- Journal List
- HHS Author Manuscripts
- PMC6679992
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice
J Surg Res. Author manuscript; available in PMC 2020 Oct 1.
Published in final edited form as:
J Surg Res. 2019 Oct; 242: 136–144.
Published online 2019 May 8. doi:10.1016/j.jss.2019.04.047
PMCID: PMC6679992
NIHMSID: NIHMS1527145
PMID: 31077945
Natalie Liu, MD,1,* Meghan C. Cusack, BA,2,* Manasa Venkatesh, MA, MS,1 Anisa L. Pontes,3 Grace Shea, BS,3 Dillon C. Svoboda, BS,3 Jacob A. Greenberg, MD, EdM,1 Anne O. Lidor, MD, MPH,1 and Luke M. Funk, MD, MPH1,4
Author information Copyright and License information PMC Disclaimer
The publisher's final edited version of this article is available at J Surg Res
Associated Data
- Supplementary Materials
Abstract
Background:
Intraoperative testing of anastomoses and staples lines is commonly performed to minimize the risk of postoperative leaks in bariatric surgery, but its impact is unclear. The aim of this study was to determine the association between leak testing and 30-day postoperative leak, bleed, reoperation, and readmission rates for patients undergoing laparoscopic sleeve gastrectomy (LSG) or Roux-en-Y gastric bypass (RYGB).
Methods:
This is a retrospective observational study utilizing 2015–2016 data from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) database. Postoperative outcomes were compared using χ2-test. Multivariable logistic regression was used to identify factors associated with 30-day outcomes.
Results:
We included 237,081 patients. Leak testing was performed on 73.0% and 92.1% of LSG and RYGB patients, respectively. LSG was associated with lower rates of leak, bleed, reoperation, and readmission than RYGB. On multivariable analysis, intraoperative leak testing was associated with increased rates of postoperative leak for LSG and RYGB (OR 1.48 and 1.90 respectively) and lower rates of bleed for LSG (OR 0.76). There were no significant associations between leak testing and rates of reoperation or readmission.
Conclusion:
Use of intraoperative leak testing was not associated with improved outcomes for either LSG or RYGB. A prospective trial investigating leak testing is warranted to better elucidate its impact.
Keywords: bariatric surgery, anastomotic leak, reoperation, readmission, bleed, intraoperative provocative testing
Introduction
Bariatric surgery is the most effective treatment for severe obesity, defined as having a BMI ≥ 40 kg/m2, or ≥ 35 kg/m2 with an obesity-related health condition (e.g. diabetes, cardiovascular disease, high blood pressure, obstructive sleep apnea, gastroesophageal reflux disease). It results in greater rates of weight loss, resolution of comorbidities, decreased overall mortality, and improved quality of life when compared to medical weight loss strategies alone.1 The two most commonly performed bariatric procedures are laparoscopic sleeve gastrectomy (LSG) and laparoscopic Roux-en-Y gastric bypass (RYGB). The safety profiles of bariatric surgery have improved significantly over the past several decades. Mortality rates have decreased ten-fold, from 3% in the 1990s to roughly 0.3% presently.2,3
Despite this improvement, complications such as postoperative anastomotic and staple line leaks and bleeds continue to occur. Rates of leak and bleed have been estimated to be as high as 8.3% and 4.4%, respectively, depending on procedure performed.4,5 One technique used to minimize the risk of anastomotic/staple line leak is intraoperative provocative testing, or “leak testing.” The purpose of intraoperative leak testing is to evaluate the integrity of anastomoses and staple lines and, when endoscopy is used, to assess for intraluminal bleeding. Methods for intraoperative leak testing include endoscopic or nasogastric insufflation with air, saline, or methylene blue dye. Evidence to support the use of leak testing to prevent postoperative complications is mixed. Numerous retrospective studies have shown an association between leak testing and lower rates of anastomotic leak postoperatively.6,7 In contrast, a recently published retrospective study using national data to determine predictors of leak after bariatric surgery demonstrated an association between leak testing and increased rates of postoperative leak.8 Additionally, there is a gap in the literature on the impact of leak testing on complications beyond anastomotic leaks; no studies have examined the association between leak testing and other surgical outcomes, such as postoperative bleeding, reoperation, and readmission.
The objective of this study was to evaluate the association between intraoperative leak testing and 30-day outcomes in a national sample of bariatric surgery patients. We also sought to identify factors associated with postoperative leak, bleed, reoperation, and readmission in patients undergoing LSG or RYGB.
Methods
Data Source
We used data from the Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP) database. The MBSAQIP database reports more than 200 variables containing preoperative, intraoperative, and 30-day postoperative information on bariatric surgery patients from over 700 accredited institutions across the United States and Canada.9 Data integrity was maintained through prospective data entry by trained reviewers and through periodic institutional audits. All data were de-identified, and therefore this study was deemed exempt from the University of Wisconsin Institutional Review Board approval process.
