Test-retest variability of the full-field stimulus test in patients with retinitis pigmentosa: REPEAT Study Report No. 4.

Purpose: To evaluate test-retest variability (TRV) of the full-field stimulus test (FST) in patients with retinitis pigmentosa (RP) and poor best-corrected visual acuity (BCVA; ≤20/50 Snellen; ≥0.40 logMAR), and to assess the reliability of FST as a clinical endpoint in future RP trials.

Methods: In this prospective cohort study, 28 patients with clinically diagnosed RP performed FST (Diagnosys) on two different days, 2 weeks apart. Retinal sensitivity estimates on FST for blue, red, and white stimuli were determined, and the difference between blue and red thresholds was calculated to identify photoreceptor mediation type. TRV was analysed using Bland-Altman analyses and coefficients of repeatability (CoRs).

Results: Patients had a median age of 38.0 years (interquartile range [IQR]: 24.5-57.5), and a median BCVA of 1.1 logMAR (IQR: 0.6-1.2). Following Bland-Altman analyses, learning effects (biases) across the FST measurements were minimal, ranging from -0.02 to 0.02 log units. The CoRs were ± 0.38 log units (95% confidence interval [CI]: 0.30-0.52) for blue stimuli, ±0.30 log units (95%CI: 0.23-0.40) for red stimuli, ±0.59 log units (95%CI: 0.47-0.81) for white stimuli, and ± 0.30 log units (95%CI: 0.24-0.41) for the blue-red difference. Patients with mixed photoreceptor mediation had a higher TRV than those with rod or cone mediation. Age was a significant predictor of TRV on blue light FST (p = 0.028).

Conclusions: These findings support the use of FST as a reliable measurement of retinal sensitivity in patients with RP, and provide a strong foundation for developing clinical endpoints for FST based on the CoRs to improve assessment of changes in RP, for example, before and after treatment.