Chances are you will be seeing a patient with vomiting during your next shift in the ED. That vomiting may very well be caused by a case of acute gastroenteritis (AGE). You’ve probably written for, and also prescribed ondansetron (Zofran) – perhaps in a reflexive fashion. As with many therapies that have become accepted into routine practice I think that it important to delve into the evidence as to why this is the case. So, without further ado, let’s look at the use of ondansetron to treat vomiting in acute gastroenteritis.

A brief review of pharmacology

Ondansteron, AKA Zofran is a serotonin 5-HT3 receptor antagonist. These receptors are located in terminal of the vagus nerve in the gut and in the brain. Specific cells in the small intestine produce serotonin in response to vomit-inducing stimuli. Blocking the serotonin receptors thus preventing vomiting. The dose is generally 0.15 mg/kg PO or IV. Oral forms are available as a solution or oral dissolving tablet. If you want to keep things simple you could dose in the following manner:

2 mg for children weighing 8 to 15 kg
4 mg for children weighing >15 to 30 kg
8 mg for children weighing >30 kg

The side effects include diarrhea, urticaria, headache, constipation, and fever. It is contraindicated in patients at risk for malignant hyperthermia and in those with history of prolonged QTc. EKG changes can occur with all 5-HT3 antagonists. Patients especially at risk include those with hypokalemia or hypomagnesemia, heart failure, bradyarrhythmias and those on other medications that increase the risk of QTc prolongation.

The evidence

It is not surprising (especially to readers of the blog) that I would first cite a systematic review. DeCamp et al in JAMA Pediatrics, 2008 identified 11 trials of antiemetics to reduce vomiting in kids with AGE. Six of these compared ondansetron to placebo. Let’s look at each in turn, in chronological order, before moving on to the conclusions.

Cubeddu et al. Aliment Pharmacol Ther, 1997

Three groups of 12 patients were randomized in a parallel fashion to get a single i.v. dose of ondansetron (0.3 mg/kg), metoclopramide (0.3 mg/kg) or placebo (saline). Patients received primarily oral rehydration via fluids over a 24 hour period. In this small trial the authors found that:

Patients that received ondansetron vomited less – mean of 2 episodes versus 5 with placebo (p = 0.048)
Patients that received ondansetron were more likely to not vomit at all 58% versus 17% placebo (p – 0.039)
The antiemetic groups saw more diarrhea, and fewer treatment failures versus placebo overall

Reeves et al. Pediatrics, 2002

This randomized, double blind, placebo-controlled trial in a pediatric ED of patient 1 month to 22 years included patients needing IV fluids for AGE. They were randomized to either IV ondansetron or placebo. The investigators assessed for frequency of vomiting, admission rate, and occurrence of complications.

More ondansetron patients stopped vomiting (70% vs 51%)
More placebo patients were admitted (30% vs 26%)
Subanalysis of patients with CO2 ≥15 mEq/L showed a greater proportion of the placebo group requiring admission (23% vs 7%)

Ramsook et al. Ann Emerg Med, 2002

This double-blind RCT prospectively examined children between the ages of 6 months and 12 years who had vomited at least 5 times in the last 24 hours. They received oral ondansetron or placebo. ORT started 15 minutes later at 5 mL/min. Oral rehydration was commenced 15 minutes later at 5 mL/min. Treatment failure was defined as patients necessitating admission. Kids that were sent home did so with recommendations to do oral rehydration at home with the introduction of a bananas, rice, applesauce, and toast (BRAT) diet after the first 24 hours. The discharged patients also got 5 doses of either ondansetron or placebo. Phone follow up occurred at 24 and 48 hours and parents were queried on the number of episodes of vomiting and diarrhea. Parents also kept a diary of the same. Ultimately they found the following:

Ondansetron patients vomited less in the ED (p =.001)
There was no significant difference in the number of episodes of vomiting during the first 48 hour follow up period
Ondansetron patients had more diarrhea during the follow up period
Ondansetron patients needed IV fluids less frequently (p =.015)
The admission rate was lower in the ondansetron group (p =.007)
The revisit rate was higher in the ondansetron group vs placebo (p =.047)

It is important to note that in either group the median number of episodes of vomiting int he ED and at home was actually zero. They assessed for the rank sum (comparing the continuous distribution of number of episodes of vomiting). This was a viable alternative since so few patients vomited.

