The high-risk factors of urinary infectious stones: a retrospective study

Abstract

Background

Urinary stones are a common condition with increasing prevalence worldwide. Predicting the type of urinary stones is essential for guiding treatment, yet complex imaging models are not always accessible. This study aims to identify simpler clinical predictors and explore the risk factors for infected urinary stones using statistical and receiver operating characteristic (ROC) analyses.

Methods

A retrospective analysis was conducted on 1,067 patients with urinary stones who underwent surgical treatment between 2018 and 2023. Patients were classified into infected and non-infected stone groups based on stone composition. Logistic regression analysis adjusted for age, gender, body mass index (BMI), and serum electrolytes was performed to identify significant predictors, with sensitivity analyses using inverse probability weighting (IPW) to address cohort imbalance. The predictive performance of key factors was assessed using ROC curves.

Results

Of the total cohort, 686 (64.3%) had infectious stones, and 381 (35.7%) had non-infectious stones. Infectious stones were more common in females, younger patients, and those with ureteral stones. Preoperative urine cultures revealed Proteus mirabilis (24.3%) and Escherichia coli (18.2%) as predominant pathogens. Multivariate analysis identified three independent predictors for infectious stones: alkaline urine pH [hazard ratio (HR) 2.54, 95% confidence interval (CI): 1.33–4.88, P<0.001], ureteral stone location (HR 5.60, 95% CI: 2.38–13.17, P<0.001), and absence of diabetes mellitus (HR 4.74, 95% CI: 1.50–15.03, P=0.01). Sensitivity analyses confirmed robustness (adjusted HRs: 2.49, 5.58 and 4.65, respectively). Among these, ureteral stone location had the best predictive performance [area under the curve (AUC) =0.782, sensitivity 84.0%, specificity 71.5%], followed by urine pH (AUC =0.766, sensitivity 79.3%, specificity 68.7%), while diabetes status showed weaker predictive ability (AUC =0.623).

Conclusions

Alkaline urine pH and ureteral stone location are strong predictors of infected urinary stones, while diabetes status is less predictive. These findings highlight the importance of integrating simple clinical parameters to improve the preoperative assessment and management of patients with urinary stones, particularly in resource-limited settings.

Highlight box.

Key findings

• This study analyzed 1,067 patients with urinary stones to identify predictors of infectious stones. Infectious stones (64.3%) were more common in females, younger patients, and those with ureteral stones. Multivariate analysis revealed three independent predictors: alkaline urine pH, ureteral stone location, and absence of diabetes mellitus. Ureteral stone location had the highest predictive performance, followed by urine pH, while diabetes status was less predictive.

What is known and what is new?

• Urinary stones are a prevalent global condition, and predicting stone type is crucial for treatment guidance. However, complex imaging models for prediction are often inaccessible, especially in resource-limited settings. Infected urinary stones are associated with specific risk factors, but simpler clinical predictors for identifying them remain underexplored.

• This study identifies alkaline urine pH, ureteral stone location, and absence of diabetes mellitus as independent predictors of infected urinary stones. Ureteral stone location [area under the curve (AUC) =0.782] and alkaline urine pH (AUC =0.766) demonstrated strong predictive performance, while diabetes status was less predictive (AUC =0.623). These findings emphasize the utility of simple, accessible clinical parameters for preoperative assessment, offering a practical approach to managing urinary stones, particularly in settings with limited resources.

What is the implication, and what should change now?

• Alkaline urine pH and ureteral stone location are robust predictors of infectious stones, especially valuable where advanced diagnostics are unavailable. Clinicians should integrate these parameters for improved preoperative risk stratification, optimized treatment planning, and reduced infectious complications in resource-limited settings.

