Abstract
Background
Epidural anesthesia has been associated with a decrease in cardiopulmonary complications and a decrease in blood loss in orthopedic procedures. Its influence on the outcomes of patients receiving radical cystectomies is unknown. We aim to use the large national database from the National Surgical Quality Improvement Project (NSQIP) to examine whether postoperative complications may be affected by the use of epidural anesthesia during radical cystectomy.
Methods
Data were collected from the 2014–2016 participant user files of the NSQIP database. Patients receiving radical cystectomy were identified by CPT code and further stratified by anesthesia type. Demographics, length of stay, and 30-day complications including death were collected and analyzed using univariable and multivariable analysis.
Results
A total of 6448 patients met the inclusion criteria for analysis. Between 2014 and 2016, 5064 patients received general anesthesia only (GA) and 1384 patients received general and epidural anesthesia (GEA). Statistical analysis showed an overall increase in major complications (17.8% vs 18.5%) in the GEA group (p = 0.0046). Subgroup analysis showed major complications to be more likely in patients older than 75 years receiving GEA instead of GA (p = 0.0301).
Conclusions
Elderly patients (age > 75) undergoing radical cystectomy may experience more major complications with the use of epidural anesthesia. This may be due to end-organ effects from the hemodynamic changes of epidural anesthesia which are poorly tolerated in the elderly population. Further single intervention epidural studies need to be performed to isolate the effects of epidural anesthesia on individual surgical procedures.
Keywords: Anesthesia, Urology, Outcomes
Introduction
Bladder cancer is the fourth most common cancer in men with an estimated 81,000 new cases in the United States in 2018. Radical cystectomy continues to be the treatment of choice for muscle-invasive disease. Despite advances in equipment and technique, the perioperative morbidity remains significant ranging up to 64% [1].
Perioperative techniques aimed at reducing postoperative complications have been recently introduced in the form of enhanced recovery protocols. Regional anesthesia continues to be a significant component of most enhanced recovery pathways. Epidural anesthesia has been routinely used for superior postoperative pain control in major abdominal and orthopedic surgeries. Further, epidural anesthesia has been associated with an improvement in short term outcomes such as cardiopulmonary complications, the return of bowel function, deep vein thrombosis, and intraoperative bleeding [2-4]. However, other studies have found no difference in short term and long term outcomes [5]. Most of the literature focuses on epidural use in colorectal and orthopedic procedures. There is a paucity of data regarding epidural use for patients receiving radical cystectomy. Due to the complex nature of the procedure, high morbidity rate, and need to operate on multiple organ systems, one cannot extrapolate and apply data from colorectal surgery to radical cystectomy.
Using a large national database, we aim to investigate whether epidural anesthesia can influence short term perioperative outcomes in patients receiving radical cystectomy. We hypothesize that patients will experience fewer complications due to the physiologic blunting of sympathetic response with epidural anesthesia. Identification of benefits in addition to pain control may allow the use of epidural anesthesia to attenuate complications in high-risk patients undergoing complex urologic surgeries.
Methods
The American College of Surgeons-National Surgical Quality Improvement Project (NSQIP) was used for data regarding patients receiving radical cystectomy. The NSQIP program is a nationally validated, risk-adjusted, outcomes-based program which is used to measure and improve the quality of care. Literature has shown that analysis of NSQIP data has contributed to reduction of complications and cost [6]. NSQIP currently collects information from over 680 institutions thereby providing data from over 1 million cases per year. For each case over 130 variables including demographics information, intraoperative and postoperative data are collected up to 30 days after surgery.
Using the participant user files from the NSQIP database, 3 years of data (2014–2016) were reviewed. Exemption from Institutional Review Board approval and informed consent were granted by the Protocol Review and Regulatory Affairs Committee due to use of only deidentified patient data. All patients with CPT codes between 51,570 and 51,597, and whose primary anesthesia was ‘general anesthesia’ were selected for analysis. The group was further divided into patients receiving general anesthesia alone and patients receiving general and epidural anesthesia. Descriptive statistics were used to summarize the continuous and categorical variables for the whole data set and stratified by type of anesthesia. Demographics and outcome variables of interest were compared between the two groups using either the Chi-square of Fischer’s Exact test for categorical variables and the non-parametric Kruskal–Wallis test for numeric variables. All tests were deemed significant at the 0.05 significance level. No corrections for multiple testing were employed. SAS version 9.4 software was used to analyze the data.
