Article, Gastroenterology

Nasogastric decompression not associated with a reduction in surgery or bowel ischemia for acute small bowel obstruction

a b s t r a c t

Objectives: Small bowel obstructions (SBOs) occur 300,000 times annually leading to $1.3 billion in cost. Approx- imately 20% of patients require a laparotomy to manage the obstruction and either prevent or treat intestinal is- chemia. Early management may play a role in reducing these complications. Nasogastric decompression is commonly used for early management. Our primary objective was to determine if NGD was associated with lower rates of surgery, Bowel ischemia or length of stay.

Methods: We retrospectively enrolled 181 ED patients with SBO from 9/2013 to 9/2015 in order to determine if nasogastric decompression was associated with a reduction in rates of surgery, bowel ischemia or hospital length of stay.

Results: Our subject population was 46% female, median age of 60.27% of patients received surgery. Nasogastric decompression was used in 51% of patients. There was no association with a reduction in rates of surgery (p = 0.20) or bowel resection (p = 0.41) with patients receiving Nasogastric decompression, and no difference in baseline characteristics. Nasogastric decompression was associated with a two-day increase in hospital length of stay. Factors that were significantly associated with surgical exploration of SBO were: female (OR 2.32 (95% CI: 1.01-5.31)) and “definite SBO” on CT (OR 3.29 (95% CI: 1.18-9.20)). Abnormal vital signs, obstipation, and lab values were not predictors of surgery.

Conclusion: Nasogastric decompression is not associated with a reduction in need for surgery or bowel resection, but is associated with a 2-day increase in median LOS. Women were more likely to receive surgery than men.

(C) 2017

Introduction

Small bowel obstruction (SBO) is a common cause of surgical admis- sions, occurring nearly 300,000 times annually and costing the US healthcare system an estimated 1.3 billion dollars annually (1:2). The traditional approach to small bowel obstruction was to operate on an urgent basis in accord with maxim to “never let the sun rise and set on a SBO…” Mortality was observed to increase from 2% to 25% in SBO patients with underlying bowel ischemia who received delayed surgical care [3].

However, recent studies have shown that only 18% of SBOs receive surgery and the average time to surgery was two days (4:5:6). Success- ful conservative management may be advantageous because surgery it- self increases the risk of SBO recurrence due to the development of adhesions [7]. Thus, it is a paradox of SBO management that the

* Corresponding author at: George Washington University, 2120 L Street NW, Suite 450, Washington 20037, DC, United States.

E-mail address: [email protected] (A.C. Meltzer).

treatment is also the leading cause of the disease (8:9). Nasogastric de- compression has been traditionally recommended for early manage- ment of acute SBO, but there is significant variability regarding its use today. The primary objective of this study was to determine if Nasogas- tric decompression was associated with the following clinical out- comes: rate of surgery, rate of bowel resection and hospital length of stay. As a secondary objective, we aimed to identify early predictors for surgery or bowel resection.

Methods

Identification of subjects occurred by two steps. First, subjects were screened broadly with all likely ICD-9 codes that may have been linked to SBO including intussusception (560.0), paralytic ileus (560.1), volvu- lus (560.2), impaction of intestine unspecified (560.3), gallstone ileus (560.31), fecal impaction (560.32), other impaction of intestine (560.39), intestinal or Peritoneal adhesions with obstruction (560.8), other specified intestinal obstruction (560.89), or unspecified intestinal obstruction (560.9). Based on those criteria, 296 patient visits were

http://dx.doi.org/10.1016/j.ajem.2017.08.029

0735-6757/(C) 2017

1920 D.J. Berman et al. / American Journal of Emergency Medicine 35 (2017) 19191921

identified from September 2013 to September 2015. Second, we reviewed the treating radiologist report of each CT scan and classified

Table 1

Nasogastric decompression.

who specialized in interpreting CT scans of the abdomen and thorax. In- terpretation was performed as part of normal clinical care. Ultimately,

70 +? 37% 19%

Triage SBP (mm hg) b90 or N 160 25% 15%

the likelihood of SBO as “definite” or “likely/ early” or “no evidence.” Pa-

NGD (+) N = 93

NGD (-) N = 88

tients without confirmatory CT scans and patients with “no evidence of

SBO were excluded. In addition, patients with an alternative diagnosis

Female Age: 18-29

49%

3%

43%

10%

such as fecal impaction, constipation or large bowel obstructions were

20-49

18%

30%

excluded. Radiologists were all board certified attending physicians

50-69

42%

41%

Triage Temp (Celsius), b36 or N 38 13% 9%

181 patients were included in study.

