- Email: [email protected]
Application of a Novel Severity Grading System for Surgical Complications after Colorectal Resection
Application of a Novel Severity Grading System for Surgical Complications after Colorectal Resection Haggi Mazeh, MD, Yacov Samet, MD, Bassam Abu-Wasel, MD, Nahum Beglaibter, MD, Ronit Grinbaum, MD, Tzeela Cohen, MD, Meir Pinto, MD, Tamar Hamburger, BA, Herbert R Freund, MD, Aviram Nissan, MD Uniform and accurate reporting of surgical complications is the basis for quality control. We developed a computerized system for reporting and grading surgical complications in colorectal surgery. This study was conducted to evaluate this computerized reporting system. STUDY DESIGN: A retrospective chart review was conducted of all surgical complications in patients who underwent resection of the colon or rectum at our institution between the years 1999 and 2004 (n ⫽ 408). All complications were recorded using the computerized reporting system and compared with complications reported in the literature. RESULTS: Elective operations were performed in 75.7% of patients, and 24.3% required emergency operations. Of the 408 patients in the study, 239 (58.6%) had an uneventful recovery without complications. At least 1 complication was recorded in 169 (41.4%) patients. Grades 1 and 2 complications were recorded in 83 (20.3%) and 105 (25.7%) patients, respectively, requiring observation or medical treatment only, and 59 patients (14.5%) had grades 3 to 5 complications. The three leading complications were surgical site infection, intraabdominal abscess, and hemorrhage requiring blood transfusion. The grades 3 to 5 complication rate was within the range described in the literature, and the rate of grades 1 and 2 complications was substantially higher. These grades 1 and 2 complications were associated with a substantially longer hospital stay. CONCLUSIONS: This novel complication reporting system was found feasible and proved to have a higher sensitivity for recording minor but meaningful complications that tend to prolong hospital stay. (J Am Coll Surg 2009;208:355–361. © 2009 by the American College of Surgeons) BACKGROUND:
a powerful driving force, which might curb health-care costs and improve quality.3 Accurate and uniform reporting of surgical complications is essential for quality-of-care measurement. Currently, there is no uniform reporting system for surgical complications. Adverse events reporting systems are used for clinical trials4 but are rarely used to report surgical complications. Institutional reporting is based mainly on a venerable tradition of weekly morbidity and mortality conferences, and clinical trial complications reporting is inconsistent and in many trials even underreported. Short-term surgical outcomes, such as length of hospital stay, 30-day mortality rate, operating time, and estimated blood loss, are routinely included in the data reported. Conclusive assessments of surgical procedures remain limited by the lack of consensus on how to define complications and to stratify them by severity.5 A surgical complication grading system must be precise, simple, and reproducible. Absence of a common grading system encouraged us to develop a five-tiered grading system for surgical complications. Based on systems reported for gastric and pancreatic
Growing demand for health care, rising costs, constrained resources, and variations in clinical practice have triggered interest in measuring and improving the quality of health care delivery. For a valuable quality assessment, relevant data on outcomes must be obtained in a standardized and reproducible fashion to allow comparison among different centers, between different therapies, and within a center over time.1,2 Objective and reliable outcomes data are also increasingly requested by patients and their representatives, health-care providers, and national and international health organizations. In addition, health policy makers point out that the availability of comparative data on individual institution’s and physician’s performance represents Disclosure Information: Nothing to disclose. Received October 9, 2008; Revised November 26, 2008; Accepted December 2, 2008. From the Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus, Jerusalem, Israel. Correspondence address: Aviram Nissan, MD, Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus, POB 24035, Jerusalem, 91240 Israel. email: [email protected]
© 2009 by the American College of Surgeons Published by Elsevier Inc.
355
ISSN 1072-7515/09/$36.00 doi:10.1016/j.jamcollsurg.2008.12.008
356
Mazeh et al
A Novel Grading System for Surgical Complications
operations,6,7 we adopted the same principles of a fivetiered grading system to develop a surgical complication reporting system. This study was undertaken to validate this surgical complication reporting system. We elected to validate the system on a cohort of patients who underwent colorectal operations, because colon or rectal resections are related to a variety of complications, are conducted in elective and emergency settings, and can be performed as a laparoscopic or open procedure. This would allow us to compare subgroups of patients and, as a result, learn more about the validity of the system suggested. This is a retrospective single-institution report of complications after colorectal operations based on a five-tiered system in a large cohort of patients.
METHODS A computerized complication reporting database was created using Microsoft Access software (Microsoft Corp). A five-tiered scale grading severity of complications was incorporated. All colon and rectal resections performed between January 1999 and December 2004 at the Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus were retrieved from the operating theater database. All patients’ hospital and outpatient clinic charts were reviewed. Data recorded included age, gender, past medical history, medications, comorbidities, and indications for operation. Operative details recorded were type of operation, extent of resection, laparoscopic versus open operations, emergency versus elective procedure, and need for a stoma. Mechanical bowel preparation was performed in all elective patients using sodium phosphate solution, and bowel decontamination was performed with oral neomycin and metronidazole. Prophylactic antibiotics were administrated to all elective and emergent patients within 30 minutes of skin incision and included IV ampicillin, gentamycin, and metronidazole, unless contraindicated. Intermittent compression stockings were applied in all cases and subcutaneous low-molecularweight heparin was initiated 6 hours after operation. Variables we recorded included length of operation, blood loss and the need for transfusion, and bowel content spillage. Hospital and outpatient clinic charts were reviewed for detailed postoperative course, including all complications and interventions, length of stay, and readmissions within 30 days of the surgical procedure. All complications were classified and graded according to the Surgical Complication Severity Scoring System (Table 1) (Appendix A online). Complications were defined as any deviation from the normal postoperative course. This broad definition also includes asymptomatic documented complications. Preex-
J Am Coll Surg
Table 1. Examples of Complication Grades Grades
1
2
3 4
Examples
Episode of atrial fibrillation not requiring medical therapy Atelectasis proved by chest x-ray Wound infection opened at bedside Arrhythmia requiring -blockers for control Urinary tract infection requiring antibiotics Pneumonia Thromboembolic episode requiring anticoagulation therapy Prolonged ileus resulting in long hospital stay Bleeding requiring blood transfusion Gastrointestinal bleeding controlled by endoscopy Anastomotic leak drained by interventional radiology Anastomotic leak and sepsis requiring surgical repair Evisceration Bleeding requiring operation for control
isting medical conditions that did not change during the hospital course were not considered as adverse events. All complications were graded by severity according to the Surgical Complications Severity Scoring System. Grade 1 included minor asymptomatic complications that did not require medical therapy or radiologic or surgical intervention. Grade 2 complications were defined as complications requiring pharmacological therapy, minor interventions, or any prolongation of hospital stay (as defined in Appendix A online). Grade 3 complications were defined as complications that required intervention other than operations, including interventional radiology or endoscopy. Grade 4 included complications that required surgical intervention and complications resulting in permanent loss of function of an organ. Grade 5 complications included complications resulting in death. Statistical analysis
Summary statistics were performed using established methods. Association between categorical factors was studied with Fisher’s exact test or chi-square test, as appropriate. Association between noncategorical (continuous) variables was performed using Student’s t-test. SPSS statistical package version 13.0 (SPSS, Inc) was used for statistical analysis. Two-sided p value ⱕ 0.05 was set as significant.
RESULTS During the study period, a total of 408 patients underwent colon or rectal resections. The median age was 65 years (range, 17 to 96 years), 1 with a slight female predominance, 52%, versus 48% male. At least 1 comorbid condition was recorded for 62.3% of the patients. The most common indication for operation was malignancy in
Vol. 208, No. 3, March 2009
Mazeh et al
Table 2. Patient and Procedure Characteristics (n ⫽ 408) Age (y), median (range) 65 (17⫺96) n % Gender Male 196 48 Female 212 52 Patients with comorbidities ⱕ1 261 63 ⬎1 147 37 Indication for procedure Malignancy 316 77.5 Benign 92 22.5 IBD 21 5.1 Diverticulitis 37 9 Volvulus 12 2.9 Other 22 5.4 Procedure setting Elective 309 75.9 Emergency 99 24.3 Technique Open 323 79.2 Laparoscopic 85 20.8 Procedure type Right hemicolectomy 130 31.8 Left hemicolectomy 23 5.6 Sigmoid resection 97 23.7 Sleeve resection 5 1.2 Subtotal colectomy 68 16.6 Low anterior resection 73 17.8 Abdominoperineal resection 1 0.2 Total proctocolectomy 11 2.6 IBD, inflammatory bowel disease.
