Post-Operative Infections among Patients Undergoing Radical Cystectomy at a Tertiary Center

Abstract

Background:

Post-operative infection after radical cystectomy is a common complication. It is essential to identify modifiable risk factors that can predict post-operative infection to reduce the incidence of infection after radical cystectomy. We investigated the incidence of post-operative infection, associated pathogens, and risk factors for post-operative infection after radical cystectomy.

Patients and Methods:

Patients who underwent radical cystectomy for bladder cancer in a single urologic department from 2014 to 2016 were included. Age, gender, length of stay, body mass index (BMI), diabetes mellitus, hypertension, pre-operative estimated glomerular filtration rate, hydronephrosis, smoking status, neoadjuvant chemotherapy, pathologic tumor stage, pathologic nodal stage, types of urinary diversion, laparoscopic approach, operative time, and receipt of peri-operative blood transfusion were assessed to determine the association of these factors with the risk of infection within 30 days of radical cystectomy. Risk factors were assessed for correlation with any kind of infection and separately for urinary tract infection (UTI), blood infection/septic shock (BI), and surgical site infection (SSI). Pathogens were identified in all cases of infection.

Results:

A total of 134 patients were included in the analysis. Post-operative infection was diagnosed in 10.45%. Urinary tract infection, BI, and SSI were diagnosed in 3.73%, 4.48%, and 7.46%, respectively. Co-infections and mixed infection were diagnosed in 50% of patients with infection. Enterococcus was the most common pathogen. On multivariable analysis, age, BMI ≥30 kg/m², and laparoscopic approach were associated with the risk of infection; odds ratios (ORs) were 1.27 (95% confidence interval [CI] 1.01–1.59), 13.07 (1.39–122.88), and 0.07 (0.01–0.78), respectively.

Conclusion:

One-tenth of patients undergoing radical cystectomy developed an infection within 30 days of surgery. Fifty percent of patients had co-infection with UTI, SSI, and/or BI and 50% of infections were polymicrobial. Increased BMI is the strongest factor associated independently with all types of infection. Therefore, obese patients should be encouraged to lose weight pre-operatively.

Radical cystectomy with urinary diversion is a urologic operative procedure with a significant rate of associated morbidity and mortality. Post-operative infection after radical cystectomy is one of the most common complications. Risk of infectious complications was estimated to be between 25% and 44% [1,2]. Infection is strongly associated with early re-admission. In patients undergoing radical cystectomy, the re-admission rate within 30 days is the highest among all major urologic oncology procedures, and approximately 14% of these re-admissions are the result of infection [3].

In previous studies, three primary types of infection after radical cystectomy were assessed: urinary tract infection (UTI), blood infection/septic shock (BI), and surgical site infection (SSI). However, most studies focused only on UTI, and there is still a lack of data concerning BI and SSI [2,4,5]. Urinary tract infection is the most common infection among patients after radical cystectomy and ileal neobladder; the risk of infection is the highest within the first 30 days after radical cystectomy. Limited data exist concerning the organisms responsible for these infections and some patients have multiple pathogen infections or are infected by drug-resistant pathogens. Worldwide, the most frequently isolated pathogen is Escherichia coli, but this changes during follow-up [5].

Post-operative infection has been shown to result in frequent adverse events, such as renal failure, wound dehiscence, and prolonged hospital stay. In a study performed by Meyer et al. [6], all patients with wound dehiscence after radical cystectomy had SSI [6]. Treatment for post-operative infection is associated with a substantial increase in the cost of treatment, whereas current treatment for bladder cancer is one of the most expensive therapies among all oncologic diseases [7]. Thus, identifying modifiable risk factors for post-operative infection is essential for reducing the rate of infection after radical cystectomy. In this study, we investigated the incidence of post-operative infection in our institution as well as its association with pathogens and risk factors for post-operative infection after radical cystectomy.

