- Email: [email protected]
Risk factors affecting morbidity and mortality following emergency laparotomy for small bowel obstruction: A retrospective cohort study
International Journal of Surgery 28 (2016) 63e68
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original research
Risk factors affecting morbidity and mortality following emergency laparotomy for small bowel obstruction: A retrospective cohort study € genur b Maja Haunstrup Jeppesen a, b, *, Mai-Britt Tolstrup a, Sara Kehlet Watt b, Ismail Go a b
Department of Gastroenterology, Surgical Section, Copenhagen University Hospital Herlev, Herlev Ringvej 75, 2730 Herlev, Denmark Department of Surgery, Zealand University Hospital Køge, Lykkebæksvej 1, 4600 Koege, Denmark
h i g h l i g h t s 30-day morbidity and mortality rates were 28% and 13%, respectively. Six independent risk factors for morbidity were identified. Five independent risk factors for mortality were identified. Identified risk factors included chronic nephropathy and low physical performance.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 November 2015 Accepted 6 February 2016 Available online 18 February 2016
Introductions: To identify risk factors for increased 30-day morbidity and mortality, using standardized measuring tools for the characterization of complications after emergency surgery for small bowel obstruction. Methods: A retrospective cohort study including patients treated with emergency laparotomy for small bowel obstruction at a Copenhagen University Hospital (2009e2013). Complications were evaluated according to the Clavien-Dindo classification. Results: A total of 323 patients were included. The overall 30-day morbidity and mortality rates were 28% and 13%, respectively. Six covariates were identified as independent risk factors associated with morbidity by multiple logistic regression analysis. The highest odds for morbidity were seen in patients with chronic nephropathy (Odds Ratio [OR] ¼ 3.9; 95% CI 1.3e15.1), and in patients with a daily use of steroids (OR ¼ 3.5; 95% CI 1.2e10.4). Five independent risk factors were associated with increased odds for mortality. Patients with low physical performance (OR ¼ 3.4; 95% CI 1.3e8.2) or metabolic disorders (OR ¼ 3.2; 95% CI 1.2e8.5) had the highest risk of mortality. Conclusions: Morbidity and mortality rates were high in this study compared with other studies. Several comorbid conditions were associated with morbidity and mortality. These results may aid the acute care surgeon in identifying patients with a high-risk for postoperative complications and fatal outcomes. © 2016 IJS Publishing Group Limited. Published by Elsevier Ltd. All rights reserved.
Keywords: Emergency laparotomy Small bowel obstruction Outcomes Mortality Morbidity
1. Introduction Small bowel obstruction (SBO) is one of the most frequent indications for emergency laparotomy surgery, and is known as a high-risk procedure with morbidity and mortality rates at 20e30% and 3e5%, respectively [1e4]. Adhesions are the most frequent
* Corresponding author. Department of Surgery, Zealand University Hospital Køge, Lykkebæksvej 1, 4600 Koege, Denmark. E-mail addresses: [email protected] (M.H. Jeppesen), m.tolstrup@dadlnet. dk (M.-B. Tolstrup), [email protected] (S. Kehlet Watt), igo@regionsjaelland. €genur). dk (I. Go
reason for SBO, accounting for almost 75% of all cases, followed by malignancies, inflammatory bowel disease and hernias [5]. In recent years more interest has been taken in describing postoperative outcomes after abdominal emergencies, and in searching for preoperative factors that predispose poor outcome [6e8]. This information may aid the acute care surgeon in identifying patients with a high-risk for postoperative complications and fatal outcomes and eventually initiate interventions to prevent these adverse outcomes. Previous studies have identified potential risk factors such as surgical delay, increased age, high ASA score (American Society of Anaesthesiologists), septic shock, obesity, and comorbidities as factors associated with morbidity and mortality
http://dx.doi.org/10.1016/j.ijsu.2016.02.059 1743-9191/© 2016 IJS Publishing Group Limited. Published by Elsevier Ltd. All rights reserved.
64
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
[1e4,9]. There is a need for large studies that characterize these patients with focus on the consequences of their complications in a standardized method. The aim of this study was to use standardized measures for the characterization of complications after emergency surgery for SBO, and to investigate potential risk factors for increased 30-day morbidity and mortality. 2. Methods 2.1. Study design and patients This study is a retrospective cohort study conducted at a tertiary university hospital situated in the greater Copenhagen area, serving a population of 432.000 citizens. All patients are covered by public health insurance and pay no visit fees. Patients were included if any gastrointestinal emergency surgical procedure, laparoscopy or laparotomy, was performed from May 2009 to April 2013. The study was approved by the Danish Data Collection Agency (HEH-2013034 I-Suite no: 02336). 2.2. Variables A group of seven researchers collected pre- and postoperative data by review of medical records. This was done as part of creating a standardized database containing perioperative information of all emergency abdominal procedures performed at the hospital. Data for this study population consisted of all patients diagnosed with SBO, and treated with emergency laparotomy. These patients were identified and extracted from the database by the NOMESCO Classification of Surgical Procedure (NCSP) codes KJFB00 (small bowel resection), KJFK00 (division of adhesive band in intestinal obstruction), and KJFK10 (freeing of adhesions in intestinal obstruction). Exclusion criterias were death during surgery and patients aged <18 yr. Patients who had any other procedure simultaneously with the SBO (e.g. ileostomy) were excluded. If a patient had more than one incidence of SBO managed surgically during the study period, only data from the index procedure was analysed for complications. Intraoperative data was collected from the Danish Anaesthesia Database and from the surgery scheduling software. Pre-existing comorbidities and conditions were registered from the admission report, and were considered as existing if the condition was medically treated or if previous treatment was described. The comorbidities were defined as follows: chronic nephropathy; elevated se-creatinine or dialysis treatment, obstructive pulmonary disease; chronic obstructive pulmonary disease (COPD) or asthma, malignant disease; actual malignant disease (disseminated and not disseminated), metabolic disorder; diabetes or thyroid disease, cardiovascular disease; hypertension, atrial fibrillation, ischaemic heart disease or any history of cerebrovascular disease. Performance score was evaluated by the Eastern Cooperative Oncology Group (EC0G) score [10], which is a six step scale ranging from 0 to 5 (fully active to death). Excessive alcohol consumption was defined as a weekly alcohol intake of >7 units per week for women and >14 units per week for men. The time from the decision to operate was made, and surgical intervention was initiated, was registered. Perioperative blood transfusion was defined as being transfused with blood products between 7 days before and 30 days after surgery. Surgical contamination was evaluated by the wound classification described by Berard et al. [11]. Readmission related to the surgical procedure was also registered. Postoperative complications were registered according to the Clavien-Dindo (CD) classification [12,13]. The CD classification is a validated ranking system, which classifies postoperative complications based on the therapy used to treat the complication. Any
complication is graded on a 5-step scale. CD score 1 is any deviation from the normal postoperative course, without the need for pharmacological treatment or surgical, endoscopic, and radiological interventions. CD score 5 is death of the patient [13]. Complications were divided into two categories: surgical and medical. Surgical complications include bleeding, wound dehiscence, postoperative ileus, wound infection, intraabdominal abscesses and anastomosis leakage. Medical complications were evaluated according to the involved organ system (cardiac, respiratory, neurological, gastrointestinal, renal and thromboembolic). The researchers were instructed in registration of postoperative complications according to the CD classification prior to data collection by clinical examples of complication grades. The researchers were tested by two questionnaires to assess the reproducibility [13]. The collectors were not approved before reproducibility over 90% was achieved. The first author of this study collected additional variables of interest regarding the populations of patients who had a procedure for SBO performed. This was done by review of medical records and included data regarding peri-operative blood transfusion, daily use of analgesics, previous abdominal surgery, and wound contamination. 2.3. Outcomes Outcome measures included 30-day mortality and morbidity. We defined postoperative morbidity as a patient having one or more severe complications classified as CD 3e5 [12,13]. A severe complication included any condition demanding, surgical, endoscopic, or radiological intervention (without general anaesthesia ¼ CD 3a, in general anaesthesia ¼ CD 3b), ICUmanagement (with organ failure ¼ CD 4a, with multi-organ failure ¼ CD 4b) or the death of the patient (CD 5). In case of more than one complication, the higher graded complication was used for further analysis. Mortality was defined by in-hospital death within 30 days (CD 5). 2.4. Statistical analysis Demographical data were presented with descriptive statistics. Parametric nominal data were presented as mean with standard deviations (SD), and non-parametric data as median with interquartile range (IQR). Categorical variables were reported as absolute numbers and percentages. Baseline variables with missing data were reported. Chi-square or Fischer's exact tests were used to compare categorical data or dichotomous data respectively. Non-parametric nominal data were compared by Mann-Whitney U tests. Once statistical significance was determined by univariate analyses, all factors with a P-value < 0.2 were included in two separate multivariate backward stepwise logistic regression models to identify independent risk factors for 30-day mortality and morbidity, respectively. Both models were controlled for age (10-year intervals) and gender. Outcomes were presented as odds ratios (OR) with 95% confidence limits. P 0.05 was considered statistically significant. All data analyses were performed using SPSS statistics for Windows, version 22.0 (SPSS, Inc.). 3. Results During the four-year period, 343 patients with SBO underwent an emergency laparotomy. One patient died during surgery, and 14 patients had more than one procedure for SBO in the study period. Four patients were excluded because of an incorrect surgical code. A total of 323 patients were included in the study.
