Factors and outcomes leading to postoperative emergency department visits after ureteroneocystostomy

Abstract

BACKGROUND

Ureteroneocystostomy (UNC) is considered the gold standard for pediatric vesicoureteral reflux (VUR) treatment. While UNC lowers the likelihood of needing additional VUR procedures within 12 months, patients also have high 30-day and 90-day readmission rates and emergency department (ED) visits. The most common causes of an ED visit following any urologic procedure are urinary tract infections (UTIs) and catheter/drain concerns. Prior studies are limited in identifying predisposing factors to help mitigate complications of UNC and improve patient outcomes.

AIM

To identify modifiable characteristics at the time of discharge after UNC that predict subsequent unplanned ED visits.

METHODS

The 2020 American College of Surgeons National Surgical Quality Improvement Program Pediatric data was analyzed for patients undergoing UNC for VUR. A total of 1742 patients were evaluated, with 1495 meeting inclusion criteria. Patients with an ED visit within 30 days following an anti-reflux procedure (n = 164) were compared to those who did not return to the ED (n = 1331). Basic statistics and logistic regression analysis were performed to find predictive factors associated with postoperative ED visits after UNC.

RESULTS

Among the 1495 patients, 11.0% visited the ED within the 30-day postoperative period. Patients who returned to the ED visit following UNC were more likely to have had a longer mean operative time, surgical site infection, postoperative UTI, postoperative sepsis, history of prior readmission, unplanned reoperation, blood transfusion, or unplanned urinary catheter placement. Multivariate analysis revealed postoperative UTI (P < 0.001), superficial surgical site infection (P = 0.022), unplanned procedure (P < 0.001), unplanned urinary catheter (P < 0.001), and prematurity (35-36 weeks gestation) (P = 0.004) as independent risk factors for postoperative ED visits.

CONCLUSION

Utmost caution is needed prior to discharge after UNC to forestall a return to the ED. Postoperative infection remains a primary risk for ED visits in the acute postoperative period.

Key Words: Ureteroneocystostomy; Pediatric urology; Vesicoureteral reflux; Urinary tract infection; Postoperative complication; Surgical site infections

Core Tip: Ureteroneocystostomy (UNC) is the gold standard for pediatric vesicoureteral reflux treatment but has high postoperative emergency department (ED) visit rates. Analyzing 2020 data from 1495 patients, we identified key risk factors for ED visits within 30 days post-UNC, including postoperative urinary tract infections, surgical site infections, unplanned procedures, urinary catheter placements, and prematurity. These findings underscore the necessity for stringent discharge protocols to reduce postoperative ED visits, emphasizing the management of infections and other modifiable risk factors to enhance patient outcomes and minimize complications.

INTRODUCTION

Vesicoureteral reflux (VUR) is a common pediatric urologic disorder with an estimated incidence of 10% in the United States[1]. Ureteroneocystostomy (UNC) is considered the gold standard surgical treatment for VUR, with the primary goal of preventing recurrent infections and renal scarring[2,3]. However, there is a lack of data evaluating risk factors for adverse outcomes leading to postoperative emergency department (ED) visits after UNC in the pediatric population.

The most common reasons for returning to the ED following a urologic procedure are concerns for a urinary tract infection (UTI) or issues with a catheter or drain[4]. Specific urologic surgeries in the pediatric population have also been identified to have higher rates of return to the ED, such as penile and lower urinary tract procedures[5]. Further, return to the ED within the acute postoperative period is an often underutilized primary endpoint; most studies use rates of readmission as a primary variable of interest when discussing surgical risk, but this fails to include the cohort of patients who return to the hospital for a procedure-related concern that doesn’t warrant admission. The previously cited study identified that only 15.5% of pediatric patients who returned to the ED after a urologic procedure required readmission, while the rest were managed conservatively[5].

As the current literature lacks a comparative breadth of articles focusing on postoperative returns to the ED rather than those who become readmitted to the hospital, the primary objective of this study is to identify predictive factors that lead to higher rates of post-discharge ED visits after UNC in the pediatric population.

