Changes in Antiemetic Overuse in Response to Choosing Wisely Recommendations

Key Points

Question  Did the American Society of Clinical Oncology’s antiemetic recommendations in the 2013 Choosing Wisely campaign reduce overuse of expensive antiemetics in patients receiving chemotherapy?

Findings  In an observational study applying multivariable logistic regression analysis to longitudinal health insurance claims data for 678 220 patients receiving chemotherapy before and after the October 29, 2013, announcement of the Choosing Wisely guidelines, the odds of antiemetic overuse decreased by 7% for the 6 months after the announcement. However, the odds of antiemetic overuse increased by 7% after 6 months compared with baseline; 24% of patients receiving chemotherapy overused costly antiemetics.

Meaning  The Choosing Wisely campaign had a short-term effect on oncologists.

Importance  Antiemetics are used to prevent chemotherapy-induced nausea and vomiting in patients with cancer. Newer antiemetic agents (serotonin and neurokinin-1 receptor antagonists) have increased efficacy but are expensive. The American Society of Clinical Oncology’s first guideline in the 2013 Choosing Wisely (CW) campaign discouraged overuse of expensive antiemetics in patients with low risk of chemotherapy-induced nausea and vomiting. However, little is known about patterns or trends in antiemetic overuse or whether any change has occurred with the publication of the CW recommendations.

Objective  To estimate the baseline prevalence, trends, determinants, and costs of antiemetic overuse from January 1, 2008, through March 31, 2015.

Design, Setting, and Participants  From January 1, 2008, through March 31, 2015, this observational study applied descriptive (univariate and bivariate) and multivariable logistic regression analyses to longitudinal health insurance enrollment and nationwide MarketScan insurance claims data for 678 220 privately insured patients receiving chemotherapy before and after the October 29, 2013, announcement of the CW guidelines. The baseline prevalence, trends, determinants, and costs of antiemetic overuse were estimated in cases stratified by risk for chemotherapy-induced nausea and vomiting.

Main Outcomes and Measures  Antiemetic use, overuse measure, and expenses before and after the publication of the CW recommendation, with adjustment for patient and health care professional characteristics.

Results  The sample included 678 220 adults who started chemotherapy during the observation period. The average age of the sample was 59.5 years, with 58.2% (n = 394 724) female. Antiemetic overuse occurred in 24.1% (n = 163 451) of patients, with highest rates among those receiving intravenous chemotherapy with high chemotherapy-induced nausea and vomiting risk (32.4% [n = 106 795]). Compared with baseline before the CW, patients had 7.0% lower odds of antiemetic overuse (95% CI, 4.4%-9.5%) during the 6 months after the CW, but this decrease was transitory: the odds of antiemetic overuse were 7.4% (95% CI, 4.6%-10.2%) higher than baseline at 6 months after the CW. Low-risk intravenous chemotherapy agents had overuse that continued to decrease 6 months after the CW. Antiemetic overuse was associated with higher costs. Reducing antiemetic overuse could have paid for 6.1% (95% CI, 5.8%-6.4%) of the chemotherapy drug costs.

Conclusions and Relevance  Antiemetic overuse is prevalent and results in unnecessary spending associated with systemic chemotherapy treatment. Short-term decreases in antiemetic overuse were associated with the CW recommendation, but sustained decreases occurred in only one risk group.

Cancer is among the most costly chronic conditions in the United States, with an estimated expense of $144 billion in 2012.¹ Cancer-related expense in the United States is projected to increase 30% to 40% by 2020.² To promote cost reductions while improving the quality and value of cancer care, the American Society of Clinical Oncology (ASCO) Value in Cancer Care Task Force issued a top five list in 2012 and 2013 for the American Board of Internal Medicine Foundation’s Choosing Wisely (CW) campaign to identify instances of overuse of low-value cancer care.^3-6 For example, the 2012 recommendations warned against the use of computed tomography, positron emission tomography, or radionuclide bone scans for patients with breast or prostate cancer at low risk for metastasis. The Quality Oncology Practice Initiative of the ASCO has worked to implement these recommendations among community-based oncologists, whereas the Consumers Union is developing and posting consumer-oriented information about low-value care and specific CW recommendations.⁷

From the beginning, there was little empirical information on the prevalence of the low-value practices, data essential to assess reductions in overuse and cost savings that may result from implementing ASCO’s CW recommendations. With time, several studies have documented baseline adherence to the 2012 recommendations.^8-14 For example, Ramsey et al⁸ reported between 51% and 78% adherence to the five 2012 ASCO guidelines among 1 large insurer. However, only 1 study directly assessed changes in low-value care in the period after the CW announcement, and, to our knowledge, there is no evidence specific to the ASCO recommendations.¹⁵