Patients and Methods

A review of medical records in a single urology department was performed to identify patients who underwent radical cystectomy and urinary diversion for bladder cancer from January 2014 to July 2017. The goal of the study was identification of all patients with SSI, UTI, and BI during 30-day follow-up after radical cystectomy. Additionally, the number of patients with co-infection and mixed infection were noted. Co-infection was defined as the occurrence of two separate infections, for example, UTI and BI. Mixed infection was defined as a single infection (only UTI, BI, or SSI) caused by a variety of bacterial species that simultaneous caused the same infection. The following exclusion criteria were applied: patients with documented infection at the time of radical cystectomy, patients undergoing concurrent organ resection such as nephrectomy or nephroureterectomy, and patients undergoing partial cystectomy.

When infection was suspected based on symptoms and/or elevation of inflammatory parameters, patients were admitted and treated with intravenous broad spectrum antibiotic agents. The treatment was verified after detection of pathogen susceptibility.

Operations before August 2016 were performed via the open surgical approach, whereas the laparoscopic approach was performed after August 2016. In the analyzed period, three types of urinary diversion were performed: single/double-sided ureterostomy, ileal conduit, and orthotopic neobladder. After surgery, all patients were treated with prophylactic antibiotic agents, which usually included seven days of a second-generation cephalosporin and, in case of ileal conduit or orthotopic neobladder, an additional five days of intravenous metronidazole. Pre-operative bowel preparation was not performed routinely [8]. Drainage tubes were removed as indicated by the surgeon. All patients were trained in stoma care or intermittent self-catheterization by a skilled nurse before leaving the hospital.

Characteristics, including age, gender, length of hospital stay, body mass index (BMI), diabetes mellitus, hypertension, estimated glomerular filtration rate, pre-operative hydronephrosis, smoking status, neoadjuvant chemotherapy, pathologic stage, lymph node involvement, type of urinary diversion, operative method (open/laparoscopic), operative time, and receipt of peri-operative blood transfusion were collected. Body mass index was analyzed as a dichotomous variable (<30 kg/m² vs. ≥30 kg/m²), as in previous radical cystectomy series [2]. Cigarette smoking status was defined before surgery as never smoker, former smoker, and active smoker. The presence of hydronephrosis was assessed one day before radical cystectomy by ultrasonography. Operative time was categorized into two ranges (<240 min vs. ≥240 min). All of the abovementioned potential risk factors were assessed for association with any kind of infection and separately for UTI, BI, and SSI.

Statistica software, version 13.5 (StatSoft, Inc., Tulsa, OK) was used for all statistical analyses. A p value <0.05 was considered significant and all p values were two-sided.

Results

A total of 146 patients underwent surgery because of bladder cancer during the analyzed period. Three patients were excluded because of a documented infection at the time of radical cystectomy; all three patients had Escherichia coli UTI. Five and four patients were excluded because of treatment with concurrent organ resection and partial cystectomy, respectively. A total of 134 patients were included in our analysis. Mean age was 65.9 years and ranged from 44 to 85 years. The female to male ratio was 31:103. Table 1 highlights the clinical variables of the cohort.

Table 1.

Baseline Clinicopathologic Characteristics, Stratified by Perioperative Infection Status

Characteristic	No infection	Infection	Total
No. of patients	120	14	134
Age (mean)	65.9	69.3	65.9
Gender
Male	93	10	103
Female	27	4	31
Median length of stay (IQR)	8 (7–9)	9 (6–12)
BMI, kg/m ²
25–<30	108	8	116
≥30	12	6	18
Diabetes mellitus
No	100	8	108
Yes	20	6	26
Hypertension
No	56	5	61
Yes	61	9	70
Pre-operative eGFR
≥90	33	3	36
60–89	51	3	54
30–59	26	7	33
<30	10	1	11
Hydronephrosis
No	60	7	67
Yes	60	7	67
Smoking
Never	16	1	17
Former	75	10	85
Active	29	3	32
Neoadjuvant chemotherapy
No	108	10	118
Yes	18	4	22
Pathologic T stage
pT0	13	1	14
only CIS	1	1	2
Ta	0	0	0
T1	12	0	12
T2	23	1	24
T3	48	9	57
T4	23	2	25
Pathologic N stage
N0	82	8	90
N+	38	6	44
Urinary diversion
Conduit	55	8	63
Ureterocutaneostomy	61	6	67
Orthotopic neobladder	4	0	4
Laparoscopic approach
No	77	12	89
Yes	43	2	45
Operative time (min)
<240	70	10	80
≥240	50	4	54
PBI
No	54	3	57
Yes	66	11	77

IQR = interquartile range; eGFR = estimated glomerular filtration rate; BMI: body mass index; PBT = peri-operative blood transfusion.

