6

The Effect of Benzodiazepine Co-Prescribing¹ on All-Cause Mortality, Including Suicide Mortality, Among Individuals on Consistent Opioid Pharmacotherapy

INTRODUCTION

The studies explored in the previous two chapters compared the effect of newly dispensed² opioid pharmacotherapy compared to non-opioid pain pharmacotherapy on all-cause mortality and suicide mortality in individuals who are opioid naïve (Chapter 4), as well as the effects of initial opioid dosage and dosage escalation strategies in opioid-naïve individuals on all-cause mortality and suicide mortality (Chapter 5). However, among individuals consistently dispensed opioid pharmacotherapy, or long-term opioid treatment (LTOT), the prescribing of other pharmacotherapies for the treatment of co-occurring physical and mental health conditions may increase the risk of harm. Of particular concern to individuals, clinicians, and policy makers is the potential for harm when individuals receiving consistent LTOT are prescribed benzodiazepine pharmacotherapy. In this chapter, the committee focused on the effect of newly dispensed benzodiazepine pharmacotherapy compared to alternative non-benzodiazepine pharmacotherapy in individuals consistently dispensed opioid pharmacotherapy on all-cause mortality and suicide mortality. The committee identified alternative non-benzodiazepine pharmacotherapy as those that treat insomnia or anxiety.

Clinical Use of Benzodiazepine Pharmacotherapy

Benzodiazepines are prescription sedatives typically used to treat anxiety, insomnia, panic disorders, alcohol withdrawal, seizures, and muscle spasms as well as for sedation in the perioperative period (DEA, 2020; FDA, 2020; Lembke et al., 2018; Olfson et al., 2015). As of 2020, the Food and Drug Administration (FDA) approved (on-label use) select benzodiazepines in clinical settings to treat generalized anxiety disorder (GAD), insomnia, seizures, social phobia, and panic disorder (FDA, 2020). Some benzodiazepines, such as diazepam, lorazepam, and clonazepam, are not FDA approved for GAD in adults, but they are frequently used off label for this purpose (Strawn et al., 2018). Similarly, other select benzodiazepines associated with sedative and hypnotic effects (including clonazepam and lorazepam) are frequently used off label to treat adult insomnia (Ghiasi et al., 2024; Basit and Kahwaji, 2023; Madari et al., 2021). Similar to opioids, benzodiazepines can cause mild to severe depression

___________________

¹ Operationalized as co-dispensing in the analysis.

² In the committee’s studies “newly dispensed” was defined as pharmacotherapy of interest not previously dispensed to the individual in the last 90 days before the start date.

of the central nervous system (see Box 1-1). Long-term use of benzodiazepines can lead to an increased tolerance and the need for higher dosages, contributing to dependence and the potential development of a substance use disorder (Fluyau et al., 2018). However, short-term exposure to benzodiazepines, or changes in dose, may have adverse effects related to respiratory depression and oversedation. When combined with other prescribed pharmacotherapies, short-term use of benzodiazepines can have adverse effects resulting in death (Hernandez et al., 2018; Passaro et al., 2000).

Prescribing Benzodiazepine Pharmacotherapy

The percentage of U.S. adults prescribed benzodiazepines increased during the study 3 period (2007–2019). Based on an analysis of the Medical Expenditure Panel Survey, the proportion of adults with a filled benzodiazepine prescription increased from 4.1 percent in 1996 to 5.6 in 2013 (Bachhuber et al., 2016). Moreover, data from the National Ambulatory Medical Care Survey showed a marked increase in the proportion of individuals prescribed benzodiazepines in the outpatient setting, from 3.8 percent in 2003 to 7.4 in 2015 (Agarwal and Landon, 2019). An analysis of 2015–2016 data from the National Survey on Drug Use and Health found that 12.6 percent of U.S. adults reported benzodiazepine use in the past year (Maust et al., 2019). Toward the end of the study 3 period and beyond, 2017–2021, benzodiazepine dispensing rates declined—from 31.2 per 100 persons in 2017 to 25.8 per 100 persons in 2021.³

Risks Associated with Benzodiazepine Use

Risks associated with benzodiazepines include deaths due to overdose and suicide, as well as falls, fractures, cognitive dysfunction, motor vehicle crashes, non-fatal overdose, anterograde amnesia, of which older individuals are especially at increased risk (Edinoff et al., 2021; AGS, 2019; Dodds, 2017; Glass et al., 2005; Stewart, 2005; Cumming and Couteur, 2003; Wang et al., 2001; Hemmelgarn et al., 1997; Ray et al., 1992; Curran, 1991). Furthermore, chronic benzodiazepine use is associated with an elevated risk of dementia and Alzheimer’s disease (de Gage et al., 2014, 2012). Additionally, Kripke and colleagues (2012) reported greater than a threefold increase in likelihood of mortality among individuals taking benzodiazepines and other hypnotics.

Further, although clinical guidelines of the American Geriatrics Society recommend that older adults avoid benzodiazepines, studies continue to demonstrate that use of benzodiazepines was about three times more prevalent in adults over 50 compared to younger adults (AGS, 2019; Maust et al., 2019; Olfson et al., 2015; Wysowski and Baum, 1991).

Data from the Centers for Disease Control and Prevention (CDC) highlights that benzodiazepine-related drug overdose deaths increased from 1,135 in 1999 to 11,537 in 2017 and declined to 9,711 in 2019 (NIDA, 2023). Between 2019 and 2021, deaths rose again to 12,499 before beginning to decline again in 2022 (NIDA, 2023). In overdose deaths with benzodiazepines and opioids (2000–2019), 86.2 percent were classified as accidental, 8.5 percent as suicides, and 5.2 percent as undetermined intent (Kleinman and Weiss, 2022). Among overdose deaths involving benzodiazepines, 55.9 percent were classified as accidental, 36.2 percent as due to suicide, and 7.7 percent as underdetermined intent (Kleinman and Weiss, 2022). Overdose deaths with both benzodiazepines and antidepressants (e.g., tricyclic antidepressants (TCAs), monoamine oxidase inhibitors, and other unspecified antidepressants) were categorized as suicide in 51.4 percent of cases from 2000–2019 (Kleinman and Weiss, 2022).

Alternative Pharmacotherapies to Benzodiazepines

Alternative pharmacotherapies for anxiety, insomnia, and other indications for benzodiazepines include TCAs, tetracyclic antidepressants (TeCAs), serotonin-norepinephrine reuptake inhibitor (SNRIs), buspirone, and selective serotonin reuptake inhibitors, which are sometimes used to treat depression, anxiety, and other mental health conditions (Garakani et al., 2020; Krystal et al., 2019; Dale et al., 2015; Bespalov et al., 2009; Wilson and Nutt, 2007).

