- Email: [email protected]
Factors affecting morbidity in emergency general surgery
The American Journal of Surgery (2011) 201, 456 – 462
Society of Black Academic Surgeons
Factors affecting morbidity in emergency general surgery Felix Akinbami, M.B.B.S., M.S., Reza Askari, M.D., Jill Steinberg, R.N., M.P.H., Maria Panizales, R.N., M.S.N., Selwyn O. Rogers Jr, M.D., M.P.H.* Department of Surgery and Center for Surgery and Public Health, Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115, USA KEYWORDS: Postoperative complications; General surgery; Emergency; Outcomes
Abstract BACKGROUND: Emergency status adversely affects surgical outcomes. Predictors of increased morbidity of emergency general surgery are unknown. We determined predictors of postoperative complications of emergency general surgery. METHODS: We conducted a retrospective study of Brigham and Women’s Hospital American College of Surgeons National Surgical Quality Improvement Program patients who had an emergency general surgery procedure from January 1, 2007, to December 31, 2009. Additional non–American College of Surgeons National Surgical Quality Improvement Program variables were collected. Our primary outcome was postoperative complications within 30 days. RESULTS: Of 819 cases, 24.7% had 1 or more complications, with 8.9% mortality within 30 days. Common complications were respiratory (47%) and wound occurrences (18%). Age, sex, blood glucose level, creatinine level, albumin level, surgery duration, and smoking were independent predictors of morbidity. CONCLUSIONS: Emergency general surgery patients with postoperative complications are likely to be older, male, smokers, have increased blood glucose and creatinine levels, lower albumin levels, and longer surgical times. Fluid resuscitation and experienced surgical teams are putative targets to improve outcomes. © 2011 Elsevier Inc. All rights reserved.
Emergency surgery is defined as nonelective surgery that is performed with the aim to prevent morbid or fatal health consequences of a surgically treatable condition. According to the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP), emergency surgery is defined as a case performed within 12 hours of presentation to the hospital or after onset of related preoperative symptomatology and/or cases defined by the surgeon or anesthesiologist as emergent.1 These procedures are per* Corresponding author. Tel.: ⫹1-617-732-8042; fax: ⫹1-617-5826047. E-mail address: [email protected] Manuscript received September 3, 2010; revised manuscript November 9, 2010
0002-9610/$ - see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2010.11.007
formed with the purpose to save a patient’s life or prevent a patient’s health deterioration. Emergency surgeries represent a significant proportion of surgeries and cut across the various surgical specialties. Over the past decade, there has been a significant focus on reducing postoperative complications. The Surgical Care Improvement Program supported by the Center for Medicare and Medicaid Services set a goal of reducing surgical complications by 25% by 2010.2 The NSQIP has been associated with a 45% reduction in morbidity and a 27% reduction in mortality at Veterans Affairs hospitals.3 There was also a reduction of 11% in morbidity and 17% in mortality seen in 118 ACS NSQIP–participating hospitals from 2006 to 2007.4 Despite the focus on quality improvement in surgery, studies continue to show that emergency status contributes
F. Akinbami et al.
Morbidity in emergency general surgery
significantly to morbidity and mortality of general surgeries.5,6 Complications resulting from emergency surgeries lead to worsened clinical status, emotional and financial burden to patients and families, and additional hospital costs.7 However, the reasons why emergency status confers a negative outcome remains poorly elucidated. The primary aim of this study was to determine the factors that contribute to morbidity, associated with emergency status in general surgery. We also aim to further identify those risk factors that are potentially modifiable, which may lead to targets to reduce surgical complications of emergency general surgery.
Methods Patient population We performed a retrospective observational study of 819 major emergency general surgery patients at the Brigham and Women’s Hospital who were enrolled in the ACS NSQIP over a 3-year period (January 1, 2007, to December 31, 2009) with prior approval from the Human Subjects Office at our institution. Randomization and eligibility criteria were per ACS NSQIP guidelines.1,8 –15 Trauma cases and those patients who had a prior surgery within 30 days were excluded. Additional variables considered potential predictors were 18 non–ACS NSQIP preoperative variables, which were collected from chart review. Some of these included blood glucose levels closest to the time of surgical incision, length of time to surgery from emergency department arrival, contact with the emergency department attending physician to time of incision, length of time from obtaining a computed tomography scan or abdominal ultrasound to time of incision, and time of antibiotic or deep venous thrombosis prophylaxis administration before incision. The outcome of interest was occurrence of any of the predefined ACS NSQIP postoperative occurrences within 30 days, which included surgical site infections, sepsis, respiratory (unplanned intubation, pulmonary embolism, pneumonia, on ventilator ⬎48 h), urinary (acute renal failure, urinary tract infection), neurologic (cerebrovascular accident), and cardiac (myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation) occurrences.
Statistical analysis Univariate analysis was performed using the Wilcoxon signed rank sum test for continuous variables and the chisquare test or the Fisher exact test, where appropriate, for categoric variables. Multivariate analysis using logistic regression was performed by calculating odds ratios and 95% confidence intervals (CIs) to identify significant independent predictors of morbidity (P ⬍ .05). C-statistic and Hosmer–Lemeshow goodness-of-fit measurements were calculated to determine the fit of the model.
457 Table 1
Baseline demographics of patient population
Characteristics
Patients, n (%)
Male sex Race Asian Black White Other Unknown Smokers Dyspnea at rest Diabetes mellitus Worsening functional health status Blood glucose level, mg/dL ⬍98 98–115 116–141 ⬎142 Missing Duration of surgery, min ⬍50 50–76 77–119 ⱖ120 Missing Transfer to hospital Admitted directly from home Admitted from acute care hospital or nursing home Surgical procedures Appendectomy Colectomy Enterectomy Hernia repair Cholecystectomy Others
344 (42) 17 91 576 134 1 144 46 47 99
(2) (11) (70) (16) (⬍.1) (18) (6) (6) (12)
193 181 180 182 83
(26.2) (24.6) (24.5) (24.7)
210 200 204 204 1
(25.7) (24.5) (24.9) (24.9)
689 (84) 130 (16) 216 112 88 37 55 311
(26) (14) (11) (4) (7) (38)
Length of surgery (in minutes) and blood glucose levels (in mg/dL) were categorized into 4 levels based on quartiles for simplification and to reduce the effects of outliers.