Introduction

Urinary tract stones affect approximately 5–10% of the general population worldwide, with increasing incidence, particularly in regions with warmer climates. The global incidence has been on the rise due to factors such as diet, lifestyle changes, and metabolic conditions. Men have a higher lifetime risk, ranging from 10–15%, compared to 5–7% in women (1,2). In developed countries, there is an observed peak incidence around the age of 30–50 years, with a subsequent rise in older populations as well. Urinary tract stones are often linked with conditions like dehydration, hypercalciuria, and metabolic syndromes, with certain geographic areas exhibiting higher prevalence due to environmental and dietary factors (3). The choice of treatment and prevention program for urinary stones is determined by the location, size, composition, presence or absence of symptoms, the patient’s wishes, and the clinician’s ability; therefore, the choice of treatment and prevention program for treating different components of urinary stones in the body may also vary (4). Based on current research, we have found that knowledge of the composition of stones not only helps us to choose the appropriate treatment program to treat stones, but also helps us to choose the appropriate medication or lifestyle to prevent stones from recurring (5,6). Urinary stones can be categorized into four main groups according to the cause of formation, including: non-infectious stones, infectious stones, genetic cause infected stones, and medicinal causes of stones. Among them, non-infectious stones include: calcium oxalate stones, calcium hydrogen phosphate dihydrate, uric acid stones; infectious stones include: magnesium ammonium phosphate stones, carbonate apatite, ammonium urate stones (7). Infectious stones carry a higher surgical risk than other types of stones, and severe urogenital sepsis may be induced during surgery, leading to shock (8). In addition to this, infectious stones have a higher rate of recurrence and would lead to the loss of kidney function than other types of stones (9). The search for effective treatment and prevention of infectious stones has become the focus of our research.

There have been many studies on the risk factors for urinary stones, and these studies generally agree that obesity, diabetes, dietary and metabolic problems would lead to an increased risk of urinary stone development (10,11). For urinary stones, it has been suggested that gender, urine culture results, and urine turbidity can indicate the infection status of the stones (12). One study tried to predict stone composition from urine composition (1). There have also been attempts to predict infected versus non-infectious stones preoperatively from imaging data as a way to help clinicians improve their treatment plans (13). Kazemi et al. constructed an artificial neural network (ANN) model based on clinical data of uric acid, blood calcium, history of hypertension, and history of diabetes mellitus from 936 patients with renal stones predicted stone composition with 97.1% accuracy (14). An ANN model for predicting stone composition based on dual-energy computed tomography (CT) images of in vitro stone specimens had an accuracy of nearly 90% in predicting stone composition (15). Although the predictive efficacy of these imaging models is significant, not all patients with stones undergo CT, which would limit the use of the above models in clinical practice. Therefore, we need simpler predictive methods to evaluate stone types. The aim of this study is to explore the risk factors for infected urinary stones. This study collected clinical data from a certain number of patients with urinary stones. These data were analyzed by statistical methods to screen the risk factors associated with the risk of infectious urinary stones. By using statistical and receiver operating characteristic (ROC) analysis to analyze a large number of patients, this study aims to bridge the gap between complex imaging models and simple, widely available clinical evaluations, ultimately improving preoperative management of urinary tract stones. We present this article in accordance with the STROBE reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-244/rc).

Methods

Study design and participants

This study included patients with urinary stones in our center from January 2018 to January 2023. The patients who met the following criteria were included in the study: (I) diagnosed with urinary stones; (II) underwent urinary stone extraction surgery. Subjects meeting any of the following criteria were excluded: (I) missing identity information; (II) incomplete preoperative clinical data; (III) missing results of stone composition analysis; (IV) patients with multiple stone sites were excluded to minimize confounding from heterogeneous stone etiologies (e.g., mixed metabolic and infectious components). The sample size was determined by all consecutive patients meeting the inclusion criteria during the study period (2018–2023). Post hoc power analysis was performed using G*Power 3.1.9.7 with the following parameters: effect size (Cohen’s f²=0.25), α=0.05, and sample size =1,067. The achieved power exceeded 99% for all primary predictors. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Xiangya Hospital Zhuzhou Central South University (No. 2023-08116) and individual consent for this retrospective analysis was waived.

Detection methods

Urine samples were analyzed within 30 minutes of collection to prevent bacterial overgrowth. Urine pH was determined using a colorimetric assay (URiT-11 Automated Urine Analyzer, Dingsheng Group, Jinan, China). Urine specimens were collected via spontaneous urination within 24 hours preoperatively and analyzed immediately to minimize bacterial overgrowth. Urine cultures were performed using standard clinical protocols: midstream urine samples were inoculated onto blood agar and MacConkey agar plates (BioMérieux, France) and incubated at 37 ℃ for 24–48 hours. Bacterial identification was conducted via MALDI-TOF mass spectrometry (Bruker Daltonics, Germany) or biochemical testing (VITEK 2 Compact System, BioMérieux), with bacterial identification based on colony morphology and biochemical testing. After removing the stone lesions from patients with stones, specimens were made and Automatic Infrared Spectrum Analysis System (SUN-3G, Dingsheng Group) was used to detect the composition of the stones. Subsequently, we divided the samples into two groups based on stone composition, i.e., the infected stone group and the non-infected stone group. Stone compositions were classified as infectious stones if they contained magnesium ammonium phosphate, carbonate apatite, or ammonium urate, following standardized infrared spectral criteria (Table S1). Reference spectra were validated against the Clinical and Laboratory Standards Institute (CLSI) guidelines (16). Infectious stones were defined exclusively by composition (magnesium ammonium phosphate, carbonate apatite, or ammonium urate), regardless of preoperative urinary tract infection (UTI) status. Non-infectious stones (e.g., calcium oxalate, uric acid) were classified separately even if associated with urinary infections. The data were analyzed statistically.

