Introduction

Chemotherapy-induced nausea and vomiting (CINV) is a common side effect experienced by cancer patients undergoing chemotherapy. CINV can be very unpleasant, and it can interfere with a patient’s ability to carry out normal everyday activities [5]. Acute CINV is defined as nausea and/or vomiting that occurs within the first 24 h after receiving chemotherapy [8]. Delayed CINV occurs more than 24 h after chemotherapy administration and may last for several days. Several risk factors are associated with occurrence of CINV. Notably, females and younger patients are at higher risk of having CINV [11]. Additionally, some chemotherapy agents are associated with increased risk of CINV. Highly emetogenic chemotherapy (HEC) drugs include cyclophosphamide (>1,500 mg/m2), cisplatin, and carmustine, and moderately emetogenic chemotherapy (MEC) drugs include doxorubicin, cyclophosphamide (<1,500 mg/m2), epirubicin, and oxaliplatin [6].

Without the use of prophylactic antiemetic therapy, chemotherapy (especially with highly emetogenic agents) could result in very severe nausea and vomiting, and in some cases, patients could elect to discontinue chemotherapy due to CINV. A number of antiemetic drugs for preventing CINV are available, and they are usually taken immediately before, and for several days after, chemotherapy. Prevention of CINV is the desired goal when using antiemetics, and they are frequently prescribed as prophylactics; however if prevention of CINV is not achieved, other antiemetics may be prescribed as rescue medication. Corticosteroids, 5-hydroxytryptamine-3 receptor antagonists (5-HT3 RAs), and neurokinin-1 receptor antagonists (NK-1 RAs) are commonly used antiemetic agents [8]. 5-HT3 RAs, which first became available in the early 1990s, target the serotonin receptor and are effective at controlling acute CINV but are generally not considered as effective for controlling delayed CINV [4]. The first NK-1 RA was approved more recently, in 2003, and appears to be effective at controlling both acute and delayed CINV [14]. Because 5-HT3 RAs and NK-1 RAs target different neurotransmitters, it is recommended that 5-HT3 RAs and NK-1 RAs be prescribed together, along with corticosteroids, for best control of CINV [6, 10].

Despite the availability of prophylactic antiemetic medication, significant numbers of patients may still experience CINV. CINV can exert a substantial burden on patients and on the health care system due to associated direct and indirect costs and due to the impact CINV can have on a patient’s quality of life [2]. Direct medical costs associated with managing CINV may include the cost of antiemetic medications, unscheduled doctor visits, emergency room (ER) visits, and hospitalizations, and indirect costs may include reduced productivity at work and missed time at work for both patients and caregivers. Additionally, CINV may impact a patient’s quality of life in many ways, from making it more difficult to enjoy meals to making it harder to socialize with friends.

In this study, we prospectively assessed the prevalence of CINV in a group of US patients undergoing a first course of chemotherapy in 2007–2008. Data regarding health care utilization, missed work, and impact on daily functioning was gathered from patients’ diaries and from case report forms completed by physicians. We used this information to determine the indirect and direct costs that resulted from CINV and the impact that CINV had on patients’ daily functioning. Because the patients in our study were treated in 2007–2008 (after the approval of the first NK-1 RA), the findings reported here should reflect a current picture of the burden of CINV in the USA.

Methods

Study design and patients

This was an observational study in which patients were recruited in 32 oncology specialty care settings in the USA. The sites were located in 19 different states and were found in all regions of the country (the West, Northeast, South, and Midwest). Study approval was obtained from the Copernicus Group Institutional Review Board on June 20, 2007; approval for the revised patient diary was received on August 28, 2007. Written informed consent was required from each patient who participated in the study. The study physicians and designated staff were responsible for recruiting patients, obtaining informed consent, and collecting baseline patient information. Patient recruitment and data collection occurred from September 19, 2007 to April 2, 2008.

