How Good Are Our Tests for Acute Pancreatitis?

A recent cross-sectional pediatric study (SH Orkin et al. J Pediatr 2019; 213: 143-8),  with a prospective clinical database, provides data on children presenting with acute pancreatitis, n=112 (2013-16).

Acute pancreatitis (AP): requires at least 2 of 3 criteria:

  1. Abdominal pain consistent with AP
  2. Serum amylase and/or lipase activity at least 3 times ULN
  3. Imaging findings compatible with AP

Key points:

  • Among AP patients who had a lipase level, the sensitivity was 95% whereas the sensitivity for amylase was 39%.
  • Among AP patients who had an ultrasound, the sensitivity was 52%.  In those with either CT or MRI, the sensitivity was 78%.
  • In this cohort, 5.4% did not meet diagnostic criteria based on biochemical elevation (amylase or lipase) and instead relied on imaging along with signs/symptoms.

The authors note that lipase has a delayed peak and longer duration of elevation with AP.   Amylase normalizes more rapidly.

My take: This study reinforces the view that an elevated lipase is more sensitive than amylase and that imaging (especially ultrasound) is frequently normal in AP.

Related blog posts:

Island Ford Park, Chattahoochee River


ER Evaluations of Visits Related to Inflammatory Bowel Disease

A short communication (C Pant et al. JPGN 2015; 61: 282-84, ed 267-8) provides some insight into what is happening in the ER when patient with inflammatory bowel disease are evaluated.

The authors analyzed a national database, the Healthcare Cost and Utilizationa Porject Nationwide Emergency Department Sample (NEDS), from 2006-2010.

Key findings:

  • ED visits for children 5-19 years of age increased from 14,527 (2006) –>18,193 (2010)
  • Frequency of computerized tomography (CT) imaging increased 80.43% (P<0.01) for Crohn’s disease and 59.26% (P<0.01) for ulcerative colitis
  • Overall rate of hospital admissions decreased by 14.32% (P<0.01)

The associated editorial by Jennifer Dotson and Michael Kappelman explain the limitations of relying on data that are derived mainly for billing rather than research and comment on the “alarming increase” in CT  usage due to the risk of radiation exposure and the potential alternatives (eg. ultrasound, MRE).  With regard to limitations, they likely included inaccurate and incomplete entries.  In addition, the database included 950 hospitals, but may have under-sampled large freestanding children’s hospitals.  This could skew the data.  Finally, the data uses “visit-level” data rather than “patient-level” data; thus, one ‘cannot distinguish between a single patient who is seen on three separate visits or three unique patients.’

My take:  There has been increasing use of ER visits and CT scanning for pediatric patients with IBD.  One of my colleagues in town (Cary Sauer) has a humorous slide of a CT scanner as the entrance door to the ER.  The message from this study indicates that we should be working on changing practice by working with our ER colleagues along with radiologists to minimize CT scans in favor of more gently imaging.

Related blog posts:

Atlanta Botanical Gardens, Bruce Munro exhibit

Atlanta Botanical Gardens, Bruce Munro exhibit

Imaging in NAFLD -Don’t Rely on Ultrasound

A recent article (Clin Gastroenterol Hepatol 2014; 12: 765-73) notes that ultrasound is not accurate for diagnosing NAFLD:

The link to the article:, video abstract: (from Jeff Schwimmer’s twitter feed) ,and an excerpt from the article’s abstract:


We analyzed 9 studies comprising 610 children; 4 studies assessed ultrasonography and 5 studies assessed magnetic resonance imaging (MRI). Ultrasonography was used in the diagnosis of fatty liver with positive predictive values of 47% to 62%. There was not a consistent relationship between ultrasound steatosis score and the reference measurement of hepatic steatosis. Liver fat as measurements by MRI or by spectroscopy varied with the methodologies used. Liver fat measurements by MRI correlated with results from histologic analyses, but sample size did not allow for an assessment of diagnostic accuracy.


Available evidence does not support the use of ultrasonography for the diagnosis or grading of fatty liver in children. Although MRI is a promising approach, the data are insufficient to make evidence-based recommendations regarding its use in children for the assessment of hepatic steatosis.

Related blog posts:

Ultrasound Unreliable to Exclude Fatty Liver

More information on the sensitivity of ultrasonography for detecting fatty liver is available in the setting of living-related liver transplantation (Transplantation 2013; 95: DOI: 10.1097/TP.0b013e31828d1588).

In this study the authors retrospectively examined the degree of steatosis from 492 living liver donors who had normal ultrasounds and normal aminotransferase levels. The median age of the donors was 30.1 year and the median BMI was 22.4 kg/meter-squared.

Background: According to the authors, if liver histology shows severe macrosteatosis (>60%), transplantation is canceled.  Furthermore, in cases of moderate macrosteatosis (30-59%), the risks/benefits need to be considered on an individual