Secondary Logo

Journal Logo

General Articles: Review Article

Risk Factors for Postoperative Nausea and Vomiting


Author Information
doi: 10.1213/01.ANE.0000219597.16143.4D
  • Free


Postoperative nausea and vomiting (PONV) is the most frequent side effect after anesthesia (1), occurring in ∼30% of unselected inpatients and up to 70% of “high-risk” inpatients during the 24 h after emergence (2). Its incidence may be less frequent in ambulatory surgery than in inpatient surgery, but PONV may be under-recognized in the outpatient setting, where patients quickly leave direct medical oversight (2,3). Although PONV is almost always self-limiting and non-fatal (4), it can cause significant morbidity, including dehydration, electrolyte imbalance, suture tension and dehiscence, venous hypertension and bleeding, esophageal rupture, and life-threatening airway compromise, although the more severe complications are rare (5,6). Each vomiting episode delays discharge from the recovery room by ∼20 min (7).

In a preoperative survey, patients ranked emesis as the most undesirable and nausea as the fourth most undesirable of 10 negative postoperative outcomes; postoperative pain ranked third in this study (8). In another study, patients were, on average, willing to pay $56 out-of-pocket to avoid PONV; the figure increased to $73 and $100 in patients who had experienced postoperative nausea or postoperative vomiting, respectively (9). Because patients find PONV so highly unpleasant, it has been proposed that PONV management, similar to pain management, could be considered an end unto itself (6).

However, interventions to prevent PONV are not needed in the majority of the general patient population who, even without prophylaxis, will not suffer these symptoms. In addition, current interventions may cause side effects and may entail substantial effort or expense (7,10). Moreover, current interventions lack universal efficacy, especially as monotherapy (11).

Therefore, it is important to direct interventions to the patients most likely to experience PONV (7,10), especially in the case of combination therapy or “multimodal management,” which are the most effective management strategies now available but which are also more costly and can have added side effects. Knowledge of PONV risk factors is essential to this process.

However, current understanding of risk factors for PONV is incomplete, in part because much remains to be elucidated about the pathophysiology of these symptoms, particularly their molecular biology. Obtaining understanding of PONV risk factors and pathophysiology is complicated by the multifactorial nature of PONV because of the involvement of multiple receptors and stimuli (12,13). At least 7 neurotransmitter types are documented, or believed to be implicated, in PONV, namely serotonin, dopamine, muscarine, acetylcholine, neurokinin-1, histamine, and opioids. Stimulation of the vestibular-cochlear, glossopharyngeal, or vagus nerves also may be involved.

Nonetheless, understanding of PONV risk factors has dramatically improved since the advent in the early 1990s of clinical studies that used more sophisticated multivariable statistical analyses and stratification. The wider use of meta-analyses and systematic reviews also has added knowledge. In addition, the development and validation of predictive scoring systems based on the results of the aforementioned clinical studies, and the publication of trials using the scoring systems to allocate and tailor prophylaxis, provide guidance for the application of risk factors to everyday practice.

The aim of this qualitative review is to summarize the findings of the past decade and a half of research on PONV risk factors in both adults and children. The review focuses on key prospective clinical studies (based on author opinion) involving consecutive adult or pediatric patients or both, published in peer-reviewed journals, that controlled for multiple variables, i.e., patient characteristics and types of anesthesia or surgery, and in most cases, did not report having given anti-PONV prophylaxis. Important meta-analyses and systematic reviews also are discussed. These publications were identified based on searches of the PubMed® and Medline® databases for the period January 1990–June 2005 under the keywords “postoperative nausea and vomiting” or “postoperative nausea” or “postoperative vomiting” and “risk factors.” Additional publications were identified based on citations in papers found in the database searches and from the author’s knowledge. The present review also seeks to present the clinical implications of the recent PONV research and to suggest avenues for further investigation. The review begins by explaining key PONV terminology and classifications.

Definition and Classification of PONV

PONV encompasses three main symptoms that may occur separately or in combination after surgery. Nausea is the subjective sensation of an urge to vomit, in the absence of expulsive muscular movements; when severe, it is associated with increased salivary secretion, vasomotor disturbances, and sweating (14). Vomiting or emesis is the forcible expulsion through the mouth of the gastric contents. Vomiting results from coordinated activity of the abdominal, intercostal, laryngeal, and pharyngeal muscles, including retrograde giant contraction of the intestines, relaxation of the gastric fundus, closure of the glottis, and elevation of the soft palate. This activity is associated with increased heart rate and breathing and with sweating (15). Retching is an unproductive effort to vomit (14). Retching and vomiting are collectively termed emetic episodes.

PONV may take place in single or multiple episodes, which may last minutes, hours, or even days (6). It is classified as early, occurring up to 2 to 6 h after surgery, or late, occurring up to 24 or 48 h after surgery, with the exact cut-off times depending upon the individual investigator’s definition. As may be inferred from this lack of a standard cut-off time, the delineation is somewhat arbitrary and is related to the patient’s location at the time of evaluation for the symptoms, e.g., the postanesthesia care unit, surgical or other ward, or home. However, there are suggestions that early and late PONV may differ at least somewhat in their pathogenesis. The use of volatile anesthetics may be a main cause of early PONV (5,6). Opioid-induced symptoms and motion sickness caused by transportation from the PACU to the ward or from the hospital to the home may account for much of late PONV (16–19). However, for the most part, PONV research has focused on identifying risk factors themselves rather than their time of activity.

Risk Factors for PONV


PONV risk factors have been described in the literature since the late 1800s (20). Traditionally, investigation focused on a single potential factor at a time, with little to no attempt to control for other variables, i.e., to account for the possible independent effects of additional factors (21,22). In studies with these drawbacks, the true influence of the investigated risk factor remained unclear.

