Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Mar 1.
Published in final edited form as: J Occup Environ Med. 2023 Dec 14;66(3):226–233. doi: 10.1097/JOM.0000000000003029

Productivity Loss Among Opioid and Benzodiazepine Users in the United States: A Medical Expenditure Panel Survey from 2010–2019

J Douglas Thornton a,b,*, Tyler Varisco a,b, Bhatt Prachet a,b, Olajumoke Olateju a,b, Mina Shrestha a,b, Chan Shen c
PMCID: PMC10922134  NIHMSID: NIHMS1948569  PMID: 38151973

Abstract

Objective:

To estimate the association between productivity losses and the use of prescription opioids and benzodiazepines among employed US adults with painful conditions.

Methods:

Using Medical Expenditures Panel Survey (2010–2019), we employed two-part (logistic regression and generalized linear model with zero-truncated negative binomial link) model to compare missed workdays due to illness or injury among employed adults with a painful condition.

Results:

Of the eligible sample of 57,413 working US individuals, 14.65% were prescription opioid users, 2.95% were benzodiazepine users, and 1.59% were both opioid and benzodiazepine users. The predicted missed workdays were 5.75 (95%CL:5.58–5.92) days for benzodiazepine users, 13.06 (95%CL:12.88–13.23) days among opioid users, and 15.18 (95%CL:14.46–15.90) days for opioid and benzodiazepine concomitant users.

Conclusions:

Concomitant use of prescription opioids and benzodiazepines was significantly associated with having more missed workdays among employed adults with documented painful conditions.

Keywords: Opioids, Benzodiazepine, Productivity Loss, Absenteeism, MEPS

Introduction

Opioids are commonly prescribed for the treatment of acute, severe, and chronic pain1. Their use for chronic non-cancer pain (CNCP) is controversial, especially when alternatives such as non-opioid analgesics or nonpharmacologic interventions can equally benefit patients1,2. This is because extended or inappropriate use of prescription opioids has medical implications such as overdose, serious health problems, and death, as well as economic implications such as increased medical care costs and workplace productivity losses3. However, a substantial number of patients still receive opioid prescriptions for their pain; about 20% of patients in the United States with noncancer pain-related diagnoses or symptoms receive an opioid prescription1. These health risks can increase with the concurrent use of Central Nervous System (CNS) depressants like benzodiazepines4–9, which is one of the most prescribed controlled substances with opioids5,10,11. The concurrent use of opioids and benzodiazepines has been shown to increase the risk of respiratory depression, oversedation and associated risks, and overdose-related deaths4–9.

The use of prescription opioids is prevalent among the US workforce, and this has been associated with decreased on-the-job productivity (presenteeism), missed days of work (absenteeism), disability, incarceration, and premature death12,13. The degree of this negative occupational hazard increases with the dose of the opioid leading to even longer recovery and more significant medical expenses14–16. The economic burden associated with opioid-related productivity loss was estimated at over $25 billion in 200713 and $78.5 billion in 201317 accounting for about 45% and 26% of the total societal costs associated with prescription opioid use disorder (OUD) in those years respectively. Although the prevalence of opioid prescribing among the US working population is reducing, there are still concerns about occupational injuries and disabilities due to their use, leading to absenteeism. This may be attributed to the negative effect of opioids on muscle strength and reaction time, judgment, coordination, memory, and attention3,18. Price et al reported a strong association between occupational injuries and opioid use in their analysis of Louisiana’s workers compensation claims19. Van Hasselt et al showed using National Survey on Drug Use and Health (NSDUH) 2008 – 2012 data that prescription drug misuse, including opioids, led to a 7.4% increase in absenteeism20.

Prior studies have suggested that measuring absenteeism can give a direct estimation of productivity loss and profitability and recommended that absenteeism be considered in assessing the impact of opioids on the population. However, there is a paucity of nationally representative studies showing how absenteeism is associated with the use of prescription opioids alone or in combination with prescription benzodiazepines; therefore, a lack of recent information on how these prescription medications may affect productivity in the workplace persists. Such findings can provide more compelling evidence for prescribers to critically examine the need to prescribe opioids, further reducing opioid drug misuse and associated health economic implications. The objective of this study was therefore to estimate the association between productivity loss and the use of prescription opioids, with or without benzodiazepine use, among a nationally representative sample of non-institutionalized US adults with a painful condition.

Methods

Study Design and Sample

This was a retrospective cross-sectional study designed to assess the association between workplace productivity loss and opioid use with or without benzodiazepine use among working adults in the US. Participants were eligible for inclusion if they were over 18-years of age, had at least one medical encounter due to a condition likely to cause chronic pain21, not diagnosed with any cancer, and were currently employed. For this study, we adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Guidelines for observational studies and the checklist is attached in the Supplementary Digital Content (SDC).

Data Source

This study used 10 years (2010–2019) of the Medical Expenditures Panel Survey (MEPS) data, a publicly available, nationally representative (of the civilian, non-institutionalized population) survey-based dataset with detailed information on the use and cost of health care22. MEPS data are released in full-year consolidated household files, medical conditions, prescription drug events, and others22. Full-year consolidated household files contain individual-level data on demographics, healthcare expenditures, healthcare use, sources of payments, and health insurance coverage22. The Medical Conditions files contain information on conditions taken from respondents which were coded to International Classification of Disease, 9th or 10th Edition, Clinical Modification (ICD-9-CM/ICD-10-CM) codes by professional coders then were converted by Agency for Healthcare Research and Quality (AHRQ) to clinical classification codes. The information in the Prescribed Medicines files is at the event level, and each record is a unique prescribed medicine including the national drug code (NDC), medicine name, and any conditions reported with the prescription processing. MEPS participants are followed for a period of two-calendar years22. The survey is weighted so that annual observations for each participant are independent. This means that each participant may have contributed two observations to the study, one for each year they met the inclusion criteria.

