Advertisement

Prompt intervention in large bowel obstruction management: A Nationwide Inpatient Sample analysis

Open AccessPublished:July 18, 2022DOI:https://doi.org/10.1016/j.amjsurg.2022.07.002

      Highlights

      • Etiology of LBO: 20% colon cancer, 15.6% extracolonic malignancy, 64.7% benign.
      • Prompt intervention (occurring within 2 days of admission) in LBO: 42.6% of cases.
      • Prompt intervention less likely in Blacks with benign LBO and women with malignant LBO.
      • Prompt intervention increased discharge to home and decreased LOS, but not mortality.

      Abstract

      Background

      Large bowel obstruction is an urgent condition which can progress to ischemia and perforation. The importance of prompt intervention has not been rigorously demonstrated.

      Methods

      Patients with bowel obstruction who underwent stoma, stent, and/or colectomy in the Nationwide Inpatient Sample were used to study prompt intervention (defined as occurring within 2 days of admission). Outcomes were inpatient mortality, discharge to home, and length of stay in an adjusted analysis.

      Results

      Among the 31,277 patients, prompt intervention occurred in 42.6%. In an adjusted analysis, prompt intervention was more likely in higher income patients and less likely in patients with comorbidities; among those with malignant obstruction, less likely in women, and among those with benign obstruction, less likely in Blacks. Inpatient mortality (6%) was not different between groups. Discharge home (71% vs 68%; p < 0.0001) and shorter LOS (−3 days) occurred in those managed promptly.

      Conclusion

      Prompt intervention in large bowel obstruction results in decreased LOS and greater likelihood of discharge to home, but not a mortality benefit. Female, Black and lower income patients were less likely to have prompt intervention.

      Keywords

      1. Introduction

      Large bowel obstruction (LBO) typically presents urgently. Frequently the diagnosis and etiology of obstruction is unknown at presentation. Patients therefore present a dual challenge--determining a management approach that addresses the acute obstruction while also appropriately addressing the underlying colon pathology, which may be malignant. Adding to the complexity of decision-making is time pressure. Delaying management of any bowel obstruction risks bowel ischemia, perforation, sepsis, prolonged recovery, or death. However, there are reasonable arguments to defer intervention beyond 24 h, including resuscitation, bowel decompression in order to allow safe induction of anesthesia, and/or limitations in operating resources or expertise. Furthermore, some patients may present with a non-toxic clinical appearance despite dilated proximal bowel on imaging, resulting in de-prioritization of these patients by the surgical team. Unfortunately, the paucity of published guidelines results in a case-by-case determination of urgency, which can lead to variability in care. Using a large administrative database, we sought to evaluate whether prompt intervention (surgery or colonic stent) in patients with large bowel obstruction is associated with better outcomes.

      2. Methods

      Patients in the Nationwide Inpatient Sample (NIS), a population-based dataset, discharged between January 1, 2010 and September 30, 2015 with ICD-9 diagnosis codes for bowel obstruction (560.89, 560.9) and who underwent stoma, colorectal resection, or colonic stent placement were studied, including only those admitted emergently (Codes shown in Supplementary Appendix). Prompt intervention, the predictor variable of interest, was defined as ostomy creation surgery, colon resection, or colonic stent placement occurring the day-of or day-after admission. Etiology of obstruction was determined through a hierarchical selection process using ICD-9 diagnoses codes to generate the following four categories: colon cancer without metastasis, metastatic colon cancer, extracolonic malignancy, and those without malignant diagnosis codes were categorized as benign obstruction.
      • Kwaan M.R.
      • Ren Y.
      • Wu Y.
      • Xirasagar S.
      Colonic stent use by indication and patient outcomes: a Nationwide inpatient sample study.
      Extracolonic malignancy included tumors with peritoneal spread that cause extrinsic compression of the bowel, such as metastatic pancreatic or ovarian cancers. Study outcomes were inpatient mortality, length of stay (LOS) and discharge to home. Procedures were defined as colonic stent placement, ostomy creation surgery, or colon resection using ICD-9 procedure codes.
      Multiple logistic/linear regression analyses were performed separately for the malignant and benign obstruction groups, to study the association of prompt intervention with each outcome, adjusted for age, sex, race, Charlson comorbidity index, etiology of obstruction, socioeconomic factors, and hospital characteristics. Regression models excluded observations with missing data on income, payer, hospital characteristics, hospital region, and transfer status (all of which had too few patients with missing data to include as a category). Observations with missing data or unknown status on other variables were included with “missing or unknown” as a category. LOS was studied using linear regression, after verifying near-normal distribution which was noted up to LOS 35. Outlier LOS values > 35 days (1442 patients, 3.4% of the sample) were set to 35 days for regression purposes. Data were analyzed in SAS (V.8.4, SAS Institute, Cary, NC). The study was exempt from IRB review as de-identified data were provided by the AHRQ Healthcare Cost and Utilization Project.