Study Population
Patients in the MBSAQIP database who underwent either laparoscopic sleeve gastrectomy (CPT code 437775) or Roux-en-Y gastric bypass (CPT codes 43644 and 43645) in 2015 and 2016 were included in the analysis. We excluded patients under 18 years of age and over 80 years of age, as well as any patients with missing 30-day follow-up information. Patients were also excluded if they underwent any of the following procedures: mini-loop gastric bypass procedure, gastric plication procedure, endoscopic therapy, intragastric balloon procedure, clinical trial/experimental procedures, previous bariatric or foregut surgery, emergency cases, concurrent procedures, and all surgical approaches besides conventional laparoscopy.
Study Variables
Patient race/ethnicity was categorized as white (non-Hispanic), black (non-Hispanic), Hispanic, other (non-Hispanic), and unknown. Surgery type was classified as either laparoscopic sleeve gastrectomy (LSG) or laparoscopic Roux-en-Y gastric bypass (RYGB). Preoperative obesity-related comorbidities including diabetes mellitus, gastroesophageal reflux, hypertension, hyperlipidemia, and obstructive sleep apnea were identified from the MBSAQIP database. Preoperative history included current smoking status, chronic obstructive pulmonary disease, oxygen dependence, chronic steroid use, and preoperative albumin levels. The American Society of Anesthesiologists’ (ASA) Physical Status Classification (I-V) was obtained from the MBSAQIP database.
30-day postoperative outcomes included leak, bleed, reoperation, and readmission. We defined postoperative leak according to previously published studies: having a drain present for more than 30 days, an organ space surgical site infection, or any leak-related 30-day readmission, reoperation, or intervention.8,10 Postoperative bleed was characterized as receiving a transfusion within 72 hours of surgery or having any bleed-related 30-day readmission, reoperation, or intervention.11 Intraoperative leak testing was defined by MBSAQIP to include any endoscopic or nasogastric air or methylene blue leak test.
Analysis
Postoperative outcomes for LSG and RYGB were compared using Pearson’s χ2 test. Bivariable logistic regression analysis was conducted comparing baseline patient characteristics and use of intraoperative leak testing (yes/no) with 30-day outcomes for both LSG and RYGB. Multivariable logistic regression analysis was subsequently used to identify factors associated with the primary outcomes of interest (30-day postoperative leak, bleed, reoperation, and readmission) for both LSG and RYGB. Multivariable analysis included variables associated with the outcome on bivariable analysis (at p ≤ 0.2), as well as the following a priori variables: intraoperative leak testing, age, sex, BMI, black race, Hispanic ethnicity, current smoking status, chronic steroid use, preoperative albumin level, diabetes, hypertension, gastroesophageal reflux, and obstructive sleep apnea. An adjusted odds ratio (OR) with a 95% confidence interval (CI) was calculated for each variable. Associations were deemed statistically significant with p ≤ 0.05. SAS version 9.4 was used to conduct the analysis. R version 3.4.0 was used to create the figures.
Results
Patient Characteristics
We included 237,081 total patients who met study inclusion criteria (Figure 1). Within our cohort, 167,954 (70.8%) and 69,127 (29.2%) underwent LSG and RYGB, respectively (Figure 1). The mean age of our cohort was 44.7 years, and mean BMI was 45.5 kg/m2 (Table 1). The majority of patients were female (79.2%) and white (74.0%). 72.9% of patients were ASA class III preoperatively. Hypertension was the most prevalent preoperative obesity-related comorbidity, affecting 49.1% of patients. The overall 30-day postoperative leak rate was 0.7% for our cohort.
Figure 1.
Study Cohort: STROBE Diagram
Table 1.
Patient Characteristics (n = 237,081)
| Characteristic | |
|---|---|
| Surgery type (n,%) | |
| Sleeve gastrectomy | 167,954 (70.84) |
| Roux-en-Y gastric bypass | 69,127 (29.16) |
| Demographics (mean, sd) | |
| Age, mean years | 44.72 (11.91) |
| BMI, mean kg/m2 | 45.48 (8.03) |
| Gender (n,%) | |
| Female | 187,819 (79.22) |
| Male | 49,262 (20.78) |
| Race/ethnicity (n,%) | |
| White, non-Hispanic | 142,440 (64.75) |
| Black, non-Hispanic | 37,211 (16.92) |
| Hispanic | 29,192 (13.27) |
| Other, non-Hispanic | 2,132 (0.97) |
| Unknown | 8,999 (4.09) |
| Obesity-related comorbidities (n,%) | |
| Diabetes mellitus | 63,093 (26.61) |
| Gastroesophageal reflux | 73,965 (31.20) |
| Hypertension | 116,445 (49.12) |
| Hyperlipidemia | 58,734 (24.77) |
| Obstructive sleep apnea | 90,148 (38.02) |
| Preoperative history | |
| Current smoker (n,%) | 20,739 (8.75) |
| Chronic obstructive pulmonary disease (n,%) | 4,161 (1.76) |
| Oxygen dependent (n,%) | 1,639 (0.69) |
| Chronic steroid use (n,%) | 3,874 (1.63) |
| Albumin (g/dl) (mean, sd) | 4.05 (0.38) |
| ASA class (n, %) | |
| I | 778 (0.33) |
| II | 54,538 (23.12) |
| III | 172,051 (72.93) |
| IV | 8,514 (3.61) |
| V | 16 (0.01) |
Open in a separate window
30-Day Postoperative Outcomes
Intraoperative leak testing was performed on 78.5% of patients, including 122,481 patients undergoing LSG (73.0%) and 63,652 patients undergoing RYGB (92.1%) (Table 2). For both LSG and RYGB patients, intraoperative leak testing was not associated with decreased rates of leak. Rather, intraoperative leak testing was associated with increased rates of leak for both LSG (0.5% with leak testing vs. 0.4%; p < 0.0001) and RYGB (1.2% with leak testing vs. 0.6%; p < 0.0001) patients. Overall, LSG patients experienced lower rates of 30-day leak, bleed, reoperation, and readmission, as well as shorter mean operative time, in comparison to RYGB patients (Table 2). For LSG, intraoperative leak testing was also associated with lower rates of postoperative bleeding (0.6% with leak testing vs. 0.8%; p < 0.001).
