Surgery for Obesity and Related Diseases 11 (2015) 828–835
Original article
Bariatric surgery in patients with reduced kidney function: an analysis of short-term outcomes Fady Saleh, M.D., M.P.H.a, S. Joseph Kim, M.D., Ph.D., M.H.S.b,c, Allan Okrainec, M.D., M.H.P.E.a,d, Timothy D. Jackson, M.D., M.P.H.a,d,* a
Division of General Surgery, University Health Network, Toronto, Ontario, Canada b Division of Nephrology, University Health Network, Toronto, Ontario, Canada c Department of Medicine, University of Toronto, Toronto, Ontario, Canada d Department of Surgery, University of Toronto, Toronto, Ontario, Canada Received October 10, 2014; accepted November 10, 2014
Abstract
Background: With rates of obesity among patients with chronic kidney disease (CKD) mirroring that of the general population, there is growing interest in offering bariatric surgery to these patients. We sought to determine the safety of bariatric surgery in this patient population. Methods: Patients who underwent selected laparoscopic bariatric procedures between 2005 and 2011. Estimated glomerular filtration rate (eGFR) was calculated and divided into stages of CKD. Procedures included Roux-en-Y gastric bypass (RYGB), laparoscopic adjustable gastric band (LAGB), and laparoscopic sleeve gastrectomy (SG). Univariable analysis and multivariable adjustment was used to compare complication rates across stages of eGFR. Results: A total of 64,589 patients were included: 64.5% underwent RYGB, 29.8% LAGB, and 5.7% SG. A total of 61.7% of patients had normal eGFR (Stage 1), 32.0% were stage 2, 5.3% were stage 3, and 1.0% were stage 4/5. After adjusting for relevant patient characteristics, there was a trend toward increasing complications from stage 1 to stage 4/5 CKD among RYGB, LAGB, and SG groups, but none were statistically significant. Similarly, major complications generally increased across stages of CKD for each procedure, but was only significant for RYGB comparing stage 3 to stage 1 (OR 1.22; 95% CI: 1.01–1.47; P ¼ .042) and risk difference .96% (95% CI: .03– 1.96). Considering only stage 4/5 CKD, overall (P ¼ .114) and major complications (P ¼ .032) were highest in the RYGB group, followed by SG and LAGB. Conclusion: More advanced stages of CKD do not appear to be statistically associated with an increased risk of 30-day postoperative complications. (Surg Obes Relat Dis 2015;11:828–835.) r 2015 Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.
Keywords:
Bariatric surgery; Chronic kidney disease; ACS NSQIP; Surgical outcomes; Patient safety
American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. * Correspondence: Timothy D. Jackson, M.D., M.P.H., Division of General Surgery, University Health Network, 399 Bathurst Street, Room 8MP-322, Toronto, Ontario, M5T 2S8 Canada. E-mail: [email protected]
The obesity epidemic is a growing problem in North America. Among U.S. adults, over 30% are obese and two thirds are overweight with morbid obesity (body mass index [BMI] 440 kg/m2) affecting approximately 5% of the population [1]. Obesity poses a serious threat to health and is associated with an increased prevalence of cardiac disease, diabetes mellitus, hypertension, and obstructive sleep apnea [2]. Furthermore, obesity increases the risk of
http://dx.doi.org/10.1016/j.soard.2014.11.012 1550-7289/r 2015 Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835
developing chronic kidney disease (CKD) 4-fold (7-fold in the morbidly obese) [3]. Not only do the obesity-related comorbidities of hypertension and diabetes account for 70% of the causes for end-stage renal disease (ESRD), obesity is also an independent predictor for developing ESRD [3,4]. Despite the well-described benefits of bariatric surgery on the amelioration of obesity-related illnesses, there appears to be a general reluctance to perform bariatric surgery in patients with ESRD likely because of safety concerns [5–7]. In a recent study, Cloyd et al. [7] demonstrated an increase in morbidity and mortality among hemodialysis patients compared to nonhemodialysis patients undergoing major abdominal surgery. In addition to perioperative risk, concerns over what is referred to as the ‘obesity paradox’, in which epidemiologic evidence suggests a survival advantage in hemodialysis patients with higher BMI, have called to question whether bariatric surgery should be offered to ESRD patients [8,9]. However, obesity is also an independent predictor of poor outcomes after kidney transplantation [10–12] and has been shown to be a barrier to accessing transplantation [13]. Bariatric surgeons are increasingly asked to see patients with impaired kidney function in the setting of multiple other medical co-morbidities and, at some centers, patients with ESRD in the setting of transplantation. There are limited data on the safety of performing bariatric surgery in patients with renal impairment. The objective of this study was to determine the effect of varying degrees of kidney dysfunction on short-term outcomes after bariatric surgery. Methods Data sources The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) Participant Use Files, from the years 2005 to 2011, were used for this study. ACS NSQIP is a prospective, multi-institutional, cohort study collecting rich clinical data on patients undergoing surgical procedures. Data are collected on preoperative, intraoperative, and postoperative variables, including 30-day outcomes. ACS NSQIP data methodology has been described in detail elsewhere [14–19]. The study protocol was approved by the institutional Research Ethics Board.
829
43770); and (3) laparoscopic sleeve gastrectomy (SG) (CPT 43775). Open bariatric procedures and revisional surgeries were excluded. Furthermore, patients were excluded if they exhibited high risk features such as documented American Society of Anesthesiologists (ASA) 5, recorded as an emergency case, a history of ascites, or a worsening cardiac condition. Patients were also excluded if a major concurrent procedure was performed at the time of their procedure. Fig. 1 outlines patient flow based on inclusion and exclusion criteria. Exposure and outcomes The exposure of interest was the estimated glomerular filtration rate (eGFR) in mL/min/1.73 m2, grouped according to stages of chronic kidney disease (CKD): stage 1: Z90 (normal reference group), stage 2: 60–89.9, stage 3: 30–59.9, and stage 4/5: o30 [20]. The eGFR was computed using the CKD-EPI formula because of its ability to more accurately characterize eGFR in bariatric patients [21]. Because patient race was missing in 9% of patients, eGFR was calculated using both with race missing and with race imputed using random assignment, with the latter presented in our results. Imputation was performed maintaining the proportions of race from the nonmissing data. CKD stages 4 and 5 were grouped together as eGFR o 30 because of small numbers of patients. All patients on dialysis were placed in this group. Our primary outcomes were 30-day overall postoperative morbidity, 30-day major postoperative morbidity, and 30day postoperative mortality. ACS NSQIP collects data on over 20 postoperative complications including woundrelated, septic or infectious, bleeding, thromboembolic, cardiorespiratory and renal complications. Both overall complications and major complications were composite outcomes of the postoperative 30-day complications listed in the data set, with major complications excluding urinary tract infections and superficial/deep wound infections. Both overall and major complications included prolonged length of stay (430 d), reoperation within 30 days, and mortality as a complication. All endpoints were binary outcome variables. Statistical analysis
Study patients All patients Z18 years old with a BMI Z30 kg/m2 who underwent a bariatric procedure during the study period were included in this study. Bariatric procedures were identified using both Current Procedural Terminology (CPT) codes and the ICD-9 diagnosis (i.e., the diagnosis was related to morbid obesity or associated condition). Three bariatric procedures were included: (1) laparoscopic Roux-en-Y gastric bypass (RYGB) (CPT 43644–5, 43846); (2) laparoscopic adjustable gastric band (LAGB) (CPT
Summary statistics were used to define the study population. Univariable analyses using the χ2 test or the Fisher’s exact test was performed to compare categorical variables and the Kruskal-Wallis test was used to compare continuous variables. Multivariable logistic regression was performed for each procedure separately and for both overall and major complications (6 total equations). Because there were too few deaths to perform multivariable regression modeling, only unadjusted results are provided for 30-day mortality. Variables controlled for in our models
830
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
Fig. 1. Patient flow diagram. ACS NSQIP ¼ American College of Surgeons National Surgical Quality Improvement Program; CPT ¼ current procedural terminology.