In the analyzed period, 10.45% (14/134) patients had post-operative infection. The rate of BI was the highest at 7.46%, whereas UTI and SSI were present in 3.73% and 4.48% of patients, respectively. We found only one incidence of intra-abdominal infection in laparoscopy approach (Enterococcus faecalis); three patients had intra-abdominal infection in open approach. Table 2 provides information about organisms recovered in the operative site infections. Co-infections were diagnosed in 50% of patients with infection. Nevertheless, only one patient had co-existing infections of the blood, urine, and incision.

Table 2.

Organisms Recovered in the Operative Site Infections and Antibiotic Resistance

	Pathogenes
Antibiotic name	Escherichia coli	Escherichia coli	Escherichia coli	Enterococcus faecalis	Enterococcus faecium	Enterococcus faecium	Staphylococcus aureus	Enterobacter cloacae	Proteus mirabilis
Ampicillin/amoxicillin	R	R	R		R	R		R	R
Amoxicillin/clavulanic acid	R	R	R					R
Piperacillin/tazobactam	R
Cloxacillin
Cefuroxime	R	R						R
Cefotaxime	R	R
Ceftazidime	R	R
Ceftriaxone	R	R
Cefepime		R
Imipenem
Meropenem
Ertapenem
Gentamicin				R					R
Amikacin
Ciprofloxacin
Levofloxacin
Erythromycin							R
Clindamycin							R
Trimethoprim/sulfamethoxazole		R							R
Vancomycin
Teikoplamina
Tigecycline

R = resistance.

In the analysis of all types of infection, Enterococcus was the most common implicated pathogen (5/14; 35.71%). However, in UTI, Enterobacter was the most frequent pathogen (3/6; 50.0%). Fifty percent of patients with post-operative infection had mixed infection. Escherichia coli and Enterococcus were the most common co-existing pathogens. Fungal infection was diagnosed in four patients.

Univariable logistic regression analysis indicated that BMI ≥30≥kg/m² and diabetes mellitus were associated with post-operative infection, with odds ratios (ORs) of 6.75 (95% confidence interval [CI], 2.00–22.75) and 3.75 (95% CI 1.17–11.99), respectively. In multivariable analysis, BMI remained associated with post-operative infection, with an OR of 13.07 (95% CI 1.39–122.88); diabetes mellitus was no longer a predictor for post-operative infection. Furthermore, in multivariable analysis, age correlated with post-operative infection (OR 1.27 [95% CI 1.01–1.59]). Laparoscopic approach was the only variable associated with a decreased risk of post-operative infection (OR 0.07 [95% CI 0.01–0.78]; Table 3).

Table 3.

Logistic Regression Analyses for Risks of Pre-Operative Infection within Thirty Days of Radical Cystectomy