Data collection

The following information was collected for each patient: name, gender, age, body mass index (BMI), history of previous diseases, urine pH, urinary occult blood, urinary protein, urinary leukocytes, urinary nitrites, urinary microscopic leukocyte counts, urinary bacterial culture results, blood electrolyte levels (blood sodium, blood potassium, blood calcium, blood phosphorus), serum creatinine, serum uric acid, serum cystatin C, serum cholesterol, serum triglycerides, serum high-density lipoprotein (HDL), serum low-density lipoprotein (LDL), stone composition. Urine specimens were collected via spontaneous urination to ensure natural composition. Urine pH was measured using first-morning void samples collected within 24 hours prior to surgery. Alkaline urine pH was defined as pH >7.0 based on clinical standards outlined in the European Association of Urology (EAU) guidelines (17).

Statistical analysis

Data analysis was conducted using the SPSS 29.0 statistical software (SPSS Inc., Chicago, IL, USA). Categorical variables were expressed as [n (%)] and the Chi-squared test was applied using the basic formula when the sample size was ≥40 and the theoretical frequency T was ≥5. When the sample size was ≥40 but the theoretical frequency was 1 ≤ T <5, the Chi-squared test was conducted using the corrected formula. For sample sizes <40 or theoretical frequencies T <1, statistical analysis was performed using Fisher’s exact probability method. The Shapiro-Wilk test was used to check the normal distribution of continuous variables. Normally distributed continuous variables were expressed as mean ± standard deviation (SD) and analyzed using the t-test with corrected variance. Non-normally distributed continuous variables were expressed in the form of median (25th percentile, 75th percentile) and analyzed using the Wilcoxon rank-sum test. After that, logistic regression analysis was used to calculate the hazard ratio (HR) of all factors with infectious stones as the dependent variable and influencing factors as the independent variables. The influencing factors related to infectious stones were screened on the basis of univariate logistic regression analysis for multivariate logistic regression analysis. Multivariate logistic regression adjusted for age, gender, BMI, serum sodium, potassium, and uric acid—variables selected based on clinical relevance and univariate significance (P<0.1). To address potential bias from cohort imbalance, sensitivity analyses using inverse probability weighting (IPW) were performed. IPW adjusted for age, gender, BMI, and serum electrolytes to balance the distribution of covariates between groups. Stability of HR was confirmed if the adjusted HRs remained statistically significant (P<0.05) and their 95% confidence intervals (CI) overlapped with the original estimates. To evaluate the predictive ability of independent influencing factors for infectious stones, ROC curves were plotted, and the area under the curve (AUC) was calculated. The AUC values were used to assess the discrimination ability of the predictive models, where a larger AUC indicated better predictive performance. Sensitivity and specificity were also derived from the ROC curves to evaluate the accuracy of each factor. A two-tailed P value <0.05 was considered statistically significant.

Results

Baseline characteristics

A total of 1,067 patients were included, including a total of 686 infectious stones and 381 non-infectious stones. The clinical characteristics of the two groups of patients are summarized in Table 1. There are significant differences between the two groups in multiple baseline features. In the infectious stone group, the proportion of women was higher, the age was lower, the urine pH value was significantly higher, the proportion of ureteral stones was significantly higher than that in the non-infectious stone group, while the proportion of diabetes in the non-infectious stone group was higher, and the levels of blood uric acid and blood potassium are also significantly higher. In addition, the blood phosphorus and sodium levels in the infectious stone group were higher than those in the non-infectious stone group (P<0.05) (Figure 1). Other indicators such as BMI, urine occult blood, urine protein, urine nitrate, urine white blood cells, urine bacterial culture, blood calcium, blood creatinine, blood cystatin C, blood cholesterol, blood triglycerides, HDL, and LDL levels showed no statistically significant differences between groups (P>0.05).