were BMI, age, ASA class, hypertensive status, and presence of diabetes. These potential confounders were selected using backward stepwise selection (P o .05 used as our cut-off) in a model assessing complications across the entire patient population. Adjusted odds ratios (OR) with 95% confidence intervals (CI) are reported along with absolute risk differences (RD). The RD was generated as average treatment effects using marginal probabilities from our logistic regression models and adjusted for potential confounding variables. A sensitivity analysis excluding patients with missing race was performed. Statistical significance was set at a 2-sided P value o .05. All statistical analyses were performed using Stata/IC, version 12.1 (Statacorp, College Station, TX). Results The characteristics of the study patients are presented in Table 1. There were 64,589 patients who met inclusion and exclusion criteria. Patients with missing baseline creatinine values (7,708 patients) were excluded but had similar baseline characteristics as the patients included (Fig. 1). The majority of patients had a normal eGFR, with stage 1 comprising 61.7% of patients followed by stage 2 with
32.0%, stage 3 with 5.3%, and stage 4/5 with 1.0%. Laparoscopic RYGB was the most common procedure (64.5%), followed by LAGB (29.8%) and SG (5.7%). The proportions of patients undergoing each of the 3 procedures were statistically different across stages of CKD (P o .001). In addition, patients with lower renal function tended to have more co-morbid conditions (such as diabetes, coronary artery disease, and hypertension), were older in age, and had a higher ASA class score (P o .001). Although there was a statistically significant difference in mean BMI across stages of CKD, the differences were quantitatively small. The majority of patients undergoing bariatric surgery were female (79.1%) and nonblack (85.8%). Overall complications, major complications, and mortality by procedure are compared across stages of CKD and are presented in Table 2. In this unadjusted comparison, there is a trend for the relative odds of overall complications to increase across categories of decreasing kidney function (with eGFR 490 as the referent). In all 3 procedures types, the greatest absolute difference in the incidence of complications across stages of CKD remains under 3%. Similarly, major complications increases across worsening eGFR, but is statistically significant only for RYGB. Mortality doubles from .12% to .25% in the RYGB comparing stage 1 and
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835
831
Table 1 Baseline patient characteristics* Variable† N ¼ 64,589 Procedure RYGB LAGB SG Mean age, yr (⫾SD) Black race Female Mean BMI, kg/m2 (⫾SD) Diabetes CAD‡ Smoker§ Hypertension ASA class ASA 1 ASA 2 ASA 3 ASA 4
eGFR Z 90 39,847 (61.7)
eGFR 60–89.9 20,665 (32.0)
eGFR 30–59.9 3,410 (5.3)
eGFR o 30 667 (1.0)
25,760 (60.5) 11,641 (60.5) 2,446 (66.2) 41.0 (10.5) 6,727 (18.7) 28,445 (79.1) 46.8 (8.1) 8,747 (24.3) 520 (1.62) 4,681 (13.0) 16,631 (46.3)
13,288 (31.9) 6,364 (33.1) 1,013 (27.4) 49.9 (10.0) 2,027 (10.6) 15,265 (79.8) 45.6 (7.6) 5,624 (29.4) 650 (3.7) 1,903 (10.0) 12,020 (62.9)
2,209 (5.3) 1,028 (5.4) 173 (4.7) 57.7 (8.5) 324 (10.1) 2,415 (75.4) 46.5 (8.2) 1,764 (55.0) 309 (10.6) 231 (7.2) 2,809 (87.6)
404 (1.0) 199 (1.0) 64 (1.7) 50.7 (11.7) 94 (25.8) 239 (65.5) 47.0 (7.4) 198 (54.3) 33 (10.1) 30 (8.2) 313 (85.8)
199 (.6) 13,514 (37.6) 21,692 (60.4) 504 (1.4)
81 (.4) 5,839 (30.6) 12,850 (67.3) 325 (1.7)
2 (.1) 571 (17.8 ) 2,485 (77.6) 145 (4.5)
0 (.0) 52 (14.3) 271 (74.5) 41 (11.3)
P value N/A o.001
o.001 o.001 o.001 o.001 o.001 o.001 o.001 o.001 o.001
eGFR ¼ estimated glomerular filtration rate; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy; BMI ¼ body mass index; CAD ¼ coronary artery disease; ASA ¼ American Society of Anesthesiologists. eGFR measured in mL/min/1.73 m2. * Patient with missing data: race 5892, gender 269, BMI 97, CAD 6444, smoker 1, ASA 67. † Unless otherwise indicated, values in brackets represent the percentage of patients in the respective subgroup. ‡ Previous history of cardiac bypass surgery or angioplasty. § Smoker within last year.
stages 4/5, and the trend across stages is statistically significant (P ¼ .001). Notably, there is no statistically significant relationship between kidney function and mortality in either the LAGB or the SG groups, although the numbers of patients in the higher stages of CKD are low. Table 3 shows the univariable and multivariable logistic regression models for overall complications by procedure across stages of CKD. Both the univariable and multivariable results are presented as OR while RD is illustrated graphically in Fig. 2, displaying results on both multiplicative and
additive scales. Considering the RYGB group first, after adjusting for important co-morbid conditions, all of the adjusted ORs reduce in magnitude and statistical significance from the unadjusted univariable ORs (i.e., closer to 1.0). For example, the unadjusted OR comparing stage 3 CKD (eGFR 30–59.9) to stage 1 (eGFR Z90) is 1.45 (95% CI: 1.25, 1.68; P o .001) and becomes 1.13 (95% CI: .96, 1.32; P ¼ .145) after adjustment with a RD .84% (95% CI: -.03, 2.00; P ¼ .145). This trend toward the null after multivariable adjustment also was seen in the LAGB group,
Table 2 30-day overall complications, major complications and mortality by procedure Procedure
Complications*
eGFR Z 90
eGFR 60–89.9
eGFR 30–59.9
eGFR o 30
P value
RYGB
N Overall Major Death N Overall Major Death N Overall Major Death
25,760 1,778 (6.9) 1,166 (4.5) 30 (.12) 11,641 278 (2.4) 174 (1.5) 3 (.03) 2,446 122 (5.0) 81 (3.3) 1 (0.04)
13,288 1,001 (7.5) 671 (5.1) 24 (.18) 6,464 182 (2.9) 121 (1.9) 1 (.02) 1,013 51 (5.0) 38 (3.8) 0 (.0)
2,209 214 (9.7) 151 (6.8) 11 (.5) 1,028 45 (4.4) 23 (2.2) 2 (.19) 173 13 (7.5) 8 (4.6) 0 (.0)
404 35 (8.7) 29 (7.2) 1 (.25) 199 8 (4.0) 4 (2.0) 0 (.0) 64 5 (7.8) 4 (6.3) 0 (.0)
N/A o.001 o.001 .001 N/A .001 .094 .081 N/A .383 .480 1.000
LAGB
SG
eGFR ¼ estimated glomerular filtration rate; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Values in brackets represent the percentage of patients in the respective subgroup.
