The critical areas of improvement in treating acute adhesive small bowel obstruction (ASBO) are the early diagnosis of intestinal ischemia, optimal surgery timing, and the duration of non-operative management (NOM). Currently, the assessment of the impact of these factors on patient survival is a topic of scientific discussion [1-4], and the optimal surgical management for ASBO has not been defined [5]. The issue of intestinal ischemia preoperative diagnosis is only partially solved; therefore, in a significant proportion of patients with strangulated bowel obstruction, NOM is performed excessively long [6, 7]. In non-strangulated bowel obstruction, sometimes early surgery is performed because it is impossible to exclude bowel ischemia [6, 8]. Although it is impossible to identify the types of ASBO, the time to surgery presented in the Russian national clinical guidelines (RNCGs) is very specific: 2 h in strangulated bowel obstruction and 12 h in non-strangulated obstruction. Failure to adhere to these timeframes is expected to be associated with poorer treatment outcomes. However, according to Matsushima et al. [9], mortality and morbidity in intestinal ischemia are related to delayed surgical intervention for more than 24 h, and the concept of prolonged NOM, adopted worldwide in the absence of intestinal ischemia, suggests non-surgical treatment for 3–5 days [10]. The PubMed database contains hundreds of relevant publications on optimal approaches in ASBO, including optimal timing of surgery and NOM, the effectiveness of diagnostic and therapeutic methods used, and statistics on complications and outcomes of suboptimal surgical management in ASBO, which are not available in domestic publications.

The objective of the study was to investigate the mortality factors of patients with ASBO.

Materials and methods

Between May 2017 and December 2019, a multicenter retrospective study included 4 city clinical hospitals in Moscow. The time frame for hospitals participating in the study ranged from 4 to 29 months. The study was approved by the Local Ethics Committee of the Pirogov Russian National Research Medical University. All study participants provided informed consent to be included in the study and to have their personal data processed.

Inclusion criteria: ASBO confirmed clinically, by ultrasound, radiography, and/or abdominal CT scan.

Exclusion criteria: early adhesive obstruction, non-adhesive small bowel obstruction, incarcerated hernia, large bowel obstruction, adynamic ileus, bowel obstruction due to abdominal cancer.

A total of 167 consecutive patients with small bowel obstruction were initially recruited for the study, ranging from 10 to 69 patients from each site.

After excluding patients who did not meet the criteria, 143 (85.6%) patients with AASBO were included in the study.

To analyze the role of obstruction type (strangulated or non-strangulated), the treatment type (surgery or NOM), and the timing of surgery (less or more than 12 h), the following main clinical groups were formed:

- early surgery (ES) — up to 12 h after admission;

- late surgery (LS) — more than 12 h after admission;

- NOM.

The surgical groups (ES and LS) were divided into strangulated (SES and SLS) and non-strangulated obstruction (NES and NLS) subgroups.

After analysis by demographics, condition severity (APACHE II and ASA) and features (number of previous surgeries, number of ASBO episodes, duration of the present episode, rate of intestinal ischemia), examination methods (laboratory data, ultrasound examination, radiography, enterography, MSCT scan) and treatment (NOM duration, nasogastric decompression, surgical technique — laparoscopic adhesiolysis [LA] or laparotomy) the formed groups were compared by the length of hospital stay and mortality.

To eliminate group heterogeneity, the second stage analyzed mortality factors and the daily probability of patients' discharge for outpatient treatment.

To assess the significance of differences in the mean values of quantitative variables between groups, we used the ANOVA test; to determine the significance of differences in the distributions of categorical variables, we used the χ² test. The exception was paired tests for categorical variables where the value of at least one cell in the contingency table was less than 5; in this case, Fisher's exact test was used.

The survival probabilities (mortality factors) were assessed using the Kaplan-Meier method. A competing risk methodology was used to eliminate heterogeneity between the surgery and NOM groups regarding mortality and time of hospital discharge. Cumulative incidence curves of mortality and daily discharge were plotted. In all cases, at a p-value less than 0.05, the groups were considered statistically significantly different from each other on the relevant characteristic at the 5% significance level.