Table 2.
30-Day Postoperative Complications With and Without Leak Testing
| Sleeve gastrectomy (n = 167,954) | Gastric bypass (n = 69,127) | |||||
|---|---|---|---|---|---|---|
| Outcome (n,%) | Leak testing (n =122,481) | No leak testing (n = 45,473) | p-value | Leak testing (n = 63,652) | No leak testing (n= 5,475) | p-value |
| Leak | 606 (0.49) | 157 (0.35) | <0.0001 | 732 (1.15) | 35 (0.64) | <0.0001 |
| Bleeding | 721 (0.59) | 343 (0.75) | <0.0001 | 1095 (1.72) | 88 (1.61) | 0.0366 |
| Reoperation | 1032 (0.84) | 416 (0.91) | 0.0086 | 1412 (2.22) | 119 (2.27) | 0.0376 |
| Readmission | 3906 (3.19) | 1463 (3.22) | 0.0119 | 3793 (5.96) | 347 (6.34) | 0.0123 |
| Operative time | 74.69 (37.48) | 66.09 (31.05) | <0.0001 | 117.6 (52.34) | 108.6 (48.74) | <0.0001 |
Open in a separate window
Predictors of 30-Day Postoperative Outcomes
Results of bivariate analysis are shown in Appendix A–D. On multivariable analysis, intraoperative leak testing was not associated with decreased rates of leak for either LSG or RYGB (Figure 2). Intraoperative leak testing was associated with decreased rates of postoperative bleed for LSG (OR 0.76, 95% CI 0.65–0.89) but not for RYGB (Figure 3). There was no association between intraoperative leak testing and 30-day reoperations or readmissions (Figures 4–5).
Figure 2.
Multivariable Logistic Regression Model for Predictors of Leak
Figure 3.
Multivariable Logistic Regression Model for Predictors of Bleed
Figure 4.
Multivariable Logistic Regression Model for Predictors of Reoperation
Figure 5.
Multivariable Logistic Regression Model for Predictors of Readmission
Hispanic patients had increased rates of leak regardless of surgical procedure (OR 2.65, 95% CI 2.19–3.21, for LSG; OR 5.81, 95% CI 4.89–7.90, for RYGB), adjusting for confounding factors (Figure 2). Black patients had increased rates of reoperations (OR 1.25, 95% CI 1.06–1.46, for LSG; OR 1.30, 95% CI 1.10–1.55, for RYGB) and readmissions (OR 1.47, 95% CI 1.36–1.60, for LSG; OR 1.44, 95% CI 1.30–1.60, for RYGB), but no increased rates of leak or bleed (Figure2–5).
Discussion
Our findings suggest that the use of intraoperative leak testing was not associated with a lower risk of postoperative leak for patients undergoing laparoscopic LSG or RYGB. Overall, patients undergoing LSG had lower rates of postoperative leak, bleed, reoperation, and readmission compared to patients undergoing RYGB. Hispanic patients undergoing either LSG or RYGB had a significantly higher rates of leak than patients of other ethnicities. Furthermore, black and Hispanic patients both had higher rates of reoperation and readmission.
