Impact Factors and an Efficient Nomogram for Predicting the Occurrence of Sepsis after Percutaneous Nephrolithotomy

Introduction To describe the clinical parameters of urinary stones and investigate the preoperative predictors of sepsis in patients following percutaneous nephrolithotomy (PCNL). Patients and Methods. A retrospective study of patients who underwent PCNL between August 2017 and December 2019 was performed. The patients were divided into the sepsis and nonsepsis groups according to whether they had sepsis, and their data were compared for further analysis. Results Fifteen (6.1%) patients matching in age, gender, body mass index (BMI), and the number of access variables were included in the sepsis group. The multivariate analysis demonstrated that the staghorn calculi (OR: 12.206, P < 0.001) and positive midstream urine culture (OR: 16.505, P < 0.001) were independent risk factors of sepsis, while preoperative renal fistula (OR: 0.122, P < 0.001) was a protective factor of sepsis. The three factors were ultimately selected to develop a nomogram to predict the probability of sepsis. The new nomogram was well calibrated and had higher diagnostic accuracy (the area under the curve: 0.916). Conclusions Our study reveals that patients with complex stones and positive bacteriuria are associated with a significantly high risk of sepsis after surgery. The removal of obstruction before operation under certain conditions might be a reliable protective factor of sepsis.


Introduction
Urinary stones are one of the most common diseases, with an incidence of approximately 5-15% around the world [1,2] and a recurrence rate reaching 50% in a decade [3,4]. Due to the advantages of high effectiveness in stone removal and quick recovery, percutaneous nephrolithotomy (PCNL) is now recognized as the first-line choice for the removal of renal stones, especially for complex stones with a high burden [5][6][7]. Nevertheless, PCNL can also induce severe adverse events occasionally.
Sepsis, a systemic inflammatory response syndrome (SIRS) clinically confirmed or highly suspicious by documented evidence of bacterial pathogens, is a notable postoperative complication after PCNL with an incidence of about 0.3% to 7.6% [8]. Based on the Sequential Organ Failure Assessment (SOFA) scoring system, sepsis is defined as life-threatening organ dysfunction caused by a dysregulated host response to infection [9]. It is the most common cause of perioperative death in PCNL in large series [7]. Some important risk factors of infec-tion after PCNL, including the stone size and operation time, have been well identified [8]. However, the evaluation of effect factors for sepsis or a few risk factors that only apply to struvite patients undergoing PCNL is rarely studied [10]. Therefore, in the present study, the related factors for urinary sepsis were evaluated, and a new nomogram was developed to predict the probability of sepsis for all post-PCNL patients.

Study Design and Data
Collection. This paper retrospectively reviewed the data of consecutive patients with the renal calculi or upper urethral calculi who underwent PCNL at the Affiliated Hospital of Jiangnan University from August 2017 to December 2019. All operations were performed in standard procedure by experienced urologists. Patients who had tumor, blood diseases, urinary tract tuberculosis, and juvenile and incomplete medical records were excluded. Patients' preoperative factors including age, sex, BMI, diabetes, surgery time, serum creatinine (SCr), midstream urine culture, stone location, the urinary sediment microscopy WBC (+ ∼ ++++), and staghorn calculi were recorded. Appropriate antibiotics would be administered as empirical therapy when patients had any focus of infection, and the use of sensitive antibiotics was guided by the positive urine culture report before operation. Sepsis was defined by a SOFA score of 2 or more consequent upon confirmed or suspected infection [9]. (Supplementary Table 1).

Percutaneous Nephrolithotomy Technique.
After induction of general anesthesia in the lithotomic position, a urethral catheter was placed, and the patient was moved to a prone position for the acquisition of percutaneous access under ultrasonographic guidance. Then, 22F-24F Amplatz working sheaths were inserted with the help of a guide wire through the track enlarged by balloon dilatation. After performing 17F nephros-copy and 1000 μm holmium laser fiber, or 22F nephroscopy and ultrasonic endoscopic lithotripter to maximize the dusting effect, the stone fragments were flushed out by irrigation, forceps, or graspers. At the end of the procedure, a 5F double-J stent and an 18F-20F nephrostomy tube were placed for drainage, which usually were left in the body for 2-4 weeks and 7 days, respectively.
2.3. Statistical Analysis. SPSS20.0 (SPSS, Chicago, IL, USA) was used for statistical analysis. Data were collected as mean plus or minus the standard deviation (SD) or the median and range. Comparisons of normally distributed continuous variables were performed using Student's t-test. Categorical variables were presented as numbers (percentages) and analyzed using the chi-square test or Fisher's exact test. The relative risk (odds ratio, OR) was used in a multivariable logistic regression by backward stepwise to determine risk factors for sepsis. Two-tailed P values <0.05 were considered statistically significant. A nomogram for sepsis was established based on the multivariate analysis results achieved by R software (Version 3.4.1; Institute for Statistics and Mathematics, Vienna, VIC, Austria) with the package rms.