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
65
3.1. Descriptive data
3.2. Preoperative comorbidities
Data regarding demographics, characteristics and comorbidities of the study population are presented in Table 1. The median age was 72 (IQR 24), and 174 (54%) were females. A total of 132 patients (41%) had a preoperative ASA score 3, and almost half of the patients were fully active with a performance score of 0, while 47 patients (14%) spent 50% or more of the day in bed (performance score 3e4).
Of the 323 patients, 217 (67%) had one or more comorbidities. The most common comorbidity was hypertension (36%), followed by ischaemic heart disease (13%), and chronic obstructive pulmonary disease (COPD) (12%). The most common use of daily medical therapy was anticoagulants (27%) and analgesics (26%). Forty-four patients (14%) had a malignancy, whereof 18 (6%) had disseminated disease. Twenty patients (6%) were in oncological treatment or had been within the past 8 weeks. 3.3. Surgical variables
Table 1 Demographics and patients' characteristics. Characteristics
n ¼ 323
Gender, female Age, yr, median (IQR) Age 60 yr. Smoking Alcohol, units per week (w/m) <7/14 7e14/14e21 >14/21 Missing BMI, mean (SD) <18.5 18.5e24.9 25e30 >30 Missing BMI 25 ASA score 1/2/3/4 ASA score 3 Performance score 0/1/2/3/4 Comorbidities Metabolic disorder Cardiovascular disorder Liver cirrhosis Chronic nephropathy Obstructive pulmonary disease Malignant disease Medication use Anticoagulants Systemic corticosteroids Statins Beta-blockers Immunomodulatoriesa Daily analgesics (total) Paracetamol NSAID Opioid Other Missing Oncological treatmentb Previous surgery Laparoscopic None 1 >1 Laparotomy None 1 2e3 >3
174 72 242 71
(54) (24) (75) (22)
239 28 22 34 24 20 138 56 21 88 77
(74) (9) (7) (11) (5) (6) (43) (18) (7) (27) (24)
Surgical characteristics are outlined in Table 2. The median length of anaesthesia and surgery were 198 min (IQR 88) and 102 min (IQR 87), respectively. Median time from decision to operate until surgery was 4 h and 21 min (IQR 6). A total of 146 patients (45%) had a bowel resection with primary anastomosis, and the surgical wound was classified as contaminated or dirty in 41 patients (13%). 3.4. Postoperative outcomes The overall 30 and 90-day mortality rates were 13% and 17%, respectively. 30-day morbidity rate (CD 3e5) was 28%, while 31% of the patients experienced minor complications (CD 1e2). The most common surgical complication was wound infection (9%), followed by wound dehiscence (6%), and postoperative ileus (6%). The most common medical complications were gastrointestinal (24%) (e.g. postoperative paralysis, infectious diarrhoea, sepsis) and pulmonary (24%), followed by cardiac complications (12%). Eighty-eight
46(14)/145(45)/104(32)/28(9) 132 (41) 156(48)/58(18)/59(18)/42(13)/5(2) 34 144 6 15 44 44
(11) (45) (2) (5) (14) (14)
88 20 54 37 13 83 28 7 45 3 5 20
(27) (6) (17) (12) (4) (26) (9) (2) (14) (1) (2) (6)
281 (87) 40 (12) 2 (1) 117 109 83 14
(36) (34) (26) (4)
Data are expressed as n (%) unless indicated otherwise. BMI, Body Mass Index; ASA, American Society of Anaesthesiologists. a Immunomodulatory therapy was defined as therapy with TNF-a inhibitors, ciclosporin, biologic IL-2 or IL-6 inhibitors, interferon beta, methotrexate or azathioprine/6-mercaptipurin. b Oncologic therapy was defined as present therapy or less than 8 weeks ago.
Table 2 Surgical characteristics. Characteristics
n ¼ 323
Surgical indication to surgery, h, median (IQR) Anaesthesia time, min, median (IQR) Surgery time, min, median (IQR) Wound classification Clean-contaminated Contaminated Dirty Missing Bowel resection Peri-operative blood transfusion Missing Medical complications No complications Grade 1 Grade 2 Grade 3a Grade 3b Grade 4a Grade 4b Grade 5 Surgical complications No complications Grade 1 Grade 2 Grade 3a Grade 3b Grade 4a Grade 4b Grade 5 30-readmission to hospital 30-day morbidity 30-day mortality 90-day mortality
4.4 (5.8) 198 (88) 102 (87)
Data are expressed as n (%) unless indicated otherwise.
279 14 27 3 146 88 17
(86) (4) (8) (1) (45) (27) (5)
159 47 58 4 8 8 7 39
(49) (15) (18) (1) (3) (3) (2) (12)
258 18 7 7 29 1 0 3 56 89 42 54
(80) (6) (2) (2) (9) (0) (0) (1) (17) (28) (13) (17)
66
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
patients (26%) had peri-operative blood transfusion, whereof 90% had a transfusion at the day of surgery or within seven days after.
3.5. Predictors of mortality and morbidity The results of the univariate risk analysis are outlined in Table 3. Several covariates were significantly associated with morbidity and mortality. Eleven covariates, besides age and gender were entered in the morbidity model (Table 4). At the end of the analysis six covariates were significantly associated with increased odds for morbidity. The highest OR for increased morbidity was seen in patients with chronic nephropathy (OR 3.9, 95% CI 1.3e15.1), and patients with a daily use of systemic corticosteroids (OR 3.5, 95% CI 1.2e10.4). In the mortality model, ten variables were entered into the analysis (Table 5). The model identified five independent risk factors associated with mortality. The highest OR for increased mortality was seen for patients with low physical performance (OR 3.4, 95% CI 1.3e8.2) and patients with a metabolic disorder (OR 3.2, 95% CI 1.2e8.5).
