- Email: [email protected]
Surgical Management of Complicated Diverticulitis: A Comparison of the Laparoscopic and Open Approaches
Surgical Management of Complicated Diverticulitis: A Comparison of the Laparoscopic and Open Approaches Tafari Mbadiwe, BASc, JD, Augustine C Obirieze, MBBS, MPH, Edward E Cornwell III, Patricia Turner, MD, FACS, Terrence M Fullum, MD, FACS
MD, FACS, FCCM,
Laparoscopy has become a commonly used method of performing colectomies, but the outcomes associated with laparoscopy in the emergency setting have not been well studied. STUDY DESIGN: The American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database was queried for patients with diverticulitis without hemorrhage who underwent a colectomy. Patient data retrieved included demographics and preoperative comorbidities. Each member of the cohort received either a primary anastomosis (PA) or a colostomy. Open and laparoscopic procedures were compared within these subgroups. Multivariate logistic regression analyses were performed to compare the risk-adjusted odds of postoperative morbidity and mortality for laparoscopic and open procedures. The riskadjusted impact of preoperative comorbidities was also assessed. RESULTS: A total of 11,981 patients in the database met the study criteria. The majority were female (53%) and Caucasian (82%), and the mean age was 58 (13) years. Comorbidities of the cardiovascular, pulmonary, or renal systems were present in 47%, 5%, and 1% of the cohort, respectively. On bivariate analysis, patients undergoing laparoscopy experienced lower rates of complications with both PA (14% vs 26%, p < 0.001) and colostomy (30% vs 37%, p ¼ 0.02). The laparoscopic approach was associated with decreased mortality rates for patients undergoing PA (0.24% vs 0.79%, p < 0.001). Multivariate analysis revealed that preoperative cardiovascular and pulmonary comorbidities were each associated with increased postoperative morbidity, and that the laparoscopic approach was associated with lower postoperative morbidity for patients undergoing PA. The reduced risk of death for patients undergoing laparoscopic PA (vs open approach) did not achieve statistical significance (odds ratio 0.68, p ¼ 0.3). A small number of patients underwent laparoscopic colostomy (n ¼ 237, 2.4%), and they did not have a significantly different risk of death. CONCLUSIONS: The laparoscopic approach is associated with lower complication rates compared with the open approach for the surgical treatment of diverticulitis with a primary anastomosis. (J Am Coll Surg 2013;216:782e790. 2013 by the American College of Surgeons)
BACKGROUND:
Colonic diverticulosis is a frequent occurrence among older Americans, and as many as 25% of these individuals may develop diverticulitis, typically as a consequence of a secondary infection stemming from diverticular luminal obstruction with inspissated fecal debris. Surgery is often indicated in the treatment of diverticulitis if the disease is complicated by perforation, fistula formation, or obstruction.1,2 Surgical intervention will usually involve resection of the diseased portion of the colon, but depending on a variety of factors specific to the patient’s condition, the surgeon may opt to create either an anastomosis or a diverting colostomy. In either case, the resection may be performed via an open or laparoscopic approach. Open approaches continue to be used in many elective
Disclosure Information: Nothing to disclose. Presented at the Southern Surgical Association 124th Annual Meeting, Palm Beach, FL, December 2012. Received February 4, 2013; Accepted February 5, 2013. From Howard University College of Medicine (Mbadiwe), the Outcomes Research Center, Department of Surgery (Obirieze, Cornwell), and the Department of Surgery (Cornwell, Fullum), Howard University College of Medicine, Washington, DC; and the Division of Minimally Invasive Surgery, University of Maryland School of Medicine, Baltimore, MD (Turner). Correspondence address: Terrence M Fullum, MD, FACS, Division of Minimally Invasive and Bariatric Surgery, Howard University Hospital, 2041 Georgia Ave Northwest, Tower Suite 4100-B, Washington DC, 20060. email: [email protected]
ª 2013 by the American College of Surgeons Published by Elsevier Inc.
782
ISSN 1072-7515/13/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.02.003
Vol. 216, No. 4, April 2013
Mbadiwe et al
Abbreviations and Acronyms
ACSNSQIP CC OR PA SSI
¼ American College of Surgeons-National Surgical Quality Improvement Program ¼ colectomy with colostomy ¼ odds ratio ¼ primary anastomosis ¼ surgical site infection
colonic resections performed in patients with diverticulitis;3 however, there is a growing body of evidence suggesting that laparoscopic approaches are associated with reduced postoperative morbidity and mortality and shorter hospital stays.4,5 In contrast, the outcomes associated with the choice of operation (primary anastomosis vs colostomy) and the approach (laparoscopic vs open) have not been well studied in the emergency setting. The aim of this study was to examine a large surgical database to compare the 30-day postoperative outcomes of patients who underwent open vs laparoscopic colon resection for complicated diverticulitis.
METHODS A retrospective study was conducted using data acquired from the American College of Surgeons National Surgical Quality Improvement (ACS-NSQIP) database between 2005 and 2009. The ACS-NSQIP database collects information from 237 academic and community hospitals nationwide. The data amassed quantify preoperative risk factors, intraoperative variables, and 30-day postoperative morbidity and mortality rates for patients undergoing surgical procedures at participating institutions.6 This study was limited to patients who received a colonic resection for treatment of complicated diverticulitis (ICD-9 code 562.11). Two types of resection procedures were analyzed: colectomy with primary anastomosis (PA) and colectomy with colostomy (CC). The timing of surgery (elective vs emergent) was also analyzed. Patients were selected for this study on the basis of American Medical Association Current Procedural Codes noted in the database: 44140, 44145, 44204, and 44207 for colectomy with primary anastomosis procedures and 44141, 44143, 44144, 44146, 44320, 44206, 44208, and 44188 for colectomy with colostomy procedures. Within each of the 2 procedure groups, the demographic information, comorbidities, and outcomes data were compared for patients who underwent an open approach (PA-Open or CC-Open) vs a laparoscopic approach (PA-Lap or CC-Lap). Patient information retrieved from the database included age, sex, race, body mass index (normal, overweight, or obese), smoking history, functional status before
Management of Complicated Diverticulitis
783
surgery (independent, partially dependent, or totally dependent), and preoperative comorbidities. Comorbidities were separated into 4 groups: cardiovascular (previous percutaneous coronary intervention, myocardial infarction within the last 6 months, history of congestive heart failure, angina within 1 month of operation, hypertension treated with medication, previous revascularization or amputation in response to peripheral vascular disease, pain at rest, or gangrene), pulmonary (current pneumonia, history of severe COPD, or dependence on ventilator), renal (acute renal failure or preoperative dialysis), and diabetes mellitus. Additionally, each procedure was classified as occurring in either an elective or emergency context. Bivariate analysis was performed using Pearson’s chisquare test for categorical variables and Student’s t-test for continuous variables. Multivariate regression analyses were performed to evaluate the association between surgical approach and patient outcomes while controlling for patient demographics, preoperative comorbidities, body mass index, functional status before surgery, smoking history, and case type. Understanding that the hemodynamic condition of the patient may have influenced the approach of the surgeon, we carried out further analysis using the propensity score method. For this, 2 separate analyses were conducted. The first used a multivariate logistic regression model to derive propensity scores regarding the probability that a patient would undergo a laparoscopic vs open approach given the emergent or elective timing of the case, and whether a PA or CC was performed. The propensity scores derived were then controlled for, in addition to other covariates, in the final outcomes analysis. Patient outcomes that were evaluated included postoperative mortality and postoperative complications. Postoperative complications were further identified as overall, major wound, respiratory, cardiovascular, and septic complications. Major complications included organ space surgical site infection (SSI), separation of layers of the surgical wound, whole blood or packed red cell transfusions (within 72 hours of surgery), ventilator-assisted respiration more than 48 hours after surgery, sepsis or septic shock, cardiac arrest, myocardial infarction, pulmonary embolism, stroke, renal insufficiency, acute renal failure, or a return to the operating room within 30 days of surgery. Respiratory complications included pneumonia, unplanned intubation, pulmonary embolism, or ventilator use for more than 48 hours. Wound complications included superficial SSI, deep incisional SSI, and organ space SSI wound disruption. Cardiovascular complications included cardiac arrest or myocardial infarction. Overall complications included any of the above complications.
