Clinical Trial of Glucose-Oral Rehydration Solution (ORS), Rice Dextrin-ORS, and Rice Flour-ORS for the Management of Children With Acute Diarrhea and Mild or Moderate Dehydration

(1)

Clinical

Trial

of

Glucose-Oral

Rehydration

Solution

(ORS),

Rice

Dextrin-ORS,

and

Rice

Flour-ORS

for

the

Management

of

Children

With

Acute

Diarrhea

and

Mild

or Moderate

Dehydration

Susana Molina, MD*; Carolina Vettorazzi, MD*; Janet M. Peerson, MS;

Noel W. Solomons, MD*; and Kenneth H. Brown, MD

ABSTRACT. Objective. To assess the effects of

glu-cose (G)-oral rehydration solution (ORS), rice dextrin

(RD)-ORS, and rice flour (RF)-ORS on fluid intake,

ra-pidity of rehydration, and stool output of children with acute diarrhea and mild or moderate dehydration.

Methods. The study was a randomized,

double-masked clinical trial. One hundred forty-six male infants, ages 3 to 36 months, were randomly assigned to one of three treatment groups. Clinical evaluations and fluid balances were conducted every 2 to 4 hours for 48 hours. Principal outcome variables were ORS consumption, re-covery of hydration status, and fecal output.

Results. The groups were similar at admission with

regard to age, nutritional status, history of the current episode, and clinical status. There were no differences in

ORS consumption by treatment group during any period

of study. During the first 6-hour period, patients in group

RF had less stool output (16 ± 14 g/kg/body weight) than those in group G (22 ± 20 g/kg) or RD (21 ± 19 g/kg P < .05). After 12 hours of hospitalization, there were no differences by treatment group. Recovery of hydration status, changes in serum sodium and potassium, and duration of diarrhea in the hospital were similar in all

three groups.

Conclusion. There was a 24% to 27% reduction in

stool output during the first 6 hours of treatment among children who received RF-ORS compared with those

who received G-ORS or RD-ORS, but this effect did not

persist after the first 12 hours of therapy. Because this

difference was of small magnitude and limited

dura-tion, it has minor clinical importance. Thus, we

con-dude that the three solutions had similar efficacy for children with acute, watery diarrhea and mild or

mod-erate dehydration. Pediatrics 1995;95:191-197; diarrhea,

rehydration therapy, electrolytes, oral rehydration so-lution, clinical trial.

Therapy for diarrheal dehydration consists of

pro-viding fluid and electrolytes either orally or

intrave-nously, both to reverse existing dehydration and to

replace ongoing fecal losses. Most diarrheal disease-control programs rely primarily on oral rehydration

solutions (ORS) composed of glucose or another

car-bohydrate and electrolytes for the treatment of

de-hydrated infants because of the low cost and ease of

preparation and administration of these solutions.

Glucose-ORS (G-ORS) remains the standard therapy

for dehydrated children whose circulatory systems

and ability to drink have not been compromised by

the underlying enteric illness.1

Several studies have found that ORS prepared

from rice flour (RF-ORS) decrease fecal purging rates

and shorten the duration of diarrhea compared with

G-ORS in children with very high stool outputs.26

More recently, Pizarro et al7 compared the efficacy of

ORS prepared from partially hydrolyzed rice starch

(rice dextrin [RDJ-ORS) with G-ORS, and found

de-creased stool output and greater retention of water

and potassium during the first 6 hours of therapy in

children who received RD-ORS. There were no

differ-ences in these outcomes during later time periods.7

However, uncertainty remains regarding the

rela-tive benefits of RD-ORS, G-ORS, and RF-ORS

be-cause other studies reported no differences in ORS

intake, fecal output, and weight gain when

malto-dextrins were used instead of glucose.8 To reassess

the relative advantages of G-ORS, RD-ORS, and

RF-ORS in children with acute diarrhea and mild or

moderate dehydration, we conducted a randomized

clinical trial comparing each of these solutions. The

results of this trial are reported herein.

ABBREVIATIONS. ORS, oral rehydration solution; G-ORS,

glucose-ORS; RD-ORS, rice dextrin-ORS; RF-ORS, rice flour-ORS;

SD, standard deviation.

From the *Center for Studies of Sensory Impairment, Aging, and

Metabo-lism (CeSSIAM), Research Branch for the National Committee for the

Blind and Deaf, Guatemala City, Guatemala, Central America, and the

Program in International Nutrition, Department of Nutrition,

Univer-sity of California, Davis, CA.

Received for publication Mar 21, 1994; accepted May 27, 1994.

Reprint requests to (K.H.B.) Department of Nutrition, University of

California, Davis, CA 95616-8669.

Acad-emy of Pediatrics.

Study Site and Subjects

METHODS

The study was designed as a randomized, double-masked

din-ical trial. One hundred forty-six male infants, ages 3 to 36 months,

with acute diarrhea and mild or moderate dehydration were

hos-pitalized in the study ward of the San Juan de Dios General

Hospital in Guatemala City. Only males were enrolled to facilitate

separate, quantitative collections of stool and urine during the

observation period. Diarrhea was defined as excretion of three or

more liquid stools during the 24 hours before admission. The

degree of dehydration was classified according to clinical signs, as

recommended by the World Health Organization.9 Children with

severe dehydration, systemic infection (pneumonia, meningitis,

sepsis), dysentery, or weight for length less than -2.5 standard

deviations (SD) with respect to the National Center for Health

Statistics reference median were excluded from the trial. Children

at Viet Nam:AAP Sponsored on September 1, 2020

www.aappublications.org/news

(2)

who received prior treatment with antibiotics were accepted into

the protocol. Written informed consent was obtained from the

patients’ parent or guardian before admission. The research

protocol was approved by the Human Rights Committee of the

Center for Studies of Sensory Impairment, Aging, and Metabolism

and by the Research Committee of the San Juan de Dios General

Hospital.

