Low pain intensity after opioid withdrawal as a first step of a comprehensive pain rehabilitation program predicts long-term nonuse of opioids in chronic noncancer pain.

OBJECTIVES: In specialized pain clinics there is an increasing number of patients with severe chronic noncancer pain (CNCP) despite long-term opioid medication. Few clinical studies show short-term pain relief after opioid withdrawal (OW). We have evaluated the relation between pain intensity after OW and long-term opioid nonuse. METHODS: One hundred two consecutive patients with severe CNCP despite opioid medication (mean treatment duration, 43 mo) reported pain intensity (numerical rating scale, 0 to 10), Pain Disability Index, mood (CES-D), and quality of life (Short Form 36) before, shortly, and 12 to 24 months after inpatient OW. Total opioid withdrawal (n = 78) or significant dose reduction (DR; n = 24, mean reduction, 82%) was performed after individual decision. Opioid intake 12 to 24 months later, respectively dose increase ≥ 100% (DR group), was considered relapse. T tests, multivariable analysis of variance, logistic regression. RESULTS: After OW current pain intensity significantly decreased on an average by 41% (6.4 ± 2.4 vs. 3.8 ± 2.5), maximal and average pain by 18% and 24%, respectively. Twelve to 24 months later 42 patients (41%) relapsed (31 of the total opioid withdrawal group, 6 of the DR group, 5 lost). Patients without later relapse showed significantly lower pain scores than the later relapsed patients already shortly after OW (5.0 ± 2.2 vs. 5.9 ± 2.1) and 12 to 24 months later (5.5 ± 2.4 vs. 6.5 ± 2.0). There was a significant relation between relapse probability and pain intensity immediately after OW. CONCLUSIONS: In many patients with severe CNCP, despite opioid medication, sustainable pain relief can be achieved if OW is included in the rehabilitation program. Consequently, we recommend OW for opioid-resistant CNCP before any opioid escalation. Lower pain intensity shortly after OW may predict the long-term opioid nonuse probability.