77.5% of the patients. Elective operations were performed in 75.7% of the patients, and 24.3% required emergency operation. Only 20.8% of patients underwent laparoscopic procedures, with 10.5% (9 patients) converted to open. These patients were included in the laparoscopic group to maintain an intention-to-treat analysis. Patient and procedure’s details are outlined in Table 2. Mean operative time was 142 minutes (range, 66 to 372 minutes). In 12 patients (2.9%), gross spillage was documented, but it was not substantially associated with a higher complication grade. Of 408 patients who were included in the study, 239 (58.6%) had an uneventful recovery without any complication. Two hundred forty-seven complications were recorded in 169 (41.4%) patients. Distribution of complications by grade is presented in Table 3. Grades 1 and 2 complications were recorded in 83 (20.3%) and 105 (25.7%) patients, respectively, and together they accounted for 76.1% of all complications. In 59 patients (14.5%), grades 3 to 5 complications were recorded (Table 3). Mortality rate
A Novel Grading System for Surgical Complications
357
Table 3. Incidence and Distribution of Complications by Grade Grade
1 2 3 4 5
n
All patients (%)
All complications (%)
83 105 25 27 7
20.3 25.7 6.1 6.6 1.7
33.6 42.5 10.1 10.9 2.8
(grade 5 complications) was 1.7%. The 10 most common complications and their incidence are presented in Table 4. Surgical site infection, blood transfusion, and atelectasis accounted for 18.6% of all complications. There was a significantly lower complication rate after elective operations compared with emergency procedures (35.9% versus 59.6%; p ⬍ 0.01). All seven patients with grade 5 complications (resulting in death) underwent emergency procedures. Complication rate in patients who underwent laparoscopic colon resections was significantly lower than in the open operation group (23.5% versus 46.1%; p ⬍ 0.0001). Because none of the emergency procedures was laparoscopic, a comparison of elective-only procedures showed no difference in the complication rate between elective open and laparoscopic procedures. There was also no significant difference in complication rate between colon and rectal procedures (42.9% versus 39.2%; p ⫽ 0.66). Complication rate was not statistically different in patients younger or older than 50 years of age (43.3% versus 34.5%; p ⫽ 0.14). Tables 5 and 6 summarize the incidence of the leading complications in the various subgroups. Number of complications was strongly associated with prolonged length of stay (Fig. 1). Median length of stay (LOS) of uncomplicated patients was 6 days. Among the 112 patients with 1 complication, median LOS was 9 days. Two and 3 complications were recorded in 39 and 16 paTable 4. Ten Most Common Complications and their Incidence Incidence Complication
n
%
Surgical site infection Intraabdominal abscess Blood transfusion Pseudomembranous colitis Atelectasis Pneumonia Anastomotic leak Arrhythmia Prolonged ileus Myocardial infarct
36 35 25 18 15 15 13 9 9 9
8.8 8.6 6.1 4.4 3.7 3.7 3.2 2.2 2.2 2.2
0.22
0.66 39.2 29 42.9 140 ⬍ 0.001 149 23.5 ⬍ 0.001
46.1
5 7 0.32 4.7 7 ⬍ 0.001
20 58.6 36.0
58
12.1 7
111 Total patients with complications
14 Grade 4
Grade 5
0.95
0.86 10.3 3 11.5 16 0.91 10.1 15 20 4 0.42 8.6 5
13 Grade 3
12.7
0.055 58.6
10.3 3
17 39.3
10 14
55 0.225
0.42 10.7
44.3 66
16 5
30 6
1 0.3
0.81 41.1
6.9 4
24 48 Grade 2
11.7
36 Grade 1
43.2
0.75
0.15 20.7 6 34.3 48 0.18 30.2 45 45 9 0.85 31 18
0.58
5 Myocardial infarct
32.4
2.7 2 2.1 7 0.12 2.8 9 — — 0.15 4.0 4
0.23
9 Prolonged ileus
1.6
1.3 1 2.4 8 0.12 2.8 9 — — 0.08 — —
5 Arrhythmia
2.9
4.0 3 1.8 6 0.12 2.8 9 — — 0.15 4.0 4
0.8
6 Pneumonia
1.6
1.3 1 3.6 12 0.06 4.0 13 — — 0.01 7.1 7
0.85
9 Anastomotic leak
1.9
2.7 2 3.3 11 0.84 3.1 10 3.5 3 0.58 4.0 4
0.64
11 Atelectasis
2.9
4.0 3 3.6 12 0.57 3.4 11 4.7 4 0.8 4.0 4
0.06
11 Pseudomembranous colitis
3.5
5.4 4 4.2 14 0.3 4.9 16 2.3 2 0.14 7.1 7
0.76
17 Blood transfusion
3.5
1.3 1 7.2 24 0.9 6.2 20 5.9 5 0.35 8.1 8
0.8
Intraabdominal abscess
5.5
9.5 7 8.4 28 0.15 9.6 31 4.7 4 0.84 8.1 8 27
Surgical site infection
8.7
p Value %
8.1 6
n %
9 30
n p Value
0.83 9.0
% n
29 8.2
% n
7 0.032
p Value %
14.1 14
n % n
22
Complication
7.1
Rectal operation (n ⴝ 74) Colon operation (n ⴝ 334) Open operation (n ⴝ 323) Laparoscopic operation (n ⴝ 85) Emergent operation (n ⴝ 99)
0.32
A Novel Grading System for Surgical Complications
Mazeh et al
Elective operation (n ⴝ 309)
Table 5. Incidence of Complications in the Different Subgroups (Elective Versus Emergent, Laparoscopic Versus Open Operation, and Colon Versus Rectal)
358
J Am Coll Surg
tients, with a median LOS of 16 and 17 days, respectively (p ⬍ 0.05). Only 2 patients had 4 complications, with a median LOS of 11.5 days. Complication grade was also associated with prolonged LOS (Fig. 2). Median LOS of grades 1 to 4 complications was 10.6, 13.8, 17.8, and 17.7 days, respectively (ANOVA, p ⬍ 0.0001). Grade 5 complications (perioperative mortality) did not necessarily result in prolonged hospitalization. Surgical site infection occurred in 8.8% of patients (n ⫽ 36). It was graded as grade 1 in 78% of the patients requiring bedside drainage of the wound. The remaining 22% were grade 2 complications, requiring antibiotic treatment for surrounding soft-tissue involvement. Median LOS in patients with surgical site infection was 12 days (range, 5 to 44 days), and it occurred as a single complication in only 44.4% of patients. Surgical site infection was more common after emergency operation compared with elective procedures (7.1% versus 14.1%; p ⬍ 0.05). Surprisingly, rate of surgical site infection was similar for laparoscopic and open operations (8.2% versus 9.0%). Intraabdominal abscess occurred in 8.6% of patients (n ⫽ 35). There were grade 2 intraabdominal abscesses treated with IV antibiotics only in 42.9% of the patients. In the remaining 57.1% patients, intraabdominal abscesses were treated by interventional radiology percutaneous drainage (grade 3). Presence of intraabdominal abscess was associated with a median LOS of 11 days (range, 5 to 45 days) and occurred as a single complication in only 42.9% of the patients. Anastomotic leak was defined as air outside the gastrointestinal lumen, fluid collection, abscess in close proximity to the anastomotic site, or leak of contrast dye at the anastomosis site by radiology studies or surgical exploration. Intraabdominal abscess was not defined as a leak unless met by the previously mentioned criteria for anastomotic leakage. Anastomotic leak occurred in 3.2% of patients (n ⫽ 13). All patients with a diagnosis of anastomotic leak were reoperated (grade 4). Anastomotic leak was associated with a median LOS of 17 days (range, 6 to 45 days), was accompanied by more than 1 other complication in the majority of the patients (61.5%), and by mortality of a single patient (7.7%). Blood transfusion was defined as a grade 2 complication if 2 U of packed red blood cells or more were administrated postoperatively. It occurred in 6.1% of the patients (n ⫽ 25) and did not require additional surgical intervention in any of the patients. It was associated with a median LOS of 8 days (range, 5 to 24 days), which was not considerably different from the cohort of patients without complica-
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
359
Table 6. Incidence of Complications in the Different Subgroups (Comorbidities and Age)
Complications
Surgical site infection Intraabdominal abscess Blood transfusion Pseudomembranous colitis Atelectasis Anastomotic leak Pneumonia Arrhythmia Prolonged ileus Myocardial infarct Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Total patients with complications
0 to 1 comorbidity (n ⴝ 261) n %
>1 Comorbidity (n ⴝ 147) n %
24 21 14 10 10 9 11 6 5 4 29 46 10 15 1
9.2 8.0 5.4 3.8 3.8 3.4 4.2 2.3 1.9 1.5 28.7 45.5 9.9 14.9 1
12 14 11 8 5 4 2 3 4 5 22 26 8 8 6
8.2 9.5 7.5 5.4 3.4 2.7 1.4 2.0 2.7 3.4 32.4 38.2 11.8 11.8 8.8
101
38.7
68
46.3
tions. It presented as a single complication in 60% of patients. Respiratory complications (atelectasis and pneumonia documented by chest x-ray) were graded as 1 and 2, respectively. Atelectasis and pneumonia each occurred in 3.7% of the patients (n ⫽ 15) and were associated with a median LOS of 8 and 12 days, respectively. Atelectasis was a single complication in 53.3% of the patients, although pneumonia was associated with more than 1 other complication in 53.3% of patients. Pneumonia was more common after emergency operation compared with elective procedures (7.1% versus 1.9%; p ⫽ 0.01).
Figure 1. Median length of stay as a function of number of complications for the entire cohort of patients (n ⫽ 408). Length of stay was considerably longer for patients with two to three complications compared with those with none to one complication.
Age older than 50 y (n ⴝ 321) n %
Age younger than 50 y (n ⴝ 87) n %
0.74 0.6 0.4 0.44 0.8 0.7 0.11 0.86 0.6 0.2 0.6 0.34 0.7 0.56 0.012
29 27 20 16 12 7 10 9 8 9 44 63 12 13 7
9 8.4 6.2 5 3.7 2.2 3.1 2.8 2.5 2.8 31.7 45.3 8.6 9.4 5
7 8 5 2 3 6 3 — 1 — 10 9 5 6 —
8.0 9.2 5.7 2.3 3.4 6.9 3.4 — 1.1 — 33.3 17.8 16.7 20.7 —
0.77 0.82 0.86 0.28 0.9 0.026 0.87 0.114 0.45 0.11 0.86 0.12 0.18 0.08 0.209
0.13
139
43.3
30
34.5
0.14
p Value
p Value
DISCUSSION The definition of a surgical complication is “any deviation from the expected steady recovery after a surgical operation.”8 This broad definition results in considerable variation in surgical complication reporting, even in the setting of controlled clinical trials. Currently, there are no uniform requirements of surgical journals for reporting complications. The importance of a uniform, simple, and reproducible reporting system for surgical complications cannot be emphasized enough. Such a system will set the basis for medical centers worldwide to capture, analyze, and report
Figure 2. Median length of stay as a function of complication grade for the entire cohort of patients (n ⫽ 408). Length of stay was significantly longer (ANOVA, p ⬍ 0.0001) for patients with highergrade complication. Grade 5 (perioperative mortality) did not necessarily result in prolonged hospitalization.
