1. Background
Percutaneous nephrolithotomy (PNL) is an effective, safe, and minimally invasive treatment method with low complications for renal and ureteral calculi refractory to extracorporeal shock wave lithotripsy therapy (ESWL) and in patients with anatomic variations (1-4). In fact, this method has been known as the gold standard of care for treating renal stone diseases (5). Despite the success rate of over 90%, there are inherent complications such as adjacent organs, sepsis, fistulas, stenosis of excretory way, fluid overload, serum electrolytes imbalance, hypothermia, irreversible renal lesion, and even death (6, 7). Although various pharmacological interventions exist, significant blood loss during or after the surgery remains a major concern for both surgeons and anesthesiologists (8), which may lead to various unwanted complications, including hemodynamic instability, need for transfusion and embolization, inability to safely remove stones, prolonged hospital stay, and rarely death (9).
Studies have shown a drop in hemoglobin (Hb) levels of 2.1 - 3.3 g/dL in cases undergoing PLN (10, 11) and the need for transfusion in 1% to 34% of them (12). Bleeding following this surgery is almost always controlled by conservative intervention, although selective embolization and arteriography are required in about 0.8% of cases (13-15). The available studies' results are inconsistent, and there is no agreement on the issue (16). Some have demonstrated that hypertension, site of puncture, and duration of surgery significantly affect the reduction of Hb during PCNL, while operative position (supine/prone), the number of punctures, and tract dilatation size have no significant effect (17). In other studies, stone size, the number of tracts, the size of Amplatz sheath, and the number of stones were reported as influencing factors (18).
In contrast, Ramon de Fata et al. found that only multiple percutaneous tracts (≥ 2) and middle calyx puncture were associated with lower blood loss and did not confirm the others reported by previous studies (4). This study was planned due to the mentioned discrepancies. Certainty, a preoperative workup is crucial to restrict the risk factors for bleeding associated with this surgery (8).
2. Objectives
This retrospective study aimed to identify the risk factors of intraoperative bleeding in patients undergoing percutaneous nephrolithotomy.
3. Methods
First, the study protocol was approved by the Ethics Committee of the Guilan University of Medical Sciences (GUMS). This retrospective study was conducted at the urology department of Razi hospital, an academic and referral center. The files of patients who underwent PNL were sorted out and screened for eligibility.
Inclusion criteria: Patients with urinary calculi staghorn type, aged 18 - 85 years, who had undergone elective PNL in the prone position were included.
Exclusion criteria: Files with incomplete data, cases with long-term use of anticoagulant agents or coagulopathy disorders like hemophilia, and emergency cases were excluded.
The responsible medical student completed a checklist including patient demographics and surgical outcomes. The eligible patients had intravenous urography, preoperative anesthesia visits as elective surgery, ultrasonography, and Computed Tomography (CT). The operation was performed under general anesthesia in the prone position. The same access site for the operation was determined based on intravenous urography and CT. Urinary ultrasound evaluated the degree of hydronephrosis, and CT measured the stone size. Urine culture and analysis were used to detect Urinary Tract Infection (UTI) that was treated before surgery. All the operations were performed by a single surgeon with 10 years of experience.
3.1. Statistical Analysis
The data analysis was performed by SPSS 16.0 (SPSS, Chicago, IL, USA). Data are presented as the mean and standard deviation (SD) or the median and range. The Chi-square test was applied for univariate analysis. Finally, multivariate logistic regression was performed. Spearman's rho correlation coefficient test was also used. A P-value less than 0.05 was considered significant.
