Background: Chronic pain is implicated as a risk factor for illicit opioid use among patients
with opioid addiction treated with methadone. However, there exists conflicting evidence that
supports and refutes this claim. These discrepancies may stem from the large variability in pain
measurement reported across studies.
Objectives: We aim to determine the clinical and demographic characteristics of patients
reporting pain and evaluate the prognostic value of different pain classification measures in a
sample of opioid addiction patients.
Study Design: Multi-center prospective cohort study.
Setting: Methadone maintenance treatment facilities for managing patients with opioid
addiction.
Methods: This study includes participants from the Genetics of Opioid Addiction (GENOA)
prospective cohort study. We assessed the prognostic value of different pain measures for
predicting opioid relapse. Pain measures include the Brief Pain Inventory (BPI) and patients’
response to a direct pain question all study participants were asked from the GENOA case
report form (CRF) “are you currently experiencing or have been diagnosed with chronic pain?”
Performance characteristics of the GENOA CRF pain measure was estimated with sensitivity and
specificity using the BPI as the gold standard reference. Prognostic value was assessed using pain
classification as the primary independent variable in an adjusted analysis using 1) the percentage
of positive opioid urine screens and 2) high-risk opioid use (≥ 50% positive opioid urine screens)
as the dependent variables in a linear and logistic regression analyses, respectively.
Results: Among participants eligible for inclusion (n = 444) the BPI was found to be highly
sensitive, classifying a large number of GENOA participants with pain (n = 281 of the 297
classified with pain, 94.6%) in comparison to the GENOA CRF (n = 154 of 297 classified with
pain, 51.8%). Participants concordantly classified as having pain according to the GENOA
CRF and BPI were found to have an estimated 7.79% increase in positive opioid urine screens
(estimated coefficient: 7.79; 95%CI 0.74, 14.85: P = 0.031) and a 4 times greater odds (odds
ratio [OR]: 4.10 P = 0.008; 95%CI: 1.44, 11.63) of engaging in a “high risk” level of illicit
opioids use. The prognostic relevance of pain classification was not maintained for the additional
participants classified by the BPI (n = 143 discordant).
Conclusion: These results suggest that while the BPI may be more sensitive in capturing pain
among patients with opioid addiction, this tool is of less value for predicting the impact of pain
on illicit opioid use for opioid addiction patients on methadone maintenance treatment. The
GENOA CRF showed high predictive ability, whereby patients classified according to the GENOA
CRF are at serious risk for opioid relapse. Using the appropriate tool to assess pain in opioid
addiction may serve to improve the current detection and management of comorbid painLimitations: We caution the interpretation of these result since they are still
reflective of participants already maintained on an opioid substitution therapy
(OST), which can largely differ from patients who drop out of methadone
maintenance treatment (MMT) or never seek treatment altogether.
Key words: Chronic pain, opioid adiction, methadone maintenance treatment,
relapse, addiction, measurement, brief pain inventory