360
Mazeh et al
A Novel Grading System for Surgical Complications
comparable surgical complications. The ability to measure quality of care delivered to a group of patients in a given institution or country is dependent not only on the efficacy of the surgical therapy but also on the morbidity and mortality it carries. The ability of a patient or a health-care provider to choose the best therapy for a given disease or disorder is not only dependent on the efficacy of the evaluated therapy but also on its related morbidity in general and at a certain institute in particular. Unlike clinical trials involving pharmaceutical agents, the morbidity of surgical procedures in clinical trials has not been subjected to a uniform reporting system. Some complications, such as anastomotic leakage or wound dehiscence, are severe and would generally be reported. But there is large variability in the reporting of many intraoperative or postoperative complications that are less severe or seemingly less important or unimportant. The need for uniform reporting of surgical complications by institutions and health-care providers is not limited to clinical trials. Growing demand for quality health care, constrained resources, and variation in clinical practice has triggered a growing interest in measuring and improving the quality of health-care delivery. Objective and reliable outcomes data are increasingly requested by patients and their health-care providers.1-3 Conclusive assessments of surgical procedures remain limited by the lack of consensus on how to define complications and stratify them by severity. The current study was undertaken to validate a surgical complication reporting system on a large enough cohort of patients to measure differences between procedures and try to set a benchmark for surgical complication reporting in the future. We adopted a five-tiered reporting system from previous reports and from the National Cancer Institute Common Terminology Criteria for Adverse Events 3.0 adverse event reporting system.4 We validated the system by reviewing a cohort of colorectal patients operated on in a predetermined time period. To validate our severity score, we compared each grade to well-established and measurable outcomes, eg, LOS. Our results show that grades 3 to 5 complications are well within the range reported in the literature. The rate of grades 1 and 2 complications was much higher than reported in the surgical literature, suggesting previous underreporting of such complications. The distribution of complications, as expected, was with a majority of grades 1 and 2 complications and with a lesser degree of severe grades 3 and 4 or fatal complications (grade 5). We were also able to show that the complications rate, and especially the mortality rate, is much lower in elective operations as compared with emergency operations.
J Am Coll Surg
In this cohort of patients, we documented a variety of complications and, by using the suggested system, we were able to define a better view of the complication rate related to colorectal operations in our institution. Reporting an overall complication rate of 41% for colorectal operations is extremely high. By using the five-tiered grading system, we were able to show that our complication rate is actually within the limits reported by others and that the additional grades 1 and 2 complications that we accurately reported are usually grossly underreported by many. We showed that grades 1 and 2 complications are associated with a substantially prolonged hospital stay (Fig. 2) and should be vigorously observed, reported, and prevented. Mortality rate reported in the present study compares favorably with other reported postoperative mortality, ranging from 2% to 9%. Morbidity rate of 41% is higher than reported morbidity after colorectal resections, ranging from 18% to 37% only.9-15 In this report, only 14.5% of the patients had grades 3 to 5 complications, which are usually reported in the literature. Our data show an increased risk for fatal (grade 5) complications in patients with 2 or more comorbid conditions. Because all our grade 5 complications occurred after emergency operations, this might be the most crucial contributing factor. Because of the small number of grade 5 complications, defining the weighted value of each variant by a multivariate analysis was not feasible. Overall complication rate was also considerably higher in patients who underwent emergency operations. This is consistent with previous studies.10,12-17 The complication rate in patients who underwent laparoscopic operations was considerably lower than in the open operation group, but none of the emergency operations was performed laparoscopically. When excluding emergency operations from the open operation group, there was no difference in the complication rate between open and laparoscopic procedures. This difference reflects the higher complication rate related to emergency procedures. There are conflicting reports in the medical literature as to whether or not laparoscopic operations result in a lower complication rate.18-20 There are reports linking complications with age or anatomic location (colon versus rectum).12,14 We did not find any statistically significant difference in complication rate associated to age older than 50 years or with rectal operations. Comorbid conditions were related to surgical complications, with a higher complication rate in patients with two or more comorbid conditions.9-15,17 Reporting of surgical complications is important for quality assurance. There are many barriers for creating a uniform and robust reporting of surgical complications. It is often conceived as reporting failure and not success, so it
Vol. 208, No. 3, March 2009
Mazeh et al
is only natural that many institutions tend to underscore their complication rate. An accurate and uniform reporting system, on the one hand, will allow surgeons to report complication centers and improve outcomes. On the other hand, this can create a basis for patients’ selections of lowrate complication centers and not centers of excellence that can treat more complicated patients with higher complication rates. Another barrier is that many and different reporting systems are currently used, most adopted from medical reporting systems, and do not cover the spectra of surgical complications. This can be overcome by presenting validated and prospective data supporting use of a Surgical Complication Grading System. We conclude that the surgical complication grading system presented is precise, simple, and reproducible. With minor modifications, it can easily be adopted and adjusted to any other health-care system. In an era of quality control, surgical quality improvement, and surgical care improvement, it offers a reliable and easy to use asset. Author Contributions Study conception and design: Mazeh, Freund, Nissan Acquisition of data: Mazeh, Samet, Abu-Wasel, Beglaibter, Grinbaum, Cohen, Pinto, Hamburger, Nissan Analysis and interpretation of data: Mazeh, Hamburger, Nissan Drafting of manuscript: Mazeh, Freund, Nissan Critical revision: Mazeh, Samet, Abu-Wasel, Beglaibter, Grinbaum, Cohen, Pinto, Hamburger, Freund, Nissan
REFERENCES 1. Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery 1992;111:518–526. 2. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–213. 3. Daley J, Henderson WG, Khuri SF. Risk-adjusted surgical outcomes. Annu Rev Med 2001;52:275–287. 4. Trotti A, Colevas AD, Setser A, et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13:176–181.
A Novel Grading System for Surgical Complications
361
5. Martin RC 2nd, Brennan MF, Jaques DP. Quality of complication reporting in the surgical literature. Ann Surg 2002;235: 803–813. 6. Martin RC 2nd, Jacques DP, Brennan MF, Karpeh M. Achieving R0 resection for locally advanced gastric cancer: is it worth the risk of multiorgan resection? J Am Coll Surg 2002;194:568– 577. 7. Grobmyer SR, Pieracci FM, Allen PJ, et al. Defining morbidity after pancreaticoduodenectomy: use of a prospective complication grading system. J Am Coll Surg 2007;204:356–364. 8. Townsend CM, Beaucham RD, Evers BM, et al, eds. Sabiston textbook of surgery: the biological basis of modern surgical practice. 17th ed. Philadelphia: WB Saunders; 2004. 9. Khuri SF, Henderson WG, DePalma RG, et al. Determinants of long-term survival after major surgery and the adverse effect of postoperative complications. Ann Surg 2005;242:326⫺341; discussion 341⫺343. 10. Prystowsky JB, Bordage G, Feinglass JM. Patient outcomes for segmental colon resection according to surgeon’s training, certification, and experience. Surgery 2002;132:663–670. 11. Cook A, Hyman N. Quality assessment and improvement in colon and rectal surgery. Dis Colon Rectum 2004;47:2195–2201. 12. Alves A, Panis Y, Mathieu P, et al. Postoperative mortality and morbidity in French patients undergoing colorectal surgery: results of a prospective multicenter study. Arch Surg 2005;140:278–283. 13. Tekkis PP, Kessaris N, Kocher HM, et al. Evaluation of POSSUM and P-POSSUM scoring systems in patients undergoing colorectal surgery. Br J Surg 2003;90:340–345. 14. Bokey EL, Chapuis PH, Fung C, et al. Postoperative morbidity and mortality following resection of the colon and rectum for cancer. Dis Colon Rectum 1995;38:480–486. 15. Ko CY, Chang JT, Chaudhry S, Kominski G. Are high-volume surgeons and hospitals the most important predictors of inhospital outcome for colon cancer resection? Surgery 2002;132: 268–273. 16. Longo WE, Virgo KS, Johnson FE, et al. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis Colon Rectum 2000;43:83–91. 17. Harmon JW, Tang DG, Gordon TA, et al. Hospital volume can serve as a surrogate for surgeon volume for achieving excellent outcomes in colorectal resection. Ann Surg 1999;230:404–411. 18. Degiuli M, Mineccia M, Bertone A, et al. Outcome of laparoscopic colorectal resection. Surg Endosc 2004;18:427–432. 19. Senagore AJ, Delaney CP. A critical analysis of laparoscopic colectomy at a single institution: lessons learned after 1000 cases. Am J Surg 2006;191:377–380. 20. Schwenk W, Haase O, Neudecker J, Muller JM. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;20:CD003145.
361.e1
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Surgical Complications Grading System. Surgical Complications Grading System Complication
Grade 1
Allergic Reaction/ Hypersensitivity
Transient drug fever or rash
Hemolysis (transfusion reaction)
Laboratory evidence without decrease in hemoglobin
Thrombocytopenia
⬍ 50,000/mm3
Cardiac arrhythmia
Asympthomatic intervention notindicated
Cardiac ischemia
Asymptomatic (transient ECG changes)
Pericardial effusion
Asymptomatic
DIC (disseminated intravascular coagulation)
Mild symptoms (hematomas, Hemoglobin decrease ⱕ 2.0gr%) , 1st Grade Asymptomatic
Burn
Rash
Decubitus Ulcer
Grade 2
Grade 3
Drug fever or rash requiring Drug fever, rash, oral/topical/intravenous bronchospasm, or medications hypotension requiring bronchoscopy or any other intervention Hemoglobin decrease ⱖ Requires massive 2.0gr% or requires transfusion or dialysis transfusion or intravenous steroids Requires platelets Associated with bleeding transfusion requiresing endoscopic or angiographic intervention Managed by medical Controlled by DC intervention without conversion or hemodynamic pacemaker associated compromise with transient hemodynamic compromise Angina pectoris managed by Angina pectoris medical intervention managed by without hemodynamic intervention compromise (catheterization, stent, PTCA) without myocardial or functional damage Mild symptoms, no Symptomatic. intervention required Pericardiocentesis required without hemodynamic compromise
Requires transfusion of blood products
Associated with bleeding requiring endoscopic or angiographic intervention
Symptoms controlled by topical treatment only. 2nd degree ⬍10% of body surface
2nd degree ⬎10% of body surface or 3rd degree ⬍10% body surface. Requires debridment or skin graft
Macular or papular Associated symptoms or eruption or lesions covering ⬍ 50% erythema without BSA associated symptoms Asymptomatic Local wound care; medical intervention indicated
Severe eruption; desquamation covering ⬎50% BSA Operative debridement or other invasive intervention indicated
Grade 4
Grade 5
Anaphylactic shock
Death
Renal failure, sever bronchospasm, or requires spelenectomy Active bleeding, or requires spelenectomy
Death
Death
Life threatening arrhythmia
Death
Myocardial infarction Iscemia associated with reduction of LV function or CHF.