Results There were 819 cases that met eligibility criteria. Demographic variables showed that 42% were male, and 70% were white (Table 1). Postoperative complications were identified in 202 (24.7%) of the cases within 30 days of the procedure; 30-day mortality was recorded in 73 (8.9%) cases. A total of 144 patients (18%) were current smokers. Five percent of the patients had severe chronic obstructive pulmonary disease (COPD), whereas 6% had dyspnea at rest. The functional status of 99 patients (12%) deteriorated before surgery. A total of 130 patients (16%) were transferred from a hospital or nursing home. The most common complication was respiratory in nature (Table 2). This included unplanned intubation, on ventilator longer than 48 hours, pneumonia, and pulmonary embolism. Infectious complications, including sepsis and
458
The American Journal of Surgery, Vol 201, No 4, April 2011
Table 2 Frequency of postoperative complications, irrespective of number of occurrences per patient Complications Respiratory (unplanned intubation, PE, PNA, vent ⬎48 h) Postoperative SSI (superficial, deep, organ space) Urinary (ARF, UTI) Cardiac (MI, cardiac arrest) Sepsis, septic shock Bleeding, DVT Neurologic
n, (%) 168 (47) 63 23 12 65 24 1
(18) (6) (3) (18) (7) (⬍ 1)
Some cases had more than 1 complication. Number of complications ⫽ 358. PE ⫽ pulmonary embolism; PNA ⫽ pneumonia; Vent ⫽ ventilator support; SSI ⫽ surgical site infection; ARF ⫽ acute renal failure; UTI ⫽ urinary tract infection; MI ⫽ myocardial infarction; DVT ⫽ deep vein thrombosis.
wound occurrences, were noted in 18% of the cases. Urinary complications (including urinary tract infections and acute renal failure) as well as bleeding complications were responsible for a combined 13% of complications. In this population, cardiac and neurologic occurrences represented a minority of occurrences. In univariate analyses, 104 (51.49%) of those identified with complications were male, whereas 61.1% of females did not have a complication (Table 3). For patients who developed complications, the mean age was 61 ⫾ 18 years compared with 50 ⫾ 18 years for those without complications. There was a longer mean length of stay for patients who developed complications compared with those who did not suffer complications (19 ⫾ 14 days vs 7 ⫾ 10 days) (Table 4). Other patient-related variables that were found to be significant for morbidity include a history of cancer, altered sensorium before surgery, current smoking status, steroid use, and presence of ascites (Table 3).
Laboratory tests Patients who developed complications had higher mean laboratory values for blood urea nitrogen (BUN) (30 ⫾ 23 vs 17 ⫾ 16), mean blood glucose (144 ⫾ 63 mg/dL vs 122 ⫾ 47 mg/dL), serum creatinine (1.56 ⫾ 1.6 vs .97 ⫾ .8), white blood cells (14 ⫾ 10 vs 12 ⫾ 6 ⫻ 1,000/mm3), and international normalized ratio (1.42 ⫾ 1 vs 1.2 ⫾ .5) than patients who did not develop complications (Table 4). Prothrombin time (PT) and Partial thromboplastin time (PTT) values also were found to be higher in patients with complications. Serum albumin level (3.3 ⫾ .8 g/dL vs 3.9 ⫾ 1 g/dL) was lower in patients with complications (Table 5).
Surgical variables The most common procedures were appendectomies, colectomies, and enterectomies (Table 1). American Society
of Anesthesiologists (ASA) class of III or higher was seen in 72% of cases with complications. The mean duration of surgery was found to be higher among those with complications (125 ⫾ 122 min vs 87 ⫾ 99 min). The time-related variables were excluded from analysis because of a significant amount of missing data.
Multivariate analysis Sex, age, smoking, and increased creatinine levels were found to be independent predictors of morbidity. Females were less likely to develop complications when compared with males (odds ratio [OR], .669; 95% CI, .445–.996), whereas older patients (OR, 1.026; 95% CI, 1.014 –1.038), smokers (OR, 1.784; 95% CI, 1.086 –2.929), and patients who had higher creatinine levels (OR, 1.300; 95% CI, 1.082–1.562) were more likely to develop complications. Those with higher albumin levels had a lower risk of developing complications (OR, .541; 95% CI, .419 –.699). Patients with blood glucose levels between 98 and 115 mg/dL were less likely to develop complications compared with those with blood glucose levels greater than 142 (OR, .505; 95% CI, .282–.906). Surgical length of time less than 76 minutes compared with surgical times of longer than 140 minutes also were less likely to develop complications. The C-statistic was .791 for the final model.
Comments In a random sample of emergent general surgeries at a tertiary care medical center, we found that the morbidity and mortality of emergency general surgery is significant. In our cohort, one quarter of patients had adverse postoperative occurrences and almost 9% of them died within 30 days of surgery. The most common postoperative complications were respiratory occurrences, which included prolonged ventilator support greater than 48 hours and unplanned intubations. Wound occurrences, including superficial incisional, deep incisional, and organ space infections, were the next most common. In univariate analyses, age, sex, length of surgery, increased blood glucose levels, increased BUN level, deteriorating functional health status, ASA class of 3 or higher, presence of COPD, and history of smoking were found to be significant predictors of postoperative complications. Multivariate analyses found male sex, increased blood glucose levels, longer surgical times, increased creatinine levels, lower serum albumin levels, and current smoking to be independent predictors of morbidity. Our study corroborates previous studies by identifying the importance of emergency status and its effects on adverse outcomes. Postoperative complications increase hospital costs.7 Furthermore, patients who suffer complications have been shown to have a higher risk of mortality.16 The rates of complications and mortality seen in our study are comparable with those seen in previous work that examined
F. Akinbami et al. Table 3
Morbidity in emergency general surgery
459
Univariate analysis for categoric variables
Variable
Category
Sex
Female Male Unknown Asian Black Other White Admitted directly from home Acute care hospital inpatient Others Clean Clean/contaminated Contaminated Dirty/infected No disease Mild disease Severe disease Life-threatening Moribund No Yes ⬍1 h to incision ⬎1 h to incision Unknown Same Unknown Worse No sepsis SIRS Sepsis Septic shock Yes No Yes No Yes No Yes No No dyspnea Dyspnea on moderate exertion Dyspnea at rest Yes No Yes No Unknown Yes No Yes No
Race
Transfer
Wound class
ASA class
DVT prophylaxis Antibiotic administration Functional health status Sepsis
Cancer DNR status Diabetes mellitus Ascites Dyspnea
Steroid use COPD Smoker Altered sensorium
Patients without complications, n (%)
Patients with complications, n (%)
377 240 1 12 71 102 431 542
(61.10) (38.90) (.16) (1.94) (11.51) (16.53) (69.85) (87.84)
98 (48.51) 104 (51.49)
62 (10.05)
41 (20.30)
P value .002 .877
5 20 32 145 147
(2.48) (9.90) (15.84) (71.78) (72.77)
13 98 209 150 160 116 273 184 39 5 152 465 302 178 137 567 2 48 50 413 22 132 38 579 6 611 33 563 42 575 554 37
(2.11) (15.88) (33.87) (24.31) (25.93) (18.80) (44.25) (29.82) (6.32) (.81) (24.64) (75.36) (48.95) (28.85) (22.20) (91.90) (.32) (7.78) (8.10) (66.94) (3.57) (21.39) (6.16) (93.84) (.97) (99.03) (5.54) (94.46) (6.81) (93.19) (89.79) (6.00)
14 20 68 54 60 6 50 96 48 2 53 149 79 72 51 150 1 51 23 103 28 48 23 179 3 199 14 166 37 165 165 17
(6.93) (9.90) (33.66) (26.73) (29.70) (2.97) (24.75) (47.52) (23.76) (.99) (26.24) (73.76) (39.11) (35.64) (25.25) (74.26) (.50) (25.25) (11.39) (50.99) (13.86) (23.76) (11.39) (88.61) (1.49) (98.51) (7.78) (92.22) (18.32) (81.68) (81.68) (8.42)
26 35 582 21 596 1 98 518 18 599
(4.21) (5.67) (94.33) (3.40) (96.60) (.16) (15.88) (83.95) (2.92) (97.08)
20 30 172 24 178
(9.90) (14.85) (85.15) (11.88) (88.12)
46 156 20 182
(22.77) (77.23) (9.90) (90.10)
<.001
.176
<.001
.648 .048 <.001 <.001
.014 .544 .2693 <.001 .004
<.001 <.001 .026 <.001
Level of significance, P ⬍ .05. DVT ⫽ deep vein thrombosis; DNR ⫽ do not resuscitate; SIRS ⫽ systemic inflammatory response syndrome.