The modern era in PONV risk factor research began in the early 1990s, with publication of the first studies that attempted to simultaneously identify multiple risk factors and, in so doing, used regression models to control for a wide variety of variables (21,22). At least 20 key multivariable studies have been published in English; they are summarized in Table 1. Nearly all these studies were prospective and relied on logistic regression analysis (23). Logistic regression analysis uses modeling in which a binary or dichotomous dependent variable, that is, an outcome comprising two possible categories (e.g., PONV: yes or no), is described as a function of one or more independent variables. Logistic regression analyses generate an odds ratio (OR) for each factor examined. The OR is the ratio of the likelihood of an outcome in a group with the given risk factor to the likelihood of the outcome in a group lacking that factor. The statistical significance of each OR is also assessed. In addition, the 95% confidence interval (CI) of the OR, i.e., that is, the range of values that is 95% likely to include the true OR in the study population, is calculated. When the lower limit of the 95% CI of its OR exceeds 1.00, there is little doubt that a given factor increases PONV risk. For a more detailed statistical explanation of logistic regression and the related terminology, readers may refer to several published references (23,24).

Table 1
Table 1:
Characteristics of Key Multivariable PONV Risk Factor Studies
Table 1
Table 1:

The potential risk factors studied thus far (Tables 2 and 3) may be classified as patient-, surgery-, or anesthesia-related and as fixed or variable, that is, amenable to change by the patient’s caregivers (16,21,25). Most patient and surgical technique-related factors are fixed; some other surgery-related factors and some anesthesia-related factors are variable.

Table 2
Table 2:
Key PONV Risk Factor Findings In Adults and Children
Table 3
Table 3:
PONV Risk Factors: Other Potential Factors Examined in Multivariable Studies and Candidates for Future Investigation

Key Risk Factor Findings to Date.

Table 2 summarizes key findings of modern risk factor studies. To facilitate application of these findings to everyday practice, Table 2 classifies risk factors as “well-established,” “possible,” or “disproved” (according to expert consensus and author opinion) as well as by relationship to the patient, surgery, or anesthesia.

Patient-Related Factors

Probably the strongest risk factor identified is female gender from puberty on: all adult studies listed in Table 1 (5,17–19,21,22,26–34) concurred in identifying female gender as a risk factor, and no study has contradicted this finding. All adult risk scoring systems include this factor (Table 4). In most studies, ORs for this predictor have ranged from 2.0–4.0, reflecting a twofold to fourfold increased PONV risk for adolescent and adult females (18,21,22,27–29,31,32,34). That prepubescent girls apparently lack increased likelihood of PONV (35,36) could imply that the risk relates to hormonal factors. However, although initial studies (37,38) reported increased susceptibility to PONV during the first week of the menstrual cycle, early stage of the menstrual cycle has been disproved as a risk factor by a subsequent study (39) and in a systematic review (40).

Table 4
Table 4:
Overview of Risk Factors Use in Risk Scoring Systems

Nonsmoking status has been identified as an independent PONV risk factor in numerous adult studies (17–19,22,27–29,34) as has history of PONV and/or motion sickness (5,17,19,21,27–32); intriguingly, a recent study in children also found history of PONV in a parent or sibling to be a risk factor (36). There have been few contradictory reports (21,31,32). Nonsmoking status is included in all but one adult risk scoring system, and history of PONV or motion sickness in all risk scoring systems (Table 4). Most studies have found ORs of ∼1.5–2.5 for nonsmoking status and of ∼1.8–3.1 for history of PONV, motion sickness, or both.

A number of investigators also have identified childhood after infancy and younger adulthood as independent PONV risk factors (5, 17, 22, 26, 27, 30, 31, 33–36). For example, 2 reports noted a >10% decreased risk for every decade of age in adults (17,31). A study in children age ≤14 yr found a sharp increase in PONV risk around age 3, with a 0.2%–0.8% per year increase in risk thereafter, depending on the presence of other risk factors (36). However, age is included in only a minority of risk scoring systems (Table 4).

Possible PONV risk factors include better ASA physical status (19,22) and a history of migraine (postoperative nausea only) (18,19). A recent adult study found higher scores on the Spielberger State-Trait Anxiety Inventory anxiety scale or on the Amsterdam Preoperative Anxiety and Information Scale anxiety section to be weak PONV risk factors (OR, 1.01; 95% CI, 1.00–1.02; P = 0.04 and OR, 1.04; 95% CI, 1.02–1.05; P = 0.02, respectively); their inclusion in the investigators’ risk scoring system did not improve its discriminating power (26). In contrast, a pediatric study found preoperative anxiety to not be a significant PONV risk factor (41). A meta-analysis of PONV after gynecological surgery (42) and studies in the laboratory-induced motion sickness setting (43) suggest that ethnicity (Dutch or English versus Scandinavian and Chinese or Asian-American versus Caucasian- or African-American, respectively) could be a PONV risk factor. However, two studies using multivariable analyses do not support a role for this characteristic (26,32).

Besides early stage of the menstrual cycle, obesity has been disproved as a patient-related PONV risk factor (44). Interestingly, the systematic review that did so found that the belief in increased body mass index as a risk factor apparently largely stemmed from a “chain reaction” of 14 review articles misquoting or misinterpreting 4 original studies.

