Abstract
Objective
Although depression and chronic pain often coexist, few studies have examined antidepressant use among people with pain. This study examines the prevalence and characteristics associated with antidepressant use among people prescribed opioids for chronic noncancer pain (CNCP).
Design
Baseline data from a prospective cohort study.
Setting
Australian community.
Subjects
A total of 1166 people prescribed opioids for CNCP.
Methods
Baseline data collection consisted of a self-completed seven-day medication diary and telephone interview to collect information on sociodemographic characteristics and mental/physical health using validated questionnaires. Logistic regression was used to examine characteristics associated with antidepressant use, reporting adjusted odds ratios (AORs) and 95% confidence intervals (CIs).
Results
Of the 1166 participants, 668 (57.3%) were female, and the median (interquartile range) age was 59 (49–68) years. About half the cohort (N = 637, 54.6%) used antidepressants. Of these, 329 (51.7%) reported moderate to severe depression. Amitriptyline was the most commonly used antidepressant (17.3%). Factors independently associated with antidepressant use were being female (AOR = 1.47, 95% CI = 1.13–1.92), more years lived in pain (AOR = 1.01, 95% CI = 1.00–1.02), and use of nonopioid analgesics (AOR = 1.34, 95% CI = 1.01–1.78), benzodiazepines and related drugs (AOR = 1.84, 95% CI = 1.36–2.49), antiepileptics (AOR = 1.86, 95% CI = 1.38–2.51), and antipsychotics (AOR = 2.15, 95% CI = 1.22–3.77).
Conclusions
Antidepressant use is common among people with CNCP prescribed opioids. Those using antidepressants were more likely to use other psychotropic medicines concurrently, highlighting that they are a high-risk population requiring comprehensive assessment to optimize outcomes and reduce potential harms from polypharmacy.
Keywords: (MeSH): Analgesics, Opioids, Antidepressants, Chronic Pain, Drug Utilization
Introduction
Antidepressant medicines have traditionally been the mainstay of therapy for the management of moderate to severe depression. However, selected antidepressants are now increasingly used for a number of other indications, such as anxiety and panic disorders, and for neuropathic pain [1]. It has been estimated that antidepressants account for about two-thirds of psychotropic medicines dispensed annually in Australia [2]. Furthermore, in 2011, selective serotonin reuptake inhibitors (SSRIs) comprised 59% of all antidepressant dispensings nationally, serotonin and noradrenaline reuptake inhibitors (SNRIs) comprised 25%, and tricyclic antidepressants (TCAs) comprised 7% [2]. Although some antidepressants are recommended for different indications (e.g., TCAs for neuropathic pain and SSRIs for anxiety disorders) [3], there is limited information on their patterns of use in pain populations.
Characterizing the use of antidepressant medicines in chronic noncancer pain (CNCP) is important because there is strong evidence that depression and chronic pain coexist [4,5]. Our previous work on an Australian cohort of people with CNCP who had been prescribed opioids found that approximately half reported current moderate to severe depression [6]. Given that an estimated 15–20% of Australians experience chronic pain in a given year [7,8], understanding this relationship is central to optimizing treatment outcomes for both depression and pain. Chronic pain patients prescribed higher doses of opioids have been found to be more likely to report higher pain severity, symptoms of depression, and psychosocial problems [9]. The presence of pain has also been identified as a barrier to achieving adequate treatment responses among people with depression [10]. Furthermore, suicide is a common cause of opioid-related mortality [11,12].
Understanding how antidepressants are used at an individual level among people with pain, particularly among those who are prescribed opioids, is important to minimizing harms. For example, guidelines for managing neuropathic pain emphasize that it may be necessary to combine TCAs, opioids, and antiepileptics [1], but the concurrent use of multiple central nervous system (CNS) depressants also increases the risks of adverse events such as excessive sedation and respiratory depression [13]. In light of limited information on the use of antidepressants among people with pain and the challenges in managing pain and depression, the overall aim of this study was to undertake a detailed examination of their use in an Australian cohort of people who had been prescribed opioids for CNCP. The specific objectives were 1) to examine the prevalence of antidepressant use, 2) to examine the types and doses of antidepressants used, and 3) to determine the individual characteristics associated with antidepressant use.
Methods
Study Design
This study uses baseline data from a subset of 1,166 participants who completed a self-reported medication diary, collected as part of a larger study entitled the Pain and Opioids IN Treatment (POINT) study. The POINT study is a prospective national cohort study of 1,514 individuals in Australia taking pharmaceutical opioids for CNCP. The full study protocol and eligibility criteria have been published elsewhere [14]. In summary, community pharmacists referred patients who had been experiencing chronic pain for at least three months and who were prescribed Schedule 8 opioids (buprenorphine, codeine [as a single ingredient product], fentanyl, methadone, morphine, oxycodone, or hydromorphone) for longer than six weeks (and hence were less likely to be using opioids for an acute indication product). In Australia, a medicine is classified as a controlled (Schedule 8) drug if it meets the following two criteria: 1) The medicine is included in Schedule I or II of the United Nations Single Convention on Narcotic Drugs 1961 or in Schedule II or III of the United Nations Convention on Psychotropic Substances 1971; and 2) The medicine has a high-risk of dependency, misuse, abuse, or illicit use. Patients were not eligible if they were under 18 years of age or currently prescribed opioids for cancer or receiving opioid substitution therapy for heroin dependence. Telephone interviews to work through the questionnaires with participants were conducted by trained interviewers who had a minimum three-year health or psychology degree and had been trained in suicide intervention. Interviewers were provided with glossaries of chronic pain conditions and medicines. Additionally, a self-complete questionnaire was mailed to participants, which included a medicine diary to record all prescribed and over-the-counter medicines taken in the past seven days. Participants had the option of completing this via paper or over the telephone. All participants were reimbursed AUD40 for completing the baseline interview.
