Abstract: The emergence of totally implantable venous access ports (TIVAP) greatly alleviates the pain caused by venipuncture and maximizes the quality of life of patients. Catheter-related bloodstream infection is one of the most severe complications of TIVAP, affecting the treatment and prognosis of patients. A systematic review was conducted for nomogram model for risk of TIVAP-related bloodstream infection in gastrointestinal tumor patients, logistic model for prediction of TIVAP-related bloodstream infection in breast cancer patients, machine learning model for central venous transfusion device-related bloodstream infection and data mining model for central venous transfusion device-related bloodstream infection, the study methods for risk prediction models and risk factors for TIVAP-related bloodstream infection in China and abroad were observed and compared, it comes to a conclusion that a favorable risk prediction model for TIVAP-related bloodstream infection should have four basic conditions: clear applicable population, standard model parameters, strict internal and external verification, and practical model form so as to provide experience for building up the risk prediction model for TIVAP-related bloodstream infection in China.