Table 4 Multivariable analysis of predictors of morbidity. Covariates
Odds ratio (95% confidence limits)
P
Male gender Performance score Chronic nephropathy Obstructive pulmonary disease Systemic corticosteroids Perioperative blood transfusion
1.9 2.6 3.9 2.8 3.5 3.1
0.03 0.02 0.05 0.02 0.02 <0.001
(1.1e3.5) (1.2e5.6) (1.3e15.1) (1.3e6.1) (1.2e10.4) (1.8e5.6)
Results are shown only for the variables that were significant at end of the analysis; other variables included in the multivariable analysis were age, ASA 3, cardiovascular and metabolic disease, wound contamination, previous abdominal surgery, and bowel resection.
4. Discussion This study investigated the 30-day morbidity and mortality after SBO treated with emergency laparotomy, and identified potential risk factors for adverse outcomes. A total of 28% experienced severe complications (CD 3) while 31% had minor complications (CD < 3). The mortality rate within 30 days was 13% and within 90 days 17%. Risk factors associated with the highest risk for morbidity
Table 3 Univariate analysis of clinical factors associated with morbidity and mortality. Variable
Age, yr, median (IQR) Surgical indication to surgery, h, median (IQR) Gender Female Male ASA class
No morbidity
Morbidity (CD 3)
n (%)
n (%)
Pa
b
No mortality
Mortality (CD ¼ 5)
n (%)
n (%)
Pa
71 (25) 4.2 (5.7)
75 (21) 4.5 (5.8)
0.11 0.78b
71 (25) 4.2 (5.4)
84 (15) 5.7 (6.0)
<0.001b 0.25b
136 (78) 98 (66)
38 (22) 51 (34)
0.01
155 (89) 126 (85)
19 (11) 23 (15)
0.25
154 (81) 80 (61)
37 (19) 52 (39)
<0.001
178 (93) 103 (78)
13 (7) 29 (22)
<0.001
207 (76) 25 (53)
66 (24) 22 (47)
0.001
247 (91) 31 (66)
26 (9) 16 (34)
<0.001
213 (74) 21 (62)
76 (26) 13 (38)
0.14
257 (89) 24 (71)
32 (11) 10 (29)
0.006
136 (76) 98 (68)
43 (24) 46 (32)
0.11
163 (91) 118 (82)
16 (9) 26 (18)
0.02
227 (74) 7 (47)
81 (26) 8 (53)
0.04
271 (88) 10 (67)
37 (12) 5 (33)
0.03
214 (77) 20 (46)
65 (23) 24 (55)
<0.001
252 (90) 29 (66)
27 (10) 15 (34)
<0.001
226 (75) 8 (40)
77 (25) 12 (60)
0.001
265 (88) 16 (80)
38 (13) 4 (20)
0.31
67 (67) 167 (75)
33 (33) 56 (25)
0.14
81 (81) 200 (90)
19 (19) 23 (10)
0.03
208 (75) 24 (59)
71 (25) 17 (42)
0.03
246 (88) 33 (81)
33 (12) 8 (19)
0.21
137 (77) 97 (66)
40 (23) 49 (34)
0.03
160 (90) 121 (83)
17 (10) 25 (17)
0.05
187 (80) 47 (53)
48 (20) 41 (47)
<0.001
212 (91) 68 (77)
22 (9) 20 (23)
0.002
Data are expressed as n (%) unless indicated otherwise. Variables not significant in the univariate risk analysis: smoking, excessive alcohol consumption, BMI 25, liver cirrhosis, malignant disease, oncologic treatment, analgesics, anticoagulants and time between decision to operate and surgical intervention. ASA: American Society of Anaesthesiologists. a c2 test and Fisher exact test were used as appropriate. b Mann-Whitney U test.
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68 Table 5 Multivariable analysis of predictors of mortality. Covariates
Odds ratio (95% confidence limits)
P
Age, per ten year increase Performance score Obstructive pulmonary disease Metabolic disease Perioperative blood transfusion
1.3 3.4 2.7 3.2 2.7
0.04 0.008 0.006 0.02 0.01
(1.0e1.8) (1.3e8.2) (1.1e6.6) (1.2e8.5) (1.2e5.8)
Results are shown only for the variables that were significant at end of the analysis; other variables included in the multivariable analysis were sex, ASA 3, cardiovascular and chronic nephropathy, wound contamination, previous abdominal surgery and bowel resection.
were chronic nephropathy or obstructive pulmonary disease and daily use of systemic corticosteroids, while low physical performance, metabolic disease, obstructive pulmonary disease and perioperative blood transfusion were identified as independent risk factors associated with mortality. Our results are consistent with findings obtained in previous studies [2,14]. However, we have identified risk factors that have not previously been shown as risk factors for mortality and morbidity in patients with SBO. In this study, daily use of systemic corticosteroids before surgery was significantly associated with increased risk of morbidity, but not mortality. Preoperative steroid use has, to our knowledge, not been investigated as a risk factor of adverse outcomes after SBO previously. Additionally, perioperative blood transfusion was significantly associated with both mortality and morbidity. Intraoperative blood transfusion has previously been investigated as a potential risk factor, but was not found significantly associated with neither morbidity or mortality [14]. Furthermore, chronic nephropathy and metabolic disease were independent risk factors for morbidity and mortality, respectively. These comorbidities have not previously been identified as independent risk factors, although a Norwegian study found a significant association with premorbid illness [2]. The overall mortality rate of 13% in this study, seems high compared with other studies (3e5%) [1e4,15,16]. However, most previous studies focused on patients treated for adhesive small bowel obstruction, while this study also included other aetiologies like incarcerated hernias. A study investigated death and complications after surgical treatment for SBO in 877 patients from 1961 to 1995 [2]. They found a reduction in mortality and complications in the study period, but throughout the whole study period the mortality was lower than in this study (5%). The variance in mortality might be due to differences in the study populations' rate of comorbidity and the frequency of bowel resection. In our study, 67% had one or more comorbidity, compared to 30% [2] and bowel resection was carried out in 45% of the patients, while the frequency of bowel resection was 22% [2]. Bowel strangulation and comorbidity has previous been shown significantly related to death after surgical treatment for SBO, which might explain the difference in mortality rates [17]. A significant risk factor associated with both morbidity and mortality in this study was a low preoperative performance. Dependent functional health status has previously been shown as an independent risk factor for postoperative morbidity after SBO [14], but this study is to our knowledge the first to show a significant association between mortality and low preoperative performance in this surgical population. However, differences in ranking systems of functional status makes comparison between existing studies difficult. In this study, we chose the ECOG-score, which is a widely used evaluation tool. When interpreting these results interobserver variability is an important concern. In the current study we categorized the performance score into two groups:
67
performance score 0e2 and performance score 3e4. This division has previously been shown to give rise to a higher inter-observer agreement [18], which reduces the probability of considerable variation. The beneficial effect of blood transfusions has been questioned in the recent years, which has led to the recommendation of a more restrictive use [19e23]. Perioperative blood transfusion was significantly associated with both morbidity and mortality in our study, with an OR of 3.1 (95% CI: 1.8e5.6) and 2.7 (95% CI: 1.2e5.8), respectively. It is important to emphasize the difficulty and limitations of examining the effect of blood transfusions in observational studies, because the transfused patients systematically differ from non-transfused patients. Even with the use of appropriate multivariate analysis there is still a risk of residual confounding by indication [19,24]. Despite this limitation, our results point in the direction of a more restrictive use in emergency surgery, but the association has to be examined in larger prospective randomized studies. One of the biggest strengths of the present study is the thorough classification of postoperative complications by the Clavien-Dindo system [12,13]. This system gives an opportunity to define morbidity more precisely with special emphasis on the consequences of the complication. Our study has several limitations, which should be considered when interpreting the results. First, it is a retrospective observational study, which leads to the risk of selection bias, and confounding by indication. Second, seven individual researchers collected data from medical records, which might have led to interobserver variability, although the researchers were tested before evaluation of CD scores. Third, inadequate power may explain our inability to detect significant association with some potential risk factors (e.g. bowel resection). Fourth, we only included patients with SBO managed with laparotomy, and the results may not generalize to patients managed by a laparoscopic approach. Fifth, the generalizability might be decreased as in all single-center studies. 5. Conclusion As seen in this study, laparotomy due to SBO is a high-risk procedure with considerable mortality and morbidity rates. We found that chronic nephropathy and obstructive pulmonary disease and steroid therapy were significantly associated with morbidity, while low physical performance, metabolic disease and perioperative blood transfusion were independently associated with mortality. There is an urgent need of prospective studies with interventions aimed at preventing postoperative mortality and morbidity after SBO surgery. Ethical approval The study was approved by the Danish Data Collection Agency (HEH-2013-034 I-Suite no: 02336). Source of funding The authors did not receive any external funding in connection to this study. Author contribution Maja Haunstrup Jeppesen (corresponding author): Study design, data collections, data analysis, writing and final approval. Mai-Britt Tolstrup: Data collection, data analysis, writing and final approval.