Using the 2015–2016 MBSAQIP database, we found the overall composite rate of postoperative leak within 30 days of surgery for LSG and RYGB to be 0.7%, much lower than previously cited rates of up to 8.3%.5 This finding is consistent with data from the American College of Surgeons-Bariatric Surgery Center Network accreditation program, which found rates of leak after LSG and RYGB to be 0.7% and 0.8%, respectively.12 This is further supported by a recently published retrospective cohort study also using MBSAQIP data by Alizadeh.8
We also found that leak testing was associated with higher rates of 30-day postoperative leak for LSG and RYGB patients. Bariatric patients who underwent leak testing were nearly twice as likely to experience a postoperative leak in comparison to those who did not undergo leak testing. Similarly, Alizadeh’s study demonstrated that leak testing was associated with an increased rate of postoperative leak for both LSG and RYGB patients.8 Likewise, Bingham’s retrospective cohort study of LSG patients and Sakran’s retrospective multicenter study of over 2,000 patients showed no decrease in the rate of leak with the use of leak testing.13,14 In contrast, multiple studies have shown an improvement in outcomes with leak testing. A review by Rached demonstrated that the use of intraoperative leak testing has high sensitivity and specificity for the detection of leak for LSG, with subsequent improvement in rates of postoperative leak.10 Similarly, for RYGB patients, retrospective studies of U.S. adults by Alasfar and Shin both found decreased rates of leak after RYGB using intraoperative leak testing.6,7
Our analysis also showed that overall, patients undergoing LSG had lower rates of postoperative leak, bleed, reoperation, and readmission than those undergoing RYGB, as well as shorter operative times, regardless of leak testing utilization. This finding is well supported by the literature.15–18 A meta-analysis by Li evaluated randomized controlled trials comparing LSG vs. RYGB and showed that RYGB had a higher incidence of postoperative complications, including leak, bleed, reoperation, readmission, and mortality.17 A more recent study using the MBSAQIP also showed that LSG had lower rates of leak, morbidity, and mortality.18 This is not a surprising conclusion, as technically, RYGB is a more complicated procedure with multiple anastomoses.5,19
One important finding from our study was that race and ethnicity were related to 30-day postoperative outcomes for bariatric surgery. Hispanic patients undergoing bariatric surgery were 3 to 6 times more likely to have a postoperative leak compared to other ethnicities. Black and Hispanic patients also had higher rates of reoperations and readmissions. The literature is mixed regarding racial/ethnic disparities in bariatric surgery outcomes. A prospective cohort study by Jambhekar evaluating readmission after LSG, as well as Gribben’s retrospective study of both LSG and RYGB patients using ACS-NSQIP data, showed that neither race nor ethnicity were predictors of reoperation or readmission after bariatric surgery.20,21 In contrast, a retrospective study using data from the Nationwide Inpatient Sample of 115,507 US adults undergoing any bariatric procedure found that black patients had higher in-hospital mortality than non-black patients.22 Disparities in surgical outcomes have also been reported for patients with cancer and end-organ failure, as well as patients undergoing surgery for cancer, cardiovascular disease, and orthopedic conditions.23
Our findings suggest that the use of intraoperative leak testing is not associated with improved short-term outcomes for bariatric surgery patients. Given that our study was observational, there may be unmeasured confounding, such as indications for leak testing, that biased our findings. As the data demonstrate, the vast majority of surgeons – nearly 78.5% of cases (72.9% of LSG and 92.1% of RYGB) – perform leak testing, despite leak testing being associated with increased rates of leak. Thus, more information is needed on the different types of leak testing, such as endoscopic vs. nasogastric/orogastric air insufflation vs. methylene blue dye, as this has the potential to impact utility and outcomes. It is also theoretically possible that use of leak testing inadvertently damages anastomoses and staple lines. As discussed by Alizadeh and colleagues, blind placement and manipulation of nasogastric/orogastric tubes may result in occult anastomotic/staple line injuries, subsequently leading to postoperative leaks.8 To better understand if leak testing is beneficial for bariatric surgery patients, a prospective trial that that evaluates the impact of specific types of leak testing would be useful. We also identified significant racial/ethnic disparities in 30-day outcomes after bariatric surgery for black and Hispanic patients. This may be a reflection of socioeconomic differences, resulting in unequal access to hospitals and surgeons that achieve optimal outcomes.23 Alternatively, Nguyen’s retrospective study from the Nationwide Inpatient Sample suggests that racial disparities in bariatric surgery patients may stem from delayed surgery due to cultural or systemic issues, thus resulting in greater degrees of obesity and more severe comorbidities at the time of surgery.22 To address this issue, qualitative studies investigating reasons for these racial/ethnic inequalities would be informative.
Our study has several limitations. We only report 30-day postoperative outcomes, as long-term outcomes are not available through the MBSAQIP database. Additionally, the MBSAQIP database does not allow specification of the type of leak testing. All types of reported leak testing (endoscopic vs. nasogastric, air insufflation vs. methylene blue dye) are categorized under a single “provocative testing” variable. This variable does not include additional information about indications for leak testing (routine vs. selective based on surgeon concern), nor does it indicate how leaks discovered on leak testing were managed intraoperatively. Moreover, the MBSAQIP database does not include specific outcome variables for leak and bleed. Definitions for leak and bleed were generated from composite variables available through the database. However, we used similar definitions as other published studies, so these results are comparable to other studies evaluating these outcomes.8,10,11
Conclusion
In summary, the use of intraoperative leak testing was not associated with decreased rates of postoperative leak, bleed, reoperation, or readmission for either LSG or RYGB patients. However, no definitive conclusions can be drawn about whether or not leak testing should be used as we do not have specific data on the individual types of leak testing and their impact on outcomes. Therefore, a prospective trial investigating specific types of leak testing and their impact on outcomes, is needed to determine which form, if any, of leak testing may be beneficial.