      3. Results

      Among the 31,277 study patients admitted emergently with large bowel obstruction, 54% were female, 69% White, 12% Black, and 12% other race (Table 1). The sample median age was 66 years (±16 years). Nearly 20% of patients had colon cancer, 15.6% extracolonic malignancy, and 64.7% had a benign etiology for obstruction. Among patients with benign obstruction, 16.1% had diverticular disease, 13.5% had vascular insufficiency (presumably ischemic colonic strictures), 5.7% Crohn's disease, 1.5% ulcerative colitis, and the remaining 63.2% had other diagnoses or were not specified.
      Table 1Patient, procedure, and treating hospital characteristics.
      Total patients N = 31277

      Mean (SD) or No. (%)
      Obstruction-directed intervention
      Prompt (≤2days)

      N = 13323, 42.6%

      No. (%)
      Not prompt (>2days)

      N = 17954, 57.4%

      No. (%)
      p
      Age, years
       Mean age (SD)64.6 (16.1)64.3 (16.2)64.9 (16.1)0.001
       18–443035 (9.7)1372 (10.3)1663 (9.3)<0.0001
       45-543724 (11.9)1698 (12.7)2026 (11.3)
       55-645793 (18.5)2572 (19.3)3221 (17.9)
       ≥6514147 (45.2)6148 (46.2)7999 (44.6)
      Missing4578 (14.6)1533 (11.5)3045 (17.0)
       Sex
       Male14309 (45.8)6194 (46.5)8115 (45.2)0.02
       Female16964 (54.2)7126 (53.5)9838 (54.8)
      Race
       Black3749 (12.8)1482 (11.7)2267 (13.7)<0.0001
       White21619 (74.0)9424 (74.5)12195 (73.6)
       Other3864 (13.2)1748 (13.8)2116 (12.8)
      Charlson Index
       0-218275 (58.5)8478 (63.7)9797 (54.6)<0.0001
       3-65132 (16.4)2031 (15.3)3101 (17.3)
       ≥77828 (25.1)2796 (21.0)5032 (28.1)
      Sum of comorbidities
       Median (IQR), Mean12 (8–17), 12.511 (7–16), 11.713 (8–17), 13.1<0.0001
      Etiology of obstruction
       Colon cancer2960 (9.5)1181 (8.9)1779 (9.9)<0.0001
       Metastatic colon cancer3204 (10.2)1197 (9.0)2007 (11.2)
       Extracolonic malignancy4885 (15.6)1726 (13.0)3159 (17.6)
       Benign obstruction20228 (64.7)9219 (69.2)11009 (61.3)
       Transfer status
      Not transferred28620 (92.0)12169 (91.9)16451 (92.1)0.52
       Transferred from other facility2487 (8.0)1074 (8.1)1413 (7.9)
      Primary payer
      Insurance information was missing for 4438 patients.
       Medicaid2555 (9.5)1109 (9.4)1446 (9.6)0.03
       Private/Medicare/Other22857 (85.2)10053 (84.9)12804 (85.4)
       Uninsured1427 (5.3)676 (5.7)751 (5.0)
      Zip code income quartile
       >75th percentile of income6916 (22.6)3071 (23.6)3845 (21.9)0.001
       ≤75th percentile of income23697 (77.4)9969 (76.5)13728 (78.1)
      Hospital location/teaching status
       Urban teaching16670 (53.5)7050 (53.1)9620 (53.8)<0.0001
       Urban non-teaching11414 (36.6)4739 (35.7)6675 (37.3)
       Rural3085 (9.9)1481 (11.2)1604 (9.0)
      Hospital ownership
       Public, non-federal9020 (28.9)3594 (27.1)5426 (30.3)<0.0001
       Private, non-profit18566 (59.6)8208 (61.9)10358 (57.9)
       Private, for-profit3583 (11.5)1468 (11.1)2115 (11.8)
      Stent procedure
       Yes1686 (5.4)561 (4.2)1125 (6.3)<0.0001
       No29591 (94.6)12762 (95.8)16829 (93.7)
      Stoma procedure
       Yes11954 (38.2)4993 (37.5)6961 (38.8)0.02
       No19323 (61.8)8330 (62.5)10993 (61.2)
      Colon resection
       Left-sided resection17715 (56.6)8114 (60.9)9601 (53.5)<0.0001
       Right-sided resection7514 (24.0)2995 (22.5)4519 (25.2)
       No colon resection6048 (19.3)2214 (16.6)3834 (21.4)
      Discharge disposition
       Died in hospital1876 (6.0)788 (5.9)1088 (6.1)<0.0001
       Discharged to home21658 (69.3)9449 (71.0)12209 (68.1)
       Transfer to other facility7723 (24.7)3080 (23.1)4643 (25.9)
      Length of stay of patients discharged alive
      1876 patients who died during hospitalization were excluded from length of stay analysis.
       Median (IQR), Mean10 (7–14), 11.78 (6–11), 9.811 (8–15), 13.1<0.0001
      a Insurance information was missing for 4438 patients.
      b 1876 patients who died during hospitalization were excluded from length of stay analysis.
      Prompt intervention occurred in 42.6%. Patient, admission, and hospital characteristics, classified by whether or not intervention was prompt, are also shown in Table 1.

