dextrose saline compared with normal saline.13

Dextrose Saline Compared With NormalSaline Rehydration of HyperemesisGravidarumA Randomized Controlled Trial

Peng Chiong Tan, FRCOG, Mat Jin Norazilah, MD, and Siti Zawiah Omar, MOG

OBJECTIVE: To compare 5% dextrose–0.9% saline

against 0.9% saline solution in the intravenous rehydra-

tion of hyperemesis gravidarum.

METHODS: Women at their first hospitalization for

hyperemesis gravidarum were enrolled on admission to

the ward and randomly assigned to receive either 5%

dextrose–0.9% saline or 0.9% saline by intravenous infu-

sion at a rate 125 mL/h over 24 hours in a double-blind

trial. All participants also received thiamine and an anti-

emetic intravenously. Oral intake was allowed as toler-

ated. Primary outcomes were resolution of ketonuria and

well-being (by 10-point visual numerical rating scale) at

24 hours. Nausea visual numerical rating scale scores

were obtained every 8 hours for 24 hours.

RESULTS: Persistent ketonuria rates after the 24-hour

study period were 10 of 101 (9.9%) compared with 11 of

101 (10.9%) (P..99; relative risk 0.9, 95% confidence

interval 0.4–2.2) and median (interquartile range) well-

being scores at 24 hours were 9 (8–10) compared with

9 (8–9.5) (P5.73) in the 5% dextrose–0.9% saline and 0.9%

saline arms, respectively. Repeated measures analysis of

variance of the nausea visual numerical rating scale score

as assessed every 8 hours during the 24-hour study

period showed a significant difference in favor of the

5% dextrose-0.9% saline arm (P5.046) with the superior-

ity apparent at 8 and 16 hours, but the advantage had

dissipated by 24 hours. Secondary outcomes of vomiting,

resolution of hyponatremia, hypochloremia and hypoka-

lemia, length of hospitalization, duration of intravenous

antiemetic, and rehydration were not different.

CONCLUSIONS: Intravenous rehydration with 5%

dextrose–0.9% saline or 0.9% saline solution in women

hospitalized for hyperemesis gravidarum produced

similar outcomes.

CLINICAL TRIAL REGISTRATION: ISRCTN Register,

www.controlled-trials.com/isrctn, ISRCTN65014409.

(Obstet Gynecol 2013;121:291–8)

DOI: http://10.1097/AOG.0b013e31827c5e99

LEVEL OF EVIDENCE: I

Hyperemesis gravidarum has for practical purposesbeen defined as intractable vomiting of pregnancy

severe enough to require hospital admission1 affectingup to 2.3% of pregnancies.2 Hyperemesis gravidarumis thus differentiated from the less severe nausea andvomiting of pregnancy, which affects up to 85% ofpregnancies.3 Hyperemesis gravidarum is the secondmost common indication for hospitalization in womenwith successful pregnancies.4 Patients affected byhyperemesis gravidarum are dehydrated and starvedwith associated metabolic, electrolyte, and endocrinedisturbances: hyponatremia is present in 43–49% andhypochloremia in 33–40% on hospital admission.5,6

Ketonemia and resultant ketonuria is the consequenceof the switch to an alternative energy source whendietary glucose is insufficient for metabolic needs.7

The brain in the satiated state uses glucose as the exclu-sive energy substrate.8 The average requirement by theadult brain for glucose is 100 g per day and the rec-ommended daily amount for glucose in pregnancy is175 g.9 Adults should get 45–65% of their calories fromcarbohydrates and young women in the first trimesterof pregnancy should consume 2,400 calories per day.9

From the Department of Obstetrics and Gynaecology, University of Malaya,Lembah Pantai, Kuala Lumpur, Malaysia.

Funded by the University of Malaya, Grant RG412/12HTM.

Corresponding author: Peng Chiong Tan, FRCOG, Department of Obstetricsand Gynaecology, University of Malaya, Lembah Pantai, Kuala Lumpur 50603,Malaysia; e-mail: [email protected].

Financial DisclosureThe authors did not report any potential conflicts of interest.