Supplementary Material
1
Click here to view.(50K, doc)
2
Click here to view.(50K, doc)
3
Click here to view.(50K, doc)
4
Click here to view.(50K, doc)
Acknowledgements
Efforts on this study and manuscript were made possible by an American College of Surgeons George H.A. Clowes Career Development Award and a VA Career Development Award to Dr. Funk (CDA 015–060). The views represented in this article represent those of the authors and not those of the DVA or the US Government. Additionally, this study is funded through the NIH T32 Surgical Oncology Research Training Program (T32 CA090217–17).
Efforts on this study and manuscript were made possible by an American College of Surgeons George H.A. Clowes Career Development Award and a VA Career Development Award to Dr. Funk (CDA 015–060). The views represented in this article represent those of the authors and not those of the DVA or the US Government. Additionally, this study is funded through the NIH T32 Surgical Oncology Research Training Program (T32 CA090217-17).
Appendix A. Bivariate and Multivariable Predictors of 30-Day Postoperative Leak
| Sleeve gastrectomy | Roux-en-Y gastric bypass | |||||
|---|---|---|---|---|---|---|
| Bivariate analysis | Multivariable analysis | Bivariate analysis | Multivariable analysis | |||
| Variable | Odds ratio | p-value | Odds ratio (95% CI) | Odds ratio | p-value | Odds ratio (95% CI) |
| Leak testing | 1.416 | <.0001 | 1.482 (1.202 – 1.827) | 1.747 | 0.0008 | 1.897 (1.317 – 2.732) |
| Age | 1.007 | 0.0166 | 1.000 (0.992 – 1.008) | 1.001 | 0.7624 | 1.002 (0.994 – 1.010) |
| Sex | 1.082 | 0.3543 | 0.855 (0.691 – 1.057) | 1.118 | 0.1950 | 0.869 (0.713 – 1.060) |
| BMI | 0.995 | 0.2750 | 0.991 (0.980 – 1.002) | 1.006 | 0.1727 | 0.999 (0.990 – 1.009) |
| Race | ||||||
| Black | 0.765 | 0.2276 | 0.676 (0.513 – 0.891) | 0.848 | 0.0022 | 0.728 (0.527 – 1.005) |
| Hispanic | 2.082 | <.0001 | 2.650 (2.190 – 3.206) | 4.194 | <.0001 | 5.807 (4.888 – 6.898) |
| Non-Hispanic | 0.496 | 0.1875 | 0.657 (0.210 – 2.054) | 1.230 | 0.9205 | 1.348 (0.632 – 2.873) |
| Unknown | 0.818 | 0.5967 | 0.797 (0.466 – 1.364) | 0.743 | 0.0045 | 1.162 (0.677 – 1.996) |
| Obesity-related | ||||||
| comorbidities | ||||||
| Diabetes mellitus | 1.207 | 0.0199 | 1.088 (0.894 – 1.324) | 1.012 | 0.8704 | 0.843 (0.707 – 1.006) |
| Gastroesophageal reflux | 1.193 | 0.0211 | 1.151 (0.962 – 1.378) | 0.713 | <.0001 | 0.578 (0.482 – 0.694) |
| Hypertension | 1.188 | 0.0160 | 1.164 (0.963 – 1.406) | 1.292 | 0.0004 | 1.441 (1.198 – 1.735) |
| Obstructive sleep apnea | 1.327 | <.0001 | 1.412 (1.183 – 1.684) | 1.638 | <.0001 | 2.094 (1.769 – 2.478) |
| Preoperative history | ||||||
| Current smoker | 1.231 | 0.0727 | 1.127 (0.855 – 1.484) | 0.768 | 0.0619 | 0.649 (0.460 – 0.917) |
| Chronic steroid use | 1.552 | 0.0476 | 1.630 (1.002 – 2.653) | 0.983 | 0.9565 | 0.891 (0.439 – 1.808) |
| Albumin | 1.000 | 1.000 | 0.881 (0.707 – 1.097) | 0.917 | 0.3893 | 0.849 (0.688 – 1.047) |
Open in a separate window
Appendix B. Bivariate and Multivariable Predictors of 30-Day Postoperative Bleed
| Sleeve gastrectomy | Roux-en-Y gastric bypass | |||||
|---|---|---|---|---|---|---|
| Bivariate analysis | Multivariable analysis | Bivariate analysis | Multivariable analysis | |||
| Variable | Odds ratio | p-value | Odds ratio (95% CI) | Odds ratio | p-value | Odds ratio (95% CI) |