      3.1 Interventions

      Colonic stent placement was performed in 5.4% of patients, and was less likely in the prompt intervention group both in unadjusted and adjusted analyses (Table 1, Table 2). Ostomies were created in 38.2% of patients. The majority of patients had colon resections (n = 25,229; 80.7%): 56.6% had left-sided resections and 24% had right-sided resections (Table 1).
      Table 2Factors associated with prompt intervention in acute large bowel obstruction with malignant and benign etiology.
      Predictors of prompt interventionMalignant etiology
      C = 0.577.
      (n = 10954)
      Benign etiology
      C = 0.583.
      (n = 19999)
      Adjusted O.R. (95% C.I.)pAdjusted O.R. (95% C.I.)p
      Stoma procedure
       Yes1.3 (1.2–1.4)<.00010.83 (0.78–0.88)<.0001
       No1.0 [Reference]1.0 [Reference]
      Stent procedure
       Yes0.8 (0.7–1.0)0.020.7 (0.6–0.8)<.0001
       No1.0 [Reference]1.0 [Reference]
      Age, years
       18-441.0 [Reference]1.0 [Reference]
       45-541.2 (1.0–1.5)0.111.1 (1.0–1.2)0.07
       55-641.1 (0.9–1.3)0.481.1 (1.0–1.3)0.01
       ≥651.2 (1.0–1.4)0.141.1 (0.9–1.2)0.37
       Missing0.8 (0.5–1.3)0.8 (0.5–1.1)0.17
      Sex
       Male1.1 (1.0–1.2)0.011.05 (0.99–1.11)0.10
       Female1.0 [Reference]1.0 [Reference]
      Race
       Black0.9 (0.8–1.0)0.200.86 (0.79–0.95)0.002
       Other1.1 (0.9–1.2)0.291.1 (1.0–1.2)0.01
       Unknown0.7 (0.6–0.9)0.00020.6 (0.5–0.7)<.0001
       White1.0 [Reference]1.0 [Reference]
      Charlson Comorbidity Index
       ≥70.76 (0.68–0.85)<.00010.5 (0.4–0.6)<.0001
       3-60.8 (0.7–0.9)<.00010.82 (0.76–0.89)<.0001
       0-21.0 [Reference]1.0 [Reference]
      Transfer status
       Transferred from other facility1.1 (0.9–1.2)0.471.0 (0.9–1.1)0.68
       Not transferred1.0 [Reference]1.0 [Reference]
      Primary payer
       Medicaid1.0 (0.8–1.3)0.910.9 (0.7–1.0)0.06
       Private/Medicare/Other0.9 (0.8–1.2)0.600.9 (0.8–1.0)0.11
       Unknown1.0 (0.6–1.7)0.930.8 (0.5–1.2)0.23
       Uninsured1.0 [Reference]1.0 [Reference]
      Zip code income quartile
       >75th percentile of income1.1 (1.0–1.2)0.051.14 (1.06–1.22)0.003
       Unknown1.1 (0.8–1.4)0.641.0 (0.8–1.2)0.82
       ≤75th percentile of income1.0 [Reference]1.0 [Reference]
      Hospital ownership
       Public, non-federal1.3 (1.1–1.5)0.0021.0 (0.9–1.1)0.66
       Private, non-profit1.3 (1.2–1.5)<.00011.1 (1.0–1.2)0.05
       Private, for-profit1.0 [Reference]1.0 [Reference]
      Hospital location/teaching status
       Urban teaching0.7 (0.6–0.8)<.00010.85 (0.77–0.93)0.001
       Urban non-teaching0.7 (0.6–0.8)<.00010.8 (0.7–0.9)<.0001
       Rural1.0 [Reference]1.0 [Reference]
      a C = 0.577.
      b C = 0.583.