On October 29, 2013, the ASCO published a second set of 5 CW recommendations, including the one stating, “Don’t give patients starting on a chemotherapy regimen that has a low or moderate risk of causing nausea and vomiting antiemetic drugs intended for use with a regimen that has a high risk of causing nausea and vomiting.”^4(p4362)16 This recommendation is for oncologists treating patients who have cancer, are receiving chemotherapy, and are at risk for nausea and vomiting (chemotherapy-induced nausea and vomiting [CINV]). Failure to prevent or control CINV may cause dehydration and reduced nutritional status, which can require costly medical intervention to prevent severe physiological consequences.¹⁷ Chemotherapy-induced nausea and vomiting may also impair patients’ quality of life and limit their willingness to continue with chemotherapy treatments.

There is a growing collection of antiemetic agents used for CINV prophylaxis. Older drugs still prescribed include prochlorperazine, metoclopramide hydrochloride, and corticosteroids. During the past 2 decades, several new classes of antiemetics have been developed, including the serotonin receptor antagonists (oral dolasetron [Anzemet], granisetron [Kytril and Sancuso], ondansetron [Zofran and Zuplenz], and intravenous palonosetron [Aloxi]) and the neurokinin-1 receptor antagonists (aprepitant [Emend], the prodrug fosaprepitant, and rolapitant [Varubi]). The newer agents affect specific neuroreceptors that cause the nausea response and, consequently, have been particularly effective in reducing CINV.

Chemotherapy agents vary in their CINV risk, which is commonly reported in clinical trials. Regimens with a CINV incidence of less than 10% are designated as low risk, whereas regimens with an incidence higher than 90% are considered as high risk. The National Comprehensive Cancer Network (NCCN) and ASCO have developed guidelines associated with CINV management.^18-23 As illustrated in eTable 1 in the Supplement, some chemotherapy regimens require a combination of antiemetic medications. For example, intravenous chemotherapy that has a high CINV risk requires neurokinin-1 receptor and serotonin receptor antagonists along with a corticosteroid. However, these new agents and their combinations are substantially more expensive than the older agents. The antiemetic drug costs for patients prescribed serotonin receptor antagonists can be as high as $2000 per cycle.²³ The neurokinin-1 receptor antagonists are similarly expensive. There is concern about using higher-cost antiemetic agents that are not necessary to manage the expected risk for CINV, which is the subject of the CW recommendation.⁴ We use data from a large claims database that include hundreds of large employer groups with active workers and Medicare-enrolled adults to examine patterns of antiemetic use compared with the 2015 NCCN and ASCO guidelines, the cost associated with overuse, and changes in antiemetic use associated with the CW.

We used the Truven Health Analytics MarketScan database and its Medicare supplement from January 1, 2008, through March 31, 2015. These databases include health insurance claims for 50 million insured lives for many large employers previously used in national population-based studies.²⁴ The Medicare supplement includes retirees with Medicare and private supplemental health insurance from their employer. The geographical distribution is approximately the same as the US population distribution, and spending in MarketScan is correlated with 21% of US spending.²⁵ We adjusted pricing for inflation to year 2015 US dollars using the Consumer Price Index. The data are deidentified and publicly available. Thus, the analysis does not comprise human participants’ research as defined by 45 CFR 46.102²⁶ and does not require institutional review board approval.

The study sample includes adults with at least 1 claim for oral or intravenous chemotherapy. We excluded patients who had not filled at least 1 claim with a cancer diagnosis from 12 months before or 12 months after the initial chemotherapy claim. Consistent with the CW recommendation, we concentrate on the initial chemotherapy cycle for the enrollee after a minimum 12-month observation period without chemotherapy claims.

