Despite prevalent use in electronic devices, NRS and VAS are often implemented suboptimally.
Because the numeric rating scale (NRS) and visual analog scale (VAS) are frequently used in clinical outcome assessments (COAs), these (along with the verbal rating scale, which is not addressed here) are commonly migrated to electronic devices for use by trial participants to collect endpoint data. Despite their prevalence and the existence of migration best practice recommendations,1-3 we frequently see suboptimal implementation of these response scale types that could compromise a measure’s psychometric properties and trial outcomes. This article is a primer on the implementation of the NRS and VAS on eCOA systems and includes tips for ensuring your eCOA provider can accommodate best practice recommendations.
An NRS consists of numbers representing an ordered continuum of some attribute. The NRS typically includes verbal descriptors/anchors at the ends. Respondents select the number that reflects their perceived state.1 While the range on NRSs has varied,4 the most common is the 0-10 NRS.1
FDA’s patient-reported outcomes (PRO) guidance5 defines the VAS as a line of a fixed length (on paper, typically 100mm; eCOA uses 101 distinct response points, regardless of line length), with verbal anchors at the extreme ends, and no descriptors of the scale’s intermediate positions. Typically, no numbers are presented at the anchor positions. Respondents are asked to place a mark on the line corresponding to their perceived state.1
While both NRS and VAS have been used in endpoint measures to support labeling claims,6-8 a recent review9 suggests superiority of NRS over VAS in pain measures due to ease of administration and scoring, higher acceptability, and psychometric property retention. Further, while FDA guidance suggested VRS, VAS, and NRS are all suitable to measure pain,10 more recent guidance11 states the NRS is the most appropriate measure of pain intensity. In fact, FDA’s Patient-Focused Drug Development draft guidance3 states that FDA does not recommend the use of a VAS.11 These regulatory recommendations, as well as concern about VAS comprehension, are reasons why novel instrument developers may select other response scales over the VAS. Finally, the Critical Path Institute’s (C-Path’s) PRO Consortium recommends avoiding the VAS when developing de novo PRO measures and encourages use of other response scales appropriate to the measure’s content and the patient population.1,13
Regardless, because both NRS and VAS are frequently used in COAs and need to be implemented in eCOA solutions, it is imperative to follow best practice recommendations for faithful migration to retain a scale’s measurement properties. Figure 1 below shows examples of an NRS and VAS that align with best practice recommendations.1-3
Although VAS and NRS are common, some eCOA providers struggle to implement them according to best practice recommendations and offer unacceptable suggestions and/or formatting (e.g., NRS response options wrapped over multiple lines), incorrectly insisting that VAS lines must be 100mm long despite extensive comparability evidence,14 and displaying numeric values associated with a VAS response. In part, eCOA providers should be assessed based on the provider’s study-design capabilities.
Sponsors are urged to choose providers who demonstrate their system’s ability to accommodate complex designs, conditional logic, a user-friendly interface, and faithfully migrated measures. Sponsors are cautioned to scrutinize the capabilities of providers with no proven experience in eCOA data collection in clinical trials who propose non-clinical, generic data capture systems such as commercial online survey tools to capture eCOA data. The functionality required to display scales according to best practice recommendations, let alone complex solution design or custom report generation, is often not available in those generic systems. Those basic limitations could lead to deviations from eCOA best practice recommendations and result in low-quality data that compromises trial outcomes.
Ensuring your eCOA provider is aware of and has experience in implementing eCOA best practice recommendations is pivotal to collecting data that appropriately measures endpoints and withstands regulatory scrutiny.
Authored on behalf of Critical Path Institute’s eCOA Consortium by Celeste A. Elash, Vice President, eCOA Science, YPrime; and Anna Chilongo, eCOA Project Manager, ICON plc