The initial sample size estimates indicated that 42 patients were

needed per group to detect a treatment failure rate of 25%,

assum-ing a 10% treatment failure rate in the G-ORS group; 43 patients

were needed per group to detect 35% differences in ORS intake or

stool output (alpha = 0.05, beta = 0.80).’#{176}

Initial Examination, Randomization, and

Rehydration Solutions

Immediately after admission, a clinical history and physical

examination were completed. The degree of dehydration was

assessed clinically, and the patients were stratified by

breast-feeding status and degree of dehydration (mild or moderate).

Subjects were then assigned to one of three treatment groups

using a variable-length, block-randomization scheme. Children in

group G received G-ORS, children in group RD received RD

(rice-syrup solids)-ORS (purchased commercially as Ricelyte,

Mead-Johnson, Evansville, IN), and those in group RF received

RF-ORS. All of the solutions were prepared previously in liquid

form and contained 50 mmol/L sodium, 25 mmol/L potassium,

35 mmol/L citrate, and 40 mmol/L chloride. The G-ORS

con-tamed 25 g/L glucose, the RD-ORS included 30 g of rice-syrup

solids, and the RF-ORS contained 37.5 g/L freshly cooked rice

flour. The respective osmolalities of the prepared solutions, as

measured by freezing point depression, were 332 mOsm/L for

G-ORS, 218 mOsm/L for RD-ORS, and 133 mOsm/L for RF-ORS.

The amount of ORS offered was calculated to replace the

ex-isting fluid deficit, as estimated from the initial body weight and

the clinically determined degree of dehydration. The respective

solutions were given in coded, opaque feeding bottles in an

at-tempt to prevent the clinical staff from identifying the individual

solutions. The G-ORS and RD-ORS were, in fact, visually

indis-tinguishable. However, because of the difference in physical

ap-pearance of the RF-ORS, the staff could identify this solution when

it was regurgitated or vomited.

Rehydration fluid therapy was offered during the first 4-hour

period. When a patient did not show evidence of improving

hydration status, such as weight gain, reduction in clinical signs,

or decreased urine specific gravity within the first 4 hours,

addi-tional rehydration fluids were offered according to a revised

assessment of the current degree of dehydration. If a patient

vomited more than three times during the first 4-hour period, a

nasogastric tube was used to administer fluid.

Fluid Replacement Therapy

The patients were reevaluated clinically and fluid balances

were calculated at 4, 6, and 8 hours after admission and every 4

hours thereafter. The children were weighed nude on an electronic

scale (Detecto, Webb City, MO) with a capacity of 15 kg and a

precision of 5 g. All stools and urine excreted were collected

separately, using metabolic beds. Pre-weighed clothing and

tow-els were used to absorb any vomitus, and the differences in weight

were measured using an electronic scale (Ohaus, Florham Park,

NJ), with a 1000-g capacity and 2-g precision, to estimate the

amount of vomiting. All stools and vomitus excreted after

admis-sion were replaced volume-for-volume using the same oral

solu-tion that was assigned randomly for initial rehydration. Fluid

replacements were offered in addition to the maintenance diets.

Maintenance Therapy and Antibiotic Treatment

Breast-fed infants were allowed to continue breast-feeding once

the initial rehydration therapy was completed. Non-breast-fed

and partially breast-fed patients received a soy-based formula

containing 0.67 kcal/mL (Isomil, Ross Laboratories, Columbus,

OH), which was offered at a maximum of 125 kcal/kg body

weight/d (187 mL/kg body weight/d) and was provided after the

ORS was consumed during each observation period. If after 16

hours of therapy, the formula intake was less than 33% of the

offered volume and the hydration status was not improving

do-spite appropriate intake of ORS, a nasogastric tube was used to

deliver the formula. Breast-fed infants were allowed to nurse freely

according to their demand and their mothers’ wishes. Patients were discharged from the research unit after 48 hours of study.

Antibiotics were used only when children had localized

infec-tions such as otitis media (n = 5) or pneumonia without

respira-tory distress identified after admission (n = 3). Children with

positive stool cultures for Shigella spp received antibiotics at the

time of discharge from the study, which is when the laboratory

results first became available.

Treatment Failures

Once rehydration therapy was successfully completed, children

with either a recurrence of dehydration estimated to be greater

than 5% of body weight or fecal output greater than 350 g/kg

body weight/d were classified as having a failure of therapy.

Those who were not fully rehydrated within the first 24 hours of

therapy also were considered to have failed treatment. All patients

with treatment failure were removed from the study protocol and

given intravenous therapy.

Laboratory Analyses

Blood was drawn at admission and after 12 and 48 hours of

treatment. Serum sodium and potassium were determined with

an Electrolyte Analyzer Photometer (AVL, Switzerland). Serum

bicarbonate values were analyzed using the Special Chem Micro

CO2 System (Baxter Health Care Corporation, Miami FL).

Hemat-ocrit was determined using a microhematocnt centrifuge, and

serum specific gravity was measured by refractometry.

