Data Reporting Error

In the article titled “Association of Opioid Prescriptions From Dental Clinicians for US Adolescents and Young Adults With Subsequent Opioid Use and Abuse,”¹ published in the February 2019 print issue of JAMA Internal Medicine, there was a data reporting error in the Limitations section. Because an immortal time bias was introduced, the dental opioid prescriptions ascertainment period was, on average, reduced by about half in about three-quarters of the participants, or by about three-eighths overall. Because 27% of all subsequent opioid prescriptions came from dental clinicians, it is estimated that the true estimate of opioid prescriptions at 90 to 365 days is approximately 1/(0.73 + 0.27 × 0.622), which is 11% higher than the initially reported 0.1%; however, the revised estimate of 0.111%, when rounded, is still 0.1%, meaning that the absolute risk difference between the exposed and unexposed cohorts remains unchanged at 6.8%. A letter to the editor describes these changes to the Limitations section in more detail.² This article has been corrected online.