outcomes of patients after colectomies.5,16 As in most acute care hospitals, appendectomies, colectomies, enterectomies, and hernia repairs were the most common surgeries.6 As
seen in our study, many of the predictors of postoperative morbidity are not modifiable. Males, older patients, and smokers have been shown to have a higher rate of compli-
460
The American Journal of Surgery, Vol 201, No 4, April 2011
Table 4
Univariate analysis for continuous variables
Variable
Complication, no/yes
Number of cases
Mean ⫾ standard deviation
Age, y
No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes
615 202 548 188 601 199 600 200 544 187 613 197 544 194 545 195 546 195 567 181 616 202
50.38 61.34 122.45 143.67 16.99 30.32 .97 1.56 3.91 3.30 12.08 14.19 32.74 35.26 15.15 16.73 1.20 1.42 6.96 18.96 86.89 122.73
Blood glucose level, mg/dL BUN level Serum creatinine level, mg/dL Albumin level, g/dL White blood cell count, ⫻1000/mm3 PTT, s PT, s INR Length of stay, d Duration of surgery
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
17.97 17.80 47.46 62.55 15.39 23.10 .77 1.60 .97 .80 6.39 9.53 8.64 10.95 3.33 4.13 .48 1.03 9.53 14.42 99.08 105.5
P value <.001 <.001 <.001 <.001 <.001 .056 <.001 <.001 <.001 <.001 <.001
Level of significance, P ⬍ .05. PT ⫽ Prothrombin time; PTT ⫽ Partial thromboplastin time; INR ⫽ international normalized ratio.
cations.6 As patients age, they are more likely to have comorbidities that increase their postoperative risk of complications. Similar to several prior studies, we found an association between longer surgical times and an increased risk of postoperative complications.17,18 Longer surgical times may just be a proxy for complexity of the surgery itself or surgeon-related factors such as surgeon experience. Confirming prior work by our group, we also found that increased blood sugar levels before surgery increased the risk of postoperative complications.19 –21 Hyperglycemia is known to have a deleterious effect on wound healing and has been associated with complications after vascular and cardiac surgery. It is, however, noteworthy that diabetes mellitus was not found to be significant in our study. Diabetes was identified in only 6% of our target population. Also, it can be assumed that diabetic patients are more likely to be aware of their glycemic status and actively controlling it than nondiabetic patients or those who are undiagnosed. Diabetic patients might be quick to acknowledge their status at the emergency room and hence are given necessary medical attention before surgery. The effect of glycosylated hemoglobin (HbA1c) values on developing postoperative complications might help to further enlighten us. Other studies have identified comorbidities, higher ASA classification, and a history of COPD as risk factors.5 Our study had some notable limitations that were related to the study design. Our study was a retrospective observational study. The sample size might have been inadequate to fully evaluate all the putative predictors of postoperative complications. With a larger sample size, we could have
stratified the analyses by type of surgery. Because this was a retrospective study, we encountered a significant amount of missing data regarding time intervals, thus nullifying the possibility of identifying these variables as predictors. Because of the inability to capture the time data, we excluded those patients from the analyses. Implementation of electronic documentation with date and time stamps would provide an opportunity for accurate data extraction and quality improvement.
Table 5 Multivariate analysis of independent predictors of complications after emergency general surgery Variable
OR
95% CIs
Female Age Current smokers Duration of surgery I vs IV Duration of surgery II vs IV Serum creatinine Blood glucose II vs IV Serum albumin
.669 1.026 1.784 .237 .476 1.300 .505 .541
.449–.996 1.014–1.038 1.086–2.929 .127–.442 .273–.828 1.082–1.562 .282–.906 .419–.699
Because of the large number of factors examined, only significant factors were listed in Table 4. Blood glucose levels and duration of surgery were categorized into quartiles because of outliers in the data, as follows: blood glucose I, ⬍98 mg/dL; blood glucose II, 98 –115 mg/dL; blood glucose III, 116 –141 mg/dL; and blood glucose IV, ⱖ142 mg/dL. Duration of surgery I, 0 – 49 min; duration of surgery II, 50 –76 min; duration of surgery III, 77–119 min; and surgery IV, ⱖ 120 min.