Stool specimens were obtained shortly after admission for

mi-crobiologic studies, which were completed at the Institute of

Nu-trition of Central America and Panama in Guatemala City. Fecal

samples were placed in Cary-Blair transport media, veal infusion

broth, and buffered formalin. Pathogenic bacteria were isolated in

XLD, MacConkey, and Butzler Virion media streaked from the

Cary-Blair transport medium. Shigella and other bacteria were

identified according to transport methodology.” Vibrio cholerae 01

was isolated from the samples transported in Cary-Blair medium

after 6-hour incubation in alkaline peptone water. The surface of

the latter enrichment broth was subcultured to thiosulfate citrate

bile salts sucrose (TCBS) agar and identified according to standard

methodology.1’ Campylobacter jejuni was recognized by its

mor-phologic and biochemical characteristics after isolation on Butzler

virion agar.’2 Colonies of Escherichia coli were studied further for

pathogenicity, as follows. Suckling mice’3 and mouse adrenal

tumor cells’4 were used to identify heat-stable (ST) and heat-labile

(LT) toxins of E coli, respectively. Enteropathogenic E coli were

detected serologically with antiserum produced in the Institute of

Nutrition of Central America and Panama, and adherent E coli

were detected in HEp-2 monolayers.’5 Rotavirus was identified in

the veal infusion broth by enzyme-linked immunosorbent assay.’6

Cryptosporidium spp were identified in the formalin-preserved

stools after treatment with 10% KOH, rinsing with 10% formalin,

and separation by centrifugation.’7 Two smears were prepared

from the uppermost layer of the sediment. After ethanol fixation,

the preparations were covered with 1 % safranine, heated with an

open flame for 2 minutes, and rinsed with tap water. Methylene

blue (1 %) was added for 30 seconds. Formalin-preserved stools

also were used to detect protozoal cysts and helminth eggs.

Pro-tozoan trophozoites were investigated in fresh fecal smears.

Statistical Analyses

Each of the variables was explored with descriptive statistics,

and the distributions of continuous variables were assessed for

normality. Baseline variables were compared among treatment

groups, using analysis of variance for continuous variables and

Pearson’s 2 for categorical variables.’0 Logarithmic

transforma-tions were necessary for the ORS intake and stool output data to

satisfy the assumptions of the statistical models, because these

variables were not normally distributed. Intakes of rehydration

fluid and formula diets, stool output, indicators of hydration

status, body weight, hematocrit, serum protein, and serum

elec-trolytes also were analyzed using repeated-measures analysis of

variance with treatment group, breast-feeding status, and time as

fixed main effects and subject as a random main effect, and

in-cluding all two-way interactions. The duration of diarrhea in

the hospital was compared by treatment group using survival

(3)

analyses.’#{176} The Fisher’s exact test was used to determine

differ-ences in the proportion of failures by treatment. In all cases, P

.05 was considered statistically significant.’#{176}

RESULTS

Initial Characteristics of Patients

A total of 146 children were admitted to the study.

Of these, 12 children were excluded from the

analy-ses for the following reasons: absence of diarrhea

during the first 24 hours of observation (n = 4), later

appearance of sepsis not recognized at admission

(n = 3), persistent abdominal distention (n = 2),

history of chronic diarrhea not reported at

admis-sion (n = 1), poor adaptation to the metabolic bed

(n = 1), and withdrawal of parental consent (n = 1).

There were no differences in the total numbers of

exclusions or reasons for exclusion by treatment

group.

After these exclusions, 134 children were available for the final analyses (n = 43, group G; n = 46, group

RD; n = 45, group RF). The initial characteristics of

the children are described by treatment group in

Table 1. The groups were similar with regard to the

patients’ ages, weights, lengths, duration of diarrhea

before hospitalization, and indicators of severity of

diarrhea and dehydration. Overall, the children were

moderately stunted and wasted compared with

United States reference data. The patients in group

RF had slightly greater weight for age and weight for

length 2 scores than their counterparts in group G,

although these differences were not statistically sig-nificant (P = .07). The children in group RD

report-edly vomited slightly more often on the day before

admission, but again these differences were not

sta-tistically significant (P = .09). In addition to the items

shown in Table 1, there were no differences by

treatment group in rates of fever, reported poor

appetite, or use of medications during the week

before admission.

Table 2 shows the pathogens recovered from

ad-mission stool specimens. Rotavirus was isolated

most frequently, followed by the combined category

of enteropathogenic and enterotoxigenic E coil.

En-teroadherent E coil, either alone or in combination

with other microbial agents, were found in a total of

27 samples with the following adherence patterns:

local, n = 2; diffuse, n = 8; and autoaggregative,

n = 17. Eight children had enteropathogenic E coil

serogroup 0111, five had serogroup 086, and one

child each had serogroups 018, 044, 1 19, or 128. The

eight children with enteropathogenic E coil

sero-group 01 1 1 also had an autoaggregative adherence

pattern in the HEp-2 assay, three with serogroup 086

also had a diffuse or autoaggregative adherence

pat-tern, and the one with serogroup 0128 had a diffuse

adherence pattern. Ten children had enterotoxigenic

E coil that produced LT and none had ST-producing

enterotoxigenic E coil. No differences were found in

the distribution of any enteropathogens by treatment group.

Consumption of ORS and Formula and Stool Output

Table 3 provides data for the total intake of ORS

and the maintenance formula for all patients in each

study group for the observation periods 0-6, 6-12,

12-24, and 24-48 hours. Approximately 90% of the

initially allotted rehydration solutions were

con-sumed within the first 6 hours of hospitalization, and

69% of the patients consumed all their assigned

re-hydration fluid within this period. No differences in

consumption of rehydration solution by treatment

group were noted. Likewise, there were no

signifi-cant differences in intakes of either replacement fluid or total ORS (rehydration plus replacement) by

treat-ment group during any period. Four children

re-ceived their rehydration fluids by nasogastric tube

because of persistent vomiting. These children were

distributed in each of the three treatment groups.

The mean ± (SD) formula intakes during the first

24-hour period were 26 ± 35 mL/kg body weight for

breast-fed children and 77 ± 41 mL/kg body weight

for those who were not breast-fed (P < .001). No

differences were seen in formula intake by treatment group, regardless of breast-feeding status. Five

chil-dren (three in group G and one in each of the

re-maining groups) received the formula by nasogastric

tube because of poor intakes. This difference in the

use of nasogastric tubes for formula feeding was not

statistically significant.