Outcomes of interest were complications and length of stay. Complication outcomes were further analyzed by a previously used method (Parker et al. [7]) which organized NSQIP complications into four groups: none, minor, major, and death. Minor complications included superficial incisional skin infection, pneumonia, pulmonary embolus, urinary tract infection (UTI), blood transfusion, deep vein thrombosis. Major complications included deep incisional skin infection, organ space wound infection, wound dehiscence, reintubation, ventilator, acute renal failure, cerebrovascular accident (CVA), cardiac arrest, myocardial infarction, and sepsis. McNemar’s trend test and multivariable ordinal regression were used to analyze outcomes related to complications. Length of stay (LOS) was defined as the time from the procedure to discharge. Data were examined for normality and was analyzed using log transformation. Simple t test and multiple regression analysis were used for the log length of stay data. Covariates included in the model were age, sex, American Society of Anesthesia (ASA) Physical Status score, preoperative hematocrit, diabetes, chronic obstructive pulmonary disease, ascites, congestive heart failure, hypertension, renal failure, dialysis, smoking, and body mass index.
Further analysis was done with sub-groups of age, examining complications in patients greater than and less than 75 years old. This age was chosen to ensure adequate sample size availability.
Results
Data involving 6,448 patients was reviewed and divided into a GA group consisting of 5064 patients and GEA group of 1384 patients. Demographics of both groups were reviewed and found to be similar (Table 1) except for presence of preoperative diabetes (19.3% vs 16.3%, p = 0.022), hypertension (59.6% vs 55.4%, p = 0.005), and renal failure (0.5% vs 0%, p = 0.003). This led to a significant difference in high ASA score among both groups (77% vs 72.9%, p = 0.002). Similar baseline characteristics included age (67.5 vs 68.1, p = 0.121), body mass index (BMI) (28.5 vs 28.4, p = 0.765), race (p = 0.05), male gender (75.7% vs 75.4%, p = 0.853), and smoking status (23.3% vs 23.4%, p = 0.95). Univariate analysis found epidural anesthesia to be associated with an increased rate of transfusion (37.2% vs 42.2%, p < 0.001), and increased length of stay (9.56 days vs 10.22 days, p < 0.001) (Table 2). After adjusting with covariates multivariable linear regression model no difference was found between GA and GEA for length of stay (Fig. 1), however, an association between GEA and increased rate of transfusion persisted (OR 1.44, CI 1.23–1.68, p < 0.001) (Table 3, Fig. 4).
Table 1.
Patient demographics and clinical characteristics, presented as mean (standard deviation)
| General anesthesia N = 5064 |
General and epidural anesthesia N = 1384 |
p value | |
|---|---|---|---|
| Age, mean (St dev) | 67.5 (11.1) | 68.1 (10.7) | 0.12 |
| BMI, mean (St dev) | 28.5 (5.9) | 28.4 (5.9) | 0.77 |
| Race white, N (%) | 4069 (91.5) | 826 (93.4) | 0.05 |
| Male gender, N (%) | 3833 (75.7) | 1044 (75.4) | 0.85 |
| ASA score, N (%) | |||
| 3, 4, 5 | 3887 (77) | 1008 (72.9) | 0.002 |
| 1, 2 | 1163 (23.0) | 375 (27.1) | |
| Current smoker within 1 year, N (%) | 1181 (23.3) | 324 (23.4) | 0.95 |
| Diabetes, N (%) | |||
| Insulin | 350 (6.9) | 72 (5.2) | 0.02 |
| Non-insulin | 631 (12.5) | 154 (11.1) | |
| COPD, N (%) | 369 (7.3) | 111 (8.0) | 0.357 |
| Ascites, N (%) | 3 (0.1) | 1 (0.1) | 1.0 |
| CHF, N (%) | 36 (0.7) | 6 (0.4) | 0.26 |
| HTN, N (%) | 3019 (59.6) | 767 (55.4) | 0.005 |
| Acute renal failure, N (%) | 26 (0.5) | 0 (0) | 0.003 |
| Dialysis, N (%) | 47 (0.9) | 10 (0.7) | 0.47 |
| Bleeding disorder, N (%) | 155 (3.1) | 33 (2.4) | 0.19 |
| SIRS | 87 (1.7) | 15(1.1) | 0.23 |
| Sepsis, N (%) | 14 (0.3) | 3 (0.2) | |
| Transfusion before surgery, N (%) | 87 (1.7) | 25 (1.8) | 0.82 |
| Preop hematocrit, N (%) | |||
| ≤ 40 | 3208 (64.6) | 900 (66.5) | 0.198 |
| > 40 | 1755 (35.4) | 453 (33.5) |
BMI body mass index, COPD chronic obstructive pulmonary disease, CHF congestive heart failure, HTN hypertension, SIRS systemic inflammatory response syndrome
Bold values indicate statistical significance
Table 2.