4. Results
A total of 180 files were examined, and finally, the data of 151 patients were analyzed with a mean age of 47.89 ± 12.41 years (range 16 - 81). Of them, 75% were above 40. Patients' demographics, medical history, and surgery characteristics are presented in Table 1. The mean drop in Hb levels was 1.92 ± 1.56 mg/dL, and 24.8% required transfusions. None of them underwent nephrectomy to control bleeding. Only one patient needed postoperative angiography and embolization. Surgery outcomes are presented in Table 2. In this study, we found that all cases had at least a 1 mg/dL drop in Hb levels and the highest drop was 3 mg/dL (Table 2). By using Spearman's rho correlation coefficient, borderline significant values were observed in terms of stone bulk and Hb drop (P < 0.069), but not regarding other variables, including age, BMI, GFR, surgery duration, and the number of tracts (P > 0.05). A significant association was observed between UTI (P = 0.01) and transfusion (P = 0.0001) and Hb levels drop. However, no meaningful association was observed regarding other variables (P > 0.05) (Table 3). According to logistic regression, the history of open kidney surgery (P = 0.031), nephrostomy insertion (P = 0.003), and ESWL (P = 0.041) were correlated with the increased risk of Hb drop (Table 4).
Table 1.Patients' Baseline Demographics and Surgery Characteristics
| Variables | Values a |
|---|---|
| Mean body mass index (kg/m2) | 28.0904 ± 4.69695 |
| Mean age (y) | 47.89 ± 12.41 |
| Gender | |
| Male | 78 (51.7) |
| Female | 73 (48.3) |
| History of hypertension | |
| No | 103 (68.2) |
| Yes | 48 (31.8) |
| History of diabetes mellitus | |
| No | 119 (78.8) |
| Yes | 32 (21.2) |
| History of active urinary tract infection | |
| No | 141 (93.37) |
| Yes | 10 (6.623) |
| Kidney involved | |
| Right | 71 (47.02) |
| Left | 80 (52.98) |
| Multifocal stone | |
| One place | 47 (30.7) |
| More than one place | 104 (69.3) |
| Renal secretion in pyelography | |
| Normal | 128 (84.1) |
| Decreased | 23 (15.9) |
| Thickness of renal parenchyma | |
| Increased | 2 (1.32) |
| Normal | 135 (89.41) |
| Decreased | 14 (9.27) |
| Previous ESWL | |
| No | 85 (56.3) |
| Yes | 66 (43.7) |
| Previous open surgery | |
| No | 129 (85.4) |
| Yes | 22 (14.6) |
| History of kidney involvement PNL | |
| No | 136 (90.1) |
| Yes | 15 (9.9) |
| Degree of hydronephrosis | |
| No or mild | 90 (59.60) |
| Moderate | 48 (31.79) |
| Severe | 13 (8.61) |
| Mean duration of operation (min) | 48.53 ± 22.375 |
| Calyx undergoing puncture | |
| Upper | 14 (9.27) |
| Middle | 17 (11.26) |
| Lower | 120 (79.47) |
| Number of tracts | |
| 1 | 142 (94.03) |
| 2 | 9 (5.7) |
| Nephrostomy implantation | |
| No | 75 (49.67) |
| Yes | 76 (50.33) |
| Transfusion | |
| No | 114 (75.49) |
| Yes | 37 (24.51) |
a Values are expressed as No. (%) or mean ± SD.