___________________

³ Dr. Christopher Jones. Presentation to the committee on March 9, 2023. Trends in Opioid and Benzodiazepine Use, Misuse, and Overdose.

Concomitant Use of Opioids and Benzodiazepines

Concomitant use of opioids and benzodiazepines can have significant adverse effects. This section highlights trends and studies that examine this co-prescribing in the general population and its potential risks.

Trends in Private-Sector Co-Prescribed Opioids and Benzodiazepines

Before release of the 2016 CDC Clinical Practice Guidelines (CPGs) for Prescribing Opioids for Pain, coprescription of opioids and benzodiazepines was increasing nationally, affecting a sizable percentage of individuals between 2001–2010. Data from the National Ambulatory Medical Care Survey/National Hospital Ambulatory Medical Care Survey showed that benzodiazepines and opioids were co-prescribed at 8.1 percent of acute pain visits and 15.5 percent of chronic pain visits (Larochelle et al., 2015). In a study examining more than 300,000 continuously insured individuals receiving opioid prescriptions between 2001 and 2013, the percentage of persons also receiving prescribed benzodiazepines increased from 9.0 percent in 2001 to 17.0 in 2013 (Sun et al., 2017). Analysis of data from the National Health and Nutrition Examination Survey (NHANES) showed that co-prescribing increased from 11.5 percent in 2007–2008 to 16.1 in 2013–2014, decreased to 15.0 percent in 2015–2016, and then decreased further to 8.7 percent in 2017–2018 (Rhee et al., 2021). A decrease in the percentage of individuals with concurrent prescriptions of both benzodiazepines and opioids was also observed in claims data from both commercial insurance and Medicare Advantage following the release of the CDC guidelines in 2016 (Jeffery et al., 2019).

Adverse Outcomes Associated with Opioid and Benzodiazepine Co-Prescribing

Long-term opioid and benzodiazepine treatment, both independently and together, is associated with adverse outcomes and all-cause mortality, including deaths due to accidents, suicide, and overdose, regardless of misuse (VA/DoD, 2022; Gibbons et al., 2021; Xu et al., 2020; Bohnert and Ilgen, 2019; Votaw, 2019; Gressler et al., 2018; VA/DoD, 2017; Sun et al., 2017; Park et al., 2015). Adverse outcomes associated with the concomitant use of benzodiazepines and opioids include an increased risk of overdose (Bachhuber et al., 2016; Park et al., 2015; Sharp and Melnik, 2015; Turner and Liang, 2015; Chen et al., 2014), suicide (Pfeiffer et al., 2009), and all-cause mortality (Xu et al., 2020; Gaither et al., 2016; Weich et al., 2014); the use of emergency services (Yarborough et al., 2019; Sun et al., 2017; Nielsen et al., 2015); and number of serious emergency department visits (e.g., hospitalization or death) (Yarborough et al., 2019; Kaufmann et al., 2017; Jones and McAninch, 2015; Day, 2014). Furthermore, concomitant use of benzodiazepines and opioids can cause a synergistic effect that augments the opioids’ euphoric effects in the brain, potentially increasing risk of addiction and subsequent adverse effects (Jones et al., 2012; Lintzeris and Nielsen, 2010).

All-Cause Mortality.

The concomitant use of benzodiazepines and opioids has been found to further suppress respiratory drive (the intensity of the respiratory system output, which results in breathing [Jonkman et al., 2020]), a common cause of death from opioid overdose, and compound each medication’s independent adverse effects that increase risk for all-cause mortality (Xu et al., 2020; White and Irvine, 1999). A study using NHANES and National Death Index (NDI) data, and with a reference comparator group of those prescribed selective-serotonin reuptake inhibitors, Xu and colleagues (2020) found a higher risk for all-cause mortality in those with concomitant use of prescribed benzodiazepines and opioids (hazards ratio (HR): 2.04; 95% CI: 1.65–2.52) and those prescribed only benzodiazepines (HR: 1.60; 95% CI: 1.33–1.92).

Suicide Mortality.

Suicide and overdose mortality are of particular concern with concurrent prescription of opioids and benzodiazepines (Xu et al., 2020; White and Irvine, 1999). This combination of pharmacotherapy can have adverse effects given that opioid-induced respiratory depression. The sedating effects of benzodiazepines can further suppress breathing and increase the risk for fatal overdose and mortality. However, there are challenges in distinguishing between overdose mortality and deaths due to suicide as intentional or accidental, as well as challenges in attributing the cause of death to one medication or another (Hedegaard et al., 2018). These challenges complicate the ability to reliably infer which pharmacotherapy was the leading contributor to mortality and understand if

overdose was intentional or unintentional in both the clinical setting and observational studies. Most studies described in the next section report on overdose mortality.

A high percentage of overdose deaths are associated with both benzodiazepines and opioids. Of 118,208 U.S. benzodiazepine-related overdose deaths between 2000–2019, 83.5 percent were associated with opioids (Kleinman and Weiss, 2022). A 2014 report referencing data from the National Vital Statistics System reported that benzodiazepines were involved in 31 percent of the opioid-analgesic poisoning deaths in 2011 (Chen et al., 2014). Based on the number of national drug overdose deaths, in 2018, about 20 percent of overdose deaths involving opioids also involved benzodiazepines (NIDA, 2023). By 2021, 14 percent of overdose deaths involved both opioids and benzodiazepines (NIDA, 2023). Another study based on state-controlled substance prescription drug monitoring program data demonstrated a 10 times higher risk of death from overdose among individuals receiving concomitant opioid and benzodiazepine in the past year compared to those who received only opioids (Dasgupta et al., 2016).

A study using Veteran Health Administration (VHA) data from 2004 to 2009 found that approximately 50 percent of the veterans receiving opioids who died from overdose had a concurrent benzodiazepine prescription (Park et al., 2015). From 1996 to 2010, the rate of benzodiazepine overdose deaths per 100,000 U.S. adults increased fourfold; this rate then plateaued from 2010 to 2013 (Bachhuber et al., 2016). Opioid-involved benzodiazepine overdose deaths increased from 2000 to 2017 (0.32 per 100,000 to 3.08 per 100,000) and then decreased in 2019 to 2.53 per 100,000 in the United States (Kleinman and Weiss, 2022). Between 2000 and 2019, 118,208 benzodiazepine-involved overdose deaths occurred, with opioids co-involved in 83.5 percent of the deaths. Of overdose deaths co-involving opioids and benzodiazepines, 86.2 percent were classified as accidental, 8.5 percent as suicide, and 5.2 percent as undetermined intent. In 2019, 53.3 percent of benzodiazepine-involved overdose deaths co-involved synthetic opioids, 34.6 percent co-involved natural and semisynthetic opioids, and 21.1 percent co-involved heroin (Kleinman and Weiss, 2022).

