Clinician-Prioritized Measures to Use in a Virtual Concussion Assessment: A Mixed-Methods Study (Preprint)

Abstract

There is little guidance available, and no uniform assessment battery used in either in-person and remote evaluations of people who are experiencing persistent symptoms post-concussion. There are, however, a variety of measures available to assess the physical domains affected by concussion and selecting the most appropriate measures for both in-person and virtually is challenging because of the lack of expert consensus and guidance.

This mixed-methods study used expert consensus processes to: i) identify clinical measures currently used to assess five physical domains affected by concussion (neurological examination, cervical spine, vestibular, oculomotor, effort), ii) document the available psychometric properties of the identified measures, and iii) determine the feasibility of applying the identified measures virtually.

The Delphi approach was used. In the first round, experienced clinicians were surveyed regarding their use of measures in concussion assessment. In the second round, clinicians reviewed information regarding the psychometric properties for all measures that were identified in the first round by at least 15% of participants. In the second round, experts rank-ordered the measures from most relevant to least relevant based on their clinical experience and the documented psychometric properties. A working group of 4 expert clinicians then determined feasibility to virtually administer the final set of measures.

Fifty-nine clinicians completed survey Round 1 listing all measures they used to assess the physical domains affected by a concussion. The frequency counts of the 146 different measures identified were determined. Thirty-three clinicians then completed the second-round survey and rank-ordered 22 measures that met the 15% cut-off criterion retained from Round 1. Measures ranked first (under the following domains: neurological examination, cervical spine, vestibular, oculomotor) were coordination, range of motion, Vestibular Ocular Motor Screening (VOMS), and smooth pursuits. These measures were deemed feasible to administer virtually by the working group members, however modifications for virtual administration were recommended, such as adjusting the method of measurement. There is limited information available on the psychometric properties of the Round 2 identified measures.

Clinicians ranked assessment of coordination, cervical spine range of motion, VOMS, and smooth pursuits as most relevant under their respective domains. Unfortunately, there are minimal documented psychometric properties for some of the identified measures suggesting the need for further study. Nevertheless, based on expert opinion, these clinical measures are considered feasible to administer for concussion physical examinations in the virtual context, with modifications.

RR2-10.2196/40446