Percutaneous nephrolithotomy: variables that influence hemorrhage.

OBJECTIVES Renal hemorrhage is one of the most common and worrisome complications of percutaneous nephrolithotomy (PCNL). This study attempted to identify variables that might influence hemorrhage during PCNL to help urologists establish preventative and treatment strategies for bleeding during PCNL procedures. METHODS The data of 193 patients (193 PCNL procedures) were retrospectively analyzed. Hemorrhage was estimated by the postoperative decrease in hematocrit factored by the quantity of any blood transfusion. Various preoperative and operative factors were assessed for their association with blood loss using univariate, forward multivariate regression and correlation analysis. RESULTS The mean patient age was 45.7 +/- 14.4 years (range 5 to 74). The overall stone-free rate was 85.4%. The average hematocrit decrease was 8.7% +/- 5.39% (range 0.3 to 24.7). Forward multivariate regression analysis identified five significant variables that influenced PCNL-related hemorrhage: stone type (P = 0.003), number of tracts (P = 0.010), method of dilation (P = 0.010), diabetes (P = 0.022), and stone surface area (P = 0.049). A statistically significant difference was found in relation to the occurrence of hemorrhage between patients with caliceal stones and partial staghorn stones (P = 0.008) and complete staghorn stones (P = 0.006), single tracts and multiple tracts (P = 0.038), balloon dilators and Amplatz dilators (P = 0.007), patients with small stones (1000 mm2 or smaller) and large stones (greater than 1000 mm2; P = 0.018) on univariate analysis. Also, the stone surface area (P = 0.019) and number of tracts (P = 0.024) showed a positive correlation with the mean hematocrit decrease. CONCLUSIONS Staghorn stones, multiple tracts, the presence of diabetes, and large stones were associated with increased renal hemorrhage during PCNL on multivariate analysis. However, balloon dilation was associated with decreased hemorrhage.