Based on 2013–2014 Medicare Part D data, Hernandez and colleagues (2018) found that during the first 90 days of concurrent opioid and benzodiazepine use, opioid-related overdose risk was five times higher than among those using opioids only. However, after 180 days, the overdose risk was similar between the two groups.

Guidelines on Concomitant Prescription of Opioids and Benzodiazepines

Before 2016, no explicit contraindication guidelines existed for co-prescribing opioids and benzodiazepines (an earlier VA/Department of Defense (DoD) CPG provided guidance on co-prescribing benzodiazepines with methadone and fentanyl, specifically) (VA/DoD, 2010). In 2016, the FDA added a black box warning (its highest safety warning) to benzodiazepines, opioids, and opioid-containing cough medicines indicating the dangers of independent and combined use (FDA, 2016). Subsequent CPGs, 2016 U.S. CDC CPG and the 2017 VA/DoD CPG for the Management of Opioid Therapy for Chronic Pain (last updated in 2022), included strong, evidence-based recommendations to not co-prescribe any opioid with any benzodiazepine. Additionally, the Centers for Medicare & Medicaid Services (CMS) has raised concerns and explicitly recommended against concurrent use (CMS, 2019).

As noted, there are many potential clinical reasons for co-prescribing opioids and benzodiazepines, such as the co-occurrence of anxiety and other mental health conditions, sleep disturbance in individuals experiencing pain, the cyclical and compounding effect of mental health conditions on pain, and short-term needs (e.g., dental visits, magnetic resonance imaging (MRI)) (Hirschtritt et al., 2021; Quagliato et al., 2018; Kroenke et al., 2011; Tunks et al., 2008; Chou, 2007; McWilliams et al., 2003). Hawkins and colleagues (2017) conducted a survey among prescribers to understand decision making behind co-prescribing benzodiazepines and opioids among veterans receiving care from the VHA. The study results identified several factors that contributed to co-prescribing: lack of information among providers regarding other treatments, lack of information regarding individual behavior with other treatments, insufficient time to discuss discontinuation of one of the medications, and the view that discontinuation of benzodiazepines would both be too difficult and contribute to individual discomfort (Hawkins et al., 2017).

STUDY 3 METHODS

Study 3 aimed to answer the following research question: among veterans receiving care in the VHA who were consistently⁴ dispensed⁵ opioid pharmacotherapy between 2007 and 2019, what is the effect of newly dispensed benzodiazepine versus alternative non-benzodiazepine pharmacotherapy for anxiety and other common indications for benzodiazepines on all-cause mortality (primary outcome) and suicide mortality (secondary outcome) within a 3-month follow-up period?

The committee designed this observational analysis to emulate a target trial estimating the effect of benzodiazepine pharmacotherapy compared with alternative non-benzodiazepine pharmacotherapy on all-cause mortality and suicide mortality among those who were consistently dispensed opioid pharmacotherapy. A target trial is the hypothetical randomized pragmatic trial that would have best answered the causal question. Target trial emulation (TTE) framework includes two steps: (1) specify the protocol of the target trial, and (2) specify how the target trial will be emulated using observational data. The TTE framework is useful because it helps to ensure comparison of realistic treatment strategies, avoids biases from misalignment of time anchors (time zero/index date), and uses enhanced statistical techniques for minimizing confounding (Gomes et al., 2022; Hernán and Robins, 2016).

The protocol and specifications of the target trial (study 3) and how the committee emulated the target trial in the observational data are described in the following subsection and summarized in Table 6-1.

In addition, given the complexity of each study, the committee developed Figure 6-1, which reflects key temporal aspects of the longitudinal study design, specifically the index or date of entry into the study, exposure washout period, windows for exclusion and covariate assessments, and follow-up time (Schneewiss et al., 2019).

Study Population and Data Sources

The study population is defined as veterans receiving care in the VHA, and the study period is from January 1, 2007, to December 31, 2019.⁶ The earliest index date, or start date, was January 1, 2007.⁷ The latest possible index date was October 1, 2019, to allow individuals to be followed for 3 months after a newly dispensed benzodiazepine or alternative non-benzodiazepine pharmacotherapy. To assess exclusion and eligibility in the study and follow-up, data from 2006–2019 were pulled for analyses.

Data from the national VHA data files were linked to the Department of Veterans Affairs (VA) Corporate Data Warehouse (CDW), U.S. Veterans Eligibility Trends and Statistics, NDI, and Medicare Part A, B, C, and D data from the CMS. Primary sources for individual inclusion/exclusion and covariates were VA and VHA data. CMS data were used to supplement individual data used for inclusion/exclusion and other covariates. Data files were linked based on a combination of a veteran’s unique individual integrated control number, Social Security number (SSN), and date of birth. The National Academies of Sciences, Engineering, and Medicine Institutional Review Board approved the study.

SPECIFICATIONS FOR THE EMULATED TARGET TRIAL 3

Drawing from this study population of all veterans who received care in VHA facilities, this study focused on a population consistently dispensed opioid pharmacotherapy that were newly dispensed either a benzodiazepine or an alternative non-benzodiazepine pharmacotherapy. Figure 6-2 illustrates the flowchart detailing the selection process of including eligible veterans in study 3. The following sections outline each of the target trial components for study 3.

___________________

⁴ The committee defined as three or more dispensed opioids at least 21 days apart within a 180-day period for at least 84-day supply or two dispensed opioids of 90-day supply within 180 days.

⁵ The committee notes differences in types of pharmacy data measures. Pharmacy data can be categorized into three measures: prescribed (prescriber submits a prescription to a pharmacy), a fill (pharmacy completes the requested prescription), and dispensed (individual picks up/is mailed the prescription from the pharmacy). The committee used dispensed pharmacy data in its analyses. The committee operationalized the term “initiation” in the statement of task as “newly dispensed” pharmacotherapy in analyses.

⁶ Latest possible index date was October 1, 2019, in study 3.

⁷ Given that the earliest index date is January 1, 2007, the earliest pre-index data are January 1, 2006.

TABLE 6-1 Specifications for Study 3 Target Trial Protocol

¹ MOUD formulation of buprenorphine, methadone, naltrexone or LAAM.

SOURCE: Adapted from Hernán and Robins, 2016 and NASEM, 2019.

Study 3 Eligibility Criteria

For each calendar month from January 1, 2007, to December 31, 2019, the committee identified eligible individuals who met the following criteria: (1) be a U.S. veteran, (2) be 18 years or older, (3) have a valid SSN, and (4) have at least 1-year continuous enrollment in the VHA between 2006–2019, defined as at least one encounter in the VHA in the 12-month pre-index period. Eligible individuals also included veterans with at least one encounter in the VHA in the 12-month before index date within the eligible study period but only had a pharmacy claim through Medicare Part D (no VHA pharmacy claim). In addition, those who were eligible did not have an encounter or dispensed pharmacotherapy in Guam, Manila, or Singapore VHA facilities. These veterans were excluded because mortality data collected in these regions do not appear in the NDI data. See “exclusion criteria” for further information.

