Skip navigation

Overview of chemotherapy nausea and vomiting


Several systematic reviews and meta-analyses have examined the efficacy of treatments for preventing chemotherapy-induced nausea and vomiting.

Dexamethasone and chemotherapy nausea and vomiting
5-HT3 receptor antagonists for chemotherapy-induced nausea and vomiting
Cannabinoids for chemotherapy-induced nausea and vomiting
5-HT3 receptor antagonists for radiotherapy-induced nausea and vomiting

There are major similarities. All have looked at randomised trials, that report nausea, or vomiting, or complete control of emesis within the first 24 hours after chemotherapy. Most tell us the number of patients who vomit, or juts to make things complicated, who do not vomit.

Controls are problematical, because most people having chemotherapy are likely to vomit. True placebo controls are rate - there is one trial hidden in the reviews with about 20 patients given placebo only, and 18 vomited. Control often means another treatment, or possibly no placebo or no treatment with antiemetic cover of some sort. That cover will be outside of the randomisation, so that many people may be given an antiemetic before they vomit.

That's only right and humanitarian. It just means that we are unsure about where the baseline is.

The similarity of the trials means that we can make some broad comparisons between treatments. Bandolier has done this for the outcome of patients who vomit after chemotherapy within 24 hours, with the different types of treatment. No clever statistics, because those are probably neither justified nor needed. Figure 1 shows the results for all the information available from the reviews. The actual numbers are given in Table 1 below.

Figure 1: Percentage of patients vomiting in the first 24 hours after chemotherapy


That control (taken from the largest dataset in dexamethasone comparisons) is about the same as high-dose metoclopramide or conventional treatments is little surprise, because these patients are, for the most part, receiving some treatment. Cannabinoids had very few patients, and with many adverse effects. The clear picture is that 5-HT3 receptor antagonists with dexamethasone stand out as ensuring that many fewer patients vomit.

Table 1: Number of patients who did not vomit, and total treated


Number who did not vomit





High dose metoclopramide



Conventional treatment






5HT3 alone



Dexamethasone alone



5HT3 plus dexamethasone



Should we use data like this?

Not the quality of the data, but using data drawn from different trials and meta-analyses to make indirect comparisons. Purists would probably say not, though many of their quibbles would be about trials, setting, and outcomes not being the same. Here, within limits, they are the same, so many of these quibbles do not apply. It would, of course, be great if there were large, multicentre, comparable trials of A versus B, A versus C, A versus D, B versus C etc.

But those trials would have to be huge to be useful, they won't be done, and we need to make decisions now. Indirect comparison using comparable trials is the best we can do now, in order to make the best decisions for our patients and healthcare systems. We can't afford to be precious.