832
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
Table 3 Univariable and multivariable logistic regression models of overall complications and renal function by procedure Procedure
RYGB
LAGB
SG
eGFR
Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30
Univariable analysis
Multivariable analysis*
OR
95% CI
P value
OR
95% CI
P value
1.00 1.10 1.45 1.28 1.00 1.20 1.87 1.71 1.00 1.01 1.55 1.61
Referent 1.01, 1.19 1.25, 1.68 .90, 1.82 Referent 1.00, 1.45 1.36, 2.58 .84, 3.51 Referent .72, 1.41 .85, 2.80 .64, 4.10
Referent .022 o.001 .168 Referent .055 o.001 .142 Referent .954 .149 .313
1.00 1.01 1.13 1.11 1.00 1.00 1.20 1.30 1.00 1.06 1.64 1.68
Referent .92, 1.10 .96, 1.32 .78, 1.58 Referent .81, 1.23 .84, 1.71 .63, 2.69 Referent .74, 1.51 .87, 3.12 .65, 4.32
Referent .864 .145 .552 Referent .977 .314 .473 Referent .755 .128 .280
eGFR ¼ estimated glomerular filtration rate; OR ¼ odds ratio; CI ¼ confidence interval; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Adjusting for American Society of Anesthesiologists (1 and 2 versus 3 and 4), hypertension, diabetic status, age, and body mass index.
but does not exist for overall complications for SG. Despite a slight increase in both the ORs and RDs, the results fail to achieve significance. For example, in the stage 4/5 CKD, the OR increases from 1.61 (95% CI: .64, 4.10; P ¼ .313) to 1.68 (95% CI: .65, 4.32; P ¼ .280) with an adjusted RD 3.08% (95% CI: -3.76, 9.93; P ¼ .280). Table 4 and Fig. 3 outline results of major complications, showing a similar pattern as demonstrated for overall complications. Again, for RYGB and LAGB, the OR and
RD come closer to the null after adjusting for important confounders compared to the unadjusted results, while those of the SG slightly increase. Additionally, the ORs again either remain not statistically significant or lose significance after adjustment, with the exception of Stage 3 patients compared to Stage 1 in the RYGB procedure, OR 1.22 (95% CI: 1.01, 1.47; P ¼ .042) and RD .96% (95% CI: .03, 1.96; P ¼ .042). Interestingly, in the LAGB group, the adjusted results show that patients are at comparable risk for complications across categories of kidney function. Table 5 shows an analysis of CKD stage 4/5 subcohort and compares unadjusted outcomes across the 3 procedures. Although there are too few patients to appropriately perform an adjusted analysis, it appears the LAGB have the lowest incidence of 30-day complications, followed by the SG and RYGB. When patients with missing race were excluded, the results showed similar findings to the primary analysis. Discussion Before adjusting for important patient characteristics, our univariable analysis showed a trend for increasing overall and major complications as well as death across advancing stages of CKD. Although this general trend of increasing complications remained across all procedure groups after adjusting for patient co-morbidities, the odds ratios and risk differences for complications were more attenuated in the RYGB and LAGB groups and slightly increased in the SG group. Importantly, all but 1 of the adjusted analyses did not reach statistical significance, the exception being the association between major complications between CKD stage 3 compared to CKD stage 1 in the RYGB group. In essence,
Fig. 2. Forrest plot of adjusted risk differences for overall complications. RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835 Table 4 Univariable and multivariable logistic regression models of major complications and renal function by procedure Procedure eGFR
RYGB
LAGB
SG
Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30
Univariable analysis
Multivariable analysis*
OR 95% CI
P value OR 95% CI
P value
1.00 1.12 1.55 1.63 1.00 1.28 1.51 1.35 1.00 1.14 1.42 1.94
Referent .021 o.001 .012 Referent .040 .067 .555 Referent .519 .359 .208
Referent .695 .042 .068 Referent .827 .774 1.000 Referent .375 .308 .191
Referent 1.02, 1.24 1.30, 1.84 1.11, 2.39 Referent 1.01, 1.61 .97, 2.34 .50, 3.68 Referent 0.77, 1.68 .67, 2.98 .69, 5.49
1.00 1.02 1.22 1.43 1.00 1.03 0.93 1.00 1.00 1.21 1.51 2.02
Referent .92, 1.13 1.01, 1.47 .97, 2.10 Referent .80, 1.33 .58, 1.50 .36, 2.74 Referent .80, 1.84 .68, 3.36 .70, 5.78
eGFR ¼ estimated glomerular filtration rate; OR ¼ odds ratio; CI ¼ confidence interval; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Adjusting for American Society of Anesthesiologists (1 and 2 versus 3 and 4), hypertension, diabetic status, age, and body mass index.
although kidney dysfunction may be associated with increased complications, it seems that this increase is very modest. In fact, much of the morbidity in these patients may result from case mix rather than reduced kidney function per se. Turgeon et al. have previously reported on outcomes and stages of CKD using the ACS NSQIP data between the years 2006 and 2008 [37]. Although our study is similar,
833
Table 5 Unadjusted comparison of outcomes and bariatric procedures in CKD class 4/5 patients Variable n ¼ 667
RYGB 404 (60.6)
LAGB 199 (29.8)
SG 64 (9.6)
P value NA
Any complication Major complication Death
35 (8.7) 29 (7.2) 1 (.3)
8 (4.0) 4 (2.0) 0
5 (7.8) 4 (6.3) 0
.114 .032 .722
CKD ¼ chronic kidney disease; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy.
there are a number of important differences. First, we extended the cohort by adding the years 2009–2011. Additionally, we limited procedures to the laparoscopic approach. Open bariatric surgery results in significantly higher overall morbidity and mortality [22]. Additionally, they included revisional procedures which are well known to be a very distinct surgical group with higher complication rates [23]. Finally, and most importantly, we separately examined the various bariatric procedures RYGB, SG, and LAGB, which all have different complication rates [24]. These differences may explain the stronger associations between stages of CKD and complications found in the study by Turgeon et al. A number of small case series have been published demonstrating that bariatric surgery appears to be safe in end-stage renal disease (ESRD) patients [5,25–28] while other evidence suggests higher complication rates in patients with ESRD [29,30]. In a review of bariatric surgery using the United States Renal Data System from
Fig. 3. Forrest plot of adjusted risk differences for major complications. RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; LSG ¼ laparoscopic sleeve gastrectomy.