MATERIALS AND METHODS

The National Surgical Quality Improvement Program (NSQIP) is a Health Insurance Portability and Accountability Act-compliant database created by the American College of Surgeons (ACS) to improve quality of care and surgical outcomes. NSQIP includes all major cases determined by the Current Procedural Terminology (CPT) code. With the NSQIP data consisting of de-identified patient information, this retrospective study was exempt from institutional review board approval.

The 2020 ACS NSQIP Pediatric database was analyzed for patients undergoing UNC for VUR. The principal operative CPT codes included were 50780 (ureteral reimplant), 50782 (ureteral reimplant for duplicated ureter), 50783 (ureteral reimplant with tapering), 50947 (laparoscopic ureteral reimplant with cystoscopy and stent), and 50948 (laparoscopic ureteral reimplant without cystoscopy and stent). Robot-assisted laparoscopic procedures are included under codes 50947 and 50948. Patients with unreported postoperative variables were excluded. The patients were then divided into those who did and did not return to an ED within 0-30 days following their anti-reflux procedure (Figure 1).

Figure 1 Flow diagram of inclusion and exclusion criteria. ACS: American College of Surgeons; NSQIP: National Surgical Quality Improvement Program; CPT: Current Procedural Terminology; UNC: Ureteroneocystostomy; VUR: Vesicoureteral reflux; ED: Emergency department.

Patient demographics, comorbid conditions, and postoperative variables were analyzed. Basic statistics were performed, which included Pearson χ² tests for categorical variables and t-test and Welch-tests for continuous variables. Univariate and multivariate analysis was performed using a random forest model with ED visits as the dependent variable. A variable of significance was included and measured for each independent variable, which was objectively chosen for the logistic regression. Statistical analysis was completed using R Version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was accepted at P < 0.05.

RESULTS

A total of 1495 patients from 124 hospitals met the inclusion and exclusion criteria. Among the 1495 patients who underwent UNC, 164 (11.0%) returned to the ED within the 30-day postoperative period (Table 1). Distributions of age, gender, and race were all found to be similar between the two groups. Those who returned to the ED were more likely to be of Hispanic ethnicity (14.0% vs 13.4%, P = 0.01). Patients with postoperative ED visits had an increased incidence of structural pulmonary abnormalities (4.3% vs 1.9%, P = 0.05), gastrointestinal diseases (9.1% vs 5.0%, P = 0.03), and were more likely to have had an ostomy (7.9% vs 4.4%, P = 0.04). Those who returned post-UNC were also more likely to have been born at 35-36 weeks (11.0% vs 3.9%, P < 0.01) and have minor cardiac risk factors (6.1% vs 2.7%, P = 0.02) (Table 2). All other demographic and comorbid variables were similarly distributed between groups. No difference in VUR severity was found between those who returned to the ED postoperatively and those who did not.

Table 1 Patient demographics and comorbidities for patients with an emergency department visit following ureteroneocystostomy, n (%)/mean ± SD.

