Paper #1

Accuracy of Abdominal Ultrasound for the Diagnosis of Small Bowel Obstruction in the Emergency Department

Frasure S, et al. World J Emerg Med. 2018;9(4):267-271. doi:10.5847/wjem.j.1920-8642.2018.04.005

Full text: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6117534/

• Introduction and background information
  • Patients frequently present to the ED with signs and symptoms concerning for small bowel obstruction (SBO)
  • Literature estimates that 13 million patients present to the ED annually for abdominal pain and about 2% are diagnosed with an SBO¹
  • CT is currently the gold standard diagnostic study with ultrasound gaining popularity
  • These authors cite a recent systematic review of 11 studies by Gottlieb et al from 2018 which found 92.4% sensitivity and 96.6% specificity for US in SBO; but also note only 3 studies were conducted in the ED along with lack of variety of operators²

• What was the primary objective of this study?
  • To evaluate the accuracy of US by attendings, fellows, residents and PAs compared to abdominal CT in patients with suspected SBO 

• Methods and study design
  • Retrospective, single center cohort study at large tertiary care center with >65,000 visits per year
    • Large cancer institute affiliation so many ED patients with active malignancy
  • Reviewed ultrasound database for suspected SBO
  • Protocol:
    • All participants were taught how to perform POCUS for abdominal pathology
    • All 4 quadrants scanned with curvilinear probe 
    • Criteria included dilated (>2.5 cm) fluid-filled loops of bowel with abnormal “to-and-fro” peristalsis
    • Use of US was at the discretion of the provider
    • Either had abdominal CT from same visit read as SBO by radiologist or SBO was discharge diagnosis

• Results
  • 64 patients included
    • 9 did not undergo CT and 8 had indeterminate ultrasound findings and were excluded
    • Total n = 47
  • Half of patients had prior SBO and 62.5% had history of active malignancy
  • 50% had SBO identified on CT scan
  • 12% taken to OR, remainder treated conservatively 
  • Sensitivity of 93.8% and specificity of 93.3% for US 

• Discussion
  • Authors note that historically the “algorithm” for evaluation of an SBO would be abdominal plain film → CT
  • Jang et al from 2011 found poor sen/spec of plain films 46.2% and 66.7% respectively. They found similar sen/spec for US 
  • According to the authors this is the first study which has included PAs
  • They propose that low risk patients (history SBO with well controlled pain) can be admitted for conserative management while high risk patients (toxic-appearing or inconclusive ultrasounds) should undergo CT

• Were there limitations?
  • Sample size small and data pulled from a retrospective chart review
  • Large number of patients with active malignancy which limits generalizability
  • Use of US was at the provider’s discretion which introduces selection bias
  • Difficult to find a transition point which means a surgeon may want a CT anyway for operative planning

• Take-home points
  • These authors sought to evaluate the accuracy of US for SBO with a variety of EM providers, including PAs
  • Retrospective data showed similar sensitivity and specificity with prior data
  • Significant limitations including small sample size and biases, further prospective studies are needed

References

1. Taylor M, et al. Adult Small Bowel Obstruction. Academic Emergency Medicine 2013; 20: 528– 544. https://doi.org/10.1111/acem.12150.
2. Gottlieb M, et al. Utilization of ultrasound for the evaluation of small bowel obstruction: A systematic review and meta-analysis. Am J Emerg Med. 2018;36(2):234-242. doi:10.1016/j.ajem.2017.07.085
3. Jang TB, Schindler D, Kaji AHBedside ultrasonography for the detection of small bowel obstruction in the emergency departmentEmergency Medicine Journal 2011;28:676-678.

Paper #2

A Prospective, Multicenter Evaluation of Point-of-care Ultrasound for Small-bowel Obstruction in the Emergency Department

Becker BA, et al. Acad Emerg Med. 2019;26(8):921-930. doi:10.1111/acem.13713

Full text: https://pubmed.ncbi.nlm.nih.gov/30762916/ 

• Information and background
  • 300,000 adults are hospitalized annually for SBO
  • Plain X-rays for diagnosis are falling out of favor due to poor accuracy and often inconclusive results in favor of CT, which is more sensitive and specific
  •   POCUS may accurately diagnose SBO, decrease time to diagnosis and expediting necessary consultations
  •   However, most studies that exist looking at POCUS for SBO consist of small, single-center studies

