If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Universidad de Buenos Aires, Facultad de Medicina, 1ª Cátedra de Farmacología, Buenos Aires, ArgentinaUniversidad Favaloro, Facultad de Medicina, Buenos Aires, Argentina
An exhaustive review of the main worldwide psychiatry-related literature sources shows that the present antidepressants’ classification does not have a logical and epistemic nomenclature oriented to allow a quick recognition of main adverse drug reactions (ADRs) and pharmacodynamic interactions.
Methods
The study was performed in two phases. In the first, a new classification of antidepressants was built up based on their mechanisms of action. Furthermore, relevant ADRs and pharmacodynamic interactions were grouped according to their causal mechanisms.
In the second phase a comparative, prospective, longitudinal, experimental and randomized study was performed.
312 physicians who were great prescribers of antidepressants were randomly assigned to one of two groups, A and B, having 156 physicians each. Each group was assessed with a questionnaire evaluating basic knowledge of the most important ADRs and pharmacodynamic interactions. This questionnaire was provided before and after the assignment of a standard classification (group A) or the new classification (group B). In the questionnaire provided after the assignments, some questions about different acceptance variables were included.
Results
After handing the classifications, significant differences were found (p = 0.0008) in the number of correct answers in the second questionnaire, in favor of group B. In addition, analysis of acceptance variables showed significant differences between both groups, in favor of the new classification.
Conclusion
This study shows that the new classification of antidepressants allows, in contrast with the standard classifications, to quickly inform and enable physicians to easily relate each drug to important ADRs and pharmacodynamic interactions.
Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: Section 3. Pharmacological treatments.
World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for biological treatment of unipolar depressive disorders, part 1: update 2013 on the acute and continuation treatment of unipolar depressive disorders.
], shows that the currently used antidepressants’ classification does not have a logical and epistemic nomenclature oriented to allow a quick recognition of main adverse drug reactions (ADRs) and relevant pharmacodynamic interactions.
In recent years, new classifications have been proposed. However, none of them has been universally accepted yet, as described in the World Federation of Societies of Biological Psychiatry (WFSBP) guidelines (2017) for unipolar depressive disorders treatment [
], clarifies the meaning of the concepts “logical” and “epistemic”:
•
Logical: The classification must be exhaustive (all elements must be included) and exclusive (elements must belong to only one category).
•
Epistemic: The classification must have a unique criterion, useful for the purposes sought and with heuristic fertility. In turn, the term “heuristic fertility” implies that it serves to achieve or generate knowledge.
At present Neuroscience-Based Nomenclature, the second edition (NbN2) has developed a systematic approach based on the medication’s mode of action which has improved the aforementioned problem [
The NbN’s main purpose is to reduce confusion when psychiatric medication is prescribed. In order to avoid arbitrary descriptors based on indication, the NbN aims to change this approach and relate the nomenclature with the medication’s mode of action [
]. However, this classification presents different limitations that are listed below:
1.
Some of the mechanisms of action of antidepressants are incompletely mentioned, e.g. desipramine and nortriptyline are mentioned as noradrenergic modulators. However, these drugs also act with lesser potency, as serotonergic modulators. Considering both actions is important, among other reasons, to link each drug with important adverse drug reactions (ADRs) and relevant “pharmacodynamic” interactions.
2.
Drugs are not grouped according to their specific mechanisms of action, e.g. they are grouped by their general mode of action (reuptake inhibitor) or by their pharmacological target (e.g. serotonin) separately, but not as Selective Serotonin Reuptake Inhibitors (SSRIs), which makes it confusing.
3.
Certain “multimodal” antidepressants are not included in this class. With current knowledge, drugs with action on more than one molecular target of different classes related to the antidepressant effect (such as transporters, receptors or enzymes) should be included as multimodal.
4.
Some “warnings and precautions” that are highlighted by the FDA (U.S. Food and Drug Administration) at the beginning of the full prescribing information are not considered.
5.
Like the rest of the current classifications, it does not provide the “heuristic fertility” that is essential to achieve or generate knowledge.
Although the new classification proposed, unlike the NbN, covers only antidepressant medications and not all psychotropic drugs, the limitations mentioned above are resolved, as will be seen in the development of this paper. On the other hand, the aims of this new classification are different from the main purpose mentioned for the NbN, as will also be described.
In 2017, the WHO (World Health Organization) published that depression is already the main cause of disability worldwide, being among the main global public health problems [
]. For this reason, instruction to the medical community about the basic management of these drugs through a new classification that allows to quickly recognize certain ADRs and interactions that can be risky, acquires great relevance.
1.1 Aims and hypotheses
1.1.1 Hypothesis
A new classification of existing antidepressants with a logical and epistemic denomination, based on their mechanisms of action, will allow, unlike standard classifications, to quickly instruct highly prescribing physicians of antidepressants, to link each drug with important ADRs and relevant “pharmacodynamic” interactions.
