Abstract
Background
Sedative-hypnotic medications are used to treat chemotherapy-related nausea, anxiety, and insomnia. However, prolonged sedative-hypnotic use can lead to dependence, misuse, and increased health-care use. We aimed to estimate the rates at which patients who receive adjuvant chemotherapy for breast cancer become new persistent users of sedative-hypnotic medications, specifically benzodiazepines and nonbenzodiazepine sedative-hypnotics (Z-drugs).
Methods
Using the MarketScan health-care claims database, we identified sedative-hypnotic–naïve patients who received adjuvant chemotherapy for breast cancer. Patients who filled 1 and more prescriptions during chemotherapy and 2 and more prescriptions up to 1 year after chemotherapy were classified as new persistent users. Univariate and multivariable logistic regression analyses were used to estimate odds of new persistent use and associated characteristics.
Results
We identified 22 039 benzodiazepine-naïve patients and 23 816 Z-drug–naïve patients who received adjuvant chemotherapy from 2008 to 2017. Among benzodiazepine-naïve patients, 6159 (27.9%) filled 1 and more benzodiazepine prescriptions during chemotherapy, and 963 of those (15.6%) went on to become new persistent users. Among Z-drug–naïve patients, 1769 (7.4%) filled 1 and more prescriptions during chemotherapy, and 483 (27.3%) became new persistent users. In both groups, shorter durations of chemotherapy and receipt of opioid prescriptions were associated with new persistent use. Medicaid insurance was associated with new persistent benzodiazepine use (odds ratio = 1.88, 95% confidence interval = 1.43 to 2.47) compared with commercial or Medicare insurance.
Conclusions
Patients who receive sedative-hypnotic medications during adjuvant chemotherapy for breast cancer are at risk of becoming new persistent users of these medications after chemotherapy. Providers should ensure appropriate sedative-hypnotic use through tapering dosages and encouraging nonpharmacologic strategies when appropriate.
Although the opioid epidemic in the United States is well documented (1), there is growing recognition of the prevalence of sedative-hypnotic misuse (2,3). Annually, 4.8 million people (1.8% of the population) report misuse of benzodiazepine sedative-hypnotics (4), and an additional 1.6 million (0.5%) are estimated to misuse nonbenzodiazepine sedative-hypnotics (Z-drugs) such as zolpidem (5). Rates of misuse may be even higher in patients with cancer, a population in which prolonged use of sedative-hypnotics is prevalent (6,7), which places them at risk for increased health-care use and overdose mortality (8,9).
A common setting in which patients with cancer receive an initial sedative-hypnotic prescription is during chemotherapy. Sedative-hypnotics are used for the management of chemotherapy-associated nausea and vomiting (10) as well as other symptoms common in these patients, such as such as anxiety and insomnia (11,12). However, studies have not evaluated the rates at which patients become long-term users of sedative-hypnotics after the completion of chemotherapy. Given that these symptoms are often time limited and episodic (13), defining the rates of long-term sedative-hypnotic use in this vulnerable population is important because this may represent a preventable complication of the treatment of early-stage breast cancer.
The majority of the literature regarding new persistent use of controlled substances in patients who receive cancer-related interventions is focused on opioid use in postoperative patients (14-16). In this study, we sought to define the rates of new persistent sedative-hypnotic use after adjuvant chemotherapy for breast cancer. We focused on 2 classes of sedative-hypnotics, benzodiazepines and Z-drugs, and aimed to determine factors associated with the development of new persistent use of these agents.
Methods
Database
We used data from the IBM MarketScan Research Database (17). This database consists of medical and pharmaceutical claims from 350 payers in all 50 states. It includes claims from more than 30 million privately insured patients nationwide, including Medicare beneficiaries in commercial managed care plans, and 6 million Medicaid-insured patients from 12 states. Data are deidentified, and multiple steps are taken to ensure data validity and quality control. We followed the Strengthening Reporting of Observational Studies in Epidemiology reporting guidelines for observational studies (18).
Cohort
Patients aged 18 years or older who received adjuvant chemotherapy for breast cancer from 2008 to 2017 were included. International Classification of Diseases (ICD)-9, ICD-10, and/or Current Procedural Terminology codes were used to identify the following: 1) a diagnosis of breast cancer, 2) breast cancer–related surgery (lumpectomy or mastectomy), and 3) adjuvant chemotherapy. As previously described, a combination of these types of codes is highly effective in identifying patients with incident breast cancer using health-care claims data (19). Patients with code dates that suggested neoadjuvant chemotherapy treatment were excluded. To ensure selection of patients receiving adjuvant chemotherapy, we included only patients with chemotherapy codes entered 30-180 days after their first breast cancer–related procedure. We included patients who had continuous health insurance and drug benefit coverage from 12 months before their first procedure to 12 months after completion of chemotherapy to ensure complete capture of all commercial, Medicaid, and Medicare claims. Within this cohort, we defined 3 periods: 365 to 31 days before first breast cancer–related procedure (period 1), 31 days before the start of chemotherapy to 90 days after the completion of chemotherapy (period 2), and 90 to 365 days after chemotherapy (period 3).