	Univariable		Multivariable
	OR (95% CI)	p	OR (95% CI)	p
Age (cont.)	1.06 (0.99–1.15)	0.097	1.27 (1.01–1.59)	0.038
Gender
Male	1.00 (Ref.)	—	1.00 (Ref.)	—
Female	1.38 (0.40–4.74)	0.611	1.41 (0.15–13.22)	0.764
Length of stay (cont.)	1.07 (0.98–1.16)	0.115	1.04 (0.88–1.24)	0.637
BMI, kg/m²
<30	1.00 (Ref.)	—	1.00 (Ref.)	—
≥30	6.75 (2.00–22.75)	0.002	13.07 (1.39–122.88)	0.025
Diabetes mellitus
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	3.75 (1.17–11.99)	0.026	3.75 (0.50–28.35)	0.200
Hypertension
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	1.74 (0.55–5.50)	0.345	0.82 (0.11–6.13)	0.846
Preoperative eGFR
≥90	1.00 (Ref.)	—	1.00 (Ref.)	—
60–89	0.65 (0.12–3.40)	0.607	0.40 (0.04–4.00)	0.433
30–59	2.96 (0.69–12.58)	0.141	1.18 (0.15–9.38)	0.872
<30	1.10 (0.10–11.78)		0.77 (0.03–21.62)	0.878
Hydronephrosis
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	1.00 (0.33–3.03)	1.000	0.48 (0.07–3.21)	0.447
Smoking
Never	1.00 (Ref.)	—	1.00 (Ref.)	—
Former	2.13 (0.25–17.87)	0.484	2.41(0.09–63.26)	0.597
Active	1.66 (0.16–17.25)	0.674	6.29 (0.21–188.06)	0.289
Neoadjuvant chemotherapy
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	2.27 (0.64–8.01)	0.204	3.49 (0.34–35.45)	0.290
Pathologic T stage
≤pT2	1.00 (Ref.)	—	1.00 (Ref.)	—
≥pT3	2.53 (0.67–9.54)	0.170	3.54 (0.33–38.31)	0.299
Pathologic N stage
N0	1.00 (Ref.)	—	1.00 (Ref.)	—
N+	1.62 (0.52–4.99)	0.402	1.55 (0.26–9.10)	0.630
Urinary diversion
Conduit	1.00 (Ref.)	—	1.00 (Ref.)	—
Ureterocutaneostomy	1.20 (0.39–3.68)	0.747	0.78 (0.08–7.87)	0.834
Neobladder	0.00 (0.00–0.00)	0.997	0.00 (0.00–0.00)	0.997
Laparoscopic approach
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	0.30 (0.06–1.40)	0.124	0.07 (0.01–0.78)	0.031
Operative time (min)
<240	1.00 (Ref.)	—	1.00 (Ref.)	—
≥240	0.56 (0.17–1.89)	0.349	0.36 (0.06–2.29)	0.282
PBT
No	1.00 (Ref.)	—	1.00 (Ref.)	—
Yes	3.00 (0.80–11.30)	0.105	0.94 (0.11–8.33)	0.957

BMI = body mass index; PBT = perioperative blood transfusion.

Separate analysis of different types of infection revealed that BMI ≥30≥kg/m² was the strongest predictor of post-operative infection and was independently associated with UTI (OR 11.4 [95% CI 1.76–73.85]), SSI (OR 7.53 [95% CI 1.39–40.78]), and BI (OR 5.24 [95% CI 1.32–20.86]). Diabetes mellitus was associated with an increased risk of UTI (OR 6.91 [95% CI 1.09–43.75]) and SSI (OR 9.64 [95% CI 1.66–55.92]). Additionally, a separate analysis showed that BI was associated with the length of hospital stay (OR 1.10 [95% CI 1.01–1.20]; Table 4).

Table 4.

Logistic Regression Analyses for Risks of Urinary Tract Infection, Surgical Site Infection, and Blood Infection within Thirty Days of Radical Cystectomy