© 2013 by The American College of Obstetricians and Gynecologists. Publishedby Lippincott Williams & Wilkins.ISSN: 0029-7844/13

VOL. 121, NO. 2, PART 1, FEBRUARY 2013 OBSTETRICS & GYNECOLOGY 291

Approximately 25% of patients with hyperemesis grav-idarum treated with metoclopramide and standardsaline rehydration were still ketonuric at 24 hours.10

Our objective was to estimate whether the additionof 5% dextrose to 0.9% saline rehydration solution inthe first 24 hours after hospitalization, thereby pro-viding 150 g of glucose over 24 hours intravenously(equivalent to 600 calories) when nutritional intake islikely still limited, would result in faster resolution ofketonuria and a more rapid general recovery, culmi-nating in increased perception of well-being.

PATIENTS AND METHODS

The trial was conducted in a university hospital inKuala Lumpur, Malaysia. Ethical oversight was pro-vided by the University of Malaya Medical CenterMedical Ethics Committee (approval date September22, 2010, reference number 811.8). This trial wasperformed in compliance with the Declaration ofHelsinki. The trial is registered with a public trialregistry with the identifier ISRCTN65014409. Allparticipants provided informed written consent.

Women at their first hospitalization for hyper-emesis gravidarum (intractable nausea and vomitingof pregnancy with dehydration and starvation clinicallyjudged to require hospitalization for intravenousrehydration and antiemetic drug administration) wereenrolled within 2 hours of ward admission by theirhealth care providers and randomly assigned to receiveeither 5% dextrose–0.9% saline or 0.9% saline by intra-venous infusion at a rate 125 mL/h over 24 hours ina double-blind trial. Women already under intravenousrehydration therapy were not recruited. We did notprovide outpatient intravenous rehydration and anti-emetic therapy for hyperemesis gravidarum; womenwho needed these therapies were all admitted. Otherinclusion criteria were age 18 years or older, ketonuriaby urine dipstick of at least 1+ on admission, gestation16 weeks or less, plasma glucose 110 mg/dL or less,and sodium 125 mmol/L or greater on admission.

We excluded women in the case group withmultiple gestation, established nonviable pregnancy,pre-existing medical conditions that can cause nauseaand vomiting (eg, culture-proven symptomatic urinarytract infection, dengue fever), gastrointestinal causes ofvomiting (eg, gastroenteritis, gastritis, peptic ulcer),medical causes of vomiting (eg, diabetic ketoacidosis),and women with underlying medical problems (eg,established gestational hypertension, diabetes, heartdisease, renal disease, and thyroid disorder).

In our center, as standard initial treatment, patientswith hyperemesis gravidarum received intravenousrehydration with 0.9% saline solution (potassium

chloride was added as required if hypokalemic), 10 mgoral thiamine daily, and an intravenous antiemetic(typically 10 mg metoclopramide 8 hourly).11 Oralintake was allowed as tolerated at a pace decided bythe affected women. We did not ask participants torecord their oral intake.

Participants were recruited by health care pro-viders as they were admitted to the gynecology ward.Blood was taken for renal function, plasma glucose,and full blood count. They also were given the visualnumerical rating scale for nausea to be filled out atrecruitment and 8, 16, and 24 hours after admission.

Randomization was affected by the sequentialopening of numbered, sealed, opaque envelopes stating“Protocol A” or “Protocol B.” Randomization sequenceon a one-to-one ratio was computer-generated (usinghttp://www.random.org) for 223 participants by a coau-thor (P.C.T.) who played no role in enrollment.

Intravenous solutions were prepared by a coauthor(M.J.N.) using standard hospital issue 5% dextrose–0.9% saline or 0.9% saline solution in 500-mL con-tainers with the manufacturer’s label stripped offand relabeled as solutions A or B. The solutions andcontainers were identical in appearance except for theA or B labels. Randomization was to either A or B(either 63500 mL .9% saline or 63500 mL of 5%dextrose–0.9% saline) to be given intravenously over24 hours (ie, 500 mL of solution every 4 hours). Onegram (9.5 mmol) of potassium chloride can be addedto each 500-mL solution as required to correct hypo-kalemia. Participants and health care providers werethus both blinded to the allocated solution.

A multivitamin preparation, P-Trovite, whichcontains 250 mg thiamine, was given intravenouslyto all participants before starting the trial rehydrationsolution to prevent Wernicke’s encephalopathy thatmay arise as a consequence of dextrose infusion ina severely thiamine-deficient state. Intravenous anti-emetic was also prescribed according to health careproviders to all participants for 24 hours or until sig-nificant symptom relief was established.

