Test-retest reliability and interpretability of Long COVID Burden Index: A prospective observational study

The Long COVID Burden Index (LCBI) is an 11-item patient-reported outcome measure developed to assess the burden of long COVID symptoms on daily life. Although its cross-sectional validity and internal consistency have been demonstrated, its test-retest reliability, measurement error, and minimal important change (MIC) have not yet been evaluated. We conducted a longitudinal postal survey among adults with a history of COVID-19 hospitalization. The LCBI was administered at the initial survey, 2-week follow-up, and 3-month follow-up. Test-retest reliability was assessed among participants who completed the initial and 2-week surveys and reported no change on a 7-point global rating of change question. Measurement error was evaluated using the standard error of measurement (SEM) and the smallest detectable change (SDC) was calculated. The MIC was estimated at 3 months by using anchor-based methods: the mean change method, receiver operating characteristic curve analysis, and predictive modeling. Among 1031 potential participants, 301 provided valid responses at the initial survey, 258 at the 2-week follow-up, and 208 at the 3-months follow-up. The intraclass correlation coefficient for the LCBI total score was 0.78 (95% confidence interval, 0.72-0.83). The SEM and SDC were 3.26 and 9.04 points, respectively. The MIC estimates ranged from 1.73 to 2.91 points across methods. The LCBI showed good test-retest reliability. A reduction of approximately 2-3 points may represent a minimal important improvement. However, this change is below the SDC and does not provide definitive evidence of change beyond measurement error.