1.1.2 Aims
1st To develop a logical and epistemic classification of antidepressants based on their mechanisms of action.
2nd To assess whether the new classification, in contrast with standard classifications, allows professionals who are high prescribers of antidepressants to locate the drug according to its mechanism of action and quickly relate it to relevant ADRs and pharmacodynamic interactions, including drugs recommended as “first line”.
3rd To define if high prescribers consider this classification useful and easy to apply in comparison with the standard classification.
2. Material and methods
The study was carried out in two phases:
2.1 First phase
A systematic review of the literature, comprising the main papers published in PUBMED database and in specialized Psychopharmacology and Psychiatry was performed and an updated classification of antidepressants according to their mechanisms of action was achieved.
Full prescribing information for all antidepressant drugs approved by: the FDA (U.S. Food and Drug Administration), the ANMAT (Argentine Administración Nacional de Medicamentos, Alimentos y Tecnología Médica) and/or the EMEA (European Medicines Agency), as well as the data obtained from the most widely used guidelines for the treatment of MDD in adults, psychiatry-related books and the main international Physician’s Desk Reference manuals were examined.
On the other hand, using the same sources, the most important ADRs and pharmacodynamic interactions were grouped by their causal mechanism, depending on the affected neurotransmitter system (e.g. serotonergic or noradrenergic), and categorized according to their frequency and severity, underscoring warnings and precautions. The data were then inserted into a table, so that once the drug is located according to its mechanism of action, ADRs and pharmacodynamic interactions can be rapidly inferred.
2.2 Second phase
2.2.1 Design
Once the first phase was finished, a comparative, prospective, longitudinal, experimental and randomized study was performed.
2.2.2 Population
2.2.2.1 Inclusion criteria
Physicians who are high prescribers of antidepressants were included. In Argentina these are:psychiatrists, general practitioners (GP’s), clinicians and cardiologists, as well as physicians in training in the mentioned specialties.
We included Spanish-speaking physicians willing to participate in the study, and to attend courses on treatment of depression and sponsored by scientific societies and/or healthcare institutions (public hospitals, universities).
2.2.2.2 Exclusion criteria
Professionals who might be biased because they have more information about ADRs and drug interactions of antidepressants, such as Masters in Psychoneuropharmacology and/or University Professors of Pharmacology and/or physicians who have previously participated in clinical trials with antidepressants, were excluded.
2.2.3 Sample size
To calculate the sample size we used an expected difference of 16% of correct answers between the group to which the new classification was given (group B) and the group that got the standard classification (group A), with a statistical power of 0.2 and an alpha error of 0.05, resulting in a significance level of p ≤ 0.05. Statistical analysis was performed with the WHO statistical software, EPI Info 7.
The estimated sample size, based on previously established parameters, was a minimum of 120 participants, assigning 60 participants to each group. The groups were randomized 1:1.
2.2.4 Data collection
The sample was collected from September 2016 to August 2017. A total of 318 physicians were surveyed. Professionals were recruited during 13 courses, held in five Argentinean cities (Buenos Aires City, Mar del Plata, Rosario, Córdoba and Bariloche). Of those 318 physicians, 6 were excluded because they had attended more than one course. The remaining 312 professionals were randomly assigned to one of the groups (A or B) comprising 156 physicians each.
At the beginning of the study, participants received a summary of the proposal with the invitation to participate where they had to provide their name, the date and affiliated institution.
Subsequently, all participants in both groups were given an anonymous survey to assess their knowledge prior to the intervention. In this survey they had to write their specialty, age and years of experience to provide statistical data. They also had to write two initials and a number that would be used later in the second survey to identify the professional. The first survey consisted of 19 multiple choice questions with 5 options each. Questions were related to basic knowledge about ADRs and pharmacodynamic interactions relevant to the prescription of antidepressants.
Potential completion biases were considered to enhance the validity of the responses, and therefore improve the questionnaire’s overall validity [
The survey was collected and the group A was given the classification used in the WFSBP-2013 guidelines (used in this study as a reference for the standard classification), whereas group B was given the new classification. Participants’ comprehension of the classification was verified. Together with the classifications, participants received a table with the ADRs mentioned above (see First phase). Then, the following guidelines were explained for the use of the classifications (Table 1).
Table 1Guidelines for the use of the classifications.
1. In the table of ADRs (Table 3) see if the adverse reaction is related to antidepressants with serotonergic, noradrenergic action and/or with those with both actions
2. Place the antidepressant in the classification (Table 2), and see if it is grouped within a class that includes the denomination “serotonergic”, “noradrenergic” or both
3. Consider that in the non-monoaminergic (class B) drugs the ADRs mentioned here are not detected or they occur with a very low incidence
4. Consider that MAOIs (class A1) in general are serotonergic, noradrenergic and dopaminergic modulators (Table 2 number 1)
5. For more information read the rest of the references located under the ADRs table (Table 3), as well as below the classification of antidepressants by mechanisms of action (Table 2).
