Relationship of negative affect and outcome of an opioid therapy trial among low back pain patients.

Objectives: Patients with chronic noncancer pain frequently report symptoms of depression and anxiety (negative affect), which are associated with higher ratings of pain intensity and a greater likelihood of being prescribed chronic opioid therapy. The purpose of this secondary analysis was to test the hypothesis that initial levels of negative affect can predict treatment-related outcomes in a double-blind, placebo-controlled study of extended-release (ER) hydromorphone among opioid-tolerant patients with chronic low back pain.

Methods: Four hundred fifty-nine (N = 459) patients participated in the titration/conversion phase of a multicenter study, of which 268 were randomized to receive once-daily hydromorphone or placebo. All patients completed the Hospital Anxiety and Depression Scale (HADS) at baseline and were divided evenly into Low (N = 157), Moderate (N = 155), and High (N = 147) negative affect groups based on their scores. Group differences in numerical pain intensity measures at home and in the clinic, Roland-Morris Disability ratings, and measures of symptoms from the Subjective Opiate Withdrawal Scale (SOWS) throughout the trial were analyzed.

Results: Two hundred sixty-eight of the initial 459 subjects who entered the 2 to 4-week titration/conversion phase (pretreatment) were successfully randomized to either placebo or ER hydromorphone; a total of 110 patients then completed this double-blind phase of the study. Those in the Moderate and High negative affect groups tended to drop out more often during the titration/conversion phase because of the adverse effects or lack of efficacy of their prescribed opioid than those in the Low negative mood group (P < 0.05). Overall, those patients in the Moderate and High groups reported significantly higher pain intensity scores in at-home and in-clinic pain intensity ratings (P < 0.05), greater disability on the Roland-Morris Scale (P < 0.01), and more withdrawal symptoms on the SOWS (P < 0.05) than those in the Low group. Higher negative affect scores also predicted less favorable ratings of the study drug during the titration phase (P < 0.05). Interestingly, the High negative affect group showed the most improvement in pain in the placebo condition (P < 0.05).

Conclusions: Negative affect is associated with diminished benefit during a trial of opioid therapy and is predictive of dropout in a controlled clinical trial.