TABLE 1. Admission Characteristics of Pat ients by Treatment Group*

Characteristic Dietary Group Significancet

G-ORS (n = 43) RD-ORS (n = 46) RF-ORS (n = 45)

Age, mo

Weight, kg

Length, cm

Weight for length, Z score

Length for age, Z score

Weight for age, Z score

Duration of diarrhea before admission, h

No. of stools, previous 24 h

No. of vomits, previous 24 h

Degree of dehydration

Mild (1-5%)

Moderate (>5%)

13.5 ± 6.5 7.86 ± 1.75

72.1 ± 6.9 -1.45 ±0.90 -1.66 ± 1.24 -2.21 ± 1.11

43 ± 21

10 ± 6 3.9 ± 3.5

28 (65%) 15 (35%)

12.0 ± 7.1 7.69 ± 1.71 70.8 ± 7.7

-1.16 ± 0.86

-1.44 ± 1.04

-1.85 ± 0.88

36 ± 22

11 ± 8

5.7 ± 7.2

27 (59%) 19 (41%)

11.3 ± 6.2 7.72 ± 1.85 70.1 ± 6.9 -0.99 ± 0.98 -1.44 ± 1.17

-1.70 ± 1.14

43 ± 24

10± 6 3.4 ± 4.2

28 (62%) 17(38%)

.292 .892 .405 .067 .594 .074 .242 .519 .086

.822

* Abbreviations: G-ORS, glucose-oral rehydration solution; RD-ORS, rice dextrin oral rehydration solution; RF-ORS, rice flour-oral

rehydration solution.

t Analysis of variance unless otherwise indicated.

:1:

Chi-square.

(4)

TABLE 2. Pathogens Recovered From Admission Stool Specimens by Treatment Group*

Pathogen Treatme nt Group

G-ORS RD-ORS RF-ORS All

(n = 43) (n = 45) (n = 44) (n = 132)

Rotavirus only 8 (18.6%) 9 (20.0%) 10 (22.7%) 27 (20.5%)

Eschc’richia coli (EPEC or ETEC) only 8 (18.6%) 8 (17.8%) 6 (13.6%) 22 (16.7%)

Shigella only 2 (4.6%) 1 (2.2%) 2 (4.5%) 5 (3.8%)

Cryptosporidium only 2 (4.6%) 2 (4.4%) 2 (4.5%) 6 (4.5%)

Other single infections 2 (4.6%) 1 (2.2%) 2 (4.5%) 5 (3.8%)

Mixed infections 9 (20.9%) 10 (22.2%) 6 (13.6%) 25 (18.9%)

None isolated 12 (27.9%) 14 (31.1%) 16 (36.4%) 42 (31.8%)

* Data are presented as N (%). Abbreviations: EPEC, enteropathogenic E coli; ETEC, enterotoxigenic E Coli; others as in Table 1.

TABLE 3. Intakes of Oral Rehydration Solution (ORS) and Formula and Excretion of Stoo 1 by Time Period and Tr eatment Group*

Type of Fluid Time Period, h Treatment Groupt

G-ORS RD-ORS RF-ORS

ORS intake (rehydration and replacement

fluidscombined)

0-fl 6-12 12-24

24-48

68 ± 33

22±23

30 ± 35

38 ± 39

65 ± 32

21±24

29 ± 32

39 ± 45

62 ± 28

18± 19

23 ± 21

43 ± 36

Formula intakes 0-6

6-12

12-24

24-48

5 ± 11 16±18

34 ± 29

65 ± 54

6 ± 10 12±15

27 ± 27

58 ± 53

5± 8

9±13

32 ± 27 61 ± 50

All fluid intake 0-6

6-12 12-24

24-48

74 ± 39

38±28

64 ± 49

103±71

71 ± 32

34±23 56 ± 39 97±63

67 ± 29

27±22

55 ± 34

104±59

Stool output 0-6

6-12 12-24 24-48

22 ± 20’

14±19” 25 ± 33 35 ± 36

21 ± 19’

1013a,b

24 ± 26

37 ± 41

16 ±

913b

22 ± 20

42 ± 35

* Data are presented as g/kg birth weight/period, in mean ± SD. Abbreviations as in Table 1.

t All variables differed significantly by time. Only stool output differed by treatment group. Values during a single time period with

different superscripts (a, b) were significantly different. See text for details.

: Includes breast-fed and non-breast-fed patients.

The mean fecal excretion rates by study group are

shown in Table 3. Results of the repeated-measures analysis of variance indicated that children in group

RF

had significantly less stool output during the first

6 hours of the study than children in group G

(P = .013) and group RD (P = .008). When group G

was compared to group RD. no difference was found

(P = .76). Children in group RF had reduced fecal

excretion between hours 6 to 12 compared with group G (P = .025), but there were no longer any

differences between groups RF and RD. No

signifi-cant differences by treatment group were observed

thereafter, and there were no significant differences

in cumulative stool output during the entire 48-hour study period by type of ORS received. Breast-feeding

status did not have a significant influence on stool

output and was eliminated from the final, reduced

statistical model.

To assess whether other initial characteristics of

the patients may have explained the differences in

fecal excretion during the first 6 hours, we regressed the log stool output on the initial degree of

dehydra-tion; serum bicarbonate, sodium, and potassium

con-centrations; weight for height and height for age Z

scores; number of bowel movements and number of

episodes of vomiting during the 24 hours before

admission; and treatment group. Factors that were

independently associated with stool output during

the first 6 hours were initial degree of dehydration (P

< .001) and (marginally) initial serum bicarbonate

(P = .064). Controlling for these factors, treatment

group was still a significant predictor of fecal

ex-cretion during this initial period of observation (P = .023).

