bs_bs_banner
International Journal of Urology (2014)
doi: 10.1111/iju.12563
Original Article
Chronic kidney disease and perioperative outcomes in urological oncological surgery Marianne Schmid,1,2 Praful Ravi,3 Abd-El-Rahman M Abd-El-Barr,4 Julia Klap,2 Jesse D Sammon,4 Steven L Chang,2 Mani Menon,4 Adam S Kibel,2 Margit Fisch1 and Quoc-Dien Trinh2 1 Department of Urology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 2Division of Urologic Surgery, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; 3Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK; and 4Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Health System, Detroit, Michigan, USA
Abbreviations & Acronyms ASA = American Society of Anesthesiologists BMI = body mass index BUN = blood urea nitrogen CI = confidence interval CKD = chronic kidney disease CPT = Current Procedural Terminology eGFR = estimated glomerular filtration rate LOS = length of stay OR = odds ratio OT = operative time NSQIP = National Surgical Quality Improvement Program PN = partial nephrectomy pLOS = prolonged length of stay pOT = prolonged operation time RC = radical cystectomy RN = radical nephrectomy RP = radical prostatectomy SCr = serum creatinine SD = standard deviation Correspondence: Marianne Schmid M.D., Center for Surgery and Public Health and Division of Urologic Surgery, Brigham and Women’s Hospital, Harvard Medical School, 45 Francis Street, Boston, MA 02115, USA. Email: [email protected] Received 4 March 2014; accepted 15 June 2014.
© 2014 The Japanese Urological Association
Objectives: To evaluate baseline renal dysfunction among patients undergoing urological oncological surgery and its impact on early postoperative outcomes. Methods: Between 2005 and 2011, patients who underwent minimally-invasive or open radical prostatectomy, partial nephrectomy and radical nephrectomy, or open radical cystectomy, respectively, were identified in the National Surgical Quality Improvement Program dataset. Preoperative kidney function was assessed using estimated glomerular filtration rate and staged according to National Kidney Foundation definitions. Multivariable logistic regression was used to model the association between preoperative renal function and the risk of 30-day mortality and major complications. Furthermore the impact of chronic kidney disease on operation time and length of hospital stay was assessed. Results: Overall, 13 168 patients underwent radical prostatectomy (65.4%), partial nephrectomy (10.7%) and radical nephrectomy (16.1%) and radical cystectomy (7.8%), respectively; 50.1% of evaluable patients had reduced kidney function (chronic kidney disease II), and a further 12.6, 0.7 and 0.9% were respectively classified into chronic kidney disease stages III, IV, and V. Chronic kidney disease was an independent predictor of 30-day major postoperative complications (chronic kidney disease III: odds ratio 1.61, P < 0.001; chronic kidney disease IV: odds ratio 2.24, P = 0.01), of transfusions (chronic kidney disease III: odds ratio 2.14, P < 0001), of prolonged length of stay (chronic kidney disease III: odds ratio 2.61, P < 0.001; chronic kidney disease IV: odds ratio 3.37, P < 0.001; and chronic kidney disease V: odds ratio 1.68; P = 0.03) and of 30-day mortality (chronic kidney disease III: odds ratio 4.15, P = 0.01; chronic kidney disease IV: odds ratio 10.10, P = 0.003; and chronic kidney disease V: odds ratio 17.07, P < 0.001) compared with patients with no kidney disease. Conclusions: Renal dysfunction might be underrecognized in patients undergoing urological cancer surgery. Chronic kidney disease stages III, IV and V are independent predictors for poor 30-day postoperative outcomes.
Key words: chronic kidney disease, perioperative outcome, radical and partial nephrectomy, radical cystectomy, radical prostatectomy.
Introduction CKD, defined as an eGFR of less than 60 mL/min/1.73 m2,1 is common in the adult USA population with a rising prevalence from 10.0 (1988–1994) to 13.1% (1999–2004).2 Patients undergoing urological oncology procedures, such as RN and PN, RC or RP represent a unique population at risk of acute and chronic renal dysfunction.3–5 The incidence of urological malignancies increases with age, and therefore these patients more commonly present with comorbidities, such as hypertension, diabetes mellitus and vascular disease, which are known risk factors for the development of CKD.6,7 Furthermore, up to 30% of elderly urological patients have CKD, even in the presence of a normal SCr.4 Therefore, adequate assessment and preservation of renal function is critical in the urological oncological patient, as patients undergoing RC, RN and PN, in particular, face a further insult to renal function as a result of surgery. CKD is independently associated with an increased risk of cardiovascular events, hospitalization and mortality,8,9 and has been reported as a predictor of poor early and long-term postoperative outcomes in patients undergoing cardiac, vascular, general and orthopedic surgery.9–12 To our knowledge, no study has investigated the same in the setting of urological 1
M SCHMID ET AL.