Table 1. Baseline characteristics of patients.

Variable	Overall (n=1,067)	Infectious stones cohort (n=686)	Non-infectious stones cohort (n=381)	P
Gender
Male	764 (71.60)	429 (62.54)	335 (87.93)	<0.001*
Female	303 (28.40)	257 (37.46)	46 (12.07)
Age (years)	54.99 (51.98–58.00)	51.54 (47.67–55.10)	60.71 (56.06–65.40)	<0.001*
BMI (kg/m2)	23.00±4.76	22.94±5.12	23.67±4.89	0.17
Previous disease				<0.001*
No	666 (62.42)	465 (67.78)	201 (52.76)
Diabetes	87 (8.15)	19 (2.77)	68 (17.85)
Other disease except diabetes	314 (29.43)	202 (29.45)	112 (29.39)
Urine pH value	6.44±0.53	7.31±0.71	6.25±0.72	<0.001*
Urine occult blood				0.71
−	201 (18.84)	137 (19.97)	64 (16.80)
1+	246 (23.06)	165 (24.05)	81 (21.26)
2+	113 (10.59)	67 (9.77)	46 (12.07)
3+	507 (47.52)	317 (46.21)	190 (49.87)
Urine protein				0.68
−	694 (65.04)	433 (63.12)	261 (68.50)
1+	257 (24.09)	172 (25.07)	85 (22.31)
2+	95 (8.90)	60 (8.75)	35 (9.19)
3+	21 (1.97)	21 (3.06)	0 (0.00)
Urine nitrite				0.08
−	999 (93.62)	635 (92.57)	364 (95.54)
+	68 (6.38)	51 (7.43)	17 (4.46)
Urine leukocyte				0.53
−	535 (50.14)	350 (51.02)	185 (48.56)
1+	366 (34.30)	236 (34.40)	130 (34.12)
2+	53 (4.97)	35 (5.10)	18 (4.72)
3+	113 (10.59)	65 (9.48)	48 (12.60)
Urine bacteria culture				0.46
−	726 (68.04)	462 (67.35)	264 (69.29)
+	341 (32.05)	224 (32.65)	117 (30.71)
Blood sodium (mmol/L)	141.98±2.58	142.14±2.33	141.57±2.34	0.03
Blood potassium (mmol/L)	3.87±0.39	3.85±0.39	3.94±0.36	0.04
Blood calcium (mmol/L)	2.26±0.13	2.24±0.20	2.27±0.14	0.28
Blood phosphorus (mmol/L)	1.05±0.22	1.08±0.20	0.99±0.19	0.01
Serum creatinine (μmol/L)	89.54±29.47	89.06±58.25	91.91±37.11	0.63
Serum uric acid (μmol/L)	341.41±89.26	334.45±84.74	359.08±93.89	0.01
Serum cystatin C (mg/L)	1.20±0.41	1.21±0.81	1.23±0.56	0.75
Serum cholesterol (mmol/L)	4.68 (4.08–5.26)	4.66 (4.10–5.21)	4.78 (4.06–5.35)	0.91
Serum triglycerides (mmol/L)	1.68±1.40	1.53±1.10	1.98±1.59	0.07
Serum high-density lipoprotein (mmol/L)	1.10 (0.93–1.34)	1.13 (0.96–1.34)	1.09 (0.84–1.35)	0.48
Serum low-density lipoprotein (mmol/L)	2.94 (2.33–3.40)	2.77 (2.30–3.38)	2.99 (2.37–3.41)	0.79
Stone location				<0.001*
Bladder	377 (35.33)	155 (22.59)	222 (58.27)
Ureter	384 (35.99)	342 (49.85)	42 (11.02)
Kidney	306 (28.68)	189 (27.55)	117 (30.71)

Data are presented as No. (%), mean ± standard deviation, or median (25th percentile, 75th percentile). *, statistically significant (P<0.05). BMI, body mass index.

Figure 1.

