Introduction: Several prognostic models have been proposed to predict outcomes of patients affected by renal cell carcinoma. We analyze the discriminative capabilities of Karakiewicz, Kattan and Cindolo nomograms and perform a meta-analysis to yield pooled area under the receiver operator curves (AUCs) for model comparison. The end points of interest were disease-recurrence free survival (DFS) and cancer-specific survival (CSS). Methods: An electronic search of the Medline and Embase was undertaken until July 2014. The AUC value, total number of patients, number of disease recurrence, and cancer-related deaths were extracted from the included references. AUCs of the models were converted to odds ratios (ORs). For the meta-analysis, ln(OR) was used for data pooling. For each nomogram, the combined OR was transformed back to a converted AUC (cAUC). Results: A total of 16 studies were identified including 26 710 patients. The derived comparison of cAUC values revealed better predictive capability of DFS for the postoperative Karakiewicz nomogram versus Kattan nomogram (p < 0.01), but not versus Cindolo (p = 0.432) and between Cindolo versus Kattan (p = 0.03). The Mantel-Haenszel derived comparison of cAUC values revealed better predictive capability for the preoperative Karakiewicz nomogram versus the Kattan nomogram (p < 0.01) and versus the Cindolo model (p < 0.01), but also between the postoperative Karakiewicz model versus the Kattan model (p < 0.01) and the Cindolo model (p < 0.01). The Kattan model showed better discriminative capability versus the Cindolo model (p < 0.01). Conclusions: The predictive abilities of the pre-and postoperative Karakiewicz models are higher than Kattan or Cindolo in predicting DFS and CSS
Accuracy capabilities comparisons between Karakiewicz, Kattan and Cindolo nomograms in predicting outcomes for renal cancer carcinoma: A systematic review and meta-analysis.
RUSSO, GIORGIO IVAN;CIMINO, SEBASTIANO;MORGIA, Giuseppe Maria
2015-01-01
Abstract
Introduction: Several prognostic models have been proposed to predict outcomes of patients affected by renal cell carcinoma. We analyze the discriminative capabilities of Karakiewicz, Kattan and Cindolo nomograms and perform a meta-analysis to yield pooled area under the receiver operator curves (AUCs) for model comparison. The end points of interest were disease-recurrence free survival (DFS) and cancer-specific survival (CSS). Methods: An electronic search of the Medline and Embase was undertaken until July 2014. The AUC value, total number of patients, number of disease recurrence, and cancer-related deaths were extracted from the included references. AUCs of the models were converted to odds ratios (ORs). For the meta-analysis, ln(OR) was used for data pooling. For each nomogram, the combined OR was transformed back to a converted AUC (cAUC). Results: A total of 16 studies were identified including 26 710 patients. The derived comparison of cAUC values revealed better predictive capability of DFS for the postoperative Karakiewicz nomogram versus Kattan nomogram (p < 0.01), but not versus Cindolo (p = 0.432) and between Cindolo versus Kattan (p = 0.03). The Mantel-Haenszel derived comparison of cAUC values revealed better predictive capability for the preoperative Karakiewicz nomogram versus the Kattan nomogram (p < 0.01) and versus the Cindolo model (p < 0.01), but also between the postoperative Karakiewicz model versus the Kattan model (p < 0.01) and the Cindolo model (p < 0.01). The Kattan model showed better discriminative capability versus the Cindolo model (p < 0.01). Conclusions: The predictive abilities of the pre-and postoperative Karakiewicz models are higher than Kattan or Cindolo in predicting DFS and CSSFile | Dimensione | Formato | |
---|---|---|---|
Karakiewicz, Kattan and Cindolo nomograms for renal cancer carcinoma.pdf
accesso aperto
Tipologia:
Versione Editoriale (PDF)
Dimensione
6.82 MB
Formato
Adobe PDF
|
6.82 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.