The classifications were collected from both groups. Three hours later, with the goal of recalling to long term memory and stored knowledge that can be evoked [
], participants received the second survey, which consisted of the same 19 questions of the first survey, plus 6 questions of acceptability about the classification used in each group (Fig. 1).
The collected information was uploaded into a database and analyzed with the EPI Info 7 program.
Descriptive statistics (using mean/median and percentages), and inferential statistics (chi-square tests) to establish the difference between both groups, considering a significant difference at p ≤ 0.05 were used.
3. Results and discussion
3.1 First phase
3.1.1 Scientific foundations for the assembly of the classification
3.1.1.1 Mechanisms of action of antidepressants
To understand the mechanisms of action of the antidepressant monoaminergic modulators (class A of this classification), it should be remembered that the main mechanism by which the monoamines (serotonin, norepinephrine and dopamine) finish their action is the reuptake by the same pre-synaptic terminals that released them, -called neuronal reuptake or uptake type 1 (Fig. 2).
Fig. 2Mechanism of action of monoaminergic antidepressants.
The neuronal reuptake is mediated by specific transporters (mentioned below). Once inside the neuron the monoamines can either be repackaged into secretory vesicles for reuse or undergo enzymatic degradation mainly by monoamine oxidase (MAO). For the metabolism of norepinephrine and dopamine the enzyme catechol-o-methyltrasferase (COMT) also participates [
The mechanisms of action of “class A“ antidepressants are usually pre-synaptic and include:
I.
Inhibition of MAO.
II.
Inhibition of neuronal reuptake, relatively selective for SERT (serotonin transporter), NET (norepinephrine transporter), or DAT (dopamine transporter), or non-selectively, mainly with a double action on the SERT and the NET.
III.
Antagonism of autoreceptors located in the pre-synapses.
In general, the activation of somatodendritic autoreceptors (e.g., α2-adrenergic receptor for noradrenergic neurons and 5-HT1A receptors for serotonergic neurons, among others), inhibits the firing rate of the neurons, so that their antagonism enhance the firing rate and the release of the neurotransmitter. 5-HT1A receptors are found as a somatodendritic and pre-synaptic autoreceptor, but also as post-synaptic receptors.
Most of these mechanisms of action lead to an increase of the specific monoamine in the synaptic cleft. With recurrent administration of the antidepressant, a reduction in the sensitivity of the somatodendritic and terminal autoreceptors might occur, with the consequent sustained activation of post-synaptic receptors and signalling pathways triggering medium and long term neuroplastic actions and epigenetic mechanisms. These actions include changes in some regions of the brain circuitry involved in the processing of affective information that will conclude with the antidepressant effect [
Recent works postulated that, prior to the occurrence of the neuroplastic changes, the negative bias on emotional stimuli, characteristic of depressed patients, could be modified by antidepressants [
Classification of antidepressants according to their mechanisms of action
In Table 2, the term “relatively selective” refers to the greater affinity that class AIIa drugs have on certain transporters such as SERT (for class AIIa1). This affinity, in some cases can vary increasing the dose. With higher doses, sertraline can also inhibit DAT, and paroxetine can inhibit NET [
Melatonine receptors (MT1 and MT2) agonists: Agomelatine(3)
Class C: Drugs in research and development
1. In general, they are serotonergic, noradrenergic and dopaminergic modulators. The clinical indication is for MDD refractory to class AIVc. With so many other safer antidepressants currently available, MAOIs are now indicated when several treatments have failed (see general Ref.
Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: Section 3. Pharmacological treatments.
3. With trials in depression but not approved by the FDA.
4. There are also selective norepinephrine and dopamine reuptake inhibitors, which are not included in this classification because they are not in use.
5. First-line antidepressant recommended for MDD by the most cited guidelines worldwide for proven efficacy and tolerability (see general Refs.
Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: Section 3. Pharmacological treatments.
World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for biological treatment of unipolar depressive disorders, part 1: update 2013 on the acute and continuation treatment of unipolar depressive disorders.
Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: Section 3. Pharmacological treatments.
7. With action on more than one molecular target (such as transporters, receptors, enzymes) related to the antidepressant effect. It can be relatively selective for the serotonergic system (Class AIVa), for the noradrenergic (Class AIVb), or non-selective with action on both of them. In the latter case, the action is usually less specific modulating targets of different systems (Class AIVc). Some authors considerer bupropion a multimodal for acts as a NDRI and as a norepinephrine and dopamine releaser (see general Ref.