Univariate analysis, presented as mean (standard deviation)
| General anesthesia N = 5064 |
General and epidural anesthesia N = 1384 |
p value | |
|---|---|---|---|
| Length of stay, mean (St dev) | 9.56 (7.7) | 10.22 (7.5) | < 0.001 |
| Wound occurrences, N (%) | |||
| 0 | 4770 (94.2) | 1296 (93.6) | 0.14 |
| 1 | 293 (5.8) | 86 (6.2) | |
| 2 | 1 (0) | 2 (0.1) | |
| Deep incisional SSI | |||
| 0 | 4992 (98.6) | 1362 (98.4) | 0.64 |
| 1 | 72 (1.4) | 22 (1.6) | |
| Organ/space SSI | |||
| 0 | 4715 (93.1) | 1277 (92.3) | 0.13 |
| 1 | 346 (6.8) | 104 (7.5) | |
| 2 | 3 (0.1) | 3 (0.2) | |
| Wound disruptions | |||
| 0 | 4942 (97.6) | 1341 (97) | 0.23 |
| 1 | 116 (2.3) | 42 (3) | |
| 2 | 6 (0.1) | 1 (0.1) | |
| Pneumonia occ | |||
| 0 | 4928 (97.3) | 1342 (97) | 0.49 |
| 1 | 134 (2.65) | 41 (3) | |
| 2 | 2 (0) | 1 (0.1) | |
| Unplanned intubations | |||
| 0 | 4935 (97.5) | 1353 (97.8) | 0.77 |
| 1 | 113 (2.2) | 29 (2.1) | |
| 2 | 15 (0.3) | 2 (0.1) | |
| 3 | 1 (0) | 0 (0) | |
| Pulmonary embolism | |||
| 0 | 4983 (98.4) | 1360 (98.3) | 0.73 |
| 1 | 81 (1.6) | 24 (1.7) | |
| Ventilator > 48 h | |||
| 0 | 4965 (98.1) | 1370 (99) | 0.05 |
| 1 | 97 (1.9) | 14 (1.0) | |
| 2 | 2 (0) | 0 (0) | |
| Progressive renal insufficiency | |||
| 0 | 4956 (97.9) | 1361 (98.3) | 0.27 |
| 1 | 108 (2.1) | 23 (1.7) | |
| Acute renal failure | |||
| 0 | 4992 (98.6) | 1365 (98.6) | 1.0 |
| 1 | 71 (1.4) | 19 (1.4) | |
| 2 | 1 (0) | 0 (0) | |
| Urinary tract infection | |||
| 0 | 4630 (91.4) | 1279 (92.4) | 0.20 |
| 1 | 431 (8.5) | 103 (7.4) | |
| 2 | 3 (0.1) | 2 (0.1) | |
| Stroke/CVA | |||
| 0 | 5032 (99.4) | 1376 (99.4) | 1.0 |
| 1 | 31 (0.6) | 8 (0.6) | |
| 2 | 1 (0) | 0 (0) | |
| Cardiac arrest requiring CPR | |||
| 0 | 5018 (99.1) | 1365 (98.6) | 0.35 |
| 1 | 43 (0.9) | 18 (1.3) | |
| 2 | 2 (0) | 1 (0.1) | |
| 3 | 1 (0) | 0 (0) | |
| Myocardial infarction | |||
| 0 | 4994 (98.6) | 1364 (98.6) | 0.57 |
| 1 | 69 (1.4) | 19 (1.4) | |
| 2 | 1 (0) | 1 (0.1) | |
| Bleeding transfusion | |||
| 0 | 3178 (62.8) | 800 (57.8) | < 0.001 |
| 1 | 1885 (37.2) | 583 (42.1) | |
| 2 | 1 (0) | 1 (0.1) | |
| DVT/thrombophlebitis | |||
| 0 | 4926 (97.3) | 1349 (97.5) | 0.88 |
| 1 | 135 (2.7) | 34 (2.5) | |
| 2 | 3 (0.1) | 1 (0.1) | |
| Sepsis | |||