F. Akinbami et al.
Morbidity in emergency general surgery
The results from this study may not be generalizable to other hospitals unless they have a similar mix of general surgical cases. One of the issues that arises with respect to patients who develop complications relates to our ability to rescue them from the complications. “Failure to rescue”22,23 may be a significant cause of additional morbidity and mortality after emergent surgical procedures. Failure to rescue refers to the use of interventions such as intensive care unit (ICU) admission or specific processes of care that lead to improved outcomes after complications.22,23 Another issue is the site of recovery for patients after emergent general surgical procedures. In many hospitals, emergent general surgical patients are not necessarily admitted to the surgical ICU. This is in contrast to coronary artery bypass patients who all are admitted to an ICU. Given the high rate of complications and attendant mortality, should all emergency cases receive ICU care? This would be a controversial recommendation because it may further increase hospital length of stay and health care costs but potentially may reduce complication rates. Another recommendation that may be worth testing to improve surgical outcomes relates to the association of increased BUN level with morbidity after emergency surgery. Given the increased BUN levels that we found in this study, it is possible that patients were somewhat hypovolemic and may have benefited from additional intravenous fluid resuscitation. This approach should be studied prospectively. As observed in this study, a low serum albumin level may be a marker of malnutrition. This marker would be difficult to treat in the short term. Albumin supplementation has been used in some series but its effect remains controversial.24,25 Because the baseline values for serum albumin and creatinine are not known in these patients, it is unclear if these recommendations would improve outcomes because the increased values could be a reflection of the patient’s severity of illness. Control of perioperative hyperglycemia is a simple and actionable process that may have benefit to reduce infectious complications.19 –21 Finally, composition of the surgical team that is caring for the emergent general surgical patient may be modified. Recent support for acute care surgeons who address both trauma and emergent general surgery is an avenue that needs to be explored more rigorously. Such acute care surgeons may be able to potentially decrease the length of surgery and delays to surgery through their expertise.26 Britt et al27 developed an acute care surgery model to decrease time delays to surgery by using daytime operating rooms and the experience of surgeons to improve medical decision making. Earley et al26 examined the effects of this model on outcomes in patients undergoing appendectomies and found a decrease in time to surgery, complication rates, and length of stay. Further study is needed to determine if modifying these factors improves outcomes. A larger multicenter study will be required to evaluate time variables such as delay in surgical timing as well as to enhance the generalizability of our findings to other hospitals.
461
Conclusions Emergency general surgery patients who develop postoperative complications are more likely to be older, male, smokers, and have increased perioperative blood glucose and creatinine levels and lower serum albumin levels. Those patients with complications also incur longer surgical times. Preoperative intravenous fluid administration to adequately resuscitate patients, tighter glucose control, and experienced surgical teams to decrease surgical times are putative targets to improve outcomes in emergency general surgery patients.
Acknowledgments The authors thank Stuart Lipsitz, PhD, Xiangmei Gu for statistical analysis.
References 1. American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP). Chapter 4, ACSNSQIP Variables and Definitions. Available at: https://acsnsqip.org/. 2. Bratzler DW. The surgical infection prevention and surgical care improvement projects: promises and pitfalls. Am Surg 2006;72(11): 1010 – 6. 3. Khuri SF, Daley J, Henderson WG. The comparative assessment and improvement of quality of surgical care in the Department of Veterans Affairs. Arch Surg 2002;137:20 –7. 4. Hall BL, Hamilton BH, Richards K, et al. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: an evaluation of all participating hospitals. Ann Surg 2009;250:363–76. 5. Costa G, Tomassini F, Tierno SM, et al. Emergency colonic surgery: analysis of risk factors predicting morbidity and mortality. Chir Ital 2009;61:565–71. 6. Neumayer L, Hosokawa P, El-Tamer M, et al. Multivariable predictors of postoperative surgical site infection after general and vascular surgery: results from the patient safety in surgery study. J Am Coll Surg 2007;204:1178 – 87. 7. Dimick JB, Chen SL, Taheri PA, et al. Hospital costs associated with surgical complications: a report from the private-sector National Surgical Quality Improvement Program. J Am Coll Surg 2004;199:531–7. 8. Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs’ NSQIP: the first national, validated, outcome-based, riskadjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Ann Surg 1998;228:491–507. 9. Khuri SF, Daley J, Henderson W, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:315–27. 10. Daley J, Khuri SF, Henderson W, et al. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:328 – 40. 11. Khuri SF, Henderson WG, Daley J, et al. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007;204:1089 –102.
462 12. Henderson WG, Khuri SF, Mosca C, et al. Comparison of riskadjusted 30-day postoperative mortality and morbidity in Department of Veterans Affairs hospitals and selected university medical centers: general surgical operations in men. J Am Coll Surg 2007; 204:1103–14. 13. Fink AS, Hutter MM, Campbell DC Jr, et al. Comparison of riskadjusted 30-day postoperative mortality and morbidity in Department of Veterans Affairs hospitals and selected university medical centers: general surgical operations in women. J Am Coll Surg 2007;204: 1127–36. 14. Fink AS, Campbell DA Jr, Mentzer RM Jr, et al. The National Surgical Quality Improvement Program in non-veterans administration hospitals: initial demonstration of feasibility. Ann Surg 2002;236:344 –54. 15. Daley J, Forbes MG, Young GJ, et al. Validating risk-adjusted surgical outcomes: site visit assessment of process and structure. National VA Surgical Risk Study. J Am Coll Surg 1997;185:341–51. 16. McGillicuddy EA, Schuster KM, Davis KA, et al. Factors predicting morbidity and mortality in emergency colorectal procedures in elderly patients. Arch Surg 2009;144:1162. 17. Haridas M, Malangoni MA. Predictive factors for surgical site infection in general surgery. Surgery 2008;144:496 –501. 18. Kasatpibal N, Nørgaard M, Sørensen HT, et al. Risk of surgical site infection and efficacy of antibiotic prophylaxis: a cohort study of appendectomy patients in Thailand. BMC Infect Dis 2006;6:111. 19. Ramos M, Khalpey Z, Lipsitz S, et al. Relationship of perioperative hyperglycemia and postoperative infections in patients who undergo general and vascular surgery. Ann Surg 2008;248:585–91.
The American Journal of Surgery, Vol 201, No 4, April 2011 20. Jones KW, Cain AS, Mitchell JH, et al. Hyperglycemia predicts mortality after CABG: postoperative hyperglycemia predicts dramatic increases in mortality after coronary artery bypass graft surgery. J Diabetes Complications 2008;22:365–70. 21. Lazar HL, Chipkin SR, Fitzgerald CA, et al. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation 2004; 109:1497–502. 22. Silber JH, Williams SV, Krakauer H, et al. Hospital and patient characteristics associated with death after surgery. A study of adverse occurrence and failure to rescue. Med Care 1992;30:615–29. 23. Simpson KR. Failure to rescue: implications for evaluating quality of care during labor and birth. J Perinat Neonat Nurs 2005;19:24 –34. 24. Finfer S, Bellomo R, McEvoy S, et al. Effect of baseline serum albumin concentration on outcome of resuscitation with albumin or saline in patients in intensive care units: analysis of data from the saline versus albumin fluid evaluation (SAFE) study. BMJ 2006; 333:1044. 25. Vincent JL. Relevance of albumin in modern critical care medicine. Best Pract Res Clin Anaesthesiol 2009;23:183–91. 26. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery model improves outcomes in patients with appendicitis. Ann Surg 2006;244: 498 –504. 27. Britt RC, Weireter LJ, Britt LD. Initial implementation of an acute care surgery model: implications for timeliness of care. J Am Coll Surg 2009;209:421– 4.
