatment
placebo groups was -17% to + 16%.
Complications during treatment are summarized
Clinical Various aspects
Course. of the clinical
in Table 3 and show no real differences between the
course of patients studied are shown in Table 2. two groups.
There was no significant difference in any of these
parameters. Sequelae at 2 Weeks After Treatment
All survivors, 28 in the methylprednisolone group
TABLE 1. Comparability of the Stu dy Groups on Entry and 27 placebo patients, were assessed at 10 to 14
MP Placebo
TABLE 3. Complications During Treatment and Sequelae at 2
Group Group
(n=32) (n=31) Weeks After Treatment of the Two Groups*

Age, yr No. (%) of Patients

Mean SEM 7.0 0.50 6.2 0.33
MP Placebo
Range 3-14 3-10
Ratio of boys to girls 13:19 19:12
Group Group
Ratio of referred to nonreferred 16:16 15:16 Complications during treatment n = 32 n = 31
Severity Temperature rising after deferves- 16 (50) 15(48)
Grade3 18 20 cence
Grade4 14 11 Pneumonia 6 (19) 5(16)
Duration of shock before entry, h Convulsiont 3 (9) 0(0)
Overall Pulmonary hemorrhage I (3) 1 (3)
n 32 31 Positive
hemoculture I (3) 1 (3)
Mean SEM 19.62.14 19.8 1.47 Sequalae at 2 wk after treatment n = 28 n = 27
Range 4-62 7-40 Hematoma 5(17.9) 3(11.1)
Only certain history Stiff joint 2 (7.1) 4(14.8)
n 22 20 Otitis media I (3.6) 1 (3.7)
Mean SEM 20.9 2.94 20.1 2.01 Abscess at cut-down area I (3.6) 1 (3.7)
Range 5-62 7-40 Gingivitis I (3.6) 0(0)
Only uncertain history * There was no significant difference in type of complications and
n 10 11
sequelae rates. P value was more than .3 in all factors by Fisher
Mean SEM 16.7 2.20 19.4 2.10
exact, one-tailed, test.
Range 4-24 7-26
t Convulsion was observed only in the methyiprednisolone (Ml)
* There was no significant difference between the groups for any group, but there was no significant difference from the placebo
characteristic. M1 methyiprednisolone. group (P = .12 Fisher exact, one-tailed, test).

Downloaded from by guest on October 25, 2016 ARTICLES 113

days after treatment initiation. The distribution of prednisolone may increase the risk of convulsion.
sequelae, given in Table 3, shows no significant dif- Convulsions were observed in three patients in the
ferences between the two groups. methylpredmsolone group (9%) and none in the pla-
cebo group. These differences were not statistically
DISCUSSION signfficant. Convulsions were also observed in 21%
Results of our study indicate that methylpredniso- in one study of all other steroid trials.8 But the pres-
lone in a single bolus dose of 30 mg/kg had no effect ence of convulsions was reported for all study pa-
on the mortality rate in children with severe 055. The tients. Therefore, the clinical relevance of these phe-
mortality rate in the methylpredmsolone and placebo nomena is unknown.
groups was 12.5% and 12.9%, respectively. The ob- Steroid administration has also been studied at an
served treatment difference in mortality rate of 0.4% earlier stage of DHF, when shock first develops. Four
is consistent with a true treatment effect of between studies have been reported in the English language
16% in favor of control and 17% in favor of steroids. literature.379 Randomization was applied in three of
Thus, although the observed results showed no evi- four studies. Double-blind techniques were used in
dence of a mortality advantage for steroids, the as- two studies. The type of steroid was hydrocortisone
sociated confidence interval is wide. This is mainly hemisuccinate, which has a potency one-fifth that of
because observed mortality rates in both groups methylprednisolone?- The studies show conificting
were much lower than anticipated. The study has results (Table 4).
less power than planned, and so we must be cautious Only one study indicated a benefit of hydrocorti-
in ruling out an important mortality-reducing effect sone hemisuccinate.9 Even though the design of this
of steroids. However, the lack of any observed mor- study was appropriate, its results were not replicated
tality trend in favor of steroids coupled with their in later reports. One possible explanation is that the
known adverse side effect profile makes it unlikely improved supportive care of the later studies re-
that future studies will prove them to be clinically duced the impact of steroids. We conclude from those
useful in severe dengue shock. studies that steroids have no benefit in early shock.
Our results are in direct contrast to those of a pre- Today few pediatricians use low-dose or any steroid
vious study, which showed a dramatic reduction in in early shock, but some still question whether the
the mortality rate with methylprednisolone.#{176} The use of high-dose steroids would be of benefit.
contrast may be due to differences in patient selec- Steroids have also been advocated as having a role
tion and the lack of double-blind techniques. Al- in the febrile stage of DHF. Dexamethasone, in tablet
though mortality is a hard outcome and thus not form, is thought to bring the fever down more rap-
subject to biased ascertainment, lack of blinding may idly and also prevent shock, through the inhibition of
have resulted in more supportive care being giving the antigen-antibody reaction and reduced vascular
to the methylprednisolone group. Because support- leakage if given before shock. No study has been
ive care has a crucial role in this type of ifiness, conducted to address this issue. In designing such a
imbalance in supportive care may well have biased study, one must be aware that many other febrile
the earlier study in favor of methylprednisolone. patients might be wrongly induded in a presumed
The clinical course of the two groups was also DHF population. If those febrile illnesses were due to
comparable in terms of transfusion needed, enceph- bacterial or systemic viral infection, complications
alopathy, liver failure, and coagulation defects (by could be high.
prolonged prothrombin time and partial thrombo- In summary, our study showed that the methyl-
plastin time), indicating that methylpredmsolone did prednisolone (high-dose steroid) has no role in treat-
not help convey a qualitative advantage in survivors. ing severe dengue shock syndrome. However, its role
It was anticipated that the steroid group would in mild shock has not been carefully studied. A liter-
have more complications, especially infection.24 In ature review suggests that hydrocortisone hemisuc-
our study, the frequency of episodes of temperature cinate (low-dose steroid) also has no benefit in early
rising after defervescence, pneumonia, positive shock. Supportive care is crucial in managing DHF
hemoculture, and pulmonary hemorrhage were com- patients, but pediatricians must stifi search for a ther-
parable. However, our results suggest that methyl- apy that will prevent death in severe cases.

