|| Checking for direct PDF access through Ovid
To determine (1) the natural history of prescription opioid use, (2) the predictors of long-term opioid use, and (3) the association between opioid dose and pain and function in a large cohort of workers with recent back injuries.Prospective cohort of workers with back injuries (N=1883) interviewed 18 days (median) and 1 year after claim submission. Detailed pharmacy data were obtained from computerized records of paid bills.Forty-two percent of workers (781/1843) received an opioid in the year after injury, most (694/781, 89%) at or shortly after the first medical visit for the injury. Of these, most (410/694, 59%) received opioids only within the first quarter after injury, whereas 16% (111/694) received opioids for 4 quarters. Among these long-term users, total morphine equivalent dose (MED) increased significantly (P<0.01) from the first (mean, 2364 mg; standard deviation, 4019 mg) to the fourth (mean, 3824 mg; standard deviation, 5998 mg) quarter. Improvement by at least 30% in pain and function measure scores occurred in only 26% (95% confidence interval 18%-36%) and 16% (95% confidence interval 10%-25%), respectively, of long-term users. Opioid doses increased substantially over time in all but those in whom function improved. After adjustment for baseline pain, function, and injury severity, the strongest predictor of longer term opioid prescription was total MED in the first quarter. Workers receiving at least 40 mg MED per day in the first quarter had 6-fold odds of receiving longer-term opioids.For the small group of workers with compensable back injuries who receive opioids longer-term (111/1843, 6%), opioid doses increase substantially and only a minority shows clinically important improvement in pain and function. The amount of prescribed opioid received early after injury strongly predicts long-term use. More research is needed to understand clinical decisions to continue or increase opioid therapy after back injury.