Ask Mallard

Prevalence survey

A prevalence survey estimates what proportion of a defined population has a condition at a point in time, with the goal of a precise estimate rather than a comparison between groups.
NHMRC EVIDENCE LEVEL · 2009Not yet classifiedProvisional design classification. Study quality and certainty of findings require separate appraisal.
Why this level?Specify the research question: intervention, diagnostic accuracy, prognosis, aetiology, or screening intervention. Other question types are outside this hierarchy. NHMRC 2009 source

Prevalence is not one of the five question categories in this NHMRC 2009 table. An absence of classification is not a judgement about the quality or value of a prevalence survey.

What question does this help answer?

It answers "how common is this here, now", which underpins service planning, screening decisions and the interpretation of any diagnostic test used in that setting.

When to use it

Use it when the burden of a condition in a specific population is genuinely unknown, when a representative sample of that population can be drawn, and when no comparison group is needed to answer the question.

When not to use it

Do not use it to compare groups or to study causes; ambiguous temporality makes a cross-sectional comparison a weak basis for causal claims. And do not use it when the sampling frame cannot represent the population, because precision around a biased estimate is worse than no estimate.

How it works in plain language

A sample is drawn from a defined frame, each participant is assessed, and the proportion is reported with a confidence interval. Because the quantity is a single proportion, a Wilson or Clopper-Pearson interval is appropriate; a normal-approximation interval misbehaves when the proportion is near 0 or 1 and can run outside the possible range.

From target population to respondents

From target population to respondents. The sampling frame misses part of the target population, the drawn sample differs from the frame by chance, and nonresponse removes more people, so the respondents can differ from the population the question is about at every step.From the population in the question to the people who answeredTarget populationSampling frameNot in the frame: coverage errorSample drawnDiffers by chance: sampling errorRespondentsDid not respond: nonresponseA large number of responses does not repair a poor frame or selective participation: a precise estimateabout the wrong people is still wrong.
Conceptual diagram. It shows the structure of the idea, not data from any study.

Why Mallard may recommend it

Mallard may recommend a prevalence design when the question names a proportion in a population, when no exposure comparison is described, and when precision rather than a hypothesis test is what the investigator needs.

Key assumptions and design requirements

  • The target population and the sampling frame are stated, and the gap between them is acknowledged.
  • Sample size is driven by the PRECISION wanted, not by power for a comparison that is not being made.
  • The interval is appropriate for a single proportion (Wilson or exact), not a Wald interval near the extremes.
  • Non-response is reported and its likely direction discussed.
  • If the measurement is an imperfect test, apparent prevalence is distinguished from true prevalence.

What to watch out for

A prevalence estimate is only as good as the sampling frame: a convenience sample gives a precise number about the wrong population. Test imperfection matters too, since an imperfect test measures apparent rather than true prevalence unless sensitivity and specificity are accounted for.

Clinical example

Question: What proportion of adults attending a diabetes clinic have undiagnosed depression? Design: Sample consecutive attenders over a defined period, apply a validated instrument to each, and report the proportion with a Wilson interval, sizing the study for the width of interval that would be useful.

How to report it

No reporting guideline dedicated to prevalence surveys was found in a PubMed search on 2026-09-16. STROBE's cross-sectional checklist is the guideline to use. Report the sampling frame, the response rate with its denominator, and the interval method behind the estimate, since a prevalence study is judged on precision rather than on a comparison between groups. A reporting guideline is not a design manual: use it to make the study transparent, not as a substitute for careful design. Check the target journal's own requirements as well, which this page cannot know.

Code for this method
Loading…
Recent studies that used this
Searching PubMed…
Related Learn pages
Sources & further reading
  1. Observational research methods. Research design II: cohort, cross sectional, and case-control studies (opens in a new tab) · Emerg Med J · doi 10.1136/emj.20.1.54
  2. Overview of clinical study designs (opens in a new tab) · Clin Exp Emerg Med · doi 10.15441/ceem.23.036
  3. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies (opens in a new tab) · BMJ · doi 10.1136/bmj.39335.541782.AD
These are general design references. Neither addresses adjusting an apparent prevalence for imperfect test sensitivity and specificity, which needs its own method when the measurement is a test rather than a definitive assessment.
Last source review: 2026-09-06
AI-generated educational content
Mallard uses reputable methodological and reporting sources, but this material may contain errors or omit important context. Verify important methodological decisions with the cited sources and, when appropriate, a qualified research methodologist or statistician.