To the Editor: In an ironic clinician–academic dichotomy, in the same month that the Royal Australian and New Zealand College of Psychiatrists published a survey showing that 79% of Australian psychiatrists combine antidepressants and 75% of psychiatrists believe that general practitioners should be given information on this topic,1 Keks et al chose a non-psychiatric journal to “mandate that combinations be used as a last resort, and only in specialist settings”.2

Specialists have voted with their prescription pads. That a large majority of Australian psychiatrists feel ethically and clinically obliged to use combination antidepressants speaks volumes about the poor results from the suggestions outlined by Keks et al. The multiple clinical reports and reviews of the benefits of combination antidepressants,3 the suffering and death from depression, and the very low rate of complications reported to the Adverse Drug Reactions Advisory Committee from combination antidepressants do not allow the luxury of awaiting combination therapy research which may never happen. Many combinations of antihypertensives or anti-asthma medications similarly lack such rigorous proof, but are widely used. Isolated case reports of medication complications must be seen as such.

Access to psychiatrists for combination antidepressant therapy is a well intentioned but currently impractical suggestion. Most psychiatrists have massive waiting lists, and research confirms treatment resistance and progressive cell death in the hippocampus of depressed patients while awaiting effective treatment.

Australian GPs are just as capable of using combination antidepressant therapy as their international colleagues, if given the same simple information and training. Canadian GPs read in their journals advice about using combination antidepressants. Anecdotally, many Australian GPs combine antidepressants, but express the wish that the issue could be discussed openly, without them feeling intimidated. Even textbooks of psychiatry, drafted some years ago, teach about combination antidepressants. In the United States, the National Institute of Mental Health STAR*D study of 4000 patients approved combination antidepressants such as venlafaxine with mirtazapine years ago, with no safety concerns.4 Keks et al refer to treatments that today are unacceptable to many, ranging from electroconvulsive therapy to tricyclic antidepressants, despite GPs and psychiatry trainees having been warned for years by academics that tricyclics are outdated, “dirty” and dangerous.

Informed consent requires that patients be informed of all therapies that are relevant to their care and survival, and 88% of psychiatrists believe patients should be informed of combination antidepressants.1 Recent results from the STAR*D study demonstrate the superiority of modern combination antidepressants, with no statistically based evidence that they should not be used.5

To the Editor: Keks et al make a number of important points about the place of combination antidepressant strategies in the pharmacotherapy of depression.1 However, it is important for readers to note that the vigorous repudiation of combination treatments is a peculiarly Australian preoccupation. Our colleagues in Europe and North America are not nearly so troubled.

Combination antidepressant treatments are widely used by specialists. A recent survey of Australian doctors working in psychiatry reported that 79% of respondents had used combination antidepressants and that 75% believed that general practitioners should be given information on their use.2

There is emerging evidence for the use of combination antidepressant strategies — from case series, open clinical trials, and randomised controlled trials (RCTs). The largest summation of the data is a meta-analysis which found that combination antidepressant treatment produced a 62% response rate when monotherapy had failed.3 Although this finding alone cannot be convincing because of the acknowledged lack of large sample RCTs, it is quite another matter to decry combination prescribing as clinically unsound based only on the history of augmentation treatments such as lithium and, to a lesser extent, thyroid hormone treatment when, anecdotally, they provide such clinically disappointing results.

It is not unreasonable to assert the primacy of good clinical reasoning, including sensible prescribing of combination antidepressants, over rigid adherence to evidenced-based algorithms. This sort of thinking is allowable because the evidence base for the treatment of depression is poor. Meaningful guidelines cannot be produced while the evidence is predicated on the flawed proposition that depression is an “it” (a homogenous construct).4

GPs might well be puzzled by the zeal in academic psychiatry for monotherapy. They are advised to “optimise” monotherapy, but not told what this means. They are very familiar with models of staged polypharmacy for common chronic illnesses such as hypertension, epilepsy, diabetes, and asthma, but in psychiatric pharmacotherapy this is apparently unwise or too risky.

The way such admonishments are usually framed is by reference to serious but rare adverse reactions (like the serotonin syndrome), without proper attention to the equally serious and probably more common problems with the current “simple” psychotropic drug options already used by GPs. Failure to contextualise these risks leads to a distortion of risk–benefit prescribing decisions and an unnecessary restriction of treatment choices.

We must have a commonsense approach to the treatment of depression that recognises the proper context of our knowledge base. Combination antidepressant treatments may be “beyond the evidence”, but this alone is not a sufficient justification to stop using them.

In reply: The letters by Horgan and Walters et al underline our motive for reviewing antidepressant combinations. The conclusions of the survey are at least questionable, given that the response rate was only 36%, 18% of respondents were not psychiatrists, and affirmative responders may have only used combination antidepressants once.1 In any case, should clinical popularity substitute for evidence? If so, once popular but now research-discredited treatments such as insulin coma therapy would still be used.

Equating combination antidepressants to combination drugs for asthma and hypertension is misleading. How often are two β-blockers given together in maintenance treatment? Major depression causes severe suffering, but this does not justify the use of unproven treatments ahead of those supported by evidence.

General practitioners should be informed about antidepressant combinations, but the information must be evidence-based. We described the process of dose optimisation, and stand by our advice that complex cases that require unproven treatment (such as combination antidepressants) be referred to a psychiatrist.

Patients should also be informed about combination antidepressants, including the paucity of evidence concerning efficacy and safety, the absence of information about consequences of long-term treatment, and that some combinations are lethal and others frequently unsafe.

Published data from the STAR*D study provide equivocal support for the combination of citalopram and bupropion, as we noted. Evidence of modest effectiveness (remission rate, 13.7%) for the combination of mirtazapine and venlafaxine has appeared.2

Our conclusion was that some antidepressant combinations could be used in certain clinical situations where evidence-based treatments have failed, with safeguards. Given that 17% of respondents to the survey1 observed serious complications with combination antidepressants, this is good advice.