Society of Black Academic Surgeons
Factors affecting morbidity in emergency general surgery Felix Akinbami, M.B.B.S., M.S., Reza Askari, M.D., Jill Steinberg, R.N., M.P.H., Maria Panizales, R.N., M.S.N., Selwyn O. Rogers Jr, M.D., M.P.H.* Department of Surgery and Center for Surgery and Public Health, Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115, USA KEYWORDS: Postoperative complications; General surgery; Emergency; Outcomes
Abstract BACKGROUND: Emergency status adversely affects surgical outcomes. Predictors of increased morbidity of emergency general surgery are unknown. We determined predictors of postoperative complications of emergency general surgery. METHODS: We conducted a retrospective study of Brigham and Women’s Hospital American College of Surgeons National Surgical Quality Improvement Program patients who had an emergency general surgery procedure from January 1, 2007, to December 31, 2009. Additional non–American College of Surgeons National Surgical Quality Improvement Program variables were collected. Our primary outcome was postoperative complications within 30 days. RESULTS: Of 819 cases, 24.7% had 1 or more complications, with 8.9% mortality within 30 days. Common complications were respiratory (47%) and wound occurrences (18%). Age, sex, blood glucose level, creatinine level, albumin level, surgery duration, and smoking were independent predictors of morbidity. CONCLUSIONS: Emergency general surgery patients with postoperative complications are likely to be older, male, smokers, have increased blood glucose and creatinine levels, lower albumin levels, and longer surgical times. Fluid resuscitation and experienced surgical teams are putative targets to improve outcomes. © 2011 Elsevier Inc. All rights reserved.
Emergency surgery is defined as nonelective surgery that is performed with the aim to prevent morbid or fatal health consequences of a surgically treatable condition. According to the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP), emergency surgery is defined as a case performed within 12 hours of presentation to the hospital or after onset of related preoperative symptomatology and/or cases defined by the surgeon or anesthesiologist as emergent.1 These procedures are per* Corresponding author. Tel.: ⫹1-617-732-8042; fax: ⫹1-617-5826047. E-mail address: [email protected] Manuscript received September 3, 2010; revised manuscript November 9, 2010
0002-9610/$ - see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2010.11.007
formed with the purpose to save a patient’s life or prevent a patient’s health deterioration. Emergency surgeries represent a significant proportion of surgeries and cut across the various surgical specialties. Over the past decade, there has been a significant focus on reducing postoperative complications. The Surgical Care Improvement Program supported by the Center for Medicare and Medicaid Services set a goal of reducing surgical complications by 25% by 2010.2 The NSQIP has been associated with a 45% reduction in morbidity and a 27% reduction in mortality at Veterans Affairs hospitals.3 There was also a reduction of 11% in morbidity and 17% in mortality seen in 118 ACS NSQIP–participating hospitals from 2006 to 2007.4 Despite the focus on quality improvement in surgery, studies continue to show that emergency status contributes
F. Akinbami et al.
Morbidity in emergency general surgery
significantly to morbidity and mortality of general surgeries.5,6 Complications resulting from emergency surgeries lead to worsened clinical status, emotional and financial burden to patients and families, and additional hospital costs.7 However, the reasons why emergency status confers a negative outcome remains poorly elucidated. The primary aim of this study was to determine the factors that contribute to morbidity, associated with emergency status in general surgery. We also aim to further identify those risk factors that are potentially modifiable, which may lead to targets to reduce surgical complications of emergency general surgery.
Methods Patient population We performed a retrospective observational study of 819 major emergency general surgery patients at the Brigham and Women’s Hospital who were enrolled in the ACS NSQIP over a 3-year period (January 1, 2007, to December 31, 2009) with prior approval from the Human Subjects Office at our institution. Randomization and eligibility criteria were per ACS NSQIP guidelines.1,8 –15 Trauma cases and those patients who had a prior surgery within 30 days were excluded. Additional variables considered potential predictors were 18 non–ACS NSQIP preoperative variables, which were collected from chart review. Some of these included blood glucose levels closest to the time of surgical incision, length of time to surgery from emergency department arrival, contact with the emergency department attending physician to time of incision, length of time from obtaining a computed tomography scan or abdominal ultrasound to time of incision, and time of antibiotic or deep venous thrombosis prophylaxis administration before incision. The outcome of interest was occurrence of any of the predefined ACS NSQIP postoperative occurrences within 30 days, which included surgical site infections, sepsis, respiratory (unplanned intubation, pulmonary embolism, pneumonia, on ventilator ⬎48 h), urinary (acute renal failure, urinary tract infection), neurologic (cerebrovascular accident), and cardiac (myocardial infarction, cardiac arrest requiring cardiopulmonary resuscitation) occurrences.
Statistical analysis Univariate analysis was performed using the Wilcoxon signed rank sum test for continuous variables and the chisquare test or the Fisher exact test, where appropriate, for categoric variables. Multivariate analysis using logistic regression was performed by calculating odds ratios and 95% confidence intervals (CIs) to identify significant independent predictors of morbidity (P ⬍ .05). C-statistic and Hosmer–Lemeshow goodness-of-fit measurements were calculated to determine the fit of the model.
457 Table 1
Baseline demographics of patient population
Characteristics
Patients, n (%)
Male sex Race Asian Black White Other Unknown Smokers Dyspnea at rest Diabetes mellitus Worsening functional health status Blood glucose level, mg/dL ⬍98 98–115 116–141 ⬎142 Missing Duration of surgery, min ⬍50 50–76 77–119 ⱖ120 Missing Transfer to hospital Admitted directly from home Admitted from acute care hospital or nursing home Surgical procedures Appendectomy Colectomy Enterectomy Hernia repair Cholecystectomy Others
344 (42) 17 91 576 134 1 144 46 47 99
(2) (11) (70) (16) (⬍.1) (18) (6) (6) (12)
193 181 180 182 83
(26.2) (24.6) (24.5) (24.7)
210 200 204 204 1
(25.7) (24.5) (24.9) (24.9)
689 (84) 130 (16) 216 112 88 37 55 311
(26) (14) (11) (4) (7) (38)
Length of surgery (in minutes) and blood glucose levels (in mg/dL) were categorized into 4 levels based on quartiles for simplification and to reduce the effects of outliers.
