Abstract
PURPOSE: To evaluate associations between ambient temperature and surgical failure after incisional glaucoma surgery. METHODS: Individual-level clinical data were obtained from electronic medical records. Ambient temperature was estimated via spatiotemporal bilinear interpolation. Multivariable random-effects Cox proportional hazards models, incorporating restricted cubic splines and segmented regression, assessed the temperature-outcome relationship. Stratified analyses identified potentially susceptible populations. Likelihood ratio tests examined interactive effects between air pollution and ambient temperature on both multiplicative and additive scales. RESULTS: Although conventional linear analyses did not show a significant association between ambient temperature and surgical failure risk, a spline-based nonlinear models suggested a U-shaped, asymmetric nonlinear association between ambient temperature and surgical failure risk within prolonged exposure windows (P_nonlinearity < 0.0001). Under these nonlinear specifications, exposure to temperature extremes relative to the mid-range was associated with an increased estimated risk of surgical failure. These nonlinear patterns appeared more evident in older patients and those with primary open-angle glaucoma. Air pollutants showed statistically significant effect modification in interaction analyses of the temperature-outcome association 180 days after surgery. With the exception of O₃, common air pollutants demonstrated negative interaction terms with temperature, suggesting that the modeled association between lower temperatures and surgical failure risk was stronger at higher pollutant concentrations. CONCLUSIONS: Nonlinear modeling suggested potential heterogeneity in risk across temperature ranges, with the estimated associations varying by PM₂.₅ exposure strata and other air pollutant levels. These findings are model-dependent and warrant confirmation in independent cohorts and prospective studies. TRANSLATIONAL RELEVANCE: Under spline-based nonlinear modeling, postoperative ambient temperature showed a U-shaped association with glaucoma surgical failure risk, with the estimated relationship differing across levels of concurrent air pollutant exposure. These findings are hypothesis-generating and suggest that environmental conditions may warrant consideration in postoperative risk assessment.