TABLE 4. Critical Appraisal o f Steroid Use in Early Dengue Shock Syndrome*

Year Principal Study Blinding Treatment Result Case-

Reported Investigator Type (mg/kg/d) Fatality
Died Lived Total Rate, %

1973 Pongpanich7 R . . . HH (25) 0 7 7 0

Pb 0 19 19 0
1975 Sumarmo8 . . . . . . HH (25) 6 12 18 33

Pb 4 6 10 40
1975 Min9 R DB HH (25) 9 39 48 19
Pb 22 28 50 44
1982 Sumarmo R DB HH+ 8 39 47 17
Pb 9 41 50 18
* Abbreviations: R, randomized; DB, double-blind; Pb, placebo; RH, hydrocortisone hemisuccinate.
t 50 mg/kg per dose.

114 DENGUE SHOCK SYNDROME Downloaded from by guest on October 25, 2016
ACKNOWLEDGMENTS 11. Friedman LM, Furberg CD, DeMetS DL, eds. Fundamentals of Clinical
Trials. 1st ed (3rd printing). Littleton, MA: John Wright PSG mc; 1983
This work was supported by the Rockefeller Foundation (grant
12. CliniCal trials. In: Fletcher RH, fletcher SW, Wagner EH, eds. Clinical
RF 87006#95), Khon Kaen Provincial Hospital, Ministry of Public
Epidemiology: The Essentials. Baltimore MD Williams & Wilkins;
Health and Khon Kaen University, Thailand.
We thank Dr Scott B. Haistead, virologist, of the Rockefeller
13. Randomized clinical trials. In: Feinstein AR, ed. Clinical Epidemiology:
Foundation, and Professor Robin S. Roberts, biostatistician, of Mc-
Master University for their advice during the design and execu- The Architecture of Clinical Research. Philadelphia, PA: WB Saunders
tion of this trial; the medical and nursing staffs of Khon Kaen Company; 1985:683-718
Provincial Hospital and Khon Kaen University Hospital; the phar- 14. Hoffman SL, Punjabi NH, Kumala S, et al. Reduction of mortality in
macists of the Khon Kaen Provincial Hospital; and Dr Bruce L. chioramphenicol-treated severe typhoid fever by high-dose dexameth-
Innis and Dr Anant Nisaluk of the US Armed Forces Research asone. N EngI I Med. 1984310:82-88
Institute of Medical Science in Bangkok for their laboratory sup- 15. Plum F, Posner JB, eds. The Diagnosis of Stupor and Coma. 3rd ed. Phil-
port. adeiphia, PA: PA Davis; 1980-.1-5
16. Hepatic encephalopathy. In: Colon AR, ed. Textbook of Pediatric Hepatol-
REFERENCES ogy. 2nd ed. Chicago, IL: Year Book Medical Publishers mc; 1990233-
1. Dengue Hemorrhagic Fever: Diagnosis, Treatment and Control. Geneva, 17. Fulminant hepatic failure.ed. Liver DisOrders in Childhood.
In: MowatAP,
Switzerland: World Health Organization; 1986:1-2 2nd ed. London, England: & Co Ltd; 1987:126
2. Leads from the MMWR. Dengue and dengue hemorrhagic fever in the 18. Tesh RB. Amethod for the isolationand identification of dengue viruses,
Americas, 1986. JAMA. 1988;259:1781-1782
using mosquitos cell cultures. Am I 1)vp Med Hyg. 197928:1053-1056
3. Sumarmo, Tologo W, Asrin A, Isnuhandojo B, Sahudi A. Failure of
19. Clarke DH, Casals J. Techniques for hemagglutination and hemagglu-
hydrocortisone to affect outcome in dengue shock syndrome. Pediatrics.
tination inhibition with arthropod-borne viruses. Am J Trop Med Hyg.
4. Zuner RB, Weissman G. Anti-immunologic
and anti-inflammatory ef-
20. Innis BL, Nisalak A, Nimmannitya S. et aL An enzyme-linked immun-
feds of steroid therapy. In: Azanoff DL, ed. Steroid Therapy. Philadelphia
osorbent assay to characterize dengue infections where dengue and
PA: WB Saunders Company; 1975223-237
5. Kehrl JH, Fauci AS. The clinical use of glucocorticoids. Ann Allergy.
Japanese encephalitis co-circulate. Am I lhap Med Hyg. 1989;40:418-427
21. Robson AM. Parenteral fluid therapy. In: Behrman RE, Vaughan VC Ill,
ads. Nelson TextbookofPediatrics. 13th ad. Philadelphia, PA: WBSaunders
6. The endocrine system. In: Rang HP, Dale MM, eds. Pharmacology.
Singapore: Longman Singapore Pte Ltd; 1987:395-398
Publishers Company; 1987:191-192
7. Pongpanich B, Bhanchet P. Phanichyakarn P, Valyasevi A. Studies on 22. Dean AG, Dean JA, Burton AH, Dicker RC. Epi Info, Version 5: A word
dengue hemorrhagic fever. Clinical study: an evaluation on steroids as processing, database, and statistics pi ram for epidemiology on ml-
a treatment. Med Max Thai. 197356:6-14 crocomputers. Stone Mountain; GA: USD Incorporated: 1990
8. Sumarmo, Widya MS. Martoatmodjo K. Clinical observations on den- 23. SPSS Usefs Guide. 2nd ed. New York, NY: McGraw-Hill Book Corn-
gue shock syndrome: an evaluation of steroid treatment. Paediatr in- pany; 1986
dones. 1975;15:151-160 24. Marble A, Weir GC, Selenkow HA, et aL Endocrine diseases. In: Speight
9. Mm M, Tin U, Aye M, Shwe TN, Swe 1, Aye B. Hydrocortisone in the TM, ed. Averys Drug Treatments Principles and Practice of Clinical Phar-
management of dengue shock syndrome. Southeast Asian I Trop Med macology and Therapeutics. 3rd ed. Aucidand, New Zealand: AIDS Press
Public Health. 1975;6:573-579 Limited; 1987:569
10. Futrakul P, Vasanawathana S, Poshyachinda M, Mitrakul C, Cherdboon- 25. Haynes RC Jr. Murad F. Adrenocorticotropic hormone. lit Gilinan AG,
chart V. Kanthirat V. Pulse therapy in severe form of dengue shock Goodman LS, eds. The Phar,mcological&sis ofTherapeutics. 17th ed. New
syndrome. I Med Assoc Thai. 1981;64:485-491 York, NY: McMillan Publishing Co; 1985:1475



