Document Type


Publication Date



Maine Medical Center, Medical Education, Maine Medical Center Research Institute, Urology

MeSH Headings

Nephrostomy, Percutaneous, Risk Factors, Sepsis


Purpose/Background: Post-operative sepsis is a rare, potentially devastating, risk of percutaneous nephrolithotomy (PCNL). Knowledge of pre-operative risk factors may identify which patients might benefit most from new management strategies.

Methods/Approach: Retrospective chart review was performed on 153 consecutive patients who underwent PCNL at Maine Medical Center between October 2016 and December2018. Patient demographic factors, comorbidities, infection history, culture data, and stone factors were recorded. Post-operative sepsis was defined according to SIRS criteria for severe sepsis. Multivariate logistic regression was used to evaluate categorical variables as risk factors for sepsis. Fischer exact and student t tests were used to evaluate variables in patients with positive pre-op urine culture.

Results: 14 of 153 patients developed post-operative sepsis. Septic patients did not differ from others with regards age, gender, BMI, stone laterality, diabetes mellitus or renal function. Infected stone as an indication for PCNL was an independent predictor of sepsis (OR 5.66; p=0.015), as was large stone burden (a Seoul score >= 5 (OR 3.76; p=0.046) or S.T.O.N.E score >= 9 (OR 5.52; p=0.018)). Patients with limited mobility (upper or lower body) were much more likely to become septic (OR 21.4; p=0.002). Any positive pre-op culture was independently associated with sepsis (OR 13.7; p= 0.002),as were gram negative bacteria as a group (OR 5.7; p=0.025) and specifically the proteus species (OR 10.6; p=0.023). Such association was not found for gram positive bacteria. Among 45 patients with positive pre-op cultures, female gender (RR 3.9; p=0.047), infected stone as an indication (RR 3.1; p=0.047), limited lower (RR 3.5; p=0.022) and upper (RR 3.4; p=0.027) body mobility were all associated with post-op sepsis. Negative culture was protective against sepsis (OR 0.073; p= 0.002). Among patients with negative pre-op cultures, 2/109 (1.8%) had post-op sepsis, both with large complex stones (Seoul and S.T.O.N.E scores >=9).

Conclusions: Limited patient mobility and large stone burden are strong predictors of post PCNL sepsis. Positive pre-op urine culture, especially those with a gram-negative organism, also correlate with increased risk. Patients without these risk factors have a low chance of sepsis, even not accounting for intra-operative factors. Our results suggest a framework for risk-stratifying patients prior to surgery and the potential for more aggressive antimicrobial intervention in high risk patients, and less aggressive treatment in low risk patients.


2020 Costas T. Lambrew Research Retreat