To be eligible for study 3, and similar to a prior analysis (Larochelle et al., 2022), VHA veterans had to be consistently dispensed opioid pharmacotherapy (defined as three or more dispensed opioids at least 21 days apart within a 180-day period for at least 84-day supply or two dispensed opioids of 90-day supply within 180 days) and have no dispensed benzodiazepine or an alternative non-benzodiazepine pharmacotherapy (see Selection of Pharmacotherapies of Interest section for more details on the committee’s approach in selecting alternative non-benzodiazepine pharmacotherapies) in the 90-day pre-index period (the washout period). The committee established the 90-day washout period as the maximum duration of any dispensed prescription in the VHA is 90 days (except in rare circumstances [e.g., hospice]) (VHA, 2006).

Study 3 Eligibility Exclusion

The study had an exclusion assessment window of 365 days before the index date. VHA veterans who received hospice or palliative care or had a cancer diagnosis (except non-melanoma skin cancers) were excluded. International Classification of Disease (ICD) diagnosis codes were used to identify conditions (see Appendix I for list of codes). Individuals receiving palliative care and/or in hospice were excluded given that having a terminal illness (and thus being eligible for these services) is related to both death and more permissive use of opioid pharmacotherapy and thus is a confounder in the relationship between opioid pharmacotherapy and death. In addition, individuals with cancer have an increased risk of mortality and often receive an opioid pharmacotherapy to manage cancer-related pain. Decisions regarding opioid initiation, co-prescribing, and dosage increases may differ for cancer compared to non-cancer pain; policies and practices for opioid prescribing is (generally) more permissive for individuals with cancer (VA/DoD, 2022, 2003, 2017, and 2010; Dowell et al., 2016). Furthermore, study exclusion criteria are in line with published studies examining opioid pharmacotherapy, which excluded individuals with cancer (Song et al., 2022; Coyle et al., 2018; Berna et al., 2015; Turner et al., 2015; Gomes et al., 2011; Dunn et al., 2010). As a result, having cancer may be a confounder between treatment regimen and mortality; therefore, individuals with cancer, other than non-melanoma skin cancer, were excluded.

To ensure that individuals receiving medications for opioid use disorder (MOUD) treatment were excluded from the study, the committee excluded any individuals who had received ≥1 dispensed MOUD formulation of levo-alpha acetyl methadol (LAAM), methadone, buprenorphine, or naltrexone in the 90-day pre-index period. In study 3, veterans with all qualifying episodes⁸ beginning on or after October 1, 2019, were excluded to ensure that each veteran included in the study had the potential for a complete 3-month follow-up period.

The causal question for this study focuses on the effect of individuals newly dispensed benzodiazepine versus alternative non-benzodiazepine pharmacotherapy among veterans who were consistently dispensed opioid pharmacotherapy, regardless of the reason. The study did not exclude (a) veterans prescribed opioid pharmacotherapy who have an opioid use disorder (OUD) during follow-up or (b) individuals with surgery or acute painful injury within the 90 days before the index date. The committee agreed that OUD diagnosis alone is not an absolute contraindication for opioid prescribing: the OUD diagnosis may not be recent, and individuals may not actively be receiving treatment during the time of the study. In addition, OUD was not considered a covariate by the committee because it is within the causal pathway of interest. Given these reasons, the committee did not exclude individuals with an OUD diagnosis. The committee aimed to capture all opioid pharmacotherapy exposure (except cancer-related and end-of-life pain, given the increased likelihood of mortality) (see Chapter 3 for further rationale).

Study 3 excluded veterans who were dispensed both benzodiazepine and non-benzodiazepine pharmacotherapy during the same episode, which is consistent with the statement of task’s focus on comparing the two (“3. The effect of co-prescribing a benzodiazepine among patients currently receiving long-term opioids, relative to alternative treatments for anxiety and other indications for benzodiazepines”). See Figure 6-2 for the selection process of veterans for the study.

Study 3 Treatment Strategies

Drawing from this study population of eligible VHA veterans, this study considers the initiation of two mutually exclusive, clinically realistic treatment strategies:

• Of those veterans consistently dispensed opioid pharmacotherapy,
  • Treatment group: newly dispensed benzodiazepine pharmacotherapy
  • Comparator group: newly dispensed alternative non-benzodiazepine pharmacotherapy.

Table 6-2 shows the list of eligible benzodiazepine pharmacotherapy and alternative non-benzodiazepine pharmacotherapy.

___________________

⁸ All veteran episodes (regardless of treatment assignment).

TABLE 6-2 Pharmacotherapies of Interest in Study 3

¹ There are two types of opioids: full agonist and partial opioid agonist. Full opioid agonists fully activate the mu receptors in the brain, enabling the opioid to have “full” effect, such as codeine, oxycodone, and fentanyl. Partial opioid agonists also activate mu receptors in the brain but to a lesser degree, such as butorphanol or tapentadol.

² LAAM: levo-alpha acetyl methadol.

³ MOUD: medications for opioid use disorder.

Study 3 Treatment Assignment

In the hypothetical target trial, eligible individuals would be randomly assigned by clinical researchers to a treatment strategy with their consent and followed; individuals would be aware of their treatment assignment. The emulated target trial (study 3), based in retrospective observational data, treatment assignments were based on the observed prescription drug events (claims) from Medicare Part D or dispensed pharmacotherapy from the VHA electronic health records (EHRs). Thus, the analyses considered prescription dispensing records to assign individuals to a treatment strategy, unlike in the hypothetical target trial for which researchers would determine the treatment assignment, which could be different from what medications⁹ ultimately were dispensed. It is also important to distinguish the measurability of prescription dispensing data from the measurability of that which is ingested by the individual. A dispensed prescription does not necessarily mean that the medication was consumed, since these medications are commonly prescribed to be taken “as needed.” For this study, a measure of individuals ingesting the medication was not available in the dataset analyzed; the committee used pharmacy dispensing data as a proxy for medications ingested by the individual. After the first dispensed eligible pharmacotherapy, 30-day pharmacotherapy episode interval was created for eligible veterans. All episodes per veteran were selected for inclusion.

Selection of Pharmacotherapies of Interest

In study 3, the committee defined initiation as receiving ≥1 tablet of either a benzodiazepine pharmacotherapy (treatment group) or alternative non-benzodiazepine pharmacotherapy (comparator group); however, in sensitivity analyses, the committee estimated the impact of receiving ≥5 tablets.