| 0 | 4639 (91.6) | 1263 (91.3) | 0.79 |
| 1 | 423 (8.4) | 121 (8.7) | |
| 2 | 2 (0) | 0 (0) | |
| Septic shock | |||
| 0 | 4916 (97.1) | 1351 (97.6) | 0.60 |
| 1 | 146 (2.9) | 33 (2.4) | |
| 2 | 2 (0) | 0 (0) | |
Outcomes presented as number of occurrences
SSI surgical site infection, CVA cerebrovascular accident, CPR cardiopulmonary resuscitation, DVT deep venous thrombosis
Bold values indicate statistical significance
Fig. 1.
Length of stay relationships by demographics and risk factors. T-scores plotted
Table 3.
Multivariable ordinal regression model examining the association between preoperative variables and postoperative outcomes
| All complications |
Blood transfusion |
|||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p value | OR | 95% CI | p value | |
| Epidural + GAA | 1.29 | 1.13–1.48 | 0.0002 | 1.44 | 1.23–1.68 | < 0.0001 |
| Gender female | 1.42 | 1.26–1.60 | < 0.0001 | 2.03 | 1.77–2.33 | < 0.0001 |
| Age | 1.00 | 0.99–1.01 | 0.109 | 1.00 | 1.00–1.01 | 0.079 |
| Race white | 0.75 | 0.62–0.91 | 0.003 | 0.70 | 0.56–0.87 | 0.001 |
| BMI | 1.02 | 1.01–1.03 | 0.0002 | 0.99 | 0.99–1.01 | 0.469 |
| ASA (high)a | 1.27 | 1.12–1.45 | 0.0003 | 1.25 | 1.07–1.45 | 0.004 |
| Smoking | 0.95 | 0.84–1.08 | 0.444 | 0.80 | 0.69–0.92 | 0.003 |
| Diabetes | ||||||
| Insulin | 1.61 | 1.30–1.98 | 0.0002 | 0.98 | 0.77–1.25 | 0.556 |
| Non-insulin | 1.14 | 0.97–1.35 | 0.270 | 1.12 | 0.93–1.35 | 0.253 |
| COPD | 1.34 | 1.10–1.62 | 0.003 | 1.12 | 0.90–1.40 | 0.321 |
| Ascites | 1.63 | 0.20–13.19 | 0.646 | – | – | – |
| CHF | 1.62 | 0.90–2.94 | 0.108 | 1.76 | 0.88–3.51 | 0.109 |
| HTN | 1.02 | 0.91–1.14 | 0.764 | 1.01 | 0.88–1.15 | 0.929 |
| Renal failure | 1.14 | 0.51–2.53 | 0.749 | 1.11 | 0.44–2.82 | 0.823 |
| Dialysis | 1.05 | 0.61–1.80 | 0.860 | 1.47 | 0.79–2. 72 | 0.219 |
| Bleeding disorder | 1.31 | 0.97–1.77 | 0.078 | 1.26 | 0.89–1.78 | 0.186 |
| Sepsis SIRS | 1.64 | 1.10–2.43 | 0.550 | 2.29 | 1.42–3.70 | 0.895 |
| Sepsis | 4.01 | 1.37–11.78 | 0.040 | 4.65 | 0.98–22.08 | 0.162 |
| Preop transfusion | 1.63 | 1.09–2.43 | 0.016 | 2.84 | 1.70–4.72 | < 0.0001 |
| Preop hematocrit > 40 | 0.54 | 0.48–0.60 | < 0.0001 | 0.30 | 0.26–0.34 | < 0.0001 |
High ASA score = 3, 4, 5
Bold values indicate statistical significance
Fig. 4.
