A Retrospective Observational Study to Understand Medication Utilization and Lines of Treatment in Patients With Insomnia Disorder

Insomnia is a sleep disorder characterized by difficulties in initiating or maintaining sleep and is recognized as a common public health issue globally.^1,2 In Canada, it affects 9.5%–13.4%, predominantly middle-aged women.^3–5 Persistent insomnia impacts the quality of life, increases the risk of disease,^6,7 increases the economic burden,⁸ and is associated with increased mortality.⁹ Direct and indirect health care costs associated with insomnia in North America can reach $100 billion per year, which is 26%–46% higher than health care costs in those who do not have insomnia.^10,11 The high prevalence of insomnia among COVID-19 patients and its strong association with depression, posttraumatic stress disorder, anxiety, and psychological distress have further increased the health care burden.^12–16

Cognitive behavioral therapy for insomnia (CBT-I) is recommended as the primary intervention for insomnia, and pharmacotherapy can be considered alone or in combination with CBT-I.^17–19 There are several pharmacologic interventions approved by Health Canada, namely, dual orexin receptor antagonists (lemborexant, suvorexant, and daridorexant; DORA) and nonbenzodiazepines (zopiclone, eszopiclone, zaleplon, and zolpidem; Z-drugs). Lemborexant and daridorexant were approved by Health Canada in November 2020 and April 2023, respectively.^20,21 Suvorexant was approved in 2018, but not launched and subsequently removed from the market by the manufacturer in 2020 due to business reasons.²² In addition, 3 other classes, benzodiazepines (BZD), antidepressants, and antipsychotic medications, are used on-label and off-label for patients with insomnia.²³ Sedative-hypnotics such as BZD and Z-drugs bind to γ-aminobutyric acid receptors and are associated with safety concerns, such as complex motor behaviors, an increased risk of respiratory depression, abuse potential, risk of fall, withdrawal syndrome, rebound insomnia, and tolerance.²⁴ Accordingly, Health Canada and the US Food and Drug Administration have updated the safety warnings of BZD and Z-drugs patients, recommending that health care professionals limit the use of these drugs and reinforcing the risk of the combination use of opioid and BZD or other central nervous system (CNS) medications.^25,26 In Canada, drug prescriptions are mostly dispensed at community pharmacies, and physician prescribing practice is consistent across provinces.²⁷ A Canadian study showed a higher BZD dispensing rate in seniors,²⁸ and 75% of new patients’ BZD prescriptions were initiated by family physicians.²⁹

Despite the existence of treatment guidelines,^30–32 there is significant inappropriate medication usage of insomnia treatment, such as higher-than-recommended doses of sedative-hypnotics, long-term use of sedative hypnotics, concurrent use with opioids, and combination use of CNS-active drugs. Several studies have shown that this can cause serious harm such as falls and fractures and increased risk of death, to patients affected by insomnia.²⁴ Chronic use of sedative-hypnotics is associated with worse clinical outcomes and adverse events, especially in older adults,³³ and there is little evidence to prove the efficacy and safety of sedative hypnotics as long-term treatment options for patients with insomnia.³⁴ In addition to the adverse health outcomes, the associated health care cost is also higher due to the inappropriate medication usage.³⁵ Therefore, managing inappropriate drug usage could reduce adverse events, which may help reduce the burden on health care systems, as well as optimize the use of health care resources.

There is paucity of information regarding real-world treatment patterns for insomnia in Canada. The primary objective of this study was to describe the real-world utilization of medications to treat insomnia and the costs incurred, focusing on inappropriate medication usage among patients in Canada. The secondary objective was to understand the lines of treatment among patients with insomnia in Canada.