cancer surgery; CKD has mostly been measured as an outcome after urological cancer surgery, and several studies, relying primarily on renal cell carcinoma cohorts, report on the incidence of CKD in patients before urological surgery and the postoperative progression of renal dysfunction.10–12 A recent population-based analysis showed an increased risk of adverse perioperative outcomes after nephrectomy in patients with endstage renal disease,13 whereas institutional series have found similar results in patients undergoing robotic PN.14,15 Based on the previous findings, it is important to assess the impact of CKD on perioperative outcomes after major urological cancer surgery, given that they carry a significant morbidity and a risk of baseline renal function impairment.16,17 We therefore sought to specifically examine the impact of CKD on perioperative outcomes after uro-oncological surgery for prostate, renal and bladder cancer, given the lack of evidence in this area. We utilized the NSQIP database, which is an American College of Surgeons initiative that supports the collection of risk-adjusted data to facilitate the assessment of surgical outcomes and complications.18 We hypothesized that CKD would be an independent predictor of adverse perioperative outcomes after urological cancer surgery, given the findings of a recent meta-analysis that relied predominantly on studies in cardiac and vascular surgery.17
Methods Data source The current study relies on the American College of Surgeons NSQIP data. The NSQIP was developed to assess the quality of surgical care, and collects perioperative data on 135 variables including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity for patients undergoing major surgical procedures in both the inpatient and outpatient setting across the USA. Trained Surgical Clinical Reviewers prospectively collect the NSQIP data. Validated data from patients’ medical charts allows quantification of 30-day, risk-adjusted surgical outcomes, including post-discharge, when nearly 50% of complications occur. In 2011, the NSQIP included data from 315 participating sites, and more than 1.7 million cases were contributed until 2011. Specifically, NSQIP further tracks preoperative patients’ characteristics, such as SCr values, BMI and history of comorbidities, which lends itself well for estimation of preoperative CKD.19 Though limited by the lack of certain data on surgical, pathology and imaging characteristics, as well as urine analytics, this database is specifically designed to accurately capture perioperative complications.20 NSQIP therefore served as an appropriate data source to investigate the impact of CKD on urological cancer surgery outcomes.
Study population Patients undergoing open or minimally-invasive (including laparoscopic, as well as robotic assistance) RP (CPT codes: 55810, 55812, 55815, 55840, 55842, 55845, 55866), open or minimally-invasive RN (CPT codes: 50220, 50225, 50230, 50545, 50546) or open or minimally-invasive PN (CPT codes: 50240, 50542, 50543) or open RC (CPT codes: 51575, 51580, 51585, 51590, 51595, 51596, 51597; there are no CPT codes 2
for laparoscopic/robotic RC available up-to date), with concomitant International Classification of Diseases 9th edition (code of cancer diagnosis of the prostate [185], kidney [189.0] or bladder [188.x]), were identified in the NSQIP dataset between 2005 and 2011. Preoperative eGFR was calculated using the simplified Modification of Diet in Renal Disease formula,21 a four-variable equation consisting of age, sex, ethnicity and SCr levels. Kidney function was staged according to National Kidney Foundation definitions.1 Patients with an eGFR ≥90 mL/min/ 1.73 m2 were considered as having no (or slightly diminished) kidney dysfunction, and patients with an eGFR between 60 and 89 mL/min/1.73 m2 were regarded as having mild reduction of kidney function (CKD stage II).22 Accordingly, patients with a moderate eGFR of 30–59 mL/min/1.73 m2 were classified into CKD stage III, patients with severe reduction of eGFR (15– 29 mL/min/1.73 m2) into CKD stage IV and with an eGFR 3 mg/dL and increasing azotemia [increase in blood urea nitrogen]) within 24 h before surgery were excluded from further analyses (n = 29, 0.2%). With these exclusions, 13 168 cases were left for analysis.
Covariates For each patient, age at surgery, sex, BMI, race, smoking status, alcohol consumption, preoperative SCr, blood urea nitrogen, hematocrit and type of surgery were available. Preoperative ASA score, as well as previous history of hypertension and diabetes mellitus, pulmonary, cardiac, liver, cerebrovascular and peripheral vascular disease, was reported.