Benzodiazepine pharmacotherapy included in the analysis were those available for prescribing in VHA during the study period (2007–2019). Alternative non-benzodiazepine pharmacotherapy for the comparator group were empirically selected through a multistep approach. First, the most frequent mental health conditions among VHA individuals with dispensed benzodiazepine pharmacotherapy were identified using VA CDW EHR data: adjustment disorders, mood disorders, and anxiety. Second, the drug class most prevalent (based on VA drug class) in individuals with these diagnoses was identified: antidepressants. In addition, based on committee clinical expertise, CPGs, and existing literature, benzodiazepine pharmacotherapies are also used to treat sleep disorders, such as insomnia (Chun et al., 2021; VA/DoD, 2019; Sateia et al., 2017; Lie et al., 2015; Shepardson et al., 2014; Wilson and Nutt, 2007). Through this approach, alternative non-benzodiazepine pharmacotherapy (comparator)

___________________

⁹ The committee uses the term “pharmacotherapy” throughout the report instead of “medications,” “prescriptions,” or “treatment,” especially when referencing the committee’s analyses and results.

included SNRIs, TeCAs, and anxiolytics pharmacotherapies (included in the VHA drug class of antidepressants) and insomnia pharmacotherapies.

Study 3 Outcomes

The outcomes outlined in the statement of task are all-cause mortality and suicide mortality, which were identified through linked data from the NDI. The committee determined all-cause mortality as the primary outcome and suicide mortality as the secondary outcome. Suicide mortality is identified by the ICD-10 death codes X60-X84 (intentional self-harm), U03 (intentional self-harm (suicide)), and Y87.0 (sequelae of intentional self-harm) (Hirsch et al., 2016; CDC, 2002).

Study 3 Causal Contrast: Intent-to-Treat Effect

Adhering to the statement of task and the causal question, this analysis focused on the intent-to-treat (ITT) effect, which corresponds to the effect of being assigned to a treatment strategy (e.g., initiating benzodiazepine pharmacotherapy versus alternative non-benzodiazepine pharmacotherapy) at baseline, irrespective of subsequent cross-over or treatment discontinuation during follow-up. Treatment “adherence” is not an issue in this study, as the committee is only interested in the initiation of a pharmacotherapy, defined as a single, newly dispensed dose.

In ITT, only baseline weights were created. For ITT analysis, the inverse probability was a function of only the baseline covariates (Hernán and Robins, 2017; Robins et al., 2000). The general assumptions of this approach are the same as other target trial emulations–if there are no unmeasured confounders, no measurement error, and no model misspecification, the ITT analysis are unbiased estimates of causal effects (see chapter 3 for a detailed description of all the assumptions associated with this approach).

Study 3 Start and End of Follow-Up

In this study, the start date, or the index date, is when the veteran was first dispensed the qualifying ≥1 tablet supply of the benzodiazepine or alternative non-benzodiazepine pharmacotherapy and all baseline eligibility criteria were met. To facilitate computation, the index date was transformed into an index month, as the analyses were conducted using aggregated monthly data. Based on the study design that considered a 12-month pre-index period and a 3-month follow-up period, January 1, 2007, was the first eligible index date and October 1, 2019, was the last eligible index date. The committee utilized all available follow-up data (through the end of 2019) with prespecified landmark analysis at 3 months.

The committee selected a 3-month follow-up to estimate the acute effect of concomitantly dispensed opioid and benzodiazepine pharmacotherapy on mortality. For ITT, eligible individuals are followed from the time of treatment initiation (baseline) until death (or death due other causes, in analyses of suicide mortality) or the administrative end of follow-up (3 months), whichever happens first. All episodes per eligible veteran were selected for inclusion.

Study 3 Design

The study design was an active-comparator, new-user (Lund et al., 2015; Ray, 2003), retrospective cohort target trial emulation study (Hernán and Robins, 2016). The emulated trial using observational data was conducted to estimate the relative effect of being newly dispensed benzodiazepine pharmacotherapy versus alternative non-benzodiazepine pharmacotherapy on mortality for a follow-up period of 3 months.

Study 3 Covariates

To strengthen causal inferences, the committee controlled for many covariates to adjust for baseline confounding and achieve balance between the treatment and comparator groups (see chapter 3 for more details):

• Sociodemographics,
• Health conditions and behaviors,
• Supplemental health insurance to VHA coverage,
• Health care utilization,
• Specific pharmacotherapies,
• Facility,
• Facility-level characteristic, and
• Calendar month of eligible dispensed pharmacotherapy.

Additional covariates include month and the interaction of state and year as fixed effects. Baseline covariates included demographic characteristics (e.g., age, sex, race/ethnicity, marital status, disability status), supplemental insurance to the VHA coverage (e.g., Medicare, TRICARE, or private insurance), housing security (e.g., history of homelessness status), physical and mental health conditions/disorders (e.g., acute painful injury, anxiety, chronic obstructive pulmonary disease, cardiovascular disease, substance use disorders, diabetes), military status (e.g., service era, military branch, combat service), body mass index, health care utilization in prior year (e.g., VHA outpatient visits, inpatient hospital admissions, nursing home stay), VHA facility-level characteristic (e.g., urbanicity), and specific pharmacotherapies that the committee considered as potential confounders (such as migraine medications; see Table 3-4 and Table 6-2 for the detailed list). The committee evaluated candidate variables readily available in the study clinical and administrative databases. See Appendix I for health conditions and corresponding ICD codes.

The studies did not require participants to have a chronic pain diagnosis in the primary analysis. Studies have shown that individuals identified with chronic pain did not have a pain diagnosis or pain was underreported in electronic medical records compared to pain reported in a patient-based survey (Frank et al., 2019; Goulet et al., 2016) The committee controlled for a number of baseline confounders in this study. Average pain intensity score was measured solely at baseline. Given the insomnia pharmacotherapy was included in the comparator group, additional specific sleep disorders (circadian rhythm sleep disorders; hypersomnia; isolated symptoms, apparently normal variants, and unresolved issues; other psychiatric/behavioral disorders frequently encountered in the differential diagnosis of sleep disorders; other sleep disorders; parasomnia; sleep disorders associated with conditions classifiable elsewhere; and sleep-related movement disorders) were included as covariates.

Lastly, based on a priori hypotheses that effects may vary by specific individual subgroups, the committee identified a subset of variables to be evaluated for effect modification (see the “Subgroup Analyses” section) (Suda et al., 2023; Gellad et al., 2018).

Accounting for Secular Trends and Variation by Facility

Facility and calendar time (12 months) were captured as fixed effects to capture facility-specific variation (e.g., clustering of individuals within a facility, geographic variation by facility), and reflect changes over time (e.g., variation in behaviors (health care seeking, prescribing)) and prevalence of conditions during a year. The committee also included the interaction term (i.e., product term) between state and year to capture unmeasured variation over time and by states, such as when state or facility policies related to opioid prescription were implemented (e.g., when states began to participate in the Prescription Drug Monitoring Program) or secular trends in state-level policies and intensities during the opioid epidemic.