Demographics and Clinical Characteristics
Participants’ age, gender, employment status, highest level of education, marital status, and residential status (i.e., living alone or with others) were recorded. Net income was reported in 12 categories, and responses were dichotomized to two groups—more or less than AUD400 per week, consistent with the monetary value of unemployment or disability benefits.
Participants were asked about chronic pain conditions they had experienced in their lifetime and whether they had been problematic in the preceding 12 months. They were also asked to report on the number of years they had been experiencing chronic pain. Current pain severity and pain interference were measured by the Brief Pain Inventory [15] as a continuous score out of 10. Pain self-efficacy was assessed using the Pain Self-Efficacy Questionnaire [16,17].
The Short Form 12 (SF-12) Health Survey was developed from The Medical Outcomes Study: Measures of Quality of Life Survey (MOS), which measures physical function, mental health, and general well-being in patients with chronic conditions [18]. The SF-12 is based on population norms, with a mean score of 50 and a standard deviation (SD) of 10. A cutoff of two SDs below the population mean was used to indicate poor physical functioning in the current study.
Depression and generalized anxiety disorder (GAD) were identified using the Patient Health Questionnaire-9 (PHQ-9) and GAD-7 modules of the Pfizer Health Questionnaire [19]. Previously validated cutoffs were used for screening tools as follows: responses indicating moderate to severe depression were defined as a score of ≥10 on the PHQ-9 [20], responses indicating moderate to severe anxiety were defined as a score of ≥10 on the GAD-7 [21]. A score of ≥3 on the Primary Care PTSD screen (PC-PTSD) was used to indicate the presence of post-traumatic stress disorder (PTSD) [22].
Five questions concerning adverse childhood experiences were asked based on those of Sansone et al [23]. These questions, related to experiencing sexual abuse, physical neglect, witnessing violence, physical abuse, or emotional abuse as a child, and answers were combined into one variable to represent the reporting of any of these experiences.
The suicidality questions were based on questions used in the 2007 National Survey of Mental Health and Wellbeing [24]. These questions were adopted from the World Mental Health—Composite International Diagnostic Interview [25,26] suicidality module, which has been used extensively in previous international studies [27,28]. Participants were asked about lifetime and past 12-month presence of suicidal thoughts, plans, and attempts.
Lifetime and past 12-month alcohol and illicit drug use were recorded, and lifetime drug and alcohol use disorders (using ICD-10 dependence criteria) were assessed via the Composite International Diagnostic Interview alcohol and illicit drug use module [29]. Participants were also asked whether they had accessed the services of a psychologist or psychiatrist in the past 12 months.
Antidepressant and Other Medicine Use
Self-reported use of medicines was obtained from a comprehensive diary, from which names and doses of all current prescribed and over-the-counter medicines used over a seven-day period were obtained. Collection of medicine use information via this process meant that a participant’s medicine list was indicative of all medicines taken during this period rather than those prescribed or dispensed. Examples were given for each class of medicines examined to assist participants to recognize the medicines they had taken.
To take into account the daily cumulative exposure to multiple opioids, oral morphine equivalent (OME) total daily doses were estimated according to published guidelines [30,31]. First, a daily dose of each opioid medicine was calculated by dividing the total weekly amount taken by each participant by 7. All forms of each opioid were included (e.g., extended and immediate release of the same opioid). Each dose was then converted to OME milligrams, and where multiple opioids were taken, these were added together to determine the total OME dose used in milligrams per day.
Antidepressants were classified as medicines belonging to the World Health Organization’s Anatomical Therapeutic Chemical category N06 (psychoanaleptics) [32]. These were further classified into: N06AA Non-selective Monoamine Reuptake Inhibitors, which are commonly referred to as TCAs (amitriptyline, clomipramine, dothiepin/dosulepin, imipramine, nortriptyline), N06AB SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline), N06AF Non-selective Monoamine Oxidase Inhibitors (tranylcypromine), N06AG Monoamine Oxidase A Inhibitors (moclobemide), and N06AX Other Antidepressants (agomelatine, desvenlafaxine, duloxetine, mirtazapine, venlafaxine, reboxetine). Benzodiazepines and related drugs (“Z drugs”) were classified as medicines belonging to N05CD and N05CF. Antiepileptics were classified as medicines belonging to N03A, and antipsychotics were classified as medicines belonging to N05A. The concomitant use of opioids with various combinations of psychotropic medicines, including antidepressants, benzodiazepines, antipsychotics, and/or antiepileptics, was also examined using each participant’s seven-day medicine use diary.
Statistical Analyses
Data were analyzed using the Statistical Package for the Social Sciences, version 22 (IBM SPSS, Chicago, IL, USA) and STATA 15.1. Variables were summarized using means and standard deviations, medians and interquartile ranges (IQRs), and frequencies and percentages, as appropriate. Binary logistic regression was used to calculate unadjusted and adjusted odds ratios (AORs) and 95% confidence intervals (95% CIs) to compare characteristic differences between those who were prescribed antidepressants and those who were not. Variables significant at the P < 0.05 level in univariate analyses were subsequently included in a multivariable model. All variables were assessed for potential multicollinearity before inclusion in the multivariable model.
Ethical Approval
The study was approved by the Human Research Ethics Committee of UNSW Sydney, Australia (HREC reference: #HC12149).