Surgery-Related Factors

Increasing duration of surgery has been shown to be an independent PONV risk factor by a few well-conducted studies in adults (17,19,29,34) or children (36). An outpatient study found that each 30-min increase in surgery duration increased baseline PONV risk by 60% (17). However, although type of surgery has been identified as a risk factor in numerous reports (5,17–19,22,25,28,30,33,45–47), its status as such is still somewhat controversial; the specific procedures implicated as particularly emetogenic sometimes vary among studies. Types of procedures that may be viewed as possible risk factors include intraabdominal (5,11,17–19,22,25,26,28,30,33,36,45–48), laparoscopic (5,11,17–19,22,25,26,28,30,33,36,45–48), orthopedic, major gynecological (5,11,17–19, 22, 25, 26, 28, 30, 33, 36, 45–48), ear, nose and throat (ENT) (5,11,17–19,22,25,26,28,30,33,36,45–48), thyroid (28), breast (5,11,17–19,22,25,26,28,30,33,36,45–48) and plastic surgery (5,11,17–19,22,25,26,28,30,33,36,45–48), as well as neurosurgery (5,11,17–19,22,25,26,28,30,33,36,45–48), and, in children, hernia repair (46), adenotonsillectomy (46), strabismus (5,11,17–19, 22, 25, 26, 28, 30, 33, 36, 45–48), or penile surgery (5,11,17–19, 22, 25, 26, 28, 30, 33, 36, 45–48), and orchiopexy (46). Half of risk scoring systems include duration of surgery, and several incorporate one or more types of surgery (Appendix). Other possible surgery-related PONV risk factors include less preoperative or intraoperative fluid administration (49,50) or intraoperative colloid versus crystalloid administration (51), when a large volume of crystalloid in a prolonged surgery may result in gastrointestinal tissue edema leading to an increased incidence of PONV.

Anesthesia-Related Factors

Numerous anesthesia-related variables have been well established as PONV risk factors, including use of volatile anesthetics (5,11,34,52), nitrous oxide (11,26,34,53), balanced inhaled versus total IV anesthesia (11,34,52,54), and large-dose (≥2.5 mg) neostigmine (55). The choice of volatile anesthetic, e.g., isoflurane versus sevoflurane versus enflurane, appears not to affect the risk of PONV (5,11). Use of intraoperative (34,56) or postoperative (18,19,21,22,28,29,31,32,57,58) opioids and larger perioperative (59) and postoperative doses of these drugs also have been implicated as associated with PONV (60–63). However, some contradictory findings have been reported with respect to postoperative opioid use in adults (26), intraoperative or postoperative opioid use in children (36), or intraoperative opioid use in a mixed adult and pediatric population (5). Interestingly, despite the relatively large number of anesthesia-related variables identified as risk factors, most risk scoring systems do not include any, and the remainder of the systems include only a few (Table 4).

Administration of a long-acting rather than a short-acting opioid is, at best, a possible PONV risk factor. Although a small recent study observed an association between use of fentanyl versus remifentanil as an adjunct to propofol maintenance (64) and PONV, another similarly sized study found no association of alfentanil versus remifentanil use and PONV (65). Moreover, a 5199-patient multinational multifactorial designed study of anti-PONV interventions (11) failed to find fentanyl versus remifentanil as a PONV risk factor.

Far more likely, but not yet well established, anesthesia-related PONV risk factors include longer duration of anesthesia (5,17,22,27,48) or general versus other forms of anesthesia, e.g., regional or sedation (17,18,66). Together with postoperative opioid or isoflurane use, they comprise the anesthesia-related risk factors used by current risk scoring systems (Table 4). Use of standard (30%) rather than supplemental (50% or 80%) oxygen seems to have been disproved as a risk factor (11,67–69), despite early evidence of its validity as such (70,71).

Limitations of Modern Research and Suggestions for Future Investigation.

Although recent studies have vastly improved knowledge of PONV risk factors, identification of such factors remains imperfect (72,73). Five limitations of research published thus far should be borne in mind.

First, there are substantial gaps in the list of potential risk factors investigated (Table 3). Studies continue to pursue an essentially epidemiological approach, focusing on readily discernible clinical factors. However, genetic and other molecular biological patient characteristics have not been extensively examined, and even certain clinical characteristics remain under-investigated (Table 3).

For example, in a thought-provoking editorial (74), Sweeney highlights as potential PONV risk factors the degree of expression and activity of selected cytochrome P (CYP) 450 hepatocellular enzymes. CYP450 enzymes metabolize many drugs, including widely used anesthetic and analgesic drugs and antiemetics (75,76). The greater the expression and activity of a CYP450 enzyme, the more rapid the metabolism of its substrate drug(s). Based on those characteristics, individuals may be classified as “poor,” “intermediate,” “extensive,” or “ultrarapid” metabolizers of the drug(s) (77).

In addition, CYP450 enzyme synthesis may be stimulated or suppressed by environmental influences. Sweeney has speculated that the protective effect of smoking against PONV might relate to the induction of CYP450 enzymes by polycyclic aromatic hydrocarbons. These hydrocarbons are components of the “tar” portion of cigarette smoke. Other clinical characteristics that affect CYP450 enzyme expression, for example, the consumption of alcohol, of commonly prescribed medications such as cimetidine, erythromycin, or terfenadine, or of vegetables including cabbage, Brussels sprouts, cauliflower, or red pepper, should be investigated as potential PONV risk factors. Further, gender and racial differences have been documented in CYP450 enzyme expression. Other suggestions that ethnicity merits further study as a potential PONV risk factor come from a meta-analysis of PONV after gynecological surgery (42) and from motion sickness studies, in which Chinese or Asian-American subjects were significantly more susceptible to symptoms induced by a standardized drum rotation procedure than were European- or African-American subjects (43).

A second limitation of PONV risk factor research is the difficulty of controlling for subtle clinical factors, particularly in smaller or single-center studies. For example, unusual proficiency of particular anesthesiologists or surgeons might mask the nature of a procedure that would be emetogenic in less skilled hands (13).

A third limitation of recent PONV risk factor research is variation in outcomes and data collection methods. Some studies have considered nausea or emetic events as separate outcomes, some as a combined outcome. Although there is an obvious, intuitive relationship between nausea and emetic events, there are important pathophysiological differences between the two. Nausea is a subjective feeling and a conscious cortical activity; emetic episodes are an autonomic reflex and as such, directed by the brainstem (14). The symptoms are not inevitably linked: in a single-center, 671-patient study, Stadler et al. (18) observed nausea or vomiting alone in 11% and 2% of patients, respectively, and together in 8%. In another study, involving 587 adults (5), 18% of patients suffered only nausea, 4%, only vomiting, and 22% both.

