Construct validity of self-reported and interview-guided administration methods of the Danish version of the Post-COVID-19 Functional Status scale

Lotte Sørensen^1*Jane Agergaard²Trine Brøns Nielsen³Berit Schiøttz-Christensen²Cecilia Hee Laursen⁴Steffen Leth⁵Claus Vinther Nielsen⁴Lisa Gregersen Oestergaard³

• ¹Department of Physiotherapy and Occupational Therapy, Aarhus University Hospital, Aarhus, Denmark
• ²Aarhus Universitetshospital Afdeling for Infektionssygdomme, Aarhus, Denmark
• ³DEFACTUM, Central Denmark Region, Aarhus, Denmark
• ⁴Department of Social Medicine and Rehabilitation, Gødstrup Hospital, Herning, Denmark
• ⁵Department of Internal Medicine and Rehabilitation, Gødstrup Hospital, Herning, Denmark

Introduction The Post-COVID-19 Functional Status (PCFS) scale was quickly adopted into COVID-19 research and clinical practice worldwide to monitor functional status and recovery. The scale has been translated into Danish, and three different administration methods have been employed. However, clinicians have expressed concerns about the scale's ability to capture work-related functional limitations. Therefore, the purpose of this study was to evaluate the construct validity of three different administration methods of the Danish version of the PCFS scale. Methods This cross-sectional study included patients with long COVID who completed three versions of the PCFS scale: a questionnaire-based version, a flowchart-based version, and an interview-based version. The construct validity was evaluated following the Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) guidelines by testing predefined hypotheses that compared the PCFS scale with sick leave and EuroQoL Five-dimensions Five level (EQ-5D-5L). Results A total of 437 patients, with a mean age 48 years, 75% female, and 59% on sick leave, were included in this study. Statistically significant differences between the three administration methods were found. Of the 234 patients on sick leave, only 50-54% had a PCFS grade ≥3 which was below our predefined hypothesis. Furthermore, correlations between the PCFS scale and EQ-5D-5L was lower than hypothesized. Conclusion None of the three administration methods effectively captured work-related functional limitations associated with being on part-time or full-time sick leave. Additionally, correlations with quality of life were lower than expected. Overall, the construct validity of the PCFS scale was only partially supported.