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: Comments from Dr. Levin and colleagues

We thank Dr. Levin and colleagues for sharing their thoughts on our article and would like to respond to several of their comments.

Regarding patient comorbidities and the choice of diagnosis recorded on the e-prescription, there are many good reasons to believe that diagnoses recorded for antidepressants in the MOXXI system were indeed accurate and did not overestimate off-label use. First, our finding that 1.8% of antidepressant prescriptions in our study had multiple indications recorded is only slightly lower than another UK study (1) that found only 5% of antidepressant users reported taking antidepressants for multiple indications. Second, the drop-down menu of indications in the e-prescribing system lists on-label and off-label indications for a drug in alphabetical order. Thus, physicians would have to intentionally scroll past ‘depression’ to select either ‘insomnia’ or ‘pain’. For this reason, we argue that in cases where a physician suspected depression was present but recorded a comorbid condition instead, the physician may have judged that the depression was not severe enough to meet a clinical diagnosis of depression (and thus the prescription would have still been off-label even if depression had been recorded). Finally, MOXXI physicians have good reason to enter all relevant indications on the e-prescription because a) these conditions are added to the patient’s problem list where they are used to document the treatment history for the problem as well as generate alerts if potential drug-disease problems are identified, and b) MOXXI physicians operate within a universal, publicly-funded health care system where the presence of a psychiatric diagnosis in the patient’s health record does not affect physicians’ ability to obtain reimbursement or jeopardize the patient’s ability to obtain health insurance.

Regarding intra-class prescribing, Levin and colleagues wrote that, “We teach our medical students, residents and fellows that all antidepressants have similar efficacy for treating depression, and are interchangeable, side effect profile aside. This is a broadly accepted principle in psychiatry.” Yet at the same time, clinical guidelines for anxiety-related disorders such as the one by the British Association for Psychopharmacology (2) warn against assuming class effects without empirical evidence:

“The selection of a particular drug class (and of a specific drug within that class) should be determined principally by the evidence base supporting its use, and also by whether the patient has previous experience of treatment with that compound. The absence of a licensed indication does not necessarily mean an absence of evidence for the proposed treatment intervention: conversely it should not be assumed that all drugs within a class are likely to be efficacious in the treatment of a particular anxiety disorder, when one member of that class has proven efficacy.” (page 9)

Moreover, in addition to cerivastatin, there are more examples in the literature showing that within-class differences can exist between drugs. For example, Schrijnders et al. (3) discuss the unique side-effect profile of glibenclamide compared to other sulfonylureas and highlight that because important and clinically relevant within-class differences exist among the sulfonylureas, conclusions about a single drug should not be extrapolated to all drugs in the same class. Thus, even though it may be common practice in psychiatry to assume class effects, this practice does not align with either recommendations from clinical guidelines or the principles of evidence-based medicine.

Which brings us to their concluding comment. Levin and colleagues suggested that our study design should have combined on-label prescriptions with both evidence-based off-label prescriptions and off-label prescriptions with evidence extrapolated from other drugs within the same class. However, it is largely because of this exact debate about class effects that we chose to present these categories separately. In our opinion, it would have been misleading to report that 55.5% of off-label prescriptions were evidence-based when nearly three-quarters of these prescriptions would have been borrowing evidence from another drug in the same class.

Finally, we would like to highlight that our intention was not to condone intra-class prescribing. Indeed, there may be cases where physicians intentionally prescribe a certain drug because the patient was previously unsuccessful with another compound in the class. Rather, our intentions were to a) show that this phenomenon occurs frequently, b) encourage physicians to first prescribe the more evidence-based drugs in a class before prescribing the others, and c) advocate for more evidence evaluating the appropriateness of intra-class prescribing for antidepressants.

Jenna Wong, MSc, PhD Candidate
Robyn Tamblyn, PhD, MSc, MScN

1. Aarts N, Noordam R, Hofman A, Tiemeier H, Stricker BH, Visser LE. Self-reported indications for antidepressant use in a population-based cohort of middle-aged and elderly. Int J Clin Pharm. 2016 Oct;38(5):1311–7.
2. Baldwin DS, Anderson IM, Nutt DJ, Allgulander C, Bandelow B, den Boer JA, et al. Evidence-based pharmacological treatment of anxiety disorders, post-traumatic stress disorder and obsessive-compulsive disorder: a revision of the 2005 guidelines from the British Association for Psychopharmacology. J Psychopharmacol Oxf Engl. 2014 May;28(5):403–39.
3. Schrijnders D, Kleefstra N, Landman GWD. Within-class differences of the sulfonylureas should be accounted for. Diabetologia. 2015 Jun 1;58(6):1374–5.

Competing interests: No competing interests

27 April 2017
Jenna Wong
PhD candidate
Robyn Tamblyn, PhD, Scientific Director – CIHR Institute of Health Services and Policy Research, Professor – Department of Epidemiology and Biostatistics, McGill University
Department of Epidemiology, Biostatistics, and Occupational Health, McGill University
1140 Pine Ave West, Montreal, QC, H3A 1A3