834
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
1991–2004, there was a 3.5% mortality rate in the 116 patients with ESRD who had bariatric surgery, although the vast majority had bariatric surgery after transplantation and the cohort likely had a large number of open surgery given the years of the study [30]. Bariatric surgery has many attractive features as a treatment for obesity in ESRD patients. Obesity is highly prevalent in patients with ESRD and reflects the prevalence seen in the general population (i.e., 60% obese or overweight) [31]. Additionally, obese patients with ESRD who undergo kidney transplantation are subject to weight gain after transplantation with the potential to exacerbate the adverse health effects of their obesity and result in decreased long-term graft survival and poorer outcomes [12,32]. In fact, morbid obesity (BMI 4 36) was associated with a greater than 30% risk of graft loss compared to a BMI from 24–26, delayed graft function, increased cardiovascular death, and overall mortality with a functioning graft [11]. Weight loss after bariatric surgery in this patient population as well as the improvements in obesity-related illnesses mimics that of patients without CKD. Additionally, the effectiveness of transplant immunosuppressive drug regimens do not appear to be affected by the altered digestive systems of postbariatric patients given that many are driven by drug levels [5,25,33]. In addition to the many potential benefits that weight loss surgery has on the health status of obese individuals, bariatric surgery has the potential to improve kidney function in patients with CKD. In a study by Alexander and Goodman [25], 5 patients of 32 with CKD, the majority of whom were on dialysis for a mean of 2.5 years, had resolution or stabilization of their kidney disease, some of who came off the waiting list for transplantation [5]. In addition, many patients who may not have been candidates for kidney transplantation because of their weight or comorbid disease status, may become eligible after bariatric surgery [13,28]. Bariatric surgery has been shown to result in sustained weight loss [34] and results in improvement or resolution of many obesity related co-morbidities [2,25]. On average across the different bariatric procedures for all comers, resolution can be seen in up to 75% of diabetics, 60% hypertensives, and 85% of patients with obstructive sleep apnea [2] with evidence showing increased longevity in morbidly obese patients after bariatric surgery [34–36]. The extent to which these benefits extend to ESRD patients remains to be seen. There are several limitations to this study. First, because of the observational nature of this study, we are unable to control for all important confounders that could affect patient outcomes including, but not limited to, perioperative hospital care, hospital volume, and surgeon experience. Secondly, the selection of patients for surgery based on factors that are not captured in the database may lead to selection bias in the association between renal function and complications. Thirdly, ACS NSQIP does not collect data
on procedure-specific outcomes or outcomes beyond 30 days including anastomotic strictures and internal herniation but it does capture the complications relevant to this study, especially as it pertains to safety. Finally, despite the large total sample size, the number of events among patients in the most advanced stages of CKD (i.e., stages 4 and 5) was small and thus we may have missed smaller effects in the lower range of kidney function. In conclusion, more advanced stages of CKD do not appear to be statistically associated with an increased risk of 30-day postoperative complications. The excess morbidity observed in this study may be driven by the co-morbid disease status of the patients. Further research is necessary to help physicians appropriately select patients with CKD who would most benefit from bariatric surgery. Additionally, further studies are needed to assess the optimal procedure type and timing for surgery in these patients. Disclosures The authors have no commercial associations that might be a conflict of interest in relation to this article. References [1] Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 19992004. JAMA 2006;295:1549–55. [2] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724–37. [3] Kalaitzidis RG, Siamopoulos KC. The role of obesity in kidney disease: recent findings and potential mechanisms. Int Urol Nephrol 2011;43:771–84. [4] Agnani S, Vachharajani VT, Gupta R, Atray N, Vachharajani TJ. Does treating obesity stabilize chronic kidney disease? BMC Nephrol 2005;6:7. [5] Alexander JW, Goodman HR, Gersin K, et al. Gastric bypass in morbidly obese patients with chronic renal failure and kidney transplant. Transplantation 2004;78:469–74. [6] Gajdos C, Hawn MT, Kile D, Robinson TN, Henderson WG. Risk of major nonemergent inpatient general surgical procedures in patients on long-term dialysis. JAMA Surg 2013;148:137–43. [7] Cloyd JM, Ma Y, Morton JM, Tamura MK, Poultsides GA, Visser BC. Does chronic kidney disease affect outcomes after major abdominal surgery? Results from the National Surgical Quality Improvement Program. J Gastrointest Surg 2014;18:605–12. [8] Celebi-Onder S, Schmidt RJ, Holley JL. Treating the obese dialysis patient: challenges and paradoxes. Semin Dial 2012;25:311–9. [9] Degoulet P, Legrain M, Reach I, et al. Mortality risk factors in patients treated by chronic hemodialysis. Report of the Diaphane collaborative study. Nephron 1982;31:103–10. [10] el-Agroudy AE, Wafa EW, Gheith OE, Shehab el-Dein AB, Ghoneim MA. Weight gain after renal transplantation is a risk factor for patient and graft outcome. Transplantation 2004;77:1381–5. [11] Meier-Kriesche HU, Arndorfer JA, Kaplan B. The impact of body mass index on renal transplant outcomes: a significant independent risk factor for graft failure and patient death. Transplantation 2002;73: 70–4. [12] Curran SP, Famure O, Li Y, Kim SJ. Increased recipient body mass index is associated with acute rejection and other adverse outcomes after kidney transplantation. Transplantation 2014;97:64–70.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835 [13] Segev DL, Simpkins CE, Thompson RE, Locke JE, Warren DS, Montgomery RA. Obesity impacts access to kidney transplantation. J Am Soc Nephrol 2008;19:349–55. [14] Khuri SF, Henderson WG, Daley J, et al. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007;204:1089–102. [15] Khuri SF, Daley J, Henderson W, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:315–27. [16] Daley J, Khuri SF, Henderson W, et al. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:328–40. [17] Daley J, Forbes MG, Young GJ, et al. Validating risk-adjusted surgical outcomes: site visit assessment of process and structure National VA Surgical Risk Study. J Am Coll Surg 1997;185:341–51. [18] Khuri SF, Henderson WG, Daley J, et al. Successful implementation of the Department of Veterans Affairs’ National Surgical Quality Improvement Program in the private sector: the patient safety in surgery study. Ann Surg 2008;248:329–36. [19] Shiloach M, Frencher SK Jr, Steeger JE, et al. Towards robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16. [20] Levey AS, Coresh J, Balk E, et al. National Kidney Foundation: National Kidney Foundation practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Ann Intern Med 2003;139:137–47. [21] Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT. Performance of the Cockcroft-Gault, MDRD, and New CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol 2010;5:1003–9. [22] Hutter MM, Randall S, Khuri S, Henderson WG, Abbott WM, Warshaw AL. Laparoscopic versus open gastric bypass for morbid obesity: a multicenter, prospective, risk-adjusted analysis from the National Surgical Quality Improvement Program. Ann Surg 2006;243:657–62. [23] Hallowell PT, Stellato TA, Yao DA, Robinson A, Schuster MM, Graf KN. Should bariatric revisional surgery be avoided secondary to increased morbidity and mortality? Am J Surg 2009;197:391–6.
835
[24] Birkmeyer NJ, Dimick JB, Share D, et al. Hospital complication rates with bariatric surgery in Michigan. JAMA 2010;304:435–42. [25] Alexander JW, Goodman HR. Gastric bypass in chronic renal failure and renal transplant. Nutr Clin Pract 2007;22:16–21. [26] Koshy AN, Coombes JS, Wilkinson S, Fassett RG. Laparoscopic gastric banding surgery performed in obese dialysis patients prior to kidney transplantation. Am J Kidney Dis 2008;52:e15–7. [27] Newcombe V, Blanch A, Slater GH, Szoid A, Fielding GA. Laparoscopic adjustable gastric banding prior to renal transplantation. Obes Surg 2005;15:567–70. [28] Takata MC, Campos GM, Ciovica R, et al. Laparoscopic bariatric surgery improves candidacy in morbidly obese patients awaiting transplantation. Surg Obes Relat Dis 2008;4:159–65. [29] MacLaughlin HL, Hall WL, Patel AG, Macdougall IC. Laparoscopic sleeve gastrectomy is a novel and effective treatment for obesity in patients with chronic kidney disease. Obes Surg 2012;22:119–23. [30] Modanlou KA, Muthyala U, Xiao H, et al. Bariatric surgery among kidney transplant candidates and recipients: analysis of the United States renal data system and literature review. Transplantation 2009;87:1167–73. [31] Friedman AN, Miskulin DC, Rosenberg IH, Levey AS. Demographics and trends in overweight and obesity in patients at time of kidney transplantation. Am J Kidney Dis 2003;41:480–7. [32] Djamali A, Samaniego M, Muth B, et al. Medical care of kidney transplant recipients after the first posttransplant year. Clin J Am Soc Nephrol 2006;1:623–40. [33] Currie A, Chetwood A, Ahmed AR. Bariatric surgery and renal function. Obes Surg 2011;21:528–39. [34] Sjostrom L, Narbro K, Sjostrom CD, et al. Swedish Obese Subjects Study. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med 2007;357:741–52. [35] Christou NV, Sampalis JS, Liberman M, et al. Surgery decreases long-term mortality, morbidity, and health care use in morbidly obese patients. Ann Surg 2004;240:416–24. [36] Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA 2003;289:187–93. [37] Turgeon NA, Perez S, Mondestin M, et al. The impact of renal function on outcomes of bariatric surgery. J Am Soc Nephrol 2012; 23:885–94.