The target population for this study was adults who were being treated by oncologists in an outpatient setting and receiving MEC or HEC. Patients needed to meet all of the following inclusion criteria to be eligible for enrollment into the study: a diagnosis of any kind of cancer, aged 18 or older, Eastern Cooperative Oncology Group Performance Status of 0 or 1, and initiating treatment with a Hesketh level 3–5 chemotherapy regimen. HEC chemotherapy regimens included cisplatin ≥ 50 mg/m2, cyclophosphamide ≥ 1,500 mg/m2, or carmustine > 250 mg/m2; MEC chemotherapy regimens included cyclophosphamide IV (500–1,500 mg/m2) and doxorubicin IV (≥40 mg/m2), cyclophosphamide IV (500-1500 mg/m2) and epirubicin IV (≥60 mg/m2), irinotecan ≥ 100 mg/m2, or oxaliplatin ≥ 85 mg/m2. Patients were excluded from the study for the following reasons: prior use of chemotherapy, concurrent use of radiation therapy, current participation in a trial with investigational chemotherapy drugs or antiemetic agents, treatment with a multiple day chemotherapy regimen (i.e., follow-up treatment is MEC or HEC and is administered within 6 days of the initial chemotherapy treatment), a history of drug or alcohol abuse, pregnancy, or having any uncontrolled medical condition that may represent another potential etiology for emesis and nausea, such as any ongoing gastrointestinal obstruction, increased intracranial pressure or hypercalcemia.

Data collection

The study physician or designated staff collected patient information on a case report form, including contact information, demographic information, cancer diagnosis, emesis history, comorbid medical conditions, chemotherapy treatment regimen, and the types and dosages of antiemetic prophylaxis and rescue medications prescribed. On the sixth day after the initiation of chemotherapy, the case report form was updated with information about any nausea- and vomiting-related health care contacts the patient experienced (e.g., calls to physician, office visits, ER visits, hospitalizations). In addition, any new prescriptions for antiemetic rescue medications and any planned alterations in treatment plan or in antiemetic prophylaxis in subsequent chemotherapy cycles were recorded.

Patients recorded information in a diary beginning with the day chemotherapy was initiated and for the following 5 days. Most items in the diary referred to the previous 24 h, and patients were asked to report in the diary on a daily basis. Patients recorded whether they experienced any nausea and/or vomiting, the severity of their nausea (rated as none, mild, moderate, or severe), the number of vomiting episodes they had, the impact of nausea and vomiting on their daily activities, whether they took any medication for nausea and vomiting, what antiemetic medications they used (both prescription and over-the-counter medications), and what, if any, health care they utilized for nausea and vomiting (e.g., office visit, ER visit, urgent care visit, hospital admission, new prescription). On the day of chemotherapy and the fifth day after chemotherapy, patients completed the Functional Living Index—Emesis (FLIE) questionnaire to assess the impact of nausea and vomiting on daily functioning. The 18-item FLIE questionnaire is a validated questionnaire in which a patient indicates on a numerical scale the extent that nausea or vomiting affected everyday activities, such as eating or socializing. The questionnaire contains nine items about nausea and nine items about vomiting. Each item is scored on a scale of 1–7. The total score is the sum of all items (minimum score of 18, maximum of 126). An average item score of more than 6 (or total score of greater than 108) is defined as “No or minimal impact on daily life” (NIDL).Because 11 out of the 178 patients did not complete the FLIE questionnaire, FLIE scores reported here are based on N = 167 patients. On the fifth day after chemotherapy, patients completed a modified version of the Work Productivity and Assessment Inventory–Nausea and Vomiting (WPAI-NV), which assesses how nausea and vomiting affected a patient’s ability to work and perform regular activities [12].