We also examined the relation between specific

infectious etiologies and both stool output and

con-sumption of ORS. Because of the large numbers of

etiologic subgroups and relatively small numbers of

patients per subgroup, we modeled the types of

in-fections using two different aggregated categories of

infections, while controlling for the type of ORS

of-fered. In the first models, we compared children who

had none, one, or two or more pathogens identified

in their admission stool samples, and we found no

significant relation between the number of infectious

agents recovered and stool output or ORS intake

during any period. In the second set of models, we

compared children who had no infections identified,

those who had rotavirus (with or without other

in-fections), and all others. Rotavirus was marginally

associated with greater stool outputs (P = .071) and

with greater consumption of ORS (P = .077).

(5)

3

7

2 “U-. Group RD

--*-. Group RF

0

.c

a,

0

# 00 8 1:4O48

Hours since admission

Figure. Number of signs of dehydration by oral rehydration

solution (ORS)0-group and time since admission. Abbreviations:

G, glucose-ORS; RD. rice dextrin-ORS; RF, rice flour-ORS.

sion of etiologic categories in the statistical models did not affect the results by treatment group.

Recovery of Hydration Status

The recovery of adequate hydration status was

judged according to the number of clinical signs

present at the end of each period of reevaluation, the

time until disappearance of all clinical signs of

dehy-dration, urine output and specific gravity, and

changes in hematocrit and serum specific gravity.

The numbers (mean ± SD) of clinical signs of

dehy-dration that were present at each assessment time are

shown by treatment group in the Figure. There was

a significant reduction in the number of signs of

dehydration with increasing length of

hospitaliza-tion (P < .001), but no differences among study

groups were noted. The time until disappearance of

all signs of dehydration was rounded up to the end

of the observation period in which these signs were

no longer present. The time to disappearance of these

signs (mean ± SD) was 8.2 ± 4.5, 9.7 ± 10.0, and

8.3 ± 6.1 hours in groups C, RD. and RF, respectively

(P = .72). No differences were observed in the

pro-portions of children in each treatment group who

excreted any urine during the first 4 hours of

hospi-talization, or in the amount of urine excretion or

urine specific gravity at any point.

Group means (±SD) for hematocrit, serum specific

gravity, and serum electrolytes are shown by time

and treatment group in Table 4. There was a slight

but statistically significant reduction in hematocrit after admission (P < .001), but there were no

differ-ences by treatment group. Similarly, there was a

numerically small, but highly statistically significant (P < .001), reduction in serum specific gravity over

time during hospitalization; again, there were no

differences by treatment group. The serum sodium

concentrations decreased and the serum bicarbonate

and serum potassium concentrations increased

dur-ing treatment (P < .001 for each). These changes were similar for each type of ORS provided. Two children

entered the study with mild hypernatremia (serum

sodium 154 mmol/L, group G; 157 mmol/L, group

RD), and both recovered uneventfully within 12

hours. One other child had hypokalemia initially

(serum potassium 2.3 mmol/L, group RD), which

normalized by 12 hours.

Outcomes of Therapy

A total of six patients (4.5%) had treatment fail-ures. These rates of therapeutic failure were similar in all treatment groups (P = .52, Fisher’s exact test).

Three patients who received G-ORS (7.0%) had a

recurrence of dehydration, as did one child who

received RF-ORS (2.2%). Two patients in group RD

(4.3%) failed to achieve rehydration within 24 hours. Most children still excreted liquid or semi-liquid stool by 48 hours of treatment. There were no

differ-ences among the groups with regard to stool

consis-tency at 48 hours, duration of liquid stool excretion, or time to first improvement in stool consistency.

Likewise, there were no differences in changes in

body weight during hospitalization according to the

type of ORS received.

DISCUSSION

This paper reports the results of a randomized,

double-masked clinical trial of three types of ORS,

which were provided to children with acute watery

diarrhea and mild or moderate dehydration. Two of

the solutions, G-ORS and RD-ORS, were

indistin-guishable from each other, whereas the RF-ORS had

a milky appearance. Nevertheless, the strict

proto-cols of the treatment regimens and the predefined,

objective criteria for treatment failure made it

un-likely that any preexisting biases could have

influ-enced the study outcomes, even in the group

receiving RF-ORS.

The characteristics of the patients in each

treat-ment group were generally similar at admission, and

all aspects of therapy were identical except for the

types of ORS provided. Thus, any differences in

the clinical responses can be reasonably attributed

to the effects of the different solutions. Minor,

statistically nonsignificant differences in the

pa-tients’ nutritional status were identified at

admis-sion. Because previous studies have found an

as-sociation between children’s nutritional status and

the severity of purging during acute diarrhea,18 we

included anthropometric indicators as covariates

in the statistical models that related the treatment groups to the major clinical outcomes. Controlling

for these initial characteristics did not alter the

results of the study.

Children in all groups demonstrated similarly

rapid normalization of hydration status; considering

the relatively low stool outputs, there were

appro-priately few treatment failures in all groups.

Like-wise, there were no differences in the consumption

of ORS or of maintenance diets or in the duration of

diarrhea by treatment group. Children who received

RF-ORS had a 24% to 27% reduction in stool output

compared with the other groups during the first 6

hours of therapy. During the next 6-hour period,

stool outputs by patients in group RF-ORS remained

less than those of the patients who received G-ORS,

but there were no longer any differences from those

of group RD. Subsequently, there were no

differ-ences in stool output by treatment group.

