Abstract
This cohort study uses data from the National Surgical Quality Improvement Program database to examine factors associated with mortality risk among patients undergoing emergency colectomy for acute lower gastrointestinal bleeding.
Patients admitted to the hospital for an acute lower gastrointestinal bleed (LGIB) require emergency colectomy in 10% to 25% of cases.1 Existing reports of mortality after emergency surgery for LGIB come from small, single-center series and vary between 2% to 60%.2 Furthermore, few studies describe outcomes in the context of modern critical care, endoscopy, and interventional radiology.2 Prognostic uncertainty is a barrier to providing goal-concordant care in the emergency setting.3 We therefore aimed to describe the risk of mortality after emergency colectomy for acute LGIB.
Methods
We conducted a retrospective cohort study using the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database, which systematically collects data from participant hospitals and is fully described elsewhere.4 The University of Toronto research ethics board approved this study. Informed consent was waived because of the deidentified nature of the data.
We included adult patients undergoing emergency surgery during 2005 to 2017 with diagnostic and procedural codes associated with LGIB and colectomy, respectively (Figure). Emergency cases were defined using the emergency field in the NSQIP database. We excluded cancer cases and patients with sepsis.
Figure. Flow Diagram Illustrating Inclusion and Exclusion of Study Patients.
CPT indicates Current Procedural Terminology; ICD-9, International Classification of Diseases Ninth Revision; ICD-10, International Classification of Diseases Tenth Revision; NSQIP, National Surgical Quality Improvement Program.
The outcome was 30-day mortality. Mortality for the cohort was summarized as a proportion and 95% Wald confidence interval (CI). Univariable associations between patient characteristics and mortality were explored using χ2 tests and t tests. We used multivariable logistic regression to identify factors associated with mortality, including variables selected a priori (age, sex, dyspnea, functional status, hypertension, preoperative international normalized ratio, preoperative platelets, preoperative hematocrit, bleeding disorder, preoperative systemic inflammatory response syndrome, time between admission and surgery, American Society of Anesthesiologists [ASA] classification, operative approach, and procedure) based on a literature review and clinical reasoning. Odds ratios (ORs) and 95% CIs are reported for factors significantly associated with mortality on multivariable analysis.
All analyses were performed using SAS, version 9.4 (SAS Institute). Two-tailed P < .05 indicated statistical significance. Missing data were handled with a complete case analysis.
Results
A total of 1614 patients underwent emergency colectomy for acute LGIB from 2005 to 2017. Thirty-day mortality was 12.2% (95% CI, 10.6%-13.8%). Nonsurvivors were older, had more comorbidities, had a higher preoperative international normalized ratio, lower hematocrit, and higher ASA class. Nonsurvivors more often had systemic inflammatory response syndrome, underwent open surgery, and received total/subtotal colectomy (Table).
Table. Characteristics of Study Patients.
| Characteristic | No. (%) | P Valuea | |
|---|---|---|---|
| Survivors (n = 1417) | Nonsurvivors (n = 197) | ||
| Age, mean (SD), y | 71 (12) | 75 (11) | <.001 |
| Female sex | 545 (38.5) | 77 (39.1) | .87 |
| Smoker | 193 (13.6) | 25 (12.7) | .72 |
| Dyspnea at baseline | 145 (10.2) | 30 (15.2) | .03 |
| Dependent functional status | 253 (17.9) | 68 (34.5) | <.001 |
| COPD | 127 (9.0) | 41 (20.8) | <.001 |
| Ascites | 20 (1.4) | 7 (3.6) | .03 |
| CHF | 62 (4.4) | 22 (11.2) | <.001 |
| Hypertension | 1079 (76.2) | 163 (82.7) | .04 |
| Long-term steroid use | 85 (6.0) | 22 (11.2) | .01 |
| Preoperative dialysis | 100 (7.1) | 20 (10.2) | .12 |
| Bleeding disorder | 329 (23.2) | 66 (33.5) | .002 |
| Ventilator dependence | 67 (4.7) | 24 (12.2) | <.001 |
| Preoperative values | |||
| Hematocrit, %, mean (SD)b | 26.3 (5.6) | 25.0 (5.9) | .003 |
| Platelets, × 103/μL, mean (SD)b | 166 (8) | 163 (9) | .59 |
| INR, mean (SD)b | 1.22 (0.26) | 1.27 (0.27) | .01 |
| SIRS | 330 (23.3) | 61 (31.0) | .02 |
| Time from admission to surgery, mean (SD), d | 4 (5) | 4 (7) | .11 |
| ASA class 3-5 (vs 1-2) | 1291 (91.1) | 196 (99.5) | <.001 |
| Laparoscopic approach (vs open) | 207 (14.6) | 12 (6.1) | .001 |
| Total/subtotal colectomy (vs partial) | 324 (22.9) | 60 (30.5) | .02 |
Abbreviations: ASA, American Society of Anesthesiologists’ classification; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; INR, international normalized ratio; SIRS, systemic inflammatory response syndrome.