68
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
Sara Kehlet Watt: Data collection, writing and final approval. € genur: Study design, data collection, data analysis, Ismail Go writing and final approval. Conflicts of interest The authors declare no conflicts of interest. Guarantor None. Acknowledgements No funding organization did support this study. References [1] P.G. Teixeira, E. Karamanos, P. Talving, K. Inaba, L. Lam, D. Demetriades, Early operation is associated with a survival benefit for patients with adhesive bowel obstruction, Ann. Surg. 258 (2013) 459e465. [2] B.T. Fevang, J. Fevang, L. Stangeland, O. Soreide, K. Svanes, A. Viste, Complications and death after surgical treatment of small bowel obstruction: a 35year institutional experience, Ann. Surg. 231 (2000) 529e537. [3] J.E. Keenan, R.S. Turley, C.C. McCoy, J. Migaly, M.L. Shapiro, J.E. Scarborough, Trials of nonoperative management exceeding 3 days are associated with increased morbidity in patients undergoing surgery for uncomplicated adhesive small bowel obstruction, J. Trauma Acute Care Surg. 76 (2014) 1367e1372. [4] D. Schraufnagel, S. Rajaee, F.H. Millham, How many sunsets? Timing of surgery in adhesive small bowel obstruction: a study of the nationwide inpatient sample, J. Trauma Acute Care Surg. 74 (2013) 181e187 discussion 187e189. [5] G. Miller, J. Boman, I. Shrier, P.H. Gordon, Etiology of small bowel obstruction, Am. J. Surg. 180 (2000) 33e36. [6] N.R.A. Symons, K. Moorthy, A.M. Almoudaris, A. Bottle, P. Aylin, C.A. Vincent, et al., Mortality in high-risk emergency general surgical admissions, Br. J. Surg. 100 (2013) 1318e1325. [7] M.H. Al Temimi, When is death inevitable after emergency laparotomy? Analysis of the American College of Surgeons National Surgical Quality Improvement Program database, J. Am. Coll. Surg. 215 (2012) 503e511. [8] M. Smith, The importance of improving the quality of emergency surgery for a
regional quality collaborative, Ann. Surg. 257 (2013) 596e602. [9] M.-Z. Li, L. Lian, L. Xiao, W. Wu, Y. He, X. Song, Laparoscopic versus open adhesiolysis in patients with adhesive small bowel obstruction: a systematic review and meta-analysis, Am. J. Surg. 204 (2012) 779e786. [10] M.M. Oken, R.H. Creech, D.C. Tormey, J. Horton, T.E. Davis, E.T. McFadden, et al., Toxicity and response criteria of the eastern cooperative oncology group, Am. J. Clin. Oncol. 5 (1982) 649e655. [11] F. Berard, J. Gandon, Postoperative wound infections: the influence of ultraviolet irradiation of the operating room and of various other factors, Ann. Surg. 160 (1964) 1e192. [12] P.A. Clavien, J. Barkun, M.L. de Oliveira, J.N. Vauthey, D. Dindo, R.D. Schulick, et al., The Clavien-Dindo classification of surgical complications: five-year experience, Ann. Surg. 250 (2009) 187e196. [13] D. Dindo, N. Demartines, P.-A. Clavien, Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey, Ann. Surg. 240 (2004) 205e213. [14] J.A. Margenthaler, W.E. Longo, K.S. Virgo, F.E. Johnson, E.M. Grossmann, T.L. Schifftner, et al., Risk factors for adverse outcomes following surgery for small bowel obstruction, Ann. Surg. 243 (2006) 456e464. [15] N.M. Foster, M.L. McGory, D.S. Zingmond, C.Y. Ko, Small bowel obstruction: a population-based appraisal, J. Am. Coll. Surg. 203 (2006) 170e176. [16] J.J. Diaz, F. Bokhari, N.T. Mowery, J.A. Acosta, E.F.J. Block, W.J. Bromberg, et al., Guidelines for management of small bowel obstruction, J. Trauma 64 (2008) 1651e1664. [17] A.A. Deutsch, E. Eviatar, H. Gutman, R. Reiss, Small bowel obstruction: a review of 264 cases and suggestions for management, Postgrad. Med. J. 65 (1989) 463e467. [18] J.B. Sørensen, M. Klee, T. Palshof, H.H. Hansen, Performance status assessment in cancer patients. An inter-observer variability study, Br. J. Cancer 67 (1993) 773e775. [19] P.E. Marik, Efficacy of red blood cell transfusion in the critically ill: a systematic review of the literature, Crit. Care Med. 36 (2008) 2667e2674. [20] S.R. Salpeter, J.S. Buckley, S. Chatterjee, Impact of more restrictive blood transfusion strategies on clinical outcomes: a meta-analysis and systematic review, Am. J. Med. 127 (2014) 124e131. [21] J.L. Carson, P.A. Carless, P.C. Hebert, Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion, Cochrane Database Syst. Rev. 4 (2012). CD002042. [22] L.B. Holst, N. Haase, J. Wetterslev, J. Wernerman, A.B. Guttormsen, S. Karlsson, et al., Lower versus higher hemoglobin threshold for transfusion in septic shock, N. Engl. J. Med. 371 (2014) 1381e1391. [23] N.H. Roubinian, G.J. Escobar, V. Liu, B.E. Swain, M.N. Gardner, P. Kipnis, et al., Trends in red blood cell transfusion and 30-day mortality among hospitalized patients, Transfusion 54 (2014) 2678e2686. [24] R.A. Zollo, Blood transfusion in the perioperative period, Best Pract. Res. Clin. Anaesthesiol. 26 (2012) 475e484.