During the 24-hour study period, urine waschecked by dipstick 8 hourly for ketonuria andglycosuria. If there was significant glycosuria 2+ orgreater present, the patient’s capillary glucose levelwas checked with a glucometer and if reading8 mmol/L or greater, infusion fluid was to be changedto open-label standard 0.9% saline. Participants wereinstructed to mark the nausea visual numerical ratingscale (10 points, high score denoting more severe nau-sea) before initial administration of the intravenousfluid regime and then at 8, 16, and 24 hours. At24 hours, participants were also asked to mark their

292 Tan et al Rehydration in Hyperemesis Gravidarum OBSTETRICS & GYNECOLOGY

perceived well-being over the study period with a10-point visual numerical rating scale (higher score,greater well-being). At the conclusion of the 24-hourmain study period, open-label intravenous fluid wasstarted if still required.

Primary outcomes were resolution of ketonuria andwell-being (by 10-point visual numerical rating scale) at24 hours. Secondary outcomes included frequency ofvomiting in the 24-hour study period; nausea visualnumerical rating scale at 8, 16, and 24 hours, hypona-tremia (135 mmol/L or less); hypokalemia (3.5 mmol/Lor less); hypochloremia (99 mmol/L or less); hypergly-cemia (8 mmol/L or greater) at the end of the 24-hourmain study period; duration of intravenous antiemeticand intravenous rehydration during hospitalization;and the hospital admission to discharge interval.Data were extracted onto the case report form afterhospital discharge from case notes, laboratory reports,and direct contact of participants if needed.

There is no trial comparing intravenous fluidregimes for hyperemesis gravidarum to guide samplesize calculation. A previous trial of promethazinecompared with metoclopramide (using 0.9% salineas the standard intravenous fluid) in our center hasshown that 25% of patients with hyperemesis grav-idarum had not cleared their ketonuria and well-beingby visual numerical rating scale is 7.662.2 in themetoclopramide arm at 24 hours.10 Sample size calcu-lations were made with the PS program. Assuming a10% compared with 25% (relative risk 0.4) rates forpersistence of ketonuria at 24 hours for 5% dextrose–0.9% saline and 0.9% saline arms respectively,a5.05, power 80%, one-to-one recruitment ratio,and applying the x2 test, 100 women are required ineach arm. Factoring in a 10% dropout rate, a totalof (200/0.9) 223 women are needed for a suitablypowered study. Similarly, assuming that well-beingvisual numerical rating scale score will improve by 1with 5% dextrose–0.9% saline, applying the Student’st test, 77 participants will be required in each arm. Ifthe Mann-Whitney U test were to be applied insteadof the Student’s t test in the event of nonnormallydistributed visual numerical rating scale scores, a10% increase in numbers is appropriate and factoringin a 10% dropout rate, a total of only (15431.1/0.9)189 women are required for a powered study on thisoutcome. We planned to recruit 223 women.

Data were entered into SPSS 17. Analysis was byintention to treat after exclusions for criteria infringe-ments. Normality of data distribution was checkedwith the Kolmogorov-Smirnov test. Normally distrib-uted continuous data were analyzed with the Student’st test. Two-by-two categorical data sets were analyzed

with the Fisher’s exact test and larger categorical datasets with the x2 test; ordinal data and nonnormallydistributed continuous data were analyzed with theMann-Whitney U test. A repeated-measures analysisof variance was applied to the nausea visual numericalrating scale scores and to ketonuria status. All testswere two-sided and P,.05 was considered significant.

RESULTS

The trial was conducted from November 9, 2010, toFebruary 6, 2012. Recruitment was stopped on reach-ing the targeted number of participants. The recruit-ment flowchart of participants through the trial isshown in Figure 1. Exclusions were mostly the resultof study criteria infringements (shown in Fig. 1) thatwere ascertained only after randomization as a resultof our pragmatic trial process where allocated treatmentwas started before full test results were available. Allparticipants received P-Trovite intravenously and hadtheir rehydration regimen administered as allocated.