SI conversion factors: To convert hematocrit to proportion of 1.0, multiply by 0.01; for platelets to ×106/L, multiply by 1.
P values are calculated from t tests for continuous variables and χ2 tests for categorical variables.
Preoperative laboratory values were drawn closest to operative procedure time.
In the multivariable logistic regression model, older age (OR, 1.42; 95% CI, 1.20-1.68 per 10-year increase), functional dependence (OR, 2.14; 95% CI, 1.48-3.08), and higher ASA class (OR, 8.40; 95% CI, 1.15-61.59) were associated with statistically significant increases in the odds of mortality. A laparoscopic approach (OR, 0.46; 95% CI, 0.23-0.89) and higher hematocrit value (OR, 0.65; 95% CI, 0.47-0.90 per 10% increase) were associated with statistically significant reductions in the odds of mortality.
Discussion
Using a large, multi-institutional database, we found that emergency colectomy for LGIB is uncommon but is associated with high 30-day mortality compared with elective colectomy (in-hospital mortality is <1%5).
Of the examined patient and operative characteristics, high ASA class conferred the highest odds of mortality on multivariable analysis. This finding is consistent with existing literature describing an association between ASA and mortality after emergency colectomy.6 This finding, along with the association between lower preoperative hematocrit and mortality, underscores the importance of obtaining early surgical consultation to ensure patients who need surgery receive it in a timely fashion.
This study has limitations. Preoperative interventions are not documented in the database, so we could not describe associations between postoperative mortality and interventions, such as colonoscopy. Functional outcomes, while important, are also not available in the database. That laparoscopic approach was associated with a lower odds of mortality likely represents confounding by indication.
This study provides a robust estimate of the risk of mortality for patients undergoing emergency surgery for acute LGIB in the contemporary context. This prognostic information will better equip surgeons to provide goal-concordant care.
References
- 1.Hoedema RE, Luchtefeld MA. The management of lower gastrointestinal hemorrhage. Dis Colon Rectum. 2005;48(11):2010-2024. doi: 10.1007/s10350-005-0138-1 [DOI] [PubMed] [Google Scholar]
- 2.Pfeifer J. Surgical management of lower gastrointestinal bleeding. Eur J Trauma Emerg Surg. 2011;37(4):365-372. doi: 10.1007/s00068-011-0122-5 [DOI] [PubMed] [Google Scholar]
- 3.Cooper Z, Koritsanszky LA, Cauley CE, et al. Recommendations for best communication practices to facilitate goal-concordant care for seriously ill older patients with emergency surgical conditions. Ann Surg. 2016;263(1):1-6. doi: 10.1097/SLA.0000000000001491 [DOI] [PubMed] [Google Scholar]
- 4.Khuri SF. The NSQIP: a new frontier in surgery. Surgery. 2005;138(5):837-843. doi: 10.1016/j.surg.2005.08.016 [DOI] [PubMed] [Google Scholar]
- 5.Van Arendonk KJ, Tymitz KM, Gearhart SL, Stem M, Lidor AO. Outcomes and costs of elective surgery for diverticular disease: a comparison with other diseases requiring colectomy. JAMA Surg. 2013;148(4):316-321. doi: 10.1001/jamasurg.2013.1010 [DOI] [PubMed] [Google Scholar]
- 6.Ballian N, Weisensel N, Rajamanickam V, et al. Comparable postoperative morbidity and mortality after laparoscopic and open emergent restorative colectomy: outcomes from the ACS NSQIP. World J Surg. 2012;36(10):2488-2496. doi: 10.1007/s00268-012-1694-x [DOI] [PubMed] [Google Scholar]