Original article
Bariatric surgery in patients with reduced kidney function: an analysis of short-term outcomes Fady Saleh, M.D., M.P.H.a, S. Joseph Kim, M.D., Ph.D., M.H.S.b,c, Allan Okrainec, M.D., M.H.P.E.a,d, Timothy D. Jackson, M.D., M.P.H.a,d,* a
Division of General Surgery, University Health Network, Toronto, Ontario, Canada b Division of Nephrology, University Health Network, Toronto, Ontario, Canada c Department of Medicine, University of Toronto, Toronto, Ontario, Canada d Department of Surgery, University of Toronto, Toronto, Ontario, Canada Received October 10, 2014; accepted November 10, 2014
Abstract
Background: With rates of obesity among patients with chronic kidney disease (CKD) mirroring that of the general population, there is growing interest in offering bariatric surgery to these patients. We sought to determine the safety of bariatric surgery in this patient population. Methods: Patients who underwent selected laparoscopic bariatric procedures between 2005 and 2011. Estimated glomerular filtration rate (eGFR) was calculated and divided into stages of CKD. Procedures included Roux-en-Y gastric bypass (RYGB), laparoscopic adjustable gastric band (LAGB), and laparoscopic sleeve gastrectomy (SG). Univariable analysis and multivariable adjustment was used to compare complication rates across stages of eGFR. Results: A total of 64,589 patients were included: 64.5% underwent RYGB, 29.8% LAGB, and 5.7% SG. A total of 61.7% of patients had normal eGFR (Stage 1), 32.0% were stage 2, 5.3% were stage 3, and 1.0% were stage 4/5. After adjusting for relevant patient characteristics, there was a trend toward increasing complications from stage 1 to stage 4/5 CKD among RYGB, LAGB, and SG groups, but none were statistically significant. Similarly, major complications generally increased across stages of CKD for each procedure, but was only significant for RYGB comparing stage 3 to stage 1 (OR 1.22; 95% CI: 1.01–1.47; P ¼ .042) and risk difference .96% (95% CI: .03– 1.96). Considering only stage 4/5 CKD, overall (P ¼ .114) and major complications (P ¼ .032) were highest in the RYGB group, followed by SG and LAGB. Conclusion: More advanced stages of CKD do not appear to be statistically associated with an increased risk of 30-day postoperative complications. (Surg Obes Relat Dis 2015;11:828–835.) r 2015 Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.
Keywords:
Bariatric surgery; Chronic kidney disease; ACS NSQIP; Surgical outcomes; Patient safety
American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. * Correspondence: Timothy D. Jackson, M.D., M.P.H., Division of General Surgery, University Health Network, 399 Bathurst Street, Room 8MP-322, Toronto, Ontario, M5T 2S8 Canada. E-mail: [email protected]
The obesity epidemic is a growing problem in North America. Among U.S. adults, over 30% are obese and two thirds are overweight with morbid obesity (body mass index [BMI] 440 kg/m2) affecting approximately 5% of the population [1]. Obesity poses a serious threat to health and is associated with an increased prevalence of cardiac disease, diabetes mellitus, hypertension, and obstructive sleep apnea [2]. Furthermore, obesity increases the risk of
http://dx.doi.org/10.1016/j.soard.2014.11.012 1550-7289/r 2015 Published by Elsevier Inc. on behalf of American Society for Metabolic and Bariatric Surgery.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835
developing chronic kidney disease (CKD) 4-fold (7-fold in the morbidly obese) [3]. Not only do the obesity-related comorbidities of hypertension and diabetes account for 70% of the causes for end-stage renal disease (ESRD), obesity is also an independent predictor for developing ESRD [3,4]. Despite the well-described benefits of bariatric surgery on the amelioration of obesity-related illnesses, there appears to be a general reluctance to perform bariatric surgery in patients with ESRD likely because of safety concerns [5–7]. In a recent study, Cloyd et al. [7] demonstrated an increase in morbidity and mortality among hemodialysis patients compared to nonhemodialysis patients undergoing major abdominal surgery. In addition to perioperative risk, concerns over what is referred to as the ‘obesity paradox’, in which epidemiologic evidence suggests a survival advantage in hemodialysis patients with higher BMI, have called to question whether bariatric surgery should be offered to ESRD patients [8,9]. However, obesity is also an independent predictor of poor outcomes after kidney transplantation [10–12] and has been shown to be a barrier to accessing transplantation [13]. Bariatric surgeons are increasingly asked to see patients with impaired kidney function in the setting of multiple other medical co-morbidities and, at some centers, patients with ESRD in the setting of transplantation. There are limited data on the safety of performing bariatric surgery in patients with renal impairment. The objective of this study was to determine the effect of varying degrees of kidney dysfunction on short-term outcomes after bariatric surgery. Methods Data sources The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) Participant Use Files, from the years 2005 to 2011, were used for this study. ACS NSQIP is a prospective, multi-institutional, cohort study collecting rich clinical data on patients undergoing surgical procedures. Data are collected on preoperative, intraoperative, and postoperative variables, including 30-day outcomes. ACS NSQIP data methodology has been described in detail elsewhere [14–19]. The study protocol was approved by the institutional Research Ethics Board.
829
43770); and (3) laparoscopic sleeve gastrectomy (SG) (CPT 43775). Open bariatric procedures and revisional surgeries were excluded. Furthermore, patients were excluded if they exhibited high risk features such as documented American Society of Anesthesiologists (ASA) 5, recorded as an emergency case, a history of ascites, or a worsening cardiac condition. Patients were also excluded if a major concurrent procedure was performed at the time of their procedure. Fig. 1 outlines patient flow based on inclusion and exclusion criteria. Exposure and outcomes The exposure of interest was the estimated glomerular filtration rate (eGFR) in mL/min/1.73 m2, grouped according to stages of chronic kidney disease (CKD): stage 1: Z90 (normal reference group), stage 2: 60–89.9, stage 3: 30–59.9, and stage 4/5: o30 [20]. The eGFR was computed using the CKD-EPI formula because of its ability to more accurately characterize eGFR in bariatric patients [21]. Because patient race was missing in 9% of patients, eGFR was calculated using both with race missing and with race imputed using random assignment, with the latter presented in our results. Imputation was performed maintaining the proportions of race from the nonmissing data. CKD stages 4 and 5 were grouped together as eGFR o 30 because of small numbers of patients. All patients on dialysis were placed in this group. Our primary outcomes were 30-day overall postoperative morbidity, 30-day major postoperative morbidity, and 30day postoperative mortality. ACS NSQIP collects data on over 20 postoperative complications including woundrelated, septic or infectious, bleeding, thromboembolic, cardiorespiratory and renal complications. Both overall complications and major complications were composite outcomes of the postoperative 30-day complications listed in the data set, with major complications excluding urinary tract infections and superficial/deep wound infections. Both overall and major complications included prolonged length of stay (430 d), reoperation within 30 days, and mortality as a complication. All endpoints were binary outcome variables. Statistical analysis
Study patients All patients Z18 years old with a BMI Z30 kg/m2 who underwent a bariatric procedure during the study period were included in this study. Bariatric procedures were identified using both Current Procedural Terminology (CPT) codes and the ICD-9 diagnosis (i.e., the diagnosis was related to morbid obesity or associated condition). Three bariatric procedures were included: (1) laparoscopic Roux-en-Y gastric bypass (RYGB) (CPT 43644–5, 43846); (2) laparoscopic adjustable gastric band (LAGB) (CPT
Summary statistics were used to define the study population. Univariable analyses using the χ2 test or the Fisher’s exact test was performed to compare categorical variables and the Kruskal-Wallis test was used to compare continuous variables. Multivariable logistic regression was performed for each procedure separately and for both overall and major complications (6 total equations). Because there were too few deaths to perform multivariable regression modeling, only unadjusted results are provided for 30-day mortality. Variables controlled for in our models
830
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
Fig. 1. Patient flow diagram. ACS NSQIP ¼ American College of Surgeons National Surgical Quality Improvement Program; CPT ¼ current procedural terminology.