The characteristics of the participants in the twotrial arms are shown in Table 1. Participants in the twoarms have similar characteristics except for meanserum potassium in which there was a mean differ-ence of 0.1 mmol/L between the trial arms. All otherparameters from the standard renal function test, fullblood count, and random glucose as well as biometricvariables of blood pressure and pulse at recruitmentwere similar (result not shown).

Table 2 displays the primary outcomes of ketonu-ria and well-being visual numerical rating scale scoreaccording to treatment allocation. Ketonuria was stillpresent at 24 hours in 10 of 101 (9.9%) compared with11 of 101 (10.9%) (P,.99) (relative risk 0.9, 95% con-fidence interval [CI 0.4–2.2) and median (interquartilerange) well-being score was 9 (8–10) compared with 9(8–9.5) (P5.73) for 5% dextrose–0.9% saline and 0.9%saline arms, respectively, at the end of the 24-hourmain study period.

Of the secondary outcomes (Table 3), the nauseavisual numerical rating scale scores at 8 and 16 hoursshowed a significant reduction (P,.01 and P5.03,respectively) in favor of the 5% dextrose–0.9% salinearm but the difference had dissipated by 24 hours.Repeated-measures analysis of variance of between-subject nausea visual numerical rating scale scoresshowed a significant result (P5.046), but interactionof treatment and time was also significant (P5.004).Hospital stays (defined as interval in hours from ran-domization to documented medical decision to dis-charge) were 43621 compared with 48621 (P5.14)for 5% dextrose–0.9% saline and 0.9% saline arms,respectively.

VOL. 121, NO. 2, PART 1, FEBRUARY 2013 Tan et al Rehydration in Hyperemesis Gravidarum 293

There was a statistically significant difference inthe serum potassium level after completion of the 24-hour study period (mean6standard deviation): 3.96.4compared with 3.86.3 mmol/L (5% dextrose–0.9%saline compared with 0.9% saline arms; P5.04);the mean .1-mmol/L difference in serum potassiumlevel is identical to the pretreatment mean differencebetween the two trial arms and is likely to have arisenfrom the differing baseline. The proportion classifiedas hypokalemic (3.5 mmol/L or less) was not different(19.1% compared with 28.7%; P5.17, relative risk .6,95% CI 0.3–1.2 between the trial arms at 24 hours.

Post hoc, using the paired t test, for the entire trialcohort, mean glucose, serum sodium, and chloridelevels increased in the first 24 hours after hospitaliza-tion (87–100 mg/dL, 134–138, 101–106 mmol/Lrespectively, all P,.001), whereas that for serumpotassium decreased marginally from 3.92 to 3.84mmol/L (P5.014). Adjusting for antiemetic regimendid not alter the result for persistence of ketonuria

at 24 hours (adjusted odds ratio 0.8, 95% CI0.3–2.0, P5.60; 5% dextrose–0.9% saline comparedwith 0.9% saline). Restricting analysis to only themetoclopramide-exposed cases also did not alter thenonsignificant results on persistence of ketonuria(P5.81) and well-being score (P5.43); the significantresult on the nausea visual numerical rating scale byrepeated-measures analysis of variance (P5.049) infavor of the 5% dextrose–0.9% saline arm remained.

If participants excluded postrandomization asa result of enrollment criteria infringements wereincluded in the analysis, outcomes between the trialarms were still similar. Similarly, adjusting for thepotassium level also did not materially alter theresults. There was no protocol curtailment as a resultof hyperglycemia during the 24-hour study period.

DISCUSSION

To our knowledge, intravenous rehydrationregimes in the management of hyperemesis

Hospitalizations for presumedhyperemesis gravidarum

N=422

Women recruitedn=222

Randomized to dextrose-salinen=111

Available for analysisas per enrollment criteria

n=102

Women potentially eligiblefor trial entry

n=354

Rehopsitalizationsn=68

Not approached ordeclined participation

n=131

Envelope mistakenly openedand discarded (randomization

to dextrose saline)n=1

Randomizationenvelope opened

n=223

Participant withdrawaln=2

Randomized to normal salinen=111

Participant withdrawaln=1

Enrollment criteria infringement: n=7Urinary tract infections: 2High blood glucose: 2Gestational age greaterthan 16 weeks: 1

Rehospitalization: 2

Enrollment criteria infringement: n=9Urinary tract infections: 2High blood glucose: 5Gestational age greaterthan 16 weeks: 1

Multiple pregnancy: 1

Available for analysisas per enrollment criteria

n=101

Fig. 1. Recruitment flowchart for a double-blind randomized trial of intravenous rehydration with 5% dextrose–0.9% salinecompared with 0.9% saline solution at the first hospitalization for hyperemesis gravidarum.