Sedative-hypnotic medications of interest were identified from LexicompOnline. Prescriptions for the medications were identified during the 3 study periods, cross-referencing prescriptions against a preestablished list of generic drug names (Supplementary Table, available online). Patients who filled 1 or more prescriptions for a given agent during period 1 were considered previous sedative-hypnotic users and were excluded from the analysis. Remaining patients who then filled at least 1 prescription for a given agent during period 2 and at least 2 prescriptions for the same class of agent during period 3 were considered new persistent users of that agent. This definition of new persistent substance use is more conservative than used in previous studies (14,16) in that we require at least 2 prescriptions during period 3 (rather than at least 1 prescription) and we follow-up patients for up to 1 year after chemotherapy.
Additional details about sedative-hypnotic prescriptions received, such as the number of prescriptions per patient in each study period and total doses prescribed, were evaluated. We calculated the unit milligram equivalent of each sedative-hypnotic component and converted that amount to oral lorazepam equivalents using published conversion factors (20). The oral lorazepam equivalent dosage for each individual sedative-hypnotic prescription was calculated as the unit oral lorazepam equivalent exposure multiplied by the total number of pills filled in the prescription.
Patient-level factors of interest were captured from MarketScan, via ICD-9, ICD-10, and Current Procedural Terminology codes. Demographic data included age in years at surgery, sex, and insurance status (Medicaid vs commercial or Medicare insurance). Other variables of interest included type of surgery, radiation treatment, duration of chemotherapy, year of chemotherapy, and receipt of opioid prescriptions before or during chemotherapy.
Statistical Analysis
Demographic and clinical characteristics were presented descriptively, and the rate of new persistent substance use was estimated for each covariate among the substance-naïve cohorts. The unadjusted associations between risk factors and new persistent opioid use were compared using χ2 tests. Mean total lorazepam equivalent doses received were compared between groups of patients using t tests. Pooled t tests were used when group variances were equal, and Satterthwaite t tests were used when group variances were unequal. Equality of variances was determined using a folded f test. Multivariable logistic regression models were used to evaluate the association between clinically relevant risk factors and the development of new persistent substance use. All hypothesis tests were 2-sided, and statistical significance was set at P less than .05. All analyses were conducted using SAS Studio Version 3.71 (SAS Institute Inc, Cary, NC, USA).
Results
We identified 93 018 patients who received adjuvant chemotherapy for breast cancer from 2008 to 2017 (Figure 1). We generated 2 study cohorts: 22 039 benzodiazepine-naïve patients and 23 816 Z-drug–naïve patients (Table 1). The median age was 56 years for both cohorts (interquartile range = 49-62). In both the benzodiazepine-naïve and Z-drug–naïve cohorts, 99.0% of patients were female and 1.0% were male. The majority of patients had commercial or Medicare insurance (93.0% benzodiazepine-naïve, 92.6% Z-drug–naïve), received a lumpectomy (80.4% benzodiazepine-naïve, 80.1% Z-drug–naïve), and received radiation therapy (69.6% benzodiazepine-naïve, 69.4% Z-drug–naïve). Approximately one-half of patients received less than 4 months of chemotherapy (49.9% benzodiazepine-naïve, 49.9% Z-drug–naïve). More patients received opioid prescriptions before chemotherapy (58.1% benzodiazepine-naïve, 60.8% Z-drug–naïve) than during chemotherapy (26.2% benzodiazepine-naïve, 28.3% Z-drug–naïve), presumably as a result of the surgery (lumpectomy or mastectomy) received before initiation of adjuvant chemotherapy.
Figure 1.
Study cohort and sample criteria. Patients aged 18 years or older who received adjuvant chemotherapy for breast cancer between 2008 and 2017 were identified. Patients in this group who either did not have continuous insurance coverage from 1 year before surgery to 1 year after chemotherapy or did not have a full year of data before surgery or after chemotherapy were excluded. Additionally, we considered any patients who filled at least 1 prescription for a given sedative-hypnotic anytime from 365 to 31 days before their first breast cancer–related procedure (period 1) to be previous sedative-hypnotic users.
Table 1.
