The lead author and clinical colleagues were charged with converting a “same-day admission” orthopaedic sports medicine surgical procedure [arthroscopic anterior cruciate ligament reconstruction (AACLR)] to “same-day discharge home.” This preceded landmark PONV manuscripts by Dr. Christian Apfel and colleagues in 1999 [addressing predictive risk factors (9)] and 2004 [addressing multimodal therapeutic efficacy (10)]. Same-day discharge after AACLR was newly required (by market forces) in the mid-1990s, since third-party payers in the United States were no longer reimbursing hospitals for admission, instead only providing a capitated payment. The previous hospital culture of “all general anesthesia (GA) all the time” was soon transformed to primarily regional anesthesia (RA), or less commonly to combined RA-GA. One objective was to have patients emerge from anesthesia sufficiently quickly and symptom-free to reliably and safely fast-track through the traditional Phase 1 post-anaesthesia care unit (PACU) (11). Patients could then be reunited with their families sooner, en route to earlier discharge home. We learned (before any available PONVCG) that routine MM-AEPPx was central to minimizing both patient symptoms and postoperative nursing workload, en route to same-day discharge. We did this by (i) avoiding volatile agent GA (long before documented carbon footprint concerns), (ii) using propofol sedation, (iii) providing routine nerve blocks instead of opioids for analgesia, and (iv) using perphenazine and dexamethasone (and propofol sedation when GA was not used) as low-cost, off-patent MM-AEPPx, irrespective of risk factors. Phase 1 PACU Bypass fast-track criteria that were used [now called the WAKE Score (11)] emphasized multimodal prevention of pain/hyperalgesia, PONV, itching, and shivering. WAKE criteria called for routine MM-AEPPx (instead of rescue), since the postoperative routing goal of such patients was Phase 1 PACU Bypass (fast-tracking). This practice pattern of fast-tracking preceded international publications and professional societies addressing enhanced recovery after surgery (ERAS). Meanwhile, most every antiemetic study cited by the serial PONVCG publications (5–7), including the most recent (8), entailed a forced admission to the PACU Phase 1 recovery room after surgery, with no apparent option (or motivation) for PACU Bypass/fast-tracking. Meanwhile, our pre-ERAS observational data entailed ∼50% Phase 1 fast-tracking after GA including femoral nerve block, vs. ∼80% Phase 1 fast-tracking after RA (without GA) including femoral nerve block (12).

Outcomes involving shoulder surgery under RA without GA have been published (13). This population benefitted from “more than consensus-guided” MM-AEPPx (in its era), showing 94% fast-tracking of ∼450 shoulder arthroscopy patients. Those that received perphenazine-dexamethasone-propofol had a <10% PONV rate, vs. a 16% PONV rate with one or fewer antiemetics co-administered with propofol (P < 0.005) (13).

Apparent frustration shared by aforementioned bariatric surgery societies (4), regarding the lack of success of PONVCG to date, is pertinent. Their response included recommendations to “Administer at least 2 or 3 antiemetics from different pharmacologic categories even in the absence of patient-related risk factors.” One of these bariatric-recommended AEPPx categories is anticholinergics (specifically transdermal scopolamine), for which a recent contemporaneous Cochrane review (14) provided data that contradicted the bariatric societies' position. Specifically, adding usually-efficacious aprepitant to scopolamine has not been shown to yield a beneficial effect on PONV (14), which directly contradicts the bariatric societies' guidance statement “NK-1 antagonists and an anticholinergic transdermal scopolamine patch should be considered prior to the patient's arrival at the facility.” (4).

Most research addressing ITM-induced PONV originates in the obstetrical population. However, recent ERAS-related and other work in joint replacement and colorectal surgery populations are sufficiently illustrative on their own merits. For joint replacement, a recent prospective study (15) showed a 77% (27/35) PONV rate (on POD#0) after 0.2 mg ITM, and a 51% (18/35) PONV rate after 0.1 mg ITM, in the absence of any PONVCG AEPPx having been used. In GA-ITM for colorectal surgery, retrospective matched-cohort data review of ERAS patients showed a 44% PONV rate (POD#0–1) after ITM and GA with either aprepitant-ondansetron-dexamethasone or perphenazine-ondansetron-dexamethasone (16). No PONVCG ever specifically addressed ITM as its own, separately-quantifiable, emetogenic entity; it seems that 77% with no AEPPx, and 44% with consensus-guided AEPPx, are still unacceptably high PONV rates.