Determination of costs

The direct medical costs associated with CINV include the costs of medical care following chemotherapy, including the costs of antiemetic medications prescribed to patients; these costs were determined using information provided by the physician on the case report form. Resource utilization costs were determined for the 5-day period following initial treatment with chemotherapy and included physician office visits, emergency room visits, and inpatient hospitalizations. Estimated unit costs associated with these services were determined using the Centers for Medicare and Medicaid Services (CMS) Physician Fee Schedule for 2007. In addition, for determining the unit costing for ER visits, the CMS Hospital Outpatient Perspective Payment System (PPS) was consulted. Based on these resources, the following resource utilization costs were used: Family Physician Office Visit ($179.39), Oncologist Office Visit ($355.35), Emergency Room Visit ($574.85), Urgent Care Visit ($355.35), Inpatient Hospitalization Cost ($3,682.43 per day, times the cost-to-charge ratio of 0.5032). The unit costs of medications were gathered from the 2007 Red Book, which provides the average wholesale price (AWP) for a specific drug-form-strength combination. The cost for each drug a patient took for prophylaxis or rescue was calculated as (units per dose) × (doses per day) × (days supply) × (AWP).

Indirect costs were calculated using the current employment status and occupation of the patient as provided on the case report form. Days of work missed and lost productivity while at work was converted into monetary amounts utilizing average hours worked per week and average hourly earnings data from the United States Bureau of Labor Statistics, National Compensation Survey (NCS). Patients’ occupations were classified into one of nine occupational categories defined by the Bureau of Labor Statistics and used in the NCS. Mean hourly earnings by occupation category as reported from the 2005 (NCS) were adjusted to reflect 2007 dollars. Adjusted hourly earnings were multiplied by the patient-reported missed hours from work to calculate a total cost due to missed work. The WPAI-NV was used to measure the effect of nausea and vomiting on patients’ work productivity. Lost productivity as measured by the WPAI-NV was used in conjunction with the NCS mean weekly hours by occupation category to determine the reduction in number of hours worked. The reduced hours were applied to the adjusted hourly wage data to calculate costs due to lost productivity.

Results

Patients were recruited from 19 sites across the USA including nine sites in the South, five in the Northeast, three in the West, and two in the Midwest. The study was exploratory so no power calculation was conducted to determine the study sample size. The target sample size initially was 300 patients; however, during patient enrollment the target sample size was reduced due to slow recruitment. A total of 192 patients (136 receiving MEC and 56 receiving HEC) were enrolled into the study. Only seven patients did not return a patient diary, and an additional six patients returned a diary that was less than 90% complete; these 13 patients were dropped from analysis. One additional patient who did not return an informed consent form was also dropped from the study, yielding a final sample of 178 patients for analysis (Fig. 1). Patients were mostly female (67.98%), over the age of 50 (73.59%), white (85.39%), and married (67.42%; Table 1). Approximately one-third of study patients worked full time (32.02%), 10.67% worked part time, 16.82% were on medical leave, and 40.45% were unemployed (Table 1). The most frequent types of cancer were breast cancer (n = 80 patients), digestive/GI tract cancer (n = 40 patients), lung cancer (n = 19 patients), and lymphoma (n = 17 patients).

Fig. 1
figure 1

A summary of patient enrollment and diary completion for the study. ICF informed consent form

Full size image
Table 1 Patient demographic information
Full size table

Of the final 178 patients used for analysis, 125 received MEC and 53 received HEC. The HEC cohort had a higher proportion of male patients than the MEC cohort (52.83% vs. 23.20%), and the MEC cohort was slightly younger than the HEC cohort (Table 1). The MEC and HEC cohorts were similar regarding race/ethnicity, marital status, employment status, and residential status. In the HEC cohort, 44 patients received cisplatin chemotherapy, 8 received cyclophosphamide, and 1 received carmustine. In the MEC cohort, 75 patients received cyclophosphamide and doxorubicin chemotherapy, 32 received oxaliplatin, 15 received cyclophosphamide and epirubicin, and 3 received irinotecan.