Results
A total of 246 patients (184 males, 62 females) successfully treated with minimally invasive percutaneous nephrolithotomy were included in this retrospective study, with an average age of 51:22 ± 10:48 (range: 25 to 77 years). Fifty-five patients (22.4%) had diabetes mellitus. There were 127 right kidney stones, 84 left kidney stones and 35 bilateral kidney stones.    (Table 1). In the present study, 15 of 246 patients (6.1%) were found to be sepsis based on the sepsis-3 definition [9]. The patients were divided into the sepsis group and the nonsepsis group according to their complications after surgery. The mean ages of the sepsis and nonsepsis groups were 49:5 ± 10:20 and 52:47 ± 10:52 years, respectively, which showed no significant difference (P = 0:322). Besides, no statistically significant difference was found between the two groups in gender, stone location, BMI, the stone side,   BioMed Research International the stone number, serum creatinine, and surgery time (all P > 0:05). However, staghorn calculi, a larger stone size, positive midstream urine culture, diabetes mellitus, the urinary microscopy WBC (3+), and preoperative renal fistula were more frequently observed in patients with sepsis than in those without sepsis after surgery (all P < 0:05) ( Table 2).
The nomogram was depicted by three independent predictors described in the multivariate analysis ( Figure 1). The sum of points in the nomogram demonstrated the probability of sepsis (the bottom scale). The nomogram was internally validated by computing the bootstrap-corrected Harrell index and by the calibration plot. As illustrated in Figure 2, the model had moderate discriminative ability and was well calibrated. The area under the receiver-operating characteristic curve (AUC) of the new nomogram was 0.916 ( Figure 3).

Discussion
Percutaneous nephrolithotomy is a minimally invasive and effective treatment for kidney stones and usually the preferred therapy for complex upper urinary tract calculi [11]. However, this gold standard treatment still has the potential risk of causing complications, of which the incidence reaches 8-38% [12]. The stone-colonizing bacteria released during PCNL translocate through systemic circulation, contributing to the occurrence of postoperative infections and even sepsis [13]. In our research, the incidence   BioMed Research International of postoperative sepsis was 6.09%, and its outcome was similar to that found by previous studies [8].
In this retrospective study, the multivariate logistic regression analysis of patients' clinical data collected showed that staghorn calculi, preoperative positive midstream urine culture, and preoperative renal fistula were independently related to postoperative sepsis, whereas no significant association of sepsis with serum albumin and the urinary sediment microscopy WBC was observed. This result conflicted with that achieved by prior studies [10,14].
Aimed at removing stones, relieving obstruction, and protecting renal function, the PCNL surgery needed to reach the renal pelvic, renal calyx, or upper urinary tract, possibly resulting in an increased chance of infections [15,16]. Compared with small calculi, the large calculi caused relatively prolonged operation time and increased incidence of sepsis, as confirmed by our single factor analysis, but the results of the multivariate logistic regression analysis indicated that the stone location and stone size were not independent risk factors. This result was consistent with that obtained by the study of Lojanapiwat et al. [17].
Staghorn calculi are defined as the stones with multiple branches occupying a part or all of the renal pelvis and renal calyces [18]. 5 BioMed Research International demonstrated that the presence of a staghorn calculus remained independently associated with the increased risk of fever/SIRS/sepsis (OR: 3.14) [13]. Due to differences in the diagnostic criteria of sepsis and heterogeneity of the study population, some letters did not support the above conclusion [10,19] .Our multivariate analysis suggested that the staghorn calculi probably evolved into sepsis (OR: 12.206).
The preoperative midstream urine culture and intraoperative renal pelvic urine culture have been indicated as significant risk factors for post-PCNL sepsis [20]. Consistent with the result of previous studies, the positive urine culture was found to be associated with urosepsis and identified as an independent risk factor in the multivariate analysis of this study (OR: 16.505). Therefore, to avoid sepsis after PCNL patients with positive preoperative bacteriuria must be accurately treated by antibiotics sensitive for the detected pathogens before the surgical treatment.
Preoperative infections can be caused by kidney swelling (hydronephrosis) that may be associated with high renal pelvis pressure, especially in patients with serious hydronephrosis or basic kidney diseases. These patients are often suggested to remove obstruction before surgery, which may reduce the incidence of postoperative infections. As it was speculated, our results revealed that the preoperative renal fistula remained independently associated with sepsis. In addition, it was worth noting that preoperative renal fistula would be a protective factor of sepsis (OR: 0.122), which may be attributed to the removal of the bacteria culture medium. The protective role of preoperative renal fistula has not been reported in previous studies [10,14], but it only applies to specific patients.
To the best of our knowledge, nomograms have been widely used in clinical studies [21,22]. A recent study by Gao et al. [10] has investigated the predictors of sepsis among patients with struvite stones and performed the nomogram based on the calculated scores of some related factors (preoperative serum creatinine and multidrug resistance) to predict the probability of sepsis. However, this model included limited factors and was only for patients with struvite stones. Based on this model, we established a meaningful nomogram that contained two risk factors (staghorn calculi and preoperative positive midstream urine culture) and a protective factor (preoperative renal fistula) for the prediction of the probability of sepsis, and the area under the receiveroperating characteristic curve of the nomogram was calculated, which was 0.916, indicating high diagnostic accuracy.
This study has some distinct limitations. It was a retrospective design with a single institution. Besides, the sample size was small, which might be ascribed to the decreased number of more invasive PCNL surgery due to the more frequent application of flexible ureteroscope lithotripsy. What is more, our new nomogram was based on several ready predictors because some novel gene biomarkers and factors such as CaSR986 gene polymorphism and C-reactive proteins [23,24] were not included in this study due to health care restrictions in China.

Conclusion
In summary, the staghorn calculi and positive preoperative midstream urine culture are two risk factors of sepsis after percutaneous nephrolithotomy, and preoperative renal fistula is a protective factor of sepsis. The results may provide new insights for us into the prevention of postoperative complications. Furthermore, the nomogram developed in the current study is a moderately accurate tool that can predict the probability of sepsis for patients after PCNL surgery.

Data Availability
The data are available on request.

Ethical Approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This retrospective study formal consent was not required.