Home / Resources / Psychological Measures / HAM-A
Hamilton Anxiety Rating Scale:
Scoring and Interpretation
The Hamilton Anxiety Rating Scale (HAM-A) is a clinician-rated measure of anxiety severity covering 14 symptom domains. Hamilton Anxiety Rating Scale scoring is a straightforward 0 to 56 sum, so this guide spends its time on what the total can and cannot say: interpretation without a universal cutoff, somatic overlap with medical illness, SPSS and R code, and a model reporting paragraph.
What the Hamilton Anxiety Rating Scale Measures
Hamilton designed the scale to rate the severity of anxiety states, not to diagnose a particular anxiety disorder (Hamilton, 1959). Its 14 domains span psychological distress and the bodily manifestations of anxiety, so a single total combines two kinds of symptom content.
Because some items concern bodily symptoms that medical illness or medication can also produce, a total score needs care in medically complex populations. In treatment studies it can summarize change in clinician-rated symptom burden, but it should be read alongside clinical judgment and the underlying diagnosis.
Structure of the HAM-A
- Items: fourteen clinician-rated symptom domains, each scored from 0 (absent) to 4 (severe).
- Subscales: psychic and somatic clusters are often discussed, but a universally validated two-factor score should not be assumed in every sample. The standard outcome is a single total.
- Direction: no reverse-scored items. Higher scores mean greater symptom burden.
- Administration: a clinical interview is required, and the observation window should be stated. This is not equivalent to a self-report worry questionnaire.
- 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 Hamilton (1959) and authorized clinical-rating resources.
How to Score the HAM-A
- Have an appropriately trained rater complete all 14 symptom ratings using the rubric defined for the study.
- Verify that every rating is an integer from 0 to 4 and that unrated items stay missing.
- Sum the 14 ratings to obtain hama_total, which ranges from 0 to 56.
- Apply a missing-item rule set by the protocol. No developer-endorsed prorating rule was verified, so the code below requires all 14 ratings. Report how missing ratings were handled.
* Illustrative syntax: hama1 to hama14 already hold ratings coded 0 to 4.
COMPUTE hama_total = SUM.14(hama1 TO hama14).
EXECUTE.
# data: one row per rating occasion; items hama1 to hama14 coded 0 to 4.
items <- paste0("hama", 1:14)
data$hama_total <- rowSums(data[items], na.rm = FALSE)
Interpreting HAM-A Scores
| Score band | Interpretation |
|---|---|
| 0 to 56 | Continuous clinician-rated anxiety severity; higher = more severe |
| Descriptive severity thresholds | Use only if a publication for your population and version supports the cutoff |
| Universal diagnostic threshold | None verified |
Source: Hamilton (1959); no universal diagnostic threshold was verified.
Higher totals represent more severe clinician-rated anxiety symptoms. An individual score belongs next to the person’s clinical history, comorbidity and assessment setting and is not a diagnosis. Commonly quoted severity bands should not be presented as if they came from one universally applicable receiver operating characteristic analysis.
The scale has substantial somatic content. Symptoms caused by another medical problem or by medication can inflate totals and complicate comparisons between clinical groups.
Psychometric Properties
| Property | Published finding | Source |
|---|---|---|
| Original development | A clinician-rating approach for anxiety states | Hamilton (1959) |
| Construct coverage | Psychic and somatic presentations both contribute to the total | Hamilton (1959) |
| Rater agreement | Depends on interviewing and rater training; compute and report your own coefficient | Study-specific |
| Diagnostic accuracy | No single sensitivity and specificity applies across anxiety disorders | Not established universally |
These are source-specific findings and cannot be assumed to hold as reliability or accuracy estimates in a new sample.
Using the HAM-A in a Study
In intervention research, define the primary endpoint and its timing before analysis. A repeated-measures or mixed-effects model can estimate change in the total while accounting for correlation within participants. In observational work, adjust for preselected clinical confounders, and remember that the somatic items may not be specific to anxiety, which matters most when comparing medical and psychiatric groups.
Because the ratings come from clinicians, Cronbach’s alpha is not a substitute for rater agreement. An intraclass correlation or a weighted agreement statistic from a design with more than one rater is usually more relevant, and the protocol should say how many raters scored and whether they were blind to treatment. Plan sample size from the primary contrast or model and not from a generic rule based on the 14 items.
Reporting the HAM-A
Anxiety symptom severity was measured with the 14-item Hamilton Anxiety Rating Scale (HAM-A; Hamilton, 1959). A trained [clinician role] rated each item from 0 to 4, and ratings were summed to a total score (range = 0 to 56), with higher scores indicating greater clinician-rated anxiety. Assessments were completed at [time points]. The statistical model was [model], with [missing-data approach] specified in advance.
References
Hamilton, M. (1959). The assessment of anxiety states by rating. British Journal of Medical Psychology, 32(1), 50-55. https://doi.org/10.1111/j.2044-8341.1959.tb00467.x
Frequently Asked Questions
Is the HAM-A a self-report questionnaire?
The standard HAM-A is clinician rated. If a study uses a self-report adaptation, describe and validate it separately and do not treat its scores as interchangeable.
Does the HAM-A diagnose generalized anxiety disorder?
No. It measures severity and does not replace a diagnostic interview.
Can a high HAM-A total reflect physical illness?
Yes. Some ratings concern bodily symptoms, so clinical context matters when interpreting a total in medically ill samples.
How should I handle a missing item?
Do not silently count it as zero. Define the scoring and missing-data procedure before analysis and report it.
NEXT STEP
Need HAM-A 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