	UTI		SSI		Blood infection/sepsis
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Age (cont.)	0.98 (0.87–1.10)	0.731	1.03 (0.92–1.14)	0.625	1.08 (0.99–1.18)
Gender
Male	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Female	0.83 (0.09–7.67)	0.866	0.65 (0.07–5.81)	0.703	1.47 (0.36–6.06)	0.594
Length of stay (cont.)	1.11 (0.10–1.23)	0.058	1.07 (0.96–1.20)	0.232	1.10 (1.01–1.20)	0.029
BMI, kg/m²
<30	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
≥30	11.4 (1.76–73.85)	0.011	7.53 (1.39–40.78)	0.019	5.24 (1.32–20.86)	0.019
Diabetes
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	6.91 (1.09–43.75)	0.040	9.64 (1.66–55.92)	0.012	3.09 (0.80–11.88)	0.101
Hypertension
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	1.39 (0.22–8.59)	0.724	1.89 (0.33–10.62)	0.476	1.41 (0.38–5.23)	0.611
Preoperative eGFR
≥90	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
60–89	0.66 (0.04–10.91)	0.771	0.32 (0.03–3.68)	0.631	1.00 (0.16–6.30)	1.000
30–59	3.50 (0.35–35.44)	0.289	1.70 (0.27–10.87)	0.575	2.34 (0.40–13.74)	0.345
<30	0.0 (0.0–0.0)	0.997	0.0 (0.0–0.0)	0.997	1.70 (0.14–20.75)	0.678
Hydronephrosis
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	0.66 (0.11–4.06)	0.651	2.06 (0.36–11.68)	0.413	1.00 (0.28–3.63)	1.000
Smoking
Never	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Former	0.23 (0.03–1.71)	0.149	0.34 (0.05–2.20)	0.258	0.37 (0.08–1.69)	0.199
Active	0.29 (0.02–3.43)	0.326	0.29 (0.02–3.43)	0.326	0.38 (0.06–2.50)	0.312
Neoadjuvant chemotherapy
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	3.63 (0.57–23.15)	0.172	1.02 (0.11–9.17)	0.987	1.30 (0.26–6.58)	0.751
Pathologic T stage
≤pT2	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
≥pT3	1.61 (0.53–4.91)	0.396	1.28 (0.23–7.26)	0.779	1.52 (0.37–6.18)	0.555
Pathologic N stage
N0	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
N+	0.50 (0.05–4.61)	0.541	0.40 (0.04–3.49)	0.404	2.18 (0.60–7.97)	0.239
Urinary diversion
Conduit	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Ureterocutaneostomy	1.34 (0.22–8.30)	0.752	0.88 (0.17–4.52)	0.877	0.88 (0.24–3.18)	0.839
Neobladder	0.0 (0.0–0.0)	0.997	0.0 (0.0–0.0)	0.997	0.0 (0.0–0.0)	0.997
Laparoscopic approach
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	0.0 (0.0–0.0)	0.998	0.0 (0.0–0.0)	0.997	0.47 (0.10–2.32)	0.354
Operative time (min)
<240	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
≥240	0.99 (0.16–6.11)	0.989	0.73 (0.13–4.14)	0.723	0.15 (0.02–1.21)	0.745
PBT
No	1.00 (Ref.)		1.00 (Ref.)		1.00 (Ref.)
Yes	1.11 (0.18–6.90)	0.907	3.89 (0.44–34.24)	0.221	3.19 (0.65–15.63)	0.153

BMI = body mass index; PBT = perioperative blood transfusion.

Discussion

Bladder cancer is the fifth most expensive cancer in terms of total medical care expenditure. Furthermore, it is expected that in 2020, bladder cancer will be the most expensive cancer, and lifetime costs will reach $5 billion. Approximately 15%–34% of these costs are associated with radical cystectomy and trimodality treatment [9 –11]. Post-operative infection is a common complication among patients undergoing radical cystectomy and represents a key contributor to healthcare expenditures. However, there is still a lack of evidence concerning the incidence and predictors of post-operative infection in patients with bladder cancer undergoing radical cystectomy.

Several studies assessed the risk of post-operative infection after radical cystectomy. However, most of these studies considered only a few factors and did not use exclusion criteria. In this context, the present study adds vital findings for the field of urology. First, we showed that the overall incidence of post-operative infection was 10.4% (14/134). This is considerably lower than that reported from other single and multi-center series of radical cystectomy. For example, Hemelrijck et al. [12] reported that UTI/septicemia was the most common in-hospital complication after radical cystectomy, with an incidence of 67.46% (3,465/5,136). Parker et al. [2] and Pariser et al. [1] reported post-operative infection rates of 24% and 30%–41%, respectively. One plausible explanation for the lower incidence of infection in our series are the exclusion criteria we applied. All patients with a documented infection at the time of radical cystectomy were excluded. Additionally, in our study only four patients had orthotopic neobladder reconstruction; this urinary diversion predisposed to peri-operative UTI [13]. In our study, a high percentage of patients had ileal conduit diversion (44.8%), which is considered the standard for urinary diversion after radical cystectomy for bladder cancer. It is recognized as being the most cost effective, clinically satisfactory, and reliable solution in terms of long-term outcome [14].

In spite of the low incidence of post-operative infection, the present study emphasizes the high rate of co-infection and mixed infection. This suggests that treatment of patients with post-operative infection is becoming more challenging. Many of these infections required multi-drug treatment, which can result in the development and spread of multi-drug resistance pathogens with enhanced mortality after radical cystectomy.