Comparison of urine pH, diabetes status, and stone location between infected and non-infectious stones. (A) Boxplots of urine pH values for the infected stone group and non-infected stone group. The red dots represent the mean values, showing that urine pH is significantly higher in the infected stone group. (B) Bar chart comparing the proportion of patients with and without diabetes between the infected and non-infected stone groups. The infected stone group had a significantly lower proportion of diabetes patients; (C) Bar chart showing the distribution of stone location (bladder, ureter, kidney) between the infected and non-infected stone groups. Infectious stones were predominantly located in the ureter, whereas non-infectious stones were more frequent in the bladder. ***, P<0.001.

Microbiological and stone composition analysis

Among 1,067 patients, 341 (32.0%) had preoperative urine culture results available. Among these 341 patients with positive urine cultures, Proteus mirabilis (24.3% , n=83) and Klebsiella pneumoniae (15.5%, n=53) were the predominant urease-producing pathogens associated with struvite-containing stones (100% and 89%, respectively), consistent with urease-producing bacterial activity (Table 2).

Table 2. Preoperative urine culture results and associated stone types (n=341).

Bacterial species	N (%)	Associated stone type	Urease activity
Proteus mirabilis	83 (24.3)	Struvite (100%)	Yes
Escherichia coli	62 (18.2)	Mixed (65% struvite)	No
Klebsiella pneumoniae	53 (15.5)	Struvite (89%)	Yes
Enterococcus faecalis	34 (10.0)	Non-struvite (82%)	No
Other (Pseudomonas, etc.)	109 (32.0)	Non-struvite (78%)	Variable

Stone types: classified based on infrared spectroscopy. Urease activity: defined as the ability to hydrolyze urea, contributing to alkaline urine (pH >7.0). Data source: preoperative midstream urine cultures; intraoperative stone cultures were not performed.

Risk factors

The logistic regression analysis of all the factors with infectious stones is shown in Table 3. In univariate analysis, gender, age, past medical history, urine pH, serum sodium, serum potassium, serum phosphorus, blood uric acid, and location of stones showed association with infectious stones (P<0.05). A multifactorial logistic regression analysis of the above factors showed that patients without diabetes were more likely to have infectious stones (HR 4.74, 95% CI: 1.50–15.03, P=0.01). Alkaline urine pH also increased the risk of having infectious stones (HR 2.54, 95% CI: 1.33–4.88, P<0.001). The location of the stone was also related to the nature of the stone, with infectious stones being more likely to be found in the ureter (HR 5.60, 95% CI: 2.38–13.17, P<0.001). Ureteral obstruction was adjusted in multivariate analysis but did not diminish the predictive value of stone location (ureter) for infection stones (P<0.001).

Table 3. Logistic regression analysis associated with infectious of all factors.

Variable	Univariate analysis			Multivariate analysis
	HR (95% CI)	P		HR (95% CI)	P
Gender
Male	0.30 (0.17–0.51)	<0.001*		0.52 (0.23–1.17)	0.11
Female	1
Age (years)	0.97 (0.95–0.98)	<0.001*		0.99 (0.97–1.02)	0.67
BMI (kg/m2)	0.97 (0.93–1.02)	0.19
Previous disease
No	4.72 (2.14–10.42)	<0.001*		4.74 (1.50–15.03)	0.01*
Diabetes	1			1
Other disease except diabetes	4.01 (1.71–9.39)	<0.001*		4.20 (1.27–13.87)	0.052
Urine pH value	2.77 (1.64–4.68)	<0.001*		2.54 (1.33–4.88)	<0.001*
Urine occult blood
−	1.32 (0.71–2.45)	0.39
1+	1.22 (0.68–2.20)	0.51
2+	0.97 (0.45–2.07)	0.93
3+	1
Urine protein
−	1.34 (0.40–2.96)	0.33
1+	0.44 (0.05–3.97)	0.46
2+	0.34 (0.04–3.34)	0.36
3+	1
Urine nitrite
−	0.41 (0.14–1.25)	0.12
+	1
Urine leukocyte
−	0.52 (0.23–1.18)	0.52
1+	0.60 (0.26–1.40)	0.60
2+	0.64 (0.17–2.38)	0.64
3+	1
Urine bacteria culture
−	0.84 (0.51–1.38)	0.49
+	1
Blood sodium (mmol/L)	1.11 (1.01–1.22)	0.03*		1.04 (0.90–1.20)	0.62
Blood potassium (mmol/L)	0.55 (0.30–1.00)	0.054		0.51 (0.20–1.31)	0.17
Blood calcium (mmol/L)	0.45 (0.11–1.91)	0.28
Blood phosphorus (mmol/L)	8.49 (2.46–29.26)	<0.001*		4.34 (0.82–23.04)	0.09
Serum creatinine (μmol/L)	1.00 (1.00–1.00)	0.63
Serum uric acid (μmol/L)	0.97 (0.94–0.99)	0.02*		1.00 (1.00–1.00)	0.68
Serum cystatin C (mg/L)	0.95 (0.69–1.30)	0.75
Serum cholesterol (mmol/L)	1.04 (0.68–1.60)	0.84
Serum triglycerides (mmol/L)	0.77 (0.56–1.04)	0.09
Serum high-density lipoprotein (mmol/L)	1.35 (0.35–5.18)	0.66
Serum low-density lipoprotein (mmol/L)	1.02 (0.61–1.69)	0.95
Stone location
Bladder	0.51 (0.30–0.87)	0.01*		0.58 (0.27–1.22)	0.15
Ureter	3.36 (1.82–6.23)	<0.001*		5.60 (2.38–13.17)	<0.001*
Kidney	1			1