784
Mbadiwe et al
Table 1. Characteristics of Patients Who Received Open Primary Anastomosis, Laparoscopic Primary Anastomosis, Open Colectomy with Colostomy, and Laparoscopic Colectomy with Colostomy (n ¼ 11,981) Characteristic
Sex Male Female Ethnicity Caucasian Black Hispanic Other Age, y <25 25e39 40e59 60e79 80e89 Death No Yes Cardiac comorbidities No Yes Pulmonary comorbidities No Yes Renal comorbidities No Yes Diabetes No Yes Weight class Normal weight Overweight Obese Any complications No Yes Major complications No Yes Sepsis No Yes Any respiratory complications No Yes
J Am Coll Surg
Management of Complicated Diverticulitis
n
%
5,663 6,317
47.3 52.3
9,818 574 685 904
82.0 4.8 5.7 7.6
29 952 5,579 4,634 724
0.2 8.0 46.6 38.7 6.0
11,971 190
98.4 1.6
6,305 5,676
52.6 47.4
11,392 589
95.1 4.9
11,852 129
98.9 1.1
10,858 1,123
90.6 9.4
2,832 4,255 4,435
23.6 35.5 37.0
9,196 2,785
76.8 23.3
10,469 1,512
87.4 12.6
11,288 693
94.2 5.8
11,321 660
94.5 5.5 (Continued)
Table 1. Continued Characteristic
Any cardiovascular complications No Yes Wound infection No Yes Functional health status before surgery Independent Partially dependent Totally dependent
n
%
11,890 91
99.2 0.8
10,260 1,721
85.6 14.4
11,317 542 122
94.5 4.5 1.0
All statistical analyses were performed using STATA/ MP Version 11.0 (StataCorp). Statistical significance was defined as p < 0.05.
RESULTS A total of 11,981 patients in the database had surgery for diverticulitis. The majority were female (53%) and Caucasian (82%), and the mean age was 58 (13 [SEM]) years. Comorbidities of the cardiovascular, pulmonary, or renal systems were present in 47%, 5%, and 1% of the cohort, respectively. Obesity and diabetes were present in 37% and 9% of the patients, respectively (Table 1). Laparoscopic colectomies were performed on 5,342 (45%) of patients; the remaining 6,639 (55%) received open colectomies. Among these groups, 5,105 (43%) had a laparoscopic primary anastomosis (PA-Lap), 4,158 (35%) underwent an open primary anastomosis (PA-Open), 237 (2%) underwent a laparoscopic colostomy (CC-Lap), and 2,481 (21%) received an open colostomy (CC-Open). Caucasian patients received PA-Lap, PA-Open, CC-Lap, and CC-Open procedures in 44%, 34%, 2%, and 20% of cases, respectively; by comparison, African-American patients underwent PA-Lap, PA-Open, CC-Lap, and CC-Open procedures in 27%, 45%, 2%, and 26% of cases, respectively (p < 0.001). In 10,035 (84%) patients, the surgery was classified as elective; the remaining 1,946 (16%) were categorized as emergent. Patients receiving emergency surgery underwent primary anastomosis in 20% of cases; 88% of elective patients underwent PA. The overall rate of postoperative complications was 23%, and the overall mortality rate was 1.6%. On bivariate analysis, patients undergoing laparoscopy experienced lower rates of complications with both PA (14% vs 26%, p < 0.001) and colostomy (30% vs 37%, p ¼ 0.02). The laparoscopic approach was associated with decreased mortality rates for patients undergoing PA (0.24% vs
Vol. 216, No. 4, April 2013
Management of Complicated Diverticulitis
Mbadiwe et al
785
Table 2. Bivariate Analysis of Patient Demographics, Comorbidities, Risk Factors, and Postoperative Complications (n ¼ 11,981) PA
CC
PA-Open Characteristic
Sex Male Female Ethnicity Caucasian Black Hispanic Other Age, y <25 25e39 40e59 60e79 80e89 Death No Yes Cardiovascular comorbidities No Yes Pulmonary comorbidities No Yes Renal comorbidities No Yes Diabetes No Yes Weight class Normal Overweight Obese Any complications No Yes Major complications No Yes Sepsis No Yes Any respiratory complications No Yes
PA-Lap
CC-Open
CC-Lap
n
%
n
%
n
%
n
%
1,871 2,287
45.0 55.0
2,485 2,620
48.7 51.3
1,197 1,284
48.2 51.8
110 126
46.6 53.4
3,325 258 220 355
78.0 6.2 5.3 8.5
4,293 155 312 345
84.1 3.0 6.1 6.8
2,003 152 137 189
80.7 6.1 5.5 7.6
197 9 16 15
83.1 3.8 6.8 6.3
7 316 1,926 1,664 236
0.27 7.6 46.4 40.1 5.7
12 471 2,703 1,805 108
0.2 9.2 53.0 35.4 2.1
8 146 856 1,075 352
0.3 6.0 35.1 44.1 14.4
2 19 94 90 28
0.9 7.8 38.4 42.4 11.4
4,125 33
99.2 0.8
5,093 12
99.8 0.2
2,344 137
94.5 5.5
229 8
96.6 3.4
2,091 2,067
50.3 49.7
3,036 2,063
59.5 40.5
1,061 1,420
42.8 57.2
117 120
49.4 50.6
3,978 180
95.7 4.3
5,007 98
98.1 1.9
2,189 292
88.2 11.8
218 19
92.0 8.0
4,125 33
99.2 0.8
5,096 9
99.8 0.2
2,396 85
96.6 3.4
235 2
99.2 0.8
3,734 424
89.8 10.2
4,762 343
93.3 6.7
2,162 319
87.1 12.9
200 37
84.4 15.6
962 1,432 1,644
23.8 35.5 40.7
1,195 1,961 1,864
23.8 39.1 37.1
620 779 849
27.6 34.7 37.8
57 83 78
26.2 38.1 35.8
3,083 1,075
74.2 25.9
4,391 714
86.0 13.9
1,555 926
62.7 37.3
167 70
70.5 29.5
3,650 508
87.8 12.2
4,783 322
93.7 6.3
1,844 637
74.3 25.7
192 45
81.0 19.0
3,921 237
94.3 5.7
4,980 125
97.6 2.4
2,164 317
87.2 12.8
223 14
94.1 5.9
3,975 183
95.6 4.4
5,036 69
98.7 1.3
2,092 389
84.3 15.7
218 92.0 19 8.0 (Continued)
786
Table 2.