Patient demographics	Total cohort (n = 1495)	Postoperative ED visit (n = 164)	No Postoperative ED visit (n = 1331)	P value1
Age	1666 ±1208	1605.5 ± 1207	1673.3 ± 1196	0.4974
Male gender	446 (29.8)	55 (33.5)	391 (29.4)	0.2720
Non-Caucasian race	304 (20.3)	38 (23.2)	266 (20.0)	0.9070
Hispanic ethnicity	201 (13.4)	23 (14.0)	178 (13.4)	0.0109
Inpatient	1010 (67.6)	120 (73.2)	890 (65.9)	0.1038
Height	40.2 ± 8.9	39.0 ± 8.3	40.3 ± 8.6	0.4041
Weight	42.0 (24.3)	39.3 (22.5)	42.2 (24.5)	0.0696
Comorbid conditions/past medical history
Ventilator dependent	2 (0.1)	1 (0.6)	1 (0.1)	0.0773
History of asthma	25 (1.7)	5 (3.0)	20 (1.5)	0.1453
Bronchopulmonary dysplasia/chronic lung disease	14 (0.9)	3 (1.8)	11 (0.8)	0.2085
Oxygen support	10 (0.7)	2 (1.2)	8 (0.6)	0.3594
Tracheostomy	2 (0.1)	0 (0.0)	2 (0.1)	0.6195
Structural pulmonary/airway abnormalities	32 (2.1)	7 (4.3)	25 (1.9)	0.0461
Esophageal/gastric/intestinal disease	81 (5.4)	15 (9.1)	66 (5.0)	0.0254
History of cardiac surgery	30 (2.0)	3 (1.8)	27 (2.0)	0.8637
Developmental delay/impaired cognitive status	100 (6.7)	15 (9.1)	85 (6.4)	0.1820
Seizure disorder	17 (1.1)	3 (1.8)	14 (1.1)	0.3758
Cerebral palsy	6 (0.4)	2 (1.2)	4 (0.3)	0.0791
Structural CNS abnormality	65 (4.3)	9 (5.5)	56 (4.2)	0.4482
Neuromuscular disorder	35 (2.3)	7 (4.3)	28 (2.1)	0.0838
Ostomy	71 (4.7)	13 (7.9)	58 (4.4)	0.0427
Congenital malformation	1144 (76.5)	130 (79.3)	1014 (76.2)	0.6228
Childhood malignancy	4 (0.3)	0 (0.0)	4 (0.3)	0.7811
Steroid use (Within 30 days)	6 (0.4)	2 (1.2)	4 (0.3)	0.0791
Nutritional support	30 (2.0)	5 (3.0)	25 (1.9)	0.3133
Hematologic disorder	33 (2.2)	3 (1.8)	30 (2.3)	0.7270

¹The P value represents tests for significant variation from χ² analysis of categorical variables, t-test for continuous factors with equal variance.

ED: Emergency department; CNS: Central nervous system.

Table 2 Patient demographics and comorbidities for patients with an emergency department visit following ureteroneocystostomy (cont.), n (%).

	Total cohort (n = 1495)	Postoperative ED visit (n = 164)	No postoperative ED visit (n = 1331)	P value1
Premature birth
25-30 weeks	15 (1.0)	1 (0.6)	14 (1.1)	1.000
31-34 weeks	39 (2.6)	6 (3.7)	33 (2.5)	0.4303
35-36 weeks	70 (4.7)	18 (11.0)	52 (3.9)	0.0003
Cardiac risk factors
Minor	46 (3.1)	10 (6.1)	36 (2.7)	0.0176
Major	44 (2.9)	5 (3.0)	39 (2.9)	0.9324
Severe	3 (0.2)	0 (0.0)	3 (0.2)	P = 0.5429
VUR disease severity
VUR Grade 1	40 (2.7)	7 (4.3)	33 (2.5)	0.1805
VUR Grade 2/3	446 (29.8)	50 (30.5)	396 (29.8)	0.8460
VUR Grade 4/5	887 (59.3)	95 (57.9)	792 (59.5)	0.6981
ASA classification
ASA Class 1	250 (16.7)	19 (11.6)	231 (17.4)	0.0618
ASA Class 2	1060 (70.9)	119 (72.6)	941 (70.7)	0.6204
ASA Class 3	177 (11.8)	25 (15.2)	152 (11.4)	0.1528
ASA Class 4	6 (0.4)	1 (0.6)	5 (0.4)	0.6547

¹The P value represents tests for significant variation from χ² analysis of categorical variables, t-test for continuous factors with equal variance.

VUR: Vesicoureteral reflux; ASA: American society of anesthesiology; VUR: Grade determined using standardized voiding cystourethrogram; ED: Emergency department.

