Choosing optimal outcome measures maximizes statistical power, accelerates discovery and improves reliability in early-phase trials. We devised and evaluated a modification to a pragmatic measure of oxygenation function, the S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 ratio. Because of the ceiling effect in oxyhaemoglobin saturation, S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 ratio ceases to reflect pulmonary oxygenation function at high S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 values. We found that the correlation of S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 with the reference standard (S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94/S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 ratio) improves substantially when excluding S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 and refer to this measure as S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94. Using observational data from 39,765 hospitalised COVID-19 patients, we demonstrate that S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 is predictive of mortality, and compare the sample sizes required for trials using four different outcome measures. We show that a significant difference in outcome could be detected with the smallest sample size using S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94. We demonstrate that S/FS/FSpO2S/FPaO2FIO2SpO2>0.94S/F94S/F94S/F94S/F94 is an effective intermediate outcome measure in COVID-19. It is a non-invasive measurement, representative of disease severity and provides greater statistical power.
There is a need for an accurate measure of pulmonary oxygenation function that can be used as an intermediate endpoint in pragmatic clinical trials, to increase statistical power and efficiency. Here, the authors show that the S/F94, a modification of the S/F ratio, is a simple, meaningful and effective intermediate outcome measure.