Results There were 819 cases that met eligibility criteria. Demographic variables showed that 42% were male, and 70% were white (Table 1). Postoperative complications were identified in 202 (24.7%) of the cases within 30 days of the procedure; 30-day mortality was recorded in 73 (8.9%) cases. A total of 144 patients (18%) were current smokers. Five percent of the patients had severe chronic obstructive pulmonary disease (COPD), whereas 6% had dyspnea at rest. The functional status of 99 patients (12%) deteriorated before surgery. A total of 130 patients (16%) were transferred from a hospital or nursing home. The most common complication was respiratory in nature (Table 2). This included unplanned intubation, on ventilator longer than 48 hours, pneumonia, and pulmonary embolism. Infectious complications, including sepsis and
458
The American Journal of Surgery, Vol 201, No 4, April 2011
Table 2 Frequency of postoperative complications, irrespective of number of occurrences per patient Complications Respiratory (unplanned intubation, PE, PNA, vent ⬎48 h) Postoperative SSI (superficial, deep, organ space) Urinary (ARF, UTI) Cardiac (MI, cardiac arrest) Sepsis, septic shock Bleeding, DVT Neurologic
n, (%) 168 (47) 63 23 12 65 24 1
(18) (6) (3) (18) (7) (⬍ 1)
Some cases had more than 1 complication. Number of complications ⫽ 358. PE ⫽ pulmonary embolism; PNA ⫽ pneumonia; Vent ⫽ ventilator support; SSI ⫽ surgical site infection; ARF ⫽ acute renal failure; UTI ⫽ urinary tract infection; MI ⫽ myocardial infarction; DVT ⫽ deep vein thrombosis.
wound occurrences, were noted in 18% of the cases. Urinary complications (including urinary tract infections and acute renal failure) as well as bleeding complications were responsible for a combined 13% of complications. In this population, cardiac and neurologic occurrences represented a minority of occurrences. In univariate analyses, 104 (51.49%) of those identified with complications were male, whereas 61.1% of females did not have a complication (Table 3). For patients who developed complications, the mean age was 61 ⫾ 18 years compared with 50 ⫾ 18 years for those without complications. There was a longer mean length of stay for patients who developed complications compared with those who did not suffer complications (19 ⫾ 14 days vs 7 ⫾ 10 days) (Table 4). Other patient-related variables that were found to be significant for morbidity include a history of cancer, altered sensorium before surgery, current smoking status, steroid use, and presence of ascites (Table 3).
Laboratory tests Patients who developed complications had higher mean laboratory values for blood urea nitrogen (BUN) (30 ⫾ 23 vs 17 ⫾ 16), mean blood glucose (144 ⫾ 63 mg/dL vs 122 ⫾ 47 mg/dL), serum creatinine (1.56 ⫾ 1.6 vs .97 ⫾ .8), white blood cells (14 ⫾ 10 vs 12 ⫾ 6 ⫻ 1,000/mm3), and international normalized ratio (1.42 ⫾ 1 vs 1.2 ⫾ .5) than patients who did not develop complications (Table 4). Prothrombin time (PT) and Partial thromboplastin time (PTT) values also were found to be higher in patients with complications. Serum albumin level (3.3 ⫾ .8 g/dL vs 3.9 ⫾ 1 g/dL) was lower in patients with complications (Table 5).
Surgical variables The most common procedures were appendectomies, colectomies, and enterectomies (Table 1). American Society
of Anesthesiologists (ASA) class of III or higher was seen in 72% of cases with complications. The mean duration of surgery was found to be higher among those with complications (125 ⫾ 122 min vs 87 ⫾ 99 min). The time-related variables were excluded from analysis because of a significant amount of missing data.
Multivariate analysis Sex, age, smoking, and increased creatinine levels were found to be independent predictors of morbidity. Females were less likely to develop complications when compared with males (odds ratio [OR], .669; 95% CI, .445–.996), whereas older patients (OR, 1.026; 95% CI, 1.014 –1.038), smokers (OR, 1.784; 95% CI, 1.086 –2.929), and patients who had higher creatinine levels (OR, 1.300; 95% CI, 1.082–1.562) were more likely to develop complications. Those with higher albumin levels had a lower risk of developing complications (OR, .541; 95% CI, .419 –.699). Patients with blood glucose levels between 98 and 115 mg/dL were less likely to develop complications compared with those with blood glucose levels greater than 142 (OR, .505; 95% CI, .282–.906). Surgical length of time less than 76 minutes compared with surgical times of longer than 140 minutes also were less likely to develop complications. The C-statistic was .791 for the final model.
Comments In a random sample of emergent general surgeries at a tertiary care medical center, we found that the morbidity and mortality of emergency general surgery is significant. In our cohort, one quarter of patients had adverse postoperative occurrences and almost 9% of them died within 30 days of surgery. The most common postoperative complications were respiratory occurrences, which included prolonged ventilator support greater than 48 hours and unplanned intubations. Wound occurrences, including superficial incisional, deep incisional, and organ space infections, were the next most common. In univariate analyses, age, sex, length of surgery, increased blood glucose levels, increased BUN level, deteriorating functional health status, ASA class of 3 or higher, presence of COPD, and history of smoking were found to be significant predictors of postoperative complications. Multivariate analyses found male sex, increased blood glucose levels, longer surgical times, increased creatinine levels, lower serum albumin levels, and current smoking to be independent predictors of morbidity. Our study corroborates previous studies by identifying the importance of emergency status and its effects on adverse outcomes. Postoperative complications increase hospital costs.7 Furthermore, patients who suffer complications have been shown to have a higher risk of mortality.16 The rates of complications and mortality seen in our study are comparable with those seen in previous work that examined
F. Akinbami et al. Table 3
Morbidity in emergency general surgery
459
Univariate analysis for categoric variables
Variable
Category
Sex
Female Male Unknown Asian Black Other White Admitted directly from home Acute care hospital inpatient Others Clean Clean/contaminated Contaminated Dirty/infected No disease Mild disease Severe disease Life-threatening Moribund No Yes ⬍1 h to incision ⬎1 h to incision Unknown Same Unknown Worse No sepsis SIRS Sepsis Septic shock Yes No Yes No Yes No Yes No No dyspnea Dyspnea on moderate exertion Dyspnea at rest Yes No Yes No Unknown Yes No Yes No
Race
Transfer
Wound class
ASA class
DVT prophylaxis Antibiotic administration Functional health status Sepsis
Cancer DNR status Diabetes mellitus Ascites Dyspnea
Steroid use COPD Smoker Altered sensorium
Patients without complications, n (%)
Patients with complications, n (%)
377 240 1 12 71 102 431 542
(61.10) (38.90) (.16) (1.94) (11.51) (16.53) (69.85) (87.84)
98 (48.51) 104 (51.49)
62 (10.05)
41 (20.30)
P value .002 .877
5 20 32 145 147
(2.48) (9.90) (15.84) (71.78) (72.77)
13 98 209 150 160 116 273 184 39 5 152 465 302 178 137 567 2 48 50 413 22 132 38 579 6 611 33 563 42 575 554 37
(2.11) (15.88) (33.87) (24.31) (25.93) (18.80) (44.25) (29.82) (6.32) (.81) (24.64) (75.36) (48.95) (28.85) (22.20) (91.90) (.32) (7.78) (8.10) (66.94) (3.57) (21.39) (6.16) (93.84) (.97) (99.03) (5.54) (94.46) (6.81) (93.19) (89.79) (6.00)
14 20 68 54 60 6 50 96 48 2 53 149 79 72 51 150 1 51 23 103 28 48 23 179 3 199 14 166 37 165 165 17
(6.93) (9.90) (33.66) (26.73) (29.70) (2.97) (24.75) (47.52) (23.76) (.99) (26.24) (73.76) (39.11) (35.64) (25.25) (74.26) (.50) (25.25) (11.39) (50.99) (13.86) (23.76) (11.39) (88.61) (1.49) (98.51) (7.78) (92.22) (18.32) (81.68) (81.68) (8.42)
26 35 582 21 596 1 98 518 18 599
(4.21) (5.67) (94.33) (3.40) (96.60) (.16) (15.88) (83.95) (2.92) (97.08)
20 30 172 24 178
(9.90) (14.85) (85.15) (11.88) (88.12)
46 156 20 182
(22.77) (77.23) (9.90) (90.10)
<.001
.176
<.001
.648 .048 <.001 <.001
.014 .544 .2693 <.001 .004
<.001 <.001 .026 <.001
Level of significance, P ⬍ .05. DVT ⫽ deep vein thrombosis; DNR ⫽ do not resuscitate; SIRS ⫽ systemic inflammatory response syndrome.