Baseline characteristics of sedative-hypnotic–naïvea patients
| Benzodiazepine-naïve | Z-drug–naïveb | |
|---|---|---|
| (n = 22 039) |
(n = 23 816) |
|
| Demographic and clinical characteristics | No. (%) | No. (%) |
| Age, y | ||
| ≤49 | 6077 (27.6) | 6509 (27.3) |
| 50-65 | 12 408 (56.3) | 13 437 (56.4) |
| >65 | 3554 (16.1) | 3870 (16.2) |
| Sex | ||
| Female | 21 819 (99.0) | 23 583 (99.0) |
| Male | 220 (1.0) | 233 (1.0) |
| Insurance | ||
| Medicaid | 1546 (7.0) | 1769 (7.4) |
| Commercial or Medicare | 20 493 (93.0) | 22 047 (92.6) |
| Surgery | ||
| Lumpectomy | 17 728 (80.4) | 19 081 (80.1) |
| Mastectomy | 4311 (19.6) | 4735 (19.9) |
| Chemotherapy duration | ||
| <4 mo | 11 001 (49.9) | 11 875 (49.9) |
| ≥4 mo | 11 038 (50.1) | 11 941 (50.1) |
| Year of chemotherapy | ||
| 2008 or 2009 | 2768 (12.6) | 2975 (12.5) |
| 2010 | 3066 (13.9) | 3307 (13.9) |
| 2011 | 3176 (14.4) | 3439 (14.4) |
| 2012 | 2863 (12.9) | 3105 (13.0) |
| 2013 | 2546 (11.6) | 2831 (11.9) |
| 2014 | 3200 (14.5) | 3460 (14.5) |
| 2015-2016 | 4420 (20.1) | 4699 (19.7) |
| Opioid prescription before chemotherapy | ||
| Yes | 12 814 (58.1) | 14 483 (60.8) |
| No | 9225 (41.9) | 9333 (39.2) |
| Opioid prescription during chemotherapy | ||
| Yes | 5776 (26.2) | 6745 (28.3) |
| No | 16 263 (73.8) | 17 071 (71.7) |
| Radiation | ||
| Yes | 15 336 (69.6) | 16 532 (69.4) |
| No | 6703 (30.4) | 7284 (30.6) |
Sedative-hypnotic–naïve patients are patients who do not regularly fill sedative-hypnotic prescriptions in the year before adjuvant chemotherapy.
Z-drugs are nonbenzodiazepine sedative-hypnotics, such as zolpidem.
Sedative-Hypnotics
The patients in our cohorts filled a total of 125 649 benzodiazepine prescriptions and 48 433 Z-drug prescriptions during the study period (eg, during adjuvant chemotherapy and up to 1 year after chemotherapy completion). Common benzodiazepine prescriptions included alprazolam (45 550 prescriptions, 36.3% of all benzodiazepine prescriptions), lorazepam (37 411, 29.8%), clonazepam (19 747, 15.7%), and diazepam (12 601, 10.0%). Common Z-drug prescriptions included zolpidem (43 398 prescriptions, 89.6% of all Z-drug prescriptions) and eszopiclone (4375, 9.0%).
Rates of New Persistent Use
In the cohort of 22 039 benzodiazepine-naïve patients, 6159 (27.9%) filled at least 1 benzodiazepine prescription during chemotherapy, and 963 of those patients (15.6%) went on to become new persistent benzodiazepine users (Figure 2). By contrast, among the remaining 15 880 (72.1%) who filled no benzodiazepine prescriptions during chemotherapy, 123 (0.8%) began to regularly fill benzodiazepine prescriptions after chemotherapy.
Figure 2.
Rates of new persistent sedative-hypnotic use. A) (Pie chart) Proportion of benzodiazepine-naïve patients who filled 1 or more prescriptions during chemotherapy (27.9%); (bar chart) proportion of these patients who went on to become new persistent benzodiazepine users after completion of chemotherapy (15.6%). B) (Pie chart) A smaller proportion of Z-drug–naïve patients filled a Z-drug prescription during chemotherapy (7.4%); (bar chart) however, a larger proportion of these patients went on to become new persistent Z-drug users after chemotherapy (27.3%).
Similarly, in the Z-drug–naïve cohort of 23 816 patients, 1769 (7.4%) filled a Z-drug prescription during chemotherapy, and 483 of those patients (27.3%) went on to become new persistent Z-drug users. Conversely, among the remaining 22 047 (92.6%) who filled no Z-drug prescriptions during chemotherapy, 125 (0.6%) began to regularly fill Z-drug prescriptions after chemotherapy.
Rates of New Persistent Sedative-Hypnotic Use by Number of Prescriptions Received
As the number of sedative-hypnotic prescriptions dispensed during chemotherapy increased, the rates of new persistent use increased as well (Figure 3). Among 4032 patients who filled 1-2 benzodiazepine prescriptions during chemotherapy, 223 (5.5%) became new persistent users. This increased to 217 of 1050 (20.7%) of patients who filled 3-4 prescriptions, 187 of 480 (40.0%) of patients who filled 5-6 prescriptions, and 336 of 597 (56.3%) of patients who filled 7 or more prescriptions. On average, new persistent benzodiazepine users received 6.8 prescriptions during chemotherapy and 4.6 prescriptions after chemotherapy.