*, statistically significant (P<0.05). BMI, body mass index; CI, confidence interval; HR, hazard ratio.

Sensitivity analysis for cohort imbalance

To evaluate the impact of cohort imbalance, sensitivity analyses with IPW were conducted. The adjusted HRs for ureteral stone location (HR =5.58), alkaline urine pH (HR =2.49), and absence of diabetes (HR =4.65) remained statistically significant (P<0.05), confirming the robustness of our primary analysis (Table 4).

Table 4. Full sensitivity analysis with inverse probability weighting.

Variable	Original HR (95% CI)	IPW-adjusted HR (95% CI)	P
Ureteral stone location	5.60 (2.38–13.17)	5.58 (2.35–13.25)	<0.001*
Alkaline urine pH	2.54 (1.33–4.88)	2.49 (1.30–4.75)	0.002*
Absence of diabetes	4.74 (1.50–15.03)	4.65 (1.48–14.62)	0.02*

IPW: adjusted for age, gender, BMI, serum sodium, potassium, and uric acid. CI and P values were calculated using robust standard errors. *, statistically significant (P<0.05). BMI, body mass index; CI, confidence interval; HR, hazard ratio; IPW, inverse probability weighting.

Predictive factors

The predictive value of stone location, urine pH, and diabetes status for predicting infectious stones is shown in Figure 2. Among the three variables, stone location had the best predictive performance, with an AUC of 0.782, a sensitivity of 84.0%, and a specificity of 71.5% (Figure 2A). Urine pH showed moderate predictive ability (AUC =0.766, sensitivity =79.3%, specificity =68.7%) (Figure 2B), while diabetes status demonstrated relatively weak predictive performance (AUC =0.623, sensitivity =56.2%, specificity =58.1%) (Figure 2C).

Figure 2.

ROC curve and AUC comparison of predictive factors for infectious stones. (A) The red curve represents the stone location model, with an AUC of 0.782. (B) The blue curve represents the urine pH model, with an AUC of 0.766. (C) The green curve represents the diabetes status model, with an AUC of 0.623. The gray dashed line represents the random classifier (AUC =0.5). The predictive performance of these factors for infectious stones was evaluated by plotting ROC curves and calculating AUC. AUC, area under the curve; ROC, receiver operating characteristic.

Discussion

In this study, we analyzed the characteristics of infectious stones in the urinary tract and identified the high-risk factors for infectious stones in the urinary tract. From the results of our study, we can see that alkaline urine pH and ureteral stones are high-risk factors for urinary infectious stones. Our study can help clinicians determine the type of stones in patients before surgery, so that they can better plan their treatment and avoid the risk of infectious shock. Previous studies have concluded that dietary factors such as obesity and insulin resistance would lead to urinary stones (18,19). However, from our study, the odds are that urinary stones caused by insulin problems will not be infectious stones. Lower urinary pH due to insulin resistance makes it less likely that infectious urinary stones will occur (20). However, this does not mean that people with diabetes are less likely to have urinary infections; on the contrary, some studies have shown that people with diabetes are at a higher risk of developing urinary septicemia (21,22). The lower prevalence of diabetes in the infected stone group may be attributed to the acidic urinary environment in diabetic patients, which inhibits urease-producing bacteria and thus reduces struvite crystallization. This aligns with studies showing that acidic pH (<6.5) suppresses Proteus and Klebsiella spp. activity, critical pathogens in infection stone pathogenesis (1,2). However, our study did not account for potential confounders such as dietary alkalization or prophylactic antibiotics, which warrant further investigation. Previous studies have suggested that patients with diabetes have an elevated risk of having infectious urinary stones, which may be related to diabetes-induced UTIs. Further research is needed in the future to investigate the relationship between diabetes and infectious urinary stones and UTIs.