The Stadler et al. study identified overlapping but not identical risk factors for nausea or vomiting. Female gender, nonsmoking status, use of postoperative morphine analgesia, general versus regional anesthesia, and urological versus ENT procedures were independent predictors of both symptoms, but abdominal or gynecological versus ENT procedures and history of migraine were significant or near-significant predictors only for nausea (18). On the other hand, some argue that the rarity of postoperative vomiting in the absence of nausea suggests that postoperative nausea should be viewed as a symptom for potential vomiting. Further research is required to clarify the relationship between the two symptoms.

Moreover, some studies have defined PONV as recorded or volunteered symptoms, whereas many others have included in the definition symptoms reported in response to a specific query. One study (34) used administration of rescue antiemetics as its sole criterion for presence of PONV. Most studies have collected data by asking patients about specific symptoms but a few have relied on chart reviews. In a large multicenter risk factor study that examined the issue, patients reported PONV far more often than was shown in their charts (22), and it is believed that direct and specific questioning captures a larger percentage of actual PONV incidence than does spontaneous patient report (22,78). Furthermore, a heavy nursing or anesthesiologist workload may lead to under-observation of emetic episodes (17). The nature and severity of the PONV noted in a study obviously may affect the accuracy or applicability of its independent risk factor findings.

Gaps in the patient populations studied are a fourth limitation of PONV risk factor research. Only a single major multivariable study has focused on outpatients (17), and only two have focused on children (36,48). This deficiency raises at least some questions about the general applicability of findings that derive from adult inpatients.

One further limitation of current PONV risk factor research is the difficulty in separating “true” from “surrogate” risk factors (21,79). This difficulty relates to deficiencies in knowledge of PONV pathophysiology and to the peril in epidemiological research of confusing association and causality. For example, certain types of surgery, e.g., gynecological procedures, might be surrogate risk factors for the true risk factor of female gender. Postoperative opioid use might be a surrogate for pain level, or vice versa, and certain types of surgery, e.g., orthopedic shoulder procedures, also might be a surrogate for opioid use or pain level. Duration of surgery might be a surrogate for duration of anesthesia or vice versa. Thus, even if multivariable analysis identifies an independent PONV predictor, factors potentially underlying that predictor must be considered when applying the finding in clinical practice (79).

Clinical Application of Risk Factor Findings

Scoring Systems.

A number of groups have sought not only to identify independent PONV risk factors (Table 1) but to develop formulas quantifying a given patient’s likelihood of suffering nausea, emetic events, or both (17,19,21,27,29,36). They have introduced 8 major PONV risk scoring systems (Table 4 and Appendix). With 2 exceptions (17,36), these formulas preponderantly include patient-related factors and only 2 formulas include both surgery-related and anesthesia-related factors (17,26) (Table 4).

Seven of the 8 systems have been validated in additional populations, centers, or both, from those in which the formulas originally were developed. In the original studies devising 4 of the scoring systems, the overall patient populations were randomly divided into an “evaluation” subgroup, in whom the risk factors were first identified and incorporated into a formula, and a “validation” subgroup in whom the significance of the risk factors and accuracy of the scoring system were then tested (17,27,29,36). In addition, 6 of the 8 scoring systems have been validated in separate populations by the original (11,28–30,32,80) and, in some cases, also other investigators (1,31,48,81).

The accuracy of PONV scoring systems, that is, their ability to correctly discriminate between patients who will or will not suffer PONV, has most commonly been tested through calculation of the area under a given system’s receiver operating characteristic (ROC) curve. This curve plots the scoring system’s true-positive rate (sensitivity) against its false-positive rate (1 − specificity). The area under the curve is expressed as a value between 0 and 1. An area under the ROC curve of 0.50 denotes that a scoring system is correct half of the time, i.e., is no better than guessing. An area of 1.0 denotes that the scoring system is correct 100% of the time.

Given the previously discussed limitations of PONV risk factor research, as well as the limited statistical strength of predictors identified thus far (ORs generally 1.5–3.0) (79), it is not surprising that scoring systems have shown only poor to moderate accuracy, i.e., areas under the ROC curve ranging from 0.56–0.785 (Appendix). In other words, these scoring systems achieve a 12%–57% relative improvement over guesswork.

Despite the limitations in accuracy of PONV risk scoring systems, their use to better tailor antiemetic interventions has been shown to significantly reduce the incidence of PONV in general and particularly in high-risk patient populations, while avoiding the expense and potential side effects of prophylactic antiemetics in lower-risk individuals. For example, in a recent study (n = 162) involving adult inpatients undergoing surgery under general anesthesia in several departments, the overall incidence of PONV in the 24 h after surgery was reduced from an expected 47% historically to 36%, a 23% relative decrease (82). Another study in a similar population (n = 428) achieved a significant reduction in the overall PONV rate in the 24 h after surgery from 49.5% historically to 14.3% (P < 0.001), a 71% relative decrease (83). Moreover, the use of a scoring system (combination of risk factors) has been shown to have a greater discriminating power than the use of a single risk factor (80). The Apfel et al. (29) simplified scoring system predicted PONV significantly more accurately than did the single risk factors of the surgical site or a history of PONV or motion sickness: the area under the ROC curve was 0.68 (95% CI, 0.66–0.71) for the simplified scoring system versus 0.53 (0.50–0.56) for surgical site and 0.58 (0.56–0.61) (P < 0.001) for history of PONV or motion sickness.