Figure 3.
Rates of new persistent use by number of prescriptions received. As the number of benzodiazepine (A) and Z-drug (B) prescriptions dispensed during chemotherapy increased per patient, their risk of becoming a new persistent user of that substance increased linearly. For instance, of 4032 patients who received 1 or 2 benzodiazepine prescriptions during chemotherapy, 223 (5.5%) went on to become new persistent benzodiazepine users. This increased to 336 of 597 (56.3%) of patients who filled 7 or more benzodiazepine prescriptions during chemotherapy.
We observed a similar trend for Z-drugs. There were 137 of 1105 (12.4%) of patients who filled 1-2 prescriptions who went on to become Z-drug new persistent users. This increased to 134 of 327 (41.0%) of patients who filled 3-4 prescriptions, 83 of 146 (56.9%) of patients who filled 5-6 prescriptions, and 129 of 191 (67.5%) of patients who filled 7 or more prescriptions. New persistent Z-drug users received an average of 5.9 prescriptions during chemotherapy and 5.0 prescriptions after chemotherapy.
Sedative-Hypnotic Quantities Received
New persistent benzodiazepine users were prescribed statistically significantly higher mean total doses of benzodiazepines during chemotherapy than nonusers. New persistent users received a mean total dose of 4384 mg during chemotherapy relative to 847 mg received by patients who did not become new persistent benzodiazepine users (P < .001).
Similarly, new persistent Z-drug users were prescribed higher mean total doses during chemotherapy relative to those who did not become persistent Z-drug users (974 mg vs 328 mg, P < .001).
Risk Factors for New Persistent Use
Two variables were associated with both new persistent benzodiazepine and new persistent Z-drug use: shorter duration of chemotherapy and receipt of an opioid prescription during chemotherapy (Table 2). In terms of chemotherapy, patients who were treated with less than 4 months of chemotherapy became new persistent users at higher rates than those who received more than 4 months of chemotherapy, both for benzodiazepines (odds ratio [OR] = 1.30, 95% confidence interval [CI] = 1.12 to 1.51) and Z-drugs (OR = 1.47, 95% CI = 1.18 to 1.84). In terms of opioid prescriptions, those who received prescriptions during chemotherapy were more likely to become new persistent benzodiazepine users (OR = 1.94, 95% CI = 1.67 to 2.25) and new persistent Z-drug users (OR = 1.35, 95% CI = 1.07 to 1.69) than those who did not receive opioids during chemotherapy.
Table 2.
Association of patient characteristics and new persistent sedative-hypnotic usea
| New persistent benzodiazepine use |
New persistent Z-drug use |
|||
|---|---|---|---|---|
| Factor | Odds ratio (95% CI) | P | Odds ratio (95% CI) | P |
| Age, y | ||||
| 50-65 vs <49 | 1.14 (0.98 to 1.33) | .09 | 1.05 (0.83 to 1.32) | .69 |
| >65 vs <49 | 1.13 (0.88 to 1.45) | .34 | 0.70 (0.46 to 1.06) | .09 |
| Insurance | ||||
| Medicaid vs commercial/Medicare | 1.88 (1.43 to 2.47) | <.001 | 0.83 (0.50 to 1.38) | .47 |
| Surgery | ||||
| Mastectomy vs lumpectomy | 1.09 (0.83 to 1.43) | .53 | 1.12 (0.73 to 1.71) | .59 |
| Chemotherapy duration | ||||
| <4 mo vs >4 mo | 1.30 (1.12 to 1.51) | <.001 | 1.47 (1.18 to 1.84) | <.001 |
| Year of chemotherapy | ||||
| 2010 vs 2008/2009 | 1.26 (0.96 to 1.65) | .09 | 0.87 (0.61 to 1.25) | .45 |
| 2011 vs 2008/2009 | 1.12 (0.85 to 1.47) | .43 | 1.16 (0.81 to 1.65) | .42 |
| 2012 vs 2008/2009 | 1.01 (0.85 to 1.47) | .66 | 0.79 (0.53 to 1.16) | .23 |
| 2013 vs 2008/2009 | 1.19 (0.89 to 1.59) | .24 | 0.79 (0.53 to 1.16) | .13 |
| 2014 vs 2008/2009 | 1.04 (0.79 to 1.37) | .78 | 0.98 (0.67 to 1.42) | .90 |
| 2015/2016 vs 2008/2009 | 1.12 (0.86 to 1.46) | .39 | 0.77 (0.52 to 1.15) | .20 |
| Opioid prescription before chemotherapy | ||||
| Yes vs no | 0.66 (0.52 to 0.85) | <.001 | 1.05 (0.67 to 1.65) | .83 |
| Opioid prescription during chemotherapy | ||||
| Yes vs no | 1.94 (1.67 to 2.25) | <.001 | 1.35 (1.07 to 1.69) | .01 |
| Radiation | ||||
| Yes vs no | 1.08 (0.85 to 1.36) | .55 | 0.89 (0.61 to 1.30) | .55 |
aCI = confidence interval
Insurance status was associated with the development of new persistent benzodiazepine use. Specifically, Medicaid insurance was associated with new persistent benzodiazepine use relative to commercial and Medicare insurance (OR = 1.88, 95% CI = 1.43 to 2.47).