Results
Sample Characteristics
Of the 1166 participants in the cohort, 668 (57.3%) were female, and the median (IQR) age was 59 (49–68) years. More than 80% were unemployed or not currently in the workforce (e.g., retirees). The majority (75.6%) of the cohort reported having back or neck problems in the previous 12 months, followed by 63.0% reporting arthritis or rheumatism and 28.0% reporting frequent or severe headaches. The median (IQR) number of years lived in pain was 10 (5–21). Almost half (44.4%) of the cohort screened positively for moderate to severe depression, 21.5% for GAD, and 14.4% for PTSD.
Antidepressant Use
Just over half of the cohort (54.6%, N = 637) was using antidepressants (Table 1). Of all participants in the cohort, 22.6% (N = 263) were using TCAs, 20.3% (N = 237) other antidepressants, and 17.0% (N = 198) SSRIs. Amitriptyline was the most commonly used antidepressant (17.3%, N = 202), followed by duloxetine (7.9%, N = 92) and venlafaxine (5.7%, N = 67).
Table 1.
Types of antidepressants and doses used
| Antidepressant | Frequency of Use |
Daily Dose | |
|---|---|---|---|
| N | % (95% CI) | Median (IQR) | |
| N06AA Non-selective Monoamine Reuptake Inhibitors (i.e., tricyclic antidepressants) | |||
| Amitriptyline | 202 | 17.3 (15.3–19.6) | 50 (55) |
| Clomipramine | 1 | 0.1 (0.0–0.6) | 75 (0) |
| Dothiepin (dosulepin) | 22 | 1.9 (1.2–2.9) | 87.5 (101.8) |
| Doxepin | 16 | 1.4 (0.8–2.2) | 50 (290 |
| Imipramine | 9 | 0.8 (0.4–1.5) | 50 (65) |
| Nortriptyline | 14 | 1.2 (0.7–2.0) | 22.5 (40) |
| Total | 263 | 22.6 (20.2–25.0) | – |
| N06AB Selective Serotonin Reuptake Inhibitors | |||
| Citalopram | 41 | 3.5 (2.6–4.7) | 20 (20) |
| Escitalopram | 43 | 3.7 (2.7–4.9) | 20 (10) |
| Fluoxetine | 36 | 3.1 (2.2–4.3) | 40 (20) |
| Fluvoxamine | 10 | 0.9 (0.5–1.6) | 200 (100) |
| Paroxetine | 16 | 1.4 (0.8–2.2) | 38.57 (20) |
| Sertraline | 52 | 4.5 (3.4–5.8) | 100 (100) |
| Total | 198 | 17.0 (14.9–19.2) | – |
| N06AF Non-selective Monoamine Oxidase Inhibitors | |||
| Tranylcypromine | 1 | 0.1 (0.0–0.6) | 10 (0) |
| N06AG Monamine Oxidase A Inhibitors | |||
| Moclobemide | 4 | 0.3 (0.1–0.9) | 375 (431.25) |
| N06AX Other Antidepressants | |||
| Agomelatine | 5 | 0.4 (0.2–1.0) | 25 (34.8) |
| Desvenlafaxine | 25 | 2.1 (1.5–3.2) | 100 (75) |
| Duloxetine | 92 | 7.9 (6.5–9.6) | 60 (60) |
| Mirtazapine | 55 | 4.7 (3.6–6.1) | 30 (15) |
| Venlafaxine | 67 | 5.7 (4.5–7.2) | 150 (75) |
| Reboxetine | 2 | 0.2 (0.0–0.6) | 8 (0) |
| Total | 237 | 20.3 (18.1–22.7) | – |
CI = confidence interval; IQR = interquartile range.
Concomitant Use of Other Psychotropic Medicines
The frequency of use of other psychotropic medicines in conjunction with opioids is presented in Table 2. Following antidepressants, use of benzodiazepines and related drugs (N = 327, 28.0%) and use of antiepileptics (N = 325, 27.9%) were most common. Almost one in five people used two psychotropic medicines in addition to opioids: antidepressants and benzodiazepines and related drugs (N = 223, 19.1%) and antidepressants and antiepileptics (N = 224, 19.2%).
Table 2.
Concomitant use of opioids and psychotropic medicines
| Medicine Combination | N* | % (95% CI) |
|---|---|---|
| Opioid + antidepressant | 637 | 54.6 (51.7–57.5) |
| Opioid + benzodiazepine† | 327 | 28.0 (25.5–30.7) |
| Opioid + antipsychotic | 103 | 8.8 (7.3–10.6) |
| Opioid + antiepileptic | 325 | 27.9 (25.4–30.5) |
| Opioid + antidepressant + benzodiazepine† | 223 | 19.1 (17.0–21.5) |
| Opioid + antidepressant + antipsychotic | 82 | 7.0 (5.7–8.7) |
| Opioid + antidepressant + antiepileptic | 224 | 19.2 (17.0–21.6) |
| Opioid + benzodiazepine† + antipsychotic | 48 | 4.1 (3.1–5.4) |
| Opioid + benzodiazepine† + antiepileptic | 98 | 8.4 (6.9–10.1) |
| Opioid + antipsychotic + antiepileptic | 46 | 3.9 (3.0–5.2) |
| Opioid + antidepressant + benzodiazepine† + antipsychotic + antiepileptic | 19 | 1.6 (1.0–2.5) |
Categories are not mutually exclusive.
Includes benzodiazepines and related drugs.
CI = confidence interval.
