Depression, Difficult to Treat Depression and Treatment Resistant Depression: Reasons for Failing to Respond to First-Line Treatment

Journal of Psychiatry Reform vol 9 #15

Alan Eppel, MB, FRCPC1, Ana Hategan, MD, FRCPC2

Author information

1  Professor, Department of Psychiatry and Behavioural Neurosciences, Michael G. DeGroote School of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada. iD:

2  Clinical Professor, Division of Geriatric Psychiatry, Department of Psychiatry and Behavioural Neurosciences, Michael G. DeGroote School of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada. iD:


Treatment resistant depression represents a dilemma for healthcare providers. Depression is a heterogeneous disorder, as reflected by heterogeneity of treatment effect and variable nonresponse rates. Treatment resistant depression has evaded universal definition and meaning, but the term is frequently applied to those who have not responded to at least two antidepressant trials of adequate dose and duration. However, treatment response has different meanings across multiple research settings. The aim of this narrative review is to examine the challenges of the definition of treatment resistant depression and provide some caveats and recommendations for real-world clinical practice. The authors call for a strengthening of research criteria and research integrity in order to ensure increased validity of future study outcomes.


Depression is among the commonest of psychiatric disorders and constitutes a major percentage of patient presentations [1]. The individual, family and societal impact of depressive illness is enormous, with high rates suffering and incapacity [1]. Many patients present with uncomplicated symptoms and respond to first-line therapies. First-line interventions include antidepressants and/or cognitive behavioural therapy. There is agreement across various national treatment guidelines on a list of first-line antidepressant medications. Many patients, however, do not respond to first-line and may experience prolonged bouts of depression lasting months or years. Nevertheless, there are no agreed guidelines for second-line treatment interventions. The options include switching to a different antidepressant, augmentation with other medications (e.g., lithium, triiodothyronine (T3), stimulants), adding a second antidepressant, adding medications from other psychotropic classes, switching to alternate forms of psychotherapy, or using somatic therapies including electroconvulsive therapy (ECT) and repetitive transcranial magnetic stimulation [1, 2].

Treatment Resistant Depression: A Misappropriated Term

There is no agreed definition for the term “treatment resistant depression” [2]. The term is frequently used inappropriately when first or second-line treatments are not effective in cases when new medications or other therapeutic agents are being examined for efficacy. For example, in many of the studies on ketamine the definition used is the failure to respond to two courses of antidepressants without specification of antidepressant class, maximal dosage or duration of treatment [3].

More rigorous definitions of treatment resistant depression include treatment with antidepressants from different therapeutic classes at maximal doses for at least eight weeks. Choices include SSRIs, dual acting antidepressants, tricyclics and MAO inhibitors. Additional requirements should include an adequate course of ECT and an adequate course of psychotherapy.

Alternatives to the term “treatment resistant depression” have been proposed, such as “difficult to treat depression” and “refractory depression,” but these terms also lack agreed upon criteria and cannot be used to compare studies [3-5]. The bottom line is that the absence of an agreed definition of treatment resistant depression means that study populations are not comparable and conclusions from individual studies cannot be generalized.

It should be noted that national and international treatment guidelines are unable to demonstrate agreement about what to do when first-line treatments fail [2, 4-7].  What is the second step? Should the clinician switch within the same class of antidepressant, switch to another class, augment with lithium, T3, or stimulant, add an antipsychotic medication, or combine with another antidepressant from a different class?

Diagnosis and Misdiagnosis

The DSM diagnosis of major depression is broad and nonspecific resulting in a heterogeneous group of phenotypes [8-11]. The concept of comorbidity illustrates the problem of overlapping of DSM categories. Parker et al. [8] have stated that the concept of “major depression” is analogous to describing a syndrome of “major breathlessness”. If patients with asthma, cardiac failure and pneumonia were all diagnosed with “major breathlessness” and prescribed the same treatment, the outcomes would be catastrophic [8]. Clinicians are often seduced by the presentation of depressive symptoms and speedily reach for the prescription pad to prescribe a first-line antidepressant. While a rose may always be a rose, depression is very definitely not always depression. Indeed, there have been calls for a change in the paradigm for classifying and treating depression [9-11].

