Large-scale clinical trials in psychiatry are rare. That is not surprising, given how much more difficult it is to assess improvement in a psychiatric condition, compared with more simple outcomes in other medical conditions (such as death, for cardiovascular studies). The STAR*D trial was a large-scale trial funded by the US National Institute of Mental Health, whose objective was to examine treatment options in an algorithmic manner, to guide decisions for next-step strategies in patients with depression with inadequate response to an antidepressant.1 It was designed to be a real-world effectiveness study, in which entry criteria were relaxed to include patients with chronicity, comorbidity, and treatment-resistance, and thus more representative of patients in clinical practice than those usually entered in clinical trials.2 The primary outcome was chosen to be remission, defined as a score on a depression severity rating scale within a nondepressed range; for example, a score of 7 or less on the HDRS. A series of randomized studies were offered at each treatment step if patients were not in clinical remission with treatment. The clinicians treating the patients in STAR*D also used a form of measurement-based care, in which results of various rating scales were reported back to the clinician to help guide treatment decisions.3
In a Perspective article in this issue, Dr Pigott4 makes numerous criticisms of STAR*D and how the results are reported and interpreted. We (and others) agree that, in many ways, STAR*D was a failed effectiveness trial. One of the major limitations was initially considered a strength: the novel use of equipoise-stratified randomization, in which patients (with their clinician) could choose the treatments for randomization. For example, patients could choose whether to enter a randomized medication augmentation study or a medication switch study, or whether to enter a randomized cognitive therapy study. Although the concept of equipoise-stratified randomization was an attempt to mirror what happens in the real world, where some treatment options are not acceptable to patients, an unintended consequence was that very few patients were willing to be randomized to any treatment. This resulted in sample sizes for each study after step 2 that were too small to have sufficient power to detect clinically meaningful differences between treatments.
However, some of Dr Pigott’s arguments appear to be misinformed. His paper identifies 3 main objections: that STAR*D investigators incorrectly reported remission rates based on the patient-rated QIDS-SR instead of the predefined primary outcome, the clinician-rated HDRS; that remission as defined is not an acceptable goal for depression treatment; and that measurement-based care based on symptom remission is not good for patients.
The first objection reflects a misunderstanding of primary and secondary study analyses. In each of the main randomized STAR*D studies,5–10 the primary outcome is clearly specified as the HDRS, with a secondary outcome being the QIDS-SR. What Dr Pigott finds objectionable are the results of the secondary analyses describing overall QIDS-SR remission rates at various treatment steps. However, these are clearly identified as post hoc analyses, where it is acceptable to examine secondary outcomes. In this case, the investigators justified using the QIDS-SR because there were more available data points than with the HDRS. In addition, patient-rated scales are more feasible and relevant to real-world clinical practice. Hence the controversy is not about how the results are reported, but about interpreting the meaning of the results.
Second, Dr Pigott objects to how symptom remission is defined and interpreted, and whether it is a useful construct. Remission was chosen as the primary outcome in STAR*D because of numerous previous studies showing that the presence of residual symptoms, that is, lack of remission, results in poorer outcomes, including increased risk of relapse, poorer functional outcomes, and reduced quality of life.11 The importance of remission has been reinforced in subsequent reviews12,13 and, as an internal validation, poorer outcomes in nonremitters was confirmed within the STAR*D study.14
Finally, Dr Pigott propagates a skewed view of measurement-based care by assuming that symptom outcomes are the only measurements of interest. Measurement-based care consists of using validated outcome scales in routine clinical practice to help guide and monitor treatment decisions.15 Studies have shown that measurement-based care for depression can improve patient outcomes, compared with treatment as usua1.16,17 We and others have long highlighted the need to refocus the objectives of treatment beyond simple symptom relief to improvement in psychosocial functioning and quality of life—outcomes that patients regard as more important.18–21 For example, the Canadian Network for Mood and Anxiety Treatments (commonly referred to as CANMAT) depression guidelines stipulate that remission of symptoms is only the initial goal in the treatment of depression—restoration of functioning and quality of life are other goals to be targeted.22 An example of a multi-dimensional approach to measurement-based care comes from the symptom, functioning, and quality of life data collected in STAR*D, which together were better at detecting overall patient improvement than symptom measures alone.23 Comprehensive measurement-based care should incorporate validated assessments of symptoms, side effects, functioning, and quality of life.19,24,25
In summary, we should be careful about discarding a healthy baby with the murky bathwater. The STAR*D study, despite its limitations, does provide important data, both for psychiatrists and for family physicians, in the evidence-based management of patients with depression. However, regardless of the validity of Dr Pigott’s arguments, almost everyone agrees that the response and remission rates in STAR*D were modest at best. This illustrates the urgent need for better predictors of individual treatment response. Scientifically rigorous efforts to discover clinically useful response biomarkers, such as that conducted by the Canadian Biomarker Integration Network in Depression,26 may eventually revolutionize treatment for depression. Until then, measurement-based care can provide helpful tools to optimize patient outcomes with our current depression treatments.