Freedman et al. NEJM, 2006

This double blind RCT compared ondansetron ODT vs placebo in children with mild to moderate dehydration due to emesis. 15 minutes after administration they began a oral rehydration therapy based protocol. The children that received ondansetron were:

Less likely to have subsequent vomiting (14 vs 35%; p < 0.001) / Number needed to treat to prevent vomiting in one child = 5
Vomited less often (0.18 vs 0.65 episodes per child; p < 0.001)
Took more oral volume (239 vs 196 ml; p = 0.001)
Less likely to get IVF (14 vs 31%; 95% CI: 0.26-20.79) / Number needed to treat to prevent IV fluids in one child = 6
There was no difference in the rate of hospitalization or ED return visits

Stork et al. Acad. Emerg Med, 2006

This double-blinded RCT of 166 patients examined IV ondansetron in pediatric ED patients that needed IV fluids after failing ORT. All were mildly to moderately dehydrated. The subjects were randomized to IV ondansetron (0.15 mg/kg), IV dexamethasone (1 mg/kg) or placebo. Each also got IV fluids, 10-20ml/kg. They found the following:

The ondansetron group was less likely to be admitted to the hospital vs placebo (4.4 vs 20.5%; RR: 0.21; 95% CI: 0.05-00.81)
The Absolute Risk Reduction for hospitalization was 16.1%
The ondansetron subjects were also significantly more likely to tolerate PO fluids at 2 hours vs placebo (86.6 vs 67.4%; RR: 1.28; 95% CI: 1.02-01.68) – The absolute rate reduction was 18.8%

Roslund et al. Annals of Emergency Medicine, 2008

One hundred six children under the age of 10 years with mild to moderate dehydration that vomited after trying ORT in the ED were randomized to ODT ondansetron or placebo. The experimental hypothesis was that giving ondansteron would prevent children who failed ORT from needing IV fluids. After getting drug or placebo ORT was reattempted in 30 minutes. They noted the following:

The ondansetron ODT patients needed less IV fluids versus placebo (21.6 vs 54.5%; 33% absolute reduction; 95% CI: 14.54-48.37%)
The ondansetron ODT patients were admitted less often (5.9 vs 12.7%) – but this was not statistically significant
Rates of ED revisit and further episodes of vomiting and diarrhea were similar between groups

Vomiting

You already knew that ondansetron helps stop vomiting. Now that you are familiar with some of the evidence it should help you provide more accurate information to your colleagues and patients. Per the aforementioned systematic review, versus placebo, in five trials, ondansetron was associated with a decreased risk of further vomiting.

Relative risk 0.45, 95% CI 0.28-0.62

Looking at the forest plot you can see the visual representation of the Relative Risk as noted above. The diamond at the bottom represents the combined risk of all 5 studies seen above it. The width of that diamond represents the 95% confidence interval. A narrower diamond indicates greater precision. The black boxes above demonstrate the relative risk of the 5 individual studies. The thin horizontal bar behind each is representative of the 95% CI. Note that Reeves and Stork both cross 1, indicating that those results are not statistically significant. However, the 95% CI of the total (diamond) does not.

Need for IV fluids

The systematic review indicated that ondansetron when compared with placebo was associated with a reduced need for intravenous fluids.

Relative Risk 0.41, 95% CI 0.28-0.62

Admission

Ondansetron as compared with placebo was associated with a significantly decreased risk of immediate hospital admission.

Relative Risk 0.52, 95% CI 0.28-0.62

Conclusion

Giving a child with vomiting ondansetron, either orally or intravenously, in the Emergency Department does the following:

Reduces the risk of further episodes of vomiting

Reduces the risk of the patient needing IV fluids

Reduces the risk of immediate admission to the hospital

If you recall this previous post on the PEMBlog, kids that get ondansetron may have a greater risk of returning to the ED. This was demonstrated in a retrospective review by Sturm et al. of over 34,000 children with a diagnosis of AGE and vomiting. Nearly 60% of this cohort got ondansetron. The authors were looking to see if ondansetron could mask alternative diagnoses (appendicitis, intussusception, bacteremia, pyelonephritis, brain tumor) by controlling vomiting. “Of the 76 children who were hospitalized with an alternate diagnosis after returning within 72 hours of discharge from the initial ED visit, there was no difference in the percentage of patients admitted between patients who received ondansetron and those who did not at the first ED visit (15 versus 22 percent).”

It does not appear that the evidence is there for other antiemetics (promethazine, metoclopramide). So I recommend ondansetron for children. Though the aforementioned studies were done in the Pediatric Emergency Department, one could presume that giving ondansetron in the office setting to children with mild to moderate dehydration in the face of an illness with vomiting (mostly AGE).