Tan. Rehydration in Hyperemesis Gravidarum. Obstet Gynecol 2013.


gravidarum have not been previously studied. Weperformed an unrestricted online PubMed search(http://www.ncbi.nlm.nih.gov/sites/entrez) on May13, 2012, using the search terms “hyperemesis grav-idarum hydration trial” and identified only a reviewarticle. There seems to be little change since the 2004Practice Bulletin from the American College ofObstetricians and Gynecologists, which stated thatno study has compared different fluid replacementsfor nausea and vomiting of pregnancy.3

Intravenous rehydration is a mainstay of man-agement in hyperemesis gravidarum, but reflecting

the dearth of evidence, the fluid regime is typically notdefined in guidance and reviews.3,12,13 A recent reviewcautions against the use of dextrose in the rehydrationfluid in the rare event of Wernicke’s encephalopathybeing precipitated as a result of thiamine deficiencyand the overrapid correction of hyponatremia possi-bly leading to central pontine myelinolysis,14 whereasanother recommends total food avoidance and paren-teral administration of carbohydrate, vitamins, andamino acids in the process of correcting volume andelectrolyte imbalances in hyperemesis gravidarum.15

Our data suggested that intravenous multivitamins,

Table 1. Characteristics of All Trial Participants According to Randomization to 5% Dextrose–0.9% Salineor Normal Saline Solution for Rehydration of Hospitalized Hyperemesis Gravidarum Cases

Characteristic Dextrose Saline (n5111) Normal Saline (n5111) P

Age (y) 28.564.6 29.364.6 .22Gestation (wk) 9.862.8 9.862.5 .95Parity 0 (0–2) 1 (0–2) .90Prior miscarriage 15 (13.5) 16 (14.4) ..99Ethnicity .25

Malay 85 (76.6) 85 (76.6)Indian 11 (9.9) 17 (15.3)Other 15 (13.5) 9 (8.1)

Employment ..99Employed 85 (76.6) 26 (23.4)Homemaker or student 85 (76.6) 26 (23.40

Education .30Primary 6 (5.4) 2 (1.8)Secondary 45 (40.5) 51 (45.9)Tertiary 60 (54.1) 58 (523)

Weight (kg) 58.2612.2 57.3611.4 .59Body mass index (kg/m2) 24.064.5 23.764.5 .65Ketonuria (dipstick) .67

1+ 11 (9.9) 12 (10.8)2+ 14 (12.5) 13 (11.7)3+ 23 (20.7) 27 (24.3)4+ 63 (56.8) 59 (53.2)

Serum sodium (mmol/L) 13562 13462 .12Hyponatremia (135 mmol/L or less) 80 (72.1) 84 (75.7) .65Serum potassium (mmol/L) 4.060.4 3.960.4 .03Hypokalemia (3.5 mmol/L or less) 14 (12.6) 22 (19.8) .20Serum chloride (mmol/L) 10262 10163 .07Hypochloremia (99 mmol/L or less) 20 (18.0) 29 (26.1) .20Plasma glucose (mg/dL) 89611 89616 .67Hematocrit 0.3860.03 0.3860.03 .98Nausea score* 9 (7–10) 9 (7–10) .39Antiemetic regimen .053Metoclopramide 94 (85.5) 79 (72.5)Prochlorperazine 11 (10.0) 18 (16.5)Ondansetron 5 (4.5) 12 (11.0)Weight loss percentage (n5186)† 5.763.4 5.663.2 .83Weight loss 5% or greater† 55/92 (59.8) 54/94 (57.4) .77