Anesthetists are in a unique position to provide benefit for inpatients (e.g., such as those routed through ERAS programs) and outpatients. First, avoiding GA (when feasible) no longer binds the clinician to subscribe to nearly 20 years of questionably-successful iterations of GA-based (5–8) PONVCG. Similarly, ITM use (in GA or RA-only contexts) absolves clinicians from using the same GA-based PONVCG, because ITM is an otherwise unstudied (i.e., by PONVCG) high-risk emetogenic entity. The 2003–2020 evolved GA-PONVCG (5–8) may have pre-ordained a “70% PONV prevention success” rate as “acceptable,” in the possible guise of “first do no harm.” This “acceptability” threshold seems predicated on 65%–70% predictive value of applying PONVCG AEPPx strictly based on traditional risk factors (8). Why the PONVCG authors seem “accepting” of “70% success” may be related to possible apprehension of being judged as “not cost effective”, “too costly”, or perhaps “conflicted” (in a conflict-of-interest financial context).

In 2003 (5), the industry-sponsored PONV consensus panel cited 84 references addressing prophylaxis/treatment. This was followed by a 2007 (6) update (166 references), this time originating from the Society for Ambulatory Anesthesia (without industry sponsorship expressed or implied in its listing). Another from the Society for Ambulatory Anesthesia followed in 2014 [335 references (7)], followed by the 2020 (8) combined statement from the American Society of Enhanced Recovery and the Society for Ambulatory Anesthesia (430 references). The 2007 guidelines were released just before ondansetron lost patent protection (leading to a cost decrease from US$15–20/dose to <$1/dose), rendering the “new” guidelines “already” (seemingly) no longer current with respect to fiscal responsibility. The initial draft of the 2014 (7) guidelines proposed to remove the anti-dopaminergic antiemetic perphenazine from the recommendation list that had been in place in both 2003 (5) and 2007 (6) (see Supplementary Materials Item 1). Later, the 2020 Guidelines (8) featured at least 10 of its 22 authors who also authored contemporaneous (2013–2019) industry-sponsored studies of the now patent-protected amisulpride.

Our perspective is that authors from the described PONV consensus conferences (5–8) may not have appropriately relaxed their original 2003 (5) position of universal multimodal PONV prophylaxis as not being cost-effective, in light of newly inexpensive multimodal options (particularly aprepitant and palonosetron, in 2020). We propose, and have recently (2) demonstrated an alternative position, achieving ≤∼10% PONV the day-of and day-after surgery through the routine use of 5-medication MM-AEPPx (aprepitant, palonosetron, perphenazine, dexamethasone, and diphenhydramine), achieving this associated success rate amidst care plans involving emetogenic ITM.

We invite stakeholders to consider value in reducing PONV by ∼85% (2) compared to PONVCG, while not creating other frequent side effects. One does not want the side effect to be more of a nuisance (e.g., dry mouth with scopolamine) than the primary symptoms being prevented (PONV). It is conceivable that PONVCG to date have not endorsed a multimodal plan due to concerns about dysrhythmogenicity of medications in combination, or because of sparse evidence of medication interactions that is both counterintuitive and contradictory to antiemetic goals [such as the aforementioned efficacy-reducing aprepitant-scopolamine interactions (14)]. Whatever the origin, we describe low-cost clinical routines which included ITM use that have been associated with a ≤∼10% incidence of PONV and ≤11% pruritus (2) at any point throughout postoperative days zero-to-one (POD#0–1), with no apparent adverse effects.