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Original research
Risk factors affecting morbidity and mortality following emergency laparotomy for small bowel obstruction: A retrospective cohort study € genur b Maja Haunstrup Jeppesen a, b, *, Mai-Britt Tolstrup a, Sara Kehlet Watt b, Ismail Go a b
Department of Gastroenterology, Surgical Section, Copenhagen University Hospital Herlev, Herlev Ringvej 75, 2730 Herlev, Denmark Department of Surgery, Zealand University Hospital Køge, Lykkebæksvej 1, 4600 Koege, Denmark
h i g h l i g h t s 30-day morbidity and mortality rates were 28% and 13%, respectively. Six independent risk factors for morbidity were identified. Five independent risk factors for mortality were identified. Identified risk factors included chronic nephropathy and low physical performance.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 November 2015 Accepted 6 February 2016 Available online 18 February 2016
Introductions: To identify risk factors for increased 30-day morbidity and mortality, using standardized measuring tools for the characterization of complications after emergency surgery for small bowel obstruction. Methods: A retrospective cohort study including patients treated with emergency laparotomy for small bowel obstruction at a Copenhagen University Hospital (2009e2013). Complications were evaluated according to the Clavien-Dindo classification. Results: A total of 323 patients were included. The overall 30-day morbidity and mortality rates were 28% and 13%, respectively. Six covariates were identified as independent risk factors associated with morbidity by multiple logistic regression analysis. The highest odds for morbidity were seen in patients with chronic nephropathy (Odds Ratio [OR] ¼ 3.9; 95% CI 1.3e15.1), and in patients with a daily use of steroids (OR ¼ 3.5; 95% CI 1.2e10.4). Five independent risk factors were associated with increased odds for mortality. Patients with low physical performance (OR ¼ 3.4; 95% CI 1.3e8.2) or metabolic disorders (OR ¼ 3.2; 95% CI 1.2e8.5) had the highest risk of mortality. Conclusions: Morbidity and mortality rates were high in this study compared with other studies. Several comorbid conditions were associated with morbidity and mortality. These results may aid the acute care surgeon in identifying patients with a high-risk for postoperative complications and fatal outcomes. © 2016 IJS Publishing Group Limited. Published by Elsevier Ltd. All rights reserved.
Keywords: Emergency laparotomy Small bowel obstruction Outcomes Mortality Morbidity
1. Introduction Small bowel obstruction (SBO) is one of the most frequent indications for emergency laparotomy surgery, and is known as a high-risk procedure with morbidity and mortality rates at 20e30% and 3e5%, respectively [1e4]. Adhesions are the most frequent
* Corresponding author. Department of Surgery, Zealand University Hospital Køge, Lykkebæksvej 1, 4600 Koege, Denmark. E-mail addresses: [email protected] (M.H. Jeppesen), m.tolstrup@dadlnet. dk (M.-B. Tolstrup), [email protected] (S. Kehlet Watt), igo@regionsjaelland. €genur). dk (I. Go
reason for SBO, accounting for almost 75% of all cases, followed by malignancies, inflammatory bowel disease and hernias [5]. In recent years more interest has been taken in describing postoperative outcomes after abdominal emergencies, and in searching for preoperative factors that predispose poor outcome [6e8]. This information may aid the acute care surgeon in identifying patients with a high-risk for postoperative complications and fatal outcomes and eventually initiate interventions to prevent these adverse outcomes. Previous studies have identified potential risk factors such as surgical delay, increased age, high ASA score (American Society of Anaesthesiologists), septic shock, obesity, and comorbidities as factors associated with morbidity and mortality
http://dx.doi.org/10.1016/j.ijsu.2016.02.059 1743-9191/© 2016 IJS Publishing Group Limited. Published by Elsevier Ltd. All rights reserved.
64
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
[1e4,9]. There is a need for large studies that characterize these patients with focus on the consequences of their complications in a standardized method. The aim of this study was to use standardized measures for the characterization of complications after emergency surgery for SBO, and to investigate potential risk factors for increased 30-day morbidity and mortality. 2. Methods 2.1. Study design and patients This study is a retrospective cohort study conducted at a tertiary university hospital situated in the greater Copenhagen area, serving a population of 432.000 citizens. All patients are covered by public health insurance and pay no visit fees. Patients were included if any gastrointestinal emergency surgical procedure, laparoscopy or laparotomy, was performed from May 2009 to April 2013. The study was approved by the Danish Data Collection Agency (HEH-2013034 I-Suite no: 02336). 2.2. Variables A group of seven researchers collected pre- and postoperative data by review of medical records. This was done as part of creating a standardized database containing perioperative information of all emergency abdominal procedures performed at the hospital. Data for this study population consisted of all patients diagnosed with SBO, and treated with emergency laparotomy. These patients were identified and extracted from the database by the NOMESCO Classification of Surgical Procedure (NCSP) codes KJFB00 (small bowel resection), KJFK00 (division of adhesive band in intestinal obstruction), and KJFK10 (freeing of adhesions in intestinal obstruction). Exclusion criterias were death during surgery and patients aged <18 yr. Patients who had any other procedure simultaneously with the SBO (e.g. ileostomy) were excluded. If a patient had more than one incidence of SBO managed surgically during the study period, only data from the index procedure was analysed for complications. Intraoperative data was collected from the Danish Anaesthesia Database and from the surgery scheduling software. Pre-existing comorbidities and conditions were registered from the admission report, and were considered as existing if the condition was medically treated or if previous treatment was described. The comorbidities were defined as follows: chronic nephropathy; elevated se-creatinine or dialysis treatment, obstructive pulmonary disease; chronic obstructive pulmonary disease (COPD) or asthma, malignant disease; actual malignant disease (disseminated and not disseminated), metabolic disorder; diabetes or thyroid disease, cardiovascular disease; hypertension, atrial fibrillation, ischaemic heart disease or any history of cerebrovascular disease. Performance score was evaluated by the Eastern Cooperative Oncology Group (EC0G) score [10], which is a six step scale ranging from 0 to 5 (fully active to death). Excessive alcohol consumption was defined as a weekly alcohol intake of >7 units per week for women and >14 units per week for men. The time from the decision to operate was made, and surgical intervention was initiated, was registered. Perioperative blood transfusion was defined as being transfused with blood products between 7 days before and 30 days after surgery. Surgical contamination was evaluated by the wound classification described by Berard et al. [11]. Readmission related to the surgical procedure was also registered. Postoperative complications were registered according to the Clavien-Dindo (CD) classification [12,13]. The CD classification is a validated ranking system, which classifies postoperative complications based on the therapy used to treat the complication. Any
complication is graded on a 5-step scale. CD score 1 is any deviation from the normal postoperative course, without the need for pharmacological treatment or surgical, endoscopic, and radiological interventions. CD score 5 is death of the patient [13]. Complications were divided into two categories: surgical and medical. Surgical complications include bleeding, wound dehiscence, postoperative ileus, wound infection, intraabdominal abscesses and anastomosis leakage. Medical complications were evaluated according to the involved organ system (cardiac, respiratory, neurological, gastrointestinal, renal and thromboembolic). The researchers were instructed in registration of postoperative complications according to the CD classification prior to data collection by clinical examples of complication grades. The researchers were tested by two questionnaires to assess the reproducibility [13]. The collectors were not approved before reproducibility over 90% was achieved. The first author of this study collected additional variables of interest regarding the populations of patients who had a procedure for SBO performed. This was done by review of medical records and included data regarding peri-operative blood transfusion, daily use of analgesics, previous abdominal surgery, and wound contamination. 2.3. Outcomes Outcome measures included 30-day mortality and morbidity. We defined postoperative morbidity as a patient having one or more severe complications classified as CD 3e5 [12,13]. A severe complication included any condition demanding, surgical, endoscopic, or radiological intervention (without general anaesthesia ¼ CD 3a, in general anaesthesia ¼ CD 3b), ICUmanagement (with organ failure ¼ CD 4a, with multi-organ failure ¼ CD 4b) or the death of the patient (CD 5). In case of more than one complication, the higher graded complication was used for further analysis. Mortality was defined by in-hospital death within 30 days (CD 5). 2.4. Statistical analysis Demographical data were presented with descriptive statistics. Parametric nominal data were presented as mean with standard deviations (SD), and non-parametric data as median with interquartile range (IQR). Categorical variables were reported as absolute numbers and percentages. Baseline variables with missing data were reported. Chi-square or Fischer's exact tests were used to compare categorical data or dichotomous data respectively. Non-parametric nominal data were compared by Mann-Whitney U tests. Once statistical significance was determined by univariate analyses, all factors with a P-value < 0.2 were included in two separate multivariate backward stepwise logistic regression models to identify independent risk factors for 30-day mortality and morbidity, respectively. Both models were controlled for age (10-year intervals) and gender. Outcomes were presented as odds ratios (OR) with 95% confidence limits. P 0.05 was considered statistically significant. All data analyses were performed using SPSS statistics for Windows, version 22.0 (SPSS, Inc.). 3. Results During the four-year period, 343 patients with SBO underwent an emergency laparotomy. One patient died during surgery, and 14 patients had more than one procedure for SBO in the study period. Four patients were excluded because of an incorrect surgical code. A total of 323 patients were included in the study.