Data are mean6standard deviation, median (interquartile range), or n (%) unless otherwise specified.Analysis by Student’s t test for continuous variables, Mann-Whitney U test for ordinal data, Fisher’s exact test for 232 categorical data sets,

and x2 test for larger than 232 categorical data sets. All statistical tests are two-sided.* Nausea is self-scored by participants using a 10-point numeric rating score, with a score of 1–10 as nausea increases.† Weight loss defined as (prepregnancy weight–current weight)/prepregnancy weight3100—all in kilograms; prepregnancy weight as

reported by participants where available; n,203 as a result of incomplete data ascertainment.


antiemetic, and a rehydration regime comprising 3 Lof 0.9% saline over a 24-hour period in the initialmanagement of hyperemesis gravidarum is seeminglyeffective: at 24 hours, in 89%. ketonuria had resolved,hyponatremia reduced from 76% to 18% (with onlya single case of a 11-mmol/L increase in serum sodiumand another case in which serum sodium level peakedoutside our upper normal limit of 144 at 145 mmol/L),nausea visual numerical rating scale reduced froma median score of 9 to 2, and 54% (54 of 101) hadstopped vomiting altogether.

In the context of pediatric acute gastroenteritis,intravenous dextrose in the rehydration fluid on anoutpatient basis reduced the risk of a return visitrequiring admission.16 It has been postulated thatreduced carbohydrate intake leads to free fatty acidbreakdown, excess ketones, and an increased likelihoodfor continued nausea and vomiting and that the additionof dextrose to intravenous rehydration therapy reducedfree fatty acid breakdown and reduced the incidence ofoutpatient treatment failure as a result of intractablenausea and vomiting.17 Our data showed a temporary

reduction in nausea score in the 5% dextrose–0.9%saline arm at 8 and 16 hours, but the reduction didnot seem to be mediated through a reduction in keto-nuria nor was there any reduction in vomiting.

Our trial protocol did not routinely includepotassium supplementation in the rehydration solu-tion in either trial arm, although potassium could beadded to correct hypokalemia at health care pro-viders’ instruction. This resulted in a marginal drop of0.1 mmol/L in mean serum potassium level and anincrease in the proportion classified as hypokalemic(3.5 mmol/L or less) from 17% to 24% (the lowestrecorded serum potassium of 3.0 mmol/L in onewoman) at 24 hours. In 51% of participants, serumpotassium level fell in the first 24 hours. There seemsto be a reasonable case for routine supplementarypotassium in the initial management of a typical caseof hyperemesis gravidarum.

There was a small 0.1-mmol/L difference in themean serum potassium level at recruitment andsimilarly at end of the 24-hour study period betweenthe trial arms. The small difference at recruitment was

Table 2. Primary Outcomes of 5% Dextrose–0.9% Saline Compared With Normal Saline as IntravenousRehydration Solution for Hyperemesis Gravidarum

Dextrose Saline (n5102) Normal Saline (n5101) P Relative Risk (95% CI)

Well-being score at 24 h* 9 (8–10) 9 (8–9.5) .738.561.6 8.461.5 .75

Persistent ketonuria at 24 h 10 (9.9) 11 (10.9) ..99 0.9 (0.4–2.2)Ketonuria at enrollment .86

0 0 (0.0) 0 (0.0)1+ 11 (10.8) 10 (9.9)2+ 11 (10.8) 10 (9.9)3+ 21 (20.6) 26 (25.7)4+ 59 (57.8) 55 (54.5)

Ketonuria at 8 h .140 24 (23.5) 28 (27.7)1+ 26 (25.5) 15 (14.9)2+ 19 (18.6) 23 (22.8)3+ 28 (27.5) 34 (33.70)4+ 5 (4.9) 1 (1.0)

Ketonuria at 16 h .730 68 (67.3) 67 (66.3)1+ 16 (15.8) 12 (11.9)2+ 10 (9.9) 12 (11.9)3+ 7 (6.9) 10 (9.9)

Ketonuria at 24 h .420 91 (90.1) 90 (89.1)1+ 8 (7.9) 5 (5.0)2+ 1 (1.0) 4 (4.0)3+ 1 (1.0) 2 (2.0)

Repeated measures analysis of variance of ketonuria status (between-patient effects) .77

CI, confidence interval.Data are median (interquartile range), mean6standard deviation, or n (%) unless otherwise specified.Analysis by Mann-WhitneyU test for ordinal data, Fisher’s exact test for 232 categorical data, set and x2 test for larger than 232 categorical

data sets. All statistical tests are two-sided.* Well-being is assessed using a 10-point (1–10) visual numerical rating scale; higher score greater well-being.


probably a chance occurrence, which was carriedthrough to the end of the study period. A 0.1-mmol/Ldifference in mean serum potassium is unlikely to beof clinical significance. Adjustment for potassiumlevel at recruitment did not affect primary outcomeof persistence of ketonuria between trial arms.

