Treatment indications and potential off‐label use of antidepressants among older adults: A population‐based descriptive study in Denmark

Abstract Objectives Off‐label prescriptions of antidepressants may be of special concern in older‐adults. We aimed to study the potential off‐label use of antidepressants among adults ≥65 years by describing the patterns, trends, and factors associated with missing and unspecified treatment indications. Methods We used registry data to describe indications of all antidepressant prescriptions (N = 13.8 million) redeemed by older‐adults in 2006–2019. We investigated factors associated with off‐label use by considering prescriptions with missing and unspecified indications of the first antidepressant prescription using a multinomial logistic regression with the ‘depression’ indication as a reference category and reported odds ratios (ORs) with 95% confidence intervals (CI). Results Overall, 18.1% of all antidepressant prescriptions had missing indications, and 9.9% had unspecified indications. The proportion of potential off‐label use based on missing and unspecified prescriptions remained mostly consistent during 2006–2019. We identified similar associations in user characteristics whether considering missing or unspecified first prescription. ORs with 95% CI were raised in non‐western ethnicity (vs. Danish, 1.12 (0.99–1.26) for missing indication and 1.28 (1.11–1.48) for unspecified indication) and female sex (vs. male, 1.05 (1.02–1.07) and 1.05 (1.02–1.07) respectively). ORs were reduced for shorter educational attainment (vs. long, 0.90 (0.87–0.94) and 0.92 (0.88–0.96)), older age (≥81 vs. 67–70 years, 0.66 (0.65–0.71) and 0.73 (0.70–0.76)) and hospital psychiatric diagnosis (per diagnosis 0.76 (0.73–0.78) and 0.88 (0.86–0.91)). Conclusions Nearly one‐third of all antidepressant prescriptions redeemed by older‐adults in Denmark had either missing or unspecified treatment indications. Whether these prescriptions were actual off‐label use needs to be validated. Clinicians should pay special attention to patients' characteristics linking missing and unspecified indications and maintain adequate documentation while prescribing medication.


| INTRODUCTION
Major depression (depression) is the most common mental disorder in the older population (≥65 years), with a prevalence of up to 10%. 1 Antidepressants are the first-line pharmacological treatment for depression and depressive symptoms in all ages. Antidepressants are also used for treating other mental disorders, including panic disorder, social phobia, generalized anxiety disorder, obsessivecompulsive disorder (OCD), post-traumatic stress disorder, and specific types of pain. 2 Although the use of antidepressants is considered beneficial, 3 over half of older adults do not respond to initial antidepressant treatment, 4 and some may experience side effects. 5 Besides, antidepressants also increase the risk of falls, hyponatremia, stroke, and other complications. [4][5][6] Moreover, age-related pharmacokinetics and pharmacodynamics might alter the benefits of antidepressant use. 7 Although it is necessary to ensure older adults' access to antidepressant treatment, ensuring appropriate use is paramount. 8 Globally, there has been a surge in the use of antidepressants in recent decades. 9 The growing trend of unapproved/off-label use of antidepressants could be a contributing factor, among others. 10 Offlabel medication can be defined as using a specific medicine for a purpose or in a dosage not in compliance with the authorized product information. Off-label use of antidepressant prescriptions in primary care has been reported as high as 50% in the general population in Canada. 11 Like-wise, in older adults, a recent German study reported that more than 40% of all antidepressant prescriptions in 2009-2015 were off-label use. 12 We are not aware of similar data from Denmark.
However, using a proxy measure (discharge diagnosis of hospital inand outpatient contacts), recent Danish studies reported a high level of off-label use in the general population. 13,14 Nonetheless, in Denmark, although the use of antidepressants has gradually decreased in recent years in all age groups, including older adults, 15 it remains among the countries with the highest use of antidepressants. Clinicians in Denmark, including general practitioners (GPs), can prescribe off-label medications in the context when no other suitable option exists within the approved medications and with the obligation of patient caring and conscientiousness in mind. 16 When prescribing a medication, the prescriber selects an indication code from a list of approved indications or writes free text if the indication is out of the list or opts to leave the indication as empty. 17 The free text indications are yet to be ready for use in research and are coded as unspecified, and the empty indications are coded as missing indications. As a rough measure, one way to assess potential off-label by indication could be studying these prescriptions with missing and unspecified treatment indications.
Besides clinical factors, patients' socio-demographics play a critical role in the treatment because patient-related factors influence access to care, awareness of health, and decision-making capacity in the treatment process. 18 Similarly, patient-related factors may also influence the treatment indications for which one uses antidepressants, even though prescription indications should corroborate the patients' clinical manifestation of the illnesses.
Identifying and acknowledging patient-related factors concerning treatment indication of antidepressants are essential to point the direction for future interventions targeting potential sub-groups in older adults with over-or underusing of antidepressant treatments. 8 However, no recent studies have yet focused on the treatment indications of antidepressants and the potential associated factors in Danish older adults, despite the fact that the use of off-label medications is widespread in Denmark, and older adults are more vulnerable to adverse-drug-reactions from the potential unapproved drug use.
We, therefore, aimed to describe the patterns and trends of treatment indications of antidepressants redeemed by older adults at community pharmacies in Denmark and explore factors associated with potential off-label antidepressant use by considering prescriptions with missing and unspecified treatment indications. where all persons alive and residing in Denmark are registered. 19 CRS includes information on the date of birth, sex, marital status, and a unique personal identification number (CPR). 19 Using CPR, we were able to link data from the Danish National Prescription Register (DNPR), which covers information on all prescriptions redeemed at community pharmacies from 1995 onwards in Denmark. 20 DNPR contains information on dispensed drug names, the ATC Classification codes, indications, and dispensation dates. 20 We obtained socio-demographic information from the Population Statistics Register, and the Population's Education Register, 21 and clinical factors (e.g., date and diagnoses of hospital contacts) from the Danish National Patient Register. 22