Results

Surgery was performed in 66 (46.2%) patients: 39 (27.3%) patients with SBO and 27 (18.9%) patients with NBO. NOM was effective in 77 (53.8%) patients (Fig. 1).

Fig. 1. Between-group distribution of patients.

The chart shows that 24 (61.5%) patients with strangulated obstruction underwent surgery before 12 h, and 15 (38.5%) patients underwent surgery more than 12 h after hospital admission. In non-strangulated obstruction, 12 (44.4%) patients had surgery in less than 12 h and 15 (55.6%) patients in more than 12 h after hospital admission. The characteristic of the main groups is presented in Table 1.

Table 1. Characteristics of patients

Statistically significant differences between the groups were established using paired tests, correcting the levels obtained for multiple comparisons (Table 2).

Table 2. Significance of paired tests in between-group comparisons (with Holm correction)

The mean time to non-surgical resolution of the obstruction was 19.7±17.4 h (median 14 h): the obstruction resolved less than in 12 h only in 34 (44.2%) patients, after 24 h in 21 (27.3%) patients, after 36 h in 12 (15.9%) patients, after 48 h in 3 (3.9%) patients.

The average time to surgery for strangulated obstruction was 11 h (2; 75): 6.3±3.9 h in the SES group, 33.6±29.9 h in the SLS group, median 20.0 h (13; 121) after hospital admission (p<0.001). Within the time window recommended by RNCGs for SBO (2 h), 4 (10.3%) patients had surgery; 15 (38.4%) patients were operated on later than 12 h, and 7 (17.9%) had surgery one day after hospital admission. Patients with surgery after 24 h were in stable condition, with a maximum APACHE II score of 13.

In non-strangulated obstruction patients, the average time to surgery was 18 h (4; 48) after hospital admission: 7.5±3.2 h in ES, 27.6±11.5 h in LS. Seven patients were operated on after 24 h, including two patients with surgery after 48 h. All patients were stable.

Bowel resection was required in 6 (4.2%) patients, including 4 (2.8%) cases due to irreversible ischemic changes. They were patients aged 67 to 83 years with a disease duration of 10 to 120 h at the time of hospital admission. They underwent surgery between 1 and 45 h after hospital admission. Of these, one patient aged 78 died (APACHE II score 29, ASA score 4). He was admitted to the hospital 96 h from the onset and operated on 1 h after admission (ES). Another 2 (1.4%) patients with non-strangulated obstruction had bowel resection due to surgery-related injury. The bowel resection rate did not differ between the groups: 3 (8.3%) in ES and 3 (10.0%) in LS (p=0.815), including 2 (8.3%) in SES vs. 2 (13.3%) in SLS (p=0.617) and 1 bowel resection in each of NES (8.3%) and NLS (6.7%).

Complications

Surgery-related complications were reported in 15 (22.7%) patients; 9 (13.6%) had complications over grade 3A, according to the Clavien-Dindo classification. The complication rate did not differ between surgical groups (p=0.835). Iatrogenic bowel perforation occurred in 4 (8.3%) patients (only in LA). In 2 (4.1%) patients, perforation was recognized immediately during the primary operation, and in 2 (4.1%) patients, during the reoperation on day 2. Two (4.2%) of them (one in each group) underwent bowel resection (in the absence of initial ischemic changes).

Reoperation after LA was performed in another two patients due to unresolved strangulation (on day 2) and abdominal abscess (on day 7). Thus, 4 (8.3%) patients had reoperation after LA: 2 reoperations in each ES and LS group (p=0.851). There were no reoperations after primary open surgeries (n=17).

The length of hospital stay in the NOM group (3.4±2.3 days) was shorter than in the surgery group: ES — 6.2±5.7 days, LS — 8.0±4.2 days (p<0.001).