Mbadiwe et al
J Am Coll Surg
Management of Complicated Diverticulitis
Continued PA
CC
PA-Open Characteristic
Any cardiovascular complications No Yes Wound infection No Yes Functional health status before surgery Independent Partially dependent Totally dependent Timing of surgery Elective Emergent
PA-Lap
CC-Open
CC-Lap
n
%
n
%
n
%
n
%
4,133 25
99.4 0.6
5,090 15
99.7 0.3
2,432 49
98.0 2.0
235 2
99.2 0.8
3,412 746
82.1 17.9
4,618 487
90.5 9.5
2,020 461
81.4 18.6
210 27
88.6 11.4
4,037 105 16
97.1 2.5 0.4
5,073 30 2
99.4 0.6 0.0
1,999 382 100
80.6 15.4 4.0
208 25 4
87.8 10.6 1.7
3,824 334
92.0 8.0
5,041 64
98.8 1.3
1,007 1,474
40.6 59.4
163 74
68.8 31.2
CC-Lap, laparoscopic colectomy with colostomy; CC-Open, open colectomy with colostomy; PA-Lap, laparoscopic primary anastomosis; PA-Open, open primary anastomosis.
0.79%, p < 0.001), and mortality differences for patients undergoing colostomy did not achieve statistical significance (3.4% vs 5.5%, p ¼ 0.16; Table 2). Multivariate analysis revealed that the laparoscopic approach was associated with lower odds of overall complications for patients undergoing PA (odds ratio [OR] 0.52, p < 0.001). The reduced risk of death for patients undergoing laparoscopic PA (vs open approach) did not achieve statistical significance (OR 0.68, p ¼ 0.3; Table 3). A small number of patients underwent laparoscopic colostomy (n ¼ 237, 2.0%), and there was no statistically significant difference in their risk of death or postoperative complications as compared with those receiving open colostomies (Table 4). Multivariate analyses suggested that, in the context of PA operations, cardiovascular (OR 1.25, p < 0.001) and pulmonary (OR 1.81, p < 0.001) comorbidities, and functional dependence (OR 2.12, p < 0.001), each appeared to be a significant risk factor for the development of overall complications; pulmonary disease (OR 2.86, p ¼ 0.025) and functional dependence (OR 17.4, p < 0.001) were associated with a higher risk of postoperative death. Further, for CC, pulmonary disease (OR 1.75, p < 0.001), presence of diabetes mellitus (OR 1.30, p ¼ 0.042), and functional dependence (OR 2.46, p < 0.001) significantly affected the risk of overall complications, and pulmonary (OR 2.86, p < 0.001) and renal (OR 2.29, p ¼ 0.024) comorbidities and functional dependence (OR 6.55, p < 0.001) were significant risk factors for mortality. When only patients who underwent elective operations were considered, laparoscopic operations were associated
with statistically significantly lower rates of overall complications (OR 0.53, p < 0.001; Table 5). Additionally, in the emergency context, the laparoscopic approach was not associated with lower rates of postoperative complications (OR 0.78, p ¼ 0.25), but was associated with lower rates of respiratory complications (OR 0.37, p ¼ 0.02; Table 6). The results of multivariate analyses using the entire study population and accounting for the preoperative likelihood that a patient would undergo a laparoscopic procedure (as indicated by each patient’s propensity score) suggested that patients receiving laparoscopic procedures are less likely to develop postoperative complications (OR 0.55, p < 0.001; Table 7). The difference in the mortality rates between the open and laparoscopic groups was not statistically significant in this analysis.
Table 3. Risk-Adjusted Analysis of Postoperative Complications Comparing Open and Laparoscopic Primary Anastomosis Procedures for Colon Diverticulitis Complication
PA-Open
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Reference Reference Reference Reference Reference Reference Reference
PA-Lap Odds ratio 95% CI
0.52 0.59 0.68 0.50 0.41 0.69 0.50
0.47e0.58 0.51e0.69 0.32e1.41 0.44e0.57 0.31e0.55 0.35e1.36 0.40e0.63
PA-Open, open primary anastomosis; PA-Lap, laparoscopic primary anastomosis.
Vol. 216, No. 4, April 2013
Mbadiwe et al
Table 4. Risk-Adjusted Analysis of Postoperative Complications Comparing Open and Laparoscopic Colectomy with Colostomy Procedures for Colon Diverticulitis Complication
CC-Open
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Reference Reference Reference Reference Reference Reference Reference
CC-Lap Odds ratio 95% CI
0.85 0.89 1.14 0.59 0.64 0.38 0.64
0.62e1.18 0.61e1.30 0.49e2.67 0.38e0.93 0.36e1.15 0.05e2.89 0.36e1.14
CC-Lap, laparoscopic colectomy with colostomy; CC-Open, open colectomy with colostomy.