Short-term outcomes Postoperative complications were classified under the following categories: cardiovascular (including postoperative cardiac arrest, myocardial infarction or cerebrovascular accident), pulmonary (including pneumonia, need for postoperative reintubation and need for ventilatory support >48 h), thromboembolic (including deep venous thrombosis and pulmonary embolism), septic (including sepsis and septic shock), renal (including acute renal failure and progressive renal insufficiency), urinary tract infections, wound complications (including superficial, deep and organ space surgical site infections, and wound dehiscence) and bleeding complications, according to previously reported methodology.23 Overall complications were defined as occurrence of any complication. Additional outcomes reported included the need for intraoperative transfusion, pOT, pLOS, readmission and perioperative mortality. pOT and pLOS were defined as an operative time or length of stay greater than the 75th percentile © 2014 The Japanese Urological Association
Renal function and urological surgery
(≥253 min and ≥4 days respectively). The 30-day readmission and reoperation data were only available starting 1 January 2011; therefore only 1 year of data (n = 6791) was analyzed for this outcome. Perioperative mortality was defined as death within 30 days of surgery.
Statistical analyses Descriptive statistics of categorical variables focused on frequencies and proportions. Means and standard deviations, medians, and interquartile ranges were reported for continuously coded variables. Multivariable logistic regression was used to model the association between preoperative renal function and overall complications, pOT, need of postoperative transfusion, pLOS, and mortality. Covariates consisted of age at surgery, race, sex, BMI, smoking status, baseline comorbidities and type of surgery. All statistical analyses were carried out using the R statistical package (R Foundation for Statistical Computing, Vienna, Austria), with a two-sided significance level set at P < 0.05. An institutional review board waiver was obtained before carrying out the present study, in accordance with institutional regulation when dealing with de-identified administrative data.
Results Baseline characteristics and renal function assessment Table 1 describes the baseline characteristics of 13 168 patients who underwent RP (65.4%), PN (10.7%), RN (16.1%) or RC (7.8%) at NSQIP participating hospitals between 2005 and 2011. The mean SCr and eGFR were 1.1 mg/dL and 82.1 mL/ min/1.73 m2, respectively. A total of 35.7% and 50.1% of the cohort were classified as having normal (no CKD) or moderately impaired (CKD II) renal function, respectively, with 12.6%, 0.7% and 0.9% classified as CKD III, IV and V, respectively. We provide a Table S1 showing patients’ characteristics and CKD stages stratified for each of the procedures carried out.
Intraoperative and postoperative outcomes Intraoperative details and postoperative outcomes of patients undergoing major urological cancer surgery, stratified by CKD stage, are shown in Table 2. The proportion of patients who experienced pOT was significantly greater in patients with CKD II (24.8%), III (32.9%) or IV (37.8%) compared with patients with no CKD (22.8%, P < 0.001). A similar trend was seen for patients experiencing a pLOS. A total of 14.1% of patients experienced at least one complication within 30 days of surgery; the rate of experiencing any postoperative complication was significantly greater in patients with CKD II (12.5%), III (27.9%), IV (43.3%) or V (18.6%) compared with patients without CKD (10.9%, P < 0.001). There was a total frequency of any 30-day complication of 7.9, 19.8, 13.6 and 55.6% after RP, RN, PN and RC, respectively. Similarly significant trends were observed when considering each postoperative complication separately (Table S1). The overall rate of reoperation was 1.6%, and was significantly higher in individuals with CKD II (1.6%), III (2.7%) or IV (4.4%) compared with those without CKD (1.0%, © 2014 The Japanese Urological Association
P = 0.005). In total, 80 patients died within 30 days of surgery, with a significantly higher 30-day mortality rate seen in patients with CKD II (0.5%), III (1.6%), IV (3.3%) or V (2.7%) compared with patients without CKD (0.3%, P < 0.001).