Despite 150 g of dextrose by intravenous admin-istration (equivalent to 50% of the daily normalcarbohydrate intake in an average diet and 600calories) to women in the 5% dextrose–0.9% salinearm, the rise in plasma glucose was similar in magni-tude at 24 hours when compared with the 0.9% salinearm. This finding may reflect oral food intake near the24-hour mark in the 0.9% saline arm because nauseaand vomiting improved or increased glucose metabo-lism in the 5% dextrose–0.9% saline arm driven byavailability. We did not ask the participants to docu-ment oral intake in this trial.

This trial has limitations and strengths. In the0.9% saline arm, the persistent ketonuria rate was only

11% compared with an anticipated 25% as assumed inthe sample size calculation using earlier data.10 Thiswould have reduced the trial’s power to demonstratea difference. The improved ketonuria resolution ratein the 0.9% saline arm of our current trial may be theresult of the routine use of P-Trovite multivitaminsupplement or a more rigorous and consistent per-protocol administration of the prescribed intravenousrehydration solutions. On the other hand, because thestandard deviation in the well-being visual numericalrating scale was 1.5 (compared with a standard devi-ation of 2 used in sample size calculation), this wouldhave increased the power of the trial to detect a 1-pointincrease in well-being visual numerical rating scalescore to 99.7%. We believe our finding is generaliz-able for similarly managed inpatient hyperemesisgravidarum populations.

Dextrose saline solution was not superior to normalsaline solution in the initial intravenous rehydration ofwomen hospitalized for hyperemesis gravidarum on

Table 3. Secondary Outcomes of Randomized Trial of 5% Dextrose–0.9% Saline Compared With NormalSaline as Intravenous Rehydration Solution for Hyperemesis Gravidarum

Dextrose Saline(n5102)

Normal Saline(n5101) P

Relative Risk(95% CI)

Vomiting episodes* 0 (0–2) 0 (0–2) .66Nausea score at enrollment† 9 (7–10) 9 (7–10) .45Nausea score at 8 h† 6 (4–7) 7 (5–8) ,.01Nausea score at 16 h† 4 (2–5) 5 (3–6) .03Nausea score at 24 h† 2 (1–4) 2 (2–4) .39Repeated-measures analysis of variance ofnausea score (between-patient effects)

.046

Blood results at 24 hSerum sodium (mmol/L)* 13862 13762Hyponatremia (135 mmol/L or less)* 13 (13.8) 17 (18.1) .55 0.7 (0.3–1.6)Serum potassium (mmol/L)* 3.960.4 3.860.3 .04Hypokalemia (3.5 mmol/L or less)* 18 (19.1) 27 (28.7) .17 0.6 (0.3–1.2)Serum chloride (mmol/L)* 10663 10564 ,.01Hypochloremia (99 mmol/L or less)* 1 (1.1) 5 (5.3) .21 0.2 (0.02–1.7)Plasma glucose (mg/dL)* 100624 100618 .97Hyperglycemia (145 mg/dL or greater)* 3 (3.0) 2 (2.0) .68 1.5 (0.3–9.4)Duration of intravenous antiemetic use (h)‡ 24617 25615 .61Duration of intravenous rehydration (h)‡ 33614 37615 .15Recording of first oral intake (h after recruitment)§ 11.265.7 13.266.0 .022Oral intake by 8 h§ 40/101 (39.6) 24/99 (24.2) .023 2.0 (1.1–3.8)Oral intake by 16 h§ 84/101 (83.2) 78/99 (78.8) .47 1.3 (0.7–2.7)Oral intake by 24 h§ 98/101 (97.0) 93/99 (93.9) .33 2.1 (0.5–8.7)Hospital stay (h)jj 43621 48621 .14