Management of Complicated Diverticulitis
787
Table 5. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures Conducted on an Elective Basis Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Reference Reference
Laparoscopic Odds ratio 95% CI
0.53 0.61 0.84 0.50 0.45 0.69 0.51
0.48e0.60 0.52e0.71 0.44e1.61 0.45e0.57 0.34e0.60 0.36e1.31 0.41e0.64
DISCUSSION The benefits of the laparoscopic approach in the elective surgical management of patients with diverticular diseasedincluding decreased morbidity, mortality, and length of staydhave been well described.7-10 It remains an open question as to whether these benefits can be realized in the acute care surgical setting. To our knowledge, this is the largest series reported to date encompassing solely patients having surgery for diverticulitis. The statistical robustness achieved by using the ACSNSQIP database was countered by our inability to detect by diagnostic code the degree of physiologic insult experienced by patients whose diverticular disease grew sufficiently complicated to require surgical resection. Accordingly, we chose to use primary anastomosis and diverting colostomy as surrogates to represent the surgeon’s real-time judgment about patient stability. In this study, open approaches were used in 45% of patients undergoing PA and in 91% of those having CC. The notable difference in the frequencies with which laparoscopic approaches are adopted for primary anastomoses as compared with colostomies lends itself to the suggestion that many surgeons are more comfortable
using open approaches in situations in which a colostomy is required. This study demonstrated that the laparoscopic approach was associated with lower rates of postoperative complications in the PA group. The small number of patients undergoing the laparoscopic approach for CC (n ¼ 237, 2.0%) may help to explain why their lower risk of morbidity and mortality did not achieve statistical significance. Regardless of procedure, preoperative pulmonary and renal comorbidities were each associated with increased morbidity and mortality. Additional studies that more precisely delineate severity of disease on presentation are necessary, but these findings suggest to us that patients whose physiologic state will support PA in the acute management of complicated diverticulitis are candidates for the laparoscopic approach. African-American patients received laparoscopic PA less frequently than their Caucasian counterparts in our study. This discrepancy may speak to disparities in access to care and severity of disease, and merits future investigation. A limitation of this study was the unavailability of information in the ACS-NSQIP database concerning the hemodynamic status of patients at presentation. Surgeons may be predisposed to disproportionately select open (vs laparoscopic) and CC (vs PA) procedures for patients
Table 6. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures Conducted on an Emergency Basis
Table 7. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures and Adjusting for Preoperative Likelihood of Receiving a Laparoscopic Procedure, as Indicated by Propensity Scoring
Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Omitted Reference
Laparoscopic Odds ratio 95% CI
0.77 0.83 0.73 0.66 0.37
0.51e1.19 0.51e1.35 0.21e2.53 0.38e1.15 0.15e0.88
0.62
0.29e1.32
Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Reference Reference
Laparoscopic Odds ratio 95% CI
0.55 0.63 0.77 0.50 0.45 0.63 0.51
0.50e0.61 0.54e0.72 0.44e1.35 0.45e0.57 0.34e0.59 0.34e1.17 0.42e0.64
788
Mbadiwe et al
J Am Coll Surg
Discussion
with hemodynamic instability, which would bias the colostomy group toward higher rates of morbidity and mortality. The multivariate analysis performed in this study would not be able to account for such a bias, a problem we attempted to address by performing our analyses within each group (PA and CC). Also, the experience of the operating surgeon with performing laparoscopic procedures was not available.11
CONCLUSIONS In a large study of patients undergoing surgery for complicated diverticulitis, the laparoscopic approach was associated with a lower risk of total complications than the open approach for patients undergoing PA. Preoperative comorbidities are important associated risk factors for the occurrence of complications and death. These results suggest that the laparoscopic approach should be considered whenever PA is judged appropriate. Author Contributions Study conception and design: Mbadiwe, Obirieze, Cornwell, Fullum Acquisition of data: Mbadiwe, Obirieze, Cornwell, Turner, Fullum Analysis and interpretation of data: Mbadiwe, Obirieze, Cornwell, Turner, Fullum Drafting of manuscript: Mbadiwe, Obirieze, Cornwell, Fullum Critical revision: Mbadiwe, Obirieze, Cornwell, Turner, Fullum REFERENCES 1. Chapman J, Davies M, Wolff B, et al. Complicated diverticulitis: is it time to rethink the rules? Ann Surg 2005;242:576e583. 2. Rafferty J, Shellito P, Hyman NH, et al. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum 2006;49:939e944. 3. Eijsbouts QA, de Haan J, Berends F, et al. Laparoscopic elective treatment of diverticular disease. A comparison between laparoscopic-assisted and resection-facilitated techniques. Surg Endosc 2000;14:726e730. 4. Kakarla V, Nurkin S, Sharma S, et al. Elective laparoscopic versus open colectomy for diverticulosis: an analysis of ACSNSQIP database. Surg Endosc 2012;26:1837e1842. 5. Masoomi H, Buchberg B, Nguyen B, et al. Outcomes of laparoscopic versus open colectomy in elective surgery for diverticulitis. World J Surg 2011;35:2143e2148. 6. ACS NSQIP participant user data file user guide (2011). Available at: http://site.acsnsqip.org/downloads/. Accessed October 8, 2012. 7. Chen HH, Wexner SD, Weiss EG, et al. Laparoscopic colectomy for benign colorectal disease is associated with a significant reduction in disability as compared with laparotomy. Surg Endosc 1998;12:1397e1400. 8. Bergamaschi R, Arnaud JP. Immediately recognizable benefits and drawbacks after laparoscopic colon resection for benign disease. Surg Endosc 1997;11:802e804.
9. Sun J, Jiang T, Qiu Z, et al. Short-term and medium-term clinical outcomes of laparoscopic-assisted and open surgery for colorectal cancer: a single center retrospective case-control study. BMC Gastroenterol 2011;11:85. 10. Angriman I, Scarpa M, Ruffolo C. Health related quality of life after surgery for colonic diverticular disease. World J Gastroenterol 2010;16:4013e4018. 11. Waters J, Chihara R, Moreno J, et al. Laparoscopic colectomy: does the learning curve extend beyond colorectal surgery fellowship? JSLS 2010;14:325e331.
