The Hamilton Depression Rating Scale (HDRS) assigns a numerical score to the severity of a depressive episode. This score guides management, but its interpretation requires knowledge of thresholds, rating biases, and limitations that the number alone does not reveal. How does this score concretely distinguish between mild depression and a severe form, and where does its reliability end?
HDRS-17 Thresholds: Score Table and Severity Levels
The 17-item version (HDRS-17) remains the reference in clinical psychiatry and research. Each item is rated from 0 to 2 or from 0 to 4 depending on the dimension assessed. The total score, out of a maximum of 52, is divided into severity levels.
| HDRS-17 Score | Severity Level |
|---|---|
| Below 7 | No depression |
| 8 to 13 | Mild depression |
| 14 to 18 | Moderate depression |
| 19 to 22 | Severe depression |
| 23 and above | Very severe depression |
A patient rated at 12 falls into the mild range. A patient rated at 20 shifts into the severe zone. The difference may seem modest in points, but it reflects an accumulation of more intense or numerous symptoms, particularly regarding mood, psychomotor slowing, and suicidal thoughts.
Some references use slightly different cut-offs (for example, 10 to 13 for the mild level). Interpreting a score without specifying the chosen scale is akin to comparing measurements taken with different rules. Verifying the version and reference before drawing any conclusions is a non-negotiable prerequisite.
To delve deeper into the scoring logic and its practical implications, the advice from Santé Boost details each step of the evaluation process.
Somatic Weighting: The Bias that Distorts the Boundary Between Mild and Severe
Several items on the HDRS address physical dimensions: sleep disturbances, loss of appetite, fatigue, gastrointestinal symptoms, weight loss. These items can artificially inflate the score in a patient suffering from a somatic condition or taking medication that disrupts these functions.
A diabetic patient on metformin who sleeps poorly and has lost weight will accumulate points on these items without these symptoms necessarily being related to depression. Their score may exceed the threshold of 19 and indicate severe depression, even though the psychological component remains moderate.
Atypical Profiles Underestimated by the Scale
Conversely, hypersomnia and increased appetite are poorly captured by the HDRS-17. A patient who sleeps twelve hours a night and gains weight, two common markers of atypical depression, will receive a low score on the somatic items. The actual severity will be underestimated.
This weighting bias explains why the score alone is insufficient to distinguish between mild and severe. The HAS recommends complementing the numerical evaluation with a clinical interview focused on functional impact in social and professional life.
- A score of 15 in a patient whose physical illness explains part of the somatic symptoms may correspond to a milder depression than the number suggests.
- A score of 11 in a patient with marked hypersomnia and complete social withdrawal may mask a more severe form.
- The intersection between the HDRS score and the assessment of overall functioning (work, relationships, autonomy) provides a more reliable picture than the isolated score.
Inter-Rater Reliability: Why Two Clinicians Obtain Different Scores
The HDRS is a hetero-evaluation. The clinician observes, questions, and rates. Two evaluators may assign different scores to the same patient during the same interview. The discrepancy arises from the phrasing of questions, interpretation of responses, and the time spent on each item.
On an item like psychomotor slowing (rated from 0 to 4), an experienced clinician may detect a slowness of thought that their less trained colleague in Hamilton scoring might miss. A difference of one point on three or four items is enough to shift a patient from the mild range to the moderate range.
The Role of Structured Interview Guides
Structured interview guides exist to standardize the administration. They set the questions to be asked, the allowed prompts, and the precise scoring criteria for each level of each item. Reliability heavily depends on the evaluator’s training in these protocols.
Without this standardization, a score of 13 (upper limit of mild depression) or 14 (entry into the moderate zone) loses significance. The boundary area between two severity levels is the most sensitive to inter-rater variability.
HDRS vs. BDI and PHQ-9: Which Scale for Which Use
The Beck Depression Inventory (BDI) and the PHQ-9 are self-assessments: the patient fills out the questionnaire themselves. The HDRS, on the other hand, relies on the clinical judgment of the evaluator. This methodological difference changes the nature of what is measured.
- The PHQ-9 (9 items, self-assessment) is suitable for rapid screening in general medicine. It captures the patient’s subjective perception.
- The BDI (21 items, self-assessment) further explores the cognitive dimension of depression (guilt, self-esteem, pessimism).
- The HDRS-17 (17 items, hetero-evaluation) incorporates clinical observation of behavior, verbal fluency, and agitation, dimensions inaccessible to a self-questionnaire.
- The Montgomery-Asberg Depression Rating Scale (MADRS) is another hetero-evaluation, often considered more sensitive to change under antidepressant treatment.
A patient who minimizes their symptoms will obtain a low score on the BDI or PHQ-9, but a trained clinician may detect psychomotor slowing or agitation during the HDRS administration. Conversely, the HDRS captures cognitive subjective symptoms such as feelings of worthlessness or loss of hope less effectively, which are better explored by the BDI.
None of these tools alone diagnose. They quantify severity at a given moment. The choice of scale depends on the context (screening, follow-up, research) and the availability of a trained evaluator.
The boundary between mild and severe depression on the Hamilton scale hinges on a few points. These points depend on the version used, the clinician’s training, the patient’s somatic profile, and the chosen scoring reference. An HDRS score only holds value when cross-referenced with the clinical interview and functional assessment.