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
65
3.1. Descriptive data
3.2. Preoperative comorbidities
Data regarding demographics, characteristics and comorbidities of the study population are presented in Table 1. The median age was 72 (IQR 24), and 174 (54%) were females. A total of 132 patients (41%) had a preoperative ASA score 3, and almost half of the patients were fully active with a performance score of 0, while 47 patients (14%) spent 50% or more of the day in bed (performance score 3e4).
Of the 323 patients, 217 (67%) had one or more comorbidities. The most common comorbidity was hypertension (36%), followed by ischaemic heart disease (13%), and chronic obstructive pulmonary disease (COPD) (12%). The most common use of daily medical therapy was anticoagulants (27%) and analgesics (26%). Forty-four patients (14%) had a malignancy, whereof 18 (6%) had disseminated disease. Twenty patients (6%) were in oncological treatment or had been within the past 8 weeks. 3.3. Surgical variables
Table 1 Demographics and patients' characteristics. Characteristics
n ¼ 323
Gender, female Age, yr, median (IQR) Age 60 yr. Smoking Alcohol, units per week (w/m) <7/14 7e14/14e21 >14/21 Missing BMI, mean (SD) <18.5 18.5e24.9 25e30 >30 Missing BMI 25 ASA score 1/2/3/4 ASA score 3 Performance score 0/1/2/3/4 Comorbidities Metabolic disorder Cardiovascular disorder Liver cirrhosis Chronic nephropathy Obstructive pulmonary disease Malignant disease Medication use Anticoagulants Systemic corticosteroids Statins Beta-blockers Immunomodulatoriesa Daily analgesics (total) Paracetamol NSAID Opioid Other Missing Oncological treatmentb Previous surgery Laparoscopic None 1 >1 Laparotomy None 1 2e3 >3
174 72 242 71
(54) (24) (75) (22)
239 28 22 34 24 20 138 56 21 88 77
(74) (9) (7) (11) (5) (6) (43) (18) (7) (27) (24)
Surgical characteristics are outlined in Table 2. The median length of anaesthesia and surgery were 198 min (IQR 88) and 102 min (IQR 87), respectively. Median time from decision to operate until surgery was 4 h and 21 min (IQR 6). A total of 146 patients (45%) had a bowel resection with primary anastomosis, and the surgical wound was classified as contaminated or dirty in 41 patients (13%). 3.4. Postoperative outcomes The overall 30 and 90-day mortality rates were 13% and 17%, respectively. 30-day morbidity rate (CD 3e5) was 28%, while 31% of the patients experienced minor complications (CD 1e2). The most common surgical complication was wound infection (9%), followed by wound dehiscence (6%), and postoperative ileus (6%). The most common medical complications were gastrointestinal (24%) (e.g. postoperative paralysis, infectious diarrhoea, sepsis) and pulmonary (24%), followed by cardiac complications (12%). Eighty-eight
46(14)/145(45)/104(32)/28(9) 132 (41) 156(48)/58(18)/59(18)/42(13)/5(2) 34 144 6 15 44 44
(11) (45) (2) (5) (14) (14)
88 20 54 37 13 83 28 7 45 3 5 20
(27) (6) (17) (12) (4) (26) (9) (2) (14) (1) (2) (6)
281 (87) 40 (12) 2 (1) 117 109 83 14
(36) (34) (26) (4)
Data are expressed as n (%) unless indicated otherwise. BMI, Body Mass Index; ASA, American Society of Anaesthesiologists. a Immunomodulatory therapy was defined as therapy with TNF-a inhibitors, ciclosporin, biologic IL-2 or IL-6 inhibitors, interferon beta, methotrexate or azathioprine/6-mercaptipurin. b Oncologic therapy was defined as present therapy or less than 8 weeks ago.
Table 2 Surgical characteristics. Characteristics
n ¼ 323
Surgical indication to surgery, h, median (IQR) Anaesthesia time, min, median (IQR) Surgery time, min, median (IQR) Wound classification Clean-contaminated Contaminated Dirty Missing Bowel resection Peri-operative blood transfusion Missing Medical complications No complications Grade 1 Grade 2 Grade 3a Grade 3b Grade 4a Grade 4b Grade 5 Surgical complications No complications Grade 1 Grade 2 Grade 3a Grade 3b Grade 4a Grade 4b Grade 5 30-readmission to hospital 30-day morbidity 30-day mortality 90-day mortality
4.4 (5.8) 198 (88) 102 (87)
Data are expressed as n (%) unless indicated otherwise.
279 14 27 3 146 88 17
(86) (4) (8) (1) (45) (27) (5)
159 47 58 4 8 8 7 39
(49) (15) (18) (1) (3) (3) (2) (12)
258 18 7 7 29 1 0 3 56 89 42 54
(80) (6) (2) (2) (9) (0) (0) (1) (17) (28) (13) (17)
66
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
patients (26%) had peri-operative blood transfusion, whereof 90% had a transfusion at the day of surgery or within seven days after.
3.5. Predictors of mortality and morbidity The results of the univariate risk analysis are outlined in Table 3. Several covariates were significantly associated with morbidity and mortality. Eleven covariates, besides age and gender were entered in the morbidity model (Table 4). At the end of the analysis six covariates were significantly associated with increased odds for morbidity. The highest OR for increased morbidity was seen in patients with chronic nephropathy (OR 3.9, 95% CI 1.3e15.1), and patients with a daily use of systemic corticosteroids (OR 3.5, 95% CI 1.2e10.4). In the mortality model, ten variables were entered into the analysis (Table 5). The model identified five independent risk factors associated with mortality. The highest OR for increased mortality was seen for patients with low physical performance (OR 3.4, 95% CI 1.3e8.2) and patients with a metabolic disorder (OR 3.2, 95% CI 1.2e8.5).