The key outcome measures for this analysis are antiemetic use associated with the initial chemotherapy exposure, expense for antiemetics, and an indicator for whether the antiemetic regimen is consistent with guidelines. To develop these measures, we characterized the chemotherapy regimen based on the presence of therapeutic classes and selected National Drug Codes and the Healthcare Common Procedural Coding System on claims. Chemotherapy was also characterized based on administration route (oral vs intravenous), identified by the Healthcare Common Procedural Coding System or by drug information in the 2015 Red Book supplement to MarketScan. We identified all chemotherapy agents administered on the date of the first claim and assigned them to the CINV-risk category specified in the ASCO and NCCN guidelines: intravenous moderate to high risk, intravenous low risk, intravenous minimal risk, oral moderate to high risk, and oral minimal to low risk. When more than 1 agent was administered, we assigned the regimen based on the highest CINV-risk chemotherapy drug. Some chemotherapy agents are assigned to a CINV-risk category based on quantity per body surface area. Because we did not have this information, we used a conservative approach and assigned those chemotherapy drugs to the higher risk of the potential categories. In addition, some moderate-risk intravenous chemotherapy agents (eg, carboplatin, irinotecan, and methotrexate) have been highly emetogenic in some patients. Using a conservative approach, we labeled these agents as high risk for all patients. When the ASCO and NCCN differed, as with bortezomib, we assigned the chemotherapy drug to the higher risk specified.

We identified the relevant Healthcare Common Procedural Coding System or National Drug Code data for all antiemetics based on either medical or prescription drug claims completed between 2 weeks before chemotherapy through the day after chemotherapy initiation (16 days). We created person-level indicators for each antiemetic used and then examined agreement with the ASCO- and NCCN-recommended antiemetic regimens according to the highest-risk chemotherapy agent administered. The NCCN guidelines became less restrictive between 2012 and 2015, adding the use of some serotonin receptor antagonists to the recommendations for low CINV risks. Our principal measures of antiemetic overuse apply the 2015 guidelines to the entire period, but in a sensitivity analysis, we applied the 2012 guidelines to the observations from 2008 to 2012. The recommendations from the 2012 and 2015 guidelines and the operational definitions are in eTable 1 in the Supplement. The expense for antiemetics was calculated from total reimbursement (Medicare, private health insurance plan, or patient liability) for all antiemetic claims during the 16-day observation period.

A key explanatory variable in the analysis was the period when chemotherapy was started relative to the CW announcement. We defined a baseline (before the CW period) from January 1, 2008, through October 29, 2013, a 6-month transition period from October 30, 2013, through April 30, 2014, and a period after the CW from May 1, 2014, through April 1, 2015. Patient characteristics were obtained through information provided on the plan enrollment files (ie, age, sex, region, and whether the plan holder is paid hourly [vs a salary], unionized, or Medicare enrolled) or through claims (number of chronic conditions).²⁷ Plan and health care professional characteristics that may affect use and costs paid by patients included indicators for chemotherapy administered in a hospital (vs physician office), at an oncologist not in the health plan network, and the type of health plan (health maintenance organization, preferred provider organization, point of service, fee-for-service, or high-deductible plan).

All analyses were performed using Stata 14 (StataCorp LP). Univariate and bivariate analyses described sample means and proportions for the characteristics and outcomes, stratified by CINV-risk group. We used an unadjusted linear spline regression to describe trends in antiemetic overuse during the 3 periods. Finally, we estimated a logit regression model to assess the association between antiemetic overuse and the CW, adjusting for CINV-risk group and other patient, oncologist, and plan characteristics. The model included fixed effects for the 135 antineoplastic agents and 36 cancer sites. We used the Tukey-Pregibon link test to select the model specification (logit rather than probit) and to test the assumptions that antiemetic overuse is a linear combination of our predictors, and that no relevant predictors were excluded.²⁸ To test the goodness of fit, we performed a receiver operating characteristic analysis. The ratio of sensitivity to 1 minus specificity was 0.77, which satisfies the recommended threshold of 0.7, indicating that the model has good ability to discriminate antiemetic overusers from nonoverusers.²⁹ Logs of all tests can be found in the eAppendix in the Supplement.

The sample included 678 220 adults who started chemotherapy during the observation period. Sample means and proportions are reported in eTable 2 in the Supplement. Intravenous chemotherapy with high to moderate CINV risk was the most common category (48.6% [n = 329 615]), followed by 20.0% (n = 135 644) with low- to minimal-risk oral therapy. The antiemetic overuse rate was 24.1% (n = 163 451), but the rates varied by CINV-risk group, as did the source of antiemetic overuse (Figure 1). For example, we observed antiemetic overuse in almost one-third (32.4% [n = 106 795]) of individuals receiving high- to moderate-risk intravenous chemotherapy, attributed to prescriptions for medications commonly prescribed to manage breakthrough symptoms several days after chemotherapy. One-fourth of individuals receiving minimal-risk intravenous chemotherapy used antiemetic regimens associated with higher-risk therapy, including 15.2% (n = 14 845) who received serotonin receptor antagonists and 16.7% (n = 16 603) corticosteroids. When we applied the more restrictive 2012 guidelines to observations in the baseline period, we identified substantially higher rates of antiemetic overuse (30.1% [n = 204 432]), particularly in the low-risk intravenous chemotherapy group where use of serotonin receptor antagonists was already prevalent (55.6% [n = 55 710]).

