Original Article
A novel scoring system for conversion and complication in laparoscopic liver resection
Abstract
Background: Although laparoscopic liver resection (LLR) has been increasingly popular worldwide, there is lack of predictive model to evaluate the feasibility and safety of LLR. The aim of this study was to establish a scoring system for predicting the possibility of conversion and complication, which could facilitate the patient selection for clinicians and communication with patients and their relatives during the informed consent process.
Methods: Consecutively 696 patients between August 1998 and December 2016 underwent LLR were recruited. The entire cohort was divided randomly into development and validation cohorts. The scoring system for conversion and complication were established according to risk factors identified from multiple logistic analysis. Subgroup analysis was performed to assess the clinical application. And the C-index and decision curve analysis (DCA) were conducted to evaluate the discrimination in comparison with other predictive models.
Results: Six hundred and ninety-six patients were enrolled eventually. The rate of conversion in the development and validation cohorts was 8.3% and 10.3%, respectively. Compared with 12.6% complication rate in the development cohort, 12.9% was concluded in the validation cohort. Upon on the identified risk factors, the risk stratification model was established and validated. Subsequent subgroup analysis indicated low risk patients presented superior surgical outcomes compared with high risk patients. Besides, the C-index and DCA implied our models had better capacities of predicting conversion and complication in comparison with previous scoring systems.
Conclusions: This novel scoring system presents the remarkable capacities of predicting conversion, complication in LLR. And thereby, it could be a useful instrument to facilitate the patient selection for clinicians and communication with patients and their relatives during the informed consent process.
Methods: Consecutively 696 patients between August 1998 and December 2016 underwent LLR were recruited. The entire cohort was divided randomly into development and validation cohorts. The scoring system for conversion and complication were established according to risk factors identified from multiple logistic analysis. Subgroup analysis was performed to assess the clinical application. And the C-index and decision curve analysis (DCA) were conducted to evaluate the discrimination in comparison with other predictive models.
Results: Six hundred and ninety-six patients were enrolled eventually. The rate of conversion in the development and validation cohorts was 8.3% and 10.3%, respectively. Compared with 12.6% complication rate in the development cohort, 12.9% was concluded in the validation cohort. Upon on the identified risk factors, the risk stratification model was established and validated. Subsequent subgroup analysis indicated low risk patients presented superior surgical outcomes compared with high risk patients. Besides, the C-index and DCA implied our models had better capacities of predicting conversion and complication in comparison with previous scoring systems.
Conclusions: This novel scoring system presents the remarkable capacities of predicting conversion, complication in LLR. And thereby, it could be a useful instrument to facilitate the patient selection for clinicians and communication with patients and their relatives during the informed consent process.