Measures

The outcome of interest was workplace absenteeism defined by the number of days of work missed due to illness or injury by any participant across the total year of observation. This was defined as a discrete count of missed workdays directly from a variable on the consolidated data file, Days Missed Work Due to Ill/Inj (DDNWRK), representing the number of times the person lost a half-day or more from work because of illness, injury, or mental or emotional problems. The exposure of interest was exposure to opioid or benzodiazepine medication at any point in the observation period. Use was identified from the event-level prescribed medicine file by multum therapeutic code (Benzodiazepines: 69, Opioid analgesics: 60, 191). After identifying participants with a single prescription in the past year for a medication in one or both classes, a categorical variable classifying patient as users of opioids, benzodiazepines, both classes of medication, and neither in the year of observation was created.

Similar to prior studies, all models were adjusted for participant occupation, age, sex, gender, marital status, smoking status, educational status, poverty status, region of national residence, painful condition history, calendar year of follow-up, number of comorbidities, and perceived mental and physical health status.8,12,23,24 Individuals missing data in any of these indicators were excluded from the analysis. As this was a cross-sectional design, no attempt was made to establish temporality between opioid or benzodiazepine use, the exposure of interest, and days of work missed. All measures were operationalized for the entire year of follow up.

Statistical Analysis

Weighted means and frequencies were used to describe the demographics of the cohort. These were estimated using the SURVEY procedures in SAS version 9.4 and are intended to account for the complex sampling methodology used in MEPS data collection and weight all participants to represent the demographics of the US. As estimates were pooled from 10-years of data, the patient weight variable was divided by 10 to annualize all estimates.

We used a two-part model to compare the number of workdays missed annually due to illness or injury among employed adults with varying patterns of opioid or benzodiazepine use. We chose this approach because the dependent variable, days of work missed due to illness or injury, was over-dispersed and zero-inflated meaning that a traditional linear regression model would have been mis-specified25,26. In a two-part model, logistic regression is first used to estimate the probability that each participant experienced at least one day of missed work25. Next, a generalized linear model with a zero-truncated negative binomial link was used to estimate the number of workdays missed among individuals with at least one missed workday27,28. The predicted days of work missed from this model were then multiplied by the predicted probability of missing at least one-day of work due to illness or injury to account for the probability of censoring due to not missing a single day of work in the year of follow-up26,28. The weighted means and the number of days of work missed and their 95% confidence intervals (CIs) were compared between each of the four exposure groups.

Results

This study included 57,413 working US individuals, representing 65,769,891 Americans after applying survey weights, from 2010 to 2019 [Table 1]. Most of the study population were younger (94.56%) female (50.11%, 95% CI: 49.48%–50.73%), and white (71.18%, 95% CI: 69.82%–72.55%). Of the sampled population, 14.65% (95% CI: 14.23%–15.07%) were prescription opioid only users, 2.95% (95% CI: 2.72%–3.18%) were benzodiazepine only users, and 1.59% (95% CI: 1.43%–1.75%) used both opioids and benzodiazepines during observation period [Table 1]. A majority of the eligible individuals were married (57.53%, 95% CI: 56.52%–58.54%), had an income of over 200% above the federal poverty line (83.16%, 95% CI: 82.53%–83.80%), and had some form of private insurance (84.20%, 95% CI: 83.59%–84.81%) [Table 1]. Individuals with the highest categories of pain included those with a composite of limb pain, joint pain and non-systemic, non-inflammatory arthritic pain (18.38%, 95% CI: 17.49%–19.26%), pain due to fractures, contusions, sprains and strains (17.59%, 95% CI: 16.82%–18.37%), and back pain (12.60%, 95% CI: 11.80%–13.41%) and individuals with the lowest category of pain included those with orofacial, ear, and temporomandibular disorder pain (2.97%, 95% CI: 2.55%–3.40%), fibromyalgia (3.10%, 95% CI: 2.66%–3.54%), and some form of urogenital, pelvic and menstrual pain (3.15%, 95% CI: 2.74%–3.57%) [Table 1]. Nearly half (49.13%, 95% CI: 48.33%–49.93%) of the population did not have any diagnosed comorbidity, or had one (22.32%, 95% CI:21.82%–22.83%) or two (13.78%, 95% CI: 13.35%–14.20%) diagnosed comorbidities [Table 1]. The sample was further described by medication use category in Table 2.

TABLE 1.1:

Demographic Characteristics of Working Adults with a Diagnosed Painful Condition: Medical Expenditure Panel Survey 2010–2019