Statistical Analysis

The causal effect of study 3 is the ITT effect. In study 3, an unadjusted and adjusted pooled logistic regression model was used to estimate the death rates (per 100,000 person-years) as a measure of absolute risk and HRs for all-cause mortality and suicide mortality in the benzodiazepine pharmacotherapy group compared to the alternative non-benzodiazepine pharmacotherapy group (the reference). A Wald confidence interval (CI) was calculated for the overall HR. For the subgroup analyses, the Bonferroni correction was applied to calculate the 95 percent CIs and adjust for multiple comparisons. The survival curves for all-cause and suicide mortality over time for those newly

dispensed benzodiazepine pharmacotherapy compared to those newly dispensed alternative non-benzodiazepine pharmacotherapy were estimated.

All analyses were conducted using SAS 9.4 and SAS Enterprise Guide. The study results do not report data if sample sizes were 10 or fewer individuals. This helps to ensure confidentiality of individuals in the study and is in line with CMS data policy and National Center for Health Statistics (NCHS) recommendations (NCHS, 2023; HHS, 2020). Furthermore, cells are marked as unreliable where sample sizes were 20 or fewer.

Accounting for Baseline Differences: Inverse Probability Treatment Weights

To account for the baseline difference between treatment and comparator groups in each target trial, analyses were weighted using inverse probability treatment weights (IPTW) for studies 1 and 3. IPTW is a method used to adjust for covariates and potential confounding in observational studies that uses a function of the propensity score (PS) of receiving treatment based on individual factors. By weighting each individual included in the analysis by the inverse of the probability of receiving the treatment or comparator treatment helps achieve balance between the treatment and comparator groups (Xu et al., 2010; Robins, 1986).

To calculate the IPTW, the committee first performed a logistic regression to calculate the PS, or the likelihood of being the treatment versus comparator group. The PS, the dependent variable, was conditioned on the aforementioned covariates. For ITT analysis, the IPTW was a function of only the baseline covariates (Robins et al., 2000). Second, the IPTW was calculated using the inverse of the PS, or 1/PS, for those newly dispensed benzodiazepine pharmacotherapy and 1/(1-PS) for those newly dispensed alternative non-benzodiazepine pharmacotherapy group for baseline treatment strategies.

To limit the influence of extreme weights on the results, the committee winsorized weights to the middle 98 percent by applying the 1st percentile weight to all observations below the 1st percentile and the 99th percentile weight to all observations above the 99th percentile (Crump et al., 2009). The distribution of the PS by treatment strategy was depicted using histograms. In addition, the balance of baseline covariates between the two groups before and after IPTW was assessed using the absolute standardized mean difference (ASMD). An ASMD of ≤0.1 between the two groups after weighting suggests reduction of potential confounding, allowing for causal inference comparisons (Austin, 2009). Missing values were handled using the missing covariate indicator methods, which assigned a missing category in the model (e.g., Hispanic ethnicity, non-Hispanic ethnicity, missing ethnicity).

Descriptive Analyses

The committee reports the distributions of baseline demographic and clinical characteristics for the overall analytic sample before and after application of the IPTW weights.

Subgroup Analyses

Subgroup analyses were conducted by age (with two categorical comparisons: 18–64 versus ≥65 and 18–34, 35–54, 55–74, ≥75), race (White, Black, Native American/Alaskan Native, Asian, Hawaiian and Pacific Islander, more than one race, and missing), ethnicity (Hispanic, and non-Hispanic), sex (male and female), time period (2007–2012 versus 2014–2019), and by payment source of dispensed prescriptions (only Medicare Part D prescription claims, only VHA prescription claims, or both during the episode). The committee selected these subgroups to examine differences given older age groups, specific racial and ethnic groups, and males have been shown to be at increased risk for mortality (Bohnert and Ilgen, 2019; Bossarte et al., 2012). Age categories used in the national annual reports on veteran suicide prevention were included in the subgroup analysis for consistency (VA, 2018).

The committee also included time period to examine changes in mortality risk. The committee selected before or after 2013, given introduction of the VHA’s Opioid Safety Initiative that year. The committee stratified by payment sources of dispensed pharmacotherapy during an episode: individuals were grouped as having prescriptions dispensed from only the VHA, only Medicare Part D, or both. Payment source of dispensed pharmacotherapy is a proxy for care, access, and coordination.

Sensitivity Analyses

Sensitivity analyses were conducted to assess the robustness of the results. First, the period of follow-up for outcome measurement was extended from 3 to 12 months from the index date. Second, the number of dispensed tablets was increased from at least one to at least five, and thus the index date was redefined to the date when the individual was dispensed at least five tablets of a benzodiazepine or alternative non-benzodiazepine pharmacotherapy.

RESULTS

Among the 13,152,784 million unique veterans who received care in the VHA from 2007–2019, 637,793¹⁰ were consistently dispensed opioid pharmacotherapy and eligible for inclusion. The mean age and standard deviation (SD) age at first visit was 60.65 years (SD: 13.74). Most study participants were male (93.67 percent), White (71.89 percent), and non-Hispanic (89.98 percent). They accounted for 1,379,093 unique veteran episodes for a mean of 2.19 (1.84) episodes per veteran and 4,103,911 veteran-months of observation for a mean follow-up time of 2.98 (SD: 0.20) months per episode. The sociodemographic and clinical characteristics of the study sample before and after weighting are presented in Table 6-3. After weighting, the maximum ASMD was less than 0.1, suggesting good covariate balance (Figure 6-5).

From 2007 through 2019, 16,913 deaths from all causes and 303 by suicide mortality occurred in this cohort. Among deaths from all causes, 4,768 occurred during an episode involving benzodiazepine pharmacotherapy (N = 123,684), resulting in an adjusted all-cause mortality rate of 7,493 deaths per 100,000 person-years of observation. Conversely, 12,145 deaths occurred during an episode involving an alternative non-benzodiazepine pharmacotherapy (N = 565,481), resulting in an adjusted all-cause mortality rate of 4,392 deaths per 100,000 person-years. The adjusted HR for all-cause mortality was 1.71 (95% CI: 1.60–1.82; unadjusted HR: 2.72 [95% CI: 2.58–2.96]). See Table 6-4 for unadjusted and IPTW adjusted all-cause and suicide mortality results.