Finally, several variables were not associated with new persistent sedative-hypnotic use. Neither type of surgery received before chemotherapy (lumpectomy or mastectomy) nor whether patients received radiation treatment during the study period were associated with either class of new persistent use.
Overlapping Sedative-Hypnotic Use
Of the 963 new persistent benzodiazepine users, 276 (28.7%) filled a Z-drug prescription at some point as well (see Supplementary Figure 1, available online). Of these, 45 (16.3%) filled a Z-drug prescription in each study period (presumably representing long-term Z-drug users), and 112 (40.6%) filled a Z-drug prescription during chemotherapy only. For the other cohort, of 483 new persistent Z-drug users, 332 (68.7%) filled at least 1 benzodiazepine prescription during the study period. Of these, 56 (16.9%) filled benzodiazepine prescriptions throughout, and 159 (47.9%) filled a benzodiazepine prescription during chemotherapy only. Finally, there were 74 patients who became new persistent users of both benzodiazepine and Z-drug medications, representing 7.7% of the new persistent benzodiazepine users and 15.3% of the new persistent Z-drug users.
Sedative-Hypnotic Use Patterns in Nonnaïve Patients
There were 2787 patients who were not benzodiazepine-naïve before chemotherapy. Of these, 2232 (80.1%) filled prescriptions during chemotherapy, and 1330 (59.6%) continued to fill prescriptions in the year after chemotherapy. Patients who filled prescriptions in all periods (likely representing chronic users) received a mean total dose of 8080 mg during chemotherapy and 6533 mg after chemotherapy.
There were also 1022 patients who were not naïve to Z-drugs before chemotherapy. Among these, 761 (74.5%) filled prescriptions during chemotherapy, and 491 (64.5%) continued to fill prescriptions after chemotherapy. Chronic Z-drug users received a mean total dose of 1394 mg during chemotherapy and 3156 mg after chemotherapy.
Discussion
In this study, we found that a considerable number of patients who received adjuvant chemotherapy for breast cancer became new persistent users of sedative-hypnotics after treatment. More specifically, 27.9% of previously benzodiazepine-naïve patients filled at least 1 benzodiazepine prescription during chemotherapy, and 15.6% of these patients became new persistent users of benzodiazepines. Although a smaller percentage of Z-drug–naïve patients filled at least 1 Z-drug prescription during chemotherapy (7.4%), these patients went on to become new persistent Z-drug users at even higher rates (27.3%) than those in the benzodiazepine cohort. Although a proportion of this use may be indicated for psychiatric reasons, the concern is that some of this use may be nonindicated, preventable, and subject patients cured of their cancer to the myriad dangers of sedative-hypnotics.
Our study is novel in the characterization of new and long-term sedative-hypnotic use after adjuvant chemotherapy. Our findings are consistent with the limited literature on sedative-hypnotic use during chemotherapy, in which small studies report that 20%-30% of patients who receive chemotherapy for breast cancer fill a sedative-hypnotic prescription for indications such as insomnia (11,12). However, our study extends this literature to demonstrate that such use during chemotherapy can lead to prolonged use of these medications afterwards. Long-term sedative-hypnotic prescriptions have been discouraged by guidelines since the 1990s (21,22), and our study defines adjuvant chemotherapy as a strong risk factor for this type of potentially nonindicated use.
Whereas much attention has been devoted to postoperative opioid use in patients undergoing cancer-related surgery (14,16), less has been explicitly devoted to sedative-hypnotic use in the cancer population. In a retrospective study of women who underwent mastectomy and reconstructive surgery, we demonstrated that similar rates of substance-naïve women became new persistent users of opioids (17.2%) and sedative-hypnotics (17.1%) postoperatively (15). Of note, we found that receipt of chemotherapy was highly associated with new persistent sedative-hypnotic use (OR = 2.17, 95% CI = 1.94 to 2.42). This study builds on these findings to focus directly on the risk of new persistent sedative-hypnotic use after adjuvant chemotherapy.