Discussion DR MICHAEL ROTONDO (Greenville, NC): The standardization of approach and outcomes managements in the emergency surgical province has become an intense area of focus over the last several years. We have seen specialties redefine themselves, like acute care surgery, and we have seen new societies emerge to greater prominence, like the European Society for Trauma and Emergency Surgery. Certainly, with the rising emphasis on the value proposition in health care, in which the relationship between quality and cost become paramount in measuring both patient outcomes and maintaining financial viability, questions such as the one before us become increasingly important. However, this work falls somewhat short of accurately comparing the efficacy of a laparoscopic approach and an open approach in a patient with complicated diverticulitis. It suffers from a lack of resolution of the very same confounding factors that are perplexing a group of surgeons convened through the American College of Surgeons, endeavoring to create an emergency general surgery data dictionary with the specific intent of developing a research and outcomes measurement tool for these types of patients. The group has a broad base of organizations represented, including the Advisory Council on General Surgery; the Society of Gastrointestinal and Endoscopic Surgeons (SAGES); the Surgical Infection Society; the Rural Surgery Committee; the Committee on Trauma, represented by Avery Nathans and the work he’s done with the Trauma Quality Improvement Program (TQIP); the American Association for the Surgery of Trauma (AAST) represented by Shaheed Shafi; the work that Wayne Meredith has gotten started; and, of course, Cliff Ko and the National Surgical Quality Improvement Program (NSQIP). And they are steadily working to devise a system in which emergency surgical entities, such as complicated diverticulitis, can be stratified by physiology at the time of presentation, the severity of the surgical entity, and comorbidities. It’s sort of like a TNM classification system for cancer that’s evolving. Without adequate standardized description of the disease process itself, comorbidities, and physiologic perturbations at the time of presentation, it is exceedingly difficult to measure outcomes in the efficacy of therapeutic interventions. And I’m afraid that the authors have not satisfactorily controlled or analyzed these confounding factors such that plausible conclusions can be reached. The presentation did help clarify some of the questions that I had in the manuscript, so it’s hard for me to know if this was
MD, FACS, FCCM,
Laparoscopy has become a commonly used method of performing colectomies, but the outcomes associated with laparoscopy in the emergency setting have not been well studied. STUDY DESIGN: The American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database was queried for patients with diverticulitis without hemorrhage who underwent a colectomy. Patient data retrieved included demographics and preoperative comorbidities. Each member of the cohort received either a primary anastomosis (PA) or a colostomy. Open and laparoscopic procedures were compared within these subgroups. Multivariate logistic regression analyses were performed to compare the risk-adjusted odds of postoperative morbidity and mortality for laparoscopic and open procedures. The riskadjusted impact of preoperative comorbidities was also assessed. RESULTS: A total of 11,981 patients in the database met the study criteria. The majority were female (53%) and Caucasian (82%), and the mean age was 58 (13) years. Comorbidities of the cardiovascular, pulmonary, or renal systems were present in 47%, 5%, and 1% of the cohort, respectively. On bivariate analysis, patients undergoing laparoscopy experienced lower rates of complications with both PA (14% vs 26%, p < 0.001) and colostomy (30% vs 37%, p ¼ 0.02). The laparoscopic approach was associated with decreased mortality rates for patients undergoing PA (0.24% vs 0.79%, p < 0.001). Multivariate analysis revealed that preoperative cardiovascular and pulmonary comorbidities were each associated with increased postoperative morbidity, and that the laparoscopic approach was associated with lower postoperative morbidity for patients undergoing PA. The reduced risk of death for patients undergoing laparoscopic PA (vs open approach) did not achieve statistical significance (odds ratio 0.68, p ¼ 0.3). A small number of patients underwent laparoscopic colostomy (n ¼ 237, 2.4%), and they did not have a significantly different risk of death. CONCLUSIONS: The laparoscopic approach is associated with lower complication rates compared with the open approach for the surgical treatment of diverticulitis with a primary anastomosis. (J Am Coll Surg 2013;216:782e790. 2013 by the American College of Surgeons)
BACKGROUND:
Colonic diverticulosis is a frequent occurrence among older Americans, and as many as 25% of these individuals may develop diverticulitis, typically as a consequence of a secondary infection stemming from diverticular luminal obstruction with inspissated fecal debris. Surgery is often indicated in the treatment of diverticulitis if the disease is complicated by perforation, fistula formation, or obstruction.1,2 Surgical intervention will usually involve resection of the diseased portion of the colon, but depending on a variety of factors specific to the patient’s condition, the surgeon may opt to create either an anastomosis or a diverting colostomy. In either case, the resection may be performed via an open or laparoscopic approach. Open approaches continue to be used in many elective
Disclosure Information: Nothing to disclose. Presented at the Southern Surgical Association 124th Annual Meeting, Palm Beach, FL, December 2012. Received February 4, 2013; Accepted February 5, 2013. From Howard University College of Medicine (Mbadiwe), the Outcomes Research Center, Department of Surgery (Obirieze, Cornwell), and the Department of Surgery (Cornwell, Fullum), Howard University College of Medicine, Washington, DC; and the Division of Minimally Invasive Surgery, University of Maryland School of Medicine, Baltimore, MD (Turner). Correspondence address: Terrence M Fullum, MD, FACS, Division of Minimally Invasive and Bariatric Surgery, Howard University Hospital, 2041 Georgia Ave Northwest, Tower Suite 4100-B, Washington DC, 20060. email: [email protected]
ª 2013 by the American College of Surgeons Published by Elsevier Inc.
782
ISSN 1072-7515/13/$36.00 http://dx.doi.org/10.1016/j.jamcollsurg.2013.02.003
Vol. 216, No. 4, April 2013
Mbadiwe et al
Abbreviations and Acronyms
ACSNSQIP CC OR PA SSI
¼ American College of Surgeons-National Surgical Quality Improvement Program ¼ colectomy with colostomy ¼ odds ratio ¼ primary anastomosis ¼ surgical site infection
colonic resections performed in patients with diverticulitis;3 however, there is a growing body of evidence suggesting that laparoscopic approaches are associated with reduced postoperative morbidity and mortality and shorter hospital stays.4,5 In contrast, the outcomes associated with the choice of operation (primary anastomosis vs colostomy) and the approach (laparoscopic vs open) have not been well studied in the emergency setting. The aim of this study was to examine a large surgical database to compare the 30-day postoperative outcomes of patients who underwent open vs laparoscopic colon resection for complicated diverticulitis.
METHODS A retrospective study was conducted using data acquired from the American College of Surgeons National Surgical Quality Improvement (ACS-NSQIP) database between 2005 and 2009. The ACS-NSQIP database collects information from 237 academic and community hospitals nationwide. The data amassed quantify preoperative risk factors, intraoperative variables, and 30-day postoperative morbidity and mortality rates for patients undergoing surgical procedures at participating institutions.6 This study was limited to patients who received a colonic resection for treatment of complicated diverticulitis (ICD-9 code 562.11). Two types of resection procedures were analyzed: colectomy with primary anastomosis (PA) and colectomy with colostomy (CC). The timing of surgery (elective vs emergent) was also analyzed. Patients were selected for this study on the basis of American Medical Association Current Procedural Codes noted in the database: 44140, 44145, 44204, and 44207 for colectomy with primary anastomosis procedures and 44141, 44143, 44144, 44146, 44320, 44206, 44208, and 44188 for colectomy with colostomy procedures. Within each of the 2 procedure groups, the demographic information, comorbidities, and outcomes data were compared for patients who underwent an open approach (PA-Open or CC-Open) vs a laparoscopic approach (PA-Lap or CC-Lap). Patient information retrieved from the database included age, sex, race, body mass index (normal, overweight, or obese), smoking history, functional status before
Management of Complicated Diverticulitis
783
surgery (independent, partially dependent, or totally dependent), and preoperative comorbidities. Comorbidities were separated into 4 groups: cardiovascular (previous percutaneous coronary intervention, myocardial infarction within the last 6 months, history of congestive heart failure, angina within 1 month of operation, hypertension treated with medication, previous revascularization or amputation in response to peripheral vascular disease, pain at rest, or gangrene), pulmonary (current pneumonia, history of severe COPD, or dependence on ventilator), renal (acute renal failure or preoperative dialysis), and diabetes mellitus. Additionally, each procedure was classified as occurring in either an elective or emergency context. Bivariate analysis was performed using Pearson’s chisquare test for categorical variables and Student’s t-test for continuous variables. Multivariate regression analyses were performed to evaluate the association between surgical approach and patient outcomes while controlling for patient demographics, preoperative comorbidities, body mass index, functional status before surgery, smoking history, and case type. Understanding that the hemodynamic condition of the patient may have influenced the approach of the surgeon, we carried out further analysis using the propensity score method. For this, 2 separate analyses were conducted. The first used a multivariate logistic regression model to derive propensity scores regarding the probability that a patient would undergo a laparoscopic vs open approach given the emergent or elective timing of the case, and whether a PA or CC was performed. The propensity scores derived were then controlled for, in addition to other covariates, in the final outcomes analysis. Patient outcomes that were evaluated included postoperative mortality and postoperative complications. Postoperative complications were further identified as overall, major wound, respiratory, cardiovascular, and septic complications. Major complications included organ space surgical site infection (SSI), separation of layers of the surgical wound, whole blood or packed red cell transfusions (within 72 hours of surgery), ventilator-assisted respiration more than 48 hours after surgery, sepsis or septic shock, cardiac arrest, myocardial infarction, pulmonary embolism, stroke, renal insufficiency, acute renal failure, or a return to the operating room within 30 days of surgery. Respiratory complications included pneumonia, unplanned intubation, pulmonary embolism, or ventilator use for more than 48 hours. Wound complications included superficial SSI, deep incisional SSI, and organ space SSI wound disruption. Cardiovascular complications included cardiac arrest or myocardial infarction. Overall complications included any of the above complications.
