RESULTS: A total of 55 (7.2%) patients were readmitted within 3 months of hospital discharge. Ureteral obstruction was the most common cause of PNL-related HR occurring in 35 (4.6%) patients fol- lowed by urinary leakage in 10 (1.3%) patients. Other causes were re- PNL, bleeding complications and urinoma in 6 (0.8%), 3 (0.4%) and 1 (0.1%) patients, respectively. The median (range) of hospital stay was 5 days (2-13). On multivariate analysis, the presence of residual frag- ments (Odds ratio [OR]: 3.2; 95% confidence interval [CI]: 1.6-6.7; p ¼ 0.002), the need for post-PNL auxiliary procedures (OR: 5.6; 95% CI: 2- 14.9; p ¼ 0.001), solitary kidney (OR: 2.5; 95% CI: 1.2-5.5; p ¼ 0.01), length of hospital stay (OR: 1.1; 95%CI: 1-1.2; p ¼ 0.01) and high baseline serum creatinine (OR: 1.3; 95%CI: 1.1-1.6; p ¼ 0.006) were independent predictors of PNL-related HR. CONCLUSIONS: PNL-related HR occurred in 7.2% of patients with ureteral obstruction was the most common cause. Independent predictors were residual fragments, the need of auxiliary procedures, solitary kidney, longer hospital stay and high preoperative serum creatinine. Source of Funding: none MP30-16 PROGNOSTIC FACTORS OF POSTOPERATIVE COMPLICATIONS IN PERCUTANEOUS NEPHROLITHOTOMY Fabio C. Vicentini*, Carlos H. Watanabe-Silva, Vinicius Meneguete, Rodrigo Perrela, Claudio B. Murta, Joaquim F. A. Claro, S~ ao Paulo, Brazil INTRODUCTION AND OBJECTIVES: Percutaneous neph- rolithotomy (PCNL) is the gold standard treatment for large kidney stones but there are concerns regarding the complication rates. Our aim is to evaluate the peri-operatory factors associated with complications after PCNL. METHODS: We performed a retrospective chart review of a prospective database of patients who underwent PCNL between July 2011 and July 2014. Preoperative and postoperative factors, such as age, body mass index (BMI), surgical risk, stone complexity (Guy’s Stone Score - GSS), access calyx, patient positioning, blood transfusion and surgical time were evaluated as predictors of complications. All complications were graded according to the Clavien Classification. The tests used in the univariate analyzes were chi-square and Fisher Exact test for categorical measures and the Student t test for continuous measures. For the multivariate analysis we used logistic regression. RESULTS: A total of 468 patients were included in the study and 517 PCNL were performed. The mean age was 48.312.8 years, mean BMI of 27.35.0 and 57.8% were female. Of these, 82% un- derwent to PCNL in supine position and 18% in the prone position. According to GGS, 24.8% were GGS1, 28.6% GSS2, 33.8% GSS3 and GSS4 12.8%. Most punctures were on lower and middle calyx (65.8% and 24.8% respectively) and 9.3% of the punctures were supracostal. Mean surgical time 114.652.78 min. The success rate (fragments 4mm) at POD 1 on CT scan was 55.1% The overall complication rate was 18.7%. According to the modified Clavien Classification System, 5.3% Clavien I, 6.7% Clavien II, 2.0% Clavien IIIa, 2.3% Clavien IIIB, 1.6% Clavien IVa, 0.4% Clavien IVB and 0.4% Clavien V. The transfusion rate was 6.0%. On univariate analysis, the number of accesses, supracostal punc- ture, upper calyx puncture and higher GSS were predictors of Clav- ien3 (major complications). On multivariate analysis, GSS and upper calyx puncture were the only predictors statistically significant for post-operative complications Clavien 3. GSS III and IV increased risk for high grade Clavien compared to GSS I and II (OR 2.52; CI95% 1.21-5.52) and also upper pole access compared to inferior pole access (OR 4.85; CI95% 1.46-13.92). Moreover, prone position (OR 2.53; CI95%1.0-6.08), number of punctures (OR 6.11 CI95%1.58-24.04) and high GSS (OR3.01 CI1.19- 