As a frame of reference, our (2) institutional costs for aprepitant went from US$85/dose to US$10/dose (40 mg orally). Preoperative pill dosing “workload” with “one pill” encourages concurrent multiple pill dosing that can also include the non-sedating anti-dopaminergic perphenazine (most commonly 8 mg) in the absence of contraindications, even though oral premedication may be a cultural change in many centers (17). Such centers could also confidently implement the premedication concept for an antihistamine [either the long-generic oral dimenhydrinate (see below), or the inexpensive generic IV preparation diphenhydramine], along with non-sedating oral analgesics such as dextromethorphan and type-2 inhibitors of cyclo-oxygenase (such as celecoxib or meloxicam). When oral aprepitant is used in women with child-bearing potential, alternative contraceptive methods in addition to any baseline hormonal-based measures are recommended for 30 days after aprepitant use.

A palonosetron dose of 75 µg is consensus-recommended (8) in adults, and was shown to be superior for ITM-related late-onset vomiting prevention [when compared to ondansetron 4 mg (18)]. Understanding that most monotherapy PONV outcomes did not differ (53% after ondansetron, 43% after palonosetron) in this ITM study in women, it is important to note that late vomiting differences were present, favoring palonosetron (11% vs. 27%, P = 0.018) (18). Our (2) institutional cost is US$10 per single-patient 250 µg palonosetron syringe (down from US$90), vs. <US$1 for 4 mg IV ondansetron. In meta-analyses, palonosetron (75 µg) monotherapy has been more effective for general anesthesia (GA)-related PONV than ondansetron (4–8 mg) plus dexamethasone (5–8 mg) (8). We cannot imagine clinical situations favoring ondansetron over palonosetron, but inpatient hospital controls may wish to avoid any rescue 5-HT3-antagonist dosing for the first 40 h after palonosetron is given, to avoid theoretical risk of QT prolongation (for patients carrying any such risks). We assume that palonosetron is preferred over ondansetron for ITM-induced pruritus based on duration of action; the 5-HT3 antagonist drug class is accepted as efficacious for prophylaxis against ITM-related itching (19). We have since cautiously escalated this 75 µg “base dose” to 150–250 µg, in non-ITM and ITM contexts respectively, based on early internal quality improvement data (improvements in next-day PONV prevention) that we hope to report in the near future, understanding that the 250 µg palonosetron dose is already approved for chemotherapy-induced nausea and/or vomiting without apparent dysrhythmogenic risk.

Injectable perphenazine (5 mg) is non-sedating, and was reported in systematic review (20) to be in evidence-category “A1” in 2003–07 PONV guidelines (5, 6). However, injectable perphenazine has not been available in the United States, for example, since 2001. Oral perphenazine is ∼20%–40% bioavailable (21). No industry sponsor, to our knowledge, has underwritten the prospective study costs of non-sedating, off-patent oral perphenazine (US$0.39 for 8 mg) as antiemetic. If the under-70-year patient is not diagnosed with Parkinson's disease, is not taking a simultaneous antidopaminergic psychotropic medication, and has had no past extrapyramidal adverse reactions to popular antidopaminergic legacy antiemetics such as prochlorperazine, we consider the screening process as ∼99.97% successful for the avoidance of possible extrapyramidal complications after 8 mg single-dose po perphenazine (22). Additional safety research could entail similar study of patients over 70–75 years with a single 4 mg oral dose [instead of 8 mg as in (22)]. We assume similar contraindications and risks for off-patent perphenazine as those for patent-protected amisulpride. The lead author does not use perphenazine in patients with Parkinson's disease, but has used lower-dose (e.g., 4 mg PO) perphenazine uneventfully and confidently in non-Parkinsonian patients with conditions such as essential tremor and/or restless legs syndrome, without any observed sequelae to date. In patients with co-existing antidopaminergic therapy (in psychiatric contexts) that includes routine benztropine for offsetting known extrapyramidal symptoms, the lead author will not administer perphenazine. Stakeholders are reminded that for patients on other anti-dopaminergic agents (e.g., taken the day before surgery, but not yet on the day of surgery upon hospital presentation), one further recommended antiemetic (below in section 2.11) is an antihistamine, which is also successful in offsetting or preventing unwanted extrapyramidal effects (22).

We are not aware of literature addressing benzamides or derivatives (drug class of amisulpride) as having any efficacy against itching/pruritus (warranting its patent-protected higher cost).