VARIABLES Unweighted Frequency Weighted Frequency Percent [95% CL]
(N= 57,413; Weighted Population = 65,769,891)
Exposure to Controlled Substances
Neither Opioid nor Benzodiazepine 47,057 53,150,274 80.81% [80.35–81.27]
Benzodiazepine Only 1,451 1,940,067 2.95% [2.72–3.18]
Opioid Only 8,128 9,633,758 14.65% [14.23–15.07]
Both Opioid & Benzodiazepine 777 1,045,792 1.59% [1.43–1.75]
Age Category
Age 18–29 9,024 10,930,919 16.6% [15.9–17.3]
Age 30–39 12,865 14,136,324 21.5% [20.8–22.2]
Age 40–49 13,744 15,186,290 23.1% [22.5–23.7]
Age 50–64 18,569 21,740,411 33.1% [32.2–33.9]
Age 65–74 2,831 3,314,441 5.04% [0.55=0.74]
Age 75–84 351 420,964 0.64% [0.55–0.74]
Age 85 and above 29 40,542 0.06% [0.04–0.10]
Race
White 31,463 46,817,374 71.18% [69.82–72.55]
Non-White 25,950 18,952,517 28.82% [27.45–30.18]
Sex
Male 29,702 32,815,111 49.89% [49.27–50.52]
Female 27,711 32,954,780 50.11% [49.48–50.73]
Region
Northeast 8,939 11,452,257 17.41% [16.20–18.63]
Northwest 12,884 15,771,031 23.98% [22.53–25.43]
South 20,201 23,257,826 35.36% [33.83–36.90]
West 15,388 15,285,984 23.24% [22.06–24.43]
Poverty Status
< 100% Federal Poverty Line 4,005 2,994,541 4.55% [4.28–4.82]
100%–200% Federal Poverty Line 9,896 8,077,871 12.28% [11.80–12.77]
> 200% Federal Poverty Line 43,512 54,697,479 83.16% [82.53–83.80]
Marital Status
Married 31,708 37,837,308 57.53% [56.52–58.54]
Unmarried 25,705 27,932,582 42.47% [41.46–43.48]
Insurance Status
Any Private Insurance 44,815 55,379,663 84.20% [83.59–84.81]
Public Insurance or No Insurance 12,598 10,390,228 15.80% [15.19–16.41]
Education Level
High School or Less 21,799 20,546,003 31.24% [30.34–32.14]
More Than High School 35,614 45,223,887 68.76% [67.86–69.66]
Smoking Status
Current Smoker 48,273 10,261,387 15.60% [15.04–16.17]
Non-Smoker or Former Smoker 9,140 55,508,504 84.40% [83.83–84.96]
Pain Category
Abdominal and bowel pain 73,77 8,555,103 4.26% [3.82–4.69]
Back pain 21,276 25,313,793 12.60% [11.80–13.41]
Fibromyalgia 5,621 6,229,272 3.10% [2.66–3.54]
Fractures, contusions, sprains, and strains 29,785 35,335,078 17.59% [16.82–18.37]
Headache 11,535 13,238,746 6.59% [6.00–7.19]
Limb/extremity pain, joint pain, and non-systemic, non-inflammatory arthritic disorders 31,415 36,907,124 18.38% [17.49–19.26]
Musculoskeletal chest pain 6,863 7,745,245 3.86% [3.37–4.35]
Neck pain 10,008 12,181,071 6.06% [5.51–6.62]
Neuropathy 15,680 17,116,531 8.52% [7.83–9.21]
Orofacial, ear, and temporomandibular disorder pain 5,453 5,974,651 2.97% [2.55–3.40]
Other painful conditions 11,393 13,607,080 6.77% [6.14–7.41]
Systemic disorders or diseases causing pain 17,485 20,051,501 9.98% [9.13–10.83]
Urogenital, pelvic, and menstrual pain 5,427 6,334,461 3.15% [2.74–3.57]
Number of Comorbidities
No Comorbidities 28,839 32,311,817 49.13% [48.33–49.93]
One Comorbidity 12,609 14,682,186 22.32% [21.82–22.83]
Two Comorbidity 7,664 9,060,014 13.78% [13.35–14.20]
Three Comorbidity 4,721 5,503,365 8.37% [8.02–8.71]
Four or More Comorbidity 3,580 4,212,509 6.40% [6.05–6.76]
Mean [95% CI]
Perceived Physical Health Status 1 57413 65,769,891 1.09 [1.088–1.095]
Perceived Mental Health Status 1 57413 65,769,891 1.05 [1.048–1.053]
Open in a new tab
1

Health status is coded as composite value: (1) excellent, very good, good, (2) fair or poor 95% CL=Wald 95% Confidence Limits

TABLE 2:

Unadjusted, Weighted Descriptive of Missed Workdays of Employed Adults with a Painful Condition: Medical Expenditure Panel Survey 2010–2019

Exposure to Controlled Substances Adults Who Missed ZERO Days of Work Adults Who Missed AT LEAST ONE Day of Work Missed Workdays (Including ZERO Missed Workdays)
Exposure to Controlled Substances Weighted Frequency Percent [95% CL] Weighted Frequency Percent [95% CL] Mean [95% CL]
Neither Opioid nor Benzodiazepine 27,892,251 52.48% [51.71–53.25%] 25,258,023 47.52% [46.75–48.29%] 2.92 [2.82–3.02]
Benzodiazepine Only 708,255 36.51% [33.18–39.84%] 1,231,812 63.49% [60.16–66.82%] 5.41 [4.73–6.10]
Opioid Only 2,067,011 21.46% [20.28–22.64%] 7,566,747 78.54% [77.36–79.72%] 12.86 [12.29–13.43]
Both Opioid & Benzodiazepine 186,333 17.82% [14.64–21.00%] 859,459 82.18% [79.00–85.36%] 15.21 [12.97–17.44]
Open in a new tab

95% CL=Wald 95% Confidence Limits

On average, individuals not exposed to either opioids or benzodiazepines missed 2.92 (95% CI: 2.82–3.02) days of work [Table 3]. The unadjusted average missed days of work was 5.41 days (95% CI: 4.73–6.10) among benzodiazepine only users and 12.86 (95% CI: 12.29–13.43) days for opioid users [Table 3]. Combination controlled substance users missed on average 15.21 (95% CI: 12.97–17.44) days of work [Table 3]. Moreover, more individuals were prescribed both opioids and benzodiazepines (82.18%; 95% CI: 79.00%–85.36%) missed at least one day of work. [Table 3].

TABLE 3:

Two Part Model Estimates of Adults with a Painful Condition Missing Work: Medical Expenditure Panel Survey 2010–2019