| Leak testing | 0.779 | 0.0001 | 0.760 (0.648 – 0.890) | 1.072 | 0.5363 | 1.100 (0.850 – 1.423) |
| Age | 1.029 | <.0001 | 1.019 (1.012 – 1.027() | 1.012 | <.0001 | 1.006 (0.999 – 1.013) |
| Sex | 1.175 | 0.0258 | 1.097 (0.917 – 1.311) | 1.435 | <.0001 | 1.296 (1.096 – 1.534) |
| BMI | 0.991 | 0.0250 | 0.989 (0.979 – 0.999) | 0.991 | 0.0228 | 0.988 (0.979 – 0.998) |
| Race | ||||||
| Black | 0.991 | 0.6340 | 1.075 (0.877 – 1.317) | 1.039 | 0.6494 | 1.086 (0.880 – 1.339) |
| Hispanic | 0.912 | 0.7480 | 1.136 (0.904 – 1.427) | 0.999 | 0.3691 | 1.063 (0.862 – 1.312) |
| Non-Hispanic | 0.890 | 0.8365 | 0.668 (0.249 – 1.792) | 1.497 | 0.0848 | 1.374 (0.786 – 2.401) |
| Unknown | 0.935 | 0.9456 | 1.105 (0.737 – 1.657) | 0.947 | 0.2755 | 1.084 (0.717 – 1.638) |
| Obesity-related | ||||||
| comorbidities | ||||||
| Diabetes mellitus | 1.657 | <.0001 | 1.262 (1.068 – 1.490) | 1.343 | <.0001 | 1.126 (0.965 – 1.313) |
| Gastroesophageal reflux | 1.504 | <.0001 | 1.180 (1.007 – 1.382) | 1.185 | 0.0044 | 1.099 (0.951 – 1.271) |
| Hypertension | 1.847 | <.0001 | 1.303 (1.097 – 1.549) | 1.346 | <.0001 | 1.016 (0.863 – 1.196) |
| Obstructive sleep apnea | 1.518 | <.0001 | 1.180 (1.008 – 1.383) | 1.181 | 0.0047 | 1.031 (0.888 – 1.197) |
| Preoperative history | ||||||
| Current smoker | 1.107 | 0.3303 | 1.328 (1.046 – 1.686) | 1.078 | 0.4595 | 1.174 (0.927 – 1.487) |
| Chronic steroid use | 1.707 | 0.0035 | 1.398 (0.893 – 2.189) | 1.357 | 0.1611 | 1.071 (0.615 – 1.866) |
| Albumin | 0.997 | 0.9739 | 0.981 (0.806 – 1.195) | 1.155 | 0.1025 | 1.078 (0.896 – 1.295) |
Open in a separate window
Appendix C. Bivariate and Multivariable Predictors of 30-Day Reoperation
| Sleeve gastrectomy | Roux-en-Y gastric bypass | |||||
|---|---|---|---|---|---|---|
| Bivariate analysis | Multivariable analysis | Bivariate analysis | Multivariable analysis | |||
| Variable | Odds ratio | p-value | Odds ratio (95% CI) | Odds ratio | p-value | Odds ratio (95% CI) |
| Leak testing | 0.920 | 0.1548 | 0.894 (0.780 – 1.025) | 1.021 | 0.8303 | 0.970 (0.782 – 1.202) |
| Age | 1.009 | <.0001 | 1.004 (0.998 – 1.010) | 1.005 | 0.0234 | 1.005 (0.999 – 1.011) |
| Sex | 1.195 | 0.0039 | 1.160 (0.999 – 1.348) | 1.010 | 0.8714 | 1.042 (0.889 – 1.220) |
| BMI | 1.004 | 0.1856 | 1.004 (0.996 – 1.012) | 0.994 | 0.0524 | 0.991 (0.983 – 0.999) |
| Race | ||||||
| Black | 1.160 | 0.0289 | 1.247 (1.063 – 1.462) | 1.235 | 0.0041 | 1.304 (1.100 – 1.546) |
| Hispanic | 0.985 | 0.7728 | 1.135 (0.942 – 1.368) | 0.854 | 0.1043 | 0.937 (0.772 – 1.137) |
| Non-Hispanic | 0.847 | 0.6289 | 0.574 (0.237 – 1.387) | 0.899 | 0.6744 | 0.939 (0.514 – 1.715) |
| Unknown | 0.836 | 0.3411 | 0.719 (0.477 – 1.083) | 0.967 | 0.8807 | 0.906 (0.610 – 1.346) |
| Obesity-related | ||||||
| comorbidities | ||||||
| Diabetes mellitus | 1.312 | <.0001 | 1.198 (1.038 – 1.383) | 0.988 | 0.8202 | 0.895 (0.780 – 1.028) |
| Gastroesophageal reflux | 1.299 | <.0001 | 1.243 (1.088 – 1.420) | 1.296 | <.0001 | 1.265 (1.115 – 1.435) |
| Hypertension | 1.407 | <.0001 | 1.185 (1.028 – 1.366) | 1.088 | 0.1036 | 0.976 (0.847 – 1.124) |
| Obstructive sleep apnea | 1.234 | <.0001 | 1.107 (0.969 – 1.266) | 1.048 | 0.3633 | 1.066 (0.935 – 1.215) |
| Preoperative history | ||||||
| Current smoker | 1.150 | 0.1129 | 1.138 (0.927 – 1.398) | 1.169 | 0.0736 | 1.107 (0.898 – 1.365) |