Schematic of odds ratios for blood transfusion, urinary tract infection, and all listed complications when epidural anesthesia is used. Bold line represents odd ratio 0.5
Multivariable analysis found the GEA group to sustain a higher overall rate of complications (Table 3) including major and minor complications (52% vs 58.5%, p = 0.0046, Fig. 2). No significant change was found in 30 day mortality (1.7% vs 1.6%, p = 0.88). Multivariable ordinal regression model showed epidural anesthesia, higher ASA status, female gender, BMI, chronic obstructive pulmonary disease (COPD), and insulin-dependent diabetes to have an association with a higher incidence of overall complications. Patients of the white race and hematocrit > 40 g/dl had a decreased incidence of complications (Fig. 3). Further multivariable logistic regression models were built to investigate if there was a single complication within each major and minor complication group, which led to this association. The only lone complication found to have a statistically significant association with GEA was blood transfusion (Table 3, Fig. 4). Increased rate of blood transfusion was also associated with female gender, ASA score 3–5, and history of preoperative transfusion.
Fig. 2.
Postoperative complications organized by grade and type of anesthesia, (one-sided Cochran–Armitage Trend test)
Fig. 3.
Odds ratios for preoperative risk factors for postoperative complications after cystectomy. Odds ratio of 1 highlighted for emphasis
Subgroup analysis of 4846 patients with age > 75 and 1602 patients with age ≤ 75 showed an association between GEA and increase in the likelihood of complications (Table 4). Patients with age > 75 receiving GEA demonstrated an increase in major complications (19.6% vs 15.4%, p = 0.0301). Conversely, in patients ≤ 75 receiving GEA there was a decrease in major complications (18.5% vs 18.1%, p = 0.019) (Table 5, Fig. 4).
Table 4.
Multivariable ordinal regression model examining the association between preoperative variables and postoperative outcomes in age > 75 and ≤ 75
| Age > 75 |
Age ≤ 75 |
|||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p value | OR | 95% CI | p value | |
| Epidural + GAA | 1.32 | 1.00–1.75 | 0.053 | 1.29 | 1.10–1.51 | 0.002 |
| Gender female | 1.27 | 0.98–1.63 | 0.068 | 1.47 | 1.28–1.69 | < 0.0001 |
| Age | 1.01 | 0.98–1.04 | 0.407 | 1.00 | 1.00–1.01 | 0.375 |
| Race white | 0.68 | 0.44–1.06 | 0.086 | 0.76 | 0.62–0.94 | 0.010 |
| BMI | 1.02 | 1.00–1.04 | 0.120 | 1.02 | 1.01–1.03 | 0.009 |
| ASA (high)a | 1.01 | 0.76–1.36 | 0.934 | 1.34 | 1.16–1.55 | < 0.0001 |
| Smoking | 1.24 | 0.84–1.82 | 0.272 | 0.92 | 0.81–1.06 | 0.256 |
| Diabetes | ||||||
| Insulin | 1.33 | 0.87–2.04 | 0.376 | 1.73 | 1.36–2.20 | < 0.0001 |
| Non-insulin | 1.20 | 0.87–1.65 | 0.840 | 1.12 | 0.92–1.36 | 0.128 |
| COPD | 1.72 | 1.20–2.48 | 0.004 | 1.23 | 0.98–1.55 | 0.069 |
| Ascites | na | na | na | 1.70 | 0.21–13.78 | 0.619 |
| CHF | 2.55 | 0.84–7.69 | 0.097 | 1.44 | 0.71–2.92 | 0.311 |
| HTN | 0.92 | 0.72–1.17 | 0.482 | 1.06 | 0.93–1.21 | 0.398 |
| Renal failure | 0.52 | 0.10–2.53 | 0.415 | 1.42 | 0.56–3.63 | 0.465 |
| Dialysis | 1.42 | 0.43–4.70 | 0.571 | 0.98 | 0.53–1.79 | 0.945 |
| Bleeding disorder | 1.41 | 0.78–2.53 | 0.251 | 1.27 | 0.89–1.80 | 0.188 |
| Sepsis SIRS | 2.06 | 1.03–4.10 | 0.913 | 1.44 | 0.89–2.34 | 0.518 |
| Sepsis | 5.00 | 0.36–69.55 | 0.355 | 3.42 | 1.05–11.21 | 0.088 |
| Preop transfusion | 1.48 | 0.65–3.36 | 0.349 | 1.64 | 1.03–2.59 | 0.036 |
| Preop hematocrit > 40 | 0.40 | 0.31–0.51 | < 0.0001 | 0.59 | 0.52–0.67 | < 0.0001 |
High ASA score = 3, 4, 5
Bold values indicate statistical significance
Table 5.
