SF-12v2 Health Survey: physical and mental health scores

What this assessment is

The SF-12v2 Health Survey is a 12-item adult self-report measure of general physical and mental health.

Its main results are a Physical Component Summary score and a Mental Component Summary score. These are often shortened to PCS and MCS.

There is no standard 12-item total.

The scores use weighted combinations of responses. They are not simple sums, percentages, or averages.

This page covers version 2. The original SF-12 and the current SF-12v2 differ in wording, response structure, and scoring.

Version must match.

Use v2.

The scores describe reported health. They do not diagnose a medical or mental health condition.

Who it was designed for

John Ware and colleagues created the original SF-12 by selecting 12 items from the SF-36. The 1996 construction study evaluated how well the shorter survey reproduced physical and mental summary scores.

SF-12v2 followed in 2000 and is the version currently identified in the official instrument record.

The survey was designed for adults across health conditions and the general population. It is used in clinical research, population studies, service evaluation, and outcome monitoring.

The person reports their own health. Interview administration can be used with the matching validated mode.

The survey is broad. It is useful when a short general health summary is needed, but it is not designed to replace a detailed condition-specific measure.

Breadth limits detail.

How it is administered

The respondent answers 12 questions about health and daily function.

The standard version uses a four-week recall period. An acute version uses the previous week. Other specialized forms also exist.

Record the recall version.

Recall matters.

The survey usually takes two or three minutes. Use one validated language and administration mode.

Do not combine responses from SF-12 version 1, SF-12v2, SF-36, or RAND-36. Similar names do not make the forms interchangeable.

The scoring method depends on the exact item coding and licensed coefficients. Keep item identifiers and response values intact when collecting data.

Weights matter.

For repeated assessment, use the same version, recall period, language, mode, and scoring norm.

What it measures

The 12 questions sample eight broad health areas:

• physical functioning
• role limitations due to physical health
• bodily pain
• general health
• vitality
• social functioning
• role limitations due to emotional problems
• mental health

The short form uses this information mainly to estimate two broad components.

PCS summarizes the pattern most closely related to physical health. MCS summarizes the pattern most closely related to mental health.

Two outputs.

Both scores use information from several areas. They are not pure measures of one body system or one mental health condition.

They are summaries.

Some scoring systems can estimate the eight domain scales, but those estimates are less precise than results from the longer SF-36v2. State exactly which outputs were calculated.

How scoring works

Responses are first recoded so their direction is consistent with better reported health.

The scoring algorithm then applies different weights to the responses. It calculates PCS and MCS separately.

The standard norm-based results use a T-score metric. Under QualityMetric's norm guidance, a score of 50 is the named reference mean and 10 points equal one reference standard deviation.

Fifty is not a cutoff.

Higher scores indicate better reported health. The achievable range is not 0 to 100. It depends on the exact survey form, scoring coefficients, norm set, and missing-data method.

Do not calculate PCS or MCS by adding selected answers. Do not label a raw sum as an SF-12v2 score.

Use licensed scoring.

Use the official scoring method and state the norm set. QualityMetric's guidance explains why norm-based values do not have 0 and 100 as their endpoints.

Do not improvise.

Follow the selected scoring system's missing-data rule. Do not invent a local prorating method.

How to interpret SF-12v2 scores

Read PCS and MCS together.

A higher PCS indicates better reported physical health relative to the selected norm. A higher MCS indicates better reported mental health relative to that norm.

A value below 50 means below the reference mean. It does not prove illness or impairment.

Scores need context.

There are no universal mild, moderate, or severe bands. Interpretation depends on the reference population, condition, purpose, and measurement uncertainty.

Small differences should not be treated as certain individual change without considering reliability and context.

When tracking progress, keep the survey and scoring settings fixed. A switch from four-week to one-week recall can change the result even when health is stable.

Compare like with like.

The two components can move differently. That pattern can guide follow-up questions, but it cannot identify a diagnosis or cause.

Ask what changed.

What the scores cannot tell you

SF-12v2 cannot diagnose depression, anxiety, disability, chronic pain, or physical illness.

MCS is not a psychiatric screening result. PCS is not a test of fitness or physical capacity.

The survey cannot show which symptom or condition caused a low score. It also cannot identify which treatment caused a change.

A broad component can hide one important problem. A person may have a near-average PCS while still reporting severe difficulty in an area represented by only one question.

The scores do not establish work capacity, treatment need, or eligibility for services.

Use them with a history, focused assessment, and the person's goals.

Evidence and limitations

The main strength of SF-12v2 is efficiency. It provides two general health summaries with little respondent burden.

The tradeoff is detail. Twelve questions cannot describe eight health areas as precisely as a longer survey.

The weighted component model can also be hard to explain. PCS and MCS are not direct counts of physical and mental problems. Their values depend on cross-domain weights and the selected reference norms.

Different countries and populations can have different score distributions. Report the norm set rather than assuming every value is directly comparable.

Use SF-12v2 for a broad health summary. Add a focused measure when the clinical or research question needs detail.

Short measures answer broad questions.