METHODS

Data Source

This retrospective observational study used private and public administrative claims data from the Canadian IQVIA National Private Drug Plan (PDP) and Ontario Drug Benefit (ODB) databases. IQVIA PDP is the largest private drug plan claim database in Canada and is composed of private insurance carriers, third-party administrators, and benefit plan managers. PDP data capture approximately 80% of all pay-direct private drug plan claims in Canada, with over 12 million active claimants and over 129 million drug claims. The ODB database contains 100% of the fully adjudicated publicly funded prescription claims from Ontario. The ODB program covers most of the drug costs for patients aged ≥65 years, patients aged ≤24 years, and other eligible populations including individuals in long-term care home or on social assistance.³⁶

These databases provide information of drugs dispensed in nonhospital setting related to patient demographics, drugs dispensed, quantity dispensed, days supplied, service date, pharmacy location, cost, and the specialty of the prescribing physician. Diagnosis information is not available; therefore, indication is inferred based on drug claim history and physician specialty. Ethics approval was not required for this study.

Study Design

For the primary objective, a cross-sectional study was conducted to assess utilization in each analysis year over a 3-year period from 1 January 2018 to 31 December 2020 (Figure 1). For the secondary objective of lines of treatment analysis, a longitudinal study was conducted to track patients’ progress through various treatment options over time. The lines of treatment refer to the sequence of different treatments patients received during the analysis period (Supplementary Methods and Supplementary Figure 1). Each time there was a change in existing treatment, it is considered a new “line” of treatment and counted towards the overall count of treatment lines.

Five medication classes, DORA (only lemborexant), Z-drugs, BZD drugs, antidepressant drugs, and antipsychotic drugs, were defined as interventions for insomnia based on the Canadian Agency for Drugs and Technologies in Health.³⁷ Of note, daridorexant was excluded from this study, as it was approved after the study period. Suvorexant was excluded due to low sample size.

Study Population

In this study, patients who had at least 1 claim of lemborexant and/or Z-drugs within their full history were inferred as “patients with insomnia.” Only lemborexant and Z-drugs were used to infer patients with insomnia because they are only indicated for insomnia,^38,39 while other classes can be used on-label or off-label for insomnia and/or other indications. For the primary objective, patients with insomnia who were aged ≥18 years, active in the plan, and had at least 1 claim of defined intervention for insomnia within the analysis year were selected. Selected patients were further grouped by age (adult patients: 18–65 years; senior patients: >65 years) and by payer type (PDP patients; ODB patients). For the secondary objective, patients with insomnia who were aged ≥18 years at the first claim of defined intervention for insomnia, active during the study period, and had at least 1 claim of defined intervention for insomnia during the selection period were selected.

Study Outcomes

In the primary objective, patient demographic characteristics, insomnia medications claimed, duration of supply per patient, prescribing daily dose per patient, daily units per patient, and drug cost per patient were reported. Within demographic characteristics, polypharmacy was defined as the count of distinct medication classes (Anatomical Therapeutic Chemical [ATC] level 2) that patients received during each analysis year. This includes all medications including those unrelated to insomnia. Patients may or may not have received the medications simultaneously. For the inappropriate medication usage analysis, patients with inappropriate medication usage and total cost on inappropriate medication usage were reported. In the secondary objective, patient demographic characteristics, treatment history per patient, and treatment lines were reported. A detailed description of the study outcomes is provided in Supplementary Table 1.

In the inappropriate medication usage analysis, 4 different types of inappropriate medication usage were defined as follows: (1) elevated daily dose was defined as patients exceeding the recommended maximum daily dose, as per the product monograph; (2) extended duration was defined as patients taking BZD and/or Z-drugs for a >30-day duration of supply, per the product monograph and therapeutic guidelines^30,31; (3) combination use was defined as adult patients (18–65 years) who received a ≥1-day combination of any drug molecules within BZD and/or Z-drug; seniors (>65 years) who received a ≥1-day combination of any of the defined medications within the BZD, Z-drugs, antidepressants, and/or antipsychotics classes, as per therapeutic guidelines; and (4) opioid overlap was defined as patients taking BZD and/or Z-drugs overlapping with consecutive claims (≤7-day gap between claims) of opioid drugs for >30 days, as per the product monograph and therapeutic guidelines.³¹ The number of patients with each type of inappropriate medication usage, duration, and total cost of inappropriate usage were calculated in each analysis year.

Statistical Analyses

Descriptive statistics were reported as mean (SD) for continuous variables, and counts and proportions for categorical variables. Comparison between groups is numerical only as this is a descriptive study. Data management and data analysis were performed using SAS version 9.4 (SAS Institute Inc, Cary, NC).