Characteristics of Antidepressant Users and Nonusers
Univariate comparisons of differences in characteristics between those who were and were not using antidepressants are presented in Table 3. In unadjusted analyses, individuals were more likely to be using an antidepressant if they had moderate to severe depression (OR = 1.92, 95% CI = 1.52–2.43), current GAD (OR = 1.95, 95% CI = 1.45–2.62), were female (OR = 1.39, 95% CI = 1.10–1.75), unemployed (OR = 1.19, 95% CI = 0.89–1.60), or had poor physical functioning (OR = 1.37, 95% CI = 1.07–1.76). Other factors associated with antidepressant use included having seen a psychologist (OR = 1.83, 95% CI = 1.43–2.35) or psychiatrist (OR = 2.03, 95% CI = 1.56–2.64) in the past 12 months and the concomitant use of antipsychotics (OR = 3.57, 95% CI = 2.18–5.86), benzodiazepines (OR = 2.20, 95% CI = 1.68–2.88), and antiepileptics (OR = 2.30, 95% CI = 1.75–3.01).
Table 3.
Characteristics of people who were and were not prescribed antidepressants
| Characteristic | Total Sample (N = 1166) | Prescribed ADs (N = 637) | Not Prescribed ADs (N = 529) | OR (95% CI) | P Value |
|---|---|---|---|---|---|
| Demographics | |||||
| % female (95% CI) | 57.3 (54.4–60.1) | 60.9 (35.4–42.9) | 47.1 (57.1–64.6) | 1.39 (1.10–1.75) | 0.006 |
| Median age (IQR), y | 59 (49–68) | 59 (48–67) | 60 (50–70) | 0.99 (0.98–1.00) | 0.025 |
| % did not complete year 10 (95% CI) | 20.0 (17.8–22.4) | 19.9 (17.0–23.2) | 20.0 (16.8–23.7) | 0.99 (0.74–1.33) | 0.966 |
| % living alone (95% CI) | 25.3 (22.9–27.9) | 24.5 (21.3–28.0) | 26.3 (22.7–30.2) | 0.91 (0.70–1.19) | 0.485 |
| % income below AUD400 per week (95% CI) | 54.7 (51.8–57.6) | 55.4 (51.5–59.2) | 53.9 (49.6–58.1) | 1.06 (0.84–1.34) | 0.599 |
| % unemployed/not currently in the workforce (i.e., retired) (95% CI) | 81.0 (78.7–83.2) | 82.3 (79.1–85.0) | 79.6 (75.9–82.8) | 1.19 (0.89–1.60) | 0.246 |
| % married/de facto (95% CI) | 54.8 (51.9–57.6) | 55.1 (51.2–58.9) | 54.4 (50.2–58.7) | 1.03 (0.81–1.29) | 0.822 |
| Clinical characteristics | |||||
| Median duration of pain (IQR), y | 10 (5–21) | 12 (5–22) | 10 (4–20) | 1.00 (1.00–1.00) | 0.002 |
| Mean pain severity (SD) | 5.0 (1.8) | 5.2 (1.7) | 4.8 (1.9) | 1.14 (1.07–1.21) | <0.001 |
| Mean pain interference (SD) | 5.6 (2.3) | 5.9 (2.2) | 5.2 (2.3) | 1.13 (1.07–1.19) | <0.001 |
| Median pain self-efficacy (IQR) | 30 (21–40) | 28 (18–38) | 32 (23–43) | 0.98 (0.97–0.98) | <0.001 |
| % poor physical functioning (SF-12 <30) (95% CI) | 69.6 (66.9–72.2) | 72.7 (69.1–76.0) | 66.0 (61.8–70.0) | 1.37 (1.07–1.76) | 0.013 |
| % current moderate to severe depression (95% CI) | 44.4 (41.6–47.3) | 51.7 (47.8–55.5) | 35.7 (31.7–39.9) | 1.92 (1.52–2.43) | <0.001 |
| % current generalized anxiety disorder (95% CI) | 21.5 (19.2–24.0) | 26.5 (23.2–30.0) | 15.6 (12.7–18.9) | 1.95 (1.45–2.62) | <0.001 |
| % post-traumatic stress disorder (95% CI) | 14.4 (12.5–16.5) | 17.4 (14.7–20.6) | 10.8 (8.4–13.7) | 1.75 (1.24–2.46) | 0.001 |
| % adverse childhood experience* (95% CI) | 37.4 (34.7–40.2) | 42.1 (38.3–45.9) | 31.8 (27.9–35.9) | 1.56 (1.23–1.99) | <0.008 |
| % lifetime suicide thoughts (95% CI) | 44.3 (41.4–47.1) | 51.6 (47.8–55.5) | 35.3 (31.4–39.5) | 1.95 (1.54–2.47) | <0.001 |
| % lifetime suicide attempt (95% CI) | 19.9 (17.7–22.3) | 24.7 (21.4–28.2) | 14.2 (11.5–17.4) | 1.98 (1.46–2.68) | <0.001 |
| % ICD-10 any drug dependence† (95% CI) | 28.6 (25.6–31.7) | 26.3 (22.4–30.7) | 31.1 (26.7–30.7) | 0.79 (0.58–1.07) | 0.131 |
| % ICD-10 lifetime alcohol use dependence (95% CI) | 16.5 (14.1–19.1) | 15.9 (12.9–19.5) | 17.1 (13.7–21.3) | 0.91 (0.64–1.31) | 0.623 |
| % seen psychiatrist in past 12 mo (95% CI) | 30.0 (27.5–32.7) | 36.6 (32.9–40.4) | 22.1 (18.8–25.9) | 2.03 (1.56–2.64) | <0.001 |
| % seen psychologist in past 12 mo (95% CI) | 34.6 (31.9–37.3) | 40.7 (36.9–44.5) | 27.2 (23.6–31.2) | 1.83 (1.43–2.35) | <0.001 |
| % pain condition in past 12 mo (95% CI) | |||||