The most common isolated pathogens in our cohort study was Enterococcus, which was responsible for 50% of SSIs, 33.3% of UTIs, and 20% of BIs, respectively. Enterobacter was the most common pathogen isolated from the urine. Both pathogens were responsible for 71.4% of post-operative infections. Similar results were reported by Koie et al. [15], who found that Enterococcus was the bacteria isolated most commonly in SSI cases after radical cystectomy in Asian countries. Kim et al. [5] indicated that Enterococcus was the most commonly implicated pathogen associated with febrile UTI after radical cystectomy and ileal neobladder reconstruction during the first 30 days after surgery (18/33; 54.5%). The present results suggest that appropriate peri-operative antibiotic prophylaxis needs to target both Enterococcus and Enterobacter pathogens to prevent the occurrence of post-operative infection. Escherichia coli, the most common pathogen worldwide, was isolated in 21.4% of patients with post-operative infection in our study. A previous study showed that the role of Escherichia coli in post-radical cystectomy infection is increasing after 30 days of surgery [5]. Specific targeting of the antimicrobial prophylaxis has been studied also in several small series, in which antimicrobial agents and duration of prophylaxis were adjusted according to the result of a terminal ileum content swab taken at the time of reconstruction. It has been demonstrated that small bowel is colonized in approximately 50% of radical cystectomy patients, and more than 65% of isolated bacterial strains were resistant to antibiotic agents administered during routine prophylaxis. Interestingly the positive culture of intestinal swab had no influence on general number of post-operative complications, however, the wound infections were more frequent in patients with positive bowel swab culture. These promising findings need to interpreted with caution, given the small number of patients in the study group [16].

In the present study, obesity was the predictor associated with all types of post-operative infection. Hegde et al. [17] hypothesized a close inter-relationship between adipose tissue, inflammatory response, the immune system, and infections. The authors suggested that proliferating pre-adipocytes share the same embryonic origin with immune cells and have phagocytic activity. Therefore, in infections, adipose tissue can expand, similar to the expansion of cells of the immune system. For this reason, obese patients with impaired adipose tissue immune function may experience exacerbation of infections [17].

Adipose tissue also produces a variety of anti-inflammatory and pro-inflammatory factors, such as leptin, adiponectin, chemokines, and cytokines [18]. Leptin plays an important role in wound healing by regulation of monocyte/macrophage activation, activation of polymorphonuclear neutrophils, impacts on cytokine production, and proliferative and anti-apoptotic effects on T lymphocytes [19,20]. Patients with genetic defects in leptin, such as those observed in ob/ob mice, show a severe hereditary obese phenotype. Furthermore, these patients have reduced T-cell function and increased sensitivity to pro-inflammatory monocyte-macrophage–activating stimuli, as well as impairment of phagocytic functions [21]. Therefore, defects in leptin production can explain the association between obesity and high susceptibility to bacterial infections.

The incidence of SSI caused by obesity may be explained additionally by local changes, such as increased local tissue trauma related to retraction, prolonged operative time, and disturbance of total body homeostatic balance [22]. Also, in obese patients, oxygenation of subcutaneous tissue is reduced, increasing the risk of SSI [23].

Current evidence gathered using the Enhanced Recovery After Surgery (ERAS) protocol demonstrated that a period after pre-operative chemotherapy and before surgery should be utilized for nutritional optimization of patients and promoting physical activity that constitute important “pre-habilitation” part of the ERAS protocol aiming at hastening post-operative recovery. These measures may lead to a desired weight loss in obese patients (safe rate being approximately 1 kg per week) and decrease risk of the post-operative infection; however, this is a secondary benefit and is usually not strictly requested from cystectomy candidates [8].

In multivariable analysis, we found an age-related risk of infection after radical cystectomy. Numerous previous studies analyzed aging as a risk factor for infections. Most of these studies agree that immune function decreases in the elderly, and there is also consensus that immunosenescence is caused by functional alterations in cellular innate or adaptive immunity rather than a decrease in the number of immune cells. There is discrepancy between two scenarios involved in immunosenescence: a gradual decline in the function of all elements of the host defense system versus a decline in the function of limited parts of the system [24].