Probability of Missing at Least One Day of Work
(Part 1: Logistic Regression Estimates)
(Sample = 57,413)		 Number of Workdays Missed Among Those Who Missed At least One Day of Work
(Sample = 29,816)
VARIABLES Adjusted Odd Ratio 95% CL Point Estimates 95% CL
Exposure to Controlled Substances
No Opioid or Benzodiazepine Use (REF) (REF) (REF) (REF)
Benzodiazepine Use 1.41 1.22–1.63 0.24 0.17–0.31
Opioid Use 3.68 3.43–3.95 0.92 0.88–0.95
Opioid & Benzodiazepine Use 3.36 2.74–4.14 0.87 0.79–0.96
Age Category
Age 18–29 (REF) (REF) (REF) (REF)
Age 30–39 0.86 0.80–0.93 0.05 0.01–0.09
Age 40–49 0.71 0.66–0.77 −0.01 (−0.05)–0.03
Age 50–64 0.61 0.56–0.65 0.05 0.01–0.09
Age 65–74 0.44 0.38–.50 0.17 0.10–0.24
Age 75–84 0.42 0.31–0.58 0.41 0.24–0.59
Age 85 and above 0.63 0.20–1.99 0.16 (−0.41)–0.72
Race
White (REF) (REF) (REF) (REF)
Non-white 0.94 0.89–0.99 0.20 0.17–0.23
Sex
Male (REF) (REF) (REF) (REF)
Female 1.42 1.36–1.50 0.17 0.14–0.19
Region
Northeast (REF) (REF) (REF) (REF)
Northwest 1.01 0.93–1.10 −0.16 (−0.20)–(−0.12)
South 0.96 0.88–1.03 −0.12 (−0.16)–(−0.08)
West 1.05 0.96–1.15 −0.08 (−0.12)–(−0.04)
Poverty Status
< 100% Federal Poverty Line (REF) (REF) (REF) (REF)
100%–200% Federal Poverty Line 1.07 0.94–1.21 −0.11 (−0.16)–(−0.05)
> 200% Federal Poverty Line 1.04 0.93–1.17 −0.24 (−0.29)–(−0.18)
Marital Status
Married (REF) (REF) (REF) (REF)
Unmarried 1.21 1.15–1.28 −0.07 (−0.10)–(−0.05)
Insurance Status
Any Private Insurance (REF) (REF) (REF) (REF)
Public Insurance or No Insurance 0.81 0.75–0.87 −0.07 (−0.11)–(−0.03)
Education Level
High School or Less (REF) (REF) (REF) (REF)
More Than High School 1.05 0.93–1.17 −0.03 (−0.06)–0.003
Smoking Status
Non-Smoker or Former Smoker (REF) (REF) (REF) (REF)
Current Smoker 0.81 0.76–0.87 0.04 0.005–0.07
Pain Category
No Pain (REF) (REF) (REF) (REF)
Abdominal and bowel pain 1.20 1.11–1.30 0.10 0.06–0.14
Back pain 1.08 1.01–1.15 −0.06 (−0.09)–(−0.03)
Fibromyalgia 1.07 0.96–1.19 0.03 (−0.03)–0.09
Fractures, contusions, sprains and strains 1.200 1.14–1.27 0.16 0.14–0.19
Headaches 1.33 1.24–1.42 −0.05 (−0.09)–(−0.02)
Limb/extremity pain, joint pain and non-systemic, non-inflammatory arthritic disorders 1.01 0.95–1.08 0.01 (−0.03)–0.04
Musculoskeletal chest pain 1.06 0.98–1.14 0.0206 (−0.02)–0.06
Neck pain 1.04 0.96–1.12 0.0778 0.04–0.12
Neuropathy 0.94 0.93–1.06 0.0690 0.03–0.11
Orofacial, ear, and temporomandibular disorder pain 1.11 1.01–1.22 −0.0220 (−0.07)–0.02
Other painful conditions 1.14 1.07–1.21 0.0995 0.07–0.13
Systemic disorders or diseases causing pain 0.93 0.87–1.00 −0.1270 (−0.17)–(−0.09)
Urogenital, pelvic and menstrual pain 1.18 1.08–1.29 −0.0335 (−0.08)–0.01
Number of Comorbidities
None or One Comorbidity (REF) (REF) (REF) (REF)
Two Comorbidity 1.45 1.36–1.55 0.16 0.12–0.19
Three Comorbidity 1.53 1.41–1.67 0.17 0.13–0.22
Four or More Comorbidity 1.86 1.68–2.06 0.35 0.30–0.41
Ever Having a Cancer Diagnosis
Yes (REF) (REF)
No 0.12 0.10–0.13
Occupational Category
Management, Business, & Financial Occupation (REF) (REF) (REF) (REF)
Professional & Related Occupation 1.14 1.06–1.23 −0.01 −0.05–0.04
Services Occupation 0.92 0.85–0.99 0.16 0.11–0.20
Sales Related Occupation 0.84 0.75–0.93 0.08 0.03–0.14
Office & Administrative Support Occupation 1.20 1.09–1.32 0.13 0.08–0.18
Farming, Fishing, & Forestry Occupation 0.81 0.59–1.11 0.46 0.28–0.63
Construction, Extraction, & Maintenance Occupation 1.13 1.01–1.27 0.27 0.21–0.33
Production & Transportation Based Occupation 1.04 0.95–1.15 0.32 0.27–0.37
Military Specific Occupation 1.13 0.70–1.81 0.03 −0.31–0.37
Year
2010 (REF) (REF) (REF) (REF)
2011 1.05 0.95–1.15 0.14 0.08–0.20
2012 1.02 0.93–1.12 −0.05 −0.11–0.01
2013 1.09 0.99–.1.20 −0.07 −0.12–(−0.01)
2014 1.02 0.93–1.12 −0.15 −0.21–(−0.09)
2015 1.18 1.07–1.30 −0.16 −0.22–(−0.11)
2016 1.11 1.00–1.22 −0.17 −0.23–(−0.09)
2017 1.09 0.97–1.21 −0.15 −0.21–(−0.10)
2018 1.39 1.23–1.57 0.04 −0.03–0.06
2019 1.44 1.28–1.62 −0.01 −0.07–0.06
Perceived Physical Health Status 1 1.73 1.60–1.87 0.41 0.37–0.45
Perceived Mental Health Status 1 1.44 1.28–1.63 0.09 0.04–0.15
Open in a new tab
1

Health status is coded as composite value: (1) excellent, very good, or good, (2) fair or poor 95% CL = Wald 95% Confidence Limits

The two-part model with zero-truncated negative binomial model found that exposure to benzodiazepine and opioid was associated with higher odds and the number of missing workdays when compared to those who did not use any of these controlled substances [Table 3]. Opioid and benzodiazepine users had higher odds (3.36; [95% CI: 2.74–4.14]) of missed workdays when compared to patients who did not use any of these controlled substances. Similarly, opioid only users had 3.68 (95% CI: 3.43–3.95) and benzodiazepine only users had 1.41 (95% CI: 1.22–1.63) times higher odds of missed workdays [Table 3]. An increase in age was negatively associated with missing a day of work (aOR(Age 18–29 vs Age 30–39) 0.86, 95% CI: 0.80–0.93) [Table 3]. Individuals who worked in a professional occupation (aOR 1.14, 95% CI: 1.06–1.23), an office or administrative support capacity (aOR 1.20, 95% CI: 1.09–1.32), or construction position were more likely to miss a day of work (aOR 1.13, 95% CI: 1.01–1.27) compared to individuals who were employed in a management, business, or financial occupation [Table 3].

