We strenuously reject the PLOS ONE Editors’ decision to retract our work, given that all the issues they have raised would usually warrant critique in a systematic review at most, not retraction.

Our team answered all points raised during post-publication discussions in extensive detail. In the interests of transparency, accuracy and fairness, the most relevant points are summarized below, so that an objective reader can make up their own mind about the rigour of our study and whether the PLOS ONE Editors’ decision to retract our work is justified:

Claim 1 – Insufficient reporting details provided to allow for replication of the treatment used and assessment of the clinical validity of treatment choices.

All additional details requested by the Editors were provided i.e. how consultations were conducted; how the choice of medicine for each patient was made according to their clinical indications; and how dosage (either 30C or 200C) was selected according to well-established homeopathic prescribing guidelines. We also clarified that all patients were prescribed 200C, except for five patients who received a 30C prescription at the end of the observation period.

Although we understand the Editors’ curiosity regarding these details, our published study provides sufficient detail for those trained in the specific prescribing technique used (‘Individualized Homeopathic Treatment’ - ‘Classical Homeopathy’) to replicate and validate the intervention given.

Claim 2 – Our study design was insufficient to rule out placebo effects, so our conclusions were not supported by our data

We reject this point of argument entirely. Our study utilized a double blind, double dummy, randomized, placebo-controlled, three-armed study design: a gold-standard study design according to CONSORT and SPIRIT 2013 guidelines. To our knowledge there is no better clinical trial design to account for placebo effects.

Furthermore, fluoxetine’s efficacy in depression in adults is well established. As our study detected a clinically beneficial, and statistically significant effect of fluoxetine on depressive symptoms compared to the matched placebo, this positive control arm demonstrates that the study design was indeed capable of generating and detecting a predicted beneficial treatment effect beyond placebo.

Disturbingly, during the post-publication discussion the PLOS ONE Editors did not explain in what way/s they considered our study design to be inadequate. Rather, they simply stated that because the homeopathic treatments included 30C (i.e. 10^-60) and 200C (10^-400) potencies of the homeopathic medicines, any positive effect seen must have been a placebo effect.

Whilst homeopathic medicines cannot have a standard pharmacological mode of action based on interaction at the biochemical level, this does not preclude other, as yet poorly understood, modes of action other than placebo.

We also explained to the Editors that Individualized Homeopathic Treatment (IHT) has already been shown to provide beneficial clinical effects beyond placebo in a systematic review and meta-analysis of 32 published studies, from 26 independent research teams, from 12 countries using IHT for many conditions within randomized control trials. Mathie et al. 2014 concluded that IHT is up to twice as likely to provide clinical benefit as placebo (OR = 1.98, 95% CI 1.16 to 3.38); this result fully withstands sensitivity analysis and directly undermines the PLOS ONE Editors’ argument.

Claim 3 – We did not provide enough information about how patients were diagnosed, and the Editors questioned the reliability of our diagnoses

The diagnosis of Major Depressive Disorder (MDD) was made using Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria from the American Psychiatric Association (stated in the Methods section of the article) and we provided a clear checklist of symptoms that were used. Each participant was also assessed at baseline using two long-established, validated, and standardized scales for assessing depression severity (Hamilton Rating Scale for Depression and Beck Depression Inventory). Moreover, according to the eligibility criteria (page 5 of the article) only women with a 17-item Hamilton Rating Scale for Depression score of 14-24 (corresponding to a moderate to severe depression) were included, confirming the MDD diagnosis.

The PLOS ONE Editors provided us with no clear reason as to why this was insufficient to address their concerns.

Claim 4 – Insufficient information provided as to ‘how blinding was achieved during the diagnostic assessments’

The precise meaning of this point is unclear, but if the claim is that blinding could have been broken during the study, this speculation is completely incorrect.

Blinding was achieved using the most rigorous ‘double dummy’ design. All patients in the study received the same treatment experience – a homeopathic consultation, followed by two prescriptions i.e. a homeopathic medicine (verum or placebo) and Fluoxetine (verum or placebo). The clinical assessments at week 0, 4 and 6 were made independently from the prescribing physician, by a psychologist with no training in homeopathy. Finally, as stated in our publication, blinding was broken only once, due to a patient (taking Fluoxetine) experiencing adverse events.

We therefore reject this point outright as a reason to retract our study.

For further details visit: https://www.HRI-Research....

Dr. Emma Macias-Cortes et al.