8.6) were associated with blood transfusions on multivariate analysis. CONCLUSIONS: The number of punctures, supracostal ac- cess, high GSS and prone position were factors associated with major complications and blood transfusions after PCNL. Source of Funding: none MP30-17 NATURAL HISTORY OF RESIDUAL FRAGMENTS FOLLOWING PERCUTANEOUS NEPHROSTOLITHOTOMY Kyle Wood*, Marc Colaco, Juan Mainez, Ilya Gorbachinsky, Miguel Osorio, Eliud Sanchez, Majid Mirzazadeh, Winston-Salem, NC; Dean Assimos, Birmingham, AL; Jorge Gutierrez-Aceves, Winston- Salem, NC INTRODUCTION AND OBJECTIVES: Percutaneous Neph- rostolithotomy (PCNL) is the method of choice to treat large kidney stones. Large stone burdens require staged PCNL procedures to remove residual fragments. This study evaluates the natural history of residual fragments after PCNL. METHODS: Data was retrospectively collected from all patients whom underwent PCNL between January 2009 and January 2014. For patients that underwent a multiple stage procedures, data was collected from the final stage of their treatment. Demographic information regarding age, gender, BMI, medical comorbidity, and stone composition were collected. Computerized tomography (CT) scans were taken after the final stage. Each CT was analyzed as to remnant number, size, and location. All cases were then cross-referenced for other stone-related events (defined as symptomatic presentation resulting in the need for an additional stone treatment) occurring within 1 year of initial treatment. RESULTS: A total of 515 patients were included of which 420 patients had complete data. There was no significant difference in gender or rates of co-morbidities between groups. Patients with no remnant stone demonstrated a significantly lower reintervention rate than those in whom remnants were found (7.7% reintervention rate versus 28.9%, p<0.0001). These differences remained true when controlled for age, gender, and comorbidity status. For patients that did demonstrate rem- nants, substratification by number of stone remnants showed that pa- tients with multiple stone remnants had significantly more reintervention than patients with a single stone remnant regardless of total stone burden (36.8% reintervention versus 21.0%, p¼0.05). Single stone remnants found in the lower calyces demonstrated significantly lower reinterven- tion rates than those left in the middle or upper calyces (p<0.01). Rein- tervention rates were 16.75, 28.6%, 36.8%, 45.8% for lower, upper, pelvic/multiple, middle calyx respectively. A total of 300 patients had BMI data. Average BMI of all patients was 30.3. Patients with a higher BMI (>25) had more need for second procedures (22.8% versus 11.2%, p¼0.02). CaOx stones are less likely to require a second treatment than CaP (9.4% versus 22.2%, p<0.01). CaOx stones had significantly less residuals than CaP stones (42.5% versus 52.7%, p < 0.05). CONCLUSIONS: This study defined parameters seen on postsurgical CT scans as well as patient characteristics and stone type that can help in defining risk of future stone events and determine need for staged PCNL. Using these variables, the odds ratio of a future stone event can be calculated. Source of Funding: None MP30-18 COMPLIANCE WITH GUIDELINES FOR ANTIBIOTICS AFTER PERCUTANEOUS NEPHROLITHOTOMY DOES NOT INCREASE INFECTIOUS COMPLICATIONS: A RETROSPECTIVE COHORT STUDY Sameer Deshmukh*, Seth Bechis, Boston, MA; Kevan Sternberg, Benjamin King, Burlington, VT; Brian Eisner, Boston, MA INTRODUCTION AND OBJECTIVES: The American Urological Association guildelines committee recommends  24 hours of antibi- otics for percutaneous nephrolithotomy. The purpose of this study was Vol. 193, No. 4S, Supplement, Saturday, May 16, 2015 THE JOURNAL OF UROLOGY â e353