      3.2 Patient, procedure, and hospital factors associated with prompt intervention

      In an adjusted analysis, among both the malignant and benign groups, prompt intervention was more likely in patients residing in high-income zip codes (top quartile of median income within zip code) and those managed at rural (versus urban) and private non-profit (versus investor-owned) hospitals (Table 2). Prompt intervention was less likely in patients with more comorbidities and those managed with colonic stents. Insurance status/type was not associated with prompt intervention. In the malignant obstruction group, prompt intervention was more likely in men, and showed greater likelihood of stoma procedures. In the benign obstruction group, prompt intervention was more likely in patients of “other” race and less likely among Blacks (versus Whites), and was associated with lower likelihood of stoma procedures.

      3.3 Outcomes

      Inpatient mortality was 6%, similar in the prompt intervention group and those who had intervention beyond the first two days (5.9% versus 6.1%; p = 0.58, Table 1). In adjusted analyses (Table 3), prompt intervention was not associated with inpatient mortality in patients with either benign or malignant disease. Of note, among those with a benign etiology of obstruction, adjusted mortality risk was higher for Blacks compared with Whites, and was lower for patients from high income zip codes. In benign and malignant obstruction groups, teaching hospitals showed lower mortality. Insurance was not associated with mortality.The median LOS for patients discharged alive was 10 days (IQR 7–14), and was 3 days shorter in those who were managed with prompt intervention (median 8 [IQR 6–11] versus median 11 [IQR 8–15]; p < 0.0001; Table 1). In adjusted analyses, the LOS was 3 days shorter in patients with malignant disease who had prompt intervention (β (SE): 3 (0.1); p < 0.0001), and was 3.1 d shorter in those with benign disease who had prompt intervention (β (SE): 3.1 (0.1); p < 0.0001). Other factors contributing to LOS are shown in the supplementary Table.Discharge to home occurred for 69% of all patients, 70.9% for the prompt intervention group, and 68% for patients who had intervention beyond the first two days. In adjusted analyses, prompt intervention was associated with greater likelihood of discharge to home, in both malignant and benign obstruction groups [Malignant: adjusted OR 1.2 (1.1–1.3), p = 0.003, Benign 1.2 (1.1–1.3), p < 0.0001]. Other factors predicting discharge to home are shown in the supplementary Table.
      Table 3Factors associated with inpatient death among acute bowel obstruction patients.
      Predictors of inpatient deathMalignant etiology
      C = 0.805.
      (n = 10952)
      Benign etiology
      C = 0.844.
      (n = 19994)
      Adjusted O.R. (95% C.I.)pAdjusted O.R. (95% C.I.)p
      Prompt intervention
       Yes1.0 (0.8–1.2)0.921.0 (0.9–1.2)0.74
       No1.0 [Reference]1.0 [Reference]<.0001
      Age, years
       18-441.0 [Reference]1.0 [Reference]
       45-541.0 (0.6–1.9)0.901.3 (0.8–2.1)0.25
       55-641.1 (0.7–1.9)0.662.2 (1.5–3.4)0.0002
       ≥652.1 (1.3–3.5)0.015.6 (3.7–8.4)<.0001
       Missing2.2 (0.7–6.8)0.177.8 (3.4–17.9)<.0001
      Sex