outcomes of patients after colectomies.5,16 As in most acute care hospitals, appendectomies, colectomies, enterectomies, and hernia repairs were the most common surgeries.6 As
seen in our study, many of the predictors of postoperative morbidity are not modifiable. Males, older patients, and smokers have been shown to have a higher rate of compli-
460
The American Journal of Surgery, Vol 201, No 4, April 2011
Table 4
Univariate analysis for continuous variables
Variable
Complication, no/yes
Number of cases
Mean ⫾ standard deviation
Age, y
No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes No Yes
615 202 548 188 601 199 600 200 544 187 613 197 544 194 545 195 546 195 567 181 616 202
50.38 61.34 122.45 143.67 16.99 30.32 .97 1.56 3.91 3.30 12.08 14.19 32.74 35.26 15.15 16.73 1.20 1.42 6.96 18.96 86.89 122.73
Blood glucose level, mg/dL BUN level Serum creatinine level, mg/dL Albumin level, g/dL White blood cell count, ⫻1000/mm3 PTT, s PT, s INR Length of stay, d Duration of surgery
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
17.97 17.80 47.46 62.55 15.39 23.10 .77 1.60 .97 .80 6.39 9.53 8.64 10.95 3.33 4.13 .48 1.03 9.53 14.42 99.08 105.5
P value <.001 <.001 <.001 <.001 <.001 .056 <.001 <.001 <.001 <.001 <.001
Level of significance, P ⬍ .05. PT ⫽ Prothrombin time; PTT ⫽ Partial thromboplastin time; INR ⫽ international normalized ratio.
cations.6 As patients age, they are more likely to have comorbidities that increase their postoperative risk of complications. Similar to several prior studies, we found an association between longer surgical times and an increased risk of postoperative complications.17,18 Longer surgical times may just be a proxy for complexity of the surgery itself or surgeon-related factors such as surgeon experience. Confirming prior work by our group, we also found that increased blood sugar levels before surgery increased the risk of postoperative complications.19 –21 Hyperglycemia is known to have a deleterious effect on wound healing and has been associated with complications after vascular and cardiac surgery. It is, however, noteworthy that diabetes mellitus was not found to be significant in our study. Diabetes was identified in only 6% of our target population. Also, it can be assumed that diabetic patients are more likely to be aware of their glycemic status and actively controlling it than nondiabetic patients or those who are undiagnosed. Diabetic patients might be quick to acknowledge their status at the emergency room and hence are given necessary medical attention before surgery. The effect of glycosylated hemoglobin (HbA1c) values on developing postoperative complications might help to further enlighten us. Other studies have identified comorbidities, higher ASA classification, and a history of COPD as risk factors.5 Our study had some notable limitations that were related to the study design. Our study was a retrospective observational study. The sample size might have been inadequate to fully evaluate all the putative predictors of postoperative complications. With a larger sample size, we could have
stratified the analyses by type of surgery. Because this was a retrospective study, we encountered a significant amount of missing data regarding time intervals, thus nullifying the possibility of identifying these variables as predictors. Because of the inability to capture the time data, we excluded those patients from the analyses. Implementation of electronic documentation with date and time stamps would provide an opportunity for accurate data extraction and quality improvement.
Table 5 Multivariate analysis of independent predictors of complications after emergency general surgery Variable
OR
95% CIs
Female Age Current smokers Duration of surgery I vs IV Duration of surgery II vs IV Serum creatinine Blood glucose II vs IV Serum albumin
.669 1.026 1.784 .237 .476 1.300 .505 .541
.449–.996 1.014–1.038 1.086–2.929 .127–.442 .273–.828 1.082–1.562 .282–.906 .419–.699
Because of the large number of factors examined, only significant factors were listed in Table 4. Blood glucose levels and duration of surgery were categorized into quartiles because of outliers in the data, as follows: blood glucose I, ⬍98 mg/dL; blood glucose II, 98 –115 mg/dL; blood glucose III, 116 –141 mg/dL; and blood glucose IV, ⱖ142 mg/dL. Duration of surgery I, 0 – 49 min; duration of surgery II, 50 –76 min; duration of surgery III, 77–119 min; and surgery IV, ⱖ 120 min.
F. Akinbami et al.
Morbidity in emergency general surgery
The results from this study may not be generalizable to other hospitals unless they have a similar mix of general surgical cases. One of the issues that arises with respect to patients who develop complications relates to our ability to rescue them from the complications. “Failure to rescue”22,23 may be a significant cause of additional morbidity and mortality after emergent surgical procedures. Failure to rescue refers to the use of interventions such as intensive care unit (ICU) admission or specific processes of care that lead to improved outcomes after complications.22,23 Another issue is the site of recovery for patients after emergent general surgical procedures. In many hospitals, emergent general surgical patients are not necessarily admitted to the surgical ICU. This is in contrast to coronary artery bypass patients who all are admitted to an ICU. Given the high rate of complications and attendant mortality, should all emergency cases receive ICU care? This would be a controversial recommendation because it may further increase hospital length of stay and health care costs but potentially may reduce complication rates. Another recommendation that may be worth testing to improve surgical outcomes relates to the association of increased BUN level with morbidity after emergency surgery. Given the increased BUN levels that we found in this study, it is possible that patients were somewhat hypovolemic and may have benefited from additional intravenous fluid resuscitation. This approach should be studied prospectively. As observed in this study, a low serum albumin level may be a marker of malnutrition. This marker would be difficult to treat in the short term. Albumin supplementation has been used in some series but its effect remains controversial.24,25 Because the baseline values for serum albumin and creatinine are not known in these patients, it is unclear if these recommendations would improve outcomes because the increased values could be a reflection of the patient’s severity of illness. Control of perioperative hyperglycemia is a simple and actionable process that may have benefit to reduce infectious complications.19 –21 Finally, composition of the surgical team that is caring for the emergent general surgical patient may be modified. Recent support for acute care surgeons who address both trauma and emergent general surgery is an avenue that needs to be explored more rigorously. Such acute care surgeons may be able to potentially decrease the length of surgery and delays to surgery through their expertise.26 Britt et al27 developed an acute care surgery model to decrease time delays to surgery by using daytime operating rooms and the experience of surgeons to improve medical decision making. Earley et al26 examined the effects of this model on outcomes in patients undergoing appendectomies and found a decrease in time to surgery, complication rates, and length of stay. Further study is needed to determine if modifying these factors improves outcomes. A larger multicenter study will be required to evaluate time variables such as delay in surgical timing as well as to enhance the generalizability of our findings to other hospitals.