Data on the number of patients who were prescribed antiemetic medications were obtained from case report forms, whereas data on patients reporting taking various antiemetic medications were obtained from patients’ diaries. Some of the most frequently prescribed antiemetic medications were the corticosteroid dexamethasone, the 5-HT3 RA palonosetron, and the NK-1 RA aprepitant (Table 2). In many cases, it appears that more patients were prescribed a medication than actually reported taking it. For example, 147 physicians reported prescribing dexamethasone, but only 140 patients reported using it. The same was found for other frequently prescribed medications, including palonosetron (111 physician reports vs. 101 patient reports) and aprepitant (66 physician reports vs. 63 patient reports). Interestingly, patients took dexamethasone less frequently for the delayed phase than for the acute phase; 72.5% of patients reported taking dexamethasone the day of treatment, but this number dropped to 19.1% of patients on the second day following chemotherapy, to 12.4% of patients on the third day following chemotherapy, and to 7.9% of patients on the fourth day following chemotherapy. It may be that the decline in dexamethasone use in the days following chemotherapy is due to patients’ previous experience of side effects (e.g., anxiety/mania/depression, stomach upset, increase in blood pressure, exacerbations of glucose control). Patients may also decide to wait to take the drug until they experience nausea rather than taking it as a prophylactic therapy.

Table 2 Physician- and patient-reported prophylactic and rescue antiemetic medications
Full size table

Health outcomes

Overall, a total of 109 patients (61.2%) reported experiencing either acute or delayed CINV (Table 3). Nausea, which affected 60.7% of patients, was more common than vomiting, which only affected 16.9% of patients. Additionally, delayed CINV, reported by 58.4% of patients, was more common than acute CINV, reported by 34.3% of patients (Table 3). On average, patients reported CINV 1.76 days following chemotherapy. Sixty-six percent of patients treated with HEC and 59.2% of patients treated with MEC reported CINV. Data collected day by day for the 5 days following chemotherapy show that patients more frequently experienced nausea alone than vomiting alone or both nausea and vomiting (Fig. 2). However, patients treated with HEC were more likely to report both nausea and vomiting than patients treated with MEC (Fig. 2).

Table 3 Prevalence of CINV, nausea, and vomiting
Full size table
Fig. 2
figure 2

Numbers of patients being treated with MEC or HEC who experienced nausea alone, vomiting alone, or both nausea and vomiting

Full size image

Costs

The direct medical costs associated with CINV were determined by calculating the resource utilization costs (for hospitalizations, ER visits, and doctor visits) and the costs of antiemetic medications. The average physician-reported cost for health resource utilization was roughly one-third that reported by patients ($31.94 per patient vs. $107.16 per patient; Table 4). There were some discrepancies between the amount of health care utilization reported by doctors and by patients. Overall, two emergency room visits, one urgent care visit, and one inpatient hospitalization that were reported by patients were not reported by physicians (Table 4). We also found that patients who reported severe nausea had higher average costs due to health care utilization ($802.40 per patient) than patients who reported moderate ($32.30 per patient) or mild nausea ($6.70 per patient). However, some of this difference was due to hospitalization costs for one patient with severe nausea that totaled slightly over $14,000. The average health care costs of patients who reported vomiting was $71.07 per patient. Medical costs were determined for antiemetic medications based on the amount of medication that doctors reported prescribing (Table 5). Of frequently prescribed prophylactic medications, generic dexamethasone was least expensive as compared to palonosetron and aprepitant, with an average prescription cost of $2.97 for dexamethasone, in comparison to an average cost of $390.00 for palonosetron and $333.04 for aprepitant. The average cost for a medication varied depending on whether it was prescribed for rescue or prophylaxis because of differences in size of the prescription. In total, the average direct medical costs for patients (based on prescribed medications and physician-reported health care utilization) were $732.14. On average, the total direct medical costs for patients receiving HEC were higher than for patients receiving MEC ($819.16 vs. $674.05, p = 0.11).

