Skip to content
Back to search results

SF-12 Health Survey

Introduced by
John E. Ware Jr., Mark Kosinski, and Susan Keller
Year
1996
Type
Brief self-report health status instrument
Family
Health-related quality of life
Font size

What does it do?

The SF-12 is a brief, 12-item version of the SF-36 health survey developed by Ware, Kosinski, and Keller in 1996.

Designed to reduce respondent burden while maintaining psychometric validity, it has become the standard instrument for large-scale surveys, epidemiological studies, and health outcomes research where administration time is critical.

The intuition behind the method

The SF-12 captures the essential health dimensions of the longer SF-36 by selecting one or two items from each of its eight domains. The resulting 12 items are scored into two composite measures: the Physical Component Summary (PCS) and Mental Component Summary (MCS). This ultra-brief design makes it ideal for time-constrained settings like general practice, telephone surveys, and large population studies, while retaining the psychometric rigor needed for clinical decision-making.

History

The SF-12 was developed in 1996 by Ware, Kosinski, and Keller at the New England Medical Center as a response to the need for shorter health measures in large surveys. It was designed to preserve the conceptual framework and psychometric properties of the SF-36 while reducing response burden by approximately 75%. The SF-12v2 was released in 2002 with improved item wording and updated scoring algorithms based on larger normative samples. The SF-12 has since become the standard health measure in many national health surveillance programs.

When to use it

The SF-12 is ideal for large surveys, epidemiological studies, and clinical settings where brevity is essential. It is the preferred instrument for national health surveillance (e.g., Behavioral Risk Factor Surveillance System, National Health Interview Survey) and cohort studies requiring repeated assessments. The SF-36 should be used instead when detailed domain-specific information is needed or when the few minutes saved by the SF-12 are not a critical constraint.

Limitations and considerations

  • Loss of domain-specific information compared to SF-36; cannot assess individual dimensions like pain or vitality in detail
  • Reduced sensitivity to change in some specific health domains due to single-item representation
  • Scoring requires access to specific regression algorithms; manual scoring is complex without software tools

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Minimal administration time (2–3 minutes) enables use in high-volume clinical and survey settings
  • Maintains strong correlation with SF-36 domain scores despite brevity
  • Extensively validated in diverse populations, languages, and health conditions
  • Two summary scores (physical and mental) provide clear interpretation of health burden
  • Ideal for longitudinal tracking in resource-constrained health systems

Applications

  • National health surveillance: Behavioral Risk Factor Surveillance System (BRFSS) uses SF-12 to track population health trends annually
  • Large epidemiological cohorts: longitudinal tracking of health outcomes in thousands of participants with minimal burden
  • Insurance claims data: embedding SF-12 scores in electronic health records for claims-based outcome measurement
  • Telehealth and remote monitoring: brief screen during virtual visits to assess overall health and guide clinical prioritization
  • Vaccine and pharmaceutical trials: rapid health status assessment in multi-site, high-enrollment clinical trials

How it works

How it works

1

The SF-12 consists of 12 items with response formats identical to the SF-36: some items use 5-point Likert scales, others use 3-point or yes/no binary responses. The instrument is self-completed in 2–3 minutes, making it highly practical for busy clinical and survey contexts.

2

The 12 items are selected from the eight SF-36 dimensions and weighted using regression coefficients derived from population surveys. Items are coded, reverse-scored where necessary, and aggregated into two standardized component scores: Physical Component Summary (PCS) and Mental Component Summary (MCS). Both scores have a mean of 50 and standard deviation of 10 in the general US population.

3

Each component score ranges from 0 to 100, with higher scores indicating better health status. Scores are interpreted relative to population norms: a score of 50 represents the average for the general population. Scores below 40 suggest substantial health impairment; above 60 indicates better than average health. The two scores can be examined separately to identify whether limitations are primarily physical or mental.

Common pitfalls

  • Using raw item scores directly instead of converting to standardized component scores; the 12 items cannot be summed like a traditional questionnaire
  • Confusing SF-12 PCS/MCS with SF-36 subscale scores; they use different calculation methods and normative frameworks
  • Administering SF-12 and interpreting results without reference to the specific version (v1 vs v2); scoring algorithms differ between versions

Frequently asked questions

Why is SF-12 scoring more complicated than a simple sum of items?

The SF-12 uses weighted regression coefficients to convert raw item responses into standardized component scores. This approach preserves the psychometric structure of the longer SF-36 and ensures scores are comparable across populations. Simplified scoring would lose this validity; software tools are recommended for accurate scoring.

When should I use SF-12 instead of SF-36?

Use SF-12 when administration time is critical (surveys, high-volume screening, telephone administration) or when you only need overall physical and mental health summaries. Use SF-36 when detailed domain assessment is important, such as understanding whether pain, vitality, or social functioning drives low scores.

What does a Physical Component Score of 45 mean?

A PCS of 45 is half a standard deviation below the population mean of 50, indicating slightly below-average physical health status. This might reflect mild activity limitations, pain, or fatigue. In context, a 45 in a 75-year-old patient may represent typical aging, while the same score in a 40-year-old suggests health concern.

How sensitive is SF-12 to detecting health changes compared to SF-36?

The SF-12 is slightly less sensitive to small changes in specific health dimensions due to using fewer items per domain. For studies expecting small effect sizes, SF-36 may be preferable. However, for detecting clinically substantial changes (≥5 points in component scores), SF-12 performs adequately.

Related methods

Sources

  1. Ware, J. E., Kosinski, M., & Keller, S. D. (1996). A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Medical Care, 34(3), 220–233.

    Open DOI reference
    Copied.
  2. Ware, J. E., Kosinski, M., & Keller, S. D. (2002). SF-12: How to score the SF-12 Physical and Mental Health Summary Scales (3rd ed.). QualityMetric.

    Open publication reference
    Copied.
  3. Gandek, B., Ware, J. E., Aaronson, N. K., et al. (2004). Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries. Journal of Clinical Epidemiology, 51(11), 1171–1178.

    Open DOI reference
    Copied.