| Back or neck problems | 75.6 (73.1–78.0) | 76.9 (73.5–80.0) | 74.1 (70.2–77.7) | 1.16 (0.89–1.52) | 0.264 |
| Arthritis or rheumatism | 63.0 (60.2–65.8) | 61.8 (58.0–65.6) | 64.5 (60.3–68.4) | 0.89 (0.70–1.14) | 0.358 |
| Frequent severe/headaches | 28.0 (25.5–30.6) | 32.3 (28.8–36.1) | 22.7 (19.3–26.5) | 1.63 (1.25–2.12) | <0.001 |
| Visceral pain | 21.9 (19.6–24.3) | 20.9 (17.9–24.2) | 23.1 (19.7–26.9) | 0.88 (0.67–1.16) | 0.370 |
| Fibromyalgia | 5.5 (4.3–7.0) | 7.1 (5.3–9.3) | 3.6 (2.3–5.6) | 2.04 (1.18–3.53) | 0.011 |
| Mean No. of chronic conditions in last 12 mo (SD) | 2.3 (1.1) | 2.3 (1.1) | 2.2 (1.0) | 1.09 (0.97–1.21) | 0.143 |
| Concomitant medicine use in the past week | |||||
| Total opioid dose as OME mg/d, median (IQR) | 75 (36–150) | 75 (37.8–156.9) | 66.5 (30.4–120.0) | 1.00 (1.00–1.00) | 0.012 |
| % nonopioid analgesics (95% CI) | 56.3 (53.4–59.1) | 59.3 (55.5–63.1) | 52.6 (48.3–56.8) | 1.32 (1.04– 1.66) | 0.020 |
| % benzodiazepine and related drugs (95% CI) | 28.0 (25.5–30.7) | 35.0 (31.4–38.8) | 19.7 (16.5–23.3) | 2.20 (1.68–2.88) | <0.001 |
| % antiepileptics (95% CI) | 27.9 (25.4–30.5) | 35.2 (31.5–39.0) | 19.1 (16.0–22.7) | 2.30 (1.75–3.01) | <0.001 |
| % antipsychotics (95% CI) | 8.8 (7.3–10.6) | 12.9 (10.5–15.7) | 4.0 (2.6–6.0) | 3.57 (2.18–5.86) | <0.001 |
AD = antidepressant; CI = confidence interval; ICD-10 = International Classification of Diseases 10th Revision; IQR = interquartile range; OME = oral morphine equivalent; OR = odds ratio SD = standard deviation; SF-12 = Short Form 12 Health Survey.
Includes sexual abuse, physical neglect, witnessing violence, physical abuse, or emotional abuse.
Includes amphetamines, benzodiazepines, cannabis, cocaine, ecstasy, hallucinogens, heroin, inhalants, and pharmaceutical opioids.
bolded values indicate significance at P<0.05
The results of the adjusted analysis are presented in Table 4. Independent characteristics associated with antidepressant use included being female (AOR = 1.47, 95% CI = 1.13–1.92), more years lived in pain (AOR = 1.01, 95% CI = 1.00–1.00), and use of nonopioid analgesics (AOR = 1.34, 95% CI = 1.01–1.78), benzodiazepines and related drugs (AOR = 1.84, 95% CI = 1.36–2.49), antiepileptics (AOR = 1.86, 95% CI = 1.38–2.51), and antipsychotics (AOR = 2.15, 95% CI = 1.22–3.77).
Table 4.
Multivariable model of the factors associated with antidepressant use
| Characteristic | Adjusted OR | 95% CI | P Value |
|---|---|---|---|
| Demographics | |||
| Female | 1.47 | 1.13–1.92 | 0.004 |
| Age | 1.01 | 1.00–1.02 | 0.224 |
| Unemployed/not in labor force | 1.01 | 0.70–1.44 | 0.973 |
| Clinical characteristics | |||
| No. of years lived in pain | 1.01 | 1.00–1.02 | 0.028 |
| Pain severity score | 1.05 | 0.96–1.15 | 0.263 |
| Pain interference score | 0.95 | 0.87–1.04 | 0.299 |
| Poor pain self-efficacy score | 0.99 | 0.98–1.00 | 0.120 |
| Poor physical functioning (SF-12 <30) | 1.23 | 0.91–1.66 | 0.182 |
| Current moderate to severe depression | 1.24 | 0.89–1.73 | 0.208 |
| Current moderate to severe anxiety | 1.11 | 0.75–1.64 | 0.601 |
| Lifetime suicidal thoughts | 1.34 | 0.96–1.87 | 0.082 |
| Lifetime suicide attempt | 1.02 | 0.68–1.54 | 0.906 |
| Post-traumatic stress disorder | 0.97 | 0.64–1.47 | 0.869 |
| Adverse childhood experience∗ | 1.00 | 0.75–1.32 | 0.974 |
| Frequent/severe headaches | 1.22 | 0.90–1.66 | 0.205 |
| Fibromyalgia | 1.24 | 0.69–2.25 | 0.473 |
| Seen psychiatrist (past 12 mo) | 1.32 | 0.94–1.84 | 0.106 |
| Seen psychologist (past 12 mo) | 1.25 | 0.92–1.70 | 0.153 |
| Concomitant medicine use in the past week | |||
| Total opioid dose as OME mg/d | 1.00 | 1.00–1.00 | 0.181 |
| Nonopioid analgesics | 1.34 | 1.01–1.78 | 0.046 |
| Benzodiazepines and related drugs | 1.84 | 1.36–2.49 | <0.001 |
| Antiepileptics | 1.86 | 1.38–2.51 | <0.001 |
| Antipsychotics | 2.15 | 1.22–3.77 | 0.008 |
Includes sexual abuse, physical neglect, witnessing violence, physical abuse, or emotional abuse.