Multivariable analyses of complications and mortality Multivariable analyses showing predictors of experiencing any postoperative complication, 30-day mortality, receipt of an intraoperative transfusion, pOT and pLOS are shown in Tables 3–5. After controlling for age, race, sex, BMI, smoking status, baseline comorbid disease and type of surgery, the odds of experiencing any postoperative complication was significantly higher in patients with CKD III (OR 1.61, 95% CI 1.32–1.97, P < 0.001) and CKD IV (OR 2.24, 95% CI 1.27– 3.95, P = 0.005) compared with patients without CKD. Other independent predictors of experiencing any postoperative complication included ASA score ≥3 (OR 1.49, 95% CI 1.29–1.74, P < 0.001), history of pulmonary (OR 1.45, 95% CI 1.09–1.92, P = 0.012) or liver disease (OR 11.00, 95% CI 1.98–61.08, P = 0.006). Furthermore, RC was independently associated with a 5.5-fold higher risk of experiencing any complications compared with the other procedures (i.e. RN, PN and RP; P = 0.002). The presence of CKD III (OR 4.15, 95% CI 1.51–11.40, P = 0.006), IV (OR 10.10, 95% CI 2.22–45.99, P = 0.003) or V (OR 17.07, 95% CI 3.62–80.44, P < 0.001) was also an independent predictor of 30-day mortality. Additional independent predictors of 30-day mortality included age ≥75 years (OR 4.67, 95% CI 2.41–9.05, P < 0.001), ASA score ≥3 (OR 2.67, 95% CI 1.21–5.89, P = 0.015), baseline history of liver disease (OR 85.72, 95% CI 13.07–562.61, P < 0.001) and a positive smoking status (OR 2.52, 95% CI 1.28–4.97, P = 0.008). CKD III (OR 2.62, 95% CI 2.23–3.07, P < 0.001), IV (OR 3.37, 95% CI 1.86–6.09, P < 0.001) and V (OR 1.68, 95% CI 1.04–2.70, P = 0.033) predicted increased odds of experiencing a pLOS on multivariable analysis. Male sex (OR 3.87, 95% CI 3.35–4.47, P < 0.001), being of black or other race (OR 1.22, 95% CI 1.04–1.44, P = 0.015), age ≥75 years (OR 3.83, 95% CI 3.22–4.56, P < 0.001), ASA score ≥3 (OR 2.22, 95% CI 1.99– 2.49, P < 0.001), being a smoker (OR 1.69, 95% CI 1.48–1.93, P < 0.001) and having a history of diabetes (OR 1.25, 95% CI 1.08–1.44, P = 0.003) or pulmonary disease (OR 2.64, 95% CI 2.04–3.42, P < 0.001) also independently predicted a higher odds of pLOS. In similar multivariable analyses, CKD II (OR 1.12, 95% CI 1.01–1.26, P = 0.037) and III (OR 1.37, 95% CI 1.16–1.60, P < 0.001) were independent predictors of experiencing pOT, in addition to age ≥75 years (OR 1.51, 95% CI 1.28–1.79), BMI ≥30 (OR 1.27, 95% CI 1.14–1.40), ASA score ≥3 (OR 1.44, 95% CI 1.29–1.61) and a positive smoking status (OR 1.26, 95% CI 1.11–1.44, all P < 0.001). Patients with CKD V (OR 0.47, 95% CI 0.25–0.90, P = 0.022) were at lower odds of experiencing pOT compared with those without CKD, whereas male sex also independently predicted lower odds of experiencing pOT (OR 0.82, 95% CI 0.70–0.96, P = 0.011). CKD III was also an independent predictor of receipt of an intraoperative transfusion (OR 2.14, 95% CI 1.54–2.98, P < 0.001); male sex (OR 1.47, 95% CI 1.09–1.98, P = 0.011), 3
M SCHMID ET AL.
Table 1
Characteristics and level of renal function according to eGFR and CKD stages in 13 168 patients undergoing major urological surgery (RP, PN and RN, and RC)
Variables
Total cohort
No CKD*
CKD II**
CKD III***
Patients, n (%) Surgery, n (%) RP RN PN RC Mean serum creatinine, mg/dL (SD) Mean eGFR, mL/min/1.73 m2 (SD) Mean hematocrit, % (SD) Mean BUN, mg/dL (SD) Age, n (%) 18–59 years 60–74 years ≥75 years Sex, n (%) Male Female BMI, n (%) Underweight Normal Obesity Race, n (%) White Black Other ASA score, n (%) 1 2 ≥3 Comorbidities, n (%) Hypertension Diabetes mellitus Cardiac disease Peripheral vascular disease Cerebrovascular disease Pulmonary disease Liver disease Systemic steroid use Adjuvant chemotherapy# Lifestyle, n (%) Alcohol Smoking
13 168 (100)
4706 (35.7)
6598 (50.1)
1661 (12.6)
8 610 (65.4) 2 124 (16.1) 1 407 (10.7) 1 027 (7.8) 1.1 (0.8) 82.1 (27.2) 41.8 (4.7) 17.2 (7.1)
3330 (70.8) 589 (12.5) 539 (11.5) 248 (5.3) 0.8 (0.1) 105.2 (28.3) 42.2 (4.3) 14.5 (5.0)
4602 (69.7) 943 (14.3) 602 (9.1) 451 (6.8) 1.1 (0.1) 75.8 (7.8) 42.4 (4.3) 16.9 (4.9)
4 887 (37.1) 7 245 (54.3) 1 136 (8.6)
2265 (48.1) 2282 (48.5) 159 (3.4)
11 572 (87.9) 1 596 (12.1)
CKD IV****
CKD V*****
P-value
90 (0.7)
113 (0.9)
International Journal of Urology (2014)
doi: 10.1111/iju.12563
Original Article
Chronic kidney disease and perioperative outcomes in urological oncological surgery Marianne Schmid,1,2 Praful Ravi,3 Abd-El-Rahman M Abd-El-Barr,4 Julia Klap,2 Jesse D Sammon,4 Steven L Chang,2 Mani Menon,4 Adam S Kibel,2 Margit Fisch1 and Quoc-Dien Trinh2 1 Department of Urology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; 2Division of Urologic Surgery, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, USA; 3Hammersmith Hospital, Imperial College Healthcare NHS Trust, London, UK; and 4Center for Outcomes Research, Analytics and Evaluation, Vattikuti Urology Institute, Henry Ford Health System, Detroit, Michigan, USA