CI, confidence interval.Data are mean6standard deviation, median (interquartile range), or n (%) unless otherwise specified.Analysis by Student’s t test for continuous variables. Mann-Whitney U test for ordinal data, Fisher’s exact test for 232 categorical data sets,

and x2 test for larger than 232 categorical data sets. All statistical tests are two-sided.* Assessed at the end of the 24-hour study period.† Nausea is assessed using a 10-point (1–10) visual numerical rating scale: higher score signifies more nausea.‡ Over the entire course of the hospital stay.§ Oral intake of solids or fluids of any quantity as first recorded by nurses or doctors; n,203 as a result of incomplete data ascertainment.jj Interval from recruitment to documented medical decision to discharge.


a range of outcomes. However, because of the theoret-ical concern of Wernicke’s encephalopathy with dex-trose infusion when in a thiamine-deficient state,normal saline may be a better choice.

REFERENCES1. Verberg MF, Gillott DJ, Al-Fardan N, Grudzinskas JG. Hyper-

emesis gravidarum, a literature review. Hum Reprod Update2005;11:527–39.

2. Tan PC, Jacob R, Quek KF, Omar SZ. The fetal sex ratio andmetabolic, biochemical, haematological and clinical indicatorsof severity of hyperemesis gravidarum. BJOG 2006;113:733–7.

3. Nausea and vomiting of pregnancy. ACOG Practice BulletinNo. 52. American College of Obstetrics and Gynecology.Obstet Gynecol 2004;103:803–16.

4. Gazmararian JA, Petersen R, Jamieson DJ, Schild L,Adams MM, Deshpande AD, et al. Hospitalizations duringpregnancy among managed care enrollees. Obstet Gynecol2002;100:94–100.

5. Tan PC, Tan NC, Omar SZ. Effect of high levels of humanchorionic gonadotropin and estradiol on the severity of hyper-emesis gravidarum. Clin Chem Lab Med 2009;47:165–71.

6. Tan PC, Jacob R, Quek KF, Omar SZ. Indicators of prolongedhospital stay in hyperemesis gravidarum. Int J Gynaecol Obstet2006;93:246–7.

7. Comstock JP, Garber AJ. Ketonuria. In: Walker HK, Hall WD,Hurst JW, editors. Clinical methods: the history, physical, andlaboratory examinations. 3rd ed. Boston (MA): Butterworths;1990.

8. Hasselbalch SG, Knudsen GM, Jakobsen J, Hageman LP,Holm S, Paulson OB. Brain metabolism during short-term star-vation in humans. J Cereb Blood Flow Metab 1994;14:125–31.

9. Institute of Medicine. Dietary reference intakes for energy, car-bohydrate, fiber, fat, fatty acids, cholesterol, protein, and aminoacids. September 5, 2002. Available at: http://www.iom.edu/Reports/2002/Dietary-Reference-Intakes-for-Energy-Carbohydrate-Fiber-Fat-Fatty-Acids-Cholesterol-Protein-and-Amino-Acids.aspx. Retrieved May 13, 2012.

10. Tan PC, Khine PP, Vallikkannu N, Omar SZ. Promethazinecompared with metoclopramide for hyperemesis gravidarum:a randomized controlled trial. Obstet Gynecol 2010;115:975–81.

11. Tan PC, Yow CM, Omar SZ. A placebo-controlled trial of oralpyridoxine in hyperemesis gravidarum. Gynecol Obstet Invest2009;67:151–7.

12. Tan PC, Omar SZ. Contemporary approaches to hyperem-esis during pregnancy. Curr Opin Obstet Gynecol 2011;23:87–93.

13. Niebyl JR. Clinical practice. Nausea and vomiting in preg-nancy. N Engl J Med 2010;363:1544–50.

14. Jarvis S, Nelson-Piercy C. Management of nausea and vomitingin pregnancy. BMJ 2011;342:d3606.

15. Jueckstock JK, Kaestner R, Mylonas I. Managing hyperemesisgravidarum: a multimodal challenge. BMC Med 2010;8:46.

16. Levy JA, Bachur RG. Intravenous dextrose during outpatientrehydration in pediatric gastroenteritis. Acad Emerg Med 2007;14:324–30.

17. Reid SR, Losek JD. Rehydration: role for early use of intrave-nous dextrose. Pediatr Emerg Care 2009;25:49–52.


dextrose saline compared with normal saline.13

Documents