Frequently, the principal condition may be any or all of the following: bipolar disorder, depressive phase or mixed state; posttraumatic stress disorder; emotional dysregulation associated with borderline personality or antisocial personality; primary substance use disorder; or specific medical illnesses such as cancer, chronic pain, and thyroid disease.

A lack of understanding of the course of bipolar illness can lead to this being misdiagnosed particularly when there have been yet no episodes of hypomania or mania. Frequently underlying posttraumatic stress disorder may be unrecognized for years. The clinician may not have facilitated an adequate therapeutic alliance to elicit such a history. Patients presenting with emotional dysregulation may be misdiagnosed as having a depressive illness. Patients with problems and anger control may be erroneously treated with antidepressants. Contributing factors can include ongoing marital or interpersonal conflict, work demands, harassment or bullying at work or school, chronic pain, medical illness and disability. Therefore, a review of past episodes, clinical documentation of hospitalizations, and outpatient treatment are essential.

The Research Literature

Treatment failure may be due to the prescription of ineffective medications. The clinician is hampered by persistent problems within the research literature including author bias, publication bias, and pay-as-you-go journals (these are journals where the authors have to pay significant fees to have papers published). There is concern that most of these journals are motivated by profit rather than scientific discovery. As well, peer review may be inadequate.

Too narrow selection criteria for randomized controlled trials is another contributor. These studies may not reflect real-world conditions. Controlled trials often exclude more seriously ill patients, those with suicidal ideation, and comorbid disorders. This is described as the gap between “efficacy” as demonstrated in clinical trials and real “effectiveness” in clinical practice. Trials for medication treatment of depression usually have a duration of eight weeks. This is too short a time to observe full clinical response or loss of response.

Outcome measures can be a source of erroneous conclusions about treatment effectiveness. Outcome measures are generally rated by observers who are blind to the treatment condition. Small improvements on scales such as the Hamilton and the Montgomery-Asberg depression rating scales are commonly reported. Trial outcomes using these scales may be statistically significant but not clinically relevant. Yet, such trials may be reported claiming superiority over the placebo or alternate medication.

Over the past decade, huge research effort has been deployed to discover biomarkers in depression that would help predict treatment response. Such endeavours have included identifying biochemical markers, the use of machine learning and artificial intelligence, and the search for genetic commonalities [12, 13]. Thus far, this research has not reached the stage for clinical application.

Spin and Critical Appraisal

Spin refers to reporting the results of a clinical trial in a more favourable light than is justified by the evidence [14]. This may be seen when the conclusions in the abstract are not consistent with the findings in the main body of the paper. A major source of bias or even deliberate misleading conclusions are conflicts of interest. There are many sources of this including: sponsorship of the study by the manufacturer of the medication; authorship of the publication by individuals that have one or multiple relationships with the pharmaceutical company marketing the drug; authors who have received funding or other material benefits from the manufacturer; or authors who own clinics promoting the use of a particular drug. Because of these sources of bias, the traditional approach to critical appraisal is not adequate. Critical appraisal involves evaluating the patient population and controls, documenting that randomization and allocation were carried out in a blind fashion, that both treating personnel and those rating outcome are blind to the randomization process.

The following additional components to critical appraisal should be considered:

  • Review of the validity of the statistical methods and evidence of bias.
  • The article should be examined for spin; the title of the article, the abstract, and conclusions must fairly reflect the findings in the study.
  • In addition to the conflicts of interest disclosed in the published paper, journal editors and readers may need to include an independent online search to identify undisclosed conflicts.

A revised approach to critical appraisal should also include:

  • Is the research published in a reputable journal with an independent editorial board and peer review process free from conflicts of interest?
  • Is there full disclosure of the author’s conflicts of interest? What is the relationship of authors to manufacturers of medications used in the study? Are they employed by the company that makes the drug? Do they receive honoraria, research funding, assistance with the write up of the article? Do authors sit on the boards of companies to hold executive positions in companies that manufacture the drug, or own clinics dedicated to the provision of the drugs in question?


