With deaths from opioid overdose rising steeply in recent years, and a large segment of the population reporting knowing someone who has been addicted to prescription painkillers, the breadth of the opioid crisis should come as no surprise, affecting people across all incomes, ages, and regions. About four in ten people addicted to opioids are covered by private health insurance and Medicaid covers a similarly large share.
Private insurance covers nearly 4 in 10 non-elderly adults with opioid addiction
In this analysis and a corresponding chart collection, we use claims data from large employers to examine how the opioid crisis has affected people with large employer coverage, including employees and their dependents. The analysis is based on a sample of health benefit claims from the Truven MarketScan Commercial Claims and Encounters Database, which we used to calculate the amounts paid by insurance and out-of-pocket on prescription drugs from 2004 to 2016. We use a sample of between 1.2 and 19.8 million enrollees per year to analyze the change from 2004 to 2016 in opioid-related spending and utilization.
We find that opioid prescription use and spending among people with large employer coverage increased for several years before reaching a peak in 2009. Since then, use of and spending on prescription opioids in this population has tapered off and is at even lower levels than it had been more than a decade ago. The drop-off in opioid prescribing frequency since 2009 is seen across people with diagnoses in all major disease categories, including cancer, but the drop-off is pronounced among people with complications from pregnancy or birth, musculoskeletal conditions, and injuries.
Meanwhile, though, the cost of treating opioid addiction and overdose – stemming from both prescription and illicit drug use – among people with large employer coverage has increased sharply, rising to $2.6 billion in 2016 from $0.3 billion 12 years earlier, a more than nine-fold increase.
Trends in prescription opioid use & spending among people with large employer coverage
Opioid prescription use among people with large employer coverage is highest for older enrollees: 22% of people age 55-64 had at least one opioid prescription in 2016, compared to 12% of young adults and 4% of children. Women with large employer coverage are somewhat more likely to take an opioid prescription than men (15% compared to 12%). Opioid prescription use among people with large employer coverage is also higher in the South (16%) than in the West (12%) or Northeast (11%).
Among people with large employer coverage, older enrollees are more likely to have an opioid prescription
Among people with large employer coverage, the frequency of opioid prescribing increased from 2004 (when 15.7% of enrollees had an opioid prescription) to 2009 (when 17.3% did). After reaching a peak in 2009, the rate of opioid prescribing began to fall. By 2014, the share of people with large employer coverage who received an opioid prescription (15.0%) was lower than it had been a decade earlier, and by 2016, the share was even lower, at 13.6% (a 21% decline since 2009).
The share of people with large employer coverage taking opioid prescriptions is at its lowest levels in over a decade
Among people with large employer coverage, this pattern (of increasing opioid prescription use through the late 2000s, followed by a drop-off through 2016) is similar across most major disease categories. Some of the steepest declines in opioid prescription use since 2009 were among people with complications from pregnancy or childbirth, musculoskeletal conditions, and injuries. The share of people experiencing complications from pregnancy or childbirth who received an opioid prescription peaked in 2007, when 35% received an opioid prescription, but this share dropped to 26% in 2016. Similarly, in 2007, 37% of people with large employer coverage who had a musculoskeletal condition received an opioid prescription, but the share dropped to 30% by 2016. The same decline can be seen among people with large employer coverage who experienced injuries and poisonings (37% in 2009, down to 30% in 2016).
Opioid use declined across disease categories, particularly pregnancy, musculoskeletal diseases, and injuries
We also see a sharp decline in the use of opioid prescriptions among people with cancer diagnoses, particularly in the most recent couple of years. In 2016, 26% of people with large employer coverage who had a cancer diagnosis received at least one opioid prescription, down from 32% in 2007. Despite declines in opioid prescribing for musculoskeletal conditions, people with large employer coverage who have musculoskeletal diagnoses still receive opioid medications more frequently (30%) than those with cancer diagnoses (26%).
Overall in 2016, among those receiving an opioid prescription, a slightly larger share received only a single prescription in that year (61%) than did in 2006, a decade earlier (58%). The average number of prescriptions each person received also rose from 2004 until 2010 and then fell again, but this measure is imperfect because it does not adjust for the length of the supply or the strength of the drug received.
In total, large employer plans and their enrollees spent $1.4 billion in 2016 on opioid prescription painkillers, down 27% from peak spending of $1.9 billion in 2009. In 2016, $263 million, or 19% of total opioid prescription drug spending was paid out-of-pocket by enrollees.
Spending on opioid prescriptions peaked in 2009
Opioid prescriptions have represented a small share of total health spending by large employer plans and enrollees.
Treatment for Opioid Addiction & Overdose among People with Large Employer Coverage
In 2016, people with large employer coverage received $2.6 billion in services for treatment of opioid addiction and overdose, up from $0.3 billion in 2004. Of the $2.6 billion spent on treatment for opioid addiction and overdose in 2016 for people with large employer coverage, $1.3 billion was for outpatient treatment, $911 million was for inpatient care, and $435 million was for prescription drugs. In 2016, $2.3 billion in addiction and overdose services was covered by insurance and $335 million was paid out-of-pocket by patients. (This total only includes only payments for services covered at least in part by insurance, not services that are paid fully out-of-pocket and not billed to insurance, so it is likely an undercount of opioid addiction and overdose treatment expenses by this population.)
