Boring Question: How useful are plain abdominal films for bowel obstruction?

In Clinical Questions by Jatin Kaicker3 Comments

Clinical Scenario

An otherwise healthy 65 year old comes into the emergency department with an 8 hour history of abdominal pain. Her last bowel movement was 3 days ago and she’s not sure if she has passed gas.  She endorses nausea but has not vomited. She has not taken her temperature but does not believe she has been febrile. She’s tried Tylenol for analgesia with little relief. She last ate 12 hours ago and has recently only prepared home cooked meals with no new foods. Her urination has been normal. She has had a number of surgeries including a hysterectomy (1990) and cholecystectomy (1995). There is no family history of colon cancer and she had a normal screening colonoscopy two years ago after an episode of rectal bleeding.

On physical examination, her vital signs were:

Temp: 37.6, HR of 90, RR:18, BP:128/80, O2: 99% on room air.

On exam, she looks uncomfortable moving around on the stretcher. Her abdomen looks slightly distended, you didn’t auscultate her bowel sounds (see why here) and she has diffuse tenderness of the abdomen but no peritoneal signs. She does not have flank percussion tenderness. The rest of her exam is unremarkable. You wonder:

 

Boring Question:  How useful are plain abdominal films for a patient with abdominal pain and suspected small bowel obstruction?

Background

Radiographs are also ordered in the emergency department for patients with acute abdominal pain. Flat and upright abdominal films, upright chest films and lateral decubitus views can be used to screen for bowel obstruction, constipation or free air (1). Plain abdominal films may demonstrate air fluid levels which may suggest mechanical obstruction. While these levels may suggest an obstruction, they are not pathognomic for such a condition (2). Plain abdominal x-rays also expose patients to about 35 times the radiation as a dose of a chest x-ray (3).

Search Strategy

Using PubMed, three separate searches were performed. These were:

1. “abdominal radiographs” AND “small bowel obstruction”

2. “plain radiographs AND “abdominal pain”

3. “abdominal x rays” AND “acute abdominal pain” AND “emergency department”

The resulting article titles and abstracts were screened with relevant articles reviewed. In addition, the textbook ‘Tintinalli’s Emergency Medicine’ was used. For investigations highlighted in review literature, the primary studies were assessed.

The Evidence

Taylor et al published a scientific article in 2013 that identified five studies, attempting to assess the usefulness of plain radiographs in diagnosing small bowel obstruction (4). Three of these investigations were prospective case studies with two retrospective case studies. All these investigations used the criteria of SBO on x-ray to be two or more air fluid levels in dilated loops of bowel (more than 2.5 cm) (4). The positive likelihood ratio was published to be 1.64 as the collective for these investigations.

A study by Van Randen et al in 2011 discussed a multi-center trial for patients with abdominal pain lasting between 2 hours and 5 days (2). Each patient was clinically assessed with physical examination and laboratory blood work, with subsequent supine abdominal and upright chest x-rays (2). This investigation found a positive predictive value of 54% for patients with a bowel obstruction with the clinical assessment and 61% after radiographs (2).

In the late 1990s, Suri et al conducted a small prospective study, comparing abdominal x-rays with ultrasound and CT to diagnose small bowel obstructions (3, 5). The calculated positive likelihood ratio from their findings is 1.54. Maglinte et al in 1996 also attempted to assess the value of CT and abdominal radiographs in patients with suspected small bowel obstruction (3, 6). For abdominal films, the positive likelihood ratio can be calculated to be 1.60 from their findings.

Bottom Line

From the literature assessed, plain abdominal films and chest radiographs have limited added diagnostic value for patients with abdominal pain and suspected small bowel obstruction.

 

 

Review by an Attending

Patients suspected of bowel obstruction are a heterogeneous group and include patients that may have ileus, gastroparesis or gastroenteritis. The gas pattern also varies due to severity, level and duration of the obstruction. It should be no surprise, therefore, that the radiographs can be normal in 1/3 of cases [especially in the setting of partial obstructions]. That said, many of these patients could be treated conservatively with 40-75% resolution.

Radiologists look at more than just the presence [and number] of air-fluid levels and dilated loops of bowel –for example “mean differential air-fluid levels”. Most studies use rad-reads in patients known to have bowel obstruction. So the published test characteristics reflect the best possible scenario. So I believe that an emergency-read plain film will probably perform even less favorably than you have described.

However, plain films may still have a role depending on factors such as resource availability. I would be reluctant to advocate for CT-everybody approach given the risk of radiation and the current global trend of spiraling health costs and ED overcrowding. It appears that ultrasound may now also emerging as a potential alternative to CT. I would advocate for a sit-down between EM, surgery and radiology to derive an algorithmic approach to which modality would best serve patient needs in your center.

– Nadim Lalani  MD FRCPC

 

References

1. Tintinalli’s Emergency Medicine-A Comprehensive Study Guide (2011). New York. McGraw Hill Companies Inc

2. Van Randen A, Lameris W, Luitse JSK, Gorzeman M, Hesselink EJ et al. (2011). The role of plain radiographs in patients with acute abdominal pain at the ED. The American Journal of Emergency Medicine. 29 (6). 582-589.

3. Smith JE, Hall EJ. (2009) The use of plain abdominal x rays in the emergency department. Emergency Medicine Journal. 26(3).160-3.

4. Taylor MR, Lalani N. (2013). Adult Small Bowel Obstruction. Academic Emergency Medicine. 20(6). 528-544.

5. Suri S, Gupta S, Sudhakar PJ et al. (1999) Comparative evaluation of plain films, ultrasound and CT in the diagnosis of intestinal obstruction. Acta Radiol 40(4). 422-8.

6. Maglinte DD, Reyes BL, Harmon BH et al (1996). Reliability and role of plain film radiography and CT in the diagnosis of small-bowel obstruction. AJR Am J Roentgenol 167(6). 1451-5.

Reviewed by Eve Purdy

 

(Visited 369 times, 1 visits today)
BoringEM
BoringEM has been 'bringing the boring' to emergency medicine since 2012. In 2016 this Canadian blog brought its content to CanadiEM.
BoringEM
- 2 days ago
  • More useful perhaps, is what is the negative predictive value for SBO using plain films Ie, if there are no air fluid levels, no distended loops, can you say they don’t have an obstruction?

  • Rob Woods

    The pooled negative predictive value from the Taylor & Lalani study is 0.43. So it all depends on your pre-test probability. Plot your pre-test probability on the Fagan normogram, then intersect 0.43 and see how much it changes your post-test probability. You also need to ask yourself, at what probability of disease am I happy saying, ‘you can go home’? 15%, 10%, 5%, 2%, 1%? Depends on the disease.

    In some ways it is similar to d-Dimer. If the pre-test probability is low, say 10%, and the films are negative, the post-test probability is around 3%. If that is below your threshold for further tests, then it will be useful for you. So at low pre-test probabilities, plain films are potentially useful to rule out SBO.

    If the pre-test probability is higher: 25%, 50% 75%, your post-test probabilities after a NEGATIVE abdo xray are: 10%, 25% and 50% respectively. Given the potential seriousness of missing bowel obstruction, none of those scenarios really leaves me feeling comfortable sending the patient home without further tests. I’d still want a definitive test (CT or US).

    So the next time Radiology tries to prevent you from ordering a CT abdo for SBO because you didn’t order plain films first, tell them your pre-test probability is too high for a diagnostic test with a negative predictive value of 0.43.

  • Rob Woods

    Sorry I used the term wrong…0.43 is the Negative LIKELIHOOD RATIO, not the Negative PREDICTIVE VALUE.