RESULTS

Medication Utilization and Inappropriate Medication Usage

Demographic characteristics. In 2018, 2019, and 2020, a total of 610,721, 597,222, and 578,717 patients received defined interventions of insomnia, respectively. The majority (64%) were female, and over 70% were 18–64 years of age. Over 37% of the patients resided in Ontario, followed by 25% in Quebec, and 99% of the patients were from the PDP database. The majority of patients exhibited polypharmacy of 5 or more drug classes (ATC level 2) in 2018 (69.2%), 2019 (70.0%), and 2020 (52%). The demographic and baseline characteristics of the patient population were comparable across years (Table 1).

Medication utilization. The medication utilization followed similar trends between 2018 and 2020. Patients with 1 type of insomnia medication were more common in PDP compared to the ODB (66.7% vs 38.3% in 2019). There were more seniors who claimed only 1 type of insomnia medication than adult patients (70.5% vs 65.5% in 2019 in PDP and 48.3% vs 26.8% in 2019 on ODB) (Supplementary Table 2). The top 5 medications used for insomnia were zopiclone (Z-drug), lorazepam (BZD), trazodone (antidepressant), clonazepam (BZD), and quetiapine (antipsychotic) (Table 2).

Overall, the ODB patients had longer duration of supply, higher prescribed daily dose, and higher drug cost spent on medications for insomnia (Table 2; Supplementary Table 3). In 2019, the average annual drug cost for insomnia medications was $90.5 for PDP patients, compared to $283.2 for ODB patients. Among PDP patients, senior patients had approximately a month’s longer duration of supply, a lower prescribed daily dose, and a lower drug cost spent on medications compared to adult patients. In 2019, senior patients received medications for an average of 259.9 days’ supply, with the annual drug cost being $77.4 per patient, whereas adult patients received medications for an average of 222.8 days’ supply and a $94.4 cost (Table 2). The utilization of the top 5 medications is also listed in Table 2.

Inappropriate medication usage. Inappropriate medication usage for insomnia increased numerically over the 3 years (2018–2020). The total cost of all insomnia drugs in 2018 was $56.7 million, with 51.8% attributed to inappropriate drug usage ($54.8 million and 55.2% in 2019; $57.7 million and 55.8% in 2020). Between 2018 and 2020, the cost slightly increased among adult and senior patients (Table 3).

Extended duration of BZD and Z-drug use was the most frequent form of inappropriate medication usage, with the highest proportion of patients and cost of inappropriate medication usage (52.0% and 46.6% in 2019). The overlap between BZD/Z-drugs and opioids was less frequent (4.4% and 3.7% in 2018, 4.2% and 3.9% in 2019, and 4.1% and 3.7% in 2020) (Table 3).

Overall, a higher proportion of senior patients compared to adult patients showed inappropriate medication usage across all the 3 years (69.0% vs 52.2% in 2018, 69.5% vs 52.5% in 2019, and 69.4% vs 54.1% in 2020). Correspondingly, senior patients had a higher proportion of cost for inappropriate usage than adult patients (74.7% vs 46.5% in 2018, 77.3% vs 49.7% in 2019, and 77.7% vs 50.3% in 2020) (Table 3). The percentage of the cost of medication due to overlap between BZD/Z-drugs and opioids in adult and senior populations was 3.6% and 4.2% in 2018, 3.8% and 4.4% in 2019, and 3.6% and 4.2% in 2020. ODB patients had higher inappropriate usage than PDP patients (73.3% vs 55.7% in 2018, 82.6 vs 56.3% in 2019, and 85.4% vs 57.6% in 2020). The total cost of inappropriate medication usage was lower in ODB patients than in PDP patients (47.8% vs 55.4% in 2019) (Table 3).

Lines of Treatment

Demographic characteristics. A total of 240,820 patients met our selection criteria (study population). The majority were female (n = 148,689, 61.7%). Adult patients (18–64 years) comprised 81.8% of the cohort, with a mean age of 54 years. 99% of patients were from the PDP database, and 91,791 resided in the Ontario region (38.1%), while 61,451 (25.5%) were residents of Quebec (Supplementary Table 4).