The predicted average number of workdays missed for individuals not exposed to either opioids or benzodiazepines was 2.87 (95% CL: 2.85–2.90) days of work, for benzodiazepine only users was 5.74 (95% CI: 5.57–5.91) and opioid users was 13.08 (95% CL: 12.90–13.26) days of work in a year [Table 4]. Combination benzodiazepine and opioid users were predicted to miss on average 15.28 (95% CI: 14.56–16.01) days of work annually [Table 4].

TABLE 4:

Predicted Average Number of Workdays Missed by Adults Diagnosed with a Painful Condition (Excluding Adults Who Missed Zero Days of Work): Medical Expenditure Panel Survey 2010–2019

VARIABLES Missed Workdays (Excluding Adults Who Missed Zero Days of Work)
(N = 29,816)
Exposure to Controlled Substances Predicted Average [95% CL]
No Opioid or Benzodiazepine Use 2.87 [2.85–2.90]
Benzodiazepine Use 5.74 [5.57–5.91]
Opioid Use 13.08 [12.90–13.26]
Opioid & Benzodiazepine Use 15.28 [14.56–16.01]
Open in a new tab

95% CL=Wald 95% Confidence Limits

Discussion

This study used a nationally representative survey (MEPS) to estimate the association between productivity loss and the use of prescription opioids, with or without benzodiazepine use. We found that use of prescription opioid and benzodiazepine was significantly associated with a greater number of missing workdays. Our study found that patients who were prescribed benzodiazepines and opioids, on average, missed 5.74 workdays and 13.08 workdays, respectively. This translates to one to nearly two weeks throughout the year. Moreover, there is an increasing level of absenteeism as individuals are prescribed opioids and benzodiazepine concurrently as compared to those only prescribed either opioids or benzodiazepines. We estimated that individuals using both opioids and benzodiazepines missed on average 15.28 workdays per year which is a little more than two weeks. Goplerud et. al. used three years of data from the National Survey of Drug Use and Health and found that people with substance use disorders missed 14.8 days a year which is similar to our findings for opioid users29. However, our study used ten years of MEPS (2010–2019) data to estimate workdays missed by users of specific types of controlled substances (i.e., opioids and benzodiazepines). For individuals with lower socioeconomic status, this can contribute additional financial strain on already reduced household budgets30.

Another study analyzed the effects of opioid prescribing for chronic musculoskeletal pain conditions on lost productivity in the United States using deidentified MarketScan data from 2011–201731. The authors concluded that patients with low and moderate morphine milligrams equivalence (MME) were more likely than those on no (MME) to have a disability-related absence (RR 1.4, 95% CI: [1.3–1.5] and RR 1.2, 95% CI: [1.1–1.3], respectively31. Though our study does not incorporate opioid or benzodiazepine MME, our study is in line with this study that highlights controlled substance use being associated with increased number of missed workdays.

A worker’s occupation plays a critical role in missing a day of work. We found that workers from physically demanding professions such as construction and transportation as well as white collar professional workers are more likely to miss at least one day of work. Controlled substance use of opioids, benzodiazepines, or their use in combination can result in drowsiness and psychomotor slowing which can negatively impact job performance and inhibit absenteeism32,33. A user may choose to forgo a day of work due to their severe pain or from the side effects of using these prescription medications.

Prevalence of high-impact chronic pain that can limit patients’ work activities has been shown to increase with age.33 In contrast, our logistic regression found an inverse association in the probability of missing a workday with age. The projected number of cancer diagnosis between 2026 and 2030 for individuals aged 55–69 is approximately 1,050 per 100,000 compared to approximately 500 per 100,000 for individuals aged 40 to 5434. Cancer patients are more likely to be concomitant users of opioids and benzodiazepines making them susceptible to miss their work35,36. There might be a more complex relationship between age, employment, and increased use of opioids and benzodiazepines. However, this study neither explored the quadratic function of age with opioid and benzodiazepine use nor the interaction of age and cancer diagnosis.

A key strength of this study is that it aggregated ten years of nationally representative data to estimate worker productivity loss associated with opioid, benzodiazepine, and their concurrent use among US noninstitutionalized workers. The duration of this cross-sectional study (2010–2019) has seen significant changes in opioid and benzodiazepine prescribing guidelines. There has been an decrease in the rates of physicians co-prescribing both opioids and benzodiazepines after the release of the 2016 Centers for Disease Control (CDC) and Prevention Guidelines to mitigate prescribing and co-prescribing risk factors37. Clinicians, public health officials, and patients have become acutely aware of how complex pain management can be for a variety of at-risk patient groups (i.e. cancer vs non-cancer patients) as well as those with co-occurring mental health disorders37,38. Though these medications can result in misuse and abuse, patients, physicians, and the CDC have recognized that these medications are an effective pharmacological tool that allows patients to live their day-to-day life had they not been prescribed these medications.