Postoperatively, patients who returned had a longer mean total operation time (182 mins) on average than those who did not (165 minutes) (P = 0.01). The mean length of stay (22.38 vs 20.09, P < 0.01) and days from operation to discharge (2.13 vs 1.29, P < 0.01) also varied significantly between the two groups. Rates of laparoscopic vs open approach were similar, as were the percentage of patients who underwent unilateral vs bilateral fixation. Patients who returned for an ED visit were found to have higher rates of superficial incisional (1.8% vs 0.2%, P < 0.01) and organ space surgical site infections (SSIs) (1.2% vs 0.1%, P < 0.01). Postoperative UTIs (14.0% vs 1.9%, P < 0.1) and sepsis rates (3.7% vs 0.2%, P < 0.01) were also higher in patients who returned to the ED, although there was no difference in rates of septic shock (P = 0.73) (Table 3). Preoperative/intraoperative urine cultures did not vary between groups. The patients who visited the ED had higher rates of readmission (31.7% vs 1.7%, P < 0.01), progressive renal insufficiency (1.2% vs 0.2%, P = 0.04), unplanned reoperations (7.3% vs 0.8%, P < 0.01), blood transfusions (1.2% vs 0.2%, P = 0.01), unplanned procedures related to the initial UNC (13.4% vs 1.1%, P < 0.01), and unplanned urinary catheterization (17.7% vs 1.2%, P < 0.01).

Table 3 Postoperative outcomes for patients with an emergency department visit following ureteroneocystostomy, n (%)/mean ± SD.

	Total cohort (n = 1495)	Postoperative ED visit (n = 164)	No postoperative ED visit (n = 1331)		P value1
Operative time
Total operation time (Minutes)	166.5 ± 74.3	181.8 ± 84.2		164.6 ± 72.9	0.0130
Length of hospital stay
Length of stay (Days)	20.3 ± 26.6	22.4 ± 43.0		20.1 ± 23.8	0.5042
Days from operation to discharge (Days)	1.8 ± 1.6	2.1 ± 2.10		1.7 ± 1.47	0.0220
Operative approach
Laparoscopic/MIS	189 (12.6)	16 (9.8)		173 (13.0)	0.2387
Open or N/A	1175 (78.6)	133 (81.1)		1042 (78.3)	0.4078
Laparoscopic/MIS and open	131 (8.8)	15 (9.1)		116 (8.7)	0.8539
Unilateral procedure	692 (46.3)	73 (48.5)		619 (46.5)	0.6392
Bilateral procedure	798 (53.4)	91 (55.5)		707 (53.1)	0.5993
SSI
Superficial incisional SSI	5 (0.3)	3 (1.8)		2 (0.2)	0.0110
Deep incisional SSI	1 (0.1)	1 (0.6)		0 (0.0)	0.1097
Organ space SSI	3 (0.2)	2 (1.2)		1 (0.1)	0.0033
Preoperative/intraoperative urine culture
No bacterial growth	791 (52.9)	82 (50.0)		709 (53.3)	0.4290
Bacterial growth, not UTI	111 (7.4)	14 (8.5)		97 (7.3)	0.5650
Bacterial growth, UTI	76 (5.1)	10 (6.1)		66 (5.0)	0.5311
Postoperative outcomes
Postoperative UTI	48 (3.2)	23 (14.0)		25 (1.9)	< 0.0001
Postoperative sepsis	9 (0.6)	6 (3.7)		3 (0.2)	< 0.0001
Postoperative septic shock	1 (0.1)	0 (0.0)		1 (0.1)	0.7256
Readmission	75 (5.0)	52 (31.7)		23 (1.7)	< 0.0001
Progressive renal insufficiency	5 (0.3)	2 (1.2)		3 (0.2)	0.0375
Unplanned reoperation	23 (1.5)	12 (7.3)		11 (0.8)	< 0.0001
Blood transfusion	4 (0.3)	2 (1.2)		2 (0.2)	0.0124
Ureteral stent/catheter	752 (50.3)	95 (57.9)		657 (49.4)	0.1410
Unplanned procedure related to anti-reflux procedure	36 (2.4)	22 (13.4)		14 (1.1)	< 0.0001
Unplanned urinary catheter (Intermittent or Indwelling)	45 (3.0)	29 (17.7)		16 (1.2)	< 0.0001

¹The P value represents tests for significant variation from χ² analysis of categorical variables, t-test for continuous factors with equal variance.