(6)

Time Variable

Hematocrit, %

Serum specific gravity

Sodium, mmol/L

Potassium, mmol/L

Bicarbonate, mmol/L

* All variables differed significantly by time (P <.001), but there were no significant differences by treatment group. Abbreviations as in

Table 1.

TABLE 4. Mean Hematocrit and Serum Sodium, Potassium, and Bicarbonate by Time Since Admission and Treatment Group*

0

12

48

0

12

48

0

12

48

0

12

48

0

12

48

G-ORS

37.7 ± 4.1 37.2 ± 4.8 37.2 ± 3.7

1.0504 ± 0.0017 1.0494 ± 0.0015 1.0495 ± 0.0018

137.9 ± 5.9

136.9 ± 4.6

136.7 ± 4.3

4.2 ±0.8

4.7 ± 1.1

4.6 ± 0.7

15.7 ± 4.3 19.0 ± 6.1

19.3 ± 4.8

Treatment Group

RD-ORS

37.9 ± 7.3 37.1 ± 3.9

37.4 ± 4.2

1.0507 ± 0.0018 1.0495 ± 0.0015 1.0493 ± 0.0015

140.1 ± 6.1 136.9 ± 3.9

136.5 ± 3.6

4.4 ± 1.2

4.5± 0.7

4.6 ± 0.6

16.4 ± 4.5

19.2 ± 4.5

19.2 ± 4.6

RF-ORS

37.6 ± 4.2 35.6 ± 3.7

36.5 ± 3.7

1.0502 ± 0.0019 1.0489 ± 0.0022

1.0491 ± 0.0015

138.8 ± 4.7 137.7 ± 4.1 137.3 ± 4.8

4.2±0.7

4.6 ± 0.7

4.7±0.7

15.0± 4.7 19.0 ± 4.9 18.1 ±4.7

Other studies have found consistently lower stool

outputs with RF-ORS compared with G-ORS,

al-though these differences tend to be inversely related to the severity of diarrhea. A recent meta-analysis of studies of children with severe diarrhea associated

with cholera reported 32% reductions in stool

losses during the first 24 hours of treatment with

RF-ORS; however, the reduction in stool output

averaged only 18% with acute, non-cholera

diar-rhea.’ These findings are consistent with the

re-sults of the present study. The clinical importance

of our results for children with mild or moderate

severe diarrhea and dehydration is debatable,

par-ticularly because the apparent early advantage of

RF-ORS

persisted for only a short period and was not

associated with any difference in recovery from dehy-dration, treatment failure rate, or duration of illness. Thus, it seems that each of the three solutions can be considered of similar efficacy for patients with mild or

moderate degree of diarrheal dehydration.

A previous study in which RD-ORS was compared

with G-ORS found a slight reduction in fecal losses

during the first 6 hours with RD-ORS, but not there-after.7 The patients’ initial fecal excretion rates in that

study were nearly twice as great as in the present

one. In the current study, there was no apparent

benefit of RD-ORS compared with RF-ORS or G-ORS

during any period of observation. Factors other than

type of treatment that were associated with diarrheal

severity in the hospital during the present study

were the patients’ initial severity of dehydration and acidosis. These findings are similar to the results of previous studies of children with acute diarrhea.2#{176} In

addition, children with rotavirus tended to have

more severe illness.

The apparent benefit of rice-containing solutions

for reducing stool output has been attributed to their reduced osmolality and to the possible positive effect

of additional organic solutes, such as amino acids

and oligopeptides, on intestinal sodium transport. In

the present study, the osmolalities of the solutions were: RF-ORS, 133 mOsm/L; RD-ORS, 218 mOsm/L;

and G-ORS, 332 mOsm/L. Thus, lower osmolality of

the RF-ORS was consistent with the reduced stool out-put observed in children in this group during the initial 12 hours of observation. The RF-ORS was also the only solution that contained additional organic substrates, but these would seem to have less effect on stool output

than osmolality because most studies have failed to

demonstrate an advantage of adding amino acids to

ORS.2135 Qf the IS studies of ORS containing glycine, glycyl-glycine, or alanine that we located in the pub-lished literature, only thre&1’’33 reported a reduction of

stool output with the addition of an amino acid or

oligopeptide to the treatment solution. In the current

study, the early addition of the maintenance diets,

which obviously contained considerable amounts of

these nutrients, may have overridden any potential

advantage of the rice-containing solutions in this re-gard. One recent study, for example, found no

advan-tage of RF-ORS when rice-containing diets were

of-fered along with G-ORS.

That we found no significant effect of type of ORS

on the duration of diarrhea in the present study was

somewhat surprising, because recent studies have

noted a dramatic reduction in diarrheal duration

when fiber-containing diets were compared with

fi-ber-free ones.374#{176}The formula diets used during the

maintenance phase in the present study contained

practically no fiber, and the RF-ORS was the only

solution that contained small amounts of fiber. The

above meta-analysis of studies of rice-based ORS

found a 12% reduction in diarrheal duration when

these solutions were compared with G-ORS.19

How-ever, the median duration of diarrhea in the groups

receiving rice was 60 hours, which was longer than

the total duration of the current study. Thus, either

our study was too short to detect any modest

reduc-tions in the duration of illness, or the amount of fiber

present in the RF-ORS was too small to produce the

expected changes in stool consistency.

Final decisions on which solution to use for

pa-tients with mild to moderate dehydration in a

par-ticular setting will depend on the cost, availability,

(7)

and ease of preparation of the respective solutions, as well as the site in which treatment is provided.41 The

apparent advantage of RF-ORS, as found in the

present study and in others, must be balanced

against the increased difficulty involved in its prep-aration and its greater risk of microbial

contamina-tion after cooking. On the other hand, the

wide-spread availability of rice in many homes worldwide may increase access to its use. Based on the results of the current study, each of the respective solutions could be used successfully for children with diarrhea and mild or moderate dehydration.