The cost of treating opioid addiction and overdose has risen even as opioid prescription use has fallen
Spending on treatment for opioid addiction and overdose represents a small but growing share of overall health spending by people with large employer coverage. In 2016, treatment for opioid addiction and overdose represented about 1% of total inpatient spending by people with large employer coverage and about 0.5% of total outpatient spending. In 2004, treatment for opioid addiction and overdose represented about 0.3% of total inpatient spending and less than 0.1% of total outpatient spending. On average, inpatient and outpatient treatment for opioid addiction and overdose added about $26 per person to the annual cost of health benefits coverage for large employers in 2016, up from about $3 in 2004.
The bulk of the total $2.6 billion in spending for treatment of opioid addiction and overdose among people with large employer coverage was treatment for young adults, totaling $1.6 billion in 2016, even though young adults are prescribed opioids less often than older adults. Males also used more treatment than women ($1.6 billion vs $1.0 billion).
Spending on opioid addiction and overdose treatment is mostly concentrated among younger people
The bulk of spending by people with large employer coverage on inpatient and outpatient treatment for opioid addiction and overdose was for employees’ children (53%) or spouses (18%), while just under a third (29%) was for employees themselves.
Among people with large employer coverage who had outpatient spending on treatment for opioid addiction and overdose, their average outpatient expenses totaled $4,695 (of which $670 was paid out-of-pocket) in 2016. Among those with inpatient spending on treatment for opioid misuse, their average inpatient expenses totaled $16,104 (with $1,628 paid out-of-pocket) in 2016. On average, inpatient expenses have risen sharply, up from $5,809 in 2004.
In 2016, 342 people per 100,000 large group enrollees received treatment for opioid overdose or addiction, including 67 people per 100,000 who received treatment in an inpatient setting.
Among people with large employer coverage, utilization of opioid prescription painkillers has declined somewhat in recent years. Use of and spending on prescription opioids by this group peaked in 2009 and has since dropped to the lowest levels in over a decade. Across most major disease categories, we see a similar pattern of the frequency of opioid prescription use rising until the late 2000s and then declining through 2016.
Despite declining rates of opioid prescribing to those with employer coverage, spending on treatment for opioid addiction and overdose has increased rapidly, potentially tied to growing illicit use and increased awareness of opioid addiction. Opioid addiction and overdose treatment – the bulk of which is for dependents of employees – represents a small but growing share of overall employer health spending.
We analyzed a sample of claims obtained from the Truven Health Analytics MarketScan Commercial Claims and Encounters Database (Marketscan). The database has claims provided by large employers (those with more than 1,000 employees); this analysis does not include opioid prescription or addiction treatment for other populations (such as the uninsured or those on Medicaid or Medicare). We used a subset of claims from the years 2004 through 2016. In 2016, there were claims for almost 20 million people representing about 23% of the 85 million people in the large group market. Weights were applied to match counts in the Current Population Survey for large group enrollees by sex, age, state and whether the enrollee was a policy holder or dependent. People 65 and over were excluded.
Over 14,000 national drug codes (NDC) were defined as opiates. In general, we defined “prescription opioids” as those with a primary purpose of treating pain. Only prescriptions classified under the controlled substance act are included. We excluded from this category Methadone, Suboxone (Buprenorphine with Naloxone), and other drugs commonly used to treat addiction. We also excluded medications not commonly prescribed (such as Pentazocine). Each opiate script was counted as a single prescription regardless of the quantity or strength of that prescription. The Marketscan database only includes retail prescriptions administered in an outpatient setting. Disease categories are defined by AHRQ’s chronic condition indicators, and based on the diagnosis an enrollee receives.
In our analysis of opioid addiction and overdose treatment, we include medications used to treat overdose (e.g. Naloxone) and drugs used to treat addiction (e.g. Methadone and Suboxone). We also include inpatient and outpatient medical services to treat opioid addiction or overdose, identified by ICD-9 and ICD-10 diagnosis codes. Midway through 2015, Marketscan claims transitioned from ICD-9 to ICD-10. While both systems classify diagnoses, there is no precise crosswalk between the two. In consultation with a clinician, we selected both ICD-9 and ICD-10 codes which are overwhelmingly used for opioid addiction or signify misuse. A list of these ICD codes is available upon request. Because of the change in coding systems, it is not possible to tracks trends between 2014 and 2016. Diagnoses related to heroin abuse were included as opiate abuse.
Because there is no precise way to identify costs associated with opioid addiction and overdose treatment, some of our rules for inclusion lead to an underestimate, while others lead to an overestimate. In general, we elected a conservative approach. For example, in some cases, opioid abuse diagnoses may be classified under a broader drug abuse diagnosis and therefore are not captured. Additionally, we do not include the costs associated with diagnoses that commonly arise from opioid abuse, such as respiratory distress or endocarditis, unless an opioid abuse diagnosis was also present. However, if a claim included an opioid abuse diagnosis along with other diagnoses, we included spending for all procedures during that day, even if some of those interventions were to treat concurrent medical conditions unrelated or indirectly related to opioid abuse. If an enrollee paid fully out-of-pocket and did not use their insurance coverage, this spending is also not included. Overall, we think these assumptions lead to an underestimate of the costs associated with opioid addiction and overdose treatment for the large employer coverage population.