       Male1.0 (0.9–1.2)0.681.0 (0.9–1.1)0.74
       Female1.0 [Reference]1.0 [Reference]
      Race
       Black0.9 (0.7–1.2)0.421.3 (1.1–1.6)0.01
       Other0.8 (0.6–1.0)0.101.1 (0.9–1.3)0.47
       Unknown0.9 (0.6–1.3)0.551.2 (0.9–1.5)0.20
       White1.0 [Reference]1.0 [Reference]
      Charlson Comorbidity Index
       ≥71.5 (1.1–2.1)0.012.0 (1.4–2.6)<.0001
       3-61.4 (1.0–1.9)0.091.6 (1.3–1.8)<.0001
       0-21.0 [Reference]1.0 [Reference]
       Sum of Comorbidities1.09 (1.07–1.10)<.00011.10 (1.09–1.11)<.0001
      Perforation/peritonitis secondary diagnosis
       Yes5.4 (4.5–6.5)<.00015.3 (4.6–6.1)<.0001
       No1.0 [Reference]1.0 [Reference]
      Stoma procedure
       Yes1.3 (1.1–1.5)0.0031.1 (1.0–1.2)0.24
       No1.0 [Reference]1.0 [Reference]
      Transfer status
       Transferred from other facility1.3 (1.0–1.8)0.071.1 (0.9–1.4)0.20
       Not transferred1.0 [Reference]1.0 [Reference]
      Primary payer
       Medicaid0.9 (0.5–1.6)0.711.1 (0.6–1.8)0.78
       Private/Medicare/Other0.8 (0.5–1.3)0.370.9 (0.6–1.4)0.59
       Unknown0.5 (0.2–1.6)0.250.4 (0.2–1.1)0.07
       Uninsured1.0 [Reference]1.0 [Reference]
      Zip code income quartile
       >75th percentile of income1.1 (0.9–1.3)0.530.9 (0.7–1.0)0.05
       Unknown1.1 (0.6–2.0)0.721.0 (0.7–1.6)0.88
       ≤75th percentile of income1.0 [Reference]1.0 [Reference]
      Hospital ownership
       Public, non-federal1.2 (0.9–1.7)0.201.0 (0.8–1.3)0.68
       Private, non-profit0.8 (0.6–1.0)0.070.8 (0.7–1.0)0.02
       Private, for-profit1.0 [Reference]1.0 [Reference]
      Hospital location/teaching status
       Urban teaching0.7 (0.5–0.9)0.010.8 (0.7–1.0)0.05
       Urban non-teaching0.8 (0.6–1.1)0.160.9 (0.7–1.1)0.20
       Rural1.0 [Reference]1.0 [Reference]
      a C = 0.805.
      b C = 0.844.

      3.4 Demographic subgroups of interest

      Because race, sex and income were significant predictors of prompt intervention (Table 2), we describe patients in these demographic subgroups to provide additional context to the findings (Table 4). Blacks were younger than Whites (by 5.6 years), had more comorbidities, and were more likely to have obstructing colon cancers than Whites. Women were older, had fewer comorbidities, more extracolonic malignancies and benign obstructions. Patients residing in lower income zip codes had more benign obstructions.
      Table 4Patient characteristics and obstruction etiology by race, sex, and zip code income.
      All patients N = 31277RaceSexZip code income quartile
      White 21595 (74.0%)Black 3742 (12.8%)Other 3856 (13.2%)pFemale 16942 (54.3%)Male 14289 (45.8%)p>75th percentile≤75th percentilep
      Mean (SD) age in years64.6 (16.1)66.1 (15.7)60.5 (15.9)60.6 (17.0)*66.4 (16.0)62.5 (16.1)*65.0 (16.6)64.6 (16.0)@
      Charlson Comorbidity index
       0-258.4%59.1%54.9%58.3%*60.0%56.6%*58.5%58.5%@
       3-616.5%16.6%17.2%15.1%15.9%17.0%15.6%16.8%
       ≥725.1%24.3%27.9%26.7%24.0%26.3%26.0%24.8%
      Obstruction etiology
       Colon cancer9.5%9.1%10.0%11.0%*8.7%10.4%*9.2%9.5%*
       Metastatic colon cancer10.3%9.7%12.2%11.5%8.7%12.1%10.6%10.2%
       Extracolonic malignancy15.6%15.6%15.2%15.9%17.1%13.8%17.3%15.2%
       Benign obstruction64.7%65.6%62.6%61.6%65.5%63.7%62.9%65.2%
      *: p < 0.01; @: p < 0.05.

