Table 3.

Modified intent-to-treat analysis for the ad hoc subgroup analysis of participants with severe pain and insomnia symptoms at baseline. Multiple imputation was used (10 imputations) to accommodate missing data at baseline and 18-mo follow-up