| Chronic steroid use | 1.325 | 0.1118 | 1.127 (0.736 – 1.726) | 1.235 | 0.2932 | 0.984 (0.596 – 1.623) |
| Albumin | 0.914 | 0.2483 | 0.949 (0.806 – 1.118) | 0.866 | 0.0628 | 0.851 (0.724 – 1.000) |
Open in a separate window
Appendix D. Bivariate and Multivariable Predictors of 30-Day Readmission
| Sleeve gastrectomy | Roux-en-Y gastric bypass | |||||
|---|---|---|---|---|---|---|
| Bivariate analysis | Multivariable analysis | Bivariate analysis | Multivariable analysi | |||
| Variable | Odds ratio | p-value | Odds ratio (95% CI) | Odds ratio | p-value | Odds ratio (95% CI) |
| Leak testing | 0.991 | 0.7697 | 1.006 (0.933 – 1.084) | 0.936 | 0.2569 | 0.970 (0.850 – 1.108) |
| Age | 1.000 | 0.9659 | 0.996 (0.993 – 1.000) | 0.998 | 0.2349 | 0.998 (0.995 – 1.002) |
| Sex | 0.916 | 0.0115 | 0.893 (0.818 – 0.974) | 0.873 | 0.0011 | 0.890 (0.805 – 0.985) |
| BMI | 1.012 | <.0001 | 1.006 (1.002 – 1.010) | 1.008 | <.0001 | 1.004 (0.999 – 1.008) |
| Race | ||||||
| Black | 1.593 | <.0001 | 1.472 (1.357 – 1.596) | 1.450 | <.0001 | 1.442 (1.303 – 1.597) |
| Hispanic | 1.027 | 0.6418 | 1.092 (0.986 – 1.210) | 1.000 | 0.5545 | 0.978 (0.870 – 1.100) |
| Non-Hispanic | 1.013 | 0.7702 | 0.907 (0.617 – 1.333) | 0.750 | 0.0755 | 0.929 (0.641 – 1.347) |
| Unknown | 0.775 | 0.0001 | 0.812 (0.659 – 1.002) | 0.783 | 0.0050 | 0.776 (0.596 – 1.010) |
| Obesity-related | ||||||
| comorbidities | ||||||
| Diabetes mellitus | 1.227 | <.0001 | 1.160 (1.073 – 1.254) | 1.094 | 0.0070 | 1.102 (1.013 – 1.199) |
| Gastroesophageal reflux | 1.315 | <.0001 | 1.338 (1.246 – 1.436) | 1.313 | <.0001 | 1.301 (1.203 – 1.406) |
| Hypertension | 1.270 | <.0001 | 1.179 (1.093 – 1.272) | 1.057 | 0.0868 | 0.984 (0.901 – 1.074) |
| Obstructive sleep apnea | 1.138 | <.0001 | 1.094 (1.018 – 1.176) | 1.070 | 0.0354 | 1.076 (0.992 – 1.167) |
| Preoperative history | ||||||
| Current smoker | 1.152 | 0.0022 | 1.085 (0.971 −1.212) | 1.130 | 0.0264 | 1.146 (1.010 – 1.301) |
| Chronic steroid use | 1.806 | <.0001 | 1.640 (1.348 – 1.996) | 1.299 | 0.0331 | 1.121 (0.835 – 1.504) |
| Albumin | 0.730 | <.0001 | 0.810 (0.742 – 0.883) | 0.778 | <.0001 | 0.829 (0.751 −0.916) |
Open in a separate window
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures: The Metabolic and Bariatric Surgery Accreditation and Quality Improvement Program (MBSAQIP): the hospitals participating in the MBSAQIP 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.
References
1. Golzarand M, Toolabi K, Farid R. The bariatric surgery and weight losing: a meta-analysis in the long- and very long-term effects of laparoscopic adjustable gastric banding, laparoscopic Roux-en-Y gastric bypass and laparoscopic sleeve gastrectomy on weight loss in adults. Surg Endosc. 2017;31(11):4331–4345. [PubMed] [Google Scholar]
2. Flum DR, Salem L, Elrod JAB, Dellinger EP, Cheadle A, Chan L. Early mortality among Medicare beneficiaries undergoing bariatric surgical procedures. JAMA. 2005;294(15):1903. [PubMed] [Google Scholar]
3. Nguyen NT, Paya M, Melinda Stevens C, Mavandadi S, Zainabadi K, Wilson SE. The relationship between hospital volume and outcome in bariatric surgery at academic medical centers. Ann Surg. 2004;240:586–594. [PMC free article] [PubMed] [Google Scholar]
4. Nguyen N, Blackstone R, Morton J, Ponce J, Rosenthal R, eds. The ASMBS Textbook of Bariatric Surgery: Volume 1: Bariatric Surgery.; 2015.
