Skip to content
The Delphi platform for research, expert consensus, and live decision-making
Surveylet Login
Calibrum
CalibrumAppropriateness research
RAND/UCLA Study Design

Using RAND/UCLA for real-world cases and what-if scenarios

Appropriateness decisions often depend on context. A case-based RAND/UCLA design lets experts rate a realistic base scenario, then test how carefully controlled changes affect the median, disagreement, and final classification.

Discuss Your Appropriateness StudyView RAND/UCLA Support

Why cases can reveal more than general statements

A broad recommendation may sound reasonable until it must be applied to a patient, setting, or decision with several interacting factors. A case-based design makes those assumptions explicit. Panelists evaluate the decision under defined conditions rather than silently imagining different situations.

The aim is controlled comparison. Each what-if scenario should change only the factor or small set of factors the study intends to test. If many details change at once, the cause of any rating shift becomes difficult to interpret.

A practical case-and-variation structure

Base case

Describe the population, indication, evidence context, alternatives, and relevant assumptions in a consistent format.

Controlled variation

Change one clinically or operationally important condition—such as severity, comorbidity, timing, setting, resources, or patient preference.

Comparable classification

Apply the same 1–9 scale, median bands, and disagreement rule to every case and variation so shifts can be interpreted consistently.

How the workflow can operate

Develop and validate the casesConfirm that each case is realistic, sufficiently complete, and free of unintended clues or irrelevant differences.
Collect independent first-round ratingsPanelists rate each option before group discussion and can explain the assumptions or concerns behind their judgments.
Review distributions and disagreementSummaries show the median, spread, extremes, and whether the predefined disagreement criterion changes the classification.
Discuss uncertain or sensitive scenariosStructured discussion focuses on the factors that produced uncertainty or divergent ratings rather than repeating areas of clear agreement.
Re-rate anonymouslyA second private rating captures the panel’s considered judgment without requiring public conformity.

What the analysis should show

Within each scenario

  • Median and interquartile range
  • Full rating distribution
  • Predefined disagreement result
  • Appropriate, uncertain, or inappropriate classification
  • Qualitative rationale and unresolved concerns

Across related scenarios

  • Which changed condition shifted the median
  • Whether disagreement appeared or resolved
  • Which options changed classification
  • Whether the direction of change was expected
  • Limits on applying the result beyond the defined case
A high median does not override disagreement. Under the study’s predefined RAND/UCLA rule, a scenario may be classified as uncertain when ratings occupy both extremes, even if the median lies in the appropriate or inappropriate range.

How Surveylet and Calibrum can help

Surveylet can organize base cases and related variations, collect separate ratings and comments, support multiple rounds, and export the data required for RAND/UCLA classification. Calibrum can help design the scenarios, configure the workflow, calculate and interpret the results, and produce a traceable final report.

View Surveylet FeaturesExplore Final Reporting

Recent research example

A 2026 multidisciplinary consensus study applied RAND/UCLA to four anonymized real-world oncology cases. Each case included predefined what-if variations. Ten specialists rated treatment options independently, reviewed first-round results in a structured discussion, and completed a second anonymous vote. Some altered conditions produced meaningful changes in ratings and classification.

Reference: Borque-Fernando A, Alonso-Gordoa T, Juan-Fita MJ, et al. Multidisciplinary expert consensus on treatment intensification in metastatic androgen pathway modulation-sensitive prostate cancer: a case-based RAND/UCLA analysis. Drugs in Context. 2026;15:2026-6-2.

Test how the decision changes when the context changes

Tell us the real-world decisions, case factors, and alternatives your panel needs to evaluate. We can help translate them into a consistent, analysis-ready RAND/UCLA workflow.

Plan Your Case-Based StudyExplore Support Options