Abstract

Background: The six-minute walk distance (6MWD) is widely used to assess functional status and prognosis in interstitial lung disease (ILD). As exertional desaturation is common, indices combining distance and peripheral oxygen saturation (SpO2) may improve mortality prediction. Objectives: To evaluate the 5-year prognostic performance of the desaturation-distance ratio (original DDR), a new DDR, and the distance-saturation product (DSP). Methods: Retrospective cohort of adults with ILD completing a 6-min walk test (6MWT). Predictors included original DDR, new DDR, DSP, 6MWD, resting and end-exercise SpO2, and Delta SpO2. ROC analyses included patients with known 5-year follow-up; survival analyses included all patients with censoring. Multivariable Cox models adjusted for age, sex, forced vital capacity (FVC) %pred, and carbon monoxide lung diffusing capacity (DLCO) % pred. Results: We studied 183 patients (47% female; age 65 +/- 12 years); 166 (110 survivors, 56 non-survivors) were included in ROC analyses. AUCs were moderate (0.674-0.732), with original DDR and new DDR showing the highest discrimination (0.732 and 0.730, respectively). Exploratory cut-offs were original DDR>0.168, new DDR<29.06, DSP<377 m%, 6MWD<467 m, end-exercise SpO2 <= 85%, resting SpO2 <= 95%, and Delta SpO2 >= 9%. Dichotomized Cox models showed increased mortality risk across all high-risk strata. In continuous Cox models, original DDR and new DDR had the highest C-indices (0.706 and 0.705), slightly higher than 6MWD (0.652). In adjusted models, DSP and original DDR remained associated with mortality, with small increases in C-index over the base clinical model. Conclusions: Composite desaturation-distance indices provided moderate discrimination and modest prognostic gain beyond 6MWD and warrant external validation.