Objective A nomogram prediction model incorporating 3 clinical parameters was established and validated to accurately predict the likelihood of serofast state in HIV-negative patients with early syphilis following treatment. Methods A total of 603 patients with early syphilis and 431 patients were enrolled in a training cohort,while 172 patients were included in the validation cohort. The corresponding nomogram prediction model was drawn according to the logistic regression coefficients. Area under the curve (AUC) values were used to evaluate the discrimination. Calibration was evaluated with a calibration curve. The net benefit was performed via decision curve analysis (DCA). Results Multivariate logistic regression analysis showed that the three factors of age,RPR and sexual life time point were independent risk factors of serofast state. The nomogram showed a good performance as the AUCs of the nomogram were 0.816 in training and 0.726 in validation. Furthermore,decision curve analysis reveals a significant net benefit in utilizing this prediction model to estimate the probability of serofast state when the threshold probability ranges from 75％ to 95％. Conclusions This nomogram can provide individualized prediction for the serofast state in patients with early syphilis.
关键词
早期梅毒/血清固定/列线图/回顾性研究
Key words
early syphilis/serofast state/nomogram/retrospective study
维普
万方数据