Dexamethasone given intraoperatively, in a 4 mg IV dose (10) is long-supported; furthermore, an 8 mg IV dose does not adversely influence surgical site infections after non-urgent non-cardiac surgery (23). Although traditional dose-limited studies of dexamethasone (i.e., 4 mg) guided the primary practice pattern for the lead author for many years, this latter study (23) showing safe dose escalation to a popular standard dose entailing 8 mg (or 0.1 mg/kg) in adults appears to confer both confident antiemetic effects and meaningful non-opioid analgesic effects.

Dimenhydrinate is PONV-consensus-recommended (8), but with unknown effects on ITM-induced pruritus. If one were to reduce other concomitant sedatives during surgery, we forecast that inexpensive 25–50 mg oral dimenhydrinate would be unlikely to harm. Instead, we forecast that oral dimenhydrinate, co-administered with aprepitant/perphenazine above, and being part of the described (2) 5-medication MM-AEPPx plan will prove helpful for both PONV and pruritus. Similarly simple and inexpensive (instead of oral dimenhydrinate) would be IV diphenhydramine 12.5–20 mg, given before any emetogenic stimuli (such as ITM) are incurred [see (2)]. Even with a low-dose antihistamine and its potential sedative properties, one may wish to be cautious with usual customary dosing of typical non-opioid analgesic strategies (with unwanted sedative potential) such as dexmedetomidine or high-dose lidocaine infusions, at least until more multimodal research addressing sedation outcome data are available. Antihistamine use as part of 5-drug MM-AEPPx should not hinder dosing of other non-sedating non-opioid co-analgesics such as acetaminophen, non-steroidal anti-inflammatory drugs, cyclo-oxygenase (type 2) inhibitors, or NMDA-antagonists such as ketamine or magnesium; these co-analgesics seem to be logical and successful opioid-sparing techniques to sustain, in efforts to aim for “opioid-free”. In either case, the lead author is anecdotally (based on institutional quality data review) far more confident in the 24-h anti-emetic (and anti-itching) benefit of the described fronted-dose of antihistamine, when compared with far less incremental confidence with respect to sustained analgesia from an intraoperative dexmedetomidine infusion, for example, which is turned off typically before the end of surgery.

Recently, an enhanced recovery joint replacement study (24) reported on a complementary advance with respect to both establishing and sustaining antiemetic prophylaxis benefit. This was a retrospective case-control study achieving 50% reduction of PONV on postoperative days 1–3 (24). This study did NOT involve GA (rather, spinal anesthesia with local anesthetics, but not ITM). Consensus guidelines [2003–2020 (5–8)] have directed their attention toward GA (±opioids), not necessarily toward RA, nor toward spinal anesthesia replacing GA. The last-known systematic review of PONV in RA [2003 (25)] did not cite (as problematic for RA) then-known GA PONV risk factors of female gender, non-smoker, and/or past history of PONV (8, 10). However, PONV was not the primary outcome in most of the RA studies of this (25) systematic review. A few years later, the lead author's group analyzed secondary outcomes from a prospective randomized trial (2001–2005) of patients undergoing spinal anesthesia with femoral nerve block and propofol sedation; in the parent study, only 4% of patients (10/233) had PONV on the day of surgery before hospital discharge after 4 antiemetic drugs (perphenazine, ondansetron, dexamethasone, and propofol), irrespective of risk factors (26). More detailed secondary review (27) of PONV outcomes at home [after the parent (26) study's day-of-surgery] similarly showed “antiemetic momentum” to be valuable. Specifically, PONV on the day of surgery was predictive of PONV on postoperative day (POD)#1 [odds ratio (OR) = 16, P = 0.001], while PONV on POD#1–3 were each individually and respectively predictive of PONV on POD#2–4 (OR = 10–14, P ≤ 0.001). In other words, the previous day's PONV outcome was the only predictor of PONV outcomes on all 4 subsequent postoperative days, while usually-expected factors such as opioid consumption and female gender were each only predictive for two of the four postoperative days, with much lower odds ratios (14). Our group's recent ITM and 5-medication MM-AEPPx work also demonstrated the value of antiemetic momentum (OR = 9, P < 0.001) (2), for which we now (for our inpatients) provide daily oral doses of both aprepitant (40 mg) and perphenazine (4–8 mg), serving a multimodal “booster” function.