8. In addition to inhibiting SERT, vilazodone is a partial agonist of R5HT1A, vortioxetine, a full agonist of R5HT1A, a partial agonist of R5HT1B and an antagonist of R5HT1D, R5HT3 and R5HT7. Among all these targets, it has an important affinity for SERT and for R5HT3. Trazodone is an inhibitor of the reuptake of serotonin and an antagonist of different serotonergic receptors, among which is R5HT2A. (see general Ref.
9. Mianserin increases the release of noradrenaline by a double mechanism: antagonism of pre-synaptic Rα2 and inhibition of NET. The drug also antagonizes Rα1, a mechanism by which it can counteract some ADRs. Maprotiline produces a significant inhibition of NET. In addition, it interacts with different receptors, producing, among other mechanisms, a moderate antagonism of R5HT2 (see general Refs.
Bupropion (class AIIc) is a relatively weak inhibitor of the neuronal reuptake of norepinephrine and dopamine. While the mechanism of action of bupropion is unknown, it is presumed that this effect is mediated by noradrenergic and/or dopaminergic mechanisms [
WELLBUTRIN SR (bupropion hydrochloride) sustained-release tablets. Research Triangle Park, NC: GlaxoSmithKline. revised 05/2017. https://www.gsksource.com/wellbutrin_sr/; 2017 [accessed 6 November 2017].
The traditionally named tricyclic antidepressants (TCAs), are included in class AIV (Multimodals). These drugs are considered multimodal, because besides the inhibition of neuronal reuptake they act on different receptors related to the antidepressant effect. The inhibition of reuptake that they produce is non-selective, since they can inhibit both SERT and NET. Tertiary amines (such as amitriptyline, imipramine and chlorimipramine) are more potent inhibitors on SERT, whereas secondary amines (such as desipramine and nortriptyline) inhibit NET more strongly. These drugs bind to different receptors, which can influence their therapeutic effects: antimuscarinic actions related both to their ADRs and their antidepressant effect, are mainly produced by tertiary amines, although they can occur with all members of this class. In addition, amitriptyline and chlorimipramine produce, with a relatively high affinity, antagonism of 5HT2 receptor. Some members of this class have α2 receptor antagonism, although with a very low affinity [
3.1.1.2 Mechanisms of production of adverse drug reactions (ADRs)
To achieve the aim that once the drug is located according to its mechanism of action it can be quickly linked to ADRs and pharmacodynamic interactions, it is essential that the professional has a table where the main ADRs and pharmacodynamic interactions are grouped according to its relation to antidepressants with serotonergic, noradrenergic action or both (see Table 3 below).
Table 3Adverse Drug Reactions (ADRs).
1. These ADRs are still produced with the antidepressants recommended as “first line” (Table 2, number 5).
2. Precautions and warnings (P & W) are in red in this table. Other P & W highlighted in the full prescribing information by the regulatory agencies are:
• All antidepressants can increase the risk of suicide in children, adolescents and young adults under 24 years old.
• All antidepressants can increase the risk of switch towards hypomania or mania. This risk is greater with class AIVc first, and then with the AIIb.
• Most antidepressants decrease the convulsive threshold. This ADR has been less reported with mirtazapine and citalopram.
• Extension of the interval QTc has been advised for fluoxetine, citalopram, escitalopram, mirtazapine and mianserine.
• Elevation of hepatic enzymes and sporadic cases of toxic hepatitis has been advised for agomelatine and duloxetine.
3. See in classification antidepressants with this effect (Table 2).
4. Among the drugs with serotonergic action, this ADR is less frequent with those that modulate receptors that counteract this side effect (class AIII and AIVa).
5. With class AIVc drugs this effect is enhanced by the antimuscarinic action.
6. Among drugs with noradrenergic action, this ADR is rare with those that antagonize the α1 adrenergic receptor that counteracts this effect (class AIII, and mianserine of class AIVb), being able in some cases even to reach orthostatic hypotension, as usually seen with class AIVc.
7. Among the drugs with noradrenergic action, urinary retention has been reported, mainly with class AIIb (SNRI) and AIVb. Other drugs with noradrenergic action by different mechanisms can partially counteract this effect (class AIIc, AIII), or increase it (class AIVc, by antimuscarinic action).
8. Potentially fatal pharmacodynamic interaction that occurs when combining drugs with serotonergic action.
xxx: Frequent//xx: infrequent//x: rare//- - - does not produce it.
ADRs can be classified considering different variables. In general, they are classified by body systems taking into consideration their frequency and severity.
The most severe ADRs are informed as “warnings and precautions”. FDA highlights these at the beginning of the full prescribing information placing the most serious in black box warnings [
A large number of the ADRs grouped in this work in Table 3 are mentioned in the full prescribing information of the antidepressants within the precautions and warnings. The same happens with certain interactions such as serotonin syndrome.
