How we measure depression matters—especially in the most severely ill patients.
In a recently accepted manuscript in the Journal of Affective Disorders, our team analyzed how three widely used depression rating scales perform in patients with markedly treatment-resistant depression (TRD) enrolled in the RECOVER trial. As a co-author and investigator at UTHealth Houston, I am proud to contribute to this important work, which helps clarify how we should assess outcomes in some of the most difficult-to-treat cases in psychiatry.
The study provides critical psychometric insights that have direct implications for clinical trials, interventional psychiatry, and routine clinical care.
Why This Study Was Needed
The RECOVER trial enrolled patients with extraordinarily severe illness:
These were not typical outpatient depression cases. These were individuals with markedly treatment-resistant, chronic depression.
Yet, until now, we did not know how well commonly used symptom scales perform in this population.
The three scales evaluated were:
The question was simple but fundamental:
Do these tools reliably and sensitively measure symptom change in profoundly treatment-resistant patients?
Key Findings
1️. All Three Scales Were Psychometrically Sound
Using both classical test theory and modern item response theory analyses, the study demonstrated that:
This is reassuring. These instruments remain appropriate even in markedly resistant populations.
2️. Not All Symptoms Contribute Equally
A striking and highly consistent finding emerged:
Across all three scales, the same five symptom domains were most sensitive to:
Those domains were:
In contrast, the following symptoms consistently showed lower sensitivity:
This does not mean these symptoms are clinically unimportant. Rather, it suggests that when measuring therapeutic response in TRD, core affective-cognitive symptoms may be the most sensitive indicators of change.
3️. QIDS Performed Slightly Better Than MADRS in Detecting Change
While all scales were valid, the QIDS-C and QIDS-SR demonstrated greater sensitivity in detecting:
One likely explanation:
The MADRS does not include an explicit energy/fatigue item, whereas both QIDS versions do.
Energy/fatigability proved to be one of the most responsive symptom domains in patients who were markedly resistant.
This has important implications for clinical trial design and outcome measurement in interventional psychiatry.
Why This Matters for Interventional Psychiatry
At the Center for Interventional Psychiatry at UTHealth Houston, we treat patients with:
In these cases, a partial response is often clinically meaningful. Detecting subtle but real improvements is critical.
This study suggests that:
As interventional treatments such as neuromodulation evolve—including TMS, ECT, VNS, DBS, and other device-based therapies—precision in outcome measurement becomes increasingly important.
A Broader Concept: Core vs. Associated Symptoms
One of the most interesting conceptual implications of this work is the possibility that depression may consist of:
Core symptom domains
Associated diagnostic symptoms
Core symptoms appear most tightly linked to overall severity and treatment response. Associated symptoms may reflect additional biological systems, medication effects, or subtype differences.
Understanding these distinctions may help refine:
Important Caveats
These findings derive from a markedly treatment-resistant population treated with VNS in a randomized trial. Results may differ in:
Clinical safety monitoring—including sleep, appetite, and suicidality—remains essential, regardless of statistical sensitivity.
Final Thoughts
In the most difficult cases of depression, measurement precision is not a luxury—it is a necessity.
This study strengthens confidence in widely used depression scales in profoundly treatment-resistant patients and provides valuable guidance on which symptom domains are most informative when evaluating treatment response.
For clinicians, researchers, and policymakers, the message is clear:
When treating markedly treatment-resistant depression, mood, interest, energy, self-view, and concentration may be the most reliable signals of meaningful improvement.
As we continue advancing interventional psychiatry at UTHealth Houston, understanding how to measure change accurately is foundational to improving outcomes for our most vulnerable patients.
Reference
Rush AJ, Sackeim HA, Lee Y-C, et al. Assessing symptom severity in patients with markedly treatment-resistant depression: Insights from the RECOVER trial. Journal of Affective Disorders. 2026.
Contact
Request for Second Opinion Form: https://Go.uth.edu/CIPIntake (external link)
Phone: (713) 486-2621
Fax: (713) 500-2728
E-mail: [email protected]
Website: https://go.uth.edu/CIP (external link)
Disclaimer
This article was created with the assistance of artificial intelligence (AI) to enhance clarity and readability. All medical and scientific content has been reviewed and approved by Joao L. de Quevedo, MD, PhD, Director of the Center for Interventional Psychiatry at the John S. Dunn Behavioral Sciences Center at UTHealth Houston. This content is for educational purposes only and does not substitute for professional medical advice.