Table 4 Multivariable analysis of predictors of morbidity. Covariates
Odds ratio (95% confidence limits)
P
Male gender Performance score Chronic nephropathy Obstructive pulmonary disease Systemic corticosteroids Perioperative blood transfusion
1.9 2.6 3.9 2.8 3.5 3.1
0.03 0.02 0.05 0.02 0.02 <0.001
(1.1e3.5) (1.2e5.6) (1.3e15.1) (1.3e6.1) (1.2e10.4) (1.8e5.6)
Results are shown only for the variables that were significant at end of the analysis; other variables included in the multivariable analysis were age, ASA 3, cardiovascular and metabolic disease, wound contamination, previous abdominal surgery, and bowel resection.
4. Discussion This study investigated the 30-day morbidity and mortality after SBO treated with emergency laparotomy, and identified potential risk factors for adverse outcomes. A total of 28% experienced severe complications (CD 3) while 31% had minor complications (CD < 3). The mortality rate within 30 days was 13% and within 90 days 17%. Risk factors associated with the highest risk for morbidity
Table 3 Univariate analysis of clinical factors associated with morbidity and mortality. Variable
Age, yr, median (IQR) Surgical indication to surgery, h, median (IQR) Gender Female Male ASA class
No morbidity
Morbidity (CD 3)
n (%)
n (%)
Pa
b
No mortality
Mortality (CD ¼ 5)
n (%)
n (%)
Pa
71 (25) 4.2 (5.7)
75 (21) 4.5 (5.8)
0.11 0.78b
71 (25) 4.2 (5.4)
84 (15) 5.7 (6.0)
<0.001b 0.25b
136 (78) 98 (66)
38 (22) 51 (34)
0.01
155 (89) 126 (85)
19 (11) 23 (15)
0.25
154 (81) 80 (61)
37 (19) 52 (39)
<0.001
178 (93) 103 (78)
13 (7) 29 (22)
<0.001
207 (76) 25 (53)
66 (24) 22 (47)
0.001
247 (91) 31 (66)
26 (9) 16 (34)
<0.001
213 (74) 21 (62)
76 (26) 13 (38)
0.14
257 (89) 24 (71)
32 (11) 10 (29)
0.006
136 (76) 98 (68)
43 (24) 46 (32)
0.11
163 (91) 118 (82)
16 (9) 26 (18)
0.02
227 (74) 7 (47)
81 (26) 8 (53)
0.04
271 (88) 10 (67)
37 (12) 5 (33)
0.03
214 (77) 20 (46)
65 (23) 24 (55)
<0.001
252 (90) 29 (66)
27 (10) 15 (34)
<0.001
226 (75) 8 (40)
77 (25) 12 (60)
0.001
265 (88) 16 (80)
38 (13) 4 (20)
0.31
67 (67) 167 (75)
33 (33) 56 (25)
0.14
81 (81) 200 (90)
19 (19) 23 (10)
0.03
208 (75) 24 (59)
71 (25) 17 (42)
0.03
246 (88) 33 (81)
33 (12) 8 (19)
0.21
137 (77) 97 (66)
40 (23) 49 (34)
0.03
160 (90) 121 (83)
17 (10) 25 (17)
0.05
187 (80) 47 (53)
48 (20) 41 (47)
<0.001
212 (91) 68 (77)
22 (9) 20 (23)
0.002
Data are expressed as n (%) unless indicated otherwise. Variables not significant in the univariate risk analysis: smoking, excessive alcohol consumption, BMI 25, liver cirrhosis, malignant disease, oncologic treatment, analgesics, anticoagulants and time between decision to operate and surgical intervention. ASA: American Society of Anaesthesiologists. a c2 test and Fisher exact test were used as appropriate. b Mann-Whitney U test.
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68 Table 5 Multivariable analysis of predictors of mortality. Covariates
Odds ratio (95% confidence limits)
P
Age, per ten year increase Performance score Obstructive pulmonary disease Metabolic disease Perioperative blood transfusion
1.3 3.4 2.7 3.2 2.7
0.04 0.008 0.006 0.02 0.01
(1.0e1.8) (1.3e8.2) (1.1e6.6) (1.2e8.5) (1.2e5.8)
Results are shown only for the variables that were significant at end of the analysis; other variables included in the multivariable analysis were sex, ASA 3, cardiovascular and chronic nephropathy, wound contamination, previous abdominal surgery and bowel resection.
were chronic nephropathy or obstructive pulmonary disease and daily use of systemic corticosteroids, while low physical performance, metabolic disease, obstructive pulmonary disease and perioperative blood transfusion were identified as independent risk factors associated with mortality. Our results are consistent with findings obtained in previous studies [2,14]. However, we have identified risk factors that have not previously been shown as risk factors for mortality and morbidity in patients with SBO. In this study, daily use of systemic corticosteroids before surgery was significantly associated with increased risk of morbidity, but not mortality. Preoperative steroid use has, to our knowledge, not been investigated as a risk factor of adverse outcomes after SBO previously. Additionally, perioperative blood transfusion was significantly associated with both mortality and morbidity. Intraoperative blood transfusion has previously been investigated as a potential risk factor, but was not found significantly associated with neither morbidity or mortality [14]. Furthermore, chronic nephropathy and metabolic disease were independent risk factors for morbidity and mortality, respectively. These comorbidities have not previously been identified as independent risk factors, although a Norwegian study found a significant association with premorbid illness [2]. The overall mortality rate of 13% in this study, seems high compared with other studies (3e5%) [1e4,15,16]. However, most previous studies focused on patients treated for adhesive small bowel obstruction, while this study also included other aetiologies like incarcerated hernias. A study investigated death and complications after surgical treatment for SBO in 877 patients from 1961 to 1995 [2]. They found a reduction in mortality and complications in the study period, but throughout the whole study period the mortality was lower than in this study (5%). The variance in mortality might be due to differences in the study populations' rate of comorbidity and the frequency of bowel resection. In our study, 67% had one or more comorbidity, compared to 30% [2] and bowel resection was carried out in 45% of the patients, while the frequency of bowel resection was 22% [2]. Bowel strangulation and comorbidity has previous been shown significantly related to death after surgical treatment for SBO, which might explain the difference in mortality rates [17]. A significant risk factor associated with both morbidity and mortality in this study was a low preoperative performance. Dependent functional health status has previously been shown as an independent risk factor for postoperative morbidity after SBO [14], but this study is to our knowledge the first to show a significant association between mortality and low preoperative performance in this surgical population. However, differences in ranking systems of functional status makes comparison between existing studies difficult. In this study, we chose the ECOG-score, which is a widely used evaluation tool. When interpreting these results interobserver variability is an important concern. In the current study we categorized the performance score into two groups:
67
performance score 0e2 and performance score 3e4. This division has previously been shown to give rise to a higher inter-observer agreement [18], which reduces the probability of considerable variation. The beneficial effect of blood transfusions has been questioned in the recent years, which has led to the recommendation of a more restrictive use [19e23]. Perioperative blood transfusion was significantly associated with both morbidity and mortality in our study, with an OR of 3.1 (95% CI: 1.8e5.6) and 2.7 (95% CI: 1.2e5.8), respectively. It is important to emphasize the difficulty and limitations of examining the effect of blood transfusions in observational studies, because the transfused patients systematically differ from non-transfused patients. Even with the use of appropriate multivariate analysis there is still a risk of residual confounding by indication [19,24]. Despite this limitation, our results point in the direction of a more restrictive use in emergency surgery, but the association has to be examined in larger prospective randomized studies. One of the biggest strengths of the present study is the thorough classification of postoperative complications by the Clavien-Dindo system [12,13]. This system gives an opportunity to define morbidity more precisely with special emphasis on the consequences of the complication. Our study has several limitations, which should be considered when interpreting the results. First, it is a retrospective observational study, which leads to the risk of selection bias, and confounding by indication. Second, seven individual researchers collected data from medical records, which might have led to interobserver variability, although the researchers were tested before evaluation of CD scores. Third, inadequate power may explain our inability to detect significant association with some potential risk factors (e.g. bowel resection). Fourth, we only included patients with SBO managed with laparotomy, and the results may not generalize to patients managed by a laparoscopic approach. Fifth, the generalizability might be decreased as in all single-center studies. 5. Conclusion As seen in this study, laparotomy due to SBO is a high-risk procedure with considerable mortality and morbidity rates. We found that chronic nephropathy and obstructive pulmonary disease and steroid therapy were significantly associated with morbidity, while low physical performance, metabolic disease and perioperative blood transfusion were independently associated with mortality. There is an urgent need of prospective studies with interventions aimed at preventing postoperative mortality and morbidity after SBO surgery. Ethical approval The study was approved by the Danish Data Collection Agency (HEH-2013-034 I-Suite no: 02336). Source of funding The authors did not receive any external funding in connection to this study. Author contribution Maja Haunstrup Jeppesen (corresponding author): Study design, data collections, data analysis, writing and final approval. Mai-Britt Tolstrup: Data collection, data analysis, writing and final approval.