In our study, diabetes mellitus was associated with UTI and SSI after radical cystectomy. Similar results were reported in some previous studies. For example, in a meta-analysis performed by Martin et al. [25], the overall effect size for the association between diabetes and SSI with different types of surgery showed an odds ratio of 1.53 (95% CI 1.11–2.12). However, this study did not included urologic procedures [25]. The reason why diabetes mellitus influences SSI is unclear. It is possible that diabetes mellitus is a marker of other conditions, such as white blood cell dysfunction or vascular changes, which may put a patient at risk of infection. Higher risk of infection can also be explained by more pronounced peri-operative hyperglycemia in patients with poorly controlled diabetes mellitus, causing additional immune system dysfunction. Patients with and without diabetes mellitus can be affected by peri-operative hyperglycemia. The benefit of intensive peri-operative glycemic control is presented in both group of patients but is more evident in patients with diabetes mellitus [26,27].

The advantage of the laparoscopic approach in radical cystectomy has been proven in many aspects, such as reduced intra-operative blood loss, shorter length of hospital stay, reduced need for blood transfusion, shorter time to regular diet, and lower total costs [28,29]. Our results indicate that there is an additional benefit in reduced risk of post-operative infection. This can be explained by the minimal tissue injury and lower levels of glycemia after laparoscopic surgery [26]. Removal of all resected specimens, which can be inflamed and infected, in special extraction bags without direct contact with the surgical incision may additionally have an impact on reducing SSI [30].

In our study, we did not identify a correlation between peri-operative blood transfusion and infection. However, most previous studies indicated that peri-operative blood transfusion increases the risk of sepsis. This is explained mostly by the immunosuppressive effect [31]. This discrepancy can be explained by the small sample size of our study. Nevertheless, we proved that length of hospital stay significantly increases the risk of BI. One plausible explanation could be the necessity to maintain a central venous catheter and longer intravenous therapy.

In spite of these important findings, several limitations of our study warrant mentioning. First, our present study is restricted by constraints inherent to the retrospective nature of the data analysis. Additionally, it should be noted that because the present study was based on inpatient data, we only estimated the incidence of post-operative infection based on the records of patients admitted to the hospital. Therefore, the rate of infection could be underestimated, because many patients were also treated as outpatients. Despite limitations, we believe that taken together, our results will be helpful in the identification of patients at greatest risk of post-operative infection after radical cystectomy. The present findings may further improve pre-operative patient counseling and pre-operative decision-making for the prevention of post-operative infection, potentially resulting in better patient satisfaction and safety, as well as reduced healthcare expenditures.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Pariser

, Anderson

, Pearce

, et al. The effect of broader, directed antimicrobial prophylaxis including fungal coverage on perioperative infectious complications after radical cystectomy. Urol Oncol, 2016; 34:9–14.

Parker

, Tollefson

, Heins

, et al. Characterization of perioperative infection risk among patients undergoing radical cystectomy: Results from the national surgical quality improvement program. Urol Oncol, 2016; 34:13–19.

Schmid

, Chiang

, Sood

, et al. Causes of hospital readmissions after urologic cancer surgery. Urol Oncol, 2016; 34:1–11.

Parker

, Toussi

, Tollefson

, et al. Risk factors and microbial distribution of urinary tract infections following radical cystectomy. Urology, 2016; 94:96–101.

Kim

, Yoon

, et al. Febrile urinary tract infection after radical cystectomy and ileal neobladder in patients with bladder cancer. J Korean Med Sci, 2016; 31:1100–1104.

Meyer

, Rios Diaz

, Dalela

, et al. Wound dehiscence in a sample of 1 776 cystectomies: Identification of predictors and implications for outcomes. BJU Int, 2016; 117:95–101.

Sangar

, Ragavan

, Matanhelia

, et al. The economic consequences of prostate and bladder cancer in the UK. BJU Int, 2005; 95:59–63.

Arumainayagam

, McGrath

, Jefferson

, et al. Introduction of an enhanced recovery protocol for radical cystectomy. BJU Int, 2008; 101:698–701.