Abbreviations & Acronyms ASA = American Society of Anesthesiologists BMI = body mass index BUN = blood urea nitrogen CI = confidence interval CKD = chronic kidney disease CPT = Current Procedural Terminology eGFR = estimated glomerular filtration rate LOS = length of stay OR = odds ratio OT = operative time NSQIP = National Surgical Quality Improvement Program PN = partial nephrectomy pLOS = prolonged length of stay pOT = prolonged operation time RC = radical cystectomy RN = radical nephrectomy RP = radical prostatectomy SCr = serum creatinine SD = standard deviation Correspondence: Marianne Schmid M.D., Center for Surgery and Public Health and Division of Urologic Surgery, Brigham and Women’s Hospital, Harvard Medical School, 45 Francis Street, Boston, MA 02115, USA. Email: [email protected] Received 4 March 2014; accepted 15 June 2014.
© 2014 The Japanese Urological Association
Objectives: To evaluate baseline renal dysfunction among patients undergoing urological oncological surgery and its impact on early postoperative outcomes. Methods: Between 2005 and 2011, patients who underwent minimally-invasive or open radical prostatectomy, partial nephrectomy and radical nephrectomy, or open radical cystectomy, respectively, were identified in the National Surgical Quality Improvement Program dataset. Preoperative kidney function was assessed using estimated glomerular filtration rate and staged according to National Kidney Foundation definitions. Multivariable logistic regression was used to model the association between preoperative renal function and the risk of 30-day mortality and major complications. Furthermore the impact of chronic kidney disease on operation time and length of hospital stay was assessed. Results: Overall, 13 168 patients underwent radical prostatectomy (65.4%), partial nephrectomy (10.7%) and radical nephrectomy (16.1%) and radical cystectomy (7.8%), respectively; 50.1% of evaluable patients had reduced kidney function (chronic kidney disease II), and a further 12.6, 0.7 and 0.9% were respectively classified into chronic kidney disease stages III, IV, and V. Chronic kidney disease was an independent predictor of 30-day major postoperative complications (chronic kidney disease III: odds ratio 1.61, P < 0.001; chronic kidney disease IV: odds ratio 2.24, P = 0.01), of transfusions (chronic kidney disease III: odds ratio 2.14, P < 0001), of prolonged length of stay (chronic kidney disease III: odds ratio 2.61, P < 0.001; chronic kidney disease IV: odds ratio 3.37, P < 0.001; and chronic kidney disease V: odds ratio 1.68; P = 0.03) and of 30-day mortality (chronic kidney disease III: odds ratio 4.15, P = 0.01; chronic kidney disease IV: odds ratio 10.10, P = 0.003; and chronic kidney disease V: odds ratio 17.07, P < 0.001) compared with patients with no kidney disease. Conclusions: Renal dysfunction might be underrecognized in patients undergoing urological cancer surgery. Chronic kidney disease stages III, IV and V are independent predictors for poor 30-day postoperative outcomes.
Key words: chronic kidney disease, perioperative outcome, radical and partial nephrectomy, radical cystectomy, radical prostatectomy.
Introduction CKD, defined as an eGFR of less than 60 mL/min/1.73 m2,1 is common in the adult USA population with a rising prevalence from 10.0 (1988–1994) to 13.1% (1999–2004).2 Patients undergoing urological oncology procedures, such as RN and PN, RC or RP represent a unique population at risk of acute and chronic renal dysfunction.3–5 The incidence of urological malignancies increases with age, and therefore these patients more commonly present with comorbidities, such as hypertension, diabetes mellitus and vascular disease, which are known risk factors for the development of CKD.6,7 Furthermore, up to 30% of elderly urological patients have CKD, even in the presence of a normal SCr.4 Therefore, adequate assessment and preservation of renal function is critical in the urological oncological patient, as patients undergoing RC, RN and PN, in particular, face a further insult to renal function as a result of surgery. CKD is independently associated with an increased risk of cardiovascular events, hospitalization and mortality,8,9 and has been reported as a predictor of poor early and long-term postoperative outcomes in patients undergoing cardiac, vascular, general and orthopedic surgery.9–12 To our knowledge, no study has investigated the same in the setting of urological 1
M SCHMID ET AL.