ACKNOWLEDGMENTS

Financial support was provided by the Pediatric Research

Division of Ross Laboratories, Columbus, Ohio.

We are grateful to Drs Rebeca Cueto, Ludwig Ovalle, Sara

Tobar, Erick Rodas, and Juan Carlos Reyes for their care of the

patients and execution of the protocol. Drs Carlos Grazioso and

Manolo Mazareigos of the Center for Studies of Sensory

Impair-ment, Aging, and Metabolism (CeSSIAM) assisted with

supervi-sion of the study and logistical support. Ms Amy Burkhart, Ms

Roberta Hudson, and Ms Mary Guiney assisted with the coding

and preparation of ORS in the metabolic kitchen. Licda. Isabel de

Ramirez of CeSSIAM and Lic. Pablo Yurrita of the Guatemalan

Institute of Social Security performed the biochemical analysis.

For the microbiologic identification of stool pathogens, we

ad-knowledge the assistance of Dr Jose Ramiro Cruz, Licda.

Floridalma Cano, and Licda. Patricia Caceres of the Institute of

Nutrition of Central America and Panama.

REFERENCES

1. World Health Organization. Rice-based ORS. Geneva: World Health

Organization, Programme for Control of Diarrhoeal Diseases, update 7;

1990:1-3

2. Molla A, Ahmed 5, Creenough W. Rice-based oral rehydration solution

decreases the stool volume in acute diarrhoea. Bull WHO. 1985;63:

751-756

3. El-Mougi M, Hegazi E, Calal 0, et at. Controlled clinical trial on the

efficacy of rice powder-based oral rehydration solution on the outcome of acute diarrhea in infants. JPediatr Gastroenterol Nufr. 1988;7:572-576 4. Molla AM, Molla A, Rohde J, Creenough W. Turning off the diarrhea:

the role of food and ORS. JPediatr Gastroenterol Nutr. 1989;8:81-84 5. Ban A, Rahman A, Molla A, Creenough W. Rice-based oral rehydration

solution shown to be better than glucose-ORS as treatment of

non-dysenteric diarrhoea in children in rural Bangladesh. IDiarrhoeal Dis Res. 1989;7:1-7

6. Bhan M, Chai 0, Khoshoo V, et at. Efficacy of mung bean (lentil) and

pop rice based rehydration solutions in comparison with the standard

glucose electrolyte solution. JPediatr Gastroenferol Nutr. 1989;6:392-399

7. Pizarro D, Posada C, Sandi L, Moran J. Rice-based oral electrolyte

solutions for the management of infantile diarrhea. N EngI I Med. 1991;324:51 7-521

8. Santos P, Bravo L, Rogacion J,Battad C. A randomized double-blind

clinical trial of a maltodextrmn-containing oral rehydration solution in

acute infantile diarrhea. JPediatr Gastroenterol Nutr. 1993;16:23-28 9. World Health Organization. Guidelines for Planning Clinical Trials in

Diar-rhoeal Diseases. CDD/CMT 20. Geneva: World Health Organization; 1987 10. SAS/STAT. User’s Guide, Release 6.03 ed. Cary, NC: SAS Institute mc;

1988:641-666

1 1. World Health Organization. Manual for Laboratory Investigation of Acute Enteric Infections. CDD/3. Geneva: World Health Organization; 1983

12. Karmalli MA, Skirrow MB. Taxonomy of genus Campylobacter. In:

Butzler JP, ed. Campylobacter Infection in Man and Ani,nals. Boca Raton,

FL: CRC Press Inc. 1984

13. Morris CK, Merson MH, Sack DA, et at. Laboratory investigation of

diarrhea in travelers to Mexico: evaluation of methods for detecting enterotoxigenic Escherichia coli. IClin Microbiol. 1976;3:486-495

14. Sack DA, Sack RB. Test for enterotoxigenic Eschericlzia coli using Y-1

adrenal cells in miniculture. Infect Immun. 1974;11:334-336

15. Cravioto A, Cross RJ, Scotland SM, Rowe B. An adhesive factor found

in strains of Eschericliia coli belonging to the traditional infantile enter-opathogenic serotypes. Curr Microbiol. 1979;3:95-99

16. Yolken RH, Kim HW, Clem T, et at. Enzyme-linked immunosorbent

assay (ELISA) for detection of human reovirus-like agent of infantile gastroenteritis. Lancet. 1977;ii:263-267

17. Garcia LS, Bruckner DA, Brewer TC, Shimizu RY. Techniques for the

recovery and identification of Cryptosporidium oocyst from stool spec-imens. JCli,, Microbial. 1983;18:185-190.

18. Black RE, Merson MH, Eusof A, et at. Nutritional status, body size and severity of diarrhoea associated with rotavirus or enterotoxigenic

Esch-erichia coli. ITrop Med Hyg. 1984;87:83-89

19. Core SM, Fontaine 0, Pierce NF. Impact of rice-based oral rehydration

solution on stool output and duration of diarrhoea: meta-analysis of 13

clinical trials. BMJ. 1992;304:287-291

20. Lembcke J, Peerson J, Perez F, et al. Factors predicting severity of purging

during acute diarrhea in children (abstract). FASEB I.1992;6:A1650 21. Nalin DR. Cash RA, Rahman M, Yunus MD. Effect of glycine and

glucose on sodium and water absorption in patients with cholera. Gut. 1970;1 I :768-772

22. Patra FC, Mahalanabis D, Jalan KN, et at. In search of a super solution:

controlled trial of gtycine-glucose oral rehydration solution in infantile diarrhoea. Acta Paediatr. Scand. 1984;73:1 8-21