784
Mbadiwe et al
Table 1. Characteristics of Patients Who Received Open Primary Anastomosis, Laparoscopic Primary Anastomosis, Open Colectomy with Colostomy, and Laparoscopic Colectomy with Colostomy (n ¼ 11,981) Characteristic
Sex Male Female Ethnicity Caucasian Black Hispanic Other Age, y <25 25e39 40e59 60e79 80e89 Death No Yes Cardiac comorbidities No Yes Pulmonary comorbidities No Yes Renal comorbidities No Yes Diabetes No Yes Weight class Normal weight Overweight Obese Any complications No Yes Major complications No Yes Sepsis No Yes Any respiratory complications No Yes
J Am Coll Surg
Management of Complicated Diverticulitis
n
%
5,663 6,317
47.3 52.3
9,818 574 685 904
82.0 4.8 5.7 7.6
29 952 5,579 4,634 724
0.2 8.0 46.6 38.7 6.0
11,971 190
98.4 1.6
6,305 5,676
52.6 47.4
11,392 589
95.1 4.9
11,852 129
98.9 1.1
10,858 1,123
90.6 9.4
2,832 4,255 4,435
23.6 35.5 37.0
9,196 2,785
76.8 23.3
10,469 1,512
87.4 12.6
11,288 693
94.2 5.8
11,321 660
94.5 5.5 (Continued)
Table 1. Continued Characteristic
Any cardiovascular complications No Yes Wound infection No Yes Functional health status before surgery Independent Partially dependent Totally dependent
n
%
11,890 91
99.2 0.8
10,260 1,721
85.6 14.4
11,317 542 122
94.5 4.5 1.0
All statistical analyses were performed using STATA/ MP Version 11.0 (StataCorp). Statistical significance was defined as p < 0.05.
RESULTS A total of 11,981 patients in the database had surgery for diverticulitis. The majority were female (53%) and Caucasian (82%), and the mean age was 58 (13 [SEM]) years. Comorbidities of the cardiovascular, pulmonary, or renal systems were present in 47%, 5%, and 1% of the cohort, respectively. Obesity and diabetes were present in 37% and 9% of the patients, respectively (Table 1). Laparoscopic colectomies were performed on 5,342 (45%) of patients; the remaining 6,639 (55%) received open colectomies. Among these groups, 5,105 (43%) had a laparoscopic primary anastomosis (PA-Lap), 4,158 (35%) underwent an open primary anastomosis (PA-Open), 237 (2%) underwent a laparoscopic colostomy (CC-Lap), and 2,481 (21%) received an open colostomy (CC-Open). Caucasian patients received PA-Lap, PA-Open, CC-Lap, and CC-Open procedures in 44%, 34%, 2%, and 20% of cases, respectively; by comparison, African-American patients underwent PA-Lap, PA-Open, CC-Lap, and CC-Open procedures in 27%, 45%, 2%, and 26% of cases, respectively (p < 0.001). In 10,035 (84%) patients, the surgery was classified as elective; the remaining 1,946 (16%) were categorized as emergent. Patients receiving emergency surgery underwent primary anastomosis in 20% of cases; 88% of elective patients underwent PA. The overall rate of postoperative complications was 23%, and the overall mortality rate was 1.6%. On bivariate analysis, patients undergoing laparoscopy experienced lower rates of complications with both PA (14% vs 26%, p < 0.001) and colostomy (30% vs 37%, p ¼ 0.02). The laparoscopic approach was associated with decreased mortality rates for patients undergoing PA (0.24% vs
Vol. 216, No. 4, April 2013
Management of Complicated Diverticulitis
Mbadiwe et al
785
Table 2. Bivariate Analysis of Patient Demographics, Comorbidities, Risk Factors, and Postoperative Complications (n ¼ 11,981) PA
CC
PA-Open Characteristic
Sex Male Female Ethnicity Caucasian Black Hispanic Other Age, y <25 25e39 40e59 60e79 80e89 Death No Yes Cardiovascular comorbidities No Yes Pulmonary comorbidities No Yes Renal comorbidities No Yes Diabetes No Yes Weight class Normal Overweight Obese Any complications No Yes Major complications No Yes Sepsis No Yes Any respiratory complications No Yes
PA-Lap
CC-Open
CC-Lap
n
%
n
%
n
%
n
%
1,871 2,287
45.0 55.0
2,485 2,620
48.7 51.3
1,197 1,284
48.2 51.8
110 126
46.6 53.4
3,325 258 220 355
78.0 6.2 5.3 8.5
4,293 155 312 345
84.1 3.0 6.1 6.8
2,003 152 137 189
80.7 6.1 5.5 7.6
197 9 16 15
83.1 3.8 6.8 6.3
7 316 1,926 1,664 236
0.27 7.6 46.4 40.1 5.7
12 471 2,703 1,805 108
0.2 9.2 53.0 35.4 2.1
8 146 856 1,075 352
0.3 6.0 35.1 44.1 14.4
2 19 94 90 28
0.9 7.8 38.4 42.4 11.4
4,125 33
99.2 0.8
5,093 12
99.8 0.2
2,344 137
94.5 5.5
229 8
96.6 3.4
2,091 2,067
50.3 49.7
3,036 2,063
59.5 40.5
1,061 1,420
42.8 57.2
117 120
49.4 50.6
3,978 180
95.7 4.3
5,007 98
98.1 1.9
2,189 292
88.2 11.8
218 19
92.0 8.0
4,125 33
99.2 0.8
5,096 9
99.8 0.2
2,396 85
96.6 3.4
235 2
99.2 0.8
3,734 424
89.8 10.2
4,762 343
93.3 6.7
2,162 319
87.1 12.9
200 37
84.4 15.6
962 1,432 1,644
23.8 35.5 40.7
1,195 1,961 1,864
23.8 39.1 37.1
620 779 849
27.6 34.7 37.8
57 83 78
26.2 38.1 35.8
3,083 1,075
74.2 25.9
4,391 714
86.0 13.9
1,555 926
62.7 37.3
167 70
70.5 29.5
3,650 508
87.8 12.2
4,783 322
93.7 6.3
1,844 637
74.3 25.7
192 45
81.0 19.0
3,921 237
94.3 5.7
4,980 125
97.6 2.4
2,164 317
87.2 12.8
223 14
94.1 5.9
3,975 183
95.6 4.4
5,036 69
98.7 1.3
2,092 389
84.3 15.7
218 92.0 19 8.0 (Continued)
786
Table 2.