• Goals of the investigation
  • Conduct a multi-center evaluation of the accuracy or EM-performed POCUS for the diagnosis of SBO compared to CT
  • Compare POCUS interpretation to that of emergency ultrasound fellowship-trained experts
  • Assess the roles of the specific POCUS parameters in confirming the diagnosis of SBO

• Methods
  • Study design and setting
    • Prospective multi-center observational study which took place between July 2014-May 2017
    • Goal-directed POCUS of the abdomen for SBO evaluation
    • POCUS findings were interpreted at bedside by a physician who was blinded to lab and additional imaging, and retrospectively by an expert reviewer after deidentification and compared to abdominal CT
    • Facilities that enrolled: 2 suburban academic community hospitals, and an urban, university-based tertiary care center. Each has 3-year EM residency, US fellowships
  •   Selection of participants
    • Inclusion criteria:
      • At least 18 years old
      • Able to provide consent in English
      • Not pregnant
      • Not yet undergone radiology imaging
      • Presented with symptoms concerning for SBO based on clinical assessment
    • Exclusion criteria:
      • Did not receive a CT
  • Data collection and measurements
    • Clinical features recorded at time of POCUS: diarrhea within 24 hours, vomiting, duration of symptoms, last bowel movement, presence of diffuse abdominal pain
    •   Follow up data: demographics, discharge diagnosis, op reports, abdominal XR imaging, CT results
  • US technique and interpretation
    • Supine position
    • Standardized protocol: curvilinear probe (1-5MHz)
    • Systemic evaluation of the entire abdomen with dedicated views of the RUQ, LUQ, RLQ, LLQ, adjusting probe orientation with respect to the body, one clip in each location
    • Maximum bowel diameter (still image)
    • Transition point (video clip)
    • 4 required parameters:
      • Small-bowel dilation: diameter >25mm outer wall-outer wall
      • Abnormal peristalsis: “to and fro” shuttling or swirling
      • Small-bowel wall edema: present if plicae circulares projected into bowel lumen = keyboard sign (however no consensus of max thickness in normal bowel wall)
      • Intraperitoneal free fluid
    • Reviewed by attending ED physician (with and without US fellowship training), US fellow, or upper-level resident
    • Training: 30 min lecture on SBO scanning technique, brief hands on practice on normal bowel
    • POCUS not used in clinical decision making
    • POCUS classified as positive, negative, or indeterminate for SBO – based on abnormal peristalsis or dilation >25mm, additional 3 parameters used to augment impression
• Outcomes
  • Primary outcome: POCUS mediated dx confirmed by CT
  • Secondary outcomes: diagnosis of SBO by blinded expert interpretation, diagnostic accuracy of parameters
  • Indeterminant considered positive to pursue equivocal results with further work up, also non-committal CT considered to be positive
• Data analysis
  • Primary analysis was repeated with 1. Reclassification of all indeterminate POCUS as negative; 2. Removal of those with indeterminate CT scans
  • Predicted sensitivity 90%, specificity 91%, incidence of SBO 40%
• Results
  • Characteristics of study subjects
    • 232 initially enrolled, ultimately 217 patients at 3 sites, (only 3, 1.4% from third site). Median age 55, 52% female
    • Prevalence SBO 42.9%
  • Main results
    • 11 false-negative, 57 false-positive
    • Overall sensitivity 88%, spec 54%, +LR 1.92, -LR 0.22
    • Trainees (fellow/res) sensitive 91%, specific 51%
    • Attending sensitive 85%, specific 61%
  • Secondary results
    • Most sensitive parameters:  Bowel dilation 87%, Abnormal peristalsis 82%
    • Most specific parameters: transition point 98%, free fluid 93%, bowel wall edema 93%
    • Significant association between all parameters
    • CT interpretation: 6 false negative, 13 false positive; sensitive 93%, specific 90%
• Discussion
  • Overall EM POCUS relatively sensitive (88%) for SBO, less specific (54%), attending and trainees perform similarly
  • POCUS performed by fellowship trained physicians were more accurate
  • 2 prior studies showed greater sensitives and specificities with similar SBO prevalence. Why?
    • One study: POCUS training for participants (required 5+ prior SBO studies); other study: 6 hours of training
  • Speculates that SBO diagnosis hinges on recognizing a constellation of patterns and characteristic appearance, this may be dependent on prior POCUS experience exposure to differentiate between normal and abnormal cases, which leads to further discussion on the operator dependence of ultrasound. Users likely need adequate education (both didactic and hands on)
  • Future research: evaluate methods of education and assessing for competency
• Limitations
  • Observation design/convenience sample leads to bias
  • Low enrollment of one site
  • Variable US experience of physicians enrolling
  • Conducted at centers not currently using POCUS for SBO
  • Multiple different machines
  • Tried to standardize technique though there was not a specific stepwise algorithm
  • Compared to prior study with use of “gassy” patients with reported 84% specificity, recalculated still showed 72% in this study
• Conclusion
  • POCUS for SBO is moderately sensitive, less specific when performed by a diverse groups of EM physicians
  • Interpretation is significantly more accurate when read by US-fellowship trained physician who are familiar with the appearance SBO on POCUS