Grade of complications by type of anesthesia and age
| Age > 75 |
Age ≤ 75 |
|||
|---|---|---|---|---|
| GAA (%) | Epidural + GAA (%) | GAA (%) | Epidural + GAA (%) | |
| None | 535 (43.7) | 141 (37.3) | 1812 (47.2) | 412 (41.0) |
| Minor | 456 (37.2) | 152 (40.2) | 1275 (33.2) | 401 (39.9) |
| Major | 189 (15.4) | 74 (19.6) | 710 (18.5) | 182 (18.1) |
| Death | 44 (3.6) | 11(2.9) | 43 (1.1) | 11 (1.1) |
| p value | 0.0301 | 0.019 | ||
p value = one-sided Cochran–Armitage Trend test
Bold values indicate statistical significance
Multivariable Cox Proportional Hazards regression model found no association between epidural anesthesia and death within 30 days (Table 6). The analysis did show that patients with increased age and history of COPD did sustain an increased 30 day mortality.
Table 6.
Multivariable Cox proportional hazards regression model (outcome = 30 day mortality)
| HR | 95% CI | p value | |
|---|---|---|---|
| Epidural + GAA | 1.04 | 0.59–1.83 | 0.883 |
| Gender female | 0.66 | 0.38–1.16 | 0.147 |
| Age | 1.05 | 1.02–1.08 | 0.0003 |
| Race white | 1.35 | 0.61–2.99 | 0.464 |
| BMI | 0.99 | 0.95–1.03 | 0.468 |
| Smoking | 0.95 | 0.53–1.70 | 0.867 |
| Diabetes insulin | 1.64 | 0.81–3.36 | 0.171 |
| Non-insulin | 0.98 | 0.51–1.88 | 0.941 |
| COPD | 1.96 | 1.11–3.46 | 0.020 |
| Ascites | – | – | – |
| CHF | 2.70 | 0.81–9.07 | 0.107 |
| HTN | 1.10 | 0.67–1.81 | 0.692 |
| Renal failure | 2.65 | 0.51–13.66 | 0.245 |
| Dialysis | 1.20 | 0.15–9.69 | 0.864 |
| bleeding disorder | 1.48 | 0.61–3.61 | 0.388 |
| Sepsis SIRS | 1.68 | 0.66–4.24 | 0.275 |
| Sepsis | 0.85 | 0.08–8.64 | 0.893 |
| Transfusion | 1.72 | 0.62–4.79 | 0.297 |
| Preop hematocrit > 40 | 1.11 | 0.68–1.81 | 0.677 |
| ASA (high)a | 0.75 | 0.42–1.36 | 0.344 |
High ASA score = 3, 4, 5
Bold values indicate statistical significance
Discussion
Current enhanced recovery after surgery (ERAS) guidelines recommend epidural anesthesia for patients undergoing radical cystectomy [8]. These recommendations are based on better pain control [9], improved functional outcomes [10], and decreased cardiopulmonary complications [11]. However, the most signification limitation to these guidelines is that no prospective single-intervention study has yet to be performed evaluating epidural use in radical cystectomy. Our study has isolated epidural anesthesia and found that it may be associated with an increase in complications in certain age groups, unlike what has been found in colorectal surgery. This leads to a significant point that colorectal surgery evidence may not be able to be extrapolated and applied to the radical cystectomy population. While both are considered high-risk abdominal procedures, radical cystectomy can involve surgery on many organ systems and longer operative time leading to a larger systemic inflammatory response, increased extracellular fluid shifts, and longer period of postoperative resuscitation. Increased complications with epidural anesthesia in radical cystectomy may be due to the combination of extracellular fluid shift [12] and arterial hypotension from epidural anesthesia [13].
