Data collection for the current study was sponsored by Merck Sharpe and Dohme. Dr. Grunberg serves as a consultant to Merck & Co. Dr. Rubenstein has served as a member of the speaker's bureaus at Merck & Co. and MGI Pharma and as a consultant to Merck & Co and MGI Pharma. He has received grant support from Merck & Co. and he owns stock in MGI Pharma. Dr. Deuson, Dr. Geling, and Dr. Mavros are employees of and may own stock and/or stock options in Merck & Co.
The authors determined the incidence of acute and delayed chemotherapy-induced nausea and emesis (vomiting) (CINV) among patients receiving highly (HEC) or moderately (MEC) emetogenic chemotherapy. They also assessed whether physicians and nurses accurately recognized the incidence of acute and delayed CINV in their own practices.
A prospective, observational study of adult patients receiving HEC or MEC for the first time was performed. Before patient enrollment, medical oncologists and oncology nurses estimated the incidence of acute (Day 1) and delayed (Days 2–5) CINV after first administration of HEC and MEC in their own practices. Eligible patients from their practices then completed a 6-day diary including emetic episodes, nausea assessment, and antiemetic medication use. Observed incidence rates of acute and delayed CINV were compared with physician/nurse predictions.
Twenty-four physicians and nurses and 298 eligible patients (67 receiving HEC and 231 receiving MEC) were recruited from 14 oncology practices in 6 countries. Greater than 35% of patients overall experienced acute nausea, whereas 13% experienced acute emesis. Delayed nausea and emesis were observed in 60% and 50% of HEC patients, respectively, and in 52% and 28% of MEC patients, respectively. Delayed symptoms appeared without acute symptoms after HEC (emesis, 38%; nausea, 33%) and MEC (emesis, 19%; nausea, 21%). Physicians and nurses accurately predicted the incidence of acute CINV but underestimated the incidence of delayed nausea and emesis after HEC by 21 and 28 percentage points, respectively, and delayed nausea after MEC by 28 percentage points. Greater than 75% of physicians and nurses underestimated the incidence of delayed CINV after both HEC and MEC.
Historically, chemotherapy-induced nausea and emesis (vomiting) (CINV) have held well deserved status as the toxicities of chemotherapy most feared by patients with cancer.1 However, antiemetic outcomes markedly changed after the introduction in 1981 of high-dose metoclopramide, which reduced the amount of emesis by approximately 90%, providing complete protection for a significant number of patients.2 The development of serotonin (5-HT3) antagonists in the early 1990s, potentiated by concomitant use of corticosteroids, further improved control of emesis without significant toxicity.3
Control of acute emesis (emesis within the first 24 hours after chemotherapy) allowed more accurate observation of the overall natural history of chemotherapy-induced emesis. Patients followed after hospital discharge reported delayed nausea and emesis that could continue for several days after chemotherapy at a frequency that could exceed that observed on the day of chemotherapy itself.4 The presence of delayed nausea and emesis even in the absence of acute nausea and emesis4 and the differential effect of standard antiemetics on delayed compared with acute emesis5 suggested that the pathophysiology of acute and delayed symptoms might differ.
Although control of emesis has improved as evidenced by objective measures such as a decrease in the number of emetic episodes, antiemetic control is not yet optimal. In addition to the need for new effective antiemetic treatments, further progress in antiemetic control will depend on an accurate understanding of the continuing frequency of acute and delayed nausea and emesis. Appreciation of this problem is made more difficult by the finding that delayed nausea and emesis often occurs after hospital or clinic discharge when the patient is not available for direct observation.
The current study was undertaken to determine the incidence of acute and delayed nausea and emesis after highly (HEC) and moderately emetogenic chemotherapy (MEC) when appropriate antiemetic therapy is given and to evalaute the accuracy with which physicians and nurses perceive the incidence of nausea and emesis within their own practice settings.
MATERIALS AND METHODS
Our prospective observational study was performed in 14 medical practices in Denmark, France, Italy, Germany, the U.K., and the U.S. during 2001–2002. To assess whether physicians and nurses accurately recognize the incidence of CINV in their own practices, a survey of their predictions of the incidence of CINV was taken. One medical oncologist and one oncology nurse from each of the participating medical practices were eligible to participate.
A convenience sample of patients was then recruited from the 14 participating medical practices. Eligible patients were required to be adults (> 18 years of age) receiving HEC or MEC regimens (emetic levels 3–5 of chemotherapy as defined by the Hesketh index6) for the first time. Patients were ineligible if they had experienced emesis within the 24 hours before receipt of chemotherapy or if they were scheduled to receive multiple-day chemotherapy during the first treatment cycle.