      4. Discussion

      In this contemporary United States population-based analysis, just under half of emergently admitted patients with large bowel obstruction underwent an intervention within the first two days of hospitalization. The vast majority of patients had a colon resection, with only 38% requiring stomas. Of note, prompt intervention was associated with fewer stomas among patients with benign obstruction, but with more stomas among patients with malignant obstruction.
      Most guidelines and reviews that cover large bowel obstruction focus on patients with colorectal cancer,
      • Webster P.J.
      • Aldoori J.
      • Burke D.A.
      Optimal management of malignant left-sided large bowel obstruction: do international guidelines agree?.
      thought to be the most common etiology of large bowel obstruction.
      • Yeo H.L.
      • Lee S.W.
      Colorectal emergencies: review and controversies in the management of large bowel obstruction.
      However in our cohort, using a nationally representative and contemporary sample, 65% of cases were of benign etiology. Additionally, in our analysis, extracolonic malignancy was almost as common as colon cancer. In a recent series of large bowel obstruction cases from a single hospital in the US, extrinsic compression was also a substantial proportion of cases, reported in 29% of patients.
      • Capona R.
      • Hassab T.
      • Sapci I.
      • Aiello A.
      • et al.
      Surgical intervention for mechanical large bowel obstruction at a tertiary hospital: which patients receive a stoma and how often are they reversed?.
      Despite recognition of large bowel obstruction as an urgent surgical condition, few guidelines specifically address timing of surgery. Even in patients with a non-toxic appearance at presentation, rapid progression can result in perforation and/or ischemia of upstream bowel. The duration of cecal distension correlated more prominently with perforation than cecal diameter in a radiology series from 1985.
      • Johnson C.D.
      • Rice R.P.
      • Kelvin F.M.
      • Foster W.L.
      • Williford M.E.
      The radiologic evaluation of gross cecal distension: emphasis on cecal ileus.
      For malignant large bowel obstruction, urgent surgical intervention has been defined by the UK National Bowel Cancer Audit

      National Bowel Cancer Audit Methodology Supplemental Document 2019. January vol. 9, 2020. http://www.nboca.org.uk/content/uploads/2018/12/Methodology-Supplement-PDF.pdf. Accessed November 6, 2021.