No scoring system yet has emerged as a “gold standard” based on accuracy. The main improvements in scoring systems have consisted of simplification and, hence, increased user-friendliness, rather than performance enhancement. Koivuranta et al. (19), Apfel et al. (29), and Eberhart et al. (36) found that omission of the constants and coefficients derived from logistic regression modeling only minimally, if at all, diminishes scoring system accuracy. In addition, Koivuranta et al. and Apfel et al. came to the counter-intuitive conclusion that inclusion of more than a few risk factors attains little to no improvement in accuracy. Apfel et al. speculate that the latter observation may be attributable to the limited number of predictive factors identified that are applicable across populations (28).

Thus for adults, Apfel et al. (29) and Koivuranta et al. (19) have been able to create simplified scoring systems removing weighting of predictors and incorporating only 4 and 5 risk factors, respectively (Table 4, Appendix). More recently, Eberhart et al. (48) created a 4-item simplified scoring system for children. Van den Bosch et al. (26) have taken a somewhat different approach that could be characterized as “semi-simplification.” Their scoring system also omits constants or coefficients, and it contains only 5 items. However, rather than scoring each item 0 or 1 (“no” or “yes”), the system assigns different point values to particular alternative variables for each item, so that a nomogram is required to use the system (Appendix). Eberhart et al. (48) and Van den Bosch et al. (81) devised their new scoring systems because they found what they judged to be relatively limited discriminating power of existing formulas in pediatric patients (0.56–0.65) or in their adult patients undergoing a wide range of procedures (0.63–0.66), respectively (Appendix).

The simplified scoring systems obviate laborious calculations and may reduce the scope of required detailed history-taking but have demonstrated equivalent or superior discriminating power compared with more complex formulas (1,28,31). This has been seen in published scoring system comparisons, although these are limited in number. The comparisons also have suggested some differences in accuracy among the various systems. In adults, the Koivuranta et al. (19) simplified system has been shown to have a statistically higher predictive value than the Palazzo and Evans nonsimplified system (21) (0.71 versus 0.68 for postoperative nausea; P = 0.007 and 0.70 versus 0.64 for postoperative vomiting; P < 0.05) (28) and a numerically greater area under the ROC curve (0.66 versus 0.63) than does the Apfel et al. simplified system (29). In children, the Koivuranta et al. simplified system had a significantly larger area under the ROC curve (0.61) than did the Palazzo and Evans system (0.56; P < 0.001) or the Apfel et al. simplified (0.58) or nonsimplified (0.59) systems (P < 0.003 for both Apfel et al. systems) (48). In adults, the Apfel et al. simplified (28) or original (1) systems exhibited significantly greater accuracy than did the Palazzo and Evans formula (0.68 versus 0.64, P < 0.05 for PONV and 0.73 versus 0.68 for postoperative vomiting, P = 0.005, respectively). The Apfel et al. simplified system also showed significantly greater accuracy than the Sinclair et al. (17) nonsimplified formula in one adult study (0.71 versus 0.64; P = 0.008), but the Sinclair et al. system had a significantly larger area under the ROC curve than did either of the Apfel et al. systems in a pediatric study (0.65 versus 0.59 or 0.58; P < 0.003) (48). In that pediatric study, the Sinclair et al. system also had significantly greater discriminating power than did the Palazzo and Evans formula (0.65 versus 0.56; P < 0.001). In judging these comparisons, it should be kept in mind that unlike the other systems, the Sinclair et al. formula was developed in outpatients but all comparisons were in inpatients.

Taken as a whole, the comparisons suggest that for inpatients, the Koivuranta et al. simplified system (19) is perhaps the most accurate, but not vastly more accurate than the Apfel et al. simplified (29) or original (27) or the Sinclair et al. (17) systems. All four of these systems do seem superior to the Palazzo and Evans formula (21), however. The comparisons also suggest that the use of different scoring systems for adult versus pediatric inpatients may increase accuracy.

In conclusion, I believe that their accuracy, and, most importantly, simplicity relative to the other scoring systems make the Koivuranta et al. (19) or Apfel et al. (28) simplified scoring systems the current preferred choice for use in adults, especially inpatients, and the Eberhart et al. (36) simplified system the current preferred choice for use in children, especially inpatients. However, it should be noted that these scoring systems are only moderately accurate in predictive ability.

Clinical and Research Implications.

Modern multivariable risk factor studies have strengthened the belief in the multifactorial nature of PONV and led to the development of a so-called “multimodal approach” to better address this issue (52,84). The innovative feature of the multimodal approach is its reliance on risk factor reduction, e.g., avoidance of volatile anesthetics, in addition to prophylaxis with antiemetics.

Insofar as applying risk factor findings to PONV management, a “decision-tree” approach has been advocated in which patients are divided into “low,” “moderate,” “high,” or “extremely high-risk” populations based on the number or nature of their risk factors or their score on a formula (2,85,86). Consensus is emerging that antiemetic prophylaxis is not cost-effective in low-risk patients (<10% or <20% expected risk) and is appropriate in other patients. Consensus also is emerging that antiemetic prophylaxis may be best accomplished in moderate, high-risk, or extremely high-risk patients with combinations of drugs from different antiemetic classes or of pharmacological plus nonpharmacological interventions (e.g., acupuncture), with multimodal management, or with both (7,11,87) [(88–94)].


Knowledge of independent PONV risk factors is crucial for the optimal use of antiemetic prophylaxis and multimodal management strategies. Modern multivariable studies, meta-analyses, and systematic reviews have greatly increased such knowledge. Independent risk factors identified by modern research, such as female gender from puberty, nonsmoking status, history of PONV or motion sickness, childhood after infancy or younger adulthood, lengthy or emetogenic surgery, or administration of nitrous oxide, volatile anesthetics, or postoperative opioids, may be used in combination to predict, with moderate accuracy, the likelihood of PONV in a given patient. Further PONV research examining patient genetic characteristics and under-investigated potential clinical risk factors and involving outpatients and children should lead to predictive systems with improved discriminating power and applicability. This development, in turn, will enable anesthesiologists to better identify at-risk patients, further reduce the incidence of PONV, and increase the safety and cost-effectiveness of PONV prophylaxis.