5. Kim J, Azagury D, Eisenberg D, Demaria E, Campos GM. ASMBS position statement on prevention, detection, and treatment of gastrointestinal leak after gastric bypass and sleeve gastrectomy, including the roles of imaging, surgical exploration, and nonoperative management. Surg Obes Relat Dis. 2015;11:739–748. [PubMed] [Google Scholar]
6. Alasfar F, Chand B. Intraoperative endoscopy for laparoscopic Roux-en-Y gastric bypass: Leak test and beyond. Surg Laparosc Endosc Percutan Tech. 2010;20(6):424–427. [PubMed] [Google Scholar]
7. Shin RB. Intraoperative endoscopic test resulting in no postoperative leaks from the gastric pouch and gastrojejunal anastomosis in 366 laparoscopic Roux-en-Y gastric bypasses. Obes Surg. 2004;14:1067–1069. [PubMed] [Google Scholar]
8. Fazl Alizadeh R, Li S, Inaba C, et al. Risk factors for gastrointestinal leak after bariatric surgery: MBASQIP analysis. J Am Coll Surg. 2018;227(1):135–141. [PubMed] [Google Scholar]
9. MBSAQIP: User Guide for the 2017 Participant Use Data File (PUF).; 2018. https://www.facs.org/quality-programs/mbsaqip/participant-use.AccessedJanuary 24, 2019. [Google Scholar]
10. Rached AA, Basile M, El Masri H. Gastric leaks post sleeve gastrectomy: review of its prevention and management. World J Gastroenterol. 2014;20(38):13904–13910. [PMC free article] [PubMed] [Google Scholar]
11. Janik MR, Mustafa RR, Rogula TG, Alhaj Saleh A, Abbas M, Khaitan L. Safety of laparoscopic sleeve gastrectomy and Roux-en-Y gastric bypass in elderly patients – analysis of the MBSAQIP. Surg Obes Relat Dis. 2018;14(9):1276–1282. [PubMed] [Google Scholar]
12. Hutter MM, Schirmer BD, Jones DB, et al. First report from the American College of Surgeons -Bariatric Surgery Center Network: laparoscopic sleeve gastrectomy has morbidity and effectiveness positioned between the band and the bypass. Ann Surg. 2011;254(3):410–422. [PMC free article] [PubMed] [Google Scholar]
13. Bingham J, Lallemand M, Barron M, et al. Routine intraoperative leak testing for sleeve gastrectomy: is the leak test full of hot air?Am J Surg. 2016;211:943–947. [PubMed] [Google Scholar]
14. Sakran N, Goitein D, Raziel A, et al. Gastric leaks after sleeve gastrectomy: a multicenter experience with 2,834 patients. Surg Endosc. 2013;27:240–245. [PubMed] [Google Scholar]
15. Kehagias I, Karamanakos SN, Argentou M, Kalfarentzos F. Randomized clinical trial of laparoscopic Roux-en-Y gastric bypass versus laparoscopic sleeve gastrectomy for the management of patients with BMI<50 kg/m2. Obes Surg. 2011;21:1650–1656. [PubMed] [Google Scholar]
16. Helmiö M, Victorzon M, Ovaska J, et al. SLEEVEPASS: A randomized prospective multicenter study comparing laparoscopic sleeve gastrectomy and gastric bypass in the treatment of morbid obesity: preliminary results. Surg Endosc. 2012;26:2521–2526. [PubMed] [Google Scholar]
17. Li JF, Lai DD, Ni B, Sun KX. Comparison of laparoscopic Roux-en-Y gastric bypass with laparoscopic sleeve gastrectomy for morbid obesity or type 2 diabetes mellitus: a meta-analysis of randomized controlled trials. Can J Surg. 2013;56(6):E158–E164. [PMC free article] [PubMed] [Google Scholar]
18. Kumar SB, Hamilton BC, Wood SG, Rogers SJ, Carter JT, Lin MY. Is laparoscopic sleeve gastrectomy safer than laparoscopic gastric bypass? A comparison of 30-day complications using the MBSAQIP data registry. Surg Obes Relat Dis. 2018;14(3):264–269. [PubMed] [Google Scholar]
19. Smith MD, Adeniji A, Wahed AS, et al. Technical factors associated with anastomotic leak after Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2014;11(2):313–320. [PMC free article] [PubMed] [Google Scholar]
20. Jambhekar A, Maselli A, Lindborg R, Kabata K, Tortolani A, Gorecki P. Readmission following laparoscopic sleeve gastrectomy. JSLS. 2016;20(3):1–6. [PMC free article] [PubMed] [Google Scholar]
21. Gribben JL, Ilonzo N, Neifert S, Leitman IM. Predictors of reoperation and failure to rescue in bariatric surgery. JSLS. 2018;22(1):1–7. [PMC free article] [PubMed] [Google Scholar]
22. Nguyen GC, Patel AM. Racial disparities in mortality in patients undergoing bariatric surgery in the USA. Obes Surg. 2013;23:1508–1514. [PubMed] [Google Scholar]
23. Epstein AJ, Gray BH, Schlesinger M. Racial and ethnic differences in the use of high-volume hospitals and surgeons. Arch Surg. 2010;145(2):179–186. [PubMed] [Google Scholar]