The protocol was approved by the institutional review boards or ethics committees according to requirements in each participating country. Written and signed informed consent was obtained from all patients.
Medical oncologists and oncology nurses from participating practices who agreed to participate were asked to complete a questionnaire regarding their perceptions of the incidence of nausea and emesis after first administration of chemotherapy, as currently administered in their practice despite the use of antiemetic therapy. Separate assessments of the incidence rates of nausea and emesis were requested for the acute (Day 1) and delayed (Days 2–5) phases, and for HEC (e.g., > 50 mg/m2 of cisplatin or > 1500 mg/m2 of cyclophosphamide) and MEC (e.g., > 20 mg/m2 of doxorubicin or < 1500 mg/m2 of cyclophosphamide). Physician and nurse questionnaires were completed before the initiation of patient recruitment at each participating practice.
Patients who agreed to participate received a 6-day diary covering the day of chemotherapy administration and the following 5 days. Patients were instructed to use the diary to report emetic episodes, nausea assessment, and antiemetic medications taken each day during the reporting period. They recorded each emetic episode and provided a daily nausea assessment using a 100-mm visual analog scale (VAS) to rate the amount of nausea experienced during the preceding 24 hours.
Finally, registry forms were developed to facilitate collection of data regarding patient demographic characteristics, cancer type and stage, chemotherapy regimen, antiemetic medications prescribed, and comorbid conditions.
No nausea was defined as a VAS < 5 mm on the 100-mm scale. A patient was considered to have had acute nausea or acute emesis if nausea (VAS ≥ 5 mm) or at least one episode of emesis, respectively, was reported during the first day after chemotherapy. Similarly, a patient was considered to have had delayed nausea or emesis if nausea (VAS ≥ 5 mm) or at least one episode of emesis, respectively, was reported on any day from Day 2 through Day 5 after chemotherapy.
Descriptive statistics were used to summarize patient demographics and survey responses. Mean estimated incidence rates of nausea and emesis with 95% confidence intervals (95% CI) (as predicted by physicians and nurses) were compared with mean observed incidence rates of nausea and emesis with 95% CI (based on patient diaries). Physicians' predictions of incidence rates were not weighted for patient accrual at the individual sites. The McNemar test was used to evaluate the association between acute and delayed nausea and emesis.
Fourteen practices participated in this international study, including 5 sites in the U.S. and 9 sites in Europe (Table 1). At 10 sites, the healthcare provider questionnaire was completed by one medical oncologist and by one oncology nurse before patient diaries were distributed. At three sites, the questionnaire was only completed by one physician and at one site the questionnaire was only completed by one oncology nurse. Therefore, 24 questionnaires were completed by healthcare professionals (13 medical oncologists and 11 oncology nurses).
Table 1. Patient and Healthcare Professional Accrual by Country
No. of practices
No. of medical oncologists
No. of oncology nurses
Patients treated with highly emetogenic chemotherapy
Patients treated with moderately emetogenic chemotherapy
Diaries were completed by 300 patients. Two patients who received only a chemotherapeutic agent of low emetic potential (Hesketh Level 2) were considered ineligible. Of the 298 eligible patients, 67 patients received HEC and 231 patients received MEC. Patients were classified as receiving HEC if they received at least one chemotherapeutic agent of Hesketh Level 5 emetic potential. Patients were classified as receiving MEC if they received at least one chemotherapeutic agent of Hesketh Level 3 or 4 emetic potential and no chemotherapeutic agent of Hesketh Level 5 emetic potential. Patients receiving HEC were surveyed at seven of the study sites in four countries. A median of seven patients (range, 1–26 patients) was contributed by each of these sites. Patients receiving MEC were surveyed at all 14 study sites with a median of 13.5 patients (range, 1–46 patients) contributed by each site.
The median age of surveyed patients was 56 years (range, 19–87 years; Table 2). Subjects enrolled in the study were predominantly female (72%) and received MEC (78%), consistent with the common use of such therapy in the treatment of breast carcinoma. Greater than 90% of the patients classified as receiving HEC were given cisplatin, whereas > 70% of the patients receiving MEC were given regimens containing cyclophosphamide, doxorubicin, and/or epirubicin. A majority of the patients (83%) received combination chemotherapy.
Table 2. Patient Characteristics
No. of patients
M: mean; SD: standard deviation.
Based on the 292 patients who provided responses to antiemetic treatment questions.