Hedelin

, Holmang

, Wiman

. The cost of bladder tumour treatment and follow-up. Scand J Urol Nephrol, 2002; 36:344–347.

10.

Botteman

, Pashos

, Redaelli

, et al. The health economics of bladder cancer: a comprehensive review of the published literature. Pharmacoeconomics, 2003; 21:1315–1330.

11.

Yeung

, Dinh

, Lee

. The health economics of bladder cancer: an updated review of the published literature. Pharmacoeconomics, 2014; 32:1093–1104.

12.

van Hemelrijck

, Thorstenson

, Smith

, et al. Risk of in-hospital complications after radical cystectomy for urinary bladder carcinoma: Population-based follow-up study of 7608 patients. BJU Int, 2013; 112:1113–1120.

13.

Shigemura

, Tanaka

, Matsumoto

, et al. Post-operative infection and prophylactic antibiotic administration after radical cystectomy with orthotopic neobladder urinary diversion. J Infect Chemother, 2012; 18:479–484.

14.

Colombo

, Naspro

. Ileal conduit as the standard for urinary diversion after radical cystectomy for bladder cancer. Eur Urol Suppl, 2010; 9:736–744.

15.

Koie

, Yamamoto

, Okamoto

, et al. [Risk factors for surgical site infection in radical cystectomy: Efficacy of buried suture by absorbable thread for skin closure]. Hinyokika Kiyo, 2009; 55:75–78.

16.

Mokrzyś

, Skrzypczyk

, Szempliński

, et al. A small intestine microbial flora assessment in patients undergoing radical cystectomy. Cent Eur J Urol, 2016;(Suppl 1):104–105.

17.

Hegde

, Dhurandhar

. Microbes and obesity—Interrelationship between infection, adipose tissue and the immune system. Clin Microbiol Infect, 2013; 19:314–320.

18.

Fantuzzi

, Faggioni

. Leptin in the regulation of immunity, inflammation, and hematopoiesis. J Leukoc Biol, 2000; 68:437–446.

19.

Zarkesh-Esfahani

, Pockley

, Wu

, et al. Leptin indirectly activates human neutrophils via induction of TNF-alpha. J Immunol, 2004; 172:1809–1814.

20.

Fantuzzi

, Faggioni

. Leptin in the regulation of immunity, inflammation, and hematopoiesis. J Leukoc Biol, 2000; 68:437–446.

21.

Falagas

, Kompoti

. Obesity and infection. Lancet Infect Dis, 2006; 6:438–446.

22.

Vilar-Compte

, Mohar

, Sandoval

, et al. Surgical site infections at the National Cancer Institute in Mexico: A case-control study. Am J Infect Control, 2000; 28:14–20.

23.

Canturk

, Canturk

, Cetinarslan

, et al. Nosocomial infections and obesity in surgical patients. Obes Res, 2003; 11:769–775.

24.

Gavazzi

, Krause

. Ageing and infection. Lancet Infect Dis, 2002; 2:659–666.

25.

Martin

, Kaye

, Knott

, et al. Diabetes and risk of surgical site infection: A systematic review and meta-analysis. Infect Control Hosp Epidemiol, 2016; 37:88–99.

26.

Russo

. Perioperative glycemic control. Anesthesiol Clin, 2012; 30:445–466.

27.

, Ren

, Li

, et al. Clinical significance of post-operative hyperglycemia in nondiabetic patients undergoing definitive surgery for gastrointestinal fistula. Surg Infect, 2016; 17:491–497.

28.

Zheng

, Li

, Song

, et al. Comparison of laparoscopic and open cystectomy for bladder cancer: A single center of 110 cases report. Transl Androl Urol, 2012; 1:4–8.

29.

Tang

, Li

, Xia

, et al. Laparoscopic versus open radical cystectomy in bladder cancer: A systematic review and meta-analysis of comparative studies. PLoS One, 2014; 9:95667.

30.

Khan

, Fayyad

, Cecil

, et al. Laparoscopic versus open appendectomy: the risk of postoperative infectious complications. JSLS, 2007; 11:363–367.

31.

Liu

, Mullane

, Kates

, et al. Infectious complications in transfused patients after radical cystectomy. Can J Urol, 2016; 23:8342–8347.