were BMI, age, ASA class, hypertensive status, and presence of diabetes. These potential confounders were selected using backward stepwise selection (P o .05 used as our cut-off) in a model assessing complications across the entire patient population. Adjusted odds ratios (OR) with 95% confidence intervals (CI) are reported along with absolute risk differences (RD). The RD was generated as average treatment effects using marginal probabilities from our logistic regression models and adjusted for potential confounding variables. A sensitivity analysis excluding patients with missing race was performed. Statistical significance was set at a 2-sided P value o .05. All statistical analyses were performed using Stata/IC, version 12.1 (Statacorp, College Station, TX). Results The characteristics of the study patients are presented in Table 1. There were 64,589 patients who met inclusion and exclusion criteria. Patients with missing baseline creatinine values (7,708 patients) were excluded but had similar baseline characteristics as the patients included (Fig. 1). The majority of patients had a normal eGFR, with stage 1 comprising 61.7% of patients followed by stage 2 with
32.0%, stage 3 with 5.3%, and stage 4/5 with 1.0%. Laparoscopic RYGB was the most common procedure (64.5%), followed by LAGB (29.8%) and SG (5.7%). The proportions of patients undergoing each of the 3 procedures were statistically different across stages of CKD (P o .001). In addition, patients with lower renal function tended to have more co-morbid conditions (such as diabetes, coronary artery disease, and hypertension), were older in age, and had a higher ASA class score (P o .001). Although there was a statistically significant difference in mean BMI across stages of CKD, the differences were quantitatively small. The majority of patients undergoing bariatric surgery were female (79.1%) and nonblack (85.8%). Overall complications, major complications, and mortality by procedure are compared across stages of CKD and are presented in Table 2. In this unadjusted comparison, there is a trend for the relative odds of overall complications to increase across categories of decreasing kidney function (with eGFR 490 as the referent). In all 3 procedures types, the greatest absolute difference in the incidence of complications across stages of CKD remains under 3%. Similarly, major complications increases across worsening eGFR, but is statistically significant only for RYGB. Mortality doubles from .12% to .25% in the RYGB comparing stage 1 and
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835
831
Table 1 Baseline patient characteristics* Variable† N ¼ 64,589 Procedure RYGB LAGB SG Mean age, yr (⫾SD) Black race Female Mean BMI, kg/m2 (⫾SD) Diabetes CAD‡ Smoker§ Hypertension ASA class ASA 1 ASA 2 ASA 3 ASA 4
eGFR Z 90 39,847 (61.7)
eGFR 60–89.9 20,665 (32.0)
eGFR 30–59.9 3,410 (5.3)
eGFR o 30 667 (1.0)
25,760 (60.5) 11,641 (60.5) 2,446 (66.2) 41.0 (10.5) 6,727 (18.7) 28,445 (79.1) 46.8 (8.1) 8,747 (24.3) 520 (1.62) 4,681 (13.0) 16,631 (46.3)
13,288 (31.9) 6,364 (33.1) 1,013 (27.4) 49.9 (10.0) 2,027 (10.6) 15,265 (79.8) 45.6 (7.6) 5,624 (29.4) 650 (3.7) 1,903 (10.0) 12,020 (62.9)
2,209 (5.3) 1,028 (5.4) 173 (4.7) 57.7 (8.5) 324 (10.1) 2,415 (75.4) 46.5 (8.2) 1,764 (55.0) 309 (10.6) 231 (7.2) 2,809 (87.6)
404 (1.0) 199 (1.0) 64 (1.7) 50.7 (11.7) 94 (25.8) 239 (65.5) 47.0 (7.4) 198 (54.3) 33 (10.1) 30 (8.2) 313 (85.8)
199 (.6) 13,514 (37.6) 21,692 (60.4) 504 (1.4)
81 (.4) 5,839 (30.6) 12,850 (67.3) 325 (1.7)
2 (.1) 571 (17.8 ) 2,485 (77.6) 145 (4.5)
0 (.0) 52 (14.3) 271 (74.5) 41 (11.3)
P value N/A o.001
o.001 o.001 o.001 o.001 o.001 o.001 o.001 o.001 o.001
eGFR ¼ estimated glomerular filtration rate; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy; BMI ¼ body mass index; CAD ¼ coronary artery disease; ASA ¼ American Society of Anesthesiologists. eGFR measured in mL/min/1.73 m2. * Patient with missing data: race 5892, gender 269, BMI 97, CAD 6444, smoker 1, ASA 67. † Unless otherwise indicated, values in brackets represent the percentage of patients in the respective subgroup. ‡ Previous history of cardiac bypass surgery or angioplasty. § Smoker within last year.
stages 4/5, and the trend across stages is statistically significant (P ¼ .001). Notably, there is no statistically significant relationship between kidney function and mortality in either the LAGB or the SG groups, although the numbers of patients in the higher stages of CKD are low. Table 3 shows the univariable and multivariable logistic regression models for overall complications by procedure across stages of CKD. Both the univariable and multivariable results are presented as OR while RD is illustrated graphically in Fig. 2, displaying results on both multiplicative and
additive scales. Considering the RYGB group first, after adjusting for important co-morbid conditions, all of the adjusted ORs reduce in magnitude and statistical significance from the unadjusted univariable ORs (i.e., closer to 1.0). For example, the unadjusted OR comparing stage 3 CKD (eGFR 30–59.9) to stage 1 (eGFR Z90) is 1.45 (95% CI: 1.25, 1.68; P o .001) and becomes 1.13 (95% CI: .96, 1.32; P ¼ .145) after adjustment with a RD .84% (95% CI: -.03, 2.00; P ¼ .145). This trend toward the null after multivariable adjustment also was seen in the LAGB group,
Table 2 30-day overall complications, major complications and mortality by procedure Procedure
Complications*
eGFR Z 90
eGFR 60–89.9
eGFR 30–59.9
eGFR o 30
P value
RYGB
N Overall Major Death N Overall Major Death N Overall Major Death
25,760 1,778 (6.9) 1,166 (4.5) 30 (.12) 11,641 278 (2.4) 174 (1.5) 3 (.03) 2,446 122 (5.0) 81 (3.3) 1 (0.04)
13,288 1,001 (7.5) 671 (5.1) 24 (.18) 6,464 182 (2.9) 121 (1.9) 1 (.02) 1,013 51 (5.0) 38 (3.8) 0 (.0)
2,209 214 (9.7) 151 (6.8) 11 (.5) 1,028 45 (4.4) 23 (2.2) 2 (.19) 173 13 (7.5) 8 (4.6) 0 (.0)
404 35 (8.7) 29 (7.2) 1 (.25) 199 8 (4.0) 4 (2.0) 0 (.0) 64 5 (7.8) 4 (6.3) 0 (.0)
N/A o.001 o.001 .001 N/A .001 .094 .081 N/A .383 .480 1.000
LAGB
SG
eGFR ¼ estimated glomerular filtration rate; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Values in brackets represent the percentage of patients in the respective subgroup.
