Construction of a coagulation prediction model of the extracorporeal circulation circuit during hemodialysis with regional citrate anticoagulant (RCA).

OBJECTIVE

To construct a prediction model of coagulation in the extracorporeal circulation circuit during hemodialysis with regional citrate anticoagulant(RCA) conditions.

METHODS

This was a single-center, retrospective study. The clinical data of patients who received hemodialysis with RCA from February 2021 to March 2022 were collected. The risk predictors of coagulation in the extracorporeal circulation circuit were screened by LASSO regression. On this basis, we used multivariate logistic regression analysis to establish a nomogram prediction model.

RESULTS

A total of 98 patients received RCA hemodialysis for 362 times. Among them, 155 treatments with complete data were included in the study. Among the 155 treatments, coagulation of the extracorporeal circulation circuit occurred 12 times. The use of arteriovenous fistulas(AVF), the venous pressure at 4 h after hemodialysis initiation, blood flow velocity, dialyzer manufacturer, Systemic iCa2+ at 1 h after hemodialysis initiation, plasma albumin level, and plasma d-dimer level were influencing factors of coagulation in the extracorporeal circuit during hemodialysis with RCA (p < 0.05). A nomogram model was made out of the above indicators. The area under the receiver operating characteristic (ROC) curve for predicting coagulation in the circuit was 0.967 (95% CI: 0.935-0.998). The internal validation result of the memory testing (bootstrap method) showed that the area under the ROC curve was 0.967 (95% CI: 0.918-0.991).

CONCLUSION

The nomogram model has good discrimination and calibration and can intuitively and succinctly predict the risk of coagulation in the extracorporeal circulation circuit during hemodialysis with RCA.