Death
Life threatening or Tamponade Or pleural effusion managed by urgent pericardiocentesis or pericardial window Life threatening or uncontrolled bleeding
Death
Life threatening Requires massive or multiple debridments or 3rd degree ⬎10% body surface. Results in organ dysfunction. General exfoliative, ulcerative, or bullous dermatitis
Death
Life threatening consequences; major invasive intervention indicated
Death
Death
Death
(continued)
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
361.e2
Appendix A. Continued Complication
Grade 3
Grade 4
Grade 5
Wound (surgical site) infection
Erythema or discharge treated by stitch/ clips removal
Grade 1
Requires repeated irrigations or oral/ intravenous antibiotics
Requires bed-side debridement or percutaneous drainage or vacuum closure
Death
Wound dehiscence
Minor dehiscence demonstrated clinically without any intervention
Large dehiscence of a superficial wound (such as lumpectomy or mastectomy incision) demonstrated clinically requiring minimal intervention Requires oral/intravenous calcium supplements and vitamin D
Any dehiscence treated by VAC or other bedside procedures
Requires debridement in the operating room, vacuum closure or graft/flap for closure. Full dehiscence or any evisceration repaired in the operating room
Death
Oral or intravenous corticoor minieral-steroids indicated Mild symptoms, treated with oral or intravenous medications Mild symptoms treated with oral or intravenous medications Symptomatic, managed by intravenous antibiotics without intervention
Symptoms requiring ICU admission
Life threatening symptoms; unresponsive tetany or coma. Life-threatening; disabling
Symptoms requiring ICU admission
Life threatening consequences
Death
Symptoms requiring ICU admission
Life threatening consequences
Death
Managed by percutaneous drainage
Requires re-operation either for drainage or for diversion
Death
Managed by percutaneous drainage or ERCP
Requires re-operation
Death
Managed by percutaneous drainage
Requires re-operation
Death
Managed by percutaneous drainage, stenting, or nephrostomy
Requires re-operation
Death
Managed by percutaneous drainage
Requires re-operation ether for drainage.
Death
Complete, managed by endoscopy or intravenous TPN
Complete, managed by laparotomy or laparoscopy, adhesiolysis and / or bowel resection
Death
Hypoparathyroidism Asymptomatic (hypocalcemia)
Grade 2
Adrenal Insufficiency
Asymptomatic
Hyperthyroidism
Asymptomatic
Hypothyroidism
Asymptomatic
Anastomotic leak (gastrointestinal)
Asympthomatic, detetced by imaging without any medical or surgical intervention. Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging without intervention without any medical or surgical intervention. Confirmed by imaging Symptomatic, managed by intravenous antibiotics or or elevated amylase somatostatin without in drains’ fluid intervention Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging and Foley catheter without any medical or surgical intervention. Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging without intervention without any medical or surgical intervention. Complete, managed by Partial, managed by NPO and nasogastric tube and intravenous fluids intravenous fluids
Anastomotic leak (biliary)
Anastomotic leak (pancreatic)
Anastomotic leak (urinary)
Intrabdominal abscess/fluid collection
Intestinal obstruction
Associated with sever tetany requiring airway protections
Death
Death
(continued)
361.e3
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Continued Complication
Grade 4
Grade 5
Prolonged Ileus** or delayed gastric emptying or non functioning anastomosis Fistula (including, entero-cutaneous, entero-vesical, bilary, pancreatic) Bowel injury
Managed by NPO and Managed by NPO and total intravenous fluids parenteral nutrition for less than 30 days
Managed by NPO and total parenteral nutrition for more than 30 days
Managed by revision of anastomosis or other surgical intervention
Death
Minimal leak, spontaneous closure within 2 weeks
Minimal leak, spontaneous closure within 1 month.
Managed by NPO, somatostatin and total parenteral nutrition
Managed by surgery
Death
N/A
Identified and repaired during surgery without sequel.
Identified and repaired during surgery, additional drains required
Death
Small bowel ischemia
N/A
Managed by anticoagulation
Ischemic colitis
N/A
Managed conservatively
Managed by angiography with or without PTA Managed by angiography with or without PTA
Delayed identification requires reoperation, or results in stoma. Managed by surgery
Death
Pseudomembranous colitis
Incidental finding
Constipation
Occasional or intermittent symptoms
Managed by oral or intravenous antibiotics with or without prolongation of hospital stay* Persistent symptoms with use of laxatives
Life threatening, Associated with multi system organ failure or requiring colectomy Managed by surgery
Gastritis
Asymptomatic – radiographic or endoscopic findings only Asymptomatic – radiographic findings only Asymptomatic – radiographic or endoscopic findings only Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray
GI perforation
GI ulcer
Atelectasis
Pneumothorax
Pneumonia
Pleural effusion
Grade 1
Grade 2
Symptomatic; altered gastric function; treated with blood transfusions Medical intervention indicated Symptomatic; Medical intervention indicated
Symptomatic, managed by physiotherapy or deep suction Symptomatic, ⬍ 20% managed without intervention Managed by oral or intravenous antibiotics Symptomatic, managed without intervention
Grade 3
Severe colitis with ICU admission
Death
Death
Life threatening consequences or managed by surgery Associated with bleeding Life threatening consequences or requiring endoscopy managed by or angiographic surgery intervevtion. Requireing percutaneous Managed by surgery drainage
Death
Managed by endoscopy, angiography or radiology guided drainage Symptomatic, managed by bronchoscopy
Managed by surgery
Death
N/A
N/A
Symptomatic managed by tube thoracostomy insertion. managed by bronchoscopy or tube thoracostmy Symptomatic managed by pleural puncture or tube thoracostomy
Tension pneumothorax or requiring surgery Empyema requiring decortication
Death
Symptomatic managed by surgery.
Death
Requireing endoscopic manipulations
Death
Death
Death
(continued)
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
361.e4
Appendix A. Continued Complication
Grade 3
Grade 4
Grade 5
Pulmonary embolism
Asymptomatic, incidental finding on chest CT
Symptomatic, managed by oxygen and anticoagulation
Symptomatic, associated with hypoxia of PO2⬍60 managed by anticoagulation, fibrinolysis, or IVC filter
Death
Deep vein thrombosis
Asymptomatic, incidental finding
Symptomatic, managed by anticoagulation
Hemorrhage
Superficial hematoma or surgical site bleeding without action taken
Gastrointestinal hemorrhage
Minimal, managed without blood transfusion
Superficial hematoma or surgical site bleeding managed by suture, cautery, blood transfusion, or compression Managed with transfusion of ⬍4 units of packed RBCs
Symptomatic, managed by anticoagulation and fibrinolysis or vena cava filter Any hematoma or surgical site bleeding managed by angiography or bed side procedure
Life threatening, associated with hemodynamic compromise, right heart failure, requires ventilation or surgical embolectomy Limb threatening consequences requiring surgical management Any hemorrhage requiring surgery
Death
Cholecystitis
Asymptomatic
Managed by antibiotics
Controlled by surgery or associated with shock, massive transfusion or associated with permanent disability Managed by emergency surgery
Pancreatitis
Asymptomatic
Managed conservatively without complications or ICU admission; intravenous antibiotics
Pancreatitis requires nectosectomy, or requires ICU admission and associated with multi organ system failure
Death
Liver failure or hepatitis
Asymptomatic
Requires medical intervention
Death
Renal failure
Asymptomatic
Requires medical intervention
Fulminant hepatitis; severe encephalopathy, surgical shunt Permanent renal failure requires dialysis Urosepsis
Death
Delayed identification requires reoperation, results in sepsis
Death
Urinary tract infection
Urinary bladder injury
Grade 1
Grade 2
Managed by endoscopy, angiography or requires ⬎ than 4 units of packed RBCs
Managed by percutaneous chlecystostomy Necrotizing pancreatitis Pancreatitis associated with complications or requires ICU admission Or pancreattits that requires persutaneous or endoscopic intervension Requires TIPS or ascites drainage
Requires invasive intervention or temporary dialysis or ICU admission Asymptomatic Requires oral or intravenous Complicated antibiotics pyelonephritis or infection by multidrug resistant bacteria Identified and repaired Identified and repaired Identified and repaired during surgery. Foley during surgery. Foley during surgery catheter and drains without sequel catheter for longer than anticipated left
N/A
Death
Death
Death
(continued)
361.e5
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Continued Complication
Grade 1
Grade 2
Ureter or urtheral injury
N/A
Chyle or lymph leakage
Asymptomatic; clinical Symptomatic; medical or radiographic intervention indicated findings — Local wound care; medical intervention indicated
Soft tissue necrosis
Identified and repaired during surgery with stent
Grade 3
Identified and repaired during surgery with stent. Results in leak, urinoma or stenosis Managed by interventional radiology Bed side debridement
Grade 4
Grade 5
Delayed identification requires reoperation, results in sepsis Managed by surgery
Death
Death
Death
Laryngeal nerve dysfunction
Asymptomatic
Symptomatic; not interfering with ADL
Signs or symptoms interfering with ADL Interfering with ADL, intervention indicated
Cranial neuropathy/ neural injury Motor neuropathy (peripheral nerve injury-motor) Sensor neuropathy neuropathy (peripheral nerve injury-sensory) Seizure
Asymptomatic
Symptomatic; not interfering with ADL Symptomatic; not interfering with ADL
Symptomatic; interfering with ADL Symptomatic; interfering with ADL
Life threatening complication; requiring surgical debridement CVA; neurologic deficit ⬎ 24 hrs Life threatening; disabling Life threatening; tracheostomy indicated Life threatening; disabling Life threatening; disabling
Asymptomatic
Symptomatic; not interfering with ADL
Symptomatic; interfering with ADL
Life threatening; disabling
Death
N/A
One brief generalized seizure
Seizures in which consciousness is altered
Death
Phlebitis Compartment syndrome
N/A minimal symptoms treated by expectant observation
Present Managed by analgestics and observation
Requires phlebotomy Bedside fasciotomy
Seizures that are prolonged, repetitive, or difficult to control or seizures managed by surgical intervention N/A Requiering surgical intervention or associated with loss of limb or function
CNS cerebrovascular ischemia Encephalopathy
— —
Asymptomatic
Asymptomatic; radiographic findings Mild signs or symptoms
TIA ⬍ 24 hrs duration
Death Death Death
Death Death
N/A Death
*Prolonged hospital stay is defined as a time interval longer than twice the average length of stay for the performed procedure. **Prolonged ileus is defined as either ileus appearing after the resumption of oral intake or ileus associated with a prolonged time interval to oral intake of twice the average time interval for the performed procedure. DC, direct current; ECG, electrocardiogram; PTCA, percutaneous transluminal coronary angiography; LV, left ventricle; CHF, congestive heart failure; BSA, body surface area; VAC, vacuum assisted closure; ICU, intensive care unit; ERCP, endoscopic retrograde cholangiopancreaticography; TPN, total parenteral nutrition; NPO, nothing per os; N/A, not applicable; PTA, percutaneous transluminal angioptasty; CT, computed tomography; IVC, inferior vena cava; RBC, red blood cells; TIPS, transjugular intrahepatic portosystemic shunt; TIA, transient ischemic attack; CVA, cerebro vascular accident; ADL, activities of daily living.
