Introduction: Clinical indicators or pathological features alone cannot reliably predict renal survival in patients with biopsy-confirmed diabetic nephropathy (DN). Therefore, this analysis sought to develop and validate a predictive model incorporating both clinical and pathological markers to predict renal outcomes in patients with biopsy-confirmed DN.
Methods: A predictive nomogram was developed based upon data pertaining to 194 patients with biopsy-confirmed DN. The prognostic relevance of individual clinicopathological variables was assessed through univariate and multivariate Cox regression analyses. A prognostic nomogram was then developed and validated based upon concordance (C)-index values and calibration curves. Internal validation was conducted through bootstrap resampling, while the clinical utility of this model was assessed via a decision curve analysis (DCA) approach.
Results: Nephrotic-range 24-h proteinuria, late-stage CKD, glomerular classification III-IV, and IFTA score 2-3 were all identified as independent predictors of poor renal outcomes in DN patients and were incorporated into our final nomogram. Calibration curves revealed good agreement between predicted and actual 3- and 5-year renal survival in DN patients with the C-index value for this nomogram at 0.845 (95% CI: 0.826-0.864). DCA revealed that our nomogram was superior to models based solely upon clinical indicators.
Conclusion: A predictive nomogram incorporating clinical and pathological indicators was developed and validated for the prediction of renal survival outcomes in patients with biopsy-confirmed DN. This model will be of value to clinicians, as it can serve as an easy-to-use and reliable tool for physicians to guide patient management based on individualized risk.
Keywords: Clinical characteristic; Diabetic nephropathy; Pathology; Prediction; Renal biopsy.
Copyright © 2021 by S. Karger AG, Basel.