461
Conclusions Emergency general surgery patients who develop postoperative complications are more likely to be older, male, smokers, and have increased perioperative blood glucose and creatinine levels and lower serum albumin levels. Those patients with complications also incur longer surgical times. Preoperative intravenous fluid administration to adequately resuscitate patients, tighter glucose control, and experienced surgical teams to decrease surgical times are putative targets to improve outcomes in emergency general surgery patients.
Acknowledgments The authors thank Stuart Lipsitz, PhD, Xiangmei Gu for statistical analysis.
References 1. American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP). Chapter 4, ACSNSQIP Variables and Definitions. Available at: https://acsnsqip.org/. 2. Bratzler DW. The surgical infection prevention and surgical care improvement projects: promises and pitfalls. Am Surg 2006;72(11): 1010 – 6. 3. Khuri SF, Daley J, Henderson WG. The comparative assessment and improvement of quality of surgical care in the Department of Veterans Affairs. Arch Surg 2002;137:20 –7. 4. Hall BL, Hamilton BH, Richards K, et al. Does surgical quality improve in the American College of Surgeons National Surgical Quality Improvement Program: an evaluation of all participating hospitals. Ann Surg 2009;250:363–76. 5. Costa G, Tomassini F, Tierno SM, et al. Emergency colonic surgery: analysis of risk factors predicting morbidity and mortality. Chir Ital 2009;61:565–71. 6. Neumayer L, Hosokawa P, El-Tamer M, et al. Multivariable predictors of postoperative surgical site infection after general and vascular surgery: results from the patient safety in surgery study. J Am Coll Surg 2007;204:1178 – 87. 7. Dimick JB, Chen SL, Taheri PA, et al. Hospital costs associated with surgical complications: a report from the private-sector National Surgical Quality Improvement Program. J Am Coll Surg 2004;199:531–7. 8. Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs’ NSQIP: the first national, validated, outcome-based, riskadjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Ann Surg 1998;228:491–507. 9. Khuri SF, Daley J, Henderson W, et al. Risk adjustment of the postoperative mortality rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:315–27. 10. Daley J, Khuri SF, Henderson W, et al. Risk adjustment of the postoperative morbidity rate for the comparative assessment of the quality of surgical care: results of the National Veterans Affairs Surgical Risk Study. J Am Coll Surg 1997;185:328 – 40. 11. Khuri SF, Henderson WG, Daley J, et al. The patient safety in surgery study: background, study design, and patient populations. J Am Coll Surg 2007;204:1089 –102.
462 12. Henderson WG, Khuri SF, Mosca C, et al. Comparison of riskadjusted 30-day postoperative mortality and morbidity in Department of Veterans Affairs hospitals and selected university medical centers: general surgical operations in men. J Am Coll Surg 2007; 204:1103–14. 13. Fink AS, Hutter MM, Campbell DC Jr, et al. Comparison of riskadjusted 30-day postoperative mortality and morbidity in Department of Veterans Affairs hospitals and selected university medical centers: general surgical operations in women. J Am Coll Surg 2007;204: 1127–36. 14. Fink AS, Campbell DA Jr, Mentzer RM Jr, et al. The National Surgical Quality Improvement Program in non-veterans administration hospitals: initial demonstration of feasibility. Ann Surg 2002;236:344 –54. 15. Daley J, Forbes MG, Young GJ, et al. Validating risk-adjusted surgical outcomes: site visit assessment of process and structure. National VA Surgical Risk Study. J Am Coll Surg 1997;185:341–51. 16. McGillicuddy EA, Schuster KM, Davis KA, et al. Factors predicting morbidity and mortality in emergency colorectal procedures in elderly patients. Arch Surg 2009;144:1162. 17. Haridas M, Malangoni MA. Predictive factors for surgical site infection in general surgery. Surgery 2008;144:496 –501. 18. Kasatpibal N, Nørgaard M, Sørensen HT, et al. Risk of surgical site infection and efficacy of antibiotic prophylaxis: a cohort study of appendectomy patients in Thailand. BMC Infect Dis 2006;6:111. 19. Ramos M, Khalpey Z, Lipsitz S, et al. Relationship of perioperative hyperglycemia and postoperative infections in patients who undergo general and vascular surgery. Ann Surg 2008;248:585–91.
The American Journal of Surgery, Vol 201, No 4, April 2011 20. Jones KW, Cain AS, Mitchell JH, et al. Hyperglycemia predicts mortality after CABG: postoperative hyperglycemia predicts dramatic increases in mortality after coronary artery bypass graft surgery. J Diabetes Complications 2008;22:365–70. 21. Lazar HL, Chipkin SR, Fitzgerald CA, et al. Tight glycemic control in diabetic coronary artery bypass graft patients improves perioperative outcomes and decreases recurrent ischemic events. Circulation 2004; 109:1497–502. 22. Silber JH, Williams SV, Krakauer H, et al. Hospital and patient characteristics associated with death after surgery. A study of adverse occurrence and failure to rescue. Med Care 1992;30:615–29. 23. Simpson KR. Failure to rescue: implications for evaluating quality of care during labor and birth. J Perinat Neonat Nurs 2005;19:24 –34. 24. Finfer S, Bellomo R, McEvoy S, et al. Effect of baseline serum albumin concentration on outcome of resuscitation with albumin or saline in patients in intensive care units: analysis of data from the saline versus albumin fluid evaluation (SAFE) study. BMJ 2006; 333:1044. 25. Vincent JL. Relevance of albumin in modern critical care medicine. Best Pract Res Clin Anaesthesiol 2009;23:183–91. 26. Earley AS, Pryor JP, Kim PK, et al. An acute care surgery model improves outcomes in patients with appendicitis. Ann Surg 2006;244: 498 –504. 27. Britt RC, Weireter LJ, Britt LD. Initial implementation of an acute care surgery model: implications for timeliness of care. J Am Coll Surg 2009;209:421– 4.