Objectives To establish a machine learning logistic regression to predict the risk of severe hemorrhage after semi-tubeless percutaneous nephroscopy Methods The data of 465 patients (465 semi-tubeless PCNL procedures) were retrospectively analyzed at Shenyang Red Cross Hospital from January 1st,2016 to January 1st,2023.The following factors were analyzed: high pressure,stone score,expansion method,visualization,Guys grades,avascular area,diabetes,age,history of PCNL,duration,solid kidney,years,preoperative nephrostomy ,BMI and sex.We established a machine learning logistic regression model by using the data above and we collect 50 additional patients as the test group to verify the accuracy of the model. Results For all the 464 patients, 91 patients had postoperative haemorrhage and 373 had no haemorrhage.The average hemoglobin drop for all the procedures was 23 .5 ± 6.1g/L (range 20.1–38.1g/L), whereas the average hematocrit drop was 5.46 ± 4.08% (range 0.4–29%).We collect 5 demographic basic characteristics and 12 perioperative variables.There were three variables with an absolute value of the correlation coefficient less than 0.1.The lr.score of the model is 0.9448.Eighty-six percents of the predicted result were correct Conclusion As for the factors influencing the percutaneous nephroscopic hemorrhage,stone score,number of channels,BMI,age,duration,years,hypertension,solid kidney,guys score and diabetes are risk factors for haemorrhage. Visualization,avascular area,Preoperative nephrostomy are protective factors for haemorrhage

No abstract

p=0.004), organ/space infection (14.4% vs. 10.9%, p=0.026), re-operation (8.6% vs. 5.1%, p=0.003), DGE (26.8% vs 16.4%, p<0.001) and longer mean hospital length of stay (12.9 vs. 11.2 days, p=0.001) than those without FT. There was no difference in mortality (1.7% vs. 2.2%, p=0.488) or readmission rate (20.2% vs. 17.2%, p=0.099). In patients with DGE and POPF, FT placement was not associated with morbidity, mortality, length of stay or readmission rate (P>0.05). Conclusion: Surgeons place feeding tubes in patients with higher risk profile as a mitigation strategy. However, in a balanced cohort, patients with a feeding tube still had higher morbidity and longer hospital stay. These results do not support routine feeding tube placement at the time of PD; selective placement may be reasonable in high-risk patients.