Health preference study checklist - from brief to submission

A health preference study has dozens of decision points required the final evidence package. This checklist covers helps that process.

A practical decision-point checklist for health preference DCE studies, organised by project phase. Use it to plan, review, and quality-assure your study at every stage.

Ben White, 07.07.2026

Decision points that determine study quality

Every health preference study involves hundreds of decisions - about study design, attribute selection, respondent recruitment, data quality, and modelling. Most of these decisions are made early in the project when the team has limited information and maximum uncertainty about what will go wrong. A checklist does not make the decisions for you but ensures they are made deliberately rather than by default.

What a health preference checklist covers

A comprehensive health preference study checklist must cover all eight project phases: research framing, qualitative design, quantitative design, ethics and consent, pilot, main fieldwork, analysis, and reporting. Within each phase, the checklist identifies the key decisions that determine study quality and the criteria for a good decision.

This checklist is synthesised from the ISPOR Good Research Practices guidelines (Bridges et al., 2011), the FDA Patient Preference Information guidance (2019), and SurveyEngine's experience across hundreds of health preference studies.

TLDR Quick links

SurveyEngine health resources

Research management service

Open a free account

The health preference study checklist

RESEARCH FRAMING:
☐ Decision context defined (FDA, NICE, EMA, commercial)
☐ Target population specified with diagnostic criteria
☐ Attributes linked explicitly to regulatory benefit-risk framework
☐ Study design method selected with justification (DCE vs BWS vs CV)
☐ Sample size estimated for required precision

QUALITATIVE PHASE:
☐ Qualitative research conducted with target patients (not surrogates)
☐ Attributes identified from patient-reported evidence
☐ Attribute list reviewed by clinician and patient representative
☐ Level ranges set from clinical data or real-world evidence

QUANTITATIVE DESIGN:
☐ D-efficient design generated and D-error documented
☐ Blocking implemented if required
☐ Consistency check (dominant alternative) embedded
☐ Practice task included
☐ Payment vehicle specified (if applicable)

ETHICS AND CONSENT:
☐ Ethics approval obtained or waiver documented
☐ ICF developed at appropriate reading level
☐ Patient representative review of ICF completed
☐ Data management plan consistent with GDPR/HIPAA

PILOT PHASE:
☐ Pilot conducted with 30–50 target patients
☐ Attribute comprehension checked via cognitive interview questions
☐ Consistency check pass rate ≥ 85%
☐ Preliminary parameters in expected direction
☐ All design revisions documented

MAIN FIELDWORK:
☐ Soft launch at 10% quota
☐ 20% interim review conducted
☐ Consistency check pass rate monitored
☐ Response time monitoring active
☐ AE monitoring in place (if required)

ANALYSIS:
☐ Data quality exclusions documented with counts at each stage
☐ MNL and mixed logit estimated
☐ WTP space model estimated for primary results
☐ Sensitivity analyses run (with/without excluded respondents)
☐ Preference heterogeneity reported

REPORTING:
☐ PREFS checklist completed
☐ Sample flow diagram included
☐ Full design matrix in supplementary materials
☐ WTP estimates with 95% CI
☐ Clinical data linked to WTP estimates for benefit-risk mapping

Worked example - checklist review for NICE submission

A quality review of a patient preference study intended for a NICE submission uses this checklist. Three items are initially incomplete: (1) the pilot data quality review is not documented - the team completed the pilot but did not write up the results formally; (2) the sample flow diagram is missing; (3) WTP estimates are presented in preference space rather than WTP space, with no confidence intervals.

Addressing these three items takes one week. The NICE reviewer subsequently notes that the evidence package is 'thorough and well-documented' - consistent with the checklist-driven quality assurance process.