CI = confidence interval; OME = oral morphine equivalent; OR = odds ratio; SF-12 = Short Form 12 Health Survey.
bolded values indicate significance at P<0.05
Discussion
This study presents the first detailed examination of antidepressant use by Australians who have been prescribed opioids to manage CNCP. The prevalence of antidepressant use was 55%, and antidepressant use was more likely among those who have been living with pain for a longer period of time and those who were using various other psychotropic medicines.
The most commonly used antidepressants in the cohort (amitriptyline, duloxetine, and venlafaxine) also represent those with an evidence base for use in the management of pain [1,33]. This suggests that the likely reason for use in the cohort was as an adjuvant medicine for pain. Current Australian treatment guidelines recommend amitriptyline as first-line therapy and duloxetine as third-line for the management of neuropathic pain [1]. Although guidelines for neuropathic pain acknowledge that combination therapy with TCAs, opioids, and antiepileptics may be needed [1], there is limited evidence about the effectiveness of this treatment approach [33]. It is also not without risks [1,34]. For example, amitriptyline and opioids are both CNS depressants, and their concomitant use increases the potential for excess sedation and respiratory depression, particularly in vulnerable groups such as older adults [13]. Given the high prevalence of concurrent use of amitriptyline and opioids in this population, further research is needed to investigate the longer-term safety and effectiveness of this combination, particularly in view of studies of cancer pain that suggest that there may be only marginal improvements in pain and higher risks of adverse events [35].
Use of other psychotropic medicines was more common among those who used antidepressants than those who did not, including benzodiazepines and related drugs (35% vs 20%) and antipsychotics (13% vs 4%). The use of antipsychotics in CNCP remains controversial [36]. One Cochrane review concluded that there was some evidence of their efficacy in patients with acute or chronic pain, but the studies were too heterogeneous to make recommendations beyond ‘more research is needed’ [37]. A more recent review of antipsychotics in fibromyalgia concluded that there was very low-quality evidence for the time-limited use (four to 12 weeks) of quetiapine to reduce pain, sleep problems, depression, and anxiety in fibromyalgia patients with major depression, but its benefits need to be balanced against adverse events such as weight gain [38].
Polypharmacy is common among people with CNCP, and the challenge in managing depression in this population has been documented [39]. As seen in this [40] and other cohorts [41], concomitant use of benzodiazepines is common in people using opioids for CNCP. As described earlier, combining opioids with other psychotropic medicines may place individuals at increased risk of adverse events. In our study, one in five participants was using a combination of opioids with antidepressants and benzodiazepines, or with antidepressants and antiepileptics, highlighting the extent of psychotropic polypharmacy in this cohort. This is of particular concern given US evidence of increased emergency department visits [42] and overdose deaths involving combinations of opioid analgesics and benzodiazepines [42,43]. Data from coronial studies of fatal opioid overdoses in Australia support this finding, with concomitant benzodiazepine and antidepressant use identified in 45–69% and 26–41% of fatal oxycodone and fentanyl overdoses, respectively [44,45].
The high prevalence of polypharmacy identified in our study also highlights that there is a need to establish whether clinicians are aware of the complete range of medicines that patients are prescribed. Although we were unable to assess the appropriateness of medicine use in our cohort, there may be a potential role for de-prescribing to reduce unnecessary polypharmacy, particularly among older adults [46]. It is also important to consider the role of nonpharmacological approaches to treat mental health disorders such as depression and anxiety. For example, pharmacological therapies are generally considered second-line treatments for mild to moderate depression in most clinical guidelines (e.g.[3,47]), after psychological and other psychosocial interventions. It is therefore important to understand whether current practice reflects these recommendations. Moreover, psychological and other interventions are also strongly recommended for the management of chronic pain [1]. Our findings showed that those using antidepressants were more likely to receive care from psychologists and psychiatrists. Comprehensive mental health care plans are publicly subsidized in Australia for eligible patients, so there is scope to increase access to these services. However, in the long term, the financial burden associated with accessing these services tends to be greater than with pharmacotherapies.
Strengths and Limitations
This study examined patterns of antidepressant use among people with CNCP who have been prescribed opioids. As antidepressants have multiple indications, it is difficult to ascertain the exact indication for their use from the data collected. Furthermore, we were unable to distinguish participants with neuropathic pain based on the information collected. Accordingly, the appropriateness of antidepressant use could not be assessed in the study. Data in this study primarily relied on participant self-report, which is subject to biases such as recall and social desirability [48]. Measures were taken to minimize these by reinforcing to participants the confidential nature of the interview and that all data would be de-identified [49]. Moreover, in using this form of data collection, the information collected on medicines use allowed for complete capture of medicines used rather than those prescribed or dispensed, and it included both prescription and nonprescription medicines. Given that this study focused on baseline antidepressant use and was cross-sectional in nature, causality could not be inferred. Further studies are needed to examine any effects of antidepressant prescribing in this patient group on pain outcomes and the impact on opioid use.