cancer surgery; CKD has mostly been measured as an outcome after urological cancer surgery, and several studies, relying primarily on renal cell carcinoma cohorts, report on the incidence of CKD in patients before urological surgery and the postoperative progression of renal dysfunction.10–12 A recent population-based analysis showed an increased risk of adverse perioperative outcomes after nephrectomy in patients with endstage renal disease,13 whereas institutional series have found similar results in patients undergoing robotic PN.14,15 Based on the previous findings, it is important to assess the impact of CKD on perioperative outcomes after major urological cancer surgery, given that they carry a significant morbidity and a risk of baseline renal function impairment.16,17 We therefore sought to specifically examine the impact of CKD on perioperative outcomes after uro-oncological surgery for prostate, renal and bladder cancer, given the lack of evidence in this area. We utilized the NSQIP database, which is an American College of Surgeons initiative that supports the collection of risk-adjusted data to facilitate the assessment of surgical outcomes and complications.18 We hypothesized that CKD would be an independent predictor of adverse perioperative outcomes after urological cancer surgery, given the findings of a recent meta-analysis that relied predominantly on studies in cardiac and vascular surgery.17
Methods Data source The current study relies on the American College of Surgeons NSQIP data. The NSQIP was developed to assess the quality of surgical care, and collects perioperative data on 135 variables including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity for patients undergoing major surgical procedures in both the inpatient and outpatient setting across the USA. Trained Surgical Clinical Reviewers prospectively collect the NSQIP data. Validated data from patients’ medical charts allows quantification of 30-day, risk-adjusted surgical outcomes, including post-discharge, when nearly 50% of complications occur. In 2011, the NSQIP included data from 315 participating sites, and more than 1.7 million cases were contributed until 2011. Specifically, NSQIP further tracks preoperative patients’ characteristics, such as SCr values, BMI and history of comorbidities, which lends itself well for estimation of preoperative CKD.19 Though limited by the lack of certain data on surgical, pathology and imaging characteristics, as well as urine analytics, this database is specifically designed to accurately capture perioperative complications.20 NSQIP therefore served as an appropriate data source to investigate the impact of CKD on urological cancer surgery outcomes.
Study population Patients undergoing open or minimally-invasive (including laparoscopic, as well as robotic assistance) RP (CPT codes: 55810, 55812, 55815, 55840, 55842, 55845, 55866), open or minimally-invasive RN (CPT codes: 50220, 50225, 50230, 50545, 50546) or open or minimally-invasive PN (CPT codes: 50240, 50542, 50543) or open RC (CPT codes: 51575, 51580, 51585, 51590, 51595, 51596, 51597; there are no CPT codes 2
for laparoscopic/robotic RC available up-to date), with concomitant International Classification of Diseases 9th edition (code of cancer diagnosis of the prostate [185], kidney [189.0] or bladder [188.x]), were identified in the NSQIP dataset between 2005 and 2011. Preoperative eGFR was calculated using the simplified Modification of Diet in Renal Disease formula,21 a four-variable equation consisting of age, sex, ethnicity and SCr levels. Kidney function was staged according to National Kidney Foundation definitions.1 Patients with an eGFR ≥90 mL/min/ 1.73 m2 were considered as having no (or slightly diminished) kidney dysfunction, and patients with an eGFR between 60 and 89 mL/min/1.73 m2 were regarded as having mild reduction of kidney function (CKD stage II).22 Accordingly, patients with a moderate eGFR of 30–59 mL/min/1.73 m2 were classified into CKD stage III, patients with severe reduction of eGFR (15– 29 mL/min/1.73 m2) into CKD stage IV and with an eGFR 3 mg/dL and increasing azotemia [increase in blood urea nitrogen]) within 24 h before surgery were excluded from further analyses (n = 29, 0.2%). With these exclusions, 13 168 cases were left for analysis.
Covariates For each patient, age at surgery, sex, BMI, race, smoking status, alcohol consumption, preoperative SCr, blood urea nitrogen, hematocrit and type of surgery were available. Preoperative ASA score, as well as previous history of hypertension and diabetes mellitus, pulmonary, cardiac, liver, cerebrovascular and peripheral vascular disease, was reported.
