The association between embedded catheter implantation and hospitalization costs for peritoneal dialysis initiation: a retrospective cohort study.

BACKGROUND: Compared with the conventional peritoneal dialysis (PD) catheter insertion, embedding PD catheter implantation is one of the procedures for planned PD initiation. However, facilities where embedded PD catheter implantation is available are limited, and the impact of embedded PD catheter implantation on hospitalization cost and length of hospitalization is unknown. METHODS: This retrospective single-center cohort study included 132 patients with PD initiation between 2005 and 2020. The patients were divided into two groups: 64 patients in the embedding group and 68 patients in the conventional insertion group. We created a multivariable generalized linear model (GLM) with the gamma family and log-link function to evaluate the association among catheter embedding, the duration and medical costs of hospitalization for PD initiation. We also evaluated the effect modification between age and catheter embedding. RESULTS: Catheter embedding (ß coefficient - 0.13 [95% confidence interval - 0.21, - 0.05]) and age (per 10 years 0.08 [0.03, 0.14]) were significantly associated with hospitalization costs. Catheter embedding (- 0.21 [- 0.32, - 0.10]) and age (0.11 [0.03, 0.19]) were also identified as factors significantly associated with length of hospitalization. The difference between the embedding group and the conventional insertion group in hospitalization costs for PD initiation (P for interaction = 0.060) and the length of hospitalization (P for interaction = 0.027) was larger in young-to-middle-aged patients than in elderly patients. CONCLUSIONS: Catheter embedding was associated with lower hospitalization cost and shorter length of hospitalization for PD initiation than conventional PD catheter insertion, especially in young-to-middle-aged patients.

Gradient boosting decision tree becomes more reliable than logistic regression in predicting probability for diabetes with big data.

We sought to verify the reliability of machine learning (ML) in developing diabetes prediction models by utilizing big data. To this end, we compared the reliability of gradient boosting decision tree (GBDT) and logistic regression (LR) models using data obtained from the Kokuho-database of the Osaka prefecture, Japan. To develop the models, we focused on 16 predictors from health checkup data from April 2013 to December 2014. A total of 277,651 eligible participants were studied. The prediction models were developed using a light gradient boosting machine (LightGBM), which is an effective GBDT implementation algorithm, and LR. Their reliabilities were measured based on expected calibration error (ECE), negative log-likelihood (Logloss), and reliability diagrams. Similarly, their classification accuracies were measured in the area under the curve (AUC). We further analyzed their reliabilities while changing the sample size for training. Among the 277,651 participants, 15,900 (7978 males and 7922 females) were newly diagnosed with diabetes within 3 years. LightGBM (LR) achieved an ECE of 0.0018 ± 0.00033 (0.0048 ± 0.00058), a Logloss of 0.167 ± 0.00062 (0.172 ± 0.00090), and an AUC of 0.844 ± 0.0025 (0.826 ± 0.0035). From sample size analysis, the reliability of LightGBM became higher than LR when the sample size increased more than [Formula: see text]. Thus, we confirmed that GBDT provides a more reliable model than that of LR in the development of diabetes prediction models using big data. ML could potentially produce a highly reliable diabetes prediction model, a helpful tool for improving lifestyle and preventing diabetes.