Mbadiwe et al
J Am Coll Surg
Management of Complicated Diverticulitis
Continued PA
CC
PA-Open Characteristic
Any cardiovascular complications No Yes Wound infection No Yes Functional health status before surgery Independent Partially dependent Totally dependent Timing of surgery Elective Emergent
PA-Lap
CC-Open
CC-Lap
n
%
n
%
n
%
n
%
4,133 25
99.4 0.6
5,090 15
99.7 0.3
2,432 49
98.0 2.0
235 2
99.2 0.8
3,412 746
82.1 17.9
4,618 487
90.5 9.5
2,020 461
81.4 18.6
210 27
88.6 11.4
4,037 105 16
97.1 2.5 0.4
5,073 30 2
99.4 0.6 0.0
1,999 382 100
80.6 15.4 4.0
208 25 4
87.8 10.6 1.7
3,824 334
92.0 8.0
5,041 64
98.8 1.3
1,007 1,474
40.6 59.4
163 74
68.8 31.2
CC-Lap, laparoscopic colectomy with colostomy; CC-Open, open colectomy with colostomy; PA-Lap, laparoscopic primary anastomosis; PA-Open, open primary anastomosis.
0.79%, p < 0.001), and mortality differences for patients undergoing colostomy did not achieve statistical significance (3.4% vs 5.5%, p ¼ 0.16; Table 2). Multivariate analysis revealed that the laparoscopic approach was associated with lower odds of overall complications for patients undergoing PA (odds ratio [OR] 0.52, p < 0.001). The reduced risk of death for patients undergoing laparoscopic PA (vs open approach) did not achieve statistical significance (OR 0.68, p ¼ 0.3; Table 3). A small number of patients underwent laparoscopic colostomy (n ¼ 237, 2.0%), and there was no statistically significant difference in their risk of death or postoperative complications as compared with those receiving open colostomies (Table 4). Multivariate analyses suggested that, in the context of PA operations, cardiovascular (OR 1.25, p < 0.001) and pulmonary (OR 1.81, p < 0.001) comorbidities, and functional dependence (OR 2.12, p < 0.001), each appeared to be a significant risk factor for the development of overall complications; pulmonary disease (OR 2.86, p ¼ 0.025) and functional dependence (OR 17.4, p < 0.001) were associated with a higher risk of postoperative death. Further, for CC, pulmonary disease (OR 1.75, p < 0.001), presence of diabetes mellitus (OR 1.30, p ¼ 0.042), and functional dependence (OR 2.46, p < 0.001) significantly affected the risk of overall complications, and pulmonary (OR 2.86, p < 0.001) and renal (OR 2.29, p ¼ 0.024) comorbidities and functional dependence (OR 6.55, p < 0.001) were significant risk factors for mortality. When only patients who underwent elective operations were considered, laparoscopic operations were associated
with statistically significantly lower rates of overall complications (OR 0.53, p < 0.001; Table 5). Additionally, in the emergency context, the laparoscopic approach was not associated with lower rates of postoperative complications (OR 0.78, p ¼ 0.25), but was associated with lower rates of respiratory complications (OR 0.37, p ¼ 0.02; Table 6). The results of multivariate analyses using the entire study population and accounting for the preoperative likelihood that a patient would undergo a laparoscopic procedure (as indicated by each patient’s propensity score) suggested that patients receiving laparoscopic procedures are less likely to develop postoperative complications (OR 0.55, p < 0.001; Table 7). The difference in the mortality rates between the open and laparoscopic groups was not statistically significant in this analysis.
Table 3. Risk-Adjusted Analysis of Postoperative Complications Comparing Open and Laparoscopic Primary Anastomosis Procedures for Colon Diverticulitis Complication
PA-Open
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Reference Reference Reference Reference Reference Reference Reference
PA-Lap Odds ratio 95% CI
0.52 0.59 0.68 0.50 0.41 0.69 0.50
0.47e0.58 0.51e0.69 0.32e1.41 0.44e0.57 0.31e0.55 0.35e1.36 0.40e0.63
PA-Open, open primary anastomosis; PA-Lap, laparoscopic primary anastomosis.
Vol. 216, No. 4, April 2013
Mbadiwe et al
Table 4. Risk-Adjusted Analysis of Postoperative Complications Comparing Open and Laparoscopic Colectomy with Colostomy Procedures for Colon Diverticulitis Complication
CC-Open
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Reference Reference Reference Reference Reference Reference Reference
CC-Lap Odds ratio 95% CI
0.85 0.89 1.14 0.59 0.64 0.38 0.64
0.62e1.18 0.61e1.30 0.49e2.67 0.38e0.93 0.36e1.15 0.05e2.89 0.36e1.14
CC-Lap, laparoscopic colectomy with colostomy; CC-Open, open colectomy with colostomy.