Treatment history. The median treatment history from patients’ first drug claim for insomnia to the most recent claim was 3.5 years (interquartile range: 0.7, 8.0). Mean treatment history was 5.1 years (SD 5.2) (Supplementary Table 5).

Number of treatment lines. The mean number of treatment lines was 6.4 (SD 14.3). Most patients (33.8%) received 1 line of treatment during the observation period, and 6.2% of patients had >20 lines of treatment over their entire insomnia intervention (Figure 2; Supplementary Table 5).

DISCUSSION

This retrospective observational real-world study examined insomnia treatments in a large, representative sample of Canadian patients. We described drug utilization, inappropriate drug usage, and lines of treatment across age and payer cohorts to understand insomnia intervention patterns from 2018 to 2020.

Canada has a mix of public and private insurance plans to cover prescription drugs in the outpatient setting. The public plans are offered by provincial and territorial governments to eligible residents based on age, income, and medical conditions.^40,41

In this study, most patients were from the PDP database. Patients from the ODB database claimed multiple classes of insomnia medications more often and had a longer duration of medication use and higher treatment costs than private plan patients. This low capture of insomnia patients in public plan and difference in drug utilization between public and private patients could be explained by the restricted access to on label prescription interventions for insomnia in public plan. In Ontario, almost all BZD, antidepressants, and antipsychotic drugs are funded publicly, while Z-drugs and lemborexant are not. Zopiclone is only covered under the Exceptional Access Program (a program designed to facilitate patient access to drugs in Ontario).

There was a higher proportion of high polypharmacy patients (≥ 5 drug classes) in 2018 and 2019 compared to 2020. A lower polypharmacy in 2020 is likely a result of the COVID-19 pandemic, which reduced patient access to health care services and reduced medication prescriptions.⁴² More than 52% of patients exhibited high polypharmacy (≥5 drug classes), which can be associated with higher risks, eg, drug-drug interactions; therefore, it is important to appropriately manage treatment.³²

Inappropriate medication use, specifically for BZD and Z-drugs in the elderly, has been studied in Canada and globally. A study in Manitoba reported that 15.6%–35.1% of patients with Z-drugs had long-term use (≥90–180 days after first prescription).⁴³ Similarly, a study in United States reported 31.4% of patients with BZD long-term use (≥120 days).⁴⁴ In our study, we observed >50% of patients with extended duration of BZD and Z-drugs, which is higher than other studies. This could be explained by the different definition of long term use of 30 days in this study. Overall, the study results indicate that the prevalence of inappropriate use among patients with insomnia is high in Canada, and clinical guidelines and product monograph indications are not being followed appropriately. It is therefore important to raise awareness of inappropriate use, properly follow clinical guidelines to improve the medication prescribing practices, and explore other avenues such as deprescribing BZD and Z-drug to further reduce the potential harms among senior patients with insomnia. It needs efforts from all stakeholders including patients, physicians, pharmacists, and nurse practitioners to take action.

In the lines of treatment analysis, 6.2% of patients had >20 lines during their insomnia intervention history, suggesting that a subgroup dissatisfied with current interventions cycled through limited medication options. This indicates a need for better clinical assessment of insomnia by leveraging appropriate assessment tools and investigating patients’ pattern of drug use to provide proper management of insomnia and new treatments.^45–47

This study utilizes a large and representative patient population from the public and private databases of Canada, allowing for an overview of drug utilization, inappropriate usage of medications, and lines of treatment in both the adult and senior populations with insomnia in Canada. This is the first study, to our best knowledge, that assessed the real-world drug utilization among patients with insomnia in Canada and specifically generated real-world evidence of inappropriate medication usage for the management of insomnia.