The data shows overall major and minor complications to be increased with the GEA group, with an increase in major complications in patients > 75 years old. Of all complications, the only individual complication found to have a statistically significant association with GEA was a blood transfusion. While the increase in transfusion may account for the increase in minor complications, it is unclear which complications drive the increase in major complications. Even though none of the major complications showed a statistically significant increase in the GEA group, there was an increase in CVA, cardiac arrest requiring cardiopulmonary resuscitation, myocardial infarction, and sepsis. We can only hypothesize that unrecognized prolonged hypotension with epidural anesthesia may lead to major complications such as myocardial infarction, CVA, and cardiac arrest. We further hypothesize that the elderly population are more vulnerable to hypotension which is why they may be sustaining an increased rate of major complications.
The association between epidural anesthesia and transfusion is still controversial. Some studies have found patients with epidurals receive fewer blood transfusions due to decreased blood loss [2, 14], while other studies have found transfusion to be increased due to increased fluid administration leading to anemia [15]. Due to unavailability of blood loss data in the NSQIP data file, we are unable to provide a complete explanation for the increased rate of blood transfusion seen in the GEA group.
It is unknown why there is such a large discrepancy in the literature regarding outcomes with the use of epidural anesthesia. It might be due to the heterogeneity of the surgical population used in each study. Recently, a large meta-analysis made up of over 1 million patient records compared GEA to GA [16]. They found no difference in 30 day mortality, however, the GEA group did show a decrease in pulmonary complications, surgical site infection, blood transfusion and LOS. There was also an increase in myocardial infarction in the GEA group. The 27 studies evaluated included primarily orthopedic procedures (18/27) making results difficult to apply to non-orthopedic procedures.
Another reason for the varied results includes different anatomical placement and medications used in epidurals for abdominal and orthopedic procedures. While orthopedic surgeries typically involve placement of lumbar epidurals, abdominal surgery can involve low to high thoracic epidural placement. Thoracic epidural placement may have a more pronounced hemodynamic effect leading to increased hypotension and possibly increased complications. The epidural medications used in each study may also produce different results as local anesthesia can cause more hypotension than opioids.
Additionally, stratifying patients based on oncologic diagnosis may lead to the difference in major and minor complications seen in patients with epidurals. Oncologic patients may have a different baseline inflammatory state leading to increased risk of deep vein thrombosis and pancytopenia. Chemotherapy can induce vascular changes such as altered vascular tone and impaired endothelial function [17]. These changes can lead patients to increased risk of major and minor complications.
Finally, Memtsoudis et al. [18] has attributed the discrepancy in outcomes to underpowering of studies and not making a distinction between the use of epidural for intraoperative anesthesia versus postoperative analgesia.
Significant limitations of this study include the restrictions inherent to the NSQIP database. Unfortunately, these limitations don’t provide for the ability to grade complications using a standardized system such as the Clavien Dindo Classification system. In addition, variables such as surgical factors (type of diversion, blood loss, tumor stage, preoperative chemotherapy) and anesthetic factors (level of epidural blockade, epidural medication used, intraoperative hemodynamics) may have been helpful in explaining the trends seen in this study.
We continue to encourage epidural use for postoperative analgesia however recognize complications may be increased in the elderly population. Each procedure and patient should be studied individually before drawing conclusions on the benefits of epidural anesthesia. Further work is needed in examining single intervention epidural use in individual surgical populations.
Footnotes
Financial disclosures None.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Conflict of interest The authors declare that they have no conflict of interest.
Research involving human participats Exemption from Institutional Review Board approval.
Informed consent Informed consent was granted by the Protocol Review and Regulatory Affairs Committee due to use of only deidentified patient data.