The results of this study should be viewed with the context of the limitations. One limitation is the inability to account for the variation in the strength and intensity (morphine milligram equivalent; MME) of opioids and benzodiazepines which were prescribed and the associated pain severity it is meant to treat. Secondly, the construction of MEPS only accounts to the third digit of the ICD-9/ICD-10 diagnosis code. This limits this study’s ability to attribute the reason for why the prescription drugs were prescribed. Our study also did not identify the percentage of opioid and benzodiazepine users who were also receiving medication assisted treatment (MAT) or medications for opioid use disorder (MOUD). According to 2019 estimates, less than 35% of adults with OUD were able to receive treatment for opioid use within the past year highlighting a clear disparity in access to opioids and benzodiazepines39. Lastly our measure of absenteeism was unable to differentiate between missing work due to an illness, side effect of using opioids or benzodiazepines, or some other non-medical reason. However, this unspecified measure of absenteeism has been used frequently by prior researchers to measure the productivity loss associated with diseases like asthma, cancer, obesity, and osteoarthritis2,40–42. Finally, a parallel study may be conducted to estimate the direct medical cost associated with patients being prescribed these prescription medications individually and concurrently. This study employed a cross-sectional study design which comes with its own methodological challenges like the inability to account for prior use of other drugs including controlled substances.

Conclusion

The economic burden of controlled substance use is a growing problem in the United States, and the excess burden upon the work force is an overlooked component when prescribing such medication. Productivity losses as measured through absenteeism needs to be considered when weighing the risks of prescribing these prescription drugs. Future work needs to focus on combined effects of opioids with other controlled substances that place individuals at risk for missing more workdays due to ineffective pain management and opioid misuse and abuse.

Supplementary Material

Supplemental Digital Content- STROBE

TABLE 1.2:

Demographic Characteristics of Working Adults with a Diagnosed Painful Condition by their Use of Controlled Substances: Medical Expenditure Panel Survey 2010–2019

VARIABLES Neither Benzodiazepine Only Opioid Only Both
Missed Workdays
Yes 52.48% 36.51% 21.46% 17.82%
No 47.52% 63.49% 78.54% 82.18%
Age Category
Age 18–29 17.19% 9.98% 15.59% 9.46%
Age 30–39 21.50% 22.03% 21.57% 19.60%
Age 40–49 23.26% 22.66% 22.06% 24.91%
Age 50–64 32.61% 38.95% 33.77% 38.16%
Age 65–74 4.79% 5.16% 6.14% 7.25%
Age 75–84 0.59% 1.23% 0.79% 0.62%
Age 85 and above 0.06% 0.08%
Race
White 69.88% 84.40% 74.04% 87.00%
Non-White 30.12% 15.60% 25.96% 13.00%
Sex
Male 51.51% 36.58% 45.53% 32.48%
Female 48.49% 63.42% 54.47% 67.52%
Region
Northeast 17.84% 22.57% 14.34% 14.59%
Northwest 23.70% 25.35% 24.85% 27.84%
South 34.74% 32.78% 38.69% 41.47%
West 23.73% 19.31% 22.11% 16.11%
Poverty Status
< 100% Federal Poverty Line 4.37% 4.04% 5.63% 4.79%
100%–200% Federal Poverty Line 12.04% 11.12% 13.46% 16.02%
> 200% Federal Poverty Line 83.59% 84.84% 80.91% 79.20%
Marital Status
Married 58.04% 48.98% 56.62% 55.48%
Unmarried 41.96% 51.02% 43.38% 44.52%
Insurance Status
Any Private Insurance 84.12% 86.51% 84.24% 83.70%
Public Insurance or No Insurance 15.88% 13.49% 15.76% 16.30%
Education Level
High School or Less 30.69% 26.46% 35.28% 30.42%
More Than High School 69.31% 73.54% 64.72% 69.58%
Smoking Status
Current Smoker 14.58% 17.53% 19.67% 26.77%
Non-Smoker or Former Smoker 85.42% 82.47% 80.33% 73.23%
Pain Category
Abdominal and bowel pain 11.53% 14.04% 19.84% 23.36%
Back pain 36.97% 43.34% 44.10% 55.10%
Fibromyalgia 8.81% 11.46% 11.50% 20.60%
Fractures, contusions, sprains, and strains 52.19% 53.75% 61.14% 63.43%
Headache 19.26% 23.55% 22.92% 32.17%
Limb/extremity pain, joint pain, and non-systemic, non-inflammatory arthritic disorders 54.56% 61.99% 61.90% 71.22%
Musculoskeletal chest pain 11.25% 15.39% 13.20% 18.73%
Neck pain 17.20% 22.27% 23.56% 32.41%
Neuropathy 25.28% 27.01% 28.32% 40.71%
Orofacial, ear, and temporomandibular disorder pain 8.83% 10.99% 9.49% 14.73%
Other painful conditions 18.02% 36.07% 29.52% 46.47%
Systemic disorders or diseases causing pain 29.41% 34.15% 34.72% 39.55%
Urogenital, pelvic, and menstrual pain 8.95% 9.66% 12.94% 13.92%
Number of Comorbidities
No Comorbidities 51.30% 23.57% 44.68% 27.42%
One Comorbidity 23.08% 14.95% 20.82% 11.40%
Two Comorbidity 13.08% 23.75% 15.05% 18.93%
Three Comorbidity 7.53% 16.76% 10.28% 17.85%
Four or More Comorbidity 5.02% 20.97% 9.17% 24.41%
Perceived Physical Health Status 1 1.09 1.14 1.17 1.28
Perceived Mental Health Status 1 1.04 1.16 1.07 1.17
N 47,057 1,451 8,128 777
Weighted Population 53,150,274 1,940,067 9,633,758 1,045,792
Open in a new tab
1

Health status is coded as composite value: (1) excellent, very good, good, (2) fair or poor 95% CL=Wald 95% Confidence Limits

Learning Outcomes:

  • Understanding treatment choices for pain management and the risk of concomitant use of controlled substances.

  • Workplace productivity is critical to assess when evaluating the cost of chronic pain for the US workforce.

Acknowledgments:

We adhered to the STROBE Guidelines for observational studies.

Funding Sources:

Dr. Thornton and Dr. Shen were partially supported by a grant from the National Institute of Health, National Institute on Drug Abuse (1R03DA047597-01).