Paper #3

Clinical Ultrasound Is Safe and Highly Specific for Acute Appendicitis in Moderate to High Pre-test Probability Patients

Corson-Knowles Daniel, et al. The Western Journal of Emergency Medicine. 2018;19(3). 460-464. doi:10.5811/westjem.2018.1.36891

Full text: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5942008/

• Introduction and background information
  • Appendicitis is a common admitting diagnosis in the ED, accounting for a total of 1 million patient-days of admissions per year¹
  • For diagnosis CT scan is used in adults, with ultrasound being preferred in children due to lack of ionizing radiation
  • The authors note variability in accuracy of ultrasound between community and academic centers; citing a 2017 study of 441 community pediatric appendix ultrasounds, which found a 74% nondiagnostic rate² 
  • Prior data has shown high accuracy for highly trained sonographers
  • The goal was to determine if a heterogenous group of EPs could diagnose appendicitis on ultrasound

• What was the primary objective of this study?
  • To determine if EP sonographers could diagnose acute appendicitis using a combination of risk assessment, ultrasound and self-assessment

• Methods and study design
  • Prospective observational study of adult and pediatric patients presenting with concern for appendicitis
  • 3 large urban academic EDs (two adult, one pediatric) from 2014 – 2016
  • Inclusion criteria:
    • Suspected appendicitis patients who underwent EP ultrasound
    • Children and pregnant patients
  • Exclusion
    • If US images were obtained after radiology imaging
    • Missing information in data collection
  • Protocol:
    • Prior to US, EP was asked to score 10-point pre-test probability of appendicitis on a VAS based on gestalt
    • US performed at the discretion of the EP
    • After US, EP recorded presumed diagnosis followed by post-test 10-point VAS score based on confidence in interpretation
  • Sonographers included residents, ultrasound fellows and attendings
  • The standard for diagnosis was either surgical pathology or chart review at both discharge and 1 week post-discharge

• Results
  • 122 patients underwent ultrasound, narrowed 76 after excluding missing data and low pretest probability
  • Of the 76 patients, 36.8% had acute appendicitis
    • 24 of these were pediatric
    • 2 pregnant patients underwent US but were low probability
  • Sensitivity and specificity of EP performed ultrasound were 42.8% and 97.9% respectively
  • 16 false negative scans, interpreted as indeterminate, all had appendicitis on CT
  • 33 different sonographers performed the imaging
    • 52.6% residents who identified 41.7% of true positives with high inter-rater reliability

• Discussion
  • Authors note that a heterogeneous group of EPs can diagnose appendicitis with accuracy given a high pretest probability
  • Higher confidence in the ultrasound image acquisition and interpretation is a strong predictor for accuracy
  • Poor sensitivity so should not be used for ruling out disease

• Were there limitations?
  • Small sample size and convenience sampling which lowers accuracy
  • Spectrum bias – variability of test performance across different mixes of patient population due to broad inclusion criteria
  • Sonographers were not blinded to history or physical 
  • US was performed at the discretion of the EP which could lead to significant variability
  • No distinction between adult and pediatric data

• Take-home points
  • Authors sought to find accuracy of US for appendicitis using risk stratification
  • Prospectic convenience sampling showed high spec but low sensitivity
  • POCUS would be helpful in those with high pretest probability, should be avoided in trying to rule out appendicitis 

References

1. Craig, Sandy. “Appendicitis.” Appendicitis: Epidemiology, 12 Nov. 2019, emedicine.medscape.com/article/773895-overview#a6
2. Alter SM, Walsh B, Lenehan PJ, et al. Ultrasound for diagnosis of appendicitis in a community hospital emergency department has a high rate of nondiagnostic studies. J Emerg Med. 2017;52(6):833-8.