832
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
Table 3 Univariable and multivariable logistic regression models of overall complications and renal function by procedure Procedure
RYGB
LAGB
SG
eGFR
Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30
Univariable analysis
Multivariable analysis*
OR
95% CI
P value
OR
95% CI
P value
1.00 1.10 1.45 1.28 1.00 1.20 1.87 1.71 1.00 1.01 1.55 1.61
Referent 1.01, 1.19 1.25, 1.68 .90, 1.82 Referent 1.00, 1.45 1.36, 2.58 .84, 3.51 Referent .72, 1.41 .85, 2.80 .64, 4.10
Referent .022 o.001 .168 Referent .055 o.001 .142 Referent .954 .149 .313
1.00 1.01 1.13 1.11 1.00 1.00 1.20 1.30 1.00 1.06 1.64 1.68
Referent .92, 1.10 .96, 1.32 .78, 1.58 Referent .81, 1.23 .84, 1.71 .63, 2.69 Referent .74, 1.51 .87, 3.12 .65, 4.32
Referent .864 .145 .552 Referent .977 .314 .473 Referent .755 .128 .280
eGFR ¼ estimated glomerular filtration rate; OR ¼ odds ratio; CI ¼ confidence interval; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Adjusting for American Society of Anesthesiologists (1 and 2 versus 3 and 4), hypertension, diabetic status, age, and body mass index.
but does not exist for overall complications for SG. Despite a slight increase in both the ORs and RDs, the results fail to achieve significance. For example, in the stage 4/5 CKD, the OR increases from 1.61 (95% CI: .64, 4.10; P ¼ .313) to 1.68 (95% CI: .65, 4.32; P ¼ .280) with an adjusted RD 3.08% (95% CI: -3.76, 9.93; P ¼ .280). Table 4 and Fig. 3 outline results of major complications, showing a similar pattern as demonstrated for overall complications. Again, for RYGB and LAGB, the OR and
RD come closer to the null after adjusting for important confounders compared to the unadjusted results, while those of the SG slightly increase. Additionally, the ORs again either remain not statistically significant or lose significance after adjustment, with the exception of Stage 3 patients compared to Stage 1 in the RYGB procedure, OR 1.22 (95% CI: 1.01, 1.47; P ¼ .042) and RD .96% (95% CI: .03, 1.96; P ¼ .042). Interestingly, in the LAGB group, the adjusted results show that patients are at comparable risk for complications across categories of kidney function. Table 5 shows an analysis of CKD stage 4/5 subcohort and compares unadjusted outcomes across the 3 procedures. Although there are too few patients to appropriately perform an adjusted analysis, it appears the LAGB have the lowest incidence of 30-day complications, followed by the SG and RYGB. When patients with missing race were excluded, the results showed similar findings to the primary analysis. Discussion Before adjusting for important patient characteristics, our univariable analysis showed a trend for increasing overall and major complications as well as death across advancing stages of CKD. Although this general trend of increasing complications remained across all procedure groups after adjusting for patient co-morbidities, the odds ratios and risk differences for complications were more attenuated in the RYGB and LAGB groups and slightly increased in the SG group. Importantly, all but 1 of the adjusted analyses did not reach statistical significance, the exception being the association between major complications between CKD stage 3 compared to CKD stage 1 in the RYGB group. In essence,
Fig. 2. Forrest plot of adjusted risk differences for overall complications. RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835 Table 4 Univariable and multivariable logistic regression models of major complications and renal function by procedure Procedure eGFR
RYGB
LAGB
SG
Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30 Z90 60–89.9 30–59.9 o30
Univariable analysis
Multivariable analysis*
OR 95% CI
P value OR 95% CI
P value
1.00 1.12 1.55 1.63 1.00 1.28 1.51 1.35 1.00 1.14 1.42 1.94
Referent .021 o.001 .012 Referent .040 .067 .555 Referent .519 .359 .208
Referent .695 .042 .068 Referent .827 .774 1.000 Referent .375 .308 .191
Referent 1.02, 1.24 1.30, 1.84 1.11, 2.39 Referent 1.01, 1.61 .97, 2.34 .50, 3.68 Referent 0.77, 1.68 .67, 2.98 .69, 5.49
1.00 1.02 1.22 1.43 1.00 1.03 0.93 1.00 1.00 1.21 1.51 2.02
Referent .92, 1.13 1.01, 1.47 .97, 2.10 Referent .80, 1.33 .58, 1.50 .36, 2.74 Referent .80, 1.84 .68, 3.36 .70, 5.78
eGFR ¼ estimated glomerular filtration rate; OR ¼ odds ratio; CI ¼ confidence interval; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy. eGFR measured in mL/min/1.73 m2. * Adjusting for American Society of Anesthesiologists (1 and 2 versus 3 and 4), hypertension, diabetic status, age, and body mass index.
although kidney dysfunction may be associated with increased complications, it seems that this increase is very modest. In fact, much of the morbidity in these patients may result from case mix rather than reduced kidney function per se. Turgeon et al. have previously reported on outcomes and stages of CKD using the ACS NSQIP data between the years 2006 and 2008 [37]. Although our study is similar,
833
Table 5 Unadjusted comparison of outcomes and bariatric procedures in CKD class 4/5 patients Variable n ¼ 667
RYGB 404 (60.6)
LAGB 199 (29.8)
SG 64 (9.6)
P value NA
Any complication Major complication Death
35 (8.7) 29 (7.2) 1 (.3)
8 (4.0) 4 (2.0) 0
5 (7.8) 4 (6.3) 0
.114 .032 .722
CKD ¼ chronic kidney disease; RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; SG ¼ laparoscopic sleeve gastrectomy.
there are a number of important differences. First, we extended the cohort by adding the years 2009–2011. Additionally, we limited procedures to the laparoscopic approach. Open bariatric surgery results in significantly higher overall morbidity and mortality [22]. Additionally, they included revisional procedures which are well known to be a very distinct surgical group with higher complication rates [23]. Finally, and most importantly, we separately examined the various bariatric procedures RYGB, SG, and LAGB, which all have different complication rates [24]. These differences may explain the stronger associations between stages of CKD and complications found in the study by Turgeon et al. A number of small case series have been published demonstrating that bariatric surgery appears to be safe in end-stage renal disease (ESRD) patients [5,25–28] while other evidence suggests higher complication rates in patients with ESRD [29,30]. In a review of bariatric surgery using the United States Renal Data System from
Fig. 3. Forrest plot of adjusted risk differences for major complications. RYGB ¼ Roux-en-Y gastric bypass; LAGB ¼ laparoscopic adjustable gastric band; LSG ¼ laparoscopic sleeve gastrectomy.
