Intended for healthcare professionals

CCBYNC Open access

Off-label indications for antidepressants in primary care: descriptive study of prescriptions from an indication based electronic prescribing system

BMJ 2017; 356 doi: (Published 21 February 2017) Cite this as: BMJ 2017;356:j603

Re: Clarification of issues raised by Dr. Raj Bhui

We thank Dr. Raj Bhui for his responses to our paper. We would like to clarify two issues in his letter that we feel misrepresented our study.

Issue 1: Dr. Bhui made a comment about our use of approved indications at the end of the study period to classify prescriptions as on-label or off-label, stating that “Under this rule, a medication that was on-label when prescribed would be classified as inappropriate/off-label in this study if the approved indication changed over the 12.5 years of the study.”

It is highly unlikely that indications go from being approved to unapproved for a drug over time. In fact, this phenomenon did not occur for any of the drugs in our study. On the contrary, it is much more likely that indications go from being unapproved to approved over time, which indeed occurred for certain antidepressants in Canada over the study period. For example, in 2005, venlafaxine was additionally approved for panic disorder and paroxetine was additionally approved for premenstrual dysphoric disorder. We therefore used the list of approved indications at the end of the study to give more conservative estimates of off-label prescribing that reflected the latest regulatory standards. So, if a medication was off-label in the year it was prescribed but the indication eventually became approved, we classified the prescription as on-label.

Issue 2: We reported in our paper that “If a physician recorded multiple indications for the drug (n=1922, 1.8% of all antidepressant prescriptions), the prescription was classified as off-label only if all the indications were not approved.” In other words, a prescription with multiple indications would be considered as ‘on-label’ if at least one of the indications was approved. We used this classification criterion to identify as many on-label prescriptions as possible.

It seems that Dr. Bhui may have misinterpreted this criterion because he called it an “all or nothing” marking scheme and wrote that “In other words, a medication ordered with multiple indications would be marked as appropriate only if all of the indications were on-label at the end of the study.”

Competing interests: No competing interests

06 March 2017
Jenna Wong
PhD candidate
Department of Epidemiology, Biostatistics, and Occupational Health, McGill University
1140 Pine Ave West, Montreal, QC, H3A 1A3