Mental health symptoms are common in patients with cancer (23–28), and consequently, the use of psychotropic medications in this population is widespread (29). One-half of patients with early breast cancer have anxiety, depression, or both in the year after diagnosis (24). Insomnia is present in 42%-69% of patients up to 18 months after curative surgery for breast cancer, and breast cancer is associated with an odds ratio of 3.17 for a diagnosis of insomnia relative to other cancers (25–28). Psychotropic use appears to be particularly prevalent in the United States, where 43.8% of patients with early-stage breast cancer receive psychotropics relative to 18.2% of similar patients in Italy (30). Although medications such as sedative-hypnotics can be highly effective in the short term, the need to use them long term is questionable. One small, randomized trial found that over a 4-week period, anxiety declined statistically significantly among patients with cancer receiving alprazolam and those receiving placebo, with no difference between groups (31). Further, it has been shown that quality of life improves statistically significantly in breast cancer survivors as early as 1 year post diagnosis (32). As such, early tapering of these potentially dangerous medications seems prudent.
Our multivariable model additionally revealed that shorter durations of chemotherapy were associated with a higher risk of new persistent sedative-hypnotic use compared with chemotherapy courses lasting more than 4 months. The reason for this finding is not entirely clear. Possible explanations include more steroid exposure during shorter regimens, leading to more steroid-induced anxiety and insomnia. Additionally, patients who receive shorter regimens may have more hormone-related side effects due to the induction of menopause. Finally, it is possible that more psychological distress is associated with shorter courses of treatment, with more time off of active treatment to worry about recurrence. Regardless of the exact explanation, our findings suggest that patients who receive shorter courses of chemotherapy require more diligent monitoring for the development of long-term sedative-hypnotic use.
We also found that the minority of patients who filled an opioid prescription during chemotherapy were more likely to become new persistent users of sedative-hypnotics. Polypharmacy involving controlled substances is common in cancer survivors, who receive opioids and sedative-hypnotics at twice the rate of controls (33). Previous studies have demonstrated that preoperative sedative-hypnotic use increases the risk of long-term opioid use postoperatively (34). Our study adds the related observation that opioid exposure increases the risk of long-term sedative-hypnotic use in the context of chemotherapy. Presumably, the opioid prescriptions received during the chemotherapy period were related to cancer-related surgery and possible complications. Patients who receive both of these controlled substance types at any point require particular attention from clinicians.
Our study has several limitations. First, although our definition of new persistent use has precedence in the literature, there may be other ways of defining new persistent use, which may affect outcomes. For instance, a longer duration of follow-up beyond 1-year post chemotherapy would potentially strengthen the classification of new persistent use. Further, the inclusion of only patients with continuous insurance coverage before and after chemotherapy may have introduced selection bias. Second, because this is a retrospective analysis using health-care claims data, we are unable to determine the exact indications for which sedative-hypnotics were prescribed; thus, it is not possible to know whether prescriptions were appropriate. Third, due to limitations of the data available in MarketScan, we are unable to determine important sociodemographic characteristics for all patients, such as race, ethnicity, and geographic location.
Our study demonstrates that patients who receive adjuvant chemotherapy for breast cancer are at risk of new persistent benzodiazepine and Z-drug use. This may represent an underrecognized but potentially statistically significant complication that affects patients who achieve survivorship. Providers should be aware of this phenomenon and pursue nonpharmacologic approaches to anxiety and insomnia when possible (35,36), especially in high-risk subgroups, such as those receiving concomitant opioid prescriptions and shorter durations of chemotherapy.
Funding
This work was supported by the Breast Cancer Research Foundation (DLH) and the American Cancer Society (DLH). Support for data acquisition provided by the Herbert Irving Comprehensive Cancer Center (P30CA013696).
Notes
Role of the funder: The funders had no role in the study design, collection of data, analysis, interpretation of results, writing of the manuscript or the decision to submit for publication.
Disclosures: JDW: Merck (research support), UpToDate (royalties). JCC, RRR, MPB, MKA, EBE, AM & DLH report no conflicts of interest or disclosures. DLH, a JNCI Associate Editor and co-author on this Article, was not involved in the editorial review or decision to publish the manuscript.
Author contributions: AM: writing—review and editing. DLH: conceptualization; funding acquisition; investigation; methodology; resources; supervision; validation; visualization; writing—review and editing. EBE: writing—review and editing. MKA: conceptualization; writing—review &editing. MPB: writing—review and editing. JCC: conceptualization; data curation; formal analysis; investigation; methodology; project administration; validation; visualization; writing—original draft preparation; writing—review and editing. JDW: conceptualization; writing—review and editing. RRR: data curation; formal analysis; software. YH: data curation; formal analysis; software.
Prior presentations: An abstract for this project was presented at the 2021 San Antonio Breast Cancer Symposium Annual Meeting, San Antonio, TX, USA.
Supplementary Material
Contributor Information
Jacob C Cogan, Division of Hematology, Oncology and Transplantation, University of Minnesota, Minneapolis, MN, USA; Columbia University College of Physicians and Surgeons, New York, NY, USA; Herbert Irving Comprehensive Cancer Center, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA.
Rohit R Raghunathan, Joseph L. Mailman School of Public Health, Columbia University, New York, NY, USA.