3.1.1.2.1 ADRs related to the serotonin action
Some of the ADRs of antidepressants are related to the action of serotonin on its different receptors [
The drugs mentioned in this classification without serotonin action (class AIIc, class AIVb and class B) have very low or no serotonergic ADRs incidence, as shown in the full prescribing information approved by the FDA.
ADRs related to the serotonergic actions, more specific and better documented, are mentioned in the following section.
Sexual dysfunction
Sexual disorders are probably the most frequent adverse effects of SSRIs [
]. Mirtazapine (class AIII) that blocks this receptor and bupropion (class AIIc), which acts through a non-serotonergic mechanism, do not cause this effect or cause it with very low incidence [
Incidence of sexual dysfunction associated with antidepressant agents: a prospective multicenter study of 1022 outpatients. Spanish Working Group for the Study of Psychotropic-Related Sexual Dysfunction.
Vortioxetine for major depressive disorder: an indirect comparison with duloxetine, escitalopram, levomilnacipran, sertraline, venlafaxine, and vilazodone, using number needed to treat, number needed to harm, and likelihood to be helped or harmed.
Vilazodone for major depressive disorder: a systematic review of the efficacy and safety profile for this newly approved antidepressant – what is the number needed to treat, number needed to harm and likelihood to be helped or harmed?.
]. Although the exact mechanism has not been elucidated, this is probably due to the agonism on the serotonin 5-HT1A receptor, that could counteract this ADRs [
Association between haemorrhages and treatment with selective and non-selective serotonergic antidepressants: possible implications of quantitative signal detection.
Among the main mechanisms involved in this ADR, a reduction in the uptake of serotonin by platelets might be implicated due to the inhibition of SERT [
Association between haemorrhages and treatment with selective and non-selective serotonergic antidepressants: possible implications of quantitative signal detection.
This risk is greater in those individuals who have volume depletion. In many cases, hyponatremia seems to be the result of the syndrome of inappropriate antidiuretic hormone secretion (SIADH) [
]. It is more frequent and of greater intensity with agents with shorter half-lives, such as paroxetine and venlafaxine and is hardly observed with those that have long half-lives, such as fluoxetine and vortioxetine [
3.1.1.2.2 ADRs related to the noradrenergic action
Some of the ADRs of antidepressants are related to the sympathomimetic effect, produced by the increase in the noradrenergic tone caused by some of them.
Mydriasis and increased intraocular pressure (IOP)
The sympathetic response in the eye causes contraction of the radial muscle of the iris through activation of the α1 receptor and secretion of the aqueous humor from the ciliary processes by activation of the β2 receptors, which facilitates the increase in IOP [
As a consequence, most of the SNRIs (class AIIb) present among their precautions and warnings the cautious use in patients with high IOP or with narrow-angle glaucoma. These ADRs are also reported in the full prescribing information of other antidepressants with noradrenergic action such as mianserin and maprotiline (class AIVb), mirtazapine (class AIII) and bupropion (class AIIc), although these last two produce this ADR with a lower incidence.
With the traditionally named tricyclic antidepressants (class AIVc) this ADR also occurs. In addition, it must be considered that it could be potentiated by the antimuscarinic action [
In the full prescribing information of drugs class AIVa it is reported that angle closure glaucoma has occurred in patients with untreated anatomically narrow angles, and it is recommended to avoid their use with these patients.
Recent studies mention that long-term SSRIs use does not influence the risk of open-angle glaucoma and primary angle-closure glaucoma. Clinicians, however, should still consider the potential risk of glaucoma in certain high-risk groups [
Antidepressants with noradrenergic action, such as SNRIs (class AIIb) and bupropion (class AIIc) present the risk of increasing both blood pressure and heart rate via their sympathomimetic effects [
WELLBUTRIN SR (bupropion hydrochloride) sustained-release tablets. Research Triangle Park, NC: GlaxoSmithKline. revised 05/2017. https://www.gsksource.com/wellbutrin_sr/; 2017 [accessed 6 November 2017].
]. For this reason, the risk of hypertension is communicated within the warnings and precautions of the full prescribing information of these medications.
However, it should be considered that some antidepressants with noradrenergic action antagonize the α1-adrenergic receptor and thus counteract the increase in blood pressure, causing a very low incidence of this ADR, as is the case of mirtazapine (class AIII) and mianserin (class AIVb) [
]. In some cases, they can even cause orthostatic hypotension as usually seen with class AIVc (traditionally called tricyclic). Hypotension, together with muscarinic blockage caused by drugs of this class (AIVc), also increases the tachycardic effect [
Sympathetic activation causes urinary retention due to sphincter contraction of the urethra by activation of the α1-receptor, and relaxation of the detrusor muscle of the bladder by β2 activation [
]. As a consequence, the SNRIs (class AIIb) present this ADR, which is notified within the warnings of full prescribing information of those with greater noradrenergic action, such as duloxetine and levomilnacipran [
].This ADR is also observed with other antidepressants with noradrenergic action, such as mianserin and maprotiline (class AIVb) and drugs of class AIVc (traditionally termed tricyclics) [
Other drugs with noradrenergic action, can partially counteract this effect, as with bupropion (class AIIc) and mirtazapine (class AIII), acting through different mechanisms.