The predominance of infectious stones in the ureter (49.85%) contrasts with classical staghorn nephrolithiasis but may reflect acute scenarios where ureteral obstruction facilitates rapid bacterial overgrowth and urease-driven crystallization. This mechanism is supported by studies demonstrating that ureteral stones are more likely to be infected due to a combination of factors, including urinary stasis, anatomical characteristics, and biofilm formation (23,24). Urinary stasis, caused by the obstruction of the ureter by stones, creates an ideal environment for bacterial proliferation, particularly by urease-producing organisms like Proteus mirabilis. These bacteria contribute to biofilm formation, which further promotes infection persistence. Biofilms provide bacteria with protection against host immune responses and antibiotic treatments, complicating infection management and increasing the risk of recurrent infections (25). The anatomical structure of the ureter, with its narrow lumen and peristaltic motion, also facilitates bacterial attachment to the stone surface, further increasing the likelihood of infection. The biofilm matrix acts as a barrier to immune defenses, enabling bacteria to survive longer within the urinary tract and making infections difficult to resolve (26). The younger age and lower comorbidity burden in this subgroup further suggest a distinct clinical pathway compared to chronic staghorn stones, which are typically associated with prolonged UTIs and metabolic abnormalities. Future studies should compare microbial profiles and obstruction timelines between renal and ureteral infection stones to validate this hypothesis.

Our findings provide critical insights into the interplay between urinary pathogens, stone composition, and clinical predictors. The predominance of Proteus mirabilis (24.3%) and Klebsiella pneumoniae (15.5%) in preoperative urine cultures aligns with their well-characterized role as urease-producing bacteria, which hydrolyze urea to generate ammonia and carbonate ions, thereby elevating urine pH and promoting struvite crystallization (27,28). This mechanism directly links bacterial metabolism to stone pathogenesis. In contrast, Escherichia coli (18.2%), a non-urease-producing pathogen, was primarily associated with mixed-composition stones (65% struvite). This observation implies that E. coli may facilitate struvite formation indirectly, potentially through biofilm formation that entraps urease-active bacteria or via inflammatory processes that alter urinary chemistry (29). Similarly, the prevalence of Enterococcus faecalis and other non-urease organisms in non-struvite stones highlights the need to explore alternative pathways, such as calcium phosphate supersaturation driven by urinary stasis or systemic metabolic dysregulation (30). Clinically, our data underscore the importance of integrating urine culture results—even with their inherent limitations—into preoperative risk stratification. While intraoperative stone cultures remain the gold standard for identifying stone-specific pathogens (31), the strong correlation between preoperative Proteus/Klebsiella isolates and struvite composition supports their utility in guiding empirical antibiotic therapy, particularly in resource-limited settings. Future studies should investigate whether targeted antimicrobial prophylaxis against these pathogens reduces postoperative infectious complications.

The clinical characteristics of infected and non-infectious stones in the urinary tract are different, in which patients with infectious stones have higher serum sodium and serum phosphorus levels and lower serum potassium and serum uric acid levels. Our study did not find that serum calcium levels were significantly different in the two types of stones. Previous stone studies have generally focused on the metabolism of calcium, phosphorus, and uric acid, but little has been said about the effect of sodium and potassium metabolism on urinary stones (32). From our findings, serum sodium and serum potassium levels were significantly different in infected and non-infectious stones, which may serve as a novel predictor to assess urinary stones. Meanwhile, previous studies have suggested that a high calcium state will induce renal stones, and that modulation of calcium metabolism can be effective in preventing and controlling renal stones (6,33). In conjunction with our findings, calcium levels are not significantly associated with the development of infectious stones. Preventing urinary stones cannot simply be a matter of controlling the level of a particular electrolyte, but requires a precise prevention program. The condition of stones varies from one location to another. Our study suggests that ureteral stones are more likely to be infected than renal stones, while bladder stones are less likely to be infected. This suggests the need for careful evaluation of the general condition of patients with ureteral stones and adequate preoperative preparation to avoid postoperative infectious shock, especially in patients with ureteral obstruction. Besides, the lack of leukocyturia in some infected stone cases highlights the limitation of relying solely on leukocyte counts for diagnosis.