Melissa P Beauchemin, Columbia University College of Physicians and Surgeons, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA; Joseph L. Mailman School of Public Health, Columbia University, New York, NY, USA.
Melissa K Accordino, Columbia University College of Physicians and Surgeons, New York, NY, USA; Herbert Irving Comprehensive Cancer Center, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA.
Yongmei Huang, Columbia University College of Physicians and Surgeons, New York, NY, USA.
Elena B Elkin, Herbert Irving Comprehensive Cancer Center, New York, NY, USA; Joseph L. Mailman School of Public Health, Columbia University, New York, NY, USA.
Alexander Melamed, Columbia University College of Physicians and Surgeons, New York, NY, USA; Herbert Irving Comprehensive Cancer Center, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA.
Jason D Wright, Columbia University College of Physicians and Surgeons, New York, NY, USA; Herbert Irving Comprehensive Cancer Center, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA.
Dawn L Hershman, Columbia University College of Physicians and Surgeons, New York, NY, USA; Herbert Irving Comprehensive Cancer Center, New York, NY, USA; New York Presbyterian Hospital, New York, NY, USA; Joseph L. Mailman School of Public Health, Columbia University, New York, NY, USA.
Data Availability
Availability of Data and Material: The data that support the findings of this study are available from IBM MarketScan, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of IBM MarketScan.
Code Availability: The codes utilized during the current study are available from the corresponding author on reasonable request.
References
- 1. Center for Disease Control and Prevention. Opioid Basics: Understanding the Epidemic. https://www.cdc.gov/drugoverdose/epidemic/index.html. Accessed January 20, 2022.
- 2. Lembke A, Papac J, Humphreys K.. Our other prescription drug problem. N Engl J Med. 2018;378(8):693-695. [DOI] [PubMed] [Google Scholar]
- 3. Tardelli VS, Fidalgo TM, Santaella J, et al. Medical use, non-medical use and use disorders of benzodiazepines and prescription opioids in adults: differences by insurance status. Drug Alcohol Depend. 2019;204:107573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Substance Abuse and Mental Health Services Administration. Key Substance Use and Mental Health Indicators in the United States: Results from the 2019 National Survey on Drug Use and Health. https://www.samhsa.gov/data/sites/default/files/reports/rpt29393/2019NSDUHFFRPDFWHTML/2019NSDUHFFR1PDFW090120.pdf. Accessed: January 20, 2022.
- 5. Tardelli VS, Fidalgo TM, Martins SS.. How do medical and non-medical use of z-drugs relate to psychological distress and the use of other depressant drugs? Addict Behav. 2021;112:106606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Vaidya R, Sood R, Karlin N, et al. Benzodiazepine use in breast cancer survivors: findings from a consecutive series of 1,000 patients. Oncology. 2011;81(1):9-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Verger P, Cortaredona S, Tournier M, et al. Psychotropic drug dispensing in people with and without cancer in France. J Cancer Surviv. 2017;11(1):92-101. [DOI] [PubMed] [Google Scholar]
- 8. Bachhuber MA, Hennessy S, Cunningham CO, et al. Increasing benzodiazepine prescriptions and overdose mortality in the United States, 1996-2013. Am J Public Health. 2016;106(4):686-688. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Watanabe JH, Yang J.. Hospitalization and combined use of opioids, benzodiazepines, and muscle relaxants in the United States. Hosp Pharm. 2020;55(5):286-291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Hesketh PJ, Bohlke K, Kris MG.. Antiemetics: American Society of Clinical Oncology clinical practice guideline update summary. J Oncol Pract. 2017;13(12):825-830. [DOI] [PubMed] [Google Scholar]
- 11. Costantini C, Ale-Ali A, Helsten T.. Sleep aid prescribing practices during neoadjuvant or adjuvant chemotherapy for breast cancer. J Palliat Med. 2011;14(5):563-566. [DOI] [PubMed] [Google Scholar]
- 12. Moore TA, Berger AM, Dizona P.. Sleep aid use during and following breast cancer adjuvant chemotherapy. Psychooncology. 2011;20(3):321-325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Traeger L, Greer JA, Fernandez-Robles C, et al. Evidence-based treatment of anxiety in patients with cancer. J Clin Oncol. 2012;30(11):1197-1205. [DOI] [PubMed] [Google Scholar]
- 14. Lee JS, Hu HM, Edelman AL, et al. New persistent opioid use among patients with cancer after curative-intent surgery. J Clin Oncol. 2017;35(36):4042-4049. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Cogan JC, Raghunathan RR, Beauchemin MP, et al. New and persistent controlled substance use among patients undergoing mastectomy and reconstructive surgery. Breast Cancer Res Treat. 2021;189(2):445-454. [DOI] [PubMed] [Google Scholar]
- 16. Marcusa DP, Mann RA, Cron DC, et al. Prescription opioid use among opioid-naive women undergoing immediate breast reconstruction. Plast Reconstr Surg. 2017;140(6):1081-1090. [DOI] [PubMed] [Google Scholar]
- 17. The Truven Health MarketScan Databases for Life Sciences Researchers. https://truvenhealth.com/Portals/0/Assets/2017-MarketScan-Databases-Life-Sciences-Researchers-WP.pdf.