Management of Complicated Diverticulitis
787
Table 5. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures Conducted on an Elective Basis Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Reference Reference
Laparoscopic Odds ratio 95% CI
0.53 0.61 0.84 0.50 0.45 0.69 0.51
0.48e0.60 0.52e0.71 0.44e1.61 0.45e0.57 0.34e0.60 0.36e1.31 0.41e0.64
DISCUSSION The benefits of the laparoscopic approach in the elective surgical management of patients with diverticular diseasedincluding decreased morbidity, mortality, and length of staydhave been well described.7-10 It remains an open question as to whether these benefits can be realized in the acute care surgical setting. To our knowledge, this is the largest series reported to date encompassing solely patients having surgery for diverticulitis. The statistical robustness achieved by using the ACSNSQIP database was countered by our inability to detect by diagnostic code the degree of physiologic insult experienced by patients whose diverticular disease grew sufficiently complicated to require surgical resection. Accordingly, we chose to use primary anastomosis and diverting colostomy as surrogates to represent the surgeon’s real-time judgment about patient stability. In this study, open approaches were used in 45% of patients undergoing PA and in 91% of those having CC. The notable difference in the frequencies with which laparoscopic approaches are adopted for primary anastomoses as compared with colostomies lends itself to the suggestion that many surgeons are more comfortable
using open approaches in situations in which a colostomy is required. This study demonstrated that the laparoscopic approach was associated with lower rates of postoperative complications in the PA group. The small number of patients undergoing the laparoscopic approach for CC (n ¼ 237, 2.0%) may help to explain why their lower risk of morbidity and mortality did not achieve statistical significance. Regardless of procedure, preoperative pulmonary and renal comorbidities were each associated with increased morbidity and mortality. Additional studies that more precisely delineate severity of disease on presentation are necessary, but these findings suggest to us that patients whose physiologic state will support PA in the acute management of complicated diverticulitis are candidates for the laparoscopic approach. African-American patients received laparoscopic PA less frequently than their Caucasian counterparts in our study. This discrepancy may speak to disparities in access to care and severity of disease, and merits future investigation. A limitation of this study was the unavailability of information in the ACS-NSQIP database concerning the hemodynamic status of patients at presentation. Surgeons may be predisposed to disproportionately select open (vs laparoscopic) and CC (vs PA) procedures for patients
Table 6. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures Conducted on an Emergency Basis
Table 7. Risk-Adjusted Analysis of Postoperative Complications Comparing Laparoscopic and Open Procedures and Adjusting for Preoperative Likelihood of Receiving a Laparoscopic Procedure, as Indicated by Propensity Scoring
Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Omitted Reference
Laparoscopic Odds ratio 95% CI
0.77 0.83 0.73 0.66 0.37
0.51e1.19 0.51e1.35 0.21e2.53 0.38e1.15 0.15e0.88
0.62
0.29e1.32
Complication
Overall complications Major complications Death Wound infections Respiratory complications Cardiovascular complications Sepsis
Open
Reference Reference Reference Reference Reference Reference Reference
Laparoscopic Odds ratio 95% CI
0.55 0.63 0.77 0.50 0.45 0.63 0.51
0.50e0.61 0.54e0.72 0.44e1.35 0.45e0.57 0.34e0.59 0.34e1.17 0.42e0.64
788
Mbadiwe et al
J Am Coll Surg
Discussion
with hemodynamic instability, which would bias the colostomy group toward higher rates of morbidity and mortality. The multivariate analysis performed in this study would not be able to account for such a bias, a problem we attempted to address by performing our analyses within each group (PA and CC). Also, the experience of the operating surgeon with performing laparoscopic procedures was not available.11
CONCLUSIONS In a large study of patients undergoing surgery for complicated diverticulitis, the laparoscopic approach was associated with a lower risk of total complications than the open approach for patients undergoing PA. Preoperative comorbidities are important associated risk factors for the occurrence of complications and death. These results suggest that the laparoscopic approach should be considered whenever PA is judged appropriate. Author Contributions Study conception and design: Mbadiwe, Obirieze, Cornwell, Fullum Acquisition of data: Mbadiwe, Obirieze, Cornwell, Turner, Fullum Analysis and interpretation of data: Mbadiwe, Obirieze, Cornwell, Turner, Fullum Drafting of manuscript: Mbadiwe, Obirieze, Cornwell, Fullum Critical revision: Mbadiwe, Obirieze, Cornwell, Turner, Fullum REFERENCES 1. Chapman J, Davies M, Wolff B, et al. Complicated diverticulitis: is it time to rethink the rules? Ann Surg 2005;242:576e583. 2. Rafferty J, Shellito P, Hyman NH, et al. Practice parameters for sigmoid diverticulitis. Dis Colon Rectum 2006;49:939e944. 3. Eijsbouts QA, de Haan J, Berends F, et al. Laparoscopic elective treatment of diverticular disease. A comparison between laparoscopic-assisted and resection-facilitated techniques. Surg Endosc 2000;14:726e730. 4. Kakarla V, Nurkin S, Sharma S, et al. Elective laparoscopic versus open colectomy for diverticulosis: an analysis of ACSNSQIP database. Surg Endosc 2012;26:1837e1842. 5. Masoomi H, Buchberg B, Nguyen B, et al. Outcomes of laparoscopic versus open colectomy in elective surgery for diverticulitis. World J Surg 2011;35:2143e2148. 6. ACS NSQIP participant user data file user guide (2011). Available at: http://site.acsnsqip.org/downloads/. Accessed October 8, 2012. 7. Chen HH, Wexner SD, Weiss EG, et al. Laparoscopic colectomy for benign colorectal disease is associated with a significant reduction in disability as compared with laparotomy. Surg Endosc 1998;12:1397e1400. 8. Bergamaschi R, Arnaud JP. Immediately recognizable benefits and drawbacks after laparoscopic colon resection for benign disease. Surg Endosc 1997;11:802e804.
9. Sun J, Jiang T, Qiu Z, et al. Short-term and medium-term clinical outcomes of laparoscopic-assisted and open surgery for colorectal cancer: a single center retrospective case-control study. BMC Gastroenterol 2011;11:85. 10. Angriman I, Scarpa M, Ruffolo C. Health related quality of life after surgery for colonic diverticular disease. World J Gastroenterol 2010;16:4013e4018. 11. Waters J, Chihara R, Moreno J, et al. Laparoscopic colectomy: does the learning curve extend beyond colorectal surgery fellowship? JSLS 2010;14:325e331.
Discussion DR MICHAEL ROTONDO (Greenville, NC): The standardization of approach and outcomes managements in the emergency surgical province has become an intense area of focus over the last several years. We have seen specialties redefine themselves, like acute care surgery, and we have seen new societies emerge to greater prominence, like the European Society for Trauma and Emergency Surgery. Certainly, with the rising emphasis on the value proposition in health care, in which the relationship between quality and cost become paramount in measuring both patient outcomes and maintaining financial viability, questions such as the one before us become increasingly important. However, this work falls somewhat short of accurately comparing the efficacy of a laparoscopic approach and an open approach in a patient with complicated diverticulitis. It suffers from a lack of resolution of the very same confounding factors that are perplexing a group of surgeons convened through the American College of Surgeons, endeavoring to create an emergency general surgery data dictionary with the specific intent of developing a research and outcomes measurement tool for these types of patients. The group has a broad base of organizations represented, including the Advisory Council on General Surgery; the Society of Gastrointestinal and Endoscopic Surgeons (SAGES); the Surgical Infection Society; the Rural Surgery Committee; the Committee on Trauma, represented by Avery Nathans and the work he’s done with the Trauma Quality Improvement Program (TQIP); the American Association for the Surgery of Trauma (AAST) represented by Shaheed Shafi; the work that Wayne Meredith has gotten started; and, of course, Cliff Ko and the National Surgical Quality Improvement Program (NSQIP). And they are steadily working to devise a system in which emergency surgical entities, such as complicated diverticulitis, can be stratified by physiology at the time of presentation, the severity of the surgical entity, and comorbidities. It’s sort of like a TNM classification system for cancer that’s evolving. Without adequate standardized description of the disease process itself, comorbidities, and physiologic perturbations at the time of presentation, it is exceedingly difficult to measure outcomes in the efficacy of therapeutic interventions. And I’m afraid that the authors have not satisfactorily controlled or analyzed these confounding factors such that plausible conclusions can be reached. The presentation did help clarify some of the questions that I had in the manuscript, so it’s hard for me to know if this was