68
M.H. Jeppesen et al. / International Journal of Surgery 28 (2016) 63e68
Sara Kehlet Watt: Data collection, writing and final approval. € genur: Study design, data collection, data analysis, Ismail Go writing and final approval. Conflicts of interest The authors declare no conflicts of interest. Guarantor None. Acknowledgements No funding organization did support this study. References [1] P.G. Teixeira, E. Karamanos, P. Talving, K. Inaba, L. Lam, D. Demetriades, Early operation is associated with a survival benefit for patients with adhesive bowel obstruction, Ann. Surg. 258 (2013) 459e465. [2] B.T. Fevang, J. Fevang, L. Stangeland, O. Soreide, K. Svanes, A. Viste, Complications and death after surgical treatment of small bowel obstruction: a 35year institutional experience, Ann. Surg. 231 (2000) 529e537. [3] J.E. Keenan, R.S. Turley, C.C. McCoy, J. Migaly, M.L. Shapiro, J.E. Scarborough, Trials of nonoperative management exceeding 3 days are associated with increased morbidity in patients undergoing surgery for uncomplicated adhesive small bowel obstruction, J. Trauma Acute Care Surg. 76 (2014) 1367e1372. [4] D. Schraufnagel, S. Rajaee, F.H. Millham, How many sunsets? Timing of surgery in adhesive small bowel obstruction: a study of the nationwide inpatient sample, J. Trauma Acute Care Surg. 74 (2013) 181e187 discussion 187e189. [5] G. Miller, J. Boman, I. Shrier, P.H. Gordon, Etiology of small bowel obstruction, Am. J. Surg. 180 (2000) 33e36. [6] N.R.A. Symons, K. Moorthy, A.M. Almoudaris, A. Bottle, P. Aylin, C.A. Vincent, et al., Mortality in high-risk emergency general surgical admissions, Br. J. Surg. 100 (2013) 1318e1325. [7] M.H. Al Temimi, When is death inevitable after emergency laparotomy? Analysis of the American College of Surgeons National Surgical Quality Improvement Program database, J. Am. Coll. Surg. 215 (2012) 503e511. [8] M. Smith, The importance of improving the quality of emergency surgery for a
regional quality collaborative, Ann. Surg. 257 (2013) 596e602. [9] M.-Z. Li, L. Lian, L. Xiao, W. Wu, Y. He, X. Song, Laparoscopic versus open adhesiolysis in patients with adhesive small bowel obstruction: a systematic review and meta-analysis, Am. J. Surg. 204 (2012) 779e786. [10] M.M. Oken, R.H. Creech, D.C. Tormey, J. Horton, T.E. Davis, E.T. McFadden, et al., Toxicity and response criteria of the eastern cooperative oncology group, Am. J. Clin. Oncol. 5 (1982) 649e655. [11] F. Berard, J. Gandon, Postoperative wound infections: the influence of ultraviolet irradiation of the operating room and of various other factors, Ann. Surg. 160 (1964) 1e192. [12] P.A. Clavien, J. Barkun, M.L. de Oliveira, J.N. Vauthey, D. Dindo, R.D. Schulick, et al., The Clavien-Dindo classification of surgical complications: five-year experience, Ann. Surg. 250 (2009) 187e196. [13] D. Dindo, N. Demartines, P.-A. Clavien, Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey, Ann. Surg. 240 (2004) 205e213. [14] J.A. Margenthaler, W.E. Longo, K.S. Virgo, F.E. Johnson, E.M. Grossmann, T.L. Schifftner, et al., Risk factors for adverse outcomes following surgery for small bowel obstruction, Ann. Surg. 243 (2006) 456e464. [15] N.M. Foster, M.L. McGory, D.S. Zingmond, C.Y. Ko, Small bowel obstruction: a population-based appraisal, J. Am. Coll. Surg. 203 (2006) 170e176. [16] J.J. Diaz, F. Bokhari, N.T. Mowery, J.A. Acosta, E.F.J. Block, W.J. Bromberg, et al., Guidelines for management of small bowel obstruction, J. Trauma 64 (2008) 1651e1664. [17] A.A. Deutsch, E. Eviatar, H. Gutman, R. Reiss, Small bowel obstruction: a review of 264 cases and suggestions for management, Postgrad. Med. J. 65 (1989) 463e467. [18] J.B. Sørensen, M. Klee, T. Palshof, H.H. Hansen, Performance status assessment in cancer patients. An inter-observer variability study, Br. J. Cancer 67 (1993) 773e775. [19] P.E. Marik, Efficacy of red blood cell transfusion in the critically ill: a systematic review of the literature, Crit. Care Med. 36 (2008) 2667e2674. [20] S.R. Salpeter, J.S. Buckley, S. Chatterjee, Impact of more restrictive blood transfusion strategies on clinical outcomes: a meta-analysis and systematic review, Am. J. Med. 127 (2014) 124e131. [21] J.L. Carson, P.A. Carless, P.C. Hebert, Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion, Cochrane Database Syst. Rev. 4 (2012). CD002042. [22] L.B. Holst, N. Haase, J. Wetterslev, J. Wernerman, A.B. Guttormsen, S. Karlsson, et al., Lower versus higher hemoglobin threshold for transfusion in septic shock, N. Engl. J. Med. 371 (2014) 1381e1391. [23] N.H. Roubinian, G.J. Escobar, V. Liu, B.E. Swain, M.N. Gardner, P. Kipnis, et al., Trends in red blood cell transfusion and 30-day mortality among hospitalized patients, Transfusion 54 (2014) 2678e2686. [24] R.A. Zollo, Blood transfusion in the perioperative period, Best Pract. Res. Clin. Anaesthesiol. 26 (2012) 475e484.