      to occur “as soon as possible after resuscitation and usually within 24 h“ in contrast to “emergency” intervention, defined as “operation, usually within 2 h” We attempted to use these guidelines to define prompt intervention but could not do so because the NIS data do not provide timing of intervention in terms of hours since presentation, but instead provide calendar day difference between admission and intervention. Therefore, we defined “prompt” as “within two days” which could be the case even with a one calendar day difference depending on time of day when admitted and intervened. Recent studies of large bowel obstruction, particularly studies focusing on colonic stent placement versus surgery also demonstrate a general preference for expedited management. For example, Veld et al. reporting on a sample of Dutch patients with left sided malignant colonic obstruction, showed a median time to surgery of one day, even after excluding patients who had “emergency” surgery.
      • Veld J.V.
      • Amelung F.J.
      • Borstlap W.A.A.
      • et al.
      Comparison of decompressing stoma vs stent as a bridge to surgery for left-sided obstructive colon cancer.
      Median time to surgery was also one day in a French study comparing colonic stent placement to surgery.
      • Sabbagh
      • Browet F.
      • Diouf M.
      • et al.
      Is stenting as “a bridge to surgery” an oncologically safe strategy for the management of acute, left-sided, malignant, colonic obstruction? A comparative study with a propensity score analysis.
      In our population-based analysis, just under half of patients were treated within 2 days, with cancer patients slightly less likely to undergo prompt intervention.
      Predictors of prompt intervention included several sociodemographic factors such as race and income, with higher income patients more likely to be managed promptly. Among those with benign obstruction, Blacks and patients of unknown race were less likely to be managed promptly. Of note, although prompt intervention was not associated with mortality risk, in the benign obstruction group, Blacks had higher adjusted mortality risk. These associations warrant examination using other sources of data. Compelling clinical explanations for these demographic discrepancies are lacking, and there is no prior literature examining race- and income-associated disparities in the management of large bowel obstruction patients.
      Prompt intervention was less likely in patients with more comorbidities. This may reflect these patients’ need for preoperative optimization and risk assessments, or potentially reluctance of patients or families to consent to urgent surgery. Malignant obstructions managed beyond 2 days after admission resulted in fewer stoma procedures than those managed promptly, perhaps explained by colonic stent placement (which was less likely to be placed promptly). Colonic stents can decompress the colon and allow bowel preparation, optimizing conditions for an anastomosis. Patients managed in rural hospitals (10% of our cohort) were more likely to be managed promptly. Rural patients may have had fewer management options available (including stenting), as well as quicker assignment of an operating room, permitting expedited intervention. Additional research is needed with more clinical data that could explain the factors contributing to the timing of surgery. Explanations worth investigating include medical optimization (e.g., need for bowel decompression via nasogastric tube, fluid resuscitation), diagnostic dilemmas (suggested by endoscopy preceding the surgery), palliative care consultation or other multidisciplinary consultations, patient reluctance to consent, or surgeon/operating room availability. Delayed intervention in patients with extracolonic malignancy could be due to patient-physician deliberations on comfort measures or decompressive gastrostomy versus operative management.
      Several clinical scenarios clearly justify very prompt operative management in large bowel obstruction, such as signs of ischemic bowel (on laboratory or imaging tests) or peritonitis on examination. In these cases, operative intervention within hours may be critical to rescue these patients. However, such data were not available in NIS. Overall, patients with large bowel obstruction did not have worse mortality with delayed surgery or colonic stenting beyond the second day of hospitalization. In this large sample, inpatient mortality (6%) was somewhat lower compared to other series of large bowel obstruction. For example, a 9% mortality rate was reported in a population-based series of obstructing colon cancer by Mege et al.,
      • Mege D.
      • Manceau G.
      • Beyer L.
      • Bridoux V.
      • et al.
      Right-sided vs. left-sided obstructing colonic cancer: results of a multicenter study of the French Surgical Association in 2325 patients and literature review.
      and 14% in a smaller UK series of large bowel emergency surgery reported by Ng et al.
      • Ng H.J.
      • Yule M.
      • Twoon M.
      • et al.
      Current outcomes of emergency large bowel surgery.
      As would be expected, risk factors for mortality tracked with age and comorbidities. Patients with benign disease had the highest inpatient mortality, suggesting that cancer patients may be easier to salvage, possibly related to more gradual mechanical occlusion causing less acute presentation.
      While prompt intervention did not reduce the inpatient mortality risk, important benefits were avoiding an ostomy in patients with benign obstruction, greater likelihood of discharge to home and reduced length of stay, by 3 days, which exceeds the 1–2 days of inpatient stay saved by early intervention during the hospitalization. These benefits may reflect decreased post operative complications with prompt intervention.
      Our study has limitations related to its retrospective nature, episodic data, and use of administrative claims data. Crucial diagnostic information at presentation, such as laboratory and imaging results, are not available. Such data would enable more precise risk adjustment. For example, while all patients were admitted emergently, some patients may have had subacute obstructions with a lower risk of operative difficulties and therefore, post procedure mortality. Low-grade obstructions are less likely to be in the prompt intervention group, which may contribute to better relative outcomes of the “delayed” intervention group. Our finding that a large proportion of patients had a benign obstruction contrasts with existing single-center series. This could be partly due to incomplete diagnostic information available at discharge if, for example, pathology results were pending, which would result in designation of “benign”. However, given the LOS (median 10 days), misclassification on this account is unlikely. Our study suggests that benign colonic obstruction is understudied in the literature and warrants more attention. We chose not to evaluate postoperative complications because ICD codes in the claims do not distinguish between pre- and post-procedure events. Therefore we are unable to assess a broad array of outcomes.

