The 5-HT₃ antagonists are a unique group of drugs that were developed specifically for the management of nausea and vomiting. In general, their antiemetic (anti-vomiting) effects are stronger than their anti-nausea effects [32]. Most of the available research on this class of drugs focuses on ondansetron. A dose of 4 mg ondansetron has a number needed to treat (NNT) of approximately 6 for prevention of vomiting (0–24 h) and approximately 7 for prevention of nausea [32]. The 8 mg oral disintegrating tablet (ODT) is equally effective to the 4 mg IV dose [33]. Other 5-HT₃ antagonists which have been shown to be effective in preventing PONV include: granisetron 0.35–3 mg IV [34], tropisetron 2 mg IV [35], ramosetron 0.3 mg IV [36, 37], and palonosetron 0.075 mg IV [38, 39]. Of these, tropisetron and ramosetron are not available in the United States.

One reason 5-HT₃ antagonists are widely used for the prevention of PONV is their favorable side effect profile. All, except palonosetron, have been shown to prolong the QTc interval [10]. For ondansetron specifically, the U.S. FDA recommends that the dose should not exceed 16 mg in a single dose because of the risks to the QTc interval. The number-needed-to-harm (NNH) for a single dose of ondansetron is 36 for headache, 31 for elevated liver enzymes, and 23 for constipation [19]. All 5-HT₃ antagonists are considered to be equally safe.

Dexamethasone 4–5 mg IV has widely been used prophylactically to prevent PONV. It has been shown to be especially effective against late PONV [40]. A recent meta-analysis showed that the NNT for TIVA over 24 h was 3.7 (95 % CI, 3.0–4.7) [41]. There is also the added benefit that dexamethasone 0.1 mg/kg has been shown to be an effective adjunct in multimodal strategies to reduce postoperative pain and opioid consumption [42]. Recently, an increasing number of studies have used a higher dose of dexamethasone 8 mg IV. One such study found that preoperative dexamethasone 8 mg enhances the postdischarge quality of recovery in addition to reducing nausea, pain, and fatigue [43]. However, a recent meta-analysis showed no clinical advantage of higher dose 8–10 mg IV dexamethasone compared with 4–5 mg IV [41].

There is conflicting data concerning the safety of dexamethasone. Most studies show that a single dose of perioperative dexamethasone does not appear to increase the risk of wound infection [40, 42]. However, dexamethasone has been shown to cause significant increases in blood glucose levels that occur 6–12 h postoperatively in normal subjects [44, 45], those with impaired glucose tolerance [45], type 2 diabetics [46], and obese patients [45]. Additionally, one recent retrospective case-control study showed that patients who developed a postoperative wound infection were significantly more likely to have received a single perioperative dose of dexamethasone 4–8 mg IV, and thus concluded that dexamethasone may increase the postoperative risk of wound infection [47]. However, these patients were also less likely (p = 0.001) to have received a prophylactic antibiotics. Based on the totality of evidence, a single dose of 4–8 mg dexamethasone is not associated with an increased risk and hence is recommended [48, 49]. The increase in blood sugar is predictable, and should be monitored in labile diabetics. Methylprednisolone 40 mg IV is also effective for the prevention of late PONV, as it has a similarly long half-life [50, 51]. There is no evidence to suggest that methylprednisolone differs from dexamethasone in terms of adverse effects.

Droperidol and haloperidol are two butyrophenones that have been shown to be effective for the prophylactic prevention of PONV. An effective dose of droperidol for preventing PONV is 0.625–1.25 mg IV [52–54]. It has been shown to have similar efficacy to ondansetron, with an NNT of approximately 5 for the prevention of PONV within 24 h [54]. However, in 2001, the FDA issued a black box restriction on the use of droperidol over concerns of QTc prolongation. Despite this, studies have shown that droperidol has equal effects on the QTc interval as ondansetron [55, 56]. Furthermore, the combination of droperidol and ondansetron, while being more effective than either drug alone for the prevention of PONV, had no greater effect on the QTc interval than either drug alone [57]. It is believed, therefore, that at dosing levels appropriate for the prevention of PONV, droperidol can be safely used without significant cardiovascular events.

At low doses, 0.5–2 mg IM or IV, haloperidol effectively reduced the risk of PONV with an NNT between 4 and 6 [58]. Haloperidol carries a warning of QTc prolongation on its label; however at these low doses cardiac arrhythmias have not been reported. Nonetheless, haloperidol is not regarded as a first antiemetic of choice, as it also carries risk of extrapyramidal symptoms. The use of haloperidol for PONV and the IV route of administration is not an FDA-approved indication.

Antihistamines are older drugs with antiemetic effects. The antihistamine dimenhydrinate can be used as an antiemetic in doses of 1 mg/kg IV, where it has shown to have similar efficacy to the 5-HT₃ receptor antagonists, dexamethasone, and droperidol [59, 60]. Despite this, there are too few direct comparisons with other antiemetics; furthermore, there is not enough data to determine optimal administration timing, dose response, or side effect profile [10].

Transdermal scopolamine (TDS) has the added benefit that it can be applied the evening before surgery or 2–4 h before the start of anesthesia because of its 2- to 4-h onset of effect [61, 62]. TDS has been shown to have equal effectiveness in single drug therapy studies comparing it to ondansetron and droperidol [63]. TDS prevented nausea and vomiting up to 24 h after surgery with an NNT of 6 [64]. The most common adverse effects include visual disturbances (NNH = 5.6), dry mouth (NNH = 13), and dizziness (NNH = 50) [64].

Perphenazine and metoclopramide are two phenothiazines that have been used for the management of PONV. A review of 6 RCTs showed a relative risk reduction (RRR) of 0.5 (95 % CI, 0.37–0.67) for PONV when using 5 mg IV perphenazine [65]. This same review showed no significant increase in sedation or drowsiness when compared with a placebo [65]. Metoclopramide is not effective at 10 mg dose. However, efficacy has been demonstrated at higher doses. The NNT for metoclopramide 10, 25, and 50 mg for PONV at 24 h is 30, 16, and 11, respectively [66]. The NNH for extrapyramidal symptoms with 25 or 50 mg metoclopramide is 140 [66].

The most widely studied of the neurokinin-1 (NK-1) receptor antagonists and the only one currently available is aprepitant. Compared with ondansetron, aprepitant has a longer duration of action, with a half-life of 9–13 h [67]. While it was shown to be similar to ondansetron in achieving complete response (no vomiting and no use of rescue antiemetic) for 24 h after surgery, it was shown to be more effective than ondansetron at preventing vomiting at 24 and 48 h after surgery and in reducing nausea severity during the first 48 h after surgery [68, 69]. The side effect profile of aprepitant is similar to that of ondansetron [69]. The role of aprepitant as a routine prophylactic agent has not yet been established due to limited clinical experience and higher costs [70]. Casopitant and rolapitant are similar long-acting NK-1 antagonists which have shown comparable efficacy, but have not yet been approved for use [10].