Abstract: TH-PO491
Post Hoc Analysis of the DialCheck Tool to Predict Renal Failure in IgA Nephropathy (IgAN)
Session Information
- Glomerular Diseases: Clinical, Outcomes, Trials - I
November 03, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
Abstract Time: 10:00 AM - 12:00 PM
Category: Glomerular Diseases
- 1303 Glomerular Diseases: Clinical‚ Outcomes‚ and Trials
Authors
- Schena, Francesco Paolo, Universita degli Studi di Bari Aldo Moro, Bari, Puglia, Italy
- Anelli, Vito walter, Politecnico di Bari, Bari, Puglia, Italy
- Cornacchia, Giandomenico, Politecnico di Bari, Bari, Puglia, Italy
- Abbrescia, Daniela Isabel, Fondazione Schena, Bari, Italy
- Di noia, Tommaso, Politecnico di Bari, Bari, Puglia, Italy
Background
We have developed the DialCheck tool to predict renal failure (RF) in patients with IgAN (KI 2021). The tool is based on two different artificial neural networks. The first network predicts RF development while the second predicts the time frame to reach the outcome. The application of the tool in the clinical practice shows discordance between predicted and observed probabilities to develop KF in about 20% of IgAN patients.
We have conducted a post-hoc analysis to evaluate the discordance between predicted and observed probabilities to reach KF.
Methods
Baseline demographic and laboratory data were analyzed using mean ± standard deviation (SD) or median and interquartile range (IQR) for continuous variables. The D'Agostino-Pearson normality test was applied to the data in order to assess their distribution. The categorical variables were expressed as absolute numbers or percent frequency. Dichotomous and polychotomous baseline characteristics were compared using the Chi-square test. Continuous baseline characteristics were compared using the Student’s t-test or the Mann-Whitney U test. All analyses were performed using GraphPad Prism Software version n.5 (GraphPad Software, San Diego, CA). The statistical significance value of p< 0.05 was adopted.
Results
Tool performance was analyzed in the retrospective cohort of 1,116 European adult IgAN patients followed for a median time of 88 months (IQR 49 - 134). Discordance was observed in 216 (19.35%) IgAN patients. In the first set of patients with predicted no RF and reached outcome, subjects were olders and hypertensives. They had high value of serum creatinine, reduced renal function and moderate or severe renal lesions. Many of those patients did not receive therapy or were non-responders to therapy. In the second set of patients the tool predicted RF but this outcome was not reached, the patients were responders to therapy. Thus, in the discordant group (prediction did not match the observed outcome) therapy was strongly associated with outcome (P<0.0001).
Conclusion
The DialCheck tool could help physicians to determine the prognosis of the disease as well as could help patients to plan for their future. The non-correct prediction of RF in a small number of cases may be due to no therapy or non-responders to therapy.
Funding
- Government Support – Non-U.S.