3.1.1.3 Mechanisms of production of drug interactions
Drug interactions can be pharmacokinetic or pharmacodynamic [
]. The pharmacokinetic interaction cannot be related to the mechanisms of action of the drug. Because of this, in Table 2, number 6, the drugs that least provoke them are mentioned. The pharmacodynamic interactions have a direct correlation with the mechanisms of action. Among these, one of the most important is the serotonin syndrome [
This interaction is announced within the full prescribing information warnings of all antidepressants with serotonergic action (class AI, AIIa1, AIIb, class AIII, AIVa, AIVc).
3.2 Second phase
3.2.1 Demographic data
Table 4 shows the demographic data of the total number of people surveyed and their distribution by groups.
Table 4Demographic data.
TOTAL n = 312 (%)
GROUP A (Standard classification) n = 156 (%)
GROUP B (New classification) n = 156 (%)
p value
Age
0.95
Mean-SD
37,70 (11,32)
37,74 (11,48)
37,67 (11,19)
Years of experience
0.83
<5
170 (54,49)
81 (51,92)
88 (57,05)
6 a 10
44 (14,10)
23 (14,74)
21 (13,46)
>10
98 (31,41)
52 (33,33)
46 (29,49)
Specialty
0.96
General clinic
118 (37,82)
60 (38,46)
58 (37,18)
Cardiology
13 (4,17)
7 (4,49)
6 (3,85)
Psychiatry
170 (54,49)
83 (53,21)
87 (55,77)
Others
11 (3,53)
6 (3,85)
5 (3,21)
Place
0.11
Rosario
26 (8,33)
13 (8,33)
13 (8,33)
Córdoba
44 (14,10)
23 (14,74)
21 (13,46)
CABA
216 (69,23)
107 (68,59)
109 (73,72)
Mar del Plata
14 (4,49)
7 (4,49)
7 (4,49)
Bariloche
12 (3,85)
6 (3,85)
6 (3,85)
The average age of the total sample was 37.7 ± 11.32. More than 50% (1 7 0) had less than 5 years experience and about 30% (98) more than 10 years. Approximately 55% (170) were psychiatrists, 38% (118) clinicians/generalists, and 4% (13) cardiologists. The sample is concordant with the data of the high prescribers of antidepressants in Argentina (Close-Up: Argentina, Total specialties, Antidepressant Market, MAT period 7/13 to MAT 7/17, July 2017 edition). Most of the surveys were conducted in the city of Buenos Aires (CABA: 69%–216) followed by those realized in the city of Córdoba (14%–44). The rest of the surveys (17%–52) were carried out in three other cities, with the following decreasing order of participants: Rosario, Mar del Plata, Bariloche.
The p value referred to in the last column, considers, in group form, all the parameters set for each demographic variable (age, years of experience, specialty and place) and indicates the statistical significance between group A and B.
3.2.2 Comparison of the standard classification versus the new classification
3.2.2.1 Results of the surveys (correct answers)
Table 5 shows that there was no significant difference in the number of correct answers in survey 1 (prior to delivering the classifications) when comparing group A (which later was given the standard classification) with group B (which then received the new classification). The mean of correct answers was 9.80 ± 4.17 vs. 9.49 ± 4.14, respectively (p = 0.51).
Table 5Surveys results.
GROUP A (Standard classification) n = 156 (mean-SD)
GROUP B (New classification) n = 156 (mean-SD)
p value
Correct answers
Survey 1 (n = 156)
9.80 (4.17)
9.49 (4.14)
0.51
Survey 2 (n = 156)
13.12 (4.14)
14.60 (3.58)
0.0008
Difference in correct answers between Surveys 1 and 2
Conversely, significant differences were found when the number of correct answers was seen in survey 2, after handing the classifications (see Fig. 1: Experimental design). Those that belonged to group B (new classification), answered significantly more correct answers than those that belonged to group A (standard classification). Out of the total of 19 questions related to ADRs and pharmacodynamic interactions which are relevant by their severity or frequency, the mean of the correct answers was 14.60 ± 3.58 for group B vs. 13.12 ± 4.14 for group A (p = 0.0008). This corresponds to a percentage of correct answers of 76.84% vs. 69% on the survey conducted for group B and A respectively.
Regarding the difference of correct answers between survey 1 and 2, those that used the standard classification (group A) in survey 2 responded 3.34 ± 3.59 more correct answers than in survey 1, while those that used the new classification (group B) increased their correct answers by 5.08 ± 3.42, which represents a significant difference (p < 0.0001).
