BACKGROUND/RATIONALE:
Pain is common among patients receiving substance use disorder (SUD) specialty care. Opioid medications are often prescribed for pain, but there has been considerable unease among policymakers and clinicians regarding the safety of long-term opioid use for patients with current or past SUDs. A key adverse safety outcome related to the use of opioids for pain is opioid overdose, the rate of which has increased substantially in the United States in the past ten years. It is known that, among individuals prescribed opioids for pain, those with SUDs are at an increased risk for opioid overdose, but the extent to which specific treatment practices and other characteristics are associated with increased or decreased risk of overdose for individuals in SUD care with pain is not well understood.

OBJECTIVE(S):
The aims were: 1) To describe, for patients with pain receiving SUD specialty care, the prevalence of high-dose opioid prescribing (100 morphine-equivalent mg/day [MEM/d] or more), polypharmacy of prescription opioids with other sedating medications, and simultaneous opiate agonist treatment with opioid therapy for pain; 2) To examine, for patients receiving SUD specialty care and opioid therapy for pain, factors associated with unintentional overdose mortality; 3) To conduct a chart review of unintentional and intentional overdose decedents who had been in SUD treatment and received opioids for pain in the six months before death.

METHODS:
Secondary data analyses were based on medical records extractions and included VHA patients receiving SUD treatment and opioid therapy for pain during fiscal years 2000 to 2009. A nested case control designed resulted in 178 cases (died of an opioid overdose based on National Death Index data) and 707 matched controls (alive on the date of the case's death/index date). Two sets of detailed chart reviews were conducted for patients who died of an overdose between fiscal year 2000 and 2009 and who had been in SUD treatment and who received opioids for pain; one set for unintentional (n=122) and one for intentional (n=64) overdoses deaths.

FINDINGS/RESULTS:
Approximately 29% of cases and 14% of controls were prescribed 100 MEM/d or more during the 90 days prior to the index date. Concurrent use of benzodiazepines, anti-convulsants and anti-depressants was common; 54% of cases and 27% of controls had been prescribed a benzodiazepine, 35% percent of cases and 17% of controls has been prescribed an anti-convulsant, and 83% of cases and 70% of controls had been prescribed an anti-depressant in the 90 days prior to the index date. In regards to opiate agonist treatment (OAT), 8% of cases and 12% of controls has received OAT in the 90 days prior to the index date.

A series of multivariable conditional logistic regression models (adjusting for race, ethnicity, and age) examined the association between specific factors with case status. Having an inpatient mental health visit in the prior 90 days was associated with increased risk of opioid overdose (Odds Ratio [OR]=2.2); specialty pain treatment and OAT were not significantly associated. For past 90 days medication use, 100 MEM/d or greater (OR=1.5), concurrent benzodiazepine use (OR=2.4), and concurrent anti-convulsant use (OR=2.1) were significantly associated with increased risk of opioid overdose. Back pain was associated with opioid overdose (OR=1.7) but acute pain was not. Additionally, depressive disorders (OR=1.7) and personality disorders (OR=1.8) were significantly associated with opioid overdose, but other psychiatric disorders were not. Among specific substance use disorders, opioid use disorders were significantly associated with opioid overdose (OR=1.6), but other disorders were not. Hepatitis C diagnosis (OR=1.5) and Charlson comorbidity score of 2 or greater, compared to 0 (OR=1.7), were also associated with increased risk of overdose.

Chart review information provided greater detail on treatment experiences in the six months prior to overdose death. For VHA patients in SUD treatment and receiving opioid therapy who died by unintentional opioid overdose, 71% had detailed SUD treatment plans in their records; of those, the plan mentioned opioid analgesic misuse 48% of the time. Indicators of problems with opioid use were not uncommon - 26% had had their opioid medication discontinued, 13% had been given an opioid contract, and 58% had notes indicating provider concerns about opioid misuse (such as taking more than prescribed, using opioids obtained from non-medical sources, and psychological dependence). Of the 63% who had had a urine drug screen, 53% had tested positive for one or more drugs. In 47% of records, there was mention of a lifetime history of non-fatal overdose, and in 39% of records, a lifetime history of suicide attempt was mentioned. The medical record noted a period of abstinence from some or all substances for 50% of the sample, and among those, 52% had a note indicating relapse. Additionally, 16% had been imprisoned at some point and 43% had a record of housing instability.

Among patients in SUD treatment and receiving opioids for pain who died by intentional opioid overdose (suicide), 45% had received a suicide risk assessment (62% positive) and 63% had received a depression screen (78% positive). A lifetime history of non-fatal overdose was noted for 39%, and a lifetime history of suicide attempt was noted for 48%. Of those with a suicide attempt, 31% were on a facility "high risk" list.

IMPACT:
VHA has a number of initiatives underway to decrease the risk of overdose; a greater understanding of patient and treatment factors associated with overdose will help to inform risk stratification efforts. The results from this study provide greater details about which patients are at the greatest risk for overdose among those with pain and in specialty SUD care, as well as a greater understanding of where and how they come into contact with different aspects of the VHA care system.

---

External Links for this Project

---

PUBLICATIONS:

None at this time.

DRA: Substance Use Disorders
DRE: Epidemiology
Keywords: none
MeSH Terms: none