Acknowledgments
Dr Lam has received speaker and consultant honoraria or research funds from AstraZeneca, Brain Canada, Bristol-Myers Squibb, the Canadian Institutes of Health Research (CIHR), the Canadian Network for Mood and Anxiety Treatments, the Canadian Psychiatric Association, Eli Lilly, Johnson and Johnson, Lundbeck, Lundbeck Institute, Medscape, Merck, Mochida, Otsuka, Pfizer, Servier, St Jude Medical, Takeda, the University of British Columbia Institute of Mental Health–Coast Capital Savings, the University Health Network Foundation, and Vancouver Coastal Health Research Institute. Dr Lam also receives book royalties from Cambridge University Press, Informa Press, and Oxford University Press.
Dr Kennedy has received grant and research support from Bristol-Myers Squibb, CIHR, Clera Inc, Eli Lilly, GlaxoSmithKline, Janssen Ortho, Lundbeck, Ontario Brain Institute, Servier, and St Jude Medical. He is a consultant to AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Lundbeck, Pfizer, Servier, and St Jude Medical.
Abbreviations
- HDRS
Hamilton Depression Rating Scale
- QIDS-SR
Quick Inventory of Depressive Symptomatology—Self-Report
- STAR*D
Sequenced Treatment Alternatives to Relieve Depression
References
- 1.Fava M, Rush AJ, Trivedi MH, et al. Background and rationale for the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study. Psychiatr Clin North Am. 2003;26:457–494. doi: 10.1016/s0193-953x(02)00107-7. [DOI] [PubMed] [Google Scholar]
- 2.Sinyor M, Schaffer A, Levitt A. The Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial: a review. Can J Psychiatry. 2010;55:126–135. doi: 10.1177/070674371005500303. [DOI] [PubMed] [Google Scholar]
- 3.Trivedi MH, Rush AJ, Gaynes BN, et al. Maximizing the adequacy of medication treatment in controlled trials and clinical practice: STAR*D measurement-based care. Neuropsychopharmacology. 2007;32:2479–2489. doi: 10.1038/sj.npp.1301390. [DOI] [PubMed] [Google Scholar]
- 4.Pigott HE. The STAR*D trial: it is time to reexamine the clinical beliefs that guide the treatment of major depression. Can J Psychiatry. 2015;60(1):9–13. doi: 10.1177/070674371506000104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Rush AJ, Trivedi MH, Wisniewski SR, et al. Bupropion-SR, sertraline, or venlafaxine-XR after failure of SSRIs for depression. N Engl J Med. 2006;354:1231–1242. doi: 10.1056/NEJMoa052963. [DOI] [PubMed] [Google Scholar]
- 6.Trivedi MH, Fava M, Wisniewski SR, et al. Medication augmentation after the failure of SSRIs for depression. N Engl J Med. 2006;354:1243–1252. doi: 10.1056/NEJMoa052964. [DOI] [PubMed] [Google Scholar]
- 7.Thase ME, Friedman ES, Biggs MM, et al. Cognitive therapy versus medication in augmentation and switch strategies as second-step treatments: a STAR*D report. Am J Psychiatry. 2007;164:739–752. doi: 10.1176/ajp.2007.164.5.739. [DOI] [PubMed] [Google Scholar]
- 8.Fava M, Rush AJ, Wisniewski SR, et al. A comparison of mirtazapine and nortriptyline following two consecutive failed medication treatments for depressed outpatients: a STAR*D report. Am J Psychiatry. 2006;163:1161–1172. doi: 10.1176/ajp.2006.163.7.1161. [DOI] [PubMed] [Google Scholar]
- 9.Nierenberg AA, Fava M, Trivedi MH, et al. A comparison of lithium and T3 augmentation following two failed medication treatments for depression: a STAR*D report. Am J Psychiatry. 2006;163:1519–1530. doi: 10.1176/ajp.2006.163.9.1519. [DOI] [PubMed] [Google Scholar]
- 10.McGrath PJ, Stewart JW, Fava M, et al. Tranylcypromine versus venlafaxine plus mirtazapine following three failed antidepressant medication trials for depression: a STAR*D report. Am J Psychiatry. 2006;163:1531–1541. doi: 10.1176/ajp.2006.163.9.1531. [DOI] [PubMed] [Google Scholar]
- 11.Lam RW, Kennedy SH. Evidence-based strategies for achieving and sustaining full remission in depression: focus on metaanalyses. Can J Psychiatry. 2004;49(3 Suppl 1):17S–26S. [PubMed] [Google Scholar]