Obstipation

28%

16%

Following identification of study subjects, all charts were abstracted

Hernia

25%

27%

by a minimum of two abstractors blinded to goals of the study using structured data abstraction sheets. Any discrepancy in data abstraction was arbitrated by a third data abstractor. Twenty-one potential predic- tors were abstracted for each patient including initial triage vital signs, lab values, clinical characteristics, co-morbidities and diagnostic imag-

ing including CT scan findings. We also looked at demographic features

Diabetes

Coronary artery disease? Crohn’s disease

15%

14%

26%

11%

4%

6%

Malignancy? 30% 17%

Prior abdominal surgery 88% 85%

Prior SBO? 56% 32%

Leukocytosis (N 12) 33% 31%

such as Patient sex, obstipation for over 24 h, known history of hernia,

history of diabetes, history of CAD (CVA, MI, stent placements, or

“Definite SBO”

81%

60%

CABG), vasculitis, history of Crohn’s Disease, history of malignancy,

“Likely/Early SBO”

19%

40%

prior abdominal or pelvic surgery and history of SBO. For each patient,

Surgery this hospitalization

31%

23%

we followed hospital course to determine whether the patient underwent NG decompression, abdominal surgery including lysis of ad-

Surgery on hospital day (#)

0 (admit date)

1

25%

38%

30%

25%

hesions, bowel resection during surgery, or, death.

2

3%

20%

The primary objective was to determine the association of NG de-

3

14%

10%

Lactate b24 h, b 4 mg/dL Level of confidence on CT?

99% 96%

compression with clinical outcomes. To accomplish this objective, we compared patients who received NG decompression with those that did not, to primary endpoints of surgical intervention, bowel resection during surgery, and hospital length of stay . The secondary objec- tive was to identify early predictors of surgery. For this outcome, we performed a logistic regression to compare groups that received surgery and those that did not receive surgery based on the 21 possible predic- tors described above. A pre-planned sub-analysis was also performed to determine predictors of two other outcomes, bowel resection and death. Study was conducted in accordance with established methods of chart review [10]. This study was approved by the hospital’s Institu- tional Review Board (IRB) with a waiver of informed consent. Hospital is a single center urban tertiary care hospital associated with a medical school.

Results

Among study population, 46% were female and median age was 60 years. 49 patients received surgery during hospitalization and 132 did not. The median hospital day that surgery was performed was day two. Among the subject population, 93 (51%) of patients re- ceived NG decompression and 88 (49%) did not (Table 1). Patients who were N 70 years old were more likely to receive NG decompres- sion and patients with a prior history of coronary artery disease (CAD), malignancy or prior SBO were more likely to receive NG decompression. In patients who received NG decompression, there was no association with a reduction in death, surgery or bowel resec- tion. There was a significantly increased distribution of hospital length of stay for those receiving NG decompression (median LOS, 5 days) versus those who did not receive NG decompression (3 days), (p b 0.0001.) (Table 1)

The following factors were identified as significantly associated with surgical exploration of small bowel obstruction by laparotomy:

[1] female (OR 2.32 (95% CI: 1.01-5.31)) and [2]“definite SBO” versus “likely SBO” on CT (OR 3.29 (95% CI: 1.18-9.20)). In this model, ab- normal triage vital signs, presence of obstipation, NG decompression and lab studies were not associated with the need for surgery (Table 2). When the model was run again to the clinically significant outcome of bowel resection during surgery, no early predictors were clinically significant.

4 7% 0

5 0 10%

6+ 13% 5%

Mean hospital LOS (days)? 7.0 4.2

Median hospital LOS (days)? 5 3

Surgical bowel resection 13% 9%

Died during hospitalization 3% 2%

* p b 0.05.

Discussion

Early management by emergency physicians may play a role in long- term outcomes of patients with SBO including reducing the risk of ische- mic bowel. The mainstays in the early management include fluid resus- citation, NG decompression and urgent Surgical consultation [7]. NG decompression is intended to provide symptomatic relief and potential- ly decrease the need for surgery [14]. NG decompression is cited as part of the standard treatment in the Bologna guidelines (Level of evidence, 2b) [15]. However, contradicting current guidelines and expert opinion, we found no association between NG decompression and a reduction in rates of surgery or surgical bowel resection [14].

We did detect an association between NG decompression and a 2 day longer hospital stay. This increased length of stay may be due to

Table 2

Odds ratio of surgery (N = 180).