a powerful driving force, which might curb health-care costs and improve quality.3 Accurate and uniform reporting of surgical complications is essential for quality-of-care measurement. Currently, there is no uniform reporting system for surgical complications. Adverse events reporting systems are used for clinical trials4 but are rarely used to report surgical complications. Institutional reporting is based mainly on a venerable tradition of weekly morbidity and mortality conferences, and clinical trial complications reporting is inconsistent and in many trials even underreported. Short-term surgical outcomes, such as length of hospital stay, 30-day mortality rate, operating time, and estimated blood loss, are routinely included in the data reported. Conclusive assessments of surgical procedures remain limited by the lack of consensus on how to define complications and to stratify them by severity.5 A surgical complication grading system must be precise, simple, and reproducible. Absence of a common grading system encouraged us to develop a five-tiered grading system for surgical complications. Based on systems reported for gastric and pancreatic
Growing demand for health care, rising costs, constrained resources, and variations in clinical practice have triggered interest in measuring and improving the quality of health care delivery. For a valuable quality assessment, relevant data on outcomes must be obtained in a standardized and reproducible fashion to allow comparison among different centers, between different therapies, and within a center over time.1,2 Objective and reliable outcomes data are also increasingly requested by patients and their representatives, health-care providers, and national and international health organizations. In addition, health policy makers point out that the availability of comparative data on individual institution’s and physician’s performance represents Disclosure Information: Nothing to disclose. Received October 9, 2008; Revised November 26, 2008; Accepted December 2, 2008. From the Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus, Jerusalem, Israel. Correspondence address: Aviram Nissan, MD, Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus, POB 24035, Jerusalem, 91240 Israel. email: [email protected]
© 2009 by the American College of Surgeons Published by Elsevier Inc.
355
ISSN 1072-7515/09/$36.00 doi:10.1016/j.jamcollsurg.2008.12.008
356
Mazeh et al
A Novel Grading System for Surgical Complications
operations,6,7 we adopted the same principles of a fivetiered grading system to develop a surgical complication reporting system. This study was undertaken to validate this surgical complication reporting system. We elected to validate the system on a cohort of patients who underwent colorectal operations, because colon or rectal resections are related to a variety of complications, are conducted in elective and emergency settings, and can be performed as a laparoscopic or open procedure. This would allow us to compare subgroups of patients and, as a result, learn more about the validity of the system suggested. This is a retrospective single-institution report of complications after colorectal operations based on a five-tiered system in a large cohort of patients.
METHODS A computerized complication reporting database was created using Microsoft Access software (Microsoft Corp). A five-tiered scale grading severity of complications was incorporated. All colon and rectal resections performed between January 1999 and December 2004 at the Department of Surgery, Hadassah-Hebrew University Medical Center Mount Scopus were retrieved from the operating theater database. All patients’ hospital and outpatient clinic charts were reviewed. Data recorded included age, gender, past medical history, medications, comorbidities, and indications for operation. Operative details recorded were type of operation, extent of resection, laparoscopic versus open operations, emergency versus elective procedure, and need for a stoma. Mechanical bowel preparation was performed in all elective patients using sodium phosphate solution, and bowel decontamination was performed with oral neomycin and metronidazole. Prophylactic antibiotics were administrated to all elective and emergent patients within 30 minutes of skin incision and included IV ampicillin, gentamycin, and metronidazole, unless contraindicated. Intermittent compression stockings were applied in all cases and subcutaneous low-molecularweight heparin was initiated 6 hours after operation. Variables we recorded included length of operation, blood loss and the need for transfusion, and bowel content spillage. Hospital and outpatient clinic charts were reviewed for detailed postoperative course, including all complications and interventions, length of stay, and readmissions within 30 days of the surgical procedure. All complications were classified and graded according to the Surgical Complication Severity Scoring System (Table 1) (Appendix A online). Complications were defined as any deviation from the normal postoperative course. This broad definition also includes asymptomatic documented complications. Preex-
J Am Coll Surg
Table 1. Examples of Complication Grades Grades
1
2
3 4
Examples
Episode of atrial fibrillation not requiring medical therapy Atelectasis proved by chest x-ray Wound infection opened at bedside Arrhythmia requiring -blockers for control Urinary tract infection requiring antibiotics Pneumonia Thromboembolic episode requiring anticoagulation therapy Prolonged ileus resulting in long hospital stay Bleeding requiring blood transfusion Gastrointestinal bleeding controlled by endoscopy Anastomotic leak drained by interventional radiology Anastomotic leak and sepsis requiring surgical repair Evisceration Bleeding requiring operation for control
isting medical conditions that did not change during the hospital course were not considered as adverse events. All complications were graded by severity according to the Surgical Complications Severity Scoring System. Grade 1 included minor asymptomatic complications that did not require medical therapy or radiologic or surgical intervention. Grade 2 complications were defined as complications requiring pharmacological therapy, minor interventions, or any prolongation of hospital stay (as defined in Appendix A online). Grade 3 complications were defined as complications that required intervention other than operations, including interventional radiology or endoscopy. Grade 4 included complications that required surgical intervention and complications resulting in permanent loss of function of an organ. Grade 5 complications included complications resulting in death. Statistical analysis
Summary statistics were performed using established methods. Association between categorical factors was studied with Fisher’s exact test or chi-square test, as appropriate. Association between noncategorical (continuous) variables was performed using Student’s t-test. SPSS statistical package version 13.0 (SPSS, Inc) was used for statistical analysis. Two-sided p value ⱕ 0.05 was set as significant.
RESULTS During the study period, a total of 408 patients underwent colon or rectal resections. The median age was 65 years (range, 17 to 96 years), 1 with a slight female predominance, 52%, versus 48% male. At least 1 comorbid condition was recorded for 62.3% of the patients. The most common indication for operation was malignancy in
Vol. 208, No. 3, March 2009
Mazeh et al
Table 2. Patient and Procedure Characteristics (n ⫽ 408) Age (y), median (range) 65 (17⫺96) n % Gender Male 196 48 Female 212 52 Patients with comorbidities ⱕ1 261 63 ⬎1 147 37 Indication for procedure Malignancy 316 77.5 Benign 92 22.5 IBD 21 5.1 Diverticulitis 37 9 Volvulus 12 2.9 Other 22 5.4 Procedure setting Elective 309 75.9 Emergency 99 24.3 Technique Open 323 79.2 Laparoscopic 85 20.8 Procedure type Right hemicolectomy 130 31.8 Left hemicolectomy 23 5.6 Sigmoid resection 97 23.7 Sleeve resection 5 1.2 Subtotal colectomy 68 16.6 Low anterior resection 73 17.8 Abdominoperineal resection 1 0.2 Total proctocolectomy 11 2.6 IBD, inflammatory bowel disease.