PONV Risk Scoring Systems


1. Eberhart LH, Hogel J, Seeling W, et al. Evaluation of three risk scores to predict postoperative nausea and vomiting. Acta Anaesthesiol Scand 2000;44:480–8.
2. Gan TJ. Postoperative nausea and vomiting: can it be eliminated? JAMA 2002;287:1233–6.
3. Carroll NV, Miederhoff P, Cox FM, Hirsch JD. Postoperative nausea and vomiting after discharge from outpatient surgery centers. Anesth Analg 1995;80:903–9.
4. Tramer MR. Treatment of postoperative nausea and vomiting. BMJ 2003;327:762–3.
5. Apfel CC, Kranke P, Katz MH, et al. Volatile anaesthetics may be the main cause of early but not delayed postoperative vomiting: a randomized controlled trial of factorial design. Br J Anaesth 2002;88:659–68.
6. Scuderi PE, Conlay LA. Postoperative nausea and vomiting and outcome. Int Anesthesiol Clin 2003;41:165–74.
7. Gan TJ, Meyer T, Apfel CC, et al. Consensus guidelines for managing postoperative nausea and vomiting. Anesth Analg 2003;97:62–71.
8. Macario A, Weinger M, Carney S, Kim A. Which clinical anesthesia outcomes are important to avoid? The perspective of patients. Anesth Analg 1999;89:652–8.
9. Gan T, Sloan F, Dear Gde L, et al. How much are patients willing to pay to avoid postoperative nausea and vomiting? Anesth Analg 2001;92:393–400.
10. Habib AS, Gan TJ. Evidence-based management of postoperative nausea and vomiting: a review. Can J Anaesth 2004;51:326–41.
11. Apfel CC, Korttila K, Abdalla M, et al. A factorial trial of six interventions for the prevention of postoperative nausea and vomiting. N Engl J Med 2004;350:2441–51.
12. Habib AS, Gan TJ. Pharmacotherapy of postoperative nausea and vomiting. Expert Opin Pharmacother 2003;4:457–73.
13. Everett LL. Can the risk of postoperative nausea and vomiting be identified and lowered during the preoperative assessment? Int Anesthesiol Clin 2002;40:47–62.
14. Apfel CC, Roewer N, Korttila K. How to study postoperative nausea and vomiting. Acta Anaesthesiol Scand 2002;46:921–8.
15. Hornby PJ. Central neurocircuitry associated with emesis. Am J Med 2001;111 Suppl 8A:106S–12.
16. Watcha MF, White PF. Postoperative nausea and vomiting: its etiology, treatment, and prevention. Anesthesiology 1992;77:162–84.
17. Sinclair DR, Chung F, Mezei G. Can postoperative nausea and vomiting be predicted? Anesthesiology 1999;91:109–18.
18. Stadler M, Bardiau F, Seidel L, et al. Difference in risk factors for postoperative nausea and vomiting. Anesthesiology 2003;98:46–52.
19. Koivuranta M, Laara E, Snare L, Alahuhta S. A survey of postoperative nausea and vomiting. Anaesthesia 1997;52:443–9.
20. Raeder J. History of postoperative nausea and vomiting. Int Anesthesiol Clin 2003;41:1–12.
21. Palazzo M, Evans R. Logistic regression analysis of fixed patient factors for postoperative sickness: a model for risk assessment. Br J Anaesth 1993;70:135–40.
22. Cohen MM, Duncan PG, DeBoer DP, Tweed WA. The postoperative interview: assessing risk factors for nausea and vomiting. Anesth Analg 1994;78:7–16.
23. Lee J. An insight on the use of multiple logistic regression analysis to estimate association between risk factor and disease occurrence. Int J Epidemiol 1986;15:22–9.
24. Schechtman E. Odds ratio, relative risk, absolute risk reduction, and the number needed to treat-which of these should we use? Value Health 2002;5:430–5.
25. Lerman J. Surgical and patient factors involved in postoperative nausea and vomiting. Br J Anaesth 1992;69:24S–32S.
26. Van den Bosch JE, Moons KG, Bonsel GJ, Kalkman CJ. Does measurement of preoperative anxiety have added value for predicting postoperative nausea and vomiting? Anesth Analg 2005;100:1525–32.
27. Apfel CC, Greim CA, Haubitz I, et al. A risk score to predict the probability of postoperative vomiting in adults. Acta Anaesthesiol Scand 1998;42:495–501.
28. Apfel CC, Kranke P, Eberhart LH, et al. Comparison of predictive models for postoperative nausea and vomiting. Br J Anaesth 2002;88:234–40.
29. Apfel CC, Laara E, Koivuranta M, et al. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between two centers. Anesthesiology 1999;91:693–700.
30. Apfel CC, Greim CA, Haubitz I, et al. The discriminating power of a risk score for postoperative vomiting in adults undergoing various types of surgery. Acta Anaesthesiol Scand 1998;42:502–9.
31. Pierre S, Benais H, Pouymayou J. Apfel’s simplified score may favourably predict the risk of postoperative nausea and vomiting. Can J Anaesth 2002;49:237–42.
32. Toner CC, Broomhead CJ, Littlejohn IH, et al. Prediction of postoperative nausea and vomiting using a logistic regression model. Br J Anaesth 1996;76:347–51.
33. Fabling JM, Gan TJ, Guy J, et al. Postoperative nausea and vomiting: a retrospective analysis in patients undergoing elective craniotomy. J Neurosurg Anesthesiol 1997;9:308–12.
34. Junger A, Hartmann B, Benson M, et al. The use of an anesthesia information management system for prediction of antiemetic rescue treatment at the postanesthesia care unit. Anesth Analg 2001;92:1203–9.
35. Rowley MP, Brown TC. Postoperative vomiting in children. Anaesth Intensive Care 1982;10:309–13.
