Dose equivalents of antidepressants

Antidepressants are very commonly prescribed with the intention of relieving distress for people with depression.

However, side-effect profiles for various drugs mean that close monitoring of these prescriptions is necessary, and switching between antidepressants is common. It is therefore useful to be able to determine ‘equivalent’ doses between varying antidepressants. This is also of importance in clinical research as it allows comparison between dosage regimes to ensure ‘fair competition’.

Various methods exist to determine such dosing equivalents; for example through considering dose response, and side-effect profiles. However the amount of published research available in this area is limited.

An alternative method is to consider the dosage of a drug prescribed when clinicians are able to vary the amount within a trial framework; presumably reaching a dose that represents the best balance between tolerability and efficacy. This is the approach adopted in a review recently conducted and published in the Journal of Affective Disorders.

Aims

Through considering prescribed dosage within variable dosing trials, the researchers leading this study aimed to identify equivalent doses of varying antidepressants. Fluoxetine and paroxetine were chosen as the comparator drugs, representing the most commonly used agents in clinical trials.

The authors searched for randomised controlled trials in which different antidepressants were compared with fluoxetine or paroxetine.

Methods

Trial identification
- The researchers sought published randomised controlled trials where an active antidepressant was compared with either fluoxetine or paroxetine
Participants
- Trials were included if participants were older than 18 and with a mean age of less than 65
- Participants were sought who had a diagnosis of unipolar depression (although trials were included where up to 20% of participants had bipolar depression)
- Comorbid psychiatric diagnoses were excluded
Data extraction
- Two authors independently identified the number of participants within trials and the mean prescribed dosage of antidepressant, together with 95% confidence intervals
- Study quality was assessed in terms of participant blinding and allocation concealment
Quality appraisal
- The quality of available evidence was assessed in relation to:
  1. Methodological issues in trials
  2. Width of dosage 95% confidence intervals in relation to clinical practice
  3. Number of studies
  4. High levels of heterogeneity in findings
  5. Any variation within sensitivity analyses
Analysis
- Weighted mean dose for drugs were calculated considering the average prescribed dosage and number of participants within trial arms
- Equivalent dosage of paroxetine in relation to fluoxetine was calculated, then fluoxetine and paroxetine findings were combined together to generate the comparison group
- Dosages of antidepressants were then recalculated to find equivalence with a daily dose of 40mg of fluoxetine
- Sensitivity analyses were conducted to check the consistency of results
  1. A meta-analysis of mean doses was conducted where two or more trials were available – heterogeneity of results was calculated
  2. Trials where paroxetine was used were excluded from the comparator data set to check the validity of combining fluoxetine and paroxetine together
  3. The relative efficacy of antidepressants was compared to assess the possible impact of drug superiority on dosing

the weighted mean ratio for each drug, which was then used to define the drug's dosage equivalent to fluoxetine 40 mg/d.

The authors worked out the weighted mean ratio for each drug and then defined its equivalence to fluoxetine 40 mg/d.

Results

144 trials were identified meeting study inclusion criteria
- 83 trials allowed mean dosage extraction for comparison (total 14,131 participants)
The mean dosage of drug in comparison to fluoxetine 40mg was calculated
- This resulted in some unusual results; for example the mean dosage of mirtazapine was found to be 50.9mg, a value outside of the licensed maximum dosage in the UK
A high degree of heterogeneity in dosage prescribed was shown for many of the identified drugs
Within sensitivity calculations only sertraline was found to outperform other medications on pooled analysis, but this did not affect the average dose calculation
The confidence in the quality of evidence was cited as predominantly moderate to high
Fluoxetine 40 mg/day was equivalent to:
- Paroxetine dosage of 34.0 mg/day
- Agomelatine 53.2 mg/day
- Amitriptyline 122.3 mg/day
- Bupropion 348.5 mg/day
- Clomipramine 116.1 mg/day
- Desipramine 196.3 mg/day
- Dothiepin 154.8 mg/ day
- Doxepin 140.1 mg/day
- Escitalopram 18.0 mg/day
- Fluvoxamine 143.3 mg/day
- Imipramine 137.2 mg/day
- Lofepramine 250.2 mg/day
- Maprotiline 118.0 mg/day
- Mianserin 101.1 mg/day
- Mirtazapine 50.9 mg/day
- Moclobemide 575.2mg/day
- Nefazodone 535.2mg/day
- Nortriptyline 100.9mg/day
- Reboxetine 11.5 mg/day
- Sertraline 98.5 mg/day
- Trazodone 401.4 mg/day
- Venlafaxine 149.4 mg/day

Only a small number of trials were identified that allowed enough information for comparison to be extracted

83 trials were identified that allowed enough information for comparison to be extracted.

Discussion

The authors concluded:

The number of studies for some drugs was small. The current method assumes dose response relationship of antidepressants… Our findings can be useful for clinicians when they switch antidepressants and for researchers when they compare various antidepressants in their research.

As the authors state the findings of this study are limited by very small numbers of trials for many of the considered drugs. It is remarkable that so few trials were identified from which a mean dosage of drug prescribed could be extracted; the authors rightly call for this to become a standard within research reporting. Their findings are valuable for researchers and clinicians looking to compare findings between randomised controlled trials.

However when it comes to switching between antidepressants clinically the situation is clearly more complicated; varying side-effects, individual physical variation and a plethora of other considerations need to be taken into account, limiting the utility of simplistic dosage comparisons.

From an intellectual perspective this research is to be welcomed. It provides a standardisation between antidepressants and research studies. Caution is no doubt required within clinical settings, but the information in this study is potentially valuable and highlights areas in need of further attention.