77.5% of the patients. Elective operations were performed in 75.7% of the patients, and 24.3% required emergency operation. Only 20.8% of patients underwent laparoscopic procedures, with 10.5% (9 patients) converted to open. These patients were included in the laparoscopic group to maintain an intention-to-treat analysis. Patient and procedure’s details are outlined in Table 2. Mean operative time was 142 minutes (range, 66 to 372 minutes). In 12 patients (2.9%), gross spillage was documented, but it was not substantially associated with a higher complication grade. Of 408 patients who were included in the study, 239 (58.6%) had an uneventful recovery without any complication. Two hundred forty-seven complications were recorded in 169 (41.4%) patients. Distribution of complications by grade is presented in Table 3. Grades 1 and 2 complications were recorded in 83 (20.3%) and 105 (25.7%) patients, respectively, and together they accounted for 76.1% of all complications. In 59 patients (14.5%), grades 3 to 5 complications were recorded (Table 3). Mortality rate
A Novel Grading System for Surgical Complications
357
Table 3. Incidence and Distribution of Complications by Grade Grade
1 2 3 4 5
n
All patients (%)
All complications (%)
83 105 25 27 7
20.3 25.7 6.1 6.6 1.7
33.6 42.5 10.1 10.9 2.8
(grade 5 complications) was 1.7%. The 10 most common complications and their incidence are presented in Table 4. Surgical site infection, blood transfusion, and atelectasis accounted for 18.6% of all complications. There was a significantly lower complication rate after elective operations compared with emergency procedures (35.9% versus 59.6%; p ⬍ 0.01). All seven patients with grade 5 complications (resulting in death) underwent emergency procedures. Complication rate in patients who underwent laparoscopic colon resections was significantly lower than in the open operation group (23.5% versus 46.1%; p ⬍ 0.0001). Because none of the emergency procedures was laparoscopic, a comparison of elective-only procedures showed no difference in the complication rate between elective open and laparoscopic procedures. There was also no significant difference in complication rate between colon and rectal procedures (42.9% versus 39.2%; p ⫽ 0.66). Complication rate was not statistically different in patients younger or older than 50 years of age (43.3% versus 34.5%; p ⫽ 0.14). Tables 5 and 6 summarize the incidence of the leading complications in the various subgroups. Number of complications was strongly associated with prolonged length of stay (Fig. 1). Median length of stay (LOS) of uncomplicated patients was 6 days. Among the 112 patients with 1 complication, median LOS was 9 days. Two and 3 complications were recorded in 39 and 16 paTable 4. Ten Most Common Complications and their Incidence Incidence Complication
n
%
Surgical site infection Intraabdominal abscess Blood transfusion Pseudomembranous colitis Atelectasis Pneumonia Anastomotic leak Arrhythmia Prolonged ileus Myocardial infarct
36 35 25 18 15 15 13 9 9 9
8.8 8.6 6.1 4.4 3.7 3.7 3.2 2.2 2.2 2.2
0.22
0.66 39.2 29 42.9 140 ⬍ 0.001 149 23.5 ⬍ 0.001
46.1
5 7 0.32 4.7 7 ⬍ 0.001
20 58.6 36.0
58
12.1 7
111 Total patients with complications
14 Grade 4
Grade 5
0.95
0.86 10.3 3 11.5 16 0.91 10.1 15 20 4 0.42 8.6 5
13 Grade 3
12.7
0.055 58.6
10.3 3
17 39.3
10 14
55 0.225
0.42 10.7
44.3 66
16 5
30 6
1 0.3
0.81 41.1
6.9 4
24 48 Grade 2
11.7
36 Grade 1
43.2
0.75
0.15 20.7 6 34.3 48 0.18 30.2 45 45 9 0.85 31 18
0.58
5 Myocardial infarct
32.4
2.7 2 2.1 7 0.12 2.8 9 — — 0.15 4.0 4
0.23
9 Prolonged ileus
1.6
1.3 1 2.4 8 0.12 2.8 9 — — 0.08 — —
5 Arrhythmia
2.9
4.0 3 1.8 6 0.12 2.8 9 — — 0.15 4.0 4
0.8
6 Pneumonia
1.6
1.3 1 3.6 12 0.06 4.0 13 — — 0.01 7.1 7
0.85
9 Anastomotic leak
1.9
2.7 2 3.3 11 0.84 3.1 10 3.5 3 0.58 4.0 4
0.64
11 Atelectasis
2.9
4.0 3 3.6 12 0.57 3.4 11 4.7 4 0.8 4.0 4
0.06
11 Pseudomembranous colitis
3.5
5.4 4 4.2 14 0.3 4.9 16 2.3 2 0.14 7.1 7
0.76
17 Blood transfusion
3.5
1.3 1 7.2 24 0.9 6.2 20 5.9 5 0.35 8.1 8
0.8
Intraabdominal abscess
5.5
9.5 7 8.4 28 0.15 9.6 31 4.7 4 0.84 8.1 8 27
Surgical site infection
8.7
p Value %
8.1 6
n %
9 30
n p Value
0.83 9.0
% n
29 8.2
% n
7 0.032
p Value %
14.1 14
n % n
22
Complication
7.1
Rectal operation (n ⴝ 74) Colon operation (n ⴝ 334) Open operation (n ⴝ 323) Laparoscopic operation (n ⴝ 85) Emergent operation (n ⴝ 99)
0.32
A Novel Grading System for Surgical Complications
Mazeh et al
Elective operation (n ⴝ 309)
Table 5. Incidence of Complications in the Different Subgroups (Elective Versus Emergent, Laparoscopic Versus Open Operation, and Colon Versus Rectal)
358
J Am Coll Surg
tients, with a median LOS of 16 and 17 days, respectively (p ⬍ 0.05). Only 2 patients had 4 complications, with a median LOS of 11.5 days. Complication grade was also associated with prolonged LOS (Fig. 2). Median LOS of grades 1 to 4 complications was 10.6, 13.8, 17.8, and 17.7 days, respectively (ANOVA, p ⬍ 0.0001). Grade 5 complications (perioperative mortality) did not necessarily result in prolonged hospitalization. Surgical site infection occurred in 8.8% of patients (n ⫽ 36). It was graded as grade 1 in 78% of the patients requiring bedside drainage of the wound. The remaining 22% were grade 2 complications, requiring antibiotic treatment for surrounding soft-tissue involvement. Median LOS in patients with surgical site infection was 12 days (range, 5 to 44 days), and it occurred as a single complication in only 44.4% of patients. Surgical site infection was more common after emergency operation compared with elective procedures (7.1% versus 14.1%; p ⬍ 0.05). Surprisingly, rate of surgical site infection was similar for laparoscopic and open operations (8.2% versus 9.0%). Intraabdominal abscess occurred in 8.6% of patients (n ⫽ 35). There were grade 2 intraabdominal abscesses treated with IV antibiotics only in 42.9% of the patients. In the remaining 57.1% patients, intraabdominal abscesses were treated by interventional radiology percutaneous drainage (grade 3). Presence of intraabdominal abscess was associated with a median LOS of 11 days (range, 5 to 45 days) and occurred as a single complication in only 42.9% of the patients. Anastomotic leak was defined as air outside the gastrointestinal lumen, fluid collection, abscess in close proximity to the anastomotic site, or leak of contrast dye at the anastomosis site by radiology studies or surgical exploration. Intraabdominal abscess was not defined as a leak unless met by the previously mentioned criteria for anastomotic leakage. Anastomotic leak occurred in 3.2% of patients (n ⫽ 13). All patients with a diagnosis of anastomotic leak were reoperated (grade 4). Anastomotic leak was associated with a median LOS of 17 days (range, 6 to 45 days), was accompanied by more than 1 other complication in the majority of the patients (61.5%), and by mortality of a single patient (7.7%). Blood transfusion was defined as a grade 2 complication if 2 U of packed red blood cells or more were administrated postoperatively. It occurred in 6.1% of the patients (n ⫽ 25) and did not require additional surgical intervention in any of the patients. It was associated with a median LOS of 8 days (range, 5 to 24 days), which was not considerably different from the cohort of patients without complica-
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
359
Table 6. Incidence of Complications in the Different Subgroups (Comorbidities and Age)
Complications
Surgical site infection Intraabdominal abscess Blood transfusion Pseudomembranous colitis Atelectasis Anastomotic leak Pneumonia Arrhythmia Prolonged ileus Myocardial infarct Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Total patients with complications
0 to 1 comorbidity (n ⴝ 261) n %
>1 Comorbidity (n ⴝ 147) n %
24 21 14 10 10 9 11 6 5 4 29 46 10 15 1
9.2 8.0 5.4 3.8 3.8 3.4 4.2 2.3 1.9 1.5 28.7 45.5 9.9 14.9 1
12 14 11 8 5 4 2 3 4 5 22 26 8 8 6
8.2 9.5 7.5 5.4 3.4 2.7 1.4 2.0 2.7 3.4 32.4 38.2 11.8 11.8 8.8
101
38.7
68
46.3
tions. It presented as a single complication in 60% of patients. Respiratory complications (atelectasis and pneumonia documented by chest x-ray) were graded as 1 and 2, respectively. Atelectasis and pneumonia each occurred in 3.7% of the patients (n ⫽ 15) and were associated with a median LOS of 8 and 12 days, respectively. Atelectasis was a single complication in 53.3% of the patients, although pneumonia was associated with more than 1 other complication in 53.3% of patients. Pneumonia was more common after emergency operation compared with elective procedures (7.1% versus 1.9%; p ⫽ 0.01).
Figure 1. Median length of stay as a function of number of complications for the entire cohort of patients (n ⫽ 408). Length of stay was considerably longer for patients with two to three complications compared with those with none to one complication.
Age older than 50 y (n ⴝ 321) n %
Age younger than 50 y (n ⴝ 87) n %
0.74 0.6 0.4 0.44 0.8 0.7 0.11 0.86 0.6 0.2 0.6 0.34 0.7 0.56 0.012
29 27 20 16 12 7 10 9 8 9 44 63 12 13 7
9 8.4 6.2 5 3.7 2.2 3.1 2.8 2.5 2.8 31.7 45.3 8.6 9.4 5
7 8 5 2 3 6 3 — 1 — 10 9 5 6 —
8.0 9.2 5.7 2.3 3.4 6.9 3.4 — 1.1 — 33.3 17.8 16.7 20.7 —
0.77 0.82 0.86 0.28 0.9 0.026 0.87 0.114 0.45 0.11 0.86 0.12 0.18 0.08 0.209
0.13
139
43.3
30
34.5
0.14
p Value
p Value
DISCUSSION The definition of a surgical complication is “any deviation from the expected steady recovery after a surgical operation.”8 This broad definition results in considerable variation in surgical complication reporting, even in the setting of controlled clinical trials. Currently, there are no uniform requirements of surgical journals for reporting complications. The importance of a uniform, simple, and reproducible reporting system for surgical complications cannot be emphasized enough. Such a system will set the basis for medical centers worldwide to capture, analyze, and report
Figure 2. Median length of stay as a function of complication grade for the entire cohort of patients (n ⫽ 408). Length of stay was significantly longer (ANOVA, p ⬍ 0.0001) for patients with highergrade complication. Grade 5 (perioperative mortality) did not necessarily result in prolonged hospitalization.
360
Mazeh et al
A Novel Grading System for Surgical Complications
comparable surgical complications. The ability to measure quality of care delivered to a group of patients in a given institution or country is dependent not only on the efficacy of the surgical therapy but also on the morbidity and mortality it carries. The ability of a patient or a health-care provider to choose the best therapy for a given disease or disorder is not only dependent on the efficacy of the evaluated therapy but also on its related morbidity in general and at a certain institute in particular. Unlike clinical trials involving pharmaceutical agents, the morbidity of surgical procedures in clinical trials has not been subjected to a uniform reporting system. Some complications, such as anastomotic leakage or wound dehiscence, are severe and would generally be reported. But there is large variability in the reporting of many intraoperative or postoperative complications that are less severe or seemingly less important or unimportant. The need for uniform reporting of surgical complications by institutions and health-care providers is not limited to clinical trials. Growing demand for quality health care, constrained resources, and variation in clinical practice has triggered a growing interest in measuring and improving the quality of health-care delivery. Objective and reliable outcomes data are increasingly requested by patients and their health-care providers.1-3 Conclusive assessments of surgical procedures remain limited by the lack of consensus on how to define complications and stratify them by severity. The current study was undertaken to validate a surgical complication reporting system on a large enough cohort of patients to measure differences between procedures and try to set a benchmark for surgical complication reporting in the future. We adopted a five-tiered reporting system from previous reports and from the National Cancer Institute Common Terminology Criteria for Adverse Events 3.0 adverse event reporting system.4 We validated the system by reviewing a cohort of colorectal patients operated on in a predetermined time period. To validate our severity score, we compared each grade to well-established and measurable outcomes, eg, LOS. Our results show that grades 3 to 5 complications are well within the range reported in the literature. The rate of grades 1 and 2 complications was much higher than reported in the surgical literature, suggesting previous underreporting of such complications. The distribution of complications, as expected, was with a majority of grades 1 and 2 complications and with a lesser degree of severe grades 3 and 4 or fatal complications (grade 5). We were also able to show that the complications rate, and especially the mortality rate, is much lower in elective operations as compared with emergency operations.