Conclusions
Antidepressant use was common among people prescribed opioids for CNCP, with amitriptyline being the most frequently used antidepressant in this population. The high prevalence of polypharmacy involving combinations of CNS depressants is of concern because it may place many individuals at risk of serious adverse events. The benefits of combining amitriptyline and opioids in the treatment of pain must be established in light of the risks of polypharmacy. Longitudinal studies are needed to examine the outcomes of antidepressant prescribing, particularly in older adults prescribed opioids, in order to establish the clinical benefits and adverse effects of these complex medication profiles.
Acknowledgments
The authors would like to thank Jessica Belcher, Bianca Hoban, Anika Martin, Ranira Moodley, Kimberly Smith, and Sarah Freckleton for their contribution to data collection; and Teleri Moore, Alana Garton, and Samantha Lynch for their assistance in data preparation and analysis. We also wish to acknowledge the contributions of project investigators Fiona Shand, Fiona Blyth, and Richard Mattick; and the POINT advisory committee for their advice on the design and conduct of the larger study.
References
- 1. Therapeutic Guidelines Limited. eTG Complete: Analgesic. North Melbourne: Therapeutic Guidelines Limited; 2018. [Google Scholar]
- 2. Stephenson C, Karanges E, McGregor I. Trends in the utilisation of psychotropic medications in Australia from 2000 to 2011. Aust N Z J Psychiatry 2013;471:74–87. [DOI] [PubMed] [Google Scholar]
- 3. Therapeutic Guidelines. eTG Complete: Psychotropic. Melbourne, Australia: Therapeutic Guidelines Ltd; 2018. [Google Scholar]
- 4. Fishbain DA, Cutler R, Rosomoff HL, Rosomoff RS. Chronic pain-associated depression: Antecedent or consequence of chronic pain? A review. Clin J Pain 1997;132:116–37. [DOI] [PubMed] [Google Scholar]
- 5. Kroenke K, Wu J, Bair MJ, et al. Reciprocal relationship between pain and depression: A 12-month longitudinal analysis in primary care. J Pain 2011;129:964–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Campbell G, Nielsen S, Bruno R, et al. The pain and opioids IN treatment study: Characteristics of a cohort using opioids to manage chronic non-cancer pain. Pain 2015;1562:231–42. [DOI] [PubMed] [Google Scholar]
- 7. Miller A, Sanderson K, Bruno R, Breslin M, Neil AL. The prevalence of pain and analgesia use in the Australian population: Findings from the 2011 to 2012 Australian National Health Survey. Pharmacoepidemiol Drug Saf 2017;2611:1403–10. [DOI] [PubMed] [Google Scholar]
- 8. Blyth FM, March LM, Brnabic AJM, et al. Chronic pain in Australia: A prevalence study. Pain 2001;89(2–3):127–34. [DOI] [PubMed] [Google Scholar]
- 9. Merril JO, Von Korff M, Banta-Green CJ, et al. Prescribed opioid difficulties, depression and opioid dose among chronic opioid therapy patients. Gen Hosp Psychiatry 2012;34:581–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Fishbain DA, Cole B, Lewis JE, Gao J. Does pain interfere with antidepressant depression treatment response and remission in patients with depression and pain? An evidence-based structured review. Pain Med 2014;159:1522–39. [DOI] [PubMed] [Google Scholar]
- 11. Braden JB, Edlund MJ, Sullivan MD. Suicide deaths with opioid poisoning in the United States: 1999–2014. Am J Public Health 2017;1073:421–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Roxburgh A, Bruno R, Larance B, Burns L. Prescription of opioid analgesics and related harms in Australia. Med J Aust 2011;1955:280–4. [DOI] [PubMed] [Google Scholar]
- 13. Riediger C, Schuster T, Barlinn K, et al. Adverse effects of antidepressants for chronic pain: A systematic review and meta-analysis. Front Neurol 2017;8:307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Campbell G, Mattick R, Bruno R, et al. Cohort protocol paper: The Pain and Opioids In Treatment (POINT) study. BMC Pharmacol Toxicol 2014;15:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Cleeland CS, Ryan K. The Brief Pain Inventory. Madison: Pain Research Group, WHO Collaborating Centre for Symptom Evaluation in Cancer Care; 1991. [Google Scholar]
- 16. Nicholas MK. The Pain Self-Efficacy Questionnaire: Taking pain into account. Eur J Pain 2007;112:153–63. [DOI] [PubMed] [Google Scholar]
- 17. Nicholas MK, Asghari A, Blyth FM. What do the numbers mean? Normative data in chronic pain measures. Pain 2008;134(1–2):158–73. [DOI] [PubMed] [Google Scholar]
- 18. Stewart A, Ware JE. Measuring Functioning and Well-Being: The Medical Outcomes Study Approach. Durham, NC: Duke University Press; 1992. [Google Scholar]
- 19. Kroenke K, Spitzer RL, Williams JBW, Lowe B. The Patient Health Questionnaire Somatic, Anxiety, and Depressive Symptom Scales: A systematic review. Gen Hosp Psychiatry 2010;324:345–59. [DOI] [PubMed] [Google Scholar]
- 20. Kroenke K, Spitzer R, Williams JBW. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med 2001;169:606–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Spitzer RL, Kroenke K, Williams JBW, Lowe B. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch Intern Med 2006;16610:1092–7. [DOI] [PubMed] [Google Scholar]
- 22. Prins A, Ouimette P, Kimerling R, et al. The Primary Care PTSD screen (PC-PTSD): Development and operating characteristics. Prim Care Psychiatry 2004;91:9–14. [Google Scholar]