The predominance of infectious stones (64.3%) in our surgical cohort reflects the clinical reality that infection-associated complications (e.g., obstruction, sepsis) often necessitate urgent intervention, whereas non-infectious stones may be managed conservatively (34). While this imbalance could theoretically bias risk estimates, our sensitivity analyses using IPW confirmed the stability of key predictors, ureteral stone location, alkaline urine pH, and absence of diabetes—even after adjusting for covariates (P<0.05, overlapping CIs). This aligns with studies demonstrating that IPW effectively mitigates bias in unbalanced cohorts by reweighting observations to approximate a balanced population.

In addition, this study evaluated the predictive ability of urine pH, diabetes status and stone location on infectious stones. The results showed that the AUC of stone location was 0.782, showing a good predictive performance. Lu et al. (35) combined radiomics features and clinical data to construct a model for predicting kidney stone types using a multi-layer perceptron (MLP) algorithm, with an AUC of up to 0.95, demonstrating better predictive ability, which may be attributed to its advantage of integrating multiple features. In addition, the ensemble learning model developed by Kazemi and Mirroshandel (14), combined with clinical, demographic, and laboratory data of patients, predicted the type of kidney stones. The sensitivity and specificity of the model reached 81% and 86%, respectively, indicating that multi factor joint analysis can help improve prediction accuracy. In this study, the AUC of urine pH value was 0.766, indicating that urine alkalization has certain value in the diagnosis of infectious stones, but its predictive performance is still lower than the model combining imaging and laboratory data mentioned above. The AUC of diabetes state is 0.623, indicating that its predictive ability is relatively weak, which is consistent with the positioning of this variable as a risk factor rather than a major predictor. Based on the above results, although this study evaluated the predictive ability of infectious stones based on a single clinical feature, there is still a certain gap compared to the multi factor comprehensive model. Future research can further integrate imaging, laboratory indicators, and other biomarkers, using advanced algorithms such as machine learning to construct more accurate and comprehensive predictive models, in order to enhance clinical application value.

From our study, it is clear that the detection rate of infectious stones is not high with common clinical tests, such as urine leukocytes, urine nitrites, and urine culture. The results of these tests were not statistically significantly different in the two groups of patients. This feature may allow clinicians to incorrectly estimate the patient’s UTI and stone type, thereby increasing the risk of the patient experiencing an exacerbation of the infection postoperatively. In this regard, we need more effective clinical tests to predict infected urinary stones as a way to improve treatment planning. Our study has the following shortcomings: first, we collected less clinical data from patients, second, all of the patients that we chose underwent surgery, a premise that led to the study ignoring patients with infected urinary stones who did not undergo surgery, which led to bias, and third, the exclusion of patients with multiple stone sites may limit generalizability to individuals with complex stone disease. Future studies should explore predictors in such populations. Fourth, the small number of postoperative sepsis cases precluded meaningful analysis of the association between infectious stones and sepsis, despite known risks like preoperative infection or intrarenal pressure. Additionally, intraoperative stone cultures were not routinely performed, restricting direct microbial correlation with stone composition. Fifth, dietary habits (e.g., alkaline food intake) were not recorded, potentially confounding urine pH interpretations. Future studies should prioritize prospective designs with standardized stone cultures, comprehensive dietary assessments, and larger sepsis cohorts to address these gaps. Additionally, we did not assess bacterial colonization in non-infectious stones, which may theoretically contribute to sepsis risk during surgery. Prospective studies with standardized stone cultures are needed to compare sepsis rates across stone types.

Conclusions

We collected clinical information on patients suffering from infectious stones in the urinary tract and analyzed it. It was found that infectious stones were more common in women, and that alkaline urine pH and ureteral stones were both high-risk factors for infectious stones. Our study may help clinicians to predict the type of stone in their patients, so that they can better plan their treatment and avoid the risk of infectious shock.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Xiangya Hospital Zhuzhou Central South University (No. 2023-08116) and individual consent for this retrospective analysis was waived.

Data Sharing Statement

Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-244/dss

tau-14-08-2171-dss.pdf ^{(68.7KB, pdf)}

DOI: 10.21037/tau-2025-244