Development of a prediction model to estimate short- and long-term clinical outcomes in arteriovenous grafts for hemodialysis patients
摘要
The prognosis and quality of life for patients undergoing hemodialysis (HD) are significantly compromised by dysfunctions in arteriovenous grafts (AVG). This study aimed to develop and validate a risk prediction model for factors influencing both short- and long-term primary patency rates in HD patients with AVG.
MethodsThe single-center retrospective study comprised 354 HD patients who had undergone AVG creation. Using Kaplan–Meier survival analysis and Cox regression models, we evaluated the patency rates of AVG and identified associated independent risk factors based on retrospectively collected clinical data.
ResultsThe primary patency rates of the patients with AVG at 6 months, 1 year, and 2 years were 82.49%, 50%, and 16.95%, respectively. The multivariate Cox regression analysis revealed that the brachial arterial blood flow and draining venous diameter at 1 month post-operatively were independent predictors of newly created AVG patency at 6 months. Furthermore, at 1 month post-operatively, the brachial artery blood flow, draining vein diameter, and history of central venous catheter (HR 0.480 (0.281, 0.820), p = 0.007) were independent risk predictors for failure of newly created AVG at 12 months.
ConclusionThis study revealed that these parameters, including brachial arterial blood flow, draining vein diameter, and history of central venous catheter, could be applied to predict short- and long-term risks for AVG patency loss.