During the period before the announcement of the CW recommendation, antiemetic overuse rates were stable or decreasing slightly for all risk groups, with the exception of the oral high- to moderate-risk group (P < .05; 95% CI, 0.0002-0.0029 for growth rate per month) (Figure 2). In the 6 months after the CW announcement, antiemetic overuse decreased for the high- to moderate- and minimal-risk intravenous groups (P < .05; 95% CI, −0.0169 to −0.0098 and −0.002 to −0.0016, respectively). In the period after the CW, the low-risk intravenous group had a renewed decrease (95% CI, −0.0118 to −0.0016), whereas rates increased in the other 2 intravenous chemotherapy groups (95% CI, 0.014-0.020 and 0.012-0.029, respectively) and in the low- to minimum-risk oral group (95% CI, 0.0010-0.0052). Overall, compared with baseline before the CW, patients during the 6 months after the CW had 7.0% lower odds of overuse (95% CI, 4.4%, 9.5%), but this decrease was transitory: the odds of overuse were 7.4% (95% CI, 4.6%, 10.2%) higher than baseline after the first 6 months after the CW (eTable 3 in the Supplement).

In addition to temporal changes during the CW publication and CINV risk associated with the chemotherapy regimen, regression analysis identified other patient, health care professional, or health plan characteristics that were associated with antiemetic overuse (eTable 3 in the Supplement). Being enrolled in Medicare (odds ratio, 0.90; 95% CI, 0.87-0.92) and in a point of service plan (odds ratio, 0.91; 95% CI, 0.88-0.94) and receiving out of the health plan network (odds ratio, 0.89; 95% CI, 0.87-0.91) were associated with reduced odds of antiemetic overuse, whereas having a high-deductible plan and being an hourly or union worker were associated with increased odds. The probability of antiemetic overuse varied by region. Patients 70 years and older or with multiple chronic conditions were less likely to overuse antiemetics, but females were more likely to overuse.

The mean expense for antiemetic therapy was $299 (95% CI, $297-$301), with the highest spending for intravenous chemotherapy with high to moderate CINV risk at $523 (95% CI, $520-$527). Drug class-specific estimates confirm that the newer neurokinin-1 receptor and serotonin receptor antagonists were the most costly, with spending of $998 and $556 per regimen in year 2015 US dollars, respectively.

The mean expense for an antiemetic regimen with overuse was significantly higher than the expense of a guideline-consistent regimen for all CINV-risk groups (Table). The largest difference in spending was for the intravenous chemotherapy low CINV-risk agents: spending on guideline-consistent therapy was $77 (95% CI, $71-$84) whereas spending in the presence of antiemetic overuse was $529 (95% CI, $517-$541), a difference of $452 or 587.0% (95% CI, 585.8%-588.2%).

We examined the patterns of antiemetic overuse and changes before and after the publication of the 2013 CW recommendation to avoid overuse of antiemetics for CINV prophylaxis. Our findings reveal a substantial amount of antiemetic overuse during the period before the CW recommendation, although the rates varied by CINV-risk group. Antiemetic overuse decreased among intravenous chemotherapy agents during the 6-month period immediately after the 2013 CW recommendation, but not for oral agents. Furthermore, this effect was short lived. Six months after the recommendation, antiemetic overuse started to increase again for 4 of the 5 CINV-risk categories.

The source of antiemetic overuse varied by CINV-risk group. The most problematic cause from the CW perspective was associated with minimal-risk intravenous therapy, where no prophylaxis is recommended. For this group, we observed high rates of corticosteroids and the addition of serotonin receptor antagonists. The latter is not surprising because there are multiple agents in the class, including 2 that lost patent protection in 2008.^30,31 For high- to moderate-risk intravenous chemotherapy, there were no antiemetics that should be reserved for higher CINV-risk agents. Instead, we observed a large proportion of patients receiving drugs generally reserved for the management of breakthrough CINV. Because our measure included insurance claims for antiemetics that were provided on day 2 or earlier, these prescriptions must have been written by physicians and filled by patients in anticipation of possible breakthrough CINV. We cannot ascertain from the claims data whether the patients needed or took the drugs, but because the drugs generated potential excess cost to the system, we included them in our measure.