Table 4 Physician- and patient-reported health care utilization costs (calculated)
Full size table
Table 5 Cost of prescribed medications
Full size table

The indirect medical costs associated with CINV were determined by calculating the costs due to missed work and reduced productivity. On average, patients’ costs (for the entire patient population) were $31.57 due to missed work and $14.82 due to missed productivity. However, for the subset of patients who reported they are currently employed and who reported hours of missed work, average indirect costs were higher ($112.40 for missed work and $67.62 for reduced productivity). The severity of CINV also appeared to have an effect on indirect costs. Patients with severe vs. mild nausea had much greater indirect costs due to missed work ($379.13 vs. $8.19) and lost productivity ($456.41 vs. $85.83). Combining the total indirect and direct costs, we determined that CINV resulted in a total average cost of $778.53 per patient (Table 6). This reflects the average cost for a patient for the 5-day period following the first cycle of chemotherapy as well as the cost of antiemetic medications taken just before chemotherapy. The total costs for patients receiving HEC were slightly higher than those for patients receiving MEC, although the difference was not statistically significant ($905.31 vs. $724.78, p = 0.16).

Table 6 Overall direct and indirect costs (calculated)
Full size table

Quality of life

The impact of CINV on patients’ daily functioning was determined based on patient responses to the FLIE questionnaire. This is an 18-item questionnaire in which a patient indicates on a numerical scale the extent that nausea or vomiting affected everyday activities, such as eating or socializing. An average per item FLIE score of 6 or greater is defined as meaning that CINV had NIDL. Based on an average of FLIE score items, 37.2% of patients experienced an impact on daily life due to CINV (Fig. 3). Interestingly, patients receiving MEC and patients receiving HEC had very similar distributions of FLIE scores (62.7% of MEC patients had NIDL, 61.2% of HEC patients had NIDL). When patients’ average item FLIE scores from 5 days after chemotherapy were compared to baseline FLIE scores, we found an average decrease of 0.61 (from 6.59 before chemotherapy to 5.98 after). We also compared the average per item FLIE scores between patients who were complete responders (meaning a patient had well-managed CINV and did not experience vomiting or use rescue medications) and patients who did not have complete responder status. Of the complete responder patients, 90.1% had an average per item FLIE score indicative of NIDL, compared to only 36.0% of non-complete responder patients (Fig. 3). This suggests that patients with poorly controlled CINV are much more likely to experience a negative impact on daily functioning.

Fig. 3
figure 3

The average per item FLIE score was plotted against the percentage of patients who had that score or higher. a Average per item 5-day FLIE scores for patients receiving MEC or HEC. b Average per item 5-day FLIE scores for complete responder patients and non-complete responder patients. CR complete responder

Full size image

Discussion

We conducted a prospective study of CINV among cancer patients undergoing MEC or HEC. Importantly, all patients in this study were undergoing their first cycle of chemotherapy; most patients will typically undergo 4–6 cycles, and CINV is most common during the first cycle. Therefore, the results presented here are only relevant to the first cycle of chemotherapy. We determined the prevalence of CINV among patients, the indirect and direct costs associated with CINV, and the impact of CINV on patients’ daily functioning. In this study, 61.2% of the patients experienced some form of CINV (delayed CINV was more common than acute CINV), suggesting that CINV remains a significant problem among patients undergoing chemotherapy. Importantly, nausea was experienced much more frequently than vomiting among patients in this study, demonstrating that in the present day nausea is the main problem faced by patients with CINV. A decline in daily functioning due to CINV was reported by 37.2% of patients, and the average total cost of CINV was $778.53 per patient for the 5 days following the first cycle of chemotherapy. Notably, patients with more severe CINV had higher indirect costs than other patients due to missing more time from work and being less productive, and patients whose CINV was poorly controlled were significantly more likely to experience a decline in daily functioning due to CINV.