Footnotes

Conflict of Interest: Dr. Thornton is currently a consultant for the Plaintiff’s Steering Committee for Opioid Litigation and a member of the Texas Opioid Abatement Settlement Council for the State of Texas. Dr. Varisco is a paid consultant for HEALIX Infusion Therapy. All other authors declare no conflicting interest.

Ethics approval and consent to participate: This study was exempt from the University of Houston Institutional Review Boards’ approval because deidentified secondary data was used.

Consent for publication: Not Applicable

Availability of data and materials:

The author confirms that all data generated or analyzed during this study are included in this published article.

References

  • 1.Dowell D, Haegerich TM, Chou R. CDC Guideline For Prescribing Opioids For Chronic Pain — United States, 2016. MMWR Recomm Rep. 2016;315(15):1624–1645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Xiaohui Zhao DS, Kavita Gandhi, Wenhui Wei, Nilanjana Dwibedi, Lynn Webster & Usha Sambamoorthi. The Association Of Pain Interference And Opioid Use With Healthcare Utilization And Costs, And Wage Loss Among Adults With Osteoarthritis In The United States. Journal of Medical Economics. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Abay Asfaw TA, Brian Quay Prevalence And Expenses Of Outpatient Opioid Prescriptions, With Associated Sociodemographic, Economic, And Work Characteristics. International Journal of Social Determinants of Health and Health Services. 2020;50(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Boon M, van Dorp E, Broens S, Overdyk F. Combining Opioids And Benzodiazepines: Effects On Mortality And Severe Adverse Respiratory Events. Ann Palliat Med. 2020;9(2):542–557. [DOI] [PubMed] [Google Scholar]
  • 5.Park TW, Saitz R, Ganoczy D, Ilgen MA, Bohnert ASB. Benzodiazepine Prescribing Patterns And Deaths From Drug Overdose Among US Veterans Receiving Opioid Analgesics: Case-Cohort Study . BMJ : British Medical Journal. 2015;350:h2698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Sun EC, Dixit A, Humphreys K, Darnall BD, Baker LC, Mackey S. Association Between Concurrent Use Of Prescription Opioids And Benzodiazepines And Overdose: Retrospective Analysis. BMJ. 2017;356:j760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hernandez I, He M, Brooks MM, Zhang Y. Exposure-Response Association Between Concurrent Opioid And Benzodiazepine Use And Risk Of Opioid-Related Overdose In Medicare Part D Beneficiaries. JAMA Network Open. 2018;1(2):e180919-e180919. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Erin YL, Robyn T, Kristian BF, David LB. Concurrent Prescriptions For Opioids And Benzodiazepines And Risk Of Opioid Overdose: Protocol For A Retrospective Cohort Study Using Linked Administrative Data. BMJ Open. 2021;11(2):e042299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Jones CM, McAninch JK. Emergency department visits and overdose deaths from combined use of opioids and benzodiazepines. American Journal of Preventive Medicine. 2015;49(4):493–501. [DOI] [PubMed] [Google Scholar]
  • 10.Vadiei N, Bhattacharjee S. Concurrent Opioid And Benzodiazepine Utilization Patterns And Predictors Among Community-Dwelling Adults In The United States. Psychiatric Services. 2020;71(10):1011–1019. [DOI] [PubMed] [Google Scholar]
  • 11.Vozoris NT. Benzodiazepine And Opioid Co-Usage In The Us Population, 1999–2014: An Exploratory Analysis. Sleep. 2019;42(4). [DOI] [PubMed] [Google Scholar]
  • 12.White AG, Spittle T, Fernan C, et al. Assessment Of Work Loss Associated With Prescription-Related Opioid Use Disorder: A Retrospective Analysis Of Claims Data. Journal of Occupational and Environmental Medicine. 2020;62(3):217–222. [DOI] [PubMed] [Google Scholar]
  • 13.Birnbaum HG, White AG, Schiller M, Waldman T, Cleveland JM, Roland CL. Societal Costs Of Prescription Opioid Abuse, Dependence, And Misuse In The United States. Pain Medicine. 2011;12(4):657–667. [DOI] [PubMed] [Google Scholar]
  • 14.Gomes T, Mamdani MM, Dhalla IA, Paterson JM, Juurlink DN. Opioid Dose And Drug-Related Mortality In Patients With Nonmalignant Pain. Archives of Internal Medicine. 2011;171(7):686–691. [DOI] [PubMed] [Google Scholar]
  • 15.Lavin RA, Tao XG, Yuspeh L, Bernacki EJ. Impact of the combined use of benzodiazepines and opioids on workers’ compensation claim cost. J Occup Environ Med. 2014;56(9):973–978. [DOI] [PubMed] [Google Scholar]
  • 16.Hunt DL, Artuso RD, Kalia N, Leung N, Bernacki EJ, Tao X. Association of Opioid, Anti-Depressant, and Benzodiazepines With Workers’ Compensation Cost: A Cohort Study. Journal of Occupational and Environmental Medicine. 2019;61(5). [DOI] [PubMed] [Google Scholar]
  • 17.Florence C, Luo F, Xu L, Zhou C. The Economic Burden Of Prescription Opioid Overdose, Abuse And Dependence In The United States, 2013. Medical care. 2016;54(10):901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Kidner CL, Mayer TG, Gatchel RJ. Higher Opioid Doses Predict Poorer Functional Outcome In Patients With Chronic Disabling Occupational Musculoskeletal Disorders. The Journal of Bone and Joint Surgery. American volume. 2009;91(4):919. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Price JW. A Comparison Of Random And Post-Accident Urine Opiate And Opioid Tests. Journal of Addictive Diseases. 2015;34(1):36–42. [DOI] [PubMed] [Google Scholar]
  • 20.Van Hasselt M, Keyes V, Bray J, Miller T. Prescription Drug Abuse And Workplace Absenteeism: Evidence From The 2008–2012 National Survey On Drug Use And Health. Journal of Workplace Behavioral Health. 2015;30(4):379–392. [Google Scholar]