Our results indicate that antiemetic overuse can be costly. For the 100 132 patients who received intravenous low CINV-risk chemotherapy, 21 404 overused antiemetics with incremental spending of $9.7 million during their first chemotherapy session. A total of $159 million was spent on chemotherapy for these patients. Thus, the excess expense associated with antiemetic overuse represented 6.1% (95% CI, 5.8%-6.4%) of spending on chemotherapy. Hence, full implementation of the 2013 CW antiemetic recommendation would have generated substantial reductions in spending associated with chemotherapy.

Our study highlights the need to examine the factors motivating antiemetic overuse. We found evidence that health insurance type (Medicare vs non-Medicare) and health plan design are associated with the probability of antiemetic overuse, which suggests that insurer practices regarding utilization management, formulary inclusion, and cost sharing may be factors. For example, receiving chemotherapy through an oncologist not in the health plan network may be associated with higher cost sharing, and hence, patients may respond by reducing use perceived to be discretionary. Conversely, hourly and union workers had higher rates of antiemetic overuse, which may reflect more generous health insurance coverage.³² Lower rates of antiemetic overuse for Medicare-enrolled older adults, and with comorbidities, may be due to a variety of factors. Chemotherapy doses may be reduced to prevent other adverse effects, and CINV risk generally decreases with age. Individuals with a private health insurance supplement to Medicare also adhere to a different set of initial benefit structures, with antiemetics administered within 2 days of chemotherapy insured by the Medicare Part B rather than the prescription drug benefit. Thus, health insurance may generate some fundamental differences in antiemetic use rates and spending for those patients with Medicare as a primary health care coverage source. Our results do not indicate that oncologist incentives affect antiemetic overuse because rates did not differ between patients treated in the hospital outpatient department compared with a physician’s office. However, the large geographical variation suggests discretionary differences in treatment patterns not likely associated with patient need or outcomes.

Despite its strengths, the study has limitations regarding measurement and scope. For chemotherapy regimens in which CINV risk is dose dependent, we were unable to measure dose using insurance claims. In response, we assumed the higher CINV-risk category. As a result, fewer cases of antiemetic overuse might have been identified, resulting in conservative estimates of prevalence and spending. Our analysis was limited to prophylactic antiemetic use for the first cycle of chemotherapy, consistent with the subject of the CW recommendation. If the initial patterns persist throughout chemotherapy, our estimates of spending associated with antiemetic coverage are low. Our data do not provide information concerning cancer type, stage, or diagnosis date, so we did not assess the appropriateness of chemotherapy, only whether antiemetic regimens were consistent with guidelines according to the observed chemotherapy regimen. In addition, whereas our study did not assess for adverse effects associated with antiemetic overuse, it is an additional dimension to consider in future research.^28,29

Our results associated with this CW recommendation create the concern that voluntary measures to reduce antiemetic overuse may have limited effectiveness. Such measures may need to be included with financial performance incentives and are likely to be implemented by oncologists participating in the Medicare Shared Savings Programs, including the Oncology Care Model demonstration. The Medicare program has other initiatives that comprise this process. By 2019, all physicians participating in Medicare will be paid according to the pay-for-performance program known as the Merit-Based Incentive Payment System. The 2015 Centers for Medicare & Medicaid Services plan for the Merit-Based Incentive Payment System requested CW overuse measures to be incorporated.³³ However, because antiemetic overuse is a concept often misunderstood by consumers, patient engagement and education will also be critical to the successful implementation of these incentive-based initiatives.

Corresponding Author: William Encinosa, PhD, Center for Delivery, Organization, and Markets, Agency for Healthcare Research and Quality, 5600 Fisher’s Ln, Rockville, MD 20857 (william.encinosa@ahrq.hhs.gov).

Accepted for Publication: May 23, 2016.

Published Online: September 15, 2016. doi:10.1001/jamaoncol.2016.2530

Author Contributions: Drs Encinosa and Davidoff had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: All authors.

Obtained funding: Encinosa.

Administrative, technical, or material support: Encinosa.

Study supervision: Encinosa.

Conflict of Interest Disclosures: None reported.

Funding/Support: This research was supported by the Agency for Healthcare Research and Quality.

Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The views herein do not represent the views of the Agency for Healthcare Research and Quality or the US Department of Health and Human Services.