| RESULTS
We identified 462,657 older adults who had redeemed a total of 13,

| Frequencies and pattern of treatment indications and off-label use
The most common indication for all antidepressants was depression (58.4%), followed by tranquillizers (7.2%), anxiety (5.7%), pain

| Factors associated with missing and unspecified indications for the first prescription with antidepressants
Characteristics associated with a higher likelihood of missing indication than depression were female sex (vs. male: OR 1.05, 95% CI:

| Main findings
In this descriptive study, we found depression was the most common

| Patient-related factors and missing and unspecified indication for the first antidepressant prescription
Most of the studies on the indication of antidepressants did not include socio-demographic and clinical information. Only a few included socio-demographics but did not analyze them as a determinant. Previous studies reported that females tend to use antidepressants without reporting depression, whereas males are less likely to use antidepressants while having depression symptoms. 33,34 Moreover, females are more positive about seeking mental health care than males, suggesting higher use of treatment in females than males. 35 Although we found being female was associated with higher odds of missing/unspecified indications, the point estimate was small and was reduced to null when TCAs were excluded from the analysis. Individuals who are married or living with partners are more likely to seek health care and treatment than those who live alone or separated. 36 This could be one of the explanations that in our data, married/living with partners had higher odds of receiving prescriptions with missing indications since they might be prone to use antidepressants for uncommon purposes, or they might have very mild symptoms to be labeled with depression or similar diagnosis.
Like-wise, this presumably also explains the lower odds of unspecified prescription use in individuals who were widows/widowers/ separated.
Older age, short education, living in a city, and past psychiatric diagnoses are often considered risk factors for depression in older adults. 37  It is not surprising that the older adults with past psychiatric diagnoses were less likely to receive a prescription with missing/ unspecified indication than for depression, suggesting those with missing/unspecified indications of antidepressants might have less severe depression/psychiatric conditions. As we included prevalent prescription users, the cases in the early years are likely to be prevalent cases, whereas the cases in the later years as more incident cases. Thus, the observed calendar year effect concerning missing/unspecified indications is conceivably to reflect the prevalent versus incident case differences rather than an actual effect of the calendar year.

| Strength and limitations
The main strength of our study lies in its register-based design, with a large representative study population. The DNPR has widely been used in scientific research as a sole source of information on dispensed prescriptions in the primary care pharmacies in Denmark and has been shown to be highly accurate. 20 We included wideranging socio-demographical and clinical information from highly accurate national registers, and missing information was negligible.
The study has several limitations. First, the DNPR does not include information on diagnoses linking to each prescription. The There may be a difference in the treatment indications used by GPs/ private psychiatrists and by the prescribers at the hospitals. A Danish study assessed that up to 90% of all depression treatments initiated at non-hospital based settings. This unmeasured factor is to some extent reflected in past hospital contacts, which we included in the analyses. Therefore, prescriber differences might not have a substantial impact on our results. Moreover, we included previous so-

CONFLICT OF INTEREST
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
The data used in this study are from Statistics Denmark. They are stored in an anonymous form and are not publicly available. To use such data for research, require special permission in compliance with the Danish Data Privacy Act.

ETHICS STATEMENT
The study was approved by the Danish Data Protection Agency, and no further ethical approval is required regarding register-based research in Denmark.