834
F. Saleh et al. / Surgery for Obesity and Related Diseases 11 (2015) 828–835
1991–2004, there was a 3.5% mortality rate in the 116 patients with ESRD who had bariatric surgery, although the vast majority had bariatric surgery after transplantation and the cohort likely had a large number of open surgery given the years of the study [30]. Bariatric surgery has many attractive features as a treatment for obesity in ESRD patients. Obesity is highly prevalent in patients with ESRD and reflects the prevalence seen in the general population (i.e., 60% obese or overweight) [31]. Additionally, obese patients with ESRD who undergo kidney transplantation are subject to weight gain after transplantation with the potential to exacerbate the adverse health effects of their obesity and result in decreased long-term graft survival and poorer outcomes [12,32]. In fact, morbid obesity (BMI 4 36) was associated with a greater than 30% risk of graft loss compared to a BMI from 24–26, delayed graft function, increased cardiovascular death, and overall mortality with a functioning graft [11]. Weight loss after bariatric surgery in this patient population as well as the improvements in obesity-related illnesses mimics that of patients without CKD. Additionally, the effectiveness of transplant immunosuppressive drug regimens do not appear to be affected by the altered digestive systems of postbariatric patients given that many are driven by drug levels [5,25,33]. In addition to the many potential benefits that weight loss surgery has on the health status of obese individuals, bariatric surgery has the potential to improve kidney function in patients with CKD. In a study by Alexander and Goodman [25], 5 patients of 32 with CKD, the majority of whom were on dialysis for a mean of 2.5 years, had resolution or stabilization of their kidney disease, some of who came off the waiting list for transplantation [5]. In addition, many patients who may not have been candidates for kidney transplantation because of their weight or comorbid disease status, may become eligible after bariatric surgery [13,28]. Bariatric surgery has been shown to result in sustained weight loss [34] and results in improvement or resolution of many obesity related co-morbidities [2,25]. On average across the different bariatric procedures for all comers, resolution can be seen in up to 75% of diabetics, 60% hypertensives, and 85% of patients with obstructive sleep apnea [2] with evidence showing increased longevity in morbidly obese patients after bariatric surgery [34–36]. The extent to which these benefits extend to ESRD patients remains to be seen. There are several limitations to this study. First, because of the observational nature of this study, we are unable to control for all important confounders that could affect patient outcomes including, but not limited to, perioperative hospital care, hospital volume, and surgeon experience. Secondly, the selection of patients for surgery based on factors that are not captured in the database may lead to selection bias in the association between renal function and complications. Thirdly, ACS NSQIP does not collect data
on procedure-specific outcomes or outcomes beyond 30 days including anastomotic strictures and internal herniation but it does capture the complications relevant to this study, especially as it pertains to safety. Finally, despite the large total sample size, the number of events among patients in the most advanced stages of CKD (i.e., stages 4 and 5) was small and thus we may have missed smaller effects in the lower range of kidney function. In conclusion, more advanced stages of CKD do not appear to be statistically associated with an increased risk of 30-day postoperative complications. The excess morbidity observed in this study may be driven by the co-morbid disease status of the patients. Further research is necessary to help physicians appropriately select patients with CKD who would most benefit from bariatric surgery. Additionally, further studies are needed to assess the optimal procedure type and timing for surgery in these patients. Disclosures The authors have no commercial associations that might be a conflict of interest in relation to this article. References [1] Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 19992004. JAMA 2006;295:1549–55. [2] Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724–37. [3] Kalaitzidis RG, Siamopoulos KC. The role of obesity in kidney disease: recent findings and potential mechanisms. Int Urol Nephrol 2011;43:771–84. [4] Agnani S, Vachharajani VT, Gupta R, Atray N, Vachharajani TJ. Does treating obesity stabilize chronic kidney disease? BMC Nephrol 2005;6:7. [5] Alexander JW, Goodman HR, Gersin K, et al. Gastric bypass in morbidly obese patients with chronic renal failure and kidney transplant. Transplantation 2004;78:469–74. [6] Gajdos C, Hawn MT, Kile D, Robinson TN, Henderson WG. Risk of major nonemergent inpatient general surgical procedures in patients on long-term dialysis. JAMA Surg 2013;148:137–43. [7] Cloyd JM, Ma Y, Morton JM, Tamura MK, Poultsides GA, Visser BC. Does chronic kidney disease affect outcomes after major abdominal surgery? Results from the National Surgical Quality Improvement Program. J Gastrointest Surg 2014;18:605–12. [8] Celebi-Onder S, Schmidt RJ, Holley JL. Treating the obese dialysis patient: challenges and paradoxes. Semin Dial 2012;25:311–9. [9] Degoulet P, Legrain M, Reach I, et al. Mortality risk factors in patients treated by chronic hemodialysis. Report of the Diaphane collaborative study. Nephron 1982;31:103–10. [10] el-Agroudy AE, Wafa EW, Gheith OE, Shehab el-Dein AB, Ghoneim MA. Weight gain after renal transplantation is a risk factor for patient and graft outcome. Transplantation 2004;77:1381–5. [11] Meier-Kriesche HU, Arndorfer JA, Kaplan B. The impact of body mass index on renal transplant outcomes: a significant independent risk factor for graft failure and patient death. Transplantation 2002;73: 70–4. [12] Curran SP, Famure O, Li Y, Kim SJ. Increased recipient body mass index is associated with acute rejection and other adverse outcomes after kidney transplantation. Transplantation 2014;97:64–70.
Bariatric Surgery and Kidney Function / Surgery for Obesity and Related Diseases 11 (2015) 828–835 [13] Segev DL, Simpkins CE, Thompson RE, Locke JE, Warren DS, Montgomery RA. Obesity impacts access to kidney transplantation. J Am Soc Nephrol 2008;19:349–55. [14] Khuri SF, Henderson WG, Daley J, et al. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007;204:1089–102. [15] Khuri SF, Daley J, Henderson W, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:315–27. [16] Daley J, Khuri SF, Henderson W, et al. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:328–40. [17] Daley J, Forbes MG, Young GJ, et al. Validating risk-adjusted surgical outcomes: site visit assessment of process and structure National VA Surgical Risk Study. J Am Coll Surg 1997;185:341–51. [18] Khuri SF, Henderson WG, Daley J, et al. Successful implementation of the Department of Veterans Affairs’ National Surgical Quality Improvement Program in the private sector: the patient safety in surgery study. Ann Surg 2008;248:329–36. [19] Shiloach M, Frencher SK Jr, Steeger JE, et al. Towards robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16. [20] Levey AS, Coresh J, Balk E, et al. National Kidney Foundation: National Kidney Foundation practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Ann Intern Med 2003;139:137–47. [21] Michels WM, Grootendorst DC, Verduijn M, Elliott EG, Dekker FW, Krediet RT. Performance of the Cockcroft-Gault, MDRD, and New CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol 2010;5:1003–9. [22] Hutter MM, Randall S, Khuri S, Henderson WG, Abbott WM, Warshaw AL. Laparoscopic versus open gastric bypass for morbid obesity: a multicenter, prospective, risk-adjusted analysis from the National Surgical Quality Improvement Program. Ann Surg 2006;243:657–62. [23] Hallowell PT, Stellato TA, Yao DA, Robinson A, Schuster MM, Graf KN. Should bariatric revisional surgery be avoided secondary to increased morbidity and mortality? Am J Surg 2009;197:391–6.
835
[24] Birkmeyer NJ, Dimick JB, Share D, et al. Hospital complication rates with bariatric surgery in Michigan. JAMA 2010;304:435–42. [25] Alexander JW, Goodman HR. Gastric bypass in chronic renal failure and renal transplant. Nutr Clin Pract 2007;22:16–21. [26] Koshy AN, Coombes JS, Wilkinson S, Fassett RG. Laparoscopic gastric banding surgery performed in obese dialysis patients prior to kidney transplantation. Am J Kidney Dis 2008;52:e15–7. [27] Newcombe V, Blanch A, Slater GH, Szoid A, Fielding GA. Laparoscopic adjustable gastric banding prior to renal transplantation. Obes Surg 2005;15:567–70. [28] Takata MC, Campos GM, Ciovica R, et al. Laparoscopic bariatric surgery improves candidacy in morbidly obese patients awaiting transplantation. Surg Obes Relat Dis 2008;4:159–65. [29] MacLaughlin HL, Hall WL, Patel AG, Macdougall IC. Laparoscopic sleeve gastrectomy is a novel and effective treatment for obesity in patients with chronic kidney disease. Obes Surg 2012;22:119–23. [30] Modanlou KA, Muthyala U, Xiao H, et al. Bariatric surgery among kidney transplant candidates and recipients: analysis of the United States renal data system and literature review. Transplantation 2009;87:1167–73. [31] Friedman AN, Miskulin DC, Rosenberg IH, Levey AS. Demographics and trends in overweight and obesity in patients at time of kidney transplantation. Am J Kidney Dis 2003;41:480–7. [32] Djamali A, Samaniego M, Muth B, et al. Medical care of kidney transplant recipients after the first posttransplant year. Clin J Am Soc Nephrol 2006;1:623–40. [33] Currie A, Chetwood A, Ahmed AR. Bariatric surgery and renal function. Obes Surg 2011;21:528–39. [34] Sjostrom L, Narbro K, Sjostrom CD, et al. Swedish Obese Subjects Study. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med 2007;357:741–52. [35] Christou NV, Sampalis JS, Liberman M, et al. Surgery decreases long-term mortality, morbidity, and health care use in morbidly obese patients. Ann Surg 2004;240:416–24. [36] Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA 2003;289:187–93. [37] Turgeon NA, Perez S, Mondestin M, et al. The impact of renal function on outcomes of bariatric surgery. J Am Soc Nephrol 2012; 23:885–94.
