- Email: [email protected]
1003 Development and validation of nomograms to predict the recovery of urinary continence after radical prostatectomy: Comparisons between immediate, early, and late continence
1003
Development and validation of nomograms to predict the recovery of urinary continence after radical prostatectomy: Comparisons between immediate, early, and late continence Eur Urol Suppl 2013;12;e1003
Jeong S.J., Lee H. S., Kim K. B., Lee J. K., Lee B. K., Jeong J. W., Cha W. H., Lee S. C., Jeong C. W., Hong S. K., Byun S-S., Lee S. E. Seoul National University Bundang Hospital, Dept. of Urology, Seongnam, Gyeonggi, South Korea INTRODUCTION & OBJECTIVES: Few studies have been conducted on the serial evaluation of predictors for recovery of urinary continence (RC) after radical prostatectomy (RP) among the same cohort. We tried to develop and validate nomograms to predict immediate (≤1 month), early (≤3 months), and late (≤12 months) RC and compared each nomogram from a contemporary RP series conducted by a single surgeon in Korean men. MATERIAL & METHODS: Among consecutive men who received RP from a single surgeon between 2004 and 2011, 601 (76.6%) and 184 (23.4%) were randomly assigned to two subcohorts for the development of nomograms and for the split-sample external validation. Achievement of RC was defined as wearing no pads or an occasional security pad. Time to RC and percentage of continent men at each follow-up were calculated using Kaplan-Meier curve estimates. The final multivariate model was selected based on the stepwise procedure and by considering the clinical importance of the variables in the model. The regression coefficient-based nomograms were developed based on final multivariate logistic regression models. RESULTS:
Age at surgery, membranous urethral length and robot-assisted RP were significant for RC at 1, 3 and 12 months. Saving the neurovascular bundle (NVB) and prostate volume were significant only for RC at 12 months. Odds ratios for age and membranous urethral length were constant over time, whereas the odds ratio for robot-assisted RP decreased over time (Fig. 1). The discrimination performance of the multivariate logistic models demonstrated 67, 68, 73% accuracy for each RC time point. Each nomogram was very well fitted to the ideal line of the calibration plot. External validation of the developed nomograms indicated 66, 67, 72% accuracy for each RC time point. (Fig. 1. Nomograms based on the 601 patients in the development cohort for predicting (A) immediate (≤1 month), (B) early (≤3 months), and (C) late (≤12 months after surgery) recovery of urinary continence after radical prostatectomy). CONCLUSIONS: We developed nomograms for immediate, early, and late RC after RP from a contemporary series of RP among Korean men, and validated these nomograms adequately. Our findings suggest that saving the NVB and prostate volume may only affect late RC after RP. In contrast, age at surgery, membranous urethral length and robotassisted RP seem to be consistently associated with immediate, early, and late RC after surgery.
Development and validation of nomograms to predict the recovery of urinary continence after radical prostatectomy: Comparisons between immediate, early, and late continence Eur Urol Suppl 2013;12;e1003
Jeong S.J., Lee H. S., Kim K. B., Lee J. K., Lee B. K., Jeong J. W., Cha W. H., Lee S. C., Jeong C. W., Hong S. K., Byun S-S., Lee S. E. Seoul National University Bundang Hospital, Dept. of Urology, Seongnam, Gyeonggi, South Korea INTRODUCTION & OBJECTIVES: Few studies have been conducted on the serial evaluation of predictors for recovery of urinary continence (RC) after radical prostatectomy (RP) among the same cohort. We tried to develop and validate nomograms to predict immediate (≤1 month), early (≤3 months), and late (≤12 months) RC and compared each nomogram from a contemporary RP series conducted by a single surgeon in Korean men. MATERIAL & METHODS: Among consecutive men who received RP from a single surgeon between 2004 and 2011, 601 (76.6%) and 184 (23.4%) were randomly assigned to two subcohorts for the development of nomograms and for the split-sample external validation. Achievement of RC was defined as wearing no pads or an occasional security pad. Time to RC and percentage of continent men at each follow-up were calculated using Kaplan-Meier curve estimates. The final multivariate model was selected based on the stepwise procedure and by considering the clinical importance of the variables in the model. The regression coefficient-based nomograms were developed based on final multivariate logistic regression models. RESULTS:
Age at surgery, membranous urethral length and robot-assisted RP were significant for RC at 1, 3 and 12 months. Saving the neurovascular bundle (NVB) and prostate volume were significant only for RC at 12 months. Odds ratios for age and membranous urethral length were constant over time, whereas the odds ratio for robot-assisted RP decreased over time (Fig. 1). The discrimination performance of the multivariate logistic models demonstrated 67, 68, 73% accuracy for each RC time point. Each nomogram was very well fitted to the ideal line of the calibration plot. External validation of the developed nomograms indicated 66, 67, 72% accuracy for each RC time point. (Fig. 1. Nomograms based on the 601 patients in the development cohort for predicting (A) immediate (≤1 month), (B) early (≤3 months), and (C) late (≤12 months after surgery) recovery of urinary continence after radical prostatectomy). CONCLUSIONS: We developed nomograms for immediate, early, and late RC after RP from a contemporary series of RP among Korean men, and validated these nomograms adequately. Our findings suggest that saving the NVB and prostate volume may only affect late RC after RP. In contrast, age at surgery, membranous urethral length and robotassisted RP seem to be consistently associated with immediate, early, and late RC after surgery.