36. Eberhart LH, Geldner G, Kranke P, et al. The development and validation of a risk score to predict the probability of postoperative vomiting in pediatric patients. Anesth Analg 2004;99:1630–7.
37. Beattie WS, Lindblad T, Buckley DN, Forrest JB. Menstruation increases the risk of nausea and vomiting after laparoscopy: a prospective randomized study. Anesthesiology 1993;78:272–6.
38. Honkavaara P, Lehtinen AM, Hovorka J, Korttila K. Nausea and vomiting after gynaecological laparoscopy depends upon the phase of the menstrual cycle. Can J Anaesth 1991;38:876–9.
39. Gratz I, Allen E, Afshar M, et al. The effects of the menstrual cycle on the incidence of emesis and efficacy of ondansetron. Anesth Analg 1996;83:565–9.
40. Eberhart LH, Morin AM, Georgieff M. The menstruation cycle in the postoperative phase. Its effect of the incidence of nausea and vomiting [in German]. Anaesthesist 2000;49:532–5.
41. Wang SM, Kain ZN. Preoperative anxiety and postoperative nausea and vomiting in children: is there an association? Anesth Analg 2000;90:571–5.
42. Haigh CG, Kaplan LA, Durham JM, et al. Nausea and vomiting after gynaecological surgery: a meta-analysis of factors affecting their incidence. Br J Anaesth 1993;71:517–22.
43. Stern RM. The psychophysiology of nausea. Acta Biol Hung 2002;53:589–99.
44. Kranke P, Apefel CC, Papenfuss T, et al. An increased body mass index is no risk factor for postoperative nausea and vomiting. A systematic review and results of original data. Acta Anaesthesiol Scand 2001;45:160–6.
45. Gan TJ, Ginsberg B, Grant AP, Glass PS. Double-blind, randomized comparison of ondansetron and intraoperative propofol to prevent postoperative nausea and vomiting. Anesthesiology 1996;85:1036–42.
46. Rose JB, Watcha MF. Postoperative nausea and vomiting in paediatric patients. Br J Anaesth 1999;83:104–17.
47. Fabling JM, Gan TJ, El-Moalem HE, et al. A randomized, double-blinded comparison of ondansetron, droperidol, and placebo for prevention of postoperative nausea and vomiting after supratentorial craniotomy. Anesth Analg 2000;91:358–61.
48. Eberhart LH, Morin AM, Guber D, et al. Applicability of risk scores for postoperative nausea and vomiting in adults to paediatric patients. Br J Anaesth 2004;93:386–92.
49. Maharaj CH, Kallam SR, Malik A, et al. Preoperative intravenous fluid therapy decreases postoperative nausea and pain in high risk patients. Anesth Analg 2005;100:675–82.
50. Gan TJ, Soppitt A, Maroof M, et al. Goal-directed intraoperative fluid administration reduces length of hospital stay after major surgery. Anesthesiology 2002;97:820–6.
51. Moretti EW, Robertson KM, El-Moalem H, Gan TJ. Intraoperative colloid administration reduces postoperative nausea and vomiting and improves postoperative outcomes compared with crystalloid administration. Anesth Analg 2003;96:611–7.
52. Habib AS, White WD, Eubanks S, et al. A randomized comparison of a multimodal management strategy versus combination antiemetics for the prevention of postoperative nausea and vomiting. Anesth Analg 2004;99:77–81.
53. Tramer M, Moore A, McQuay H. Omitting nitrous oxide in general anaesthesia: meta-analysis of intraoperative awareness and postoperative emesis in randomized controlled trials. Br J Anaesth 1996;76:186–93.
54. Sneyd JR, Carr A, Byrom WD, Bilski AJ. A meta-analysis of nausea and vomiting following maintenance of anaesthesia with propofol or inhalational agents. Eur J Anaesthesiol 1998;15:433–45.
55. Tramer MR, Fuchs-Buder T. Omitting antagonism of neuromuscular block: effect on postoperative nausea and vomiting and risk of residual paralysis. A systematic review. Br J Anaesth 1999;82:379–86.
56. Sukhani R, Vazquez J, Pappas AL, et al. Recovery after propofol with and without intraoperative fentanyl in patients undergoing ambulatory gynecologic laparoscopy. Anesth Analg 1996;83:975–81.
57. Moiniche S, Romsing J, Dahl JB, Tramer MR. Nonsteroidal antiinflammatory drugs and the risk of operative site bleeding after tonsillectomy: a quantitative systematic review. Anesth Analg 2003;96:68–77.
58. Polati E, Verlato G, Finco G, et al. Ondansetron versus metoclopramide in the treatment of postoperative nausea and vomiting. Anesth Analg 1997;85:395–9.
59. Gan TJ, Joshi GP, Viscusi E, et al. Preoperative parenteral parecoxib and follow-up oral valdecoxib reduce length of stay and improve quality of patient recovery after laparoscopic cholecystectomy surgery. Anesth Analg 2004;98:1665–73.
60. Kehlet H. Postoperative opioid sparing to hasten recovery: what are the issues? Anesthesiology 2005;102:1083–5.
61. Marret E, Kurdi O, Zufferey P, Bonnet F. Effects of nonsteroidal antiinflammatory drugs on patient-controlled analgesia morphine side effects: meta-analysis of randomized controlled trials. Anesthesiology 2005;102:1249–60.
62. Redmond M, Glass PS. Opiate-induced nausea and vomiting: what is the challenge?. Anesth Analg 2005;101:1341–2.
63. Roberts GW, Bekker TB, Carlsen HH, et al. Postoperative nausea and vomiting are strongly influenced by postoperative opioid use in a dose-related manner. Anesth Analg 2005;101:1343–8.
64. Rama-Maceiras P, Ferreira TA, Molins N, et al. Less postoperative nausea and vomiting after propofol + remifentanil versus propofol + fentanyl anaesthesia during plastic surgery. Acta Anaesthesiol Scand 2005;49:305–11.