These findings demonstrate that despite recent advancements in antiemetic agents, many patients undergoing chemotherapy and receiving antiemetic therapy continue to experience CINV. It may be that physicians are not aggressive enough in prescribing prophylaxis; however, some patients may be noncompliant in taking their medications as prescribed and therefore do no reap the full benefit of prophylaxis treatment. We found that low numbers of physicians reported prescribing NK-1 RAs, suggesting that this class of drugs is underutilized and that physicians are not prescribing them in accordance with guidelines. Additionally, these results suggest that there may be a need for more intense patient education regarding compliance. Additional research is needed to more fully address these issues.

In this study, 34% and 58% of patients experienced acute and delayed CINV, respectively. Other recent prospective studies of CINV, where patients kept diaries or filled out questionnaires, reported similar findings. In a study of German patients, Ihbe-Heffinger and colleagues found that 60.7% of patients experienced delayed CINV and 32.8% experienced acute CINV [7]. In a study of patients at community oncology clinics in the USA, 59% of patients developed delayed CINV, and 36% developed acute CINV [3]. In a study of Canadian patients conducted by Lachaine and colleagues, 26% of patients reported acute CINV, and 44% reported delayed CINV [9]. Also, our results were consistent with those of Bloechl-Daum, who found that nausea was more common than vomiting among patients experiencing CINV [2].

Recent studies have also determined the indirect and direct costs associated with CINV in various countries, such as Germany, Canada, and Italy [1, 7, 9]. However, because of the marked differences in the health care system in the USA relative to that in other countries and because of the potential for salary differences in the USA compared to salaries in those countries, we feel it would be difficult to directly compare costs between these studies and ours. Tina Shih and colleagues performed a retrospective database study of indirect and direct costs due to CINV in US patients who were treated from 1997 to 2002 [13]. They found that the total costs of patients with uncontrolled CINV were on average $1,832 higher per month than patients with controlled CINV. Importantly, our cost calculations were based on the 5 days following initiation of the first cycle of chemotherapy, not an entire month; further, our study was performed more recently, after NK-1 RAs became available.

One limitation of this study is the relatively small sample size; a total of 178 patients were used in the final analysis. While a small sample still provides meaningful overall results, the generalizability of findings may be constrained when the sample is split into cohorts for comparison. Similar recently published prospective studies on CINV have also used small numbers of patients: Ballatori and colleagues used 172 patients for analysis and Ihbe-Heffinger and colleagues used 188 patients [1, 7].

Another potential drawback of this study is that dollar amounts associated with specific types of health care utilization came from external sources. Also, the antiemetic medication costs reported in this study are the average wholesale prices found in the 2007 Red Book, which may overestimate prices. Therefore, costs in this study were not actual costs but rather were calculated costs and estimated mean costs based on utilization. However, because patients in the study used different health plans, a standard measure of cost was used.

Finally, this study did not use the National Cancer Institute Common Toxicity Criteria to grade the severity of nausea or vomiting. Although it is common to grade nausea and vomiting events according to this scale in clinical trials, an average community doctor would not capture nausea or vomiting as an NCI grade. Further, chart reviews were not available in this study, and it was not possible to assign a grade on the basis of the patient or physician information obtained in this study. Grade III and IV nausea have a greater impact on patients and require more intervention, but any degree or grade of nausea and/or vomiting is distressing and negatively affects patients.

Conclusions

Although management of CINV has improved over time, both nausea and vomiting remain some of the most severe, prevalent, and worrisome chemotherapy-induced side effects. Acute CINV significantly reduces quality of life during treatment, and delayed CINV continues to be an issue after patients return home following their treatment, affecting their ability to perform daily activities. Furthermore, CINV negatively impacts daily activities, work attendance and productivity, and burdens an already stressed health care system. The need remains for more effective prophylactic antiemetics that could lead to better quality of life for patients receiving chemotherapy, as well as for increased utilization of existing drugs.