      5. Conclusions

      Prompt intervention with colonic stent placement or surgery within the first 48 h of presentation was not associated with better mortality profile in a large U.S. population-based analysis. However, prompt intervention was associated with decreased length of stay and increased likelihood of discharge to home which may indicate better recovery with prompt intervention. Our results also suggests the need for further study of large bowel obstruction management and outcomes among females, Blacks and lower income patients.

      Funding

      Grant 1711 P WA942 through The Regents of the University of California , on behalf of Department of Surgery.

      Declaration of competing interest

      All authors for the manuscript entitled “Prompt Intervention in Large Bowel Obstruction Management: A Nationwide Inpatient Sample Analysis” declare no conflicts of interest.
      Specifically, Mary R. Kwaan, MD, MPH, Yuqi Wu, MPH, MPA, Yang Ren, MS.
      And Sudha Xirasagar, MBBS, PhD have no Direct (employment, stock ownership, grants, patents) or Indirect conflicts (honoraria, consultancies to sponsoring organizations, mutual fund ownership, paid expert testimony) related to our manuscript submission to The American Journal of Surgery.
      Dated: April 10, 2022.

      Appendix A. Supplementary data

      The following are the Supplementary data to this article:

      References

        • Kwaan M.R.
        • Ren Y.
        • Wu Y.
        • Xirasagar S.
        Colonic stent use by indication and patient outcomes: a Nationwide inpatient sample study.
        J Surg Res. 2021; 265: 168-179
        • Webster P.J.
        • Aldoori J.
        • Burke D.A.
        Optimal management of malignant left-sided large bowel obstruction: do international guidelines agree?.
        World J Emerg Surg. 2019; 14: 23
        • Yeo H.L.
        • Lee S.W.
        Colorectal emergencies: review and controversies in the management of large bowel obstruction.
        J Gastrointest Surg. 2013; 17: 2007-2012
        • Capona R.
        • Hassab T.
        • Sapci I.
        • Aiello A.
        • et al.
        Surgical intervention for mechanical large bowel obstruction at a tertiary hospital: which patients receive a stoma and how often are they reversed?.
        Am J Surg. 2021; 221: 594-597
        • Johnson C.D.
        • Rice R.P.
        • Kelvin F.M.
        • Foster W.L.
        • Williford M.E.
        The radiologic evaluation of gross cecal distension: emphasis on cecal ileus.
        AJR Am J Roentgenol. 1985; 145: 1211-1217
      1. National Bowel Cancer Audit Methodology Supplemental Document 2019. January vol. 9, 2020. http://www.nboca.org.uk/content/uploads/2018/12/Methodology-Supplement-PDF.pdf. Accessed November 6, 2021.

        • Veld J.V.
        • Amelung F.J.
        • Borstlap W.A.A.
        • et al.
        Comparison of decompressing stoma vs stent as a bridge to surgery for left-sided obstructive colon cancer.
        JAMA Surg. 2020; 155: 206-215
        • Sabbagh
        • Browet F.
        • Diouf M.
        • et al.
        Is stenting as “a bridge to surgery” an oncologically safe strategy for the management of acute, left-sided, malignant, colonic obstruction? A comparative study with a propensity score analysis.
        Ann Surg. 2013; 258: 107-115
        • Mege D.
        • Manceau G.
        • Beyer L.
        • Bridoux V.
        • et al.
        Right-sided vs. left-sided obstructing colonic cancer: results of a multicenter study of the French Surgical Association in 2325 patients and literature review.
        Int J Colorectal Dis. 2019; 34: 1021-1032
        • Ng H.J.
        • Yule M.
        • Twoon M.
        • et al.
        Current outcomes of emergency large bowel surgery.
        Ann R Coll Surg Engl. 2015; 97: 151-156