- 12.Paykel ES. Partial remission, residual symptoms, and relapse in depression. Dialogues Clin Neurosci. 2008;10:431–437. doi: 10.31887/DCNS.2008.10.4/espaykel. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mauskopf JA, Simon GE, Kalsekar A, et al. Nonresponse, partial response, and failure to achieve remission: humanistic and cost burden in major depressive disorder. Depress Anxiety. 2009;26:83–97. doi: 10.1002/da.20505. [DOI] [PubMed] [Google Scholar]
- 14.Nierenberg AA, Husain MM, Trivedi MH, et al. Residual symptoms after remission of major depressive disorder with citalopram and risk of relapse: a STAR*D report. Psychol Med. 2010;40:41–50. doi: 10.1017/S0033291709006011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Culpepper L, Trivedi MH. Using measurement-based care with patient involvement to improve outcomes in depression. Prim Care Companion CNS Disord. 2013;15(6) doi: 10.4088/PCC.12075co3c. pii: PCC.12075co3c. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Harding KJ, Rush AJ, Arbuckle M, et al. Measurement-based care in psychiatric practice: a policy framework for implementation. J Clin Psychiatry. 2011;72:1136–1143. doi: 10.4088/JCP.10r06282whi. [DOI] [PubMed] [Google Scholar]
- 17.Yeung AS, Jing Y, Brenneman SK, et al. Clinical Outcomes in Measurement-based Treatment (Comet): a trial of depression monitoring and feedback to primary care physicians. Depress Anxiety. 2012;29:865–873. doi: 10.1002/da.21983. [DOI] [PubMed] [Google Scholar]
- 18.Zimmerman M, McGlinchey JB, Posternak MA, et al. How should remission from depression be defined? The depressed patient’s perspective. Am J Psychiatry. 2006;163:148–150. doi: 10.1176/appi.ajp.163.1.148. [DOI] [PubMed] [Google Scholar]
- 19.Lam RW, Filteau MJ, Milev R. Clinical effectiveness: the importance of psychosocial functioning outcomes. J Affect Disord. 2011;132(Suppl 1):S9–S13. doi: 10.1016/j.jad.2011.03.046. [DOI] [PubMed] [Google Scholar]
- 20.Greer TL, Kurian BT, Trivedi MH. Defining and measuring functional recovery from depression. CNS Drugs. 2010;24:267–284. doi: 10.2165/11530230-000000000-00000. [DOI] [PubMed] [Google Scholar]
- 21.Lam RW, Parikh SV, Michalak EE, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) consensus recommendations for functional outcomes in major depressive disorder. Ann Clin Psychiatry. Forthcoming. [PubMed] [Google Scholar]
- 22.Patten SB, Kennedy SH, Lam RW, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) clinical guidelines for the management of major depressive disorder in adults. I. Classification, burden and principles of management. J Affect Disord. 2009;117(Suppl 1):S5–S14. doi: 10.1016/j.jad.2009.06.044. [DOI] [PubMed] [Google Scholar]
- 23.Cohen RM, Greenberg JM, IsHak WW. Incorporating multidimensional patient-reported outcomes of symptom severity, functioning, and quality of life in the Individual Burden of Illness Index for Depression to measure treatment impact and recovery in MDD. JAMA Psychiatry. 2013;70:343–350. doi: 10.1001/jamapsychiatry.2013.286. [DOI] [PubMed] [Google Scholar]
- 24.Valenstein M, Adler DA, Berlant J, et al. Implementing standardized assessments in clinical care: now’s the time. Psychiatr Serv. 2009;60:1372–1375. doi: 10.1176/ps.2009.60.10.1372. [DOI] [PubMed] [Google Scholar]
- 25.Zimmerman M, Young D, Chelminski I, et al. Overcoming the problem of diagnostic heterogeneity in applying measurement-based care in clinical practice: the concept of psychiatric vital signs. Compr Psychiatry. 2012;53:117–124. doi: 10.1016/j.comppsych.2011.03.004. [DOI] [PubMed] [Google Scholar]
- 26.Kennedy SH, Downar J, Evans KR, et al. The Canadian Biomarker Integration Network in Depression (CAN-BIND): advances in response prediction. Curr Pharm Design. 2012;18:5976–5989. doi: 10.2174/138161212803523635. [DOI] [PubMed] [Google Scholar]