MIS: Minimally invasive surgery; SSI: Surgical site infection; UTI: Urinary tract Infection; ED: Emergency department.

On multivariate analysis (OR, 95%CI, P value), postoperative UTIs (6.39, 2.83-12.4, P < 0.01) and superficial SSIs (12.3, 1.48-107, 0.02) were found to be predictive, as were unplanned procedures (5.73, 2.15-15.3, < 0.01) and catheterization (8.81, 3.90-20.0, < 0.01). No association was found between returning to the ED and postoperative sepsis, organ space SSI, renal insufficiency, or blood transfusion. While cardiac risk factors were also no longer associated with ED visits, being born prematurely was, specifically at 35-36 weeks (2.75, 1.39-5.18, 0.01) (Table 4).

Table 4 Logistic regression analysis of factors leading to postoperative emergency department visits.

	Univariate analysis		Multivariate analysis
	OR (95% CI)	P value	OR (95% CI)	P value1
Patient characteristics
Hispanic ethnicity	1.06 (0.65-1.66)	0.818	1.18 (0.69-1.92)	0.529
Structural pulmonary abnormalities	2.33 (0.92-5.20)	0.052	2.60 (0.86-6.96)	0.087
Esophageal/gastric/intestinal disease	1.93 (1.04-3.38)	0.028	1.12 (0.46-2.52)	0.793
Ostomy	1.89 (0.97-3.42)	0.046	0.94 (0.36-2.22)	0.890
Postoperative variables and outcomes
Length of stay ≥ 20 days	4.05 (0.19-42.43)	0.254	1.87 (0.05-34.8)	0.704
Operative time ≥ 167 minutes	1.51 (1.09-2.09)	0.013	1.19 (0.82-1.71)	0.362
Postoperative urinary tract infection	8.52 (4.69-15.4)	<0.001	6.10 (2.93-12.4)	< 0.001
Postoperative sepsis	16.81 (4.39-80.3)	<0.001	3.28 (0.39-33.0)	0.272
Superficial incisional surgical site infection	12.38 (2.04-94.5)	0.006	12.3 (1.48-107)	0.022
Organ space surgical site infection	16.42 (1.56-355)	0.023	1.56 (0.06-58.3)	0.788
Progressive renal insufficiency	5.47 (0.72-33.2)	0.064	0.92 (0.06-11.4)	0.948
Blood transfusion	8.20 (0.98-68.8)	0.036	3.19 (0.13-214)	0.500
Unplanned procedure related to anti-reflux procedure	14.6 (7.37-29.8)	< 0.001	5.73 (2.15-15.3)	< 0.001
Unplanned urinary catheter (intermittent or indwelling)	17.7 (9.47-15.1)	< 0.001	8.81 (3.90-20.0)	< 0.001
Cardiac risk factors		0.153		0.107
None	Referent		Referent
Minor	2.34 (1.08-4.63)	0.021	1.57 (0.59-3.69)	0.333
Major	1.08 (0.37-2.54)	0.876	0.25 (0.05-0.98)	0.070
Prematurity		0.003		0.023
Full-term	Referent		Referent
25-30 completed weeks gestation	0.63 (0.03-3.18)	0.660	0.32 (0.02-2.00)	0.317
31-34 completed weeks gestation	1.61 (0.60-3.65)	0.292	1.21 (0.36-3.26)	0.734
35-36 completed weeks gestation	3.07 (1.70-5.30)	< 0.001	2.75 (1.39-5.18)	0.002

¹The P value represents tests for significant variation from univariate and multivariate analysis.