- 23. Sansone RA, Whitecar P, Wiederman MW. The prevalence of childhood trauma among those seeking buprenorphine treatment. J Addict Dis 2009;281:64–67. [DOI] [PubMed] [Google Scholar]
- 24. Slade T, Johnston A, Oakley Browne MA, Andrews G, Whiteford H. 2007 National Survey of Mental Health and Wellbeing: Methods and key findings. Aust N Z J Psychiatry 2009;437:594–605. [DOI] [PubMed] [Google Scholar]
- 25. Haro JM, Arbabzadeh-Bouchez S, Brugha TS, et al. Concordance of the Composite International Diagnostic Interview Version 3.0 (CIDI 3.0) with standardized clinical assessments in the WHO World Mental Health Surveys. Int J Methods Psychiatry Res 2006;154:167–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Kessler RC, Üstün TB. The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). Int J Methods Psychiatry Res 2004;132:93–121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Nock MK, Borges G, Bromet EJ, et al. Cross-national prevalence and risk factors for suicidal ideation, plans and attempts. Br J Psychiatry 2008;1922:98–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Nock MK, Kessler RC. Prevalence of and risk factors for suicide attempts versus suicide gestures: Analysis of the National Comorbidity Survey. J Abnorm Psychol 2006;1153:616–23. [DOI] [PubMed] [Google Scholar]
- 29. World Health Organization. Composite International Diagnostic Interview. Version 3.0. Geneva: World Health Organization; 2001. [Google Scholar]
- 30. Nielsen S, Degenhardt L, Hoban B, Gisev N. Comparing Opioids: A Guide to Estimating Oral Morphine Equivalents (OME) in Research. NDARC Technical Report No. 329. Sydney: UNSW Australia, National Drug and Alcohol Research Centre; 2014.
- 31. Nielsen S, Degenhardt L, Hoban B, Gisev N. A synthesis of oral morphine equivalents (OME) for opioid utilisation studies. Pharmacoepidemiol Drug Saf 2016;256:733–37. [DOI] [PubMed] [Google Scholar]
- 32. World Health Organization. ATC Index with DDDs. Oslo, Norway: WHO Collaborating Centre for Drug Statistics Methodology; 2003. [Google Scholar]
- 33. Finnerup NB, Attal N, Haroutounian S, et al. Pharmacotherapy for neuropathic pain in adults: A systematic review and meta-analysis. Lancet Neurol 2015;142:162–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Chaparro LE, Wiffen PJ, Moore RA, Gilron I. Combination pharmacotherapy for the treatment of neuropathic pain in adults. Cochrane Database Syst Rev 2012;7:CD008943.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Bennett MI. Effectiveness of antiepileptic or antidepressant drugs when added to opioids for cancer pain: Systematic review. Palliat Med 2011;255:553–9. [DOI] [PubMed] [Google Scholar]
- 36. Day R, Cohen M. Are antipsychotic medications suitable for chronic noncancer pain? Pain Manag Today 2017;4(1):22–24. [Google Scholar]
- 37. Seidel S, Aigner M, Ossege M, et al. Antipsychotics for acute and chronic pain in adults. Cochrane Database Syst Rev 2013;8:CD004844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Walitt B, Klose P, Uceyler N, Phillips T, Hauser W. Antipsychotics for fibromyalgia in adults. Cochrane Database Syst Rev 2016;6:CD011804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Menzies V, Thacker LI, Mayer SD, et al. Polypharmacy, opioid use, and fibromyalgia: A secondary analysis of clinical trial data. Biol Res Nurs 2017;191:97–105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Nielsen S, Lintzeris N, Bruno R, et al. Benzodiazepine use among chronic pain patients prescribed opioids: Associations with pain, physical and mental health, and health service utilization. Pain Med 2015;162:356–66. [DOI] [PubMed] [Google Scholar]
- 41. Zin CS, Ismail F. Co-prescription of opioids with benzodiazepine and other co-medications among opioid users: Differential in opioid doses. J Pain Res 2017;10:249–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Jones CM, McAninch JK. Emergency department visits and overdose deaths from combined use of opioids and benzodiazepines. Am J Prev Med 2015;494:493–501. [DOI] [PubMed] [Google Scholar]
- 43. Kandel DB, Hu M-C, Griesler P, Wall M. Increases from 2002 to 2015 in prescription opioid overdose deaths in combination with other substances. Drug Alcohol Depend 2017;178:501–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Darke S, Duflou J, Torok M. Toxicology and characteristics of fatal oxycodone toxicity cases in New South Wales, Australia 1999-2008. J Forensic Sci 2011;56(3):690–3. [DOI] [PubMed] [Google Scholar]
- 45. Roxburgh A, Burns L, Drummer OH, et al. Trends in fentanyl prescriptions and fentanyl-related mortality in Australia. Drug Alcohol Rev 2013;32(3):269–75. [DOI] [PubMed] [Google Scholar]
- 46. Patterson SM, Cadogan CA, Kerse N, et al. Interventions to improve the appropriate use of polypharmacy for older people. Cochrane Database Syst Rev 2014;5:CD008165. [DOI] [PubMed] [Google Scholar]
- 47. National Institute for Health and Care Excellence. Depression in Adults: Recognition and Management. London: NICE; 2018. [PubMed] [Google Scholar]
- 48. Althubaiti A. Information bias in health research: Definition, pitfalls, and adjustment methods. J Multidiscip Healthc 2016;9:211–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Darke S. Self-report among injecting drug users: A review. Drug Alcohol Depend 1998;513:253–63. [DOI] [PubMed] [Google Scholar]