- 18. von Elm E, Altman DG, Egger M, et al. ; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014;12(12):1495-1499. [DOI] [PubMed] [Google Scholar]
- 19. Nattinger AB, Laud PW, Bajorunaite R, et al. An algorithm for the use of Medicare claims data to identify women with incident breast cancer. Health Serv Res. 2004;39(6 Pt 1):1733-1749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ashton CH. Benzodiazepine Equivalence Table. https://www.benzo.org.uk/bzequiv.htm. Revised April 2007. Accessed January 20, 2022.
- 21. World Health Organization. WHO Programme on Substance Abuse: Rational Use of Benzodiazepines. http://apps.who.int/iris/bitstream/handle/10665/65947/WHO_PSA_96.11.pdf?sequence=1&isAllowed=y.
- 22. National Institute for Health and Care Excellence. Hypnotics. https://www.nice.org.uk/advice/ktt6/resources/hypnotics-pdf-1632173521093.
- 23. Carreira H, Williams R, Müller M, et al. Associations between breast cancer survivorship and adverse mental health outcomes: a systematic review. J Natl Cancer Inst. 2018;110(12):1311-1327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Burgess C, Cornelius V, Love S, et al. Depression and anxiety in women with early breast cancer: five year observational cohort study. BMJ. 2005;330(7493):702. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Harrold EC, Idris AF, Keegan NM, et al. Prevalence of insomnia in an oncology patient population: an Irish tertiary referral center experience. J Natl Compr Canc Netw. 2020;18(12):1623-1630. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Palesh OG, Roscoe JA, Mustian KM, et al. Prevalence, demographics, and psychological associations of sleep disruption in patients with cancer: University of Rochester Cancer Center-Community Clinical Oncology Program. J Clin Oncol. 2010;28(2):292-298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Sanford SD, Wagner LI, Beaumont JL, et al. Longitudinal prospective assessment of sleep quality: before, during, and after adjuvant chemotherapy for breast cancer. Support Care Cancer. 2013;21(4):959-967. [DOI] [PubMed] [Google Scholar]
- 28. Savard J, Ivers H, Villa J, et al. Natural course of insomnia comorbid with cancer: an 18-month longitudinal study. J Clin Oncol. 2011;29(26):3580-3586. [DOI] [PubMed] [Google Scholar]
- 29. Conte C, Rueter M, Laurent G, et al. Psychotropic drug initiation during the first diagnosis and the active treatment phase of B cell non-Hodgkin’s lymphoma: a cohort study of the French National Health Insurance Database. Support Care Cancer. 2016;24(11):4791-4799. [DOI] [PubMed] [Google Scholar]
- 30. Kaidar-Person O, Meattini I, Deal AM, et al. The use of psychological supportive care services and psychotropic drugs in patients with early-stage breast cancer: a comparison between 2 institutions on 2 continents. Med Oncol. 2017;34(8):144. [DOI] [PubMed] [Google Scholar]
- 31. Wald TG, Kathol RG, Noyes R Jr, et al. Rapid relief of anxiety in cancer patients with both alprazolam and placebo. Psychosomatics. 1993;34(4):324-332. [DOI] [PubMed] [Google Scholar]
- 32. Hsu T, Ennis M, Hood N, et al. Quality of life in long-term breast cancer survivors. J Clin Oncol. 2013;31(28):3540-3548. [DOI] [PubMed] [Google Scholar]
- 33. Murphy CC, Fullington HM, Alvarez CA, et al. Polypharmacy and patterns of prescription medication use among cancer survivors. Cancer. 2018;124(13):2850-2857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Rishel CA, Zhang Y, Sun EC.. Association between preoperative benzodiazepine use and postoperative opioid use and health care costs. JAMA Netw Open. 2020;3(10):e2018761. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Forbes E, Baker AL, Britton B, et al. Non-pharmacological approaches to procedural anxiety reduction for patients undergoing radiotherapy for cancer: systematic review protocol. BMJ Open. 2020;10(10):e035155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Zachariae R, Amidi A, Damholdt MF, et al. Internet-delivered cognitive-behavioral therapy for insomnia in breast cancer survivors: a randomized controlled trial. J Natl Cancer Inst. 2018;110(8):880-887. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Availability of Data and Material: The data that support the findings of this study are available from IBM MarketScan, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of IBM MarketScan.
Code Availability: The codes utilized during the current study are available from the corresponding author on reasonable request.