This study has several limitations worth noting. First, this study did not analyze cash-paying patients or publicly funded prescriptions outside of Ontario. Second, diagnostic information is not included in the PDP and ODB claims databases; therefore, the indication of insomnia in this study was inferred based on the prescription of 2 medication classes (lemborexant and Z-drugs) approved specifically for insomnia. The insomnia cases in Canada may be underestimated considering patients with insomnia may not take lemborexant and/or Z-drugs due to limited access or may choose other on-label drugs such as doxepin (3 mg or 6 mg) or some BZD drugs (flurazepam, nitrazepam, triazolam, and temazepam) as well as off-label options (BZD drugs and antidepressants or antipsychotic drugs). Further, this study followed a restricted and conservative approach (30-day consecutive use instead of 7–14 days, as per the product monograph) and did not include other interventions, such as alcohol and over-the-counter drugs. Considering these limitations, the insomnia population selected in this study could potentially be an under-representation of the true size.

CONCLUSION

This study described the drug utilization and the inappropriate usage of insomnia medications despite published prescribing guidelines and label recommendations in Canada. Inappropriate usage of insomnia medications was causing a substantial health care burden in Canada, especially in senior patients. This study demonstrates that there is a high prevalence of inappropriate medication usage in insomnia, particularly in seniors. Overall, this study shows a need for better education to general practitioners and the public of current guidelines^30–32 and more effective and safer treatment options as well as appropriate patient monitoring tools for physicians, particularly for seniors who may require long-term treatment.

Article Information

Published Online: September 11, 2024. https://doi.org/10.4088/JCP.23m15015
© 2024 Physicians Postgraduate Press, Inc.
Submitted: July 14, 2023; accepted July 9, 2024.
To Cite: Kamboj L, Ramos B, Haynes A, et al. A retrospective observational study to understand medication utilization and lines of treatment in patients with insomnia disorder. J Clin Psychiatry. 2024;85(4):23m15015.
Author Affiliations: Eisai Ltd, Mississauga, Canada (Kamboj, Ramos, Haynes, Sohi); IQVIA Solutions Canada, Inc, Kirkland, Canada (Yang, Ling, Barot, Millson); Division of Geriatric Psychiatry, Queen’s University, Providence Care Hospital, Kingston, Canada (Amanullah).
Corresponding Author: Laveena Kamboj, MSc, Eisai Ltd, 6925 Century Ave #701, Mississauga, ON L5N 7K2, Canada ([email protected]).
Author Contributions: All authors contributed to data analysis and interpretation. Kamboj, Ramos, Sohi, Yang, Ling, Barot, and Amanullah contributed to study conceptualization. Sohi, Yang, Ling, and Barot contributed to study design. Yang contributed to data collection. All authors contributed to drafting, editing, and critically revising the manuscript. All authors are accountable for the work and provided their agreement for submission to the journal and approval for publication.
Relevant Financial Relationships: Ms Kamboj, Dr Ramos, and Mr Haynes are employees of Eisai. Dr Sohi is a former employee of Eisai. Dr Yang, Mss Ling and Barot, and Mr Millson are employees of IQVIA Solutions Canada Inc and received study funding from Eisai. Dr Amanullah is affiliated with Providence Care Hospital and received funding for support on this study.
Funding/Support: This study was sponsored by Eisai Ltd, Canada.
Role of the Funders/Sponsors: Eisai was involved in the study design and preparation of the manuscript. IQVIA was involved in the study design and preparation of the manuscript.
Previous Presentation: The study was disseminated at the Choosing Wisely Canada 2022 Conference; May 25–26, 2022; virtual, as an abstract, and was presented at the Canadian Association for Population Therapeutics (CAPT) 2022 Conference; October 17–18, 2022; Toronto, Canada, as an abstract and poster.
Acknowledgments: The authors thank Oliver Sang, BSc (IQVIA Canada), for data analysis support and thank Gargi Pal, PhD; Shlok Kumar, MA; and Lakshman Puli, PhD (IQVIA, India), for medical writing and editorial support. Study analysis and medical writing assistance were provided by IQVIA Canada.
Supplementary Material: Available at Psychiatrist.com.

Clinical Points

• There is limited information regarding real-world treatment patterns for insomnia patients in Canada.
• There is a high prevalence of inappropriate medication usage in insomnia, and a subset of patients received more than 20 lines of insomnia interventions, highlighting a need for more effective and safer treatment options.