DISCUSSION

This study aimed to identify patient factors associated with subsequent ED visits within the acute post-UNC period. Of the total cohort, 11.0% visited the ED post-UNC. We found that postoperative UTIs increased the odds of returning to the ED within the 30-day postoperative period by 6.4 times compared to those without a UTI. Unfortunately, UTIs may occur postoperatively in an otherwise successful UNC procedure. Despite preemptive antibiotics and antiseptic measures taken perioperatively, presumed UTIs remain one of the most common reasons for return after pediatric urologic procedures[5]. In one retrospective review of 398 patients who underwent UNC, 27.2% developed a postoperative UTI, with contributing factors such as female gender, breakthrough infections, and the presence of voiding dysfunction[6]. Factors such as bacterial colonization during catheterization, incomplete bladder emptying, or pre-existing urologic abnormalities may explain some of the underlying biological mechanisms that increase the baseline risk of infection in such patients. Superficial SSIs were also associated with a higher likelihood of leading to a postoperative ED visit, although still a rare event overall, occurring in only 1.8% of patients included in the current analysis. Further investigation is needed into perioperative protocols, catheterization practices, and pre-existing urologic abnormalities to better understand the etiology of these infections.

Unplanned catheterization following UNC was another independent variable associated with an increased risk of ED return. A well-known risk of indwelling catheters is the development of catheter-associated UTIs, which alone could warrant visiting the ED, as previously mentioned[7]. It should also be acknowledged that catheterization following UNC could have been performed secondary to an unanticipated surgical complication, such as urinary retention. Such difficulties may appear resolved after initial catheterization, prompting discharge home, only to return or worsen once out of the hospital[8,9]. Additionally, patients sent home with an indwelling catheter can serve as a stressor for families and prompt return, particularly if painful or believed to be malfunctioning. Performing catheter-less UNC or removing it as soon as possible may be warranted. One retrospective study has found that catheter-less UNC reduced length of stay, rates of readmission, and the amount of postoperative medication needed without worsening complication rates[10]. However, it's important to note that these strategies are not universally adopted, and further research is needed to assess their feasibility and efficacy across diverse clinical settings. Implementing catheterization-free protocols requires careful patient selection and adherence to surgical best practices to ensure optimal outcomes. Clinicians should weigh the potential benefits against the risks and consider individual patient factors when deciding on the use of catheterization in UNC procedures.

While previous studies on UNC most always include the patients’ ages at the time of surgery, few include premature birth status as a possible risk factor in postoperative outcomes[2,6,11]. Premature birth status at 35-36 weeks of gestation was found to increase the likelihood of returning to ED postoperatively. However, children born between 25 and 34 weeks were not more likely to have a postoperative ED visit compared to their full-term counterparts. Multiple factors could explain this discrepancy. One hypothesis for why more premature infants were not more likely to return could be explained by the expected standard of care. With preterm neonates being a high-risk patient category, postoperative care may be more comprehensive, and greater caution is taken before sending them home. Increased parental instructions and education are also likely given to those born extremely preterm, thus helping to prevent ED visits.

The study leverages one of the largest, most up-to-date national databases for pediatric surgery, providing robust data for analysis. However, limitations similar to other retrospective, database-driven studies apply. The data may not fully represent the entire population, as the majority of reporting hospitals are large academic institutions, potentially leading to selection bias. This may limit the generalizability of the findings to smaller or non-academic settings. While machine learning was used to minimize subjective bias in the selection of independent variables for multivariable logistic regression, the potential for residual confounding remains. This could affect the accuracy of the identified risk factors, particularly if unmeasured variables influenced the outcomes. Furthermore, the reliance on retrospective data prevents establishing causal relationships between identified variables and postoperative ED visits.

To mitigate these limitations, future research could incorporate data from a broader range of healthcare settings, including community hospitals, to improve representativeness. Prospective studies are also needed to validate these findings and explore causal mechanisms. Additionally, integrating more granular clinical data, such as socioeconomic factors or detailed perioperative variables, not captured by NSQIP, could provide a more comprehensive understanding of modifiable risk factors. Despite these constraints, this study provides valuable insights into potentially modifiable risk factors for postoperative ED visits following pediatric surgical interventions, particularly given the low volume of UNCs performed annually at the national level.

CONCLUSION

Postoperative infections, unplanned procedures, and prematurity are primary factors contributing to ED visits within the immediate postoperative period after ureteral reimplantation. Additional studies are needed to determine if replicable contributing factors are found concerning ED visits after other anti-reflux procedures.