After transforming these last values (3.34 and 5.08) in percentage, the difference is still significant. In this way, although the percentage of the expected correct answers difference (16%) used for the calculation of the sample was not reached, in the second survey we obtained a percentage of 9% in the difference of correct answers in favor of group B, this difference being also significant with p < 0.0001.
In Fig. 3, the pink/orange boxes (survey 1) show the first line of the first row of Table 5, that is the correct answers for group A and B respectively, carried out before delivering the classifications. On the other hand, the green boxes (survey 2) show the second line of the first row. These are the correct answers for each of the groups in survey 2, made after delivering the standard classification to group A and the new classification to group B.
Fig. 3Comparison of groups according to surveys. In this Box-Plot the first row of Table 5 is represented (correct answers). Data in both surveys follow a normal distribution. Under these circumstances the mean is close to the median, so in the text description we will use the mean, instead of the median commonly used in the Box-Plot.
As mentioned for Table 5, this graph shows that in survey 1 there is no significant difference (p = 0.51) prior to giving the classifications, between the means of correct answers when we compare group A (9.80 ± 4.17) vs. group B (9.49 ± 4.14).
On the contrary, when analyzing survey 2, conducted after administering the standard classification to group A and the new one to group B, a significant difference (p = 0.0008) between the means of both groups (13.12 vs 14.60) can be observed in favor of group B. On the other hand, it should be noted that the dispersion is lower in group B (reflected in a smaller box) data that is reinforced by a lower SD (SD 3.58 vs. SD 4.14).
3.2.2.2 Comparative acceptability of classifications
The following Table 6 analyzes all the acceptance variables of the standard classification (group A) compared to the new classification (group B).
Table 6Acceptability of the proposal classification.
It was found that group B (new classification) had a significantly higher acceptance in each of the variables evaluated. The p value referred to in the last column considers the 3 parameters set for each variable in group form, and indicates the statistical significance between groups A and B.
The optimal parameter of each variable (marked in light blue) was the one that presented the greatest statistical significance in favor of group B (new classification), compared to group A (standard classification), being also the most selected parameter (almost 70% or more) in each variable by group B.
In the following Fig. 4, Fig. 5, Fig. 6, Fig. 7, Fig. 8, Fig. 9 the data are displayed separately from each of the variables used to assess acceptability (Table 6), after having used the classifications. Group A (standard classification) is represented in blue and group B (new classification) in orange. 100% of the sample of each group is distributed in the different parameters used to evaluate each acceptance variable.
Fig. 4Utility according to classification. This figure shows that for those who used the new classification, it proved to be useful in 82% vs. 57% of those who used the standard classification.
Fig. 5Understanding according to classification. It is observed that the new classification proved to be easier to understand (69%) than the standard classification (51%).
Fig. 6Practicity according to applied classification. The new classification resulted with a practicity between high and moderate in 97% of those who applied it vs. 87% for those who used the standard classification.
Fig. 7Contribution according to classification. When asked about the contribution for daily practice, those who were in group B and used the new classification found that it provides a positive contribution in 87% of the cases vs. 66% in group A that used the standard classification.
Fig. 8Interest according to classification. 71% of the participants of group B who applied the new classification found it very interesting, while only 46% of the members of group A who used the standard classification found it very interesting.
Fig. 9Application according to classification. It was observed that 79% of group B (new classification) would use it frequently, while group A would use the standard classification only in 60%.
3.2.3 General acceptability of the new classification
Group B (new classification) was analyzed separately. Fig. 10 shows the different acceptance variables studied.
Fig. 10General acceptability of the new classification (n = 156). The p value referred (p ≤ 0.0003), considers all the acceptance variables in group form, indicating that in all of them the highest parameter (orange colored bar) presents a statistically significant difference with moderate and low parameters. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
The new systematic classification of antidepressants presented here will also have a useful application in the future by allowing a large number of drugs currently in research and development (Table 2, class C) to be incorporated in a dynamic, as well as logic and epistemic way, when some of them are authorized for clinical use by the FDA. An example is esketamine (a glutamatergic modulator) recently approved for resistant depression, which has not been included in this study because it was under investigation when it was started.
Demographic data (Table 4) in this study show that no significant differences were found between those respondents assigned to group A and those assigned to group B; in terms of age, distribution by years of experience, specialty or place where the survey was conducted. Thereby, it is deduced that both groups are comparable to each other.
As can be seen in Table 5, with survey 1 approximately 50% of the 19 questions related to ADRs and pharmacodynamic interactions which are relevant by their severity or frequency were correctly answered (9.80-group A vs. 9.49-group B) in both groups. These data show that physicians who are high prescribers of antidepressants have a low level of knowledge about which antidepressants may be of risk to each particular patient. It can be deduced that the same occurs with the rest of the physicians that use antidepressants.
