Modification levels
3
Minor · Moderate · Substantial
Framework owner
ISPOR
Coons 2009 · Updated 2023
FDA position
Accept
With matched evidence
Equivalence Framework
Modify · Classify · Evidence · Document
Measurement equivalence is evidence that a patient-reported outcome (PRO) or clinical outcome assessment (COA) instrument produces psychometrically comparable data regardless of the device or mode used to administer it. The term applies whenever an instrument is migrated — from paper to electronic, between operating systems, across screen sizes, or from sponsor-provisioned to patient-owned (BYOD) devices.
The framework is owned by ISPOR. The original ePRO Good Research Practices Task Force report (Coons et al., Value in Health, 2009) defined three levels of instrument modification — minor, moderate, and substantial — each requiring a different evidence package. The 2023 update reflects mixed-modes and BYOD realities.
See also: BYOD in Clinical Trials · eCOA vs ePRO vs DHT · How to choose an eCOA vendor
Formatting, screen layout, fonts, and navigation. The item content, recall period, and response scale remain unchanged. Usability testing and cognitive interviewing are typically sufficient. A quantitative equivalence study is generally not required.
Pagination changes, response option modifications, or moving from paper checkboxes to digital radio buttons. Cognitive interviewing plus usability testing is needed; whether a quantitative equivalence study is required depends on the instrument and the population.
Changes to item wording, recall period, or scale type — for example, a paper visual analog scale (VAS) implemented as a digital slider. A quantitative equivalence study is typically required before the instrument can be used for a regulated endpoint.
If a PRO instrument was validated on paper and is moved to a smartphone without equivalence evidence, regulators may not accept the resulting data as the same construct. The endpoint can be lost — not because the data was missing, but because the instrument changed underneath it.
BYOD introduces multiple devices, operating systems, and screen sizes. FDA and ISPOR treat this as a modification question, not a hardware question. The equivalence required scales with the modification level — and most modern eCOA migrations with locked rendering fall into minor-to-moderate.
The FDA has accepted BYOD-collected ePRO data in regulatory submissions. Acceptance is conditional on demonstrating equivalence appropriate to the modification level, controlling the data system, and documenting the design in the protocol and statistical analysis plan.
Visual analog scales are the most-cited failure scenario for unscreened BYOD migration. Slider implementations on different screen sizes can change response distributions. If a VAS is part of a primary or key secondary endpoint, plan equivalence work before procurement.
Delve's eCOA/ePRO platform locks the rendering specification across iOS and Android so screen-size variation does not introduce mode effects that weren't planned for. The rendering control is documented in the system validation package.
The Delve clinical science team helps sponsors classify the modification level for each instrument before the eCOA is built — so the equivalence work scales to the instrument and the protocol, not to a generic checklist.
Where moderate or substantial modifications require it, Delve helps design the cognitive interviewing, usability testing, and small-sample quantitative equivalence study — and the documentation that goes with each.
Mixed-modes datasets stay clean in production when the support layer is real: multilingual concierge onboarding, same-day technical support, and protocol-aware patient outreach when assessments slip.
No. Validation establishes that an instrument measures the construct it claims to measure. Equivalence establishes that the same instrument, after migration, still measures that construct. A validated paper instrument can lose its validity when migrated without equivalence work.
Yes — when modifications are minor and ePRO design best practices were followed. The 2023 ISPOR update explicitly addresses this scenario. Moderate and substantial modifications still require quantitative work.
VAS, body maps, complex matrices, and instruments validated on a specific screen size. Muehlhausen et al. (2018) and the 2023 ISPOR update both flag these as high-risk for unscreened BYOD.
Typically a small-sample crossover or parallel-group design comparing the source and target modes, with predefined equivalence margins. Statistical analysis follows standard equivalence testing methods. Specifics depend on the instrument and the regulatory pathway.
Part 11 governs the electronic records and signatures — it is separate from instrument equivalence. Both apply. Part 11 covers audit trails, e-signatures, and access controls; equivalence covers whether the migrated instrument still measures the construct.
Delve helps sponsors classify the modification level, design the equivalence evidence package, and run the eCOA so the data your endpoint depends on actually arrives — across iOS, Android, and provisioned devices.
Talk to DelveRead: BYOD in Clinical Trials — When It Works, When It Fails →
Coons SJ, Gwaltney CJ, Hays RD, et al. Recommendations on evidence needed to support measurement equivalence between electronic and paper-based patient-reported outcome (PRO) measures: ISPOR ePRO Good Research Practices Task Force Report. Value in Health. 2009;12(4):419–429. · Eremenco S et al. PRO Data Collection in Clinical Trials Using Mixed Modes: ISPOR PRO Mixed Modes Good Research Practices Task Force. Value in Health. 2014. · Muehlhausen W et al. Measurement Equivalence of Patient-Reported Outcome Measures Migrated to Electronic Formats: A Review of Evidence and Recommendations for Clinical Trials and Bring Your Own Device. Therapeutic Innovation & Regulatory Science. 2018. · ISPOR 2023 Updated Recommendations on Measurement Comparability Among Modes of Data Collection. Value in Health. 2023.
Related: BYOD in Clinical Trials · eCOA vs ePRO vs DHT · How to choose an eCOA vendor · Why eCOA data fails in real-world trials