65. Dershwitz M, Michalowski P, Chang Y, et al. Postoperative nausea and vomiting after total intravenous anesthesia with propofol and remifentanil or alfentanil: how important is the opioid? J Clin Anesth 2002;14:275–8.
66. Borgeat A, Ekatodramis G, Schenker CA. Postoperative nausea and vomiting in regional anesthesia: a review. Anesthesiology 2003;98:530–47.
67. Joris JL, Poth NJ, Djamadar AM, et al. Supplemental oxygen does not reduce postoperative nausea and vomiting after thyroidectomy. Br J Anaesth 2003;91:857–61.
68. Purhonen S, Niskanen M, Wustefeld M, et al. Supplemental oxygen for prevention of nausea and vomiting after breast surgery. Br J Anaesth 2003;91:284–7.
69. Purhonen S, Turunen M, Ruohoaho UM, et al. Supplemental oxygen does not reduce the incidence of postoperative nausea and vomiting after ambulatory gynecologic laparoscopy. Anesth Analg 2003;96:91–6.
70. Goll V, Akca O, Greif R, et al. Ondansetron is no more effective than supplemental intraoperative oxygen for prevention of postoperative nausea and vomiting. Anesth Analg 2001;92:112–7.
71. Greif R, Laciny S, Rapf B, et al. Supplemental oxygen reduces the incidence of postoperative nausea and vomiting. Anesthesiology 1999;91:1246–52.
72. Tramer MR. A rational approach to the control of postoperative nausea and vomiting: evidence from systematic reviews. Part I. Efficacy and harm of antiemetic interventions, and methodological issues. Acta Anaesthesiol Scand 2001;45:4–13.
73. Tramer MR. A rational approach to the control of postoperative nausea and vomiting: evidence from systematic reviews. Part II. Recommendations for prevention and treatment, and research agenda. Acta Anaesthesiol Scand 2001;45:14–9.
74. Sweeney BP. Why does smoking protect against PONV? Br J Anaesth 2002;89:810–3.
75. Kaiser R, Sezer O, Papies A, et al. Patient-tailored antiemetic treatment with 5-hydroxytryptamine type 3 receptor antagonists according to cytochrome P-450 2D6 genotypes. J Clin Oncol 2002;20:2805–11.
76. Candiotti KA, Birnbach DJ, Lubarsky DA, et al. The impact of pharmacogenomics on postoperative nausea and vomiting: do CYP2D6 allele copy number and polymorphisms affect the success or failure of ondansetron prophylaxis? Anesthesiology 2005;102:543–9.
77. Brockmoller J, Kirchheiner J, Meisel C, Roots I. Pharmacogenetic diagnostics of cytochrome P450 polymorphisms in clinical drug development and in drug treatment. Pharmacogenomics 2000;1:125–51.
78. Apfel CC, Kranke P, Roewer N, et al. Bias in a further model for predicting PONV may not advance current knowledge. Anesthesiology 2000;92:1489–92.
79. Apfel CC, Roewer N. Risk assessment of postoperative nausea and vomiting. Int Anesthesiol Clin 2003;41:13–32.
80. Apfel CC, Kranke P, Eberhart LH. Comparison of surgical site and patient’s history with a simplified risk score for the prediction of postoperative nausea and vomiting. Anaesthesia 2004;59:1078–82.
81. van den Bosch JE, Kalkman CJ, Vergouwe Y, et al. Assessing the applicability of scoring systems for predicting postoperative nausea and vomiting. Anaesthesia 2005;60:323–31.
82. Biedler A, Wermelt J, Kunitz O, et al. A risk adapted approach reduces the overall institutional incidence of postoperative nausea and vomiting. Can J Anaesth 2004;51:13–9.
83. Pierre S, Corno G, Benais H, Apfel CC. A risk score-dependent antiemetic approach effectively reduces postoperative nausea and vomiting: a continuous quality improvement initiative. Can J Anaesth 2004;51:320–5.
84. Scuderi PE, James RL, Harris L, Mims GR 3rd. Multimodal antiemetic management prevents early postoperative vomiting after outpatient laparoscopy. Anesth Analg 2000;91:1408–14.
85. Watcha MF. The cost-effective management of postoperative nausea and vomiting. Anesthesiology 2000;92:931–3.
86. White PF, Watcha MF. Postoperative nausea and vomiting: prophylaxis versus treatment. Anesth Analg 1999;89:1337–9.
87. Habib AS, Gan TJ. Combination antiemetics: what is the evidence? Int Anesthesiol Clin 2003;41:119–44.
88. Visser K, Hassink EA, Bonsel GJ, et al. Randomized controlled trial of total intravenous anesthesia with propofol versus inhalation anesthesia with isoflurane-nitrous oxide: postoperative nausea with vomiting and economic analysis. Anesthesiology 2001;95:616–26.
89. Chimbira W, Sweeney BP. The effect of smoking on postoperative nausea and vomiting. Anaesthesia 2000;55:540–4.
90. Hough M, Sweeney B. The influence of smoking on postoperative nausea and vomiting. Anaesthesia 1998;53:932–3.
91. Riley TR 3rd, Chinchilli VM, Shoemaker M, Koch K. Is nausea associated with chronic hepatitis C infection? Am J Gastroenterol 2001;96:3356–60.
92. Chou FH, Lin LL, Cooney AT, et al. Psychosocial factors related to nausea, vomiting, and fatigue in early pregnancy. J Nurs Scholarsh 2003;35:119–25.
93. Haug TT, Mykletun A, Dahl AA. The prevalence of nausea in the community: psychological, social and somatic factors. Gen Hosp Psychiatry 2002;24:81–6.
94. Cohen MM, Cameron CB, Duncan PG. Pediatric anesthesia morbidity and mortality in the perioperative period. Anesth Analg 1990;70:160–7.
© 2006 International Anethesia Research Society