The reasons that patients diagnosed with depression fail to respond to first-line treatment are multifaceted. The term “treatment resistant depression” has no agreed meaning and, therefore, should not be used in research studies. The term provides no basis for comparison between treatments and study outcomes. The concept of depression and current psychiatric nosological systems comprise a heterogeneous set of conditions and etiologies. Future attempts should be made to identify which presentations respond best to particular treatment interventions.

Take-away points to consider:

  • Depression is a heterogeneous clinical syndrome.
  • Psychiatric assessment may take more than one session and is greatly assisted by collateral information and family interviews.
  • Failure to respond to depression treatment is often due to incorrect diagnosis.
  • There is no agreed sequence of treatment following first-line interventions in depression and clinicians must base their decisions on a review of guidelines and clinical experience.
  • There is a need for studies on depression with more narrowly defined patient populations.



  1. Malhi GS, Bell E, Bassett D, et al. The 2020 Royal Australian and New Zealand College of Psychiatrists clinical practice guidelines for mood disorders. Aust N Z J Psychiatry. 2021;55(1):7-117. doi:10.1177/0004867420979353.
  2. Fekadu A, Donocik JG, Cleare AJ. Standardisation framework for the Maudsley staging method for treatment resistance in depression. BMC Psychiatry. 2018;18(1):100. doi:10.1186/s12888-018-1679-x.
  3. Eppel A. Ketamine for treatment resistant depression: Claims are based on poor quality evidence. Journal of Psychiatry Reform. 2022;11:5.
  4. Bayes AJ, Parker GB. Comparison of guidelines for the treatment of unipolar depression: A focus on pharmacotherapy and neurostimulation. Acta Psychiatr Scand. 2018;137:459-471. doi: 10.1111/acps.12878.
  5. McAllister-Williams RH, Arango C, Blier P, et al. The identification, assessment and management of difficult-to-treat depression: An international consensus statement. J Affect Disord. 2020;267:264-282. doi:10.1016/j.jad.2020.02.023.
  6. Santaguida PL, MacQueen G, Keshavarz H, Levine M, Beyene J, Raina P. Treatment for Depression After Unsatisfactory Response to SSRIs. Rockville (MD): Agency for Healthcare Research and Quality (US); April 2012.
  7. Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163(11):1905-1917. doi:10.1176/ajp.2006.163.11.1905.
  8. Parker G, Paterson A, McCraw S, Hadzi-Pavlovic D. Targeting and transforming major depression. Acta Psychiatr Scand. 2017;135(4):310-318. doi:10.1111/acps.12681.
  9. Berry S. The ‘major’ consequences of how we understand depression. The Sydney Morning Herald. November 6, 2017.
  10. Malhi GS, Bell E, Boyce P, et al. Channelling response: A novel perspective and therapeutic paradigm. Aust N Z J Psychiatry. 2020;54(8):775-779. doi:10.1177/0004867420945363.
  11. McAllister-Williams R, Aaronson S, Conway C, et al. The ‘difficult-to-treat depression’ and the ‘response paradigm’ models: Implications and relevance to patient management. Australian & New Zealand Journal of Psychiatry. 2021;55(8):824-825. doi:10.1177/00048674211013090.
  12. Strawbridge R, Young AH, Cleare AJ. Biomarkers for depression: Recent insights, current challenges and future prospects. Neuropsychiatr Dis Treat. 2017;13:1245-1262. doi:10.2147/NDT.S114542.
  13. Nunez JJ, Nguyen TT, Zhou Y, et al. Replication of machine learning methods to predict treatment outcome with antidepressant medications in patients with major depressive disorder from STAR*D and CAN-BIND-1. PLoS One. 2021;16(6):e0253023. doi:10.1371/journal.pone.0253023.
  14. Chiu K, Grundy Q, Bero L. ‘Spin’ in published biomedical literature: A methodological systematic review. PLoS Biol. 2017;15(9):e2002173. doi:10.1371/journal.pbio.2002173.
Print Friendly, PDF & Email