Age (M ± SD)
55.5 ± 12.1
Median age (range)
Other types of cancer
No. of chemotherapy agents used per patient (M ± SD)
No. of antiemetic agents used per patient (M ± SD)
2.3 ± 1.1
Patients receiving ≥2 antiemetic agents
No. of days supplied with antiemetic treatments (M ± SD)
3.5 ± 1.5
Patients with ≥2 days supply of antiemetic treatments
No. of observations
Antiemetic therapy consistent with current guidelines was used in the treatment of most of these patients. Nearly all the patients (97%) received a 5-HT3 antagonist, usually accompanied by a corticosteroid (78%). The most common duration of 5-HT3 antagonist therapy was 3 days and the most common duration of corticosteroid therapy was also 3 days, with 71% of patients receiving a 5-HT3 antagonist and 55% of patients receiving a corticosteroid for at least that long. Two or more antiemetic agents were administered to 80% of patients. Treatment generally was given during the risk periods for both acute and delayed nausea and emesis, with 86% of patients receiving some antiemetic therapy for ≥ 2 days.
Incidence of Nausea and Emesis
In spite of the administration of antiemetic therapy, > 35% of patients overall continued to experience acute nausea, whereas acute emesis was experienced by 13% of patients (Table 3). Delayed nausea and emesis were observed more commonly than acute nausea and emesis. The incidence of delayed nausea exceeded the incidence of acute nausea by 16 percentage points for MEC and by 27 percentage points for HEC, with > 50% of patients experiencing delayed nausea. The incidence of delayed emesis exceeded the incidence of acute emesis by 15 percentage points for MEC and by 38 percentage points for HEC. Therefore, delayed emesis was observed in 2.5 times as many patients as acute emesis. There were no significant differences (P > 0.05) in the odds of experiencing acute nausea and emesis or delayed nausea among patients treated with HEC compared with patients receiving MEC. However the odds of experiencing delayed emesis were 2.59 times higher for patients treated with HEC compared with MEC (odds ratio = 2.59; 95% CI, 1.43–4.67).
Table 3. Observed Incidence Rates of Nausea and Emesis and 95% Confidence Intervals
Observed incidence rates
95% CI: 95% confidence interval.
Highly emetogenic chemotherapy patients
Moderately emetogenic chemotherapy patients
Assessment of Healthcare Providers' Prediction of Nausea/Emesis
Medical oncologists' and oncology nurses' predictions of the incidence rates of acute nausea and acute emesis after administration of HEC were quite accurate (Table 4). The predicted incidence rates of acute nausea (34%) and acute emesis (17%) were well within the 95% CIs of the observed incidence rates. The majority of physicians (62% and 100%) and nurses (55% and 100%) made predictions that were above the lower limit of the 95% CI of observed incidence rates for acute nausea and emesis, respectively. However, the incidence of delayed nausea and emesis after HEC was markedly underestimated (Fig. 1). The predicted incidence of delayed nausea was 39% (95% CI, 30–48%) whereas the observed incidence was 60% (95% CI, 48– 72%) and the predicted incidence of delayed emesis was 22% (95% CI, 12–31%) whereas the observed incidence was 50% (95% CI, 37–63%). In each case, ≥ 76% of physicians and 80% of nurses predicted rates that were below the lower limit of the 95% CI of the observed incidence rates.
Table 4. Healthcare Providers' Predictions of Incidence Rates of Nausea and Emesis
Percentage of responses that are below the lower limit of the 95% confidence interval of the estimated incidence rate reported in Table 1.
Highly emetogenic chemotherapy patients
Moderately emetogenic chemotherapy patients
Although medical oncologists and oncology nurses accurately predicted the incidence of acute emesis among patients receiving MEC, the incidence of acute nausea was underestimated. Greater than 92% of physicians and 81% of nurses predicted incidence rates below the lower limit of the 95% CI of the observed incidence rates of acute nausea (Table 4). The incidence rates of delayed nausea and emesis after MEC were underestimated as well (Fig. 2). The incidence rate of delayed nausea was predicted to be 24% (95% CI, 15–34%) whereas the observed incidence rate was 52% (95% CI, 46–59%) and the incidence rate of delayed emesis was predicted to be 15% (95% CI, 6–24%) whereas the observed incidence rate was 28% (95% CI, 22–34%). Once again, ≥ 76% of physicians and 80% of nurses predicted incidence rates that were below the lower limit of the 95% CI of the observed incidence rates of delayed nausea and emesis.