References
- 1.Shabsigh A, Korets R, Vora KC et al. (2009) Defining early morbidity of radical cystectomy for patients with bladder cancer using a standardized reporting methodology. Eur Urol 55:164–174 [DOI] [PubMed] [Google Scholar]
- 2.Ozyuvaci E, Altan A, Karadeniz T, Topsakal M, Besisik A, Yucel M (2005) General anesthesia versus epidural and general anesthesia in radical cystectomy. Urol Int 74:62–67 [DOI] [PubMed] [Google Scholar]
- 3.Popping DM, Elia N, Van Aken HK et al. (2014) Impact of epidural analgesia on mortality and morbidity after surgery: systematic review and meta-analysis of randomized controlled trials. Ann Surg 259:1056–1067 [DOI] [PubMed] [Google Scholar]
- 4.Memtsoudis S, Sun X, Chiu Y et al. (2013) Perioperative comparative effectiveness of anesthetic technique in orthopedic patients. Anesthesiology 118:1046–1058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Christopher Doiron R, Jaeger M, Booth CM, Wei X, Robert SD (2016) Is there a measureable association of epidural use at cystectomy and postoperative outcomes? A population-based study. Can Urol Assoc J 10:321–327 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Govaert JA, van Bommel AC, van Dikj WA et al. (2015) Reducing healthcare costs facilitated by surgical auditing: a systematic review. World J Surg 39:1672–1680 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Parker DC, Handorff E, Smaldone MC, Uzzo RG, Pitt H, Reese AC (2017) Race and postoperative complications following urologic cancer surgery: an ACS-NSQIP analysis. Urol Oncol 35(670):e1–670.e6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cerantola Y, Valerio M, Persson B et al. (2013) Guidelines for perioperative care after radical cystectomy for bladder cancer: enhanced recovery after surgery society recommendations. Clin Nutr 32:879–887 [DOI] [PubMed] [Google Scholar]
- 9.Toren P, Ladak S, Ma C, McCluskey S, Fleshner N (2009) Comparison of epidural and intravenous patient controlled analgesia in patients undergoing radical cystectomy. Can J Urol 4:4716–4720 [PubMed] [Google Scholar]
- 10.Maffezzini M, Campodonico F, Capponi G, Manuputty E, Gerbi G (2012) Fast track surgery and technical nuances to reduce complications after radical cystectomy and intestinal urinary diversion with the modified Indiana pouch. Surg Oncol 21:191–195 [DOI] [PubMed] [Google Scholar]
- 11.Nygren J, Thacker J, Carli F et al. (2013) Guidelines for perioperative care in elective rectal/pelvic surgery: enhanced recovery after surgery (ERAS) society recommendations. World J Surg 37:285–305 [DOI] [PubMed] [Google Scholar]
- 12.Kayilioglu S, Dinc T, Sozen I, Bostanoglu A, Cete M, Coskun F (2015) Postoperative fluid management. World J Crit Care Med 4:192–201 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Holte K, Foss N, Svensen C, Lund C et al. (2004) Epidural anesthesia, hypotension, and changes in intravascular volume. Anesthesiology 100:281–286 [DOI] [PubMed] [Google Scholar]
- 14.Mazul-Sunko B, Gilja I, Jelisavac M et al. (2014) Thoracic epidural analgesia for radical cystectomy improves bowel function even in traditional perioperative care: a retrospective study in eighty-five patients. Acta Clin Croat 53:319–325 [PubMed] [Google Scholar]
- 15.Tzimas P, Prout J, Papadopoulos G, Mallett SV (2013) Epidural anaesthesia and analgesia for liver resection. Anaesthesia 68:628–635 [DOI] [PubMed] [Google Scholar]
- 16.Smith LM, Coxowicz C, Uda Y, Memtsoudis SG, Barrington MJ (2017) Neuraxial and combined neuraxial/general anesthesia compared to general anesthesia for major truncal and lower limb surgery: a systematic review and meta-analysis. Anesth Analg 125:1931–1945 [DOI] [PubMed] [Google Scholar]
- 17.Cameron AC, Touyz RM, Lang NN (2016) Vascular complications of cancer chemotherapy. Can J Cardiol 32:852–862 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Memtsoudis SG, Liu SS (2014) Do neuraxial techniques affect perioperative outcomes. The story of vantage points and number games. Anesth Analg 119:501–502 [DOI] [PubMed] [Google Scholar]