Short-term outcomes Postoperative complications were classified under the following categories: cardiovascular (including postoperative cardiac arrest, myocardial infarction or cerebrovascular accident), pulmonary (including pneumonia, need for postoperative reintubation and need for ventilatory support >48 h), thromboembolic (including deep venous thrombosis and pulmonary embolism), septic (including sepsis and septic shock), renal (including acute renal failure and progressive renal insufficiency), urinary tract infections, wound complications (including superficial, deep and organ space surgical site infections, and wound dehiscence) and bleeding complications, according to previously reported methodology.23 Overall complications were defined as occurrence of any complication. Additional outcomes reported included the need for intraoperative transfusion, pOT, pLOS, readmission and perioperative mortality. pOT and pLOS were defined as an operative time or length of stay greater than the 75th percentile © 2014 The Japanese Urological Association
Renal function and urological surgery
(≥253 min and ≥4 days respectively). The 30-day readmission and reoperation data were only available starting 1 January 2011; therefore only 1 year of data (n = 6791) was analyzed for this outcome. Perioperative mortality was defined as death within 30 days of surgery.
Statistical analyses Descriptive statistics of categorical variables focused on frequencies and proportions. Means and standard deviations, medians, and interquartile ranges were reported for continuously coded variables. Multivariable logistic regression was used to model the association between preoperative renal function and overall complications, pOT, need of postoperative transfusion, pLOS, and mortality. Covariates consisted of age at surgery, race, sex, BMI, smoking status, baseline comorbidities and type of surgery. All statistical analyses were carried out using the R statistical package (R Foundation for Statistical Computing, Vienna, Austria), with a two-sided significance level set at P < 0.05. An institutional review board waiver was obtained before carrying out the present study, in accordance with institutional regulation when dealing with de-identified administrative data.
Results Baseline characteristics and renal function assessment Table 1 describes the baseline characteristics of 13 168 patients who underwent RP (65.4%), PN (10.7%), RN (16.1%) or RC (7.8%) at NSQIP participating hospitals between 2005 and 2011. The mean SCr and eGFR were 1.1 mg/dL and 82.1 mL/ min/1.73 m2, respectively. A total of 35.7% and 50.1% of the cohort were classified as having normal (no CKD) or moderately impaired (CKD II) renal function, respectively, with 12.6%, 0.7% and 0.9% classified as CKD III, IV and V, respectively. We provide a Table S1 showing patients’ characteristics and CKD stages stratified for each of the procedures carried out.
Intraoperative and postoperative outcomes Intraoperative details and postoperative outcomes of patients undergoing major urological cancer surgery, stratified by CKD stage, are shown in Table 2. The proportion of patients who experienced pOT was significantly greater in patients with CKD II (24.8%), III (32.9%) or IV (37.8%) compared with patients with no CKD (22.8%, P < 0.001). A similar trend was seen for patients experiencing a pLOS. A total of 14.1% of patients experienced at least one complication within 30 days of surgery; the rate of experiencing any postoperative complication was significantly greater in patients with CKD II (12.5%), III (27.9%), IV (43.3%) or V (18.6%) compared with patients without CKD (10.9%, P < 0.001). There was a total frequency of any 30-day complication of 7.9, 19.8, 13.6 and 55.6% after RP, RN, PN and RC, respectively. Similarly significant trends were observed when considering each postoperative complication separately (Table S1). The overall rate of reoperation was 1.6%, and was significantly higher in individuals with CKD II (1.6%), III (2.7%) or IV (4.4%) compared with those without CKD (1.0%, © 2014 The Japanese Urological Association
P = 0.005). In total, 80 patients died within 30 days of surgery, with a significantly higher 30-day mortality rate seen in patients with CKD II (0.5%), III (1.6%), IV (3.3%) or V (2.7%) compared with patients without CKD (0.3%, P < 0.001).
