Home / Resources / Psychological Measures / MADRS
MADRS:
Scoring and Interpretation
The Montgomery-Åsberg Depression Rating Scale (MADRS) is a 10-item clinician-rated measure designed to be sensitive to change in depressive symptoms during treatment. MADRS scoring is a 0 to 60 sum, and this guide covers interpretation without a universal cutoff, the interview guide issue, analysis choices for trials, SPSS and R code, and a model reporting paragraph.
What the MADRS Measures
Montgomery and Åsberg drew the ten items from a broader psychopathology rating scale and kept those that best detected change with treatment, which made the MADRS a widely used severity endpoint in depression trials (Montgomery & Åsberg, 1979).
Its content differs from broader scales that give more weight to physical and sleep symptoms. It still rests on clinical judgment, so interviewer training, the version and the reference period must be documented. A structured interview guide can standardize ratings, but the guide has to be named and not assumed.
Structure of the MADRS
- Items: ten rated symptom domains, each assigned an integer from 0 (absent) to 6 (most severe).
- Subscales: none required. The standard outcome is the ten-item total.
- Direction: no reverse-scored items. Higher ratings mean more severe symptoms.
- Administration: clinician administered and interpreted with the interview findings. In common clinical use the reference period is the preceding week, so confirm it in your guide.
- Versions: structured-interview variants such as SIGMA exist, and self-report adaptations should be distinguished from the clinician-rated scale.
- Translations: use a validated version for the intended population and name the language and its source. If no validated translation exists, say so and describe how the version was adapted.
The instrument itself is available from the original Montgomery and Åsberg publication or authorized rating materials.
How to Score the MADRS
- Confirm that all ten ratings were assigned with the same MADRS interview and scoring guide.
- Validate that each item rating is an integer from 0 through 6.
- Sum the ten ratings to obtain madrs_total, which ranges from 0 to 60.
- For a partly missing assessment, follow the study protocol. The original paper gives no verified universal imputation rule, so the code below requires all ten ratings. Report how missing ratings were handled.
* Illustrative syntax: madrs1 to madrs10 already hold ratings coded 0 to 6.
COMPUTE madrs_total = SUM.10(madrs1 TO madrs10).
EXECUTE.
# data: one row per rating occasion; items madrs1 to madrs10 coded 0 to 6.
items <- paste0("madrs", 1:10)
data$madrs_total <- rowSums(data[items], na.rm = FALSE)
Interpreting MADRS Scores
| Score band | Interpretation |
|---|---|
| 0 to 60 | Continuous clinician-rated depression severity; higher = more severe |
| Remission or severity ranges | Use a prespecified published threshold matched to the trial population |
| Universal diagnostic cutoff | None |
Source: Montgomery & Åsberg (1979); numerical cut points need study-specific validation.
Higher scores indicate greater clinician-rated depressive symptom severity. Treatment studies often evaluate a prespecified endpoint or change in the total. Descriptive severity ranges are not equivalent to a psychiatric diagnosis and should not determine clinical care without an evaluation.
MADRS interpretation and rater agreement depend on consistent interviews. Report whether the analysis used a structured version, such as SIGMA, or another protocol, because totals from different protocols are not automatically comparable.
Psychometric Properties
| Property | Published finding | Source |
|---|---|---|
| Change sensitivity | The original study designed and evaluated the scale for sensitivity to clinical change | Montgomery & Åsberg (1979) |
| Construct validity | Correlated with other clinician-rated depression assessments in the validation work | Montgomery & Åsberg (1979) |
| Rater agreement | Depends on interview protocol and rater training; compute and report your own coefficient | Study-specific |
| Diagnosis | A severity rating, not an independent diagnostic test | Instrument purpose |
These are source-specific findings and cannot be assumed to hold as reliability estimates in a new sample.
Using the MADRS in a Study
In controlled trials, decide between a baseline-adjusted endpoint model and a change-score model, and prespecify the primary contrast. Adjusting for the baseline score in an ANCOVA-type model is generally the more efficient choice when baseline is measured, and a mixed-effects model suits repeated visits. In multi-site studies, check for rater drift and inconsistency across visits.
Report the number analyzed at each visit and the pattern of missing outcomes. The total is bounded and can show floor effects in remission samples, which should inform sensitivity analyses. Because ratings come from clinicians, rater agreement from a multi-rater design is the relevant reliability evidence, not Cronbach’s alpha alone.
Reporting the MADRS
Depressive symptom severity was assessed with the 10-item Montgomery-Åsberg Depression Rating Scale (MADRS; Montgomery & Åsberg, 1979). Trained [raters] completed ratings at [visits], and item ratings (0 to 6) were summed to a total (range = 0 to 60), with higher scores indicating greater severity. The primary contrast was [predefined endpoint or adjusted difference]. Rating consistency was addressed by [training method], and missing values were handled by [method].
References
Montgomery, S. A., & Åsberg, M. (1979). A new depression scale designed to be sensitive to change. The British Journal of Psychiatry, 134(4), 382-389. https://doi.org/10.1192/bjp.134.4.382
Frequently Asked Questions
How is the MADRS total calculated?
Sum the ten clinician-rated items, each scored from 0 to 6, after checking that every rating is valid. The total ranges from 0 to 60.
Is the MADRS the same as the HAM-D?
No. They differ in items, scoring and change sensitivity, so do not substitute raw totals from one for the other.
What is a MADRS remission score?
Trial protocols vary. Report and justify the specific cut point your study uses, and prespecify it.
Can I use the MADRS in a repeated-measures design?
Yes. A prespecified longitudinal model is often suitable when its assumptions and missing-data strategy fit the trial.
NEXT STEP
Need MADRS Data Scored
or Checked?
OnlineSPSS is a statistical consultancy for researchers. Share your design and dataset, and a statistically trained Research Coordinator reviews it. The scope and one fixed project fee are confirmed before work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014