  • 21.Mayhew M, DeBar LL, Deyo RA, et al. Development And Assessment Of A Crosswalk Between ICD-9-CM And ICD-10-CM To Identify Patients With Common Pain Conditions. The Journal of Pain. 2019;20(12):1429–1445. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.MEPS A. MEPS Survey Background. In. Rockville, Md: 2019. [Google Scholar]
  • 23.Pensa MA, Galusha DH, Cantley LF. Patterns Of Opioid Prescribing And Predictors Of Chronic Opioid Use In An Industrial Cohort, 2003 To 2013. J Occup Environ Med. 2018;60(5):457–461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Mohamed R, Patel J, Shaikh NF, Sambamoorthi U. Absenteeism-Related Wage Loss Associated With Multimorbidity Among Employed Adults In The United States. Journal of Occupational and Environmental Medicine. 2021;63(6). [DOI] [PubMed] [Google Scholar]
  • 25.Belotti F, Deb P, Manning WG, Norton EC. Twopm: Two-Part Models. The Stata Journal. 2015;15(1):3–20. [Google Scholar]
  • 26.Tian L, Huang J. A Two‐Part Model For Censored Medical Cost Data. Statistics in Medicine. 2007;26(23):4273–4292. [DOI] [PubMed] [Google Scholar]
  • 27.Zhang W, Sun H. How To Analyze Work Productivity Loss Due To Health Problems In Randomized Controlled Trials? A Simulation Study. BMC Medical Research Methodology. 2021;21(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Olsen MK, Schafer JL. A Two-Part Random-Effects Model For Semicontinuous Longitudinal Data. Journal of the American Statistical Association. 2001;96(454):730–745. [Google Scholar]
  • 29.Goplerud E, Hodge S, Benham T. A Substance Use Cost Calculator for US Employers With an Emphasis on Prescription Pain Medication Misuse. Journal of Occupational and Environmental Medicine. 2017;59(11):1063–1071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Xu J, Zakari NM, Hamadi HY, Park S, Haley DR, Zhao M. The Financial Burden Of Opioid-Related Abuse Among Surgical And Non-Surgical Patients In Florida: A Longitudinal Study. International Journal of Environmental Research and Public Health. 2021;18(17):9127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fishman MA, Antony AB, Hunter CW, et al. The Cost Of Lost Productivity In An Opioid Utilizing Pain Sample. Journal of Pain Research. 2021:2347–2357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Barker MJ, Greenwood KM, Jackson M, Crowe SF. Cognitive Effects Of Long-Term Benzodiazepine Use: A Meta-Analysis. CNS Drugs. 2004;18:37–48. [DOI] [PubMed] [Google Scholar]
  • 33.Zelaya CE, Dahlhamer JM, Lucas JW, Connor EM. Chronic pain and high-impact chronic pain among U.S. adults, 2019. In: National Center for Health Statistics. Division of H, Nutrition Examination S, National Center for Hiv/Aids VHSTD, Hepatitis TBPDoV, eds. Hyattsville, MD: 2020. [Google Scholar]
  • 34.Rebecca D Kehm WY, Parisa Tehranifar, and Mary Beth Terry. 40 Years Of Change In Age- And Stage-Specific Cancer Incidence Rates In US Women And Men. JNCI Cancer Spectr. 2019;3(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Ali Haider AA, Kristy Nguyen, Rony Dev, Syed Mujtaba Ali Naqvi, Syed Mussadiq Ali Naqvi, Tonya Edwards, Akhila Reddy, Shalini Dalal, Kimberson C. Tanco, Joseph Arthur, Janet L. Williams, Minjeong Park, Diane D. Liu, Seyedeh S. Dibaj, Eduardo Bruera. Concurrent Use Of Opioids With Benzodiazepines Or Nonbenzodiazepine Sedatives Among Patients With Cancer Referred To An Outpatient Palliative Care Clinic. Cancer. 2019. [DOI] [PubMed] [Google Scholar]
  • 36.Devon K Check ANW, Nicole Fergestrom, Katherine E Reeder-Hayes, Joan M Neuner, and Andrew W Roberts. Concurrent Opioid And Benzodiazepine Prescriptions Among Older Women Diagnosed With Breast Cancer. In:2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Jeffery MM HW, Jena AB, Ross JS, Shah ND, Karaca-Mandic P. Rates Of Physician Coprescribing Of Opioids And Benzodiazepines After The Release Of The Centers For Disease Control And Prevention Guidelines In 2016. JAMA Netw Open. 2019;2(8). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.MD WMH. Chronic Pain And Mental Health Disorders: Shared Neural Mechanisms, Epidemiology, And Treatment. Mayo Clin Proc. 2016;7(7):955–970. [DOI] [PubMed] [Google Scholar]
  • 39.National Academies of Sciences E, and Medicine Committee on Medication-Assisted Treatment for Opioid Use Disorder. Barriers To Broader Use Of Medications To Treat Opioid Use Disorder. In: Medications for Opioid Use Disorder Save Lives. Washington (DC): National Academies Press (US); 2019. [PubMed] [Google Scholar]
  • 40.Hyun Jin Song KVB, Debbie L Wilson, Almut G Winterstein, Haesuk Park. Medical Costs And Productivity Loss Due To Mild, Moderate, And Severe Asthma In The United States. Journal of Asthma and Allergy. 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Zhiyuan Zheng KRY, Gery P. Guy Jr, Xuesong Han, Chunyu Li, Matthew P Banegas, Donatus U. Ekwueme, Ahmedin Jemal. Annual Medical Expenditure And Productivity Loss Among Colorectal, Female Breast, And Prostate Cancer Survivors In The United States. Journal of the National Cancer Institute. 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Cawley J, Biener A, Meyerhoefer C, et al. Job Absenteeism Costs Of Obesity In The United States: National And State-Level Estimates. Journal of Occupational and Environmental Medicine. 2021;63. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Digital Content- STROBE
NIHMS1948569-supplement-Supplemental_Digital_Content-_STROBE.docx (32.4KB, docx)

Data Availability Statement

The author confirms that all data generated or analyzed during this study are included in this published article.

RESOURCES