On the other hand, there was no significant difference in the number of correct answers when comparing group A with group B (p = 0.51), which implies that both groups started from a similar knowledge base.
Conversely, significant differences were found when the number of correct answers was analyzed in survey 2, after the administration of the classifications in favor of group B. In this group, with the application of the new classification, the percentage of correct answers increased from approximately 50% (9.49) to almost 80% (14.60).
Regarding the difference of correct answers between survey 1 and 2, a significant difference in favor of group B was also observed, which allows us to infer the usefulness of the new classification.
As in Table 5, Fig. 3 shows more clearly that in survey 2, group B (new classification) presents a number of correct answers significantly higher than group A (standard classification) and a lower dispersion (few cases that are separated from the average).
In relation with acceptability in Table 6, it was found that group B (new classification) had a significantly higher acceptance in each of the variables evaluated.
On the other hand, Fig. 10 shows that when the acceptance variables were analyzed separately only for group B, the highest parameter was chosen in a significant way for each variable, confirming the global acceptability of the new classification presented here.
Finally, when each of the acceptance variables that were analyzed with the demographic variables was compared within group B, no significant difference was found in any of them. It means that there was no significant difference in the opinion of the physicians who had received the new classification about utility, understanding, practicity, contribution, interest and application; in relation to age, years of experience, specialty or place where the survey was conducted.
From this, it follows that its applicability is universal.
3.3 Limitations
Some mechanisms of action are not well known. On the other hand, some relevant ADRs cannot be related to their mechanisms of action (Table 3) and others can be increased or counteracted by different mechanisms as mentioned in the text.
The pharmacokinetic interaction cannot be related to the mechanisms of action of the drug, therefore, unlike what happens with the pharmacodynamic interactions, this classification does not allow to link each drug with pharmacokinetic interactions.
It would be necessary to carry out further studies incorporating professionals who are not high prescribers of antidepressants, as well as to evaluate the variables of acceptability over time.
4. Conclusion
This study shows that, in contrast with standard classifications, this new classification of antidepressants with a logical and epistemic nomenclature based on their mechanisms of action, allows professionals to locate the drug according to this mechanism and quickly relate the drug to relevant ADRs and pharmacodynamic interactions.
The acceptability of this new classification was significantly higher, comparing it with the standard classification in all the acceptance variables analyzed. In addition, no significant changes were observed in relation to the demographic variables studied, demonstrating that its application is universal.
5. Role of the funding source
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
Authors state that there are not conflict of interest.
Acknowledgments
The authors express their gratitude to:
Alan F. Schatzberg, MD and Charles B Nemeroff, MD, PhD, for providing suggestions and editorial help.
Jorge Lerman MD, PhD, Edgardo Knopoff, MD and Jorge Medina, MD, PhD, for proof reading the article.
Roberto Diez, MD, PhD, Alberto Carli, MD, PhD, Andres Pichón Riviere, MD, PhD, from the Universidad de Buenos Aires, Facultad de Medicina, Secretaria de Ciencia y Técnica, for providing assistance in the methods used in this research.
Mariana Moncaut, MD, Laura Guelman, BSc., PhD for writing assistance.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: Section 3. Pharmacological treatments.
World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for biological treatment of unipolar depressive disorders, part 1: update 2013 on the acute and continuation treatment of unipolar depressive disorders.
WELLBUTRIN SR (bupropion hydrochloride) sustained-release tablets. Research Triangle Park, NC: GlaxoSmithKline. revised 05/2017. https://www.gsksource.com/wellbutrin_sr/; 2017 [accessed 6 November 2017].
Incidence of sexual dysfunction associated with antidepressant agents: a prospective multicenter study of 1022 outpatients. Spanish Working Group for the Study of Psychotropic-Related Sexual Dysfunction.
Vortioxetine for major depressive disorder: an indirect comparison with duloxetine, escitalopram, levomilnacipran, sertraline, venlafaxine, and vilazodone, using number needed to treat, number needed to harm, and likelihood to be helped or harmed.
Vilazodone for major depressive disorder: a systematic review of the efficacy and safety profile for this newly approved antidepressant – what is the number needed to treat, number needed to harm and likelihood to be helped or harmed?.
Association between haemorrhages and treatment with selective and non-selective serotonergic antidepressants: possible implications of quantitative signal detection.
30028-0&id=gr1.jpg){kind=link}
30028-0&id=gr2.jpg){kind=link}
30028-0&id=gr3.jpg){kind=link}
30028-0&id=gr4.jpg){kind=link}
30028-0&id=gr5.jpg){kind=link}
30028-0&id=gr6.jpg){kind=link}
30028-0&id=gr7.jpg){kind=link}
30028-0&id=gr8.jpg){kind=link}
30028-0&id=gr9.jpg){kind=link}
30028-0&id=gr10.jpg){kind=link}