Among the deaths from suicide, 64 occurred during a veteran episode involving newly dispensed benzodiazepine pharmacotherapy (N = 123,684), resulting in an adjusted suicide mortality rate of 117.0 suicides per 100,000 person-years of observation. Conversely, 239 suicide deaths occurred during a veteran episode involving newly dispensed alternative non-benzodiazepine pharmacotherapy (N = 565,481), resulting in an adjusted suicide mortality rate of 82.5 deaths per 100,000 person-years. The adjusted HR for suicide mortality was 1.42 (95% CI: 0.87–2.31; unadjusted HR: 1.85 [95% CI: 0.87–2.31]). Survival curves demonstrate lower survival over time in those newly dispensed benzodiazepine pharmacotherapy compared to those newly dispensed alternative non-benzodiazepine pharmacotherapy (see Figure 6-3).

The subgroup results, when available, are consistent with the overall estimates for both all-cause mortality and suicide mortality (Tables 6-5 and 6-6). This suggests that the primary results were not driven by a particular subgroup and that the risk across subgroups is similar. Consistent with both the CMS data policy and NCHS recommendations, the results in Tables 6-5 and 6-6 do not report estimates if sample sizes were fewer than 10 individuals, and estimates are marked as unreliable where sample sizes were fewer than 20.

Sensitivity Analyses

Two sensitivity analyses were run: (1) the follow-up time of the primary analysis extended to 12 months and (2) the index of exposure to benzodiazepine or alternative non-benzodiazepine pharmacotherapy increased to at least five tablets. The results were highly consistent (see Table 6-5).

The sensitivity analysis examining the effect of ≥5 tablets resulted in consistent results for all-cause mortality as the main analysis examining the effect of ≥1 tablets. Even when exposure to the treatment was more restrictive (e.g., required to have a higher number of tablets (≥5) compared at least one tablet), all-cause mortality results

___________________

¹⁰ The total number of unique veterans in study 3 is 637,793. During the study period (2007–2019), a veteran could enter the study more than one time and contributed more than 1 episode. Thus, the number of veterans in the benzodiazepine group (N = 123,684) and alternative non-benzodiazepine comparator (N = 565,481) will not sum to 637,793.

TABLE 6-3 Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, Baseline Characteristics of Veterans Included in the Current Analysis by Newly Dispensed Pharmacotherapy Group (Benzodiazepine Pharmacotherapy Versus Alternative Non-Benzodiazepine Pharmacotherapy) Before and After Inverse Probability Treatment Weights (IPTW) (2007–2019)

¹ Absolute standardized mean difference.

² Chronic obstructive pulmonary disease.

NOTE: No cell sizes of 10 or less are included in the reported results.

TABLE 6-4 Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, All-Cause Mortality and Suicide Mortality Rates and Hazard Ratios in those Newly Dispensed Benzodiazepine Pharmacotherapy Versus Alternative Non-Benzodiazepine Pharmacotherapy at 3 Months (2007–2019) (veterans: N = 637,793, veteran episodes: n = 1,379,093)

* N = 123,684; ** N = 565,481.

NOTES: CI = confidence interval; N = unique veterans; n = veteran episodes. No cell sizes of 10 or less are included in the reported results.

TABLE 6-5 Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, Inverse Probability Treatment Weighting Death Rates and Hazard Ratios for the All-Cause Mortality and Suicide Mortality Overall at 3 Months with Any Amount of Benzodiazepine and Sensitivity Analysis Outcomes (Those Dispensed at Least Five Tablets of Benzodiazepines Versus Alternative Non-Benzodiazepine Over 3 Months or Those Dispensed at Least One Tablet With an Extended Follow-Up (12 months) (2007–2019)

NOTES: No cells are reported representing 10 or fewer veterans. Overall indicates: ≥1 pills and 3-month follow-up.

were consistent, though less precise with wider confidence intervals due to a smaller sample size. Similarly, the suicide mortality results when analysis adjusted to ≥5 tablets was elevated.

When the analysis follow-up period was extended to 12 months, the all-cause mortality result was consistent but attenuated to 1.31 (95% CI: 1.28–1.35). The 12-month follow-up analysis for suicide mortality was consistent with the primary result at 3 months; (aHR: 1.46; 95% CI: 1.19–1.80).

See supplementary Tables 6-6, 6-7, and 6-8 and Figures 6-4, 6-5, 6-6, 6-7, and 6-8 and for further details.

DISCUSSION

Study 3 examined the effect on all-cause mortality and suicide mortality of newly dispensed benzodiazepine pharmacotherapy on veterans receiving care at the VHA. The study was conducted in response to the statement of task to examine (c) the effect of co-prescribing a benzodiazepine among individuals already receiving opioids, relative to alternative treatments for anxiety and other indications for benzodiazepines.

All-Cause Mortality and Suicide Mortality

The study findings are consistent with the literature on the risks of co-prescribing benzodiazepines with opioids. Co-prescribing is known to contribute to falls, injuries, and overdose as well as contribute to the suppression of respiratory drive.

Although benzodiazepine pharmacotherapy is effective in the treatment of several conditions, it is associated with risk of misuse, dependence, overdose, and mortality. For example, Maust and colleagues (2019) reported that of the 30.6 million U.S. adults reporting benzodiazepine use in the past year, with 17.2 percent of use reported as misuse.¹¹ Long-term benzodiazepine use increases the risk for benzodiazepine use disorder and dependence, which

___________________

¹¹ Misuse in Maust and colleages (2019) referenced the pre-2015 NSDUH definition of “nonmedical use,” but the 2015 definition was revised to include “in any way a doctor did not direct.”

is known to lead to harmful psychological and physical effects (Hirschtritt et al., 2021; Schweizer and Rickels, 1998). Furthermore, withdrawal from benzodiazepine use can result in severe, even life-threatening, symptoms (Edinoff et al., 2021; Ashton, 2005; Schweizer and Rickels, 1998; Pétursson, 1994; Ashton, 1991). Approximately 40 percent of individuals who abruptly ended long-term use (6 months or more) of long-acting benzodiazepines reported moderate to severe withdrawal (Hood et al., 2012; Murphy and Tyrer, 1988). Benzodiazepines can interact with alcohol and other psychoactive medications, especially those that act as central nervous system depressants, increasing the risk of adverse effects (Rawy et al., 2024; Hayes et al., 2020; Day, 2014; Neutel and Patten, 1997).

The committee evaluated the results based on several Bradford Hill criteria: biological plausibility, consistency of the association, time sequence, specificity, strength of the association, study design, quantitative strength, and dose response (Hill, 1965).