Multivariable analyses of complications and mortality Multivariable analyses showing predictors of experiencing any postoperative complication, 30-day mortality, receipt of an intraoperative transfusion, pOT and pLOS are shown in Tables 3–5. After controlling for age, race, sex, BMI, smoking status, baseline comorbid disease and type of surgery, the odds of experiencing any postoperative complication was significantly higher in patients with CKD III (OR 1.61, 95% CI 1.32–1.97, P < 0.001) and CKD IV (OR 2.24, 95% CI 1.27– 3.95, P = 0.005) compared with patients without CKD. Other independent predictors of experiencing any postoperative complication included ASA score ≥3 (OR 1.49, 95% CI 1.29–1.74, P < 0.001), history of pulmonary (OR 1.45, 95% CI 1.09–1.92, P = 0.012) or liver disease (OR 11.00, 95% CI 1.98–61.08, P = 0.006). Furthermore, RC was independently associated with a 5.5-fold higher risk of experiencing any complications compared with the other procedures (i.e. RN, PN and RP; P = 0.002). The presence of CKD III (OR 4.15, 95% CI 1.51–11.40, P = 0.006), IV (OR 10.10, 95% CI 2.22–45.99, P = 0.003) or V (OR 17.07, 95% CI 3.62–80.44, P < 0.001) was also an independent predictor of 30-day mortality. Additional independent predictors of 30-day mortality included age ≥75 years (OR 4.67, 95% CI 2.41–9.05, P < 0.001), ASA score ≥3 (OR 2.67, 95% CI 1.21–5.89, P = 0.015), baseline history of liver disease (OR 85.72, 95% CI 13.07–562.61, P < 0.001) and a positive smoking status (OR 2.52, 95% CI 1.28–4.97, P = 0.008). CKD III (OR 2.62, 95% CI 2.23–3.07, P < 0.001), IV (OR 3.37, 95% CI 1.86–6.09, P < 0.001) and V (OR 1.68, 95% CI 1.04–2.70, P = 0.033) predicted increased odds of experiencing a pLOS on multivariable analysis. Male sex (OR 3.87, 95% CI 3.35–4.47, P < 0.001), being of black or other race (OR 1.22, 95% CI 1.04–1.44, P = 0.015), age ≥75 years (OR 3.83, 95% CI 3.22–4.56, P < 0.001), ASA score ≥3 (OR 2.22, 95% CI 1.99– 2.49, P < 0.001), being a smoker (OR 1.69, 95% CI 1.48–1.93, P < 0.001) and having a history of diabetes (OR 1.25, 95% CI 1.08–1.44, P = 0.003) or pulmonary disease (OR 2.64, 95% CI 2.04–3.42, P < 0.001) also independently predicted a higher odds of pLOS. In similar multivariable analyses, CKD II (OR 1.12, 95% CI 1.01–1.26, P = 0.037) and III (OR 1.37, 95% CI 1.16–1.60, P < 0.001) were independent predictors of experiencing pOT, in addition to age ≥75 years (OR 1.51, 95% CI 1.28–1.79), BMI ≥30 (OR 1.27, 95% CI 1.14–1.40), ASA score ≥3 (OR 1.44, 95% CI 1.29–1.61) and a positive smoking status (OR 1.26, 95% CI 1.11–1.44, all P < 0.001). Patients with CKD V (OR 0.47, 95% CI 0.25–0.90, P = 0.022) were at lower odds of experiencing pOT compared with those without CKD, whereas male sex also independently predicted lower odds of experiencing pOT (OR 0.82, 95% CI 0.70–0.96, P = 0.011). CKD III was also an independent predictor of receipt of an intraoperative transfusion (OR 2.14, 95% CI 1.54–2.98, P < 0.001); male sex (OR 1.47, 95% CI 1.09–1.98, P = 0.011), 3
M SCHMID ET AL.
Table 1
Characteristics and level of renal function according to eGFR and CKD stages in 13 168 patients undergoing major urological surgery (RP, PN and RN, and RC)
Variables
Total cohort
No CKD*
CKD II**
CKD III***
Patients, n (%) Surgery, n (%) RP RN PN RC Mean serum creatinine, mg/dL (SD) Mean eGFR, mL/min/1.73 m2 (SD) Mean hematocrit, % (SD) Mean BUN, mg/dL (SD) Age, n (%) 18–59 years 60–74 years ≥75 years Sex, n (%) Male Female BMI, n (%) Underweight Normal Obesity Race, n (%) White Black Other ASA score, n (%) 1 2 ≥3 Comorbidities, n (%) Hypertension Diabetes mellitus Cardiac disease Peripheral vascular disease Cerebrovascular disease Pulmonary disease Liver disease Systemic steroid use Adjuvant chemotherapy# Lifestyle, n (%) Alcohol Smoking
13 168 (100)
4706 (35.7)
6598 (50.1)
1661 (12.6)
8 610 (65.4) 2 124 (16.1) 1 407 (10.7) 1 027 (7.8) 1.1 (0.8) 82.1 (27.2) 41.8 (4.7) 17.2 (7.1)
3330 (70.8) 589 (12.5) 539 (11.5) 248 (5.3) 0.8 (0.1) 105.2 (28.3) 42.2 (4.3) 14.5 (5.0)
4602 (69.7) 943 (14.3) 602 (9.1) 451 (6.8) 1.1 (0.1) 75.8 (7.8) 42.4 (4.3) 16.9 (4.9)
4 887 (37.1) 7 245 (54.3) 1 136 (8.6)
2265 (48.1) 2282 (48.5) 159 (3.4)
11 572 (87.9) 1 596 (12.1)
CKD IV****
CKD V*****
P-value
90 (0.7)
113 (0.9)