Odds ratio 95% CI

Female

2.32

1.01-5.31

Age N 60

0.42

0.16-1.07

SBP b 90 or N 160 (mm Hg)

1.45

0.54-3.87

HR b60 or N 100 (bpm)

1.43

0.57-3.54

Temp b 36 or N 38 (C)

0.61

0.18-2.10

CT confirmed SBO

“Likely”/”Early” 1.00

“Definite” 3.29 1.18-9.20

Obstipation 1.26 0.48-3.28

History of hernia 1.68 0.67-4.22

History of Crohn’s disease 0.44 0.04-5.40

Prior abdominal surgery 0.56 0.18-1.81

Prior SBO 0.77 0.27-2.17

Prior surgery for SBO 0.41 0.11-1.50

NG suction 1.70 0.74-3.94

Leukocytosis N 12 1.75 0.77-3.99

D.J. Berman et al. / American Journal of Emergency Medicine 35 (2017) 19191921 1921

patients with NG decompression getting a longer trial of conservative management before committing to surgery. Determining which pa- tients with SBO are likely to respond to conservative management and which require surgery to prevent bowel ischemia is a challenge. There was no association with initial vital signs, lab values or co-morbidities. Serum lactate and metabolic acidosis are late markers of bowel ischemia and do not become elevated until after Bowel necrosis has occurred [12]. Finally, Physical exam findings on the Abdominal examination have low rates of inter-observer agreement.

All patients in our study had a CT-confirmed SBO. Diagnosis of SBO is typically made with CT scan because CTs are highly accurate and fre- quently identify the level and cause of obstruction [7]. A limitation of CT is that they do not reliably identify or predict ischemic bowel for pa- tients with SBO. In our study, there was significant variability on specific signs that marked the presence of ischemia, such as mesenteric edema and pneumatosis, or other signs such as degree of gastric distension, the presence a small bowel fecal sign, or, the descriptors “high-grade” versus “low-grade.” Given the variability and low inter-observer agree- ment between abstractors, we were unable to reliably code by those in- dividual signs. However, the description of a “definitive SBO,” “likely/ early SBO,” or “no evidence of SBO” was reliably reported and reliably abstracted with high inter-observer agreement.

There are two important limitations of this study. The most impor- tant limitation is inherent in the retrospective study design in that the patients who received NG decompression may be categorically different from the patients that did not. Patients who received NG decompression were more likely to be older and more likely to have co-morbidities such as coronary artery disease and malignancy. The second major lim- itation is that this study was conducted at a single site and may not be generalizable to different patient populations and different provider groups.

In conclusion, NG decompression was not associated with a reduc- tion in operative management of SBOs or Major adverse events but was associated with longer hospital length of stay.

Prior presentations

SAEM17 Mid Atlantic Regional Conference, March 11, 2017 – Naso- gastric Decompression is Associated with Increased Hospital Length

of Stay but No Reduction in Need for Surgery in ED Patients with Acute Small Bowel Obstruction

SAEM17 Mid Atlantic Regional Conference, March 11, 2017 – Predicting the Need for Surgery in ED Patients with Acute Small Bowel Obstruction

Funding sources/disclosures

Funding for medical students provided GWU Gill Fellowship. All authors have no financial conflicts of interest to disclose.

Acknowledgments

Babak Sarani, MD; Robert Shesser MD

References

  1. Frasure S, Hildreth A, Takhar S, Stone M. Emergency department patients with small bowel obstruction: what is the anticipated clinical course? World J Emerg Med 2016;7:35-9.
  2. Loftus T, Moore F, VanZant E, Bala T, Brakenridge S, Croft C, et al. A protocol for the management of adhesive small bowel obstruction. J Trauma Acute Care Surg 2015; 78:13-21.
  3. Paulson EK, Thompson WM. Review of Small-bowel obstruction: the diagnosis and when to worry. Radiology 2015;275:332-42.

    [7] Maung AA, Johnson DC, Piper GL, Barbosa RR, Rowell SE, Bokhari F, et al. Evaluation and management of small-bowel obstruction: an eastern association for the surgery of trauma practice management guideline. The Journal of Trauma and Acute Care Surgery 2012;73:S362-9.

    [10] Gilbert EH, Lowenstein SR, Koziol-McLain J, Barta DC, Steiner J. Chart reviews in emergency medicine research: where are the methods? Ann Emerg Med 1996; 27:305-8.

    [12] Takeuchi K, Tsuzuki Y, Ando T, Sekihara M, Hara T, Yoshikawa M, et al. Clinical stud- ies of strangulating small bowel obstruction. Am J Surg 2004;70:40-4.

    [14] Diaz JJ, Bokhari F, Mowery NT, Acosta JA, Block EF, Bromberg WJ, et al. Guidelines for management of small bowel obstruction. J Trauma 2008;64:1651-64.

    [15] Di Saverio S, Coccolini F, Galati M, Smerieri N, Biffl WL, Ansaloni L, et al. Bologna guidelines for diagnosis and management of adhesive small bowel obstruction (ASBO): 2013 update of the Evidence-based guidelines from the world society of emergency surgery ASBO working group. World Journal of Emergency Surgery 2013;1.