23. Vesikari T, Isolauri E. Clycine supplemented oral rehydration solutions

for diarrhoea. Arc/i Dis Child. 1986;61 :372-376

24. Patra FC, Mahalanabis D, Jatan KN, et at. A controlled clinical trial of

rice and glycine-containing oral rehydration solution in acute diarrhoea

in children. IDiarrlzoeal Dis Res. 1986;4:16-19

25. Santosham M, Burns BA, Reid R, et at. Gtycine-based oral rehydration

solution: reassessment of safety and efficacy. IPediatr. 1986;109:795-801 26. Pizarro D, Levine MM, Posada C, Sandi L. Comparison of glucose/

electrolyte and glucose/glycine/etectrotyte oral rehydration solutions

in hospitalized children with diarrhea in Costa Rica. JPediatr Gastroen-terol Nutr. 1988;7:41 1-416

27. Pizarro D, Posada C, Mahalanabis D, Sandi L. Comparison of efficacy of

a glucose/glycine/gtycylglycine electrolyte solution versus the

stan-dard WHO/ORS in diarrheic dehydrated children. IPediatr

Gastroen-feral Nutr. 1988;7:882-888

28. Antony TJ, Mohan M. A comparative study of gtycine fortified oral

rehydration solution with standard WHO oral rehydration solution.

India,, JPediatr. 1989;26:1 196-1201

29. Bhattacharya 5K, Dutta P, Bhattacharya MK, at at. Efficacy and safety of glycine fortified oral rehydration solution in the treatment of acute dehydrating diarrhoea in children. Indian IMed Res. 1989;90:426-429 30. Bhan MK, Sazawat S. Bhatnagar 5, et at. Glycine, glycyl-glycine and

mattodextrmn based oral rehydration solution. Acta Paediatr Scand. 1990;

79:518-526

31. Khin-Maung-U, Myo-Khin, Nyunt-Nyunt-Wai, Tin-U. Comparison of

glycose/electrotyte and maltodextrin/gtycine/glycyt-glycine/

electrolyte oral rehydration solutions in cholera and watery diarrhoea in

adults. Ann Trap Med Parasifol. 1991;85:645-650

32. Khin-Maung-U, Myo-Khin, Nyunt-Nyunt-Wai, et at. Comparison of glucose/electrolyte and maltodextrin /glycine/glycyl-glycine/

electrolyte oral rehydration solutions in acute diarrhea in children. /

Pediatr Gastroenterol Nutr. 1 991;13:397-401

33. Patra FC, Sack DA, Islam A, et at. Oral rehydration formula containing

alanine and glucose for treatment of diarrhoea: a controlled trial. BMJ.

1989;298:1353-1 356

34. Da Costa Ribeiro H Jr. Lifshitz F. Alanine-based oral rehydration

ther-apy for infants with acute diarrhea. IPediatr. 1991;118:S86-S90 35. Sazawal 5, Bhatnagar 5, Bhan MK, et at. Atanine-based oral rehydration

solution: assessment of efficacy in acute nonchotera diarrhea among

children. / Pediafr Gastroenterol Nutr. 1991;12:461-468

36. Fayad IM, Hashem M, Duggan C, et at. Comparative efficacy of rico-based

and glucose-based oral rehydration salts. Lancet. 1993;342:772-775

37. Alarcon P, Montoya R, Perez F, Dongo J, Peerson JM, Brown KH.

Clinical trial of home available, mixed diets versus a lactose-free, soy

protein formula for the dietary management of acute childhood

diar-rhea. JPediatr Gastroenterol Nutr. 1991;12:224-232

38. Brown KH, Perez F, Castanaduy AS. Clinical trial of modified whole milk, lactose-hydrolyzed whole milk, or cereal-milk mixtures for the dietary management of acute childhood diarrhea. /Pediafr Gastroenterol Nutr. 1991;12:340-350

39. Torun B, Chew F. Recent developments in the nutritional management

of diarrhoea. 3. Practical approaches towards dietary management of

acute diarrhoea in developing communities. Traits R Soc Tro1, Med Hyg.

1991;875:12-17

40. Brown KH, Perez F, Peerson JM, et at. Effect of dietary fiber (soy

polysaccharide) on the severity, duration, and nutritional outcome of

acute, watery diarrhea in children. Pediatrics. 1993;92:241-247

41. Elliott K, Attawelt K, Wilson R, et at, eds. Cereal Based Oral Re/tydration Therapy for Diarrhoea. London: Pegasus Press; I 990

(8)

1995;95;191

Pediatrics

Brown

Susana Molina, Carolina Vettorazzi, Janet M. Peerson, Noel W. Solomons and Kenneth H.

Moderate Dehydration

Rice Flour-ORS for the Management of Children With Acute Diarrhea and Mild or

Clinical Trial of Glucose-Oral Rehydration Solution (ORS), Rice Dextrin-ORS, and

Services

Updated Information &

http://pediatrics.aappublications.org/content/95/2/191

including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or in its

Reprints

http://www.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

(9)

1995;95;191

Pediatrics

Brown

Susana Molina, Carolina Vettorazzi, Janet M. Peerson, Noel W. Solomons and Kenneth H.

Moderate Dehydration

Rice Flour-ORS for the Management of Children With Acute Diarrhea and Mild or

Clinical Trial of Glucose-Oral Rehydration Solution (ORS), Rice Dextrin-ORS, and

the World Wide Web at:

The online version of this article, along with updated information and services, is located on

been published continuously since 1948. Pediatrics is owned, published, and trademarked by the

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has

http://www.aappublications.org/site/misc/Permissions.xhtml

http://www.aappublications.org/site/misc/reprints.xhtml