Table 2.Surgery Outcomes
| Outcome | Value |
|---|---|
| Mean duration of hospitalization (days) | 3.25 ± 1.46 |
| Mean drop in hemoglobin (g/dL) | 1.92 ± 1.56 |
| Mortality | 0 |
| Need for angiography and embolization to control postoperative bleeding | 1 |
| Need for laparotomy due to postoperative colon perforation | 0 |
| Postoperative urosepsis after the operation | 1 |
| Postoperative pulmonary complications | 1 |
| Perforation of the collecting urinary tract after surgery | 0 |
| Postoperative urethra-pelvic junction injury | 1 |
| Postoperative large hematoma | 1 |
| Need for nephrectomy to control postoperative bleeding | 0 |
| Need to re-PNL | 2 |
Table 3.Correlation Between Hemoglobin Drop and Quantitative Variables
| Quantitative Variables | Age (y) | BMI | GFR Before Operation (Kidney Function) | Bulk Rock | Duration of Operation (Min) | Number of Tracts |
|---|---|---|---|---|---|---|
| Decrease in hemoglobin (g/dL) | ||||||
| Spearman's rho correlation coefficient | -0.061 | -1.071 | 0.123 | 0.149 | 0.136 | 0.008 |
| Significance level | 0.457 | 0.392 | 0.139 | 0.069 | 0.101 | 0.925 |
| Number | 149 | 148 | 146 | 150 | 146 | 151 |
Table 4.Multivariate Binary Logistic Regression Analysis for Factors Affecting Hemoglobin Drop (g/dL)
| Variables | B Regression Coefficient | S.E. Standard Deviation | Significance Level | Odds Ratio | Confidence Interval | |
|---|---|---|---|---|---|---|
| Lower Limit | Upper Limit | |||||
| Previous open kidney surgery | 1.969 | 0.911 | 0.031 | 7.162 | 1.201 | 42.698 |
| Step 4 | ||||||
| Nephrostomy implantation | 1.898 | 0.632 | 0.003 | 6.676 | 1.935 | 23.038 |
| Previous ESWL | 1.162 | 0.569 | 0.041 | 3.198 | 1.048 | 9.759 |
| Fixed value (effect of unknown factors) | -2.033 | 0.646 | 0.002 | 0.131 | ||
5. Discussion
In this study, we found that at least a 1 mg/dL drop in Hb levels was observed in all cases, and the highest drop was 3 mg/dL. A significant association was observed between UTI and transfusion and Hb drop. The history of open kidney surgery, nephrostomy insertion, and ESWL were correlated with the increased risk of Hb dropping. A similar study conducted by Said supports our results (12). Li et al., in a meta-analysis, found that hypertension, diabetes, stone type, and multiple tracts were significantly associated with severe bleeding and the need for embolization after PLN. They strongly recommended that the exact correlations must be evaluated by further studies (19).
Percutaneous nephrolithotomy is the standard and successful treatment for renal stones; however, complications occur in approximately 15% of cases. It is estimated that 9.4% develop intraoperative bleeding with different predisposing factors (10, 11). Akman et al. reported that diabetes was a significant risk factor for bleeding in these patients. They found it to be expected because diabetes leads to atherosclerosis and microangiopathy (13). Kukreja et al. found a positive relationship between hypertension and intraoperative bleeding, which was explained by arteriosclerosis (20).
In contrast, Meng et al. found no association between diabetes and hypertension, and bleeding during PNL (21). A few studies have indicated that females were more prone to bleeding; however, it was not supported by other studies (22). The other risk factor was the stone size. Larger stones were associated with prolonged operation time and, as a result, increased bleeding (21, 23). Kallidonis et al. found that the type of stone was the other risk factor (24), which was supported by two other studies (22, 25). Studies have shown that patients with renal stones with nil or mild hydronephrosis are at higher risk of severe bleeding due to the thickness of the renal cortex. The distribution of kidney vessels is scattered in higher degrees of hydronephrosis. Therefore, the risk of vessel injury during surgery will be lower (26). Kim et al. found that correct renal puncture was a significant risk factor for severe hemorrhage requiring angioembolization (27). Ullah et al. reported that operating time, female gender, and stone size were the main risk factors (28).
As mentioned above, studies have reported different risk factors for bleeding in PLN, and their findings are often inconsistent, which can be explained by differences in instruments, surgeon's experience, and surgical technique, such as the need for multiple or large access tracts. In addition, measurement tools, bleeding estimation criteria, the accuracy of laboratories, and studied populations are not the same among studies. Also, the inclusion criteria of the studies as an influencing factor differ in terms of patients' characteristics, comorbidities, age, and preoperative status (12, 17, 18, 21, 27, 29).
5.1. Limitations
The nature of a retrospective study is different from prospective studies. Selection bias occurs in retrospective studies, and the reliability of the results is disturbed by incomplete files that need to be excluded. Furthermore, the investigated items should be restricted to those recorded in patients' medical files. A single-center study could be another limitation of this survey.
5.2. Conclusions
According to the findings of this study, PLN under general anesthesia in a prone position was performed successfully and safely in this center. Also, UTI, history of open surgery, nephrostomy implantation, and ESWL were the significant risk factors for the severity of bleeding in this procedure.