Mortality analysis

The overall mortality rate was 7%. All patients (n=10) died after surgical treatment (Table 3); thus, the postoperative mortality was 15.2%.

Table 3. Analysis of mortality in acute adhesive small bowel obstruction

Note. * — bowel resection due to gangrene; ** — bowel resection due to surgery-related injury.

No effect of preoperative duration on mortality was found (p=0.287), although absolute mortality rates were higher in the ES group than in LS, with 7 (19.4%) and 3 (10.0%) patients, respectively (see Table 1, 2).

No difference in mortality was observed for strangulated obstruction and non-strangulated obstruction, with 5 (12.8%) and 5 (18.5%) patients, respectively (p=0.7). The high surgical activity in non-strangulated obstruction in the older age group (>80 years) is noteworthy: 8 (61.5%) out of 13 patients were operated on in 4–12 h after hospital admission, and 4 (50.0%) of them died.

All patients who had surgery after 24 h (n=14) survived.

There was no difference in mortality in the SES and SLS subgroups: 3 (12.5%) patients versus 2 (13.3%) patients, respectively (p=0.940). No statistically significant differences in mortality in the surgical groups were observed in non-strangulated obstruction, despite the advantage in absolute values in LS: 4 (33.3%) versus 1 (6.7%) (p=0.76).

In the Kaplan-Meier analysis (Table 4, Fig. 2, 3) surgery compared with NOM increased the risk of death by 10 days of inpatient treatment: 15.9% (95% CI: 15.5%; 16.3%] and 0.0% [95% CI: 0.0%; 0.0%] (p=0.001) (Fig. 1) and reduced the early hospital discharge rate by 71.4% [95% CI: 70.8%; 72.1%] 98.7% [95% DI: 98.6%; 98.7%] (Fig. 2). No effect of surgery timing (before or after 12 h) on adverse outcome was found for both strangulated obstruction (13.0% [95% CI: 12.0%; 14.1%] and 16.7% [95% CI: 14.2%; 19.2%], p=0.788) and non-strangulated obstruction (33.3% [95% CI: 29.4%; 37.3%] and 6.2% [95% CI: 5.4%; 7.1%], p=0.061) in which late surgery had advantages, although no statistically significant difference was observed. For surgical treatment, neither the intestinal ischemia (14.3% [95% CI: 13.6%; 15.0%] and 17.9% [95% CI: 16.8%;18.9%], p=0.613), nor bowel resection 33.3% [95% CI: 24.4%; 42.3%] and 14.0% [95% CI: 13.6%; 14.5%] (p=0.187) did not increase the risk of death.

Table 4. Overall mortality and likelihood of discharge in 5 and 10 days after surgery

Fig. 2. Cumulative in-hospital mortality in surgical and conservative approach.

Fig. 3. Cumulative discharge probability in surgical and conservative approach.

Discussion

Mortality factors are a critical area of research.

Surgery and mortality

We observed the lowest mortality rates for non-surgical treatment of ASBO. Also, the adverse outcome rate was incomparably higher with surgery. The unexpectedly high mortality in non-strangulated obstruction when unreasonably early (4–12 h) surgery was performed in elderly patients should be emphasized. Most studies in elderly patients without established intestinal ischemia recommend adhering to long-term NOM because the complication and mortality rate with surgery is too high [2, 11, 12]. We obtained similar results. Overall, it is likely that high surgical activity in non-strangulated obstruction was the main factor related to the high mortality in our study. In a similar ASBO study [11] with a surgery rate of 32.5% (46% in our study), a significantly lower (3.9%) overall mortality was reported with similar postoperative mortality. Interestingly, in 2018, 10 constituent entities of the Russian Federation did not use NOM at all in ASBO [13].