J Am Coll Surg
In this cohort of patients, we documented a variety of complications and, by using the suggested system, we were able to define a better view of the complication rate related to colorectal operations in our institution. Reporting an overall complication rate of 41% for colorectal operations is extremely high. By using the five-tiered grading system, we were able to show that our complication rate is actually within the limits reported by others and that the additional grades 1 and 2 complications that we accurately reported are usually grossly underreported by many. We showed that grades 1 and 2 complications are associated with a substantially prolonged hospital stay (Fig. 2) and should be vigorously observed, reported, and prevented. Mortality rate reported in the present study compares favorably with other reported postoperative mortality, ranging from 2% to 9%. Morbidity rate of 41% is higher than reported morbidity after colorectal resections, ranging from 18% to 37% only.9-15 In this report, only 14.5% of the patients had grades 3 to 5 complications, which are usually reported in the literature. Our data show an increased risk for fatal (grade 5) complications in patients with 2 or more comorbid conditions. Because all our grade 5 complications occurred after emergency operations, this might be the most crucial contributing factor. Because of the small number of grade 5 complications, defining the weighted value of each variant by a multivariate analysis was not feasible. Overall complication rate was also considerably higher in patients who underwent emergency operations. This is consistent with previous studies.10,12-17 The complication rate in patients who underwent laparoscopic operations was considerably lower than in the open operation group, but none of the emergency operations was performed laparoscopically. When excluding emergency operations from the open operation group, there was no difference in the complication rate between open and laparoscopic procedures. This difference reflects the higher complication rate related to emergency procedures. There are conflicting reports in the medical literature as to whether or not laparoscopic operations result in a lower complication rate.18-20 There are reports linking complications with age or anatomic location (colon versus rectum).12,14 We did not find any statistically significant difference in complication rate associated to age older than 50 years or with rectal operations. Comorbid conditions were related to surgical complications, with a higher complication rate in patients with two or more comorbid conditions.9-15,17 Reporting of surgical complications is important for quality assurance. There are many barriers for creating a uniform and robust reporting of surgical complications. It is often conceived as reporting failure and not success, so it
Vol. 208, No. 3, March 2009
Mazeh et al
is only natural that many institutions tend to underscore their complication rate. An accurate and uniform reporting system, on the one hand, will allow surgeons to report complication centers and improve outcomes. On the other hand, this can create a basis for patients’ selections of lowrate complication centers and not centers of excellence that can treat more complicated patients with higher complication rates. Another barrier is that many and different reporting systems are currently used, most adopted from medical reporting systems, and do not cover the spectra of surgical complications. This can be overcome by presenting validated and prospective data supporting use of a Surgical Complication Grading System. We conclude that the surgical complication grading system presented is precise, simple, and reproducible. With minor modifications, it can easily be adopted and adjusted to any other health-care system. In an era of quality control, surgical quality improvement, and surgical care improvement, it offers a reliable and easy to use asset. Author Contributions Study conception and design: Mazeh, Freund, Nissan Acquisition of data: Mazeh, Samet, Abu-Wasel, Beglaibter, Grinbaum, Cohen, Pinto, Hamburger, Nissan Analysis and interpretation of data: Mazeh, Hamburger, Nissan Drafting of manuscript: Mazeh, Freund, Nissan Critical revision: Mazeh, Samet, Abu-Wasel, Beglaibter, Grinbaum, Cohen, Pinto, Hamburger, Freund, Nissan
REFERENCES 1. Clavien PA, Sanabria JR, Strasberg SM. Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery 1992;111:518–526. 2. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–213. 3. Daley J, Henderson WG, Khuri SF. Risk-adjusted surgical outcomes. Annu Rev Med 2001;52:275–287. 4. Trotti A, Colevas AD, Setser A, et al. CTCAE v3.0: development of a comprehensive grading system for the adverse effects of cancer treatment. Semin Radiat Oncol 2003;13:176–181.
A Novel Grading System for Surgical Complications
361
5. Martin RC 2nd, Brennan MF, Jaques DP. Quality of complication reporting in the surgical literature. Ann Surg 2002;235: 803–813. 6. Martin RC 2nd, Jacques DP, Brennan MF, Karpeh M. Achieving R0 resection for locally advanced gastric cancer: is it worth the risk of multiorgan resection? J Am Coll Surg 2002;194:568– 577. 7. Grobmyer SR, Pieracci FM, Allen PJ, et al. Defining morbidity after pancreaticoduodenectomy: use of a prospective complication grading system. J Am Coll Surg 2007;204:356–364. 8. Townsend CM, Beaucham RD, Evers BM, et al, eds. Sabiston textbook of surgery: the biological basis of modern surgical practice. 17th ed. Philadelphia: WB Saunders; 2004. 9. Khuri SF, Henderson WG, DePalma RG, et al. Determinants of long-term survival after major surgery and the adverse effect of postoperative complications. Ann Surg 2005;242:326⫺341; discussion 341⫺343. 10. Prystowsky JB, Bordage G, Feinglass JM. Patient outcomes for segmental colon resection according to surgeon’s training, certification, and experience. Surgery 2002;132:663–670. 11. Cook A, Hyman N. Quality assessment and improvement in colon and rectal surgery. Dis Colon Rectum 2004;47:2195–2201. 12. Alves A, Panis Y, Mathieu P, et al. Postoperative mortality and morbidity in French patients undergoing colorectal surgery: results of a prospective multicenter study. Arch Surg 2005;140:278–283. 13. Tekkis PP, Kessaris N, Kocher HM, et al. Evaluation of POSSUM and P-POSSUM scoring systems in patients undergoing colorectal surgery. Br J Surg 2003;90:340–345. 14. Bokey EL, Chapuis PH, Fung C, et al. Postoperative morbidity and mortality following resection of the colon and rectum for cancer. Dis Colon Rectum 1995;38:480–486. 15. Ko CY, Chang JT, Chaudhry S, Kominski G. Are high-volume surgeons and hospitals the most important predictors of inhospital outcome for colon cancer resection? Surgery 2002;132: 268–273. 16. Longo WE, Virgo KS, Johnson FE, et al. Risk factors for morbidity and mortality after colectomy for colon cancer. Dis Colon Rectum 2000;43:83–91. 17. Harmon JW, Tang DG, Gordon TA, et al. Hospital volume can serve as a surrogate for surgeon volume for achieving excellent outcomes in colorectal resection. Ann Surg 1999;230:404–411. 18. Degiuli M, Mineccia M, Bertone A, et al. Outcome of laparoscopic colorectal resection. Surg Endosc 2004;18:427–432. 19. Senagore AJ, Delaney CP. A critical analysis of laparoscopic colectomy at a single institution: lessons learned after 1000 cases. Am J Surg 2006;191:377–380. 20. Schwenk W, Haase O, Neudecker J, Muller JM. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;20:CD003145.
361.e1
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Surgical Complications Grading System. Surgical Complications Grading System Complication
Grade 1
Allergic Reaction/ Hypersensitivity
Transient drug fever or rash
Hemolysis (transfusion reaction)
Laboratory evidence without decrease in hemoglobin
Thrombocytopenia
⬍ 50,000/mm3
Cardiac arrhythmia
Asympthomatic intervention notindicated
Cardiac ischemia
Asymptomatic (transient ECG changes)
Pericardial effusion
Asymptomatic
DIC (disseminated intravascular coagulation)
Mild symptoms (hematomas, Hemoglobin decrease ⱕ 2.0gr%) , 1st Grade Asymptomatic
Burn
Rash
Decubitus Ulcer
Grade 2
Grade 3
Drug fever or rash requiring Drug fever, rash, oral/topical/intravenous bronchospasm, or medications hypotension requiring bronchoscopy or any other intervention Hemoglobin decrease ⱖ Requires massive 2.0gr% or requires transfusion or dialysis transfusion or intravenous steroids Requires platelets Associated with bleeding transfusion requiresing endoscopic or angiographic intervention Managed by medical Controlled by DC intervention without conversion or hemodynamic pacemaker associated compromise with transient hemodynamic compromise Angina pectoris managed by Angina pectoris medical intervention managed by without hemodynamic intervention compromise (catheterization, stent, PTCA) without myocardial or functional damage Mild symptoms, no Symptomatic. intervention required Pericardiocentesis required without hemodynamic compromise
Requires transfusion of blood products
Associated with bleeding requiring endoscopic or angiographic intervention
Symptoms controlled by topical treatment only. 2nd degree ⬍10% of body surface
2nd degree ⬎10% of body surface or 3rd degree ⬍10% body surface. Requires debridment or skin graft
Macular or papular Associated symptoms or eruption or lesions covering ⬍ 50% erythema without BSA associated symptoms Asymptomatic Local wound care; medical intervention indicated
Severe eruption; desquamation covering ⬎50% BSA Operative debridement or other invasive intervention indicated
Grade 4
Grade 5
Anaphylactic shock
Death
Renal failure, sever bronchospasm, or requires spelenectomy Active bleeding, or requires spelenectomy
Death
Death
Life threatening arrhythmia
Death
Myocardial infarction Iscemia associated with reduction of LV function or CHF.
Death
Life threatening or Tamponade Or pleural effusion managed by urgent pericardiocentesis or pericardial window Life threatening or uncontrolled bleeding
Death
Life threatening Requires massive or multiple debridments or 3rd degree ⬎10% body surface. Results in organ dysfunction. General exfoliative, ulcerative, or bullous dermatitis
Death
Life threatening consequences; major invasive intervention indicated
Death
Death
Death
(continued)
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
361.e2
Appendix A. Continued Complication
Grade 3
Grade 4
Grade 5
Wound (surgical site) infection
Erythema or discharge treated by stitch/ clips removal
Grade 1
Requires repeated irrigations or oral/ intravenous antibiotics
Requires bed-side debridement or percutaneous drainage or vacuum closure
Death
Wound dehiscence
Minor dehiscence demonstrated clinically without any intervention
Large dehiscence of a superficial wound (such as lumpectomy or mastectomy incision) demonstrated clinically requiring minimal intervention Requires oral/intravenous calcium supplements and vitamin D
Any dehiscence treated by VAC or other bedside procedures
Requires debridement in the operating room, vacuum closure or graft/flap for closure. Full dehiscence or any evisceration repaired in the operating room
Death
Oral or intravenous corticoor minieral-steroids indicated Mild symptoms, treated with oral or intravenous medications Mild symptoms treated with oral or intravenous medications Symptomatic, managed by intravenous antibiotics without intervention
Symptoms requiring ICU admission
Life threatening symptoms; unresponsive tetany or coma. Life-threatening; disabling
Symptoms requiring ICU admission
Life threatening consequences
Death
Symptoms requiring ICU admission
Life threatening consequences
Death
Managed by percutaneous drainage
Requires re-operation either for drainage or for diversion
Death
Managed by percutaneous drainage or ERCP
Requires re-operation
Death
Managed by percutaneous drainage
Requires re-operation
Death
Managed by percutaneous drainage, stenting, or nephrostomy
Requires re-operation
Death
Managed by percutaneous drainage
Requires re-operation ether for drainage.