Biologically, both opioid and benzodiazepine pharmacotherapies are demonstrated to lead to respiratory depression, hypoxia, hypercapnia, which can cause cardiorespiratory arrest and eventually death. The study design follows individuals from their start date or entry into the study, based on their treatment strategies, and observes them over 3 months until when the outcome of interest, all-cause mortality or suicide mortality, occurs. Further, the results found were consistent across subgroups and sensitivity analyses. When the follow-up period was extended from 3 to 12 months, the aHR for all-cause mortality was attenuated. The committee’s findings are consistent with studies of different design and time period, such as by Xu and colleagues (2020), who found similar results when comparing groups receiving both benzodiazepines and opioids versus those who received neither (aHR: 2.04; 95% CI: 1.65–2.52). Further, Gaither and colleagues (2016) found that concomitant opioid and benzodiazepine prescribing contributed to an increased risk in a propensity-matched sample (aHR: 1.39; 95% CI: 1.12–1.66). Weisberg and colleagues (2015) found that those receiving both long-term opioid and benzodiazepine prescriptions had a higher risk for all-cause mortality (aHR: 1.51; 95% CI 1.22–1.87) compared to those who received neither pharmacotherapy.

The strength of association reflects quantitative strength, study design, and dose response. The study design emulates a randomized trial using observational data and reduces the confounders and biases inherent in observational studies. Increasing exposure from at least one tablet to at least five tablets demonstrated similar results for all-cause mortality and suicide mortality at 3 months, suggesting minimal dose response within a narrow range of exposure. As for the study design, the committee applied the target trial emulation framework, reflecting current best practices in the causal inference methodology field when using observational data.

These findings are consistent with concerns in guidelines that indicate avoidance of concurrent use of opioid and benzodiazepine pharmacotherapy (VA/DoD, 2017, 2022; Dowell et al 2016). In addition, guidelines recommending against concurrent opioid and benzodiazepine pharmacotherapy prescribing are complemented by the decrease of co-prescribing in recent years—from 31.2 per 100 persons in 2017 to 25.8 per 100 persons in 2021.¹² It is important to note that the committee’s interpretation of the results is not intended to influence restriction for any medication nor is not to be viewed as an absolute contraindication for the clinical use of benzodiazepines with opioids.

Limitations

The committee attempted to enable sufficient control for confounders; however, there remains a risk of confounding by indication (e.g., comparator group is not a sufficient comparator). Furthermore, the committee is unable to reasonably control for clinical decision making and confounding by indication. This includes the acknowledgment of risk, veteran informed consent, and the decision to use benzodiazepines anyway, perhaps especially in at-risk individuals.

Additionally, there remains an underlying acknowledgment of the reliability of medical examiner attributions regarding intentional versus non-intentional suicide. There is also the uncontrollable factor of the contribution of specific pharmacotherapies and substances in the context of multiple agents being present at the time of death.

___________________

¹² Dr. Christopher Jones. Presentation to the committee on March 9, 2023. Trends in Opioid and Benzodiazepine Use, Misuse, and Overdose.

The committee notes the sample does not include VHA veterans who are receiving care in the private sector other than Medicare (e.g., commercial insurance or Medicaid) and the limitation this poses to the generalizability of the results.

See Chapter 3 for additional study 3 limitations.

FINDINGS

The committee was tasked with quantifying the effects of opioid and benzodiazepine prescribing on the risk of death among veterans who received care from the VHA between 2007–2019. Study 3 effectively addresses the objective outlined in part (c) the effect of co-prescribing a benzodiazepine among patients already receiving opioids, relative to alternative treatments for anxiety and other indications for benzodiazepines.

Finding 3-1. Among veterans consistently dispensed opioid pharmacotherapy, the 3-month risk for all-cause mortality was 7,493 deaths per 100,000 person-years for those newly dispensed benzodiazepine pharmacotherapy and 4,392 per 100,000 person-years for those newly dispensed alternative non-benzodiazepine pharmacotherapy (aHR: 1.71, 95% CI: 1.60–1.82).¹³

Finding 3-2. Among veterans consistently dispensed opioid pharmacotherapy, the 3-month risk for suicide mortality was 117 deaths per 100,000 person-years for those newly dispensed benzodiazepine pharmacotherapy and 83 deaths per 100,000 person-years for those newly dispensed alternative non-benzodiazepine pharmacotherapy (aHR: 1.42, 95% CI: 0.87–2.31).¹⁴

CONCLUSIONS

The committee evaluated the effect of newly dispensed benzodiazepine pharmacotherapy on all-cause mortality and suicide mortality. The studies focused on effects of co-prescribing of opioids and benzodiazepine pharmacotherapies. The committee employed the TTE framework to reduce biases, such as due to immortal time, and confounding. In addition, the committee leveraged the large observational datasets (e.g., health care, clinical, and administrative data) available within the VA and VHA, supplemented by CMS data, and the NDI for mortality data. Based on the study findings, the committee makes the following conclusions for veterans who received care from the VHA between 2007–2019:

Conclusion 2. The committee concludes that there is an increased risk of all-cause mortality among veterans co-prescribed opioid and benzodiazepine pharmacotherapies compared to those co-prescribed opioid and alternative non-benzodiazepine pharmacotherapies.

Conclusion 7. The committee concludes that there is some evidence of a higher risk of suicide mortality among veterans co-prescribed opioid and benzodiazepine pharmacotherapy, although the estimate is imprecise at 3 months.

___________________

¹³ Alternative non-benzodiazepine pharmacotherapy can be used to treat conditions treated by benzodiazepines, such as anxiety.

¹⁴ Alternative non-benzodiazepine pharmacotherapy can be used to treat conditions treated by benzodiazepines, such as anxiety.

ANNEX: SUPPLEMENTARY FIGURES AND TABLES

TABLE 6-6 Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, Missingness of Variables Included in the Current Analysis by Total, By Those Newly Dispensed Benzodiazepine Pharmacotherapy or Alternative Non-Benzodiazepine Pharmacotherapy.

¹ Rx: prescription

NOTE: No cells are reported representing 10 or fewer veterans.

TABLE 6-7 Subgroup and Sensitivity Results—All-Cause Mortality. Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, Adjusted All-Cause Mortality Rates and Hazard Ratios Between Individuals Newly Dispensed Benzodiazepine Pharmacotherapy Compared to Those Newly Dispensed Alternative Non-Benzodiazepine Pharmacotherapy.

* Indicates cell sizes less than 20 veterans and/or suppressed to comply with privacy requirements.

NOTES: No cells are reported representing 10 or fewer veterans. CI = confidence interval; CMS = Centers for Medicare & Medicaid Services; VHA = Veterans Health Administration.

TABLE 6-8 Subgroup and Sensitivity Results—Suicide Mortality. Among VHA Veterans Consistently Dispensed Opioid Pharmacotherapy, Adjusted Suicide Mortality Rates and Hazard Ratios Between Individuals Newly Dispensed Benzodiazepine Pharmacotherapy Compared to Those Newly Dispensed Alternative Non-Benzodiazepine Pharmacotherapy.

* Suppressed to comply with privacy requirements

NOTES: No cells are reported representing 10 or fewer veterans. CI = confidence interval; CMS = Centers for Medicare & Medicaid Services; VHA = Veterans Health Administration.

REFERENCES

This page intentionally left blank.