Non-surgical treatment

Our findings indicate that the 12-hour NOM recommended by the RNCGs is insufficient. During this period, the obstruction resolved in a smaller percentage of our patients. It is reported that by the end of day 1 of non-surgical treatment, ASBO resolves in 46% of patients [14], and after 3 days, it resolves in 64–75.1% of patients [6, 15, 16]. A prerequisite for the use of NOM is the patient's stable condition. "If the patient remains clinically stable, the non-surgical approach can be significantly prolonged as long as nutrition can be provided" [17]. Criteria for an "unstable patient" are proposed: pH<7.2, body temperature less than 35 °C, BE less than –8 mmol/L, laboratory or clinical signs of coagulopathy, and any symptoms of sepsis or septic shock [18]. Our study results show that surgery performed after 24–48 h in stable patients at least does not worsen the outcomes. All patients, who had surgery after 24 h, survived.

Non-surgical treatment discontinuation criteria

Criteria for NOM discontinuation are a subject of debate. As an indication for surgery on day 3 (!) of non-surgical treatment, we consider discharge more than 500 mL through the nasogastric tube.2 The Italian consensus cites suspected intra-abdominal complications, high lactate level and leukocytosis (over 18∙109/L), body temperature above 38.5 °C, and an increase in creatinine (doubling the baseline) as criteria for NOM discontinuation [3]. No passage of oral water-soluble contrast agent in the colon within 24–36 h after intake is considered an indication for surgery [2].

Diagnosis of ischemia

It is generally accepted that clinical factors are not definitive in the diagnosis of bowel ischemia, even with an experienced surgeon [19, 20]. The scope of examination of patients in our study cannot be considered complete, but we can only partially count on the diagnosis improvement through laboratory tests. On the one hand, fever, leukocytosis, and hyperlactatemia are presented as reliable criteria for bowel ischemia in the publications [15, 21], and most surgical scientific communities [19, 20, 22] recommend considering these tests when choosing surgical strategy. On the other hand, it is indicated that the diagnostic value of clinical data and tests does not exceed 40–50% [19], and several studies [23, 24] did not establish the value of routine laboratory tests. Moreover, in the diagnosis of mesenteric ischemia, "routine laboratory tests, including lactate levels, reflect the dynamics of the disease and should not be used for diagnostic purposes" (!) [3]. "How many patients were erroneously operated on by doctors who relied on these test results?" asked J. Derikx et al. [25]. We had a higher number of patients with elevated lactate levels in the NES group than in the NLS group (p=0.026), which may have been a factor in the unnecessarily early surgery.

The only recognized objective method for diagnosis of bowel ischemia is CT scans [2, 3, 20] with a diagnostic value of 67–90% [26, 27]. Although in our study, the effectiveness of CT scan for bowel ischemia diagnosis was 50%, large population studies reported a slow decrease in mortality primarily due to the routine use of CT scans [28], and, in our opinion, RNCRs wrongly positioned CT scan as an additional rather than primary method of bowel ischemia diagnosis.

Bowel ischemia and mortality

Due to early diagnosis challenges, 15 to 30% of patients with strangulated obstruction undergo surgery later than 24 h after admission [5, 7, 20, 23], which is consistent with our results (17.9%). However, we operated on 44% of patients without strangulation in less than 12 h after hospital admission. It appeared that mortality in strangulated and non-strangulated obstruction was similar, and bowel ischemia did not increase mortality. Moreover, delayed surgery in stable patients with bowel ischemia was not associated with an increase in mortality, and we have no reason to state that surgery performed in 2 or 12 h improves treatment outcomes. However, the minimal time allotted for the decision to operate undoubtedly reduces the chances of a significant number of patients without ischemia for non-surgical resolution of ASBO, and surgery increases the risk of adverse outcomes.

Thus, our study and data from the literature suggest the need to consider increasing the duration of non-surgical treatment in stable patients with no apparent signs of strangulation, even though in some patients, the ischemia may not be recognized on the first day. When comparing the mortality risks from surgery and unrecognized strangulation, the choice should be made in favor of a strictly reasoned approach based on CT scans. An unreasonable increase in surgical activity leads to an increased risk of adverse outcomes and length of hospital stay in ASBO.

The authors declare no conflicts of interest.