Death
Complete, managed by endoscopy or intravenous TPN
Complete, managed by laparotomy or laparoscopy, adhesiolysis and / or bowel resection
Death
Hypoparathyroidism Asymptomatic (hypocalcemia)
Grade 2
Adrenal Insufficiency
Asymptomatic
Hyperthyroidism
Asymptomatic
Hypothyroidism
Asymptomatic
Anastomotic leak (gastrointestinal)
Asympthomatic, detetced by imaging without any medical or surgical intervention. Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging without intervention without any medical or surgical intervention. Confirmed by imaging Symptomatic, managed by intravenous antibiotics or or elevated amylase somatostatin without in drains’ fluid intervention Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging and Foley catheter without any medical or surgical intervention. Symptomatic, managed by Asympthomatic, intravenous antibiotics detetced by imaging without intervention without any medical or surgical intervention. Complete, managed by Partial, managed by NPO and nasogastric tube and intravenous fluids intravenous fluids
Anastomotic leak (biliary)
Anastomotic leak (pancreatic)
Anastomotic leak (urinary)
Intrabdominal abscess/fluid collection
Intestinal obstruction
Associated with sever tetany requiring airway protections
Death
Death
(continued)
361.e3
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Continued Complication
Grade 4
Grade 5
Prolonged Ileus** or delayed gastric emptying or non functioning anastomosis Fistula (including, entero-cutaneous, entero-vesical, bilary, pancreatic) Bowel injury
Managed by NPO and Managed by NPO and total intravenous fluids parenteral nutrition for less than 30 days
Managed by NPO and total parenteral nutrition for more than 30 days
Managed by revision of anastomosis or other surgical intervention
Death
Minimal leak, spontaneous closure within 2 weeks
Minimal leak, spontaneous closure within 1 month.
Managed by NPO, somatostatin and total parenteral nutrition
Managed by surgery
Death
N/A
Identified and repaired during surgery without sequel.
Identified and repaired during surgery, additional drains required
Death
Small bowel ischemia
N/A
Managed by anticoagulation
Ischemic colitis
N/A
Managed conservatively
Managed by angiography with or without PTA Managed by angiography with or without PTA
Delayed identification requires reoperation, or results in stoma. Managed by surgery
Death
Pseudomembranous colitis
Incidental finding
Constipation
Occasional or intermittent symptoms
Managed by oral or intravenous antibiotics with or without prolongation of hospital stay* Persistent symptoms with use of laxatives
Life threatening, Associated with multi system organ failure or requiring colectomy Managed by surgery
Gastritis
Asymptomatic – radiographic or endoscopic findings only Asymptomatic – radiographic findings only Asymptomatic – radiographic or endoscopic findings only Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray Asymptomatic, incidental finding on chest X-ray
GI perforation
GI ulcer
Atelectasis
Pneumothorax
Pneumonia
Pleural effusion
Grade 1
Grade 2
Symptomatic; altered gastric function; treated with blood transfusions Medical intervention indicated Symptomatic; Medical intervention indicated
Symptomatic, managed by physiotherapy or deep suction Symptomatic, ⬍ 20% managed without intervention Managed by oral or intravenous antibiotics Symptomatic, managed without intervention
Grade 3
Severe colitis with ICU admission
Death
Death
Life threatening consequences or managed by surgery Associated with bleeding Life threatening consequences or requiring endoscopy managed by or angiographic surgery intervevtion. Requireing percutaneous Managed by surgery drainage
Death
Managed by endoscopy, angiography or radiology guided drainage Symptomatic, managed by bronchoscopy
Managed by surgery
Death
N/A
N/A
Symptomatic managed by tube thoracostomy insertion. managed by bronchoscopy or tube thoracostmy Symptomatic managed by pleural puncture or tube thoracostomy
Tension pneumothorax or requiring surgery Empyema requiring decortication
Death
Symptomatic managed by surgery.
Death
Requireing endoscopic manipulations
Death
Death
Death
(continued)
Vol. 208, No. 3, March 2009
Mazeh et al
A Novel Grading System for Surgical Complications
361.e4
Appendix A. Continued Complication
Grade 3
Grade 4
Grade 5
Pulmonary embolism
Asymptomatic, incidental finding on chest CT
Symptomatic, managed by oxygen and anticoagulation
Symptomatic, associated with hypoxia of PO2⬍60 managed by anticoagulation, fibrinolysis, or IVC filter
Death
Deep vein thrombosis
Asymptomatic, incidental finding
Symptomatic, managed by anticoagulation
Hemorrhage
Superficial hematoma or surgical site bleeding without action taken
Gastrointestinal hemorrhage
Minimal, managed without blood transfusion
Superficial hematoma or surgical site bleeding managed by suture, cautery, blood transfusion, or compression Managed with transfusion of ⬍4 units of packed RBCs
Symptomatic, managed by anticoagulation and fibrinolysis or vena cava filter Any hematoma or surgical site bleeding managed by angiography or bed side procedure
Life threatening, associated with hemodynamic compromise, right heart failure, requires ventilation or surgical embolectomy Limb threatening consequences requiring surgical management Any hemorrhage requiring surgery
Death
Cholecystitis
Asymptomatic
Managed by antibiotics
Controlled by surgery or associated with shock, massive transfusion or associated with permanent disability Managed by emergency surgery
Pancreatitis
Asymptomatic
Managed conservatively without complications or ICU admission; intravenous antibiotics
Pancreatitis requires nectosectomy, or requires ICU admission and associated with multi organ system failure
Death
Liver failure or hepatitis
Asymptomatic
Requires medical intervention
Death
Renal failure
Asymptomatic
Requires medical intervention
Fulminant hepatitis; severe encephalopathy, surgical shunt Permanent renal failure requires dialysis Urosepsis
Death
Delayed identification requires reoperation, results in sepsis
Death
Urinary tract infection
Urinary bladder injury
Grade 1
Grade 2
Managed by endoscopy, angiography or requires ⬎ than 4 units of packed RBCs
Managed by percutaneous chlecystostomy Necrotizing pancreatitis Pancreatitis associated with complications or requires ICU admission Or pancreattits that requires persutaneous or endoscopic intervension Requires TIPS or ascites drainage
Requires invasive intervention or temporary dialysis or ICU admission Asymptomatic Requires oral or intravenous Complicated antibiotics pyelonephritis or infection by multidrug resistant bacteria Identified and repaired Identified and repaired Identified and repaired during surgery. Foley during surgery. Foley during surgery catheter and drains without sequel catheter for longer than anticipated left
N/A
Death
Death
Death
(continued)
361.e5
Mazeh et al
A Novel Grading System for Surgical Complications
J Am Coll Surg
Appendix A. Continued Complication
Grade 1
Grade 2
Ureter or urtheral injury
N/A
Chyle or lymph leakage
Asymptomatic; clinical Symptomatic; medical or radiographic intervention indicated findings — Local wound care; medical intervention indicated
Soft tissue necrosis
Identified and repaired during surgery with stent
Grade 3
Identified and repaired during surgery with stent. Results in leak, urinoma or stenosis Managed by interventional radiology Bed side debridement
Grade 4
Grade 5
Delayed identification requires reoperation, results in sepsis Managed by surgery
Death
Death
Death
Laryngeal nerve dysfunction
Asymptomatic
Symptomatic; not interfering with ADL
Signs or symptoms interfering with ADL Interfering with ADL, intervention indicated
Cranial neuropathy/ neural injury Motor neuropathy (peripheral nerve injury-motor) Sensor neuropathy neuropathy (peripheral nerve injury-sensory) Seizure
Asymptomatic
Symptomatic; not interfering with ADL Symptomatic; not interfering with ADL
Symptomatic; interfering with ADL Symptomatic; interfering with ADL
Life threatening complication; requiring surgical debridement CVA; neurologic deficit ⬎ 24 hrs Life threatening; disabling Life threatening; tracheostomy indicated Life threatening; disabling Life threatening; disabling
Asymptomatic
Symptomatic; not interfering with ADL
Symptomatic; interfering with ADL
Life threatening; disabling
Death
N/A
One brief generalized seizure
Seizures in which consciousness is altered
Death
Phlebitis Compartment syndrome
N/A minimal symptoms treated by expectant observation
Present Managed by analgestics and observation
Requires phlebotomy Bedside fasciotomy
Seizures that are prolonged, repetitive, or difficult to control or seizures managed by surgical intervention N/A Requiering surgical intervention or associated with loss of limb or function
CNS cerebrovascular ischemia Encephalopathy
— —
Asymptomatic
Asymptomatic; radiographic findings Mild signs or symptoms
TIA ⬍ 24 hrs duration
Death Death Death
Death Death
N/A Death
*Prolonged hospital stay is defined as a time interval longer than twice the average length of stay for the performed procedure. **Prolonged ileus is defined as either ileus appearing after the resumption of oral intake or ileus associated with a prolonged time interval to oral intake of twice the average time interval for the performed procedure. DC, direct current; ECG, electrocardiogram; PTCA, percutaneous transluminal coronary angiography; LV, left ventricle; CHF, congestive heart failure; BSA, body surface area; VAC, vacuum assisted closure; ICU, intensive care unit; ERCP, endoscopic retrograde cholangiopancreaticography; TPN, total parenteral nutrition; NPO, nothing per os; N/A, not applicable; PTA, percutaneous transluminal angioptasty; CT, computed tomography; IVC, inferior vena cava; RBC, red blood cells; TIPS, transjugular intrahepatic portosystemic shunt; TIA, transient ischemic attack; CVA, cerebro vascular accident; ADL, activities of daily living.