RENTON, Wash.-At the drive-up window at the McDonalds restaurant in this

Seattle suburb, 17-year-old Brandon MacRae is showing grace under pressure. He
greets customers with a bright Welcome to McDonalds and deftly fills their
orders while keeping an eye on hash browns sizzling in a nearby fryer.
To his bosses, Brandon is a model employee. But to state labor officials, he is an
overworked teen-ager whose 30-hour work-week may cut into school timethis fall.
For this reason, state labor officials plan to adopt regulations that would create the
nations strictest child labor law ...

If the regulations are adopted as written, 16- and 17-year-olds will be limited to
20-hour workweeks during the school year, rather than the 40 now permissible
under state law; and 14- and 15-year-olds wifi be limited to 15-hour workweeks
during the school year, rather than the 18 now permitted.

Child labor law to be tightened. San Jose Mercury News. August 17, 1992.

Noted by B.H., MD

Downloaded from by guest on October 25, 2016 ARTICLES 115

Failure of High-Dose Methylprednisolone in Established Dengue Shock Syndrome: A
Placebo-Controlled, Double-Blind Study
Sompon Tassniyom, Sirijitt Vasanawathana, Aroon Chirawatkul and Suntharee
Pediatrics 1993;92;111
Updated Information & including high resolution figures, can be found at:
Services /content/92/1/111
Citations This article has been cited by 14 HighWire-hosted articles:
Permissions & Licensing Information about reproducing this article in parts (figures, tables)
or in its entirety can be found online at:
Reprints Information about ordering reprints can be found online:

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it
has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the
American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007.
Copyright 1993 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005.
Online ISSN: 1098-4275.

Downloaded from by guest on October 25, 2016

Failure of High-Dose Methylprednisolone in Established Dengue Shock Syndrome: A
Placebo-Controlled, Double-Blind Study
Sompon Tassniyom, Sirijitt Vasanawathana, Aroon Chirawatkul and Suntharee
Pediatrics 1993;92;111

The online version of this article, along with updated information and services, is located on
the World Wide Web at:

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication,
it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked
by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village,
Illinois, 60007. Copyright 1993 by the American Academy of Pediatrics. All rights reserved. Print
ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from by guest on October 25, 2016