Association of Acute and Delayed Symptoms
The association between acute and delayed nausea and emesis was examined to determine the predictability of delayed events based on acute events (Table 5). The null hypothesis of no association between acute and delayed symptoms was rejected for all subcategories (P < 0.001). Overall, the percentages of concordant observations (both acute and delayed or neither acute nor delayed) were 71% and 73% for nausea and emesis, respectively. The percentages of concordant pairs for nausea and emesis were higher for patients treated with MEC (74% and 76%, respectively) than for patients treated with HEC (60% and 62%, respectively). Overall, 24% and 23% of patients reported delayed nausea and emesis, respectively, in the absence of the condition in the acute phase. Similar patterns were observed among patients receiving HEC or MEC.
Table 5. Percent of Patients with Acute and Delayed Combinations of Emesis and Nausea
Ordered pairs of + and − indicate the presence and absence, respectively, of the relevant condition (e.g., nausea or emesis). The first element of an ordered pair refers to the acute phase and the second element to the delayed phase.
Based on the McNemar test of the null hypothesis of no association between acute and delayed manifestation of a condition.
Highly emetogenic chemotherapy patients
Moderately emetogenic chemotherapy patients
When initially evaluated by Coates et al. in 1983,1 nausea and emesis were the toxicities of chemotherapy most feared by patients with cancer. This finding was confirmed by Griffin et al.7 a decade later. However, in a more recent survey, Carelle et al.8 found that the social impact of chemotherapy had assumed a more prominent role and that the impact of nausea and emesis had decreased. This could be interpreted as a validation of the development of effective antiemetics and the resolution of this problem, thus allowing attention to be turned to other facets of quality of life management.
However, there is a growing concern that the presumed decrease in the impact of chemotherapy-induced nausea and emesis may actually reflect a shift from acute emesis that can be observed easily in the clinic to delayed emesis that may occur after discharge. In various surveys in the U.S.9 and Europe,5, 10 an incidence rate of 25–38% for delayed emesis and 55–60% for delayed nausea has been observed. Of greater note is the observation of an incidence rate of 18–30% for delayed emesis and 23–36% for delayed nausea, even in the absence of acute nausea and emesis. We have confirmed these findings, with 32% of our patients having delayed emesis and 54% delayed nausea overall, including 23% with delayed emesis and 24% with delayed nausea in the absence of acute nausea and emesis (Table 6).
Table 6. Incidence of Chemotherapy-Induced Delayed Nausea and Emesis Reported in the Literature
Both De Angelis et al.11 and Rubenstein et al.12 have found that antiemetic treatment guidelines are often not followed by physicians, particularly in the days after chemotherapy when patients are at home and not under the direct supervision of clinic personnel. One could question whether inadequate pharmacologic management could explain the lack of control of delayed nausea and emesis. However, in the current study, most patients received appropriate antiemetics for multiple days, so that a lack of preventive treatment cannot be invoked as an explanation for this continuing phenomenon.
The most striking finding of the current study was the underestimation of the current incidence of delayed nausea and emesis in their own practices by experienced physicians and nurses. This held true for both patients receiving MEC and patients receiving HEC in spite of the overall accuracy of estimates of incidence of acute nausea and emesis. Effective treatment for a problem cannot be administered unless caregivers realize that the problem exists. Newell et al.13 found a sensitivity of only 57% for identification of patients who had experienced nausea and emesis by physicians in an oncology clinic. Even a theoretic knowledge of the existence of a problem is not sufficient. Mertens et al.14 demonstrated that behavior modification of physicians as evidenced by improved antiemetic prescribing patterns did not occur in spite of several objective educational interventions. Prescribing patterns changed only after suboptimal antiemetic control in their own patients was specifically demonstrated to the physicians. This lack of appreciation of a continuing problem can also adversely affect the physician–patient relationship. Patients with hyperemesis gravidarum, for example, have indicated that one of the most important factors in determining their level of satisfaction with medical care is whether their physician appears to believe that their problem is real.15
Nausea and emesis, particularly delayed nausea and emesis, continue to be significant problems for patients with cancer receiving chemotherapy. New antiemetic agents, such as the NK-1 antagonists16 and longer-acting antiemetics,17 may ameliorate these problems and help patients maintain their functional status during chemotherapy. However, an increased appreciation of the incidence and duration of delayed nausea and emesis, both through physician/nurse education and through structured reporting by patients of their experience during chemotherapy, will be necessary to achieve optimal control of this problem using currently available tools and techniques.
The authors thank Dr. James Bearden, Dr. Andrew Kellum, Dr. Bruno Heinrich, Dr. Mark Hill, Dr. Joel Bloch, Dr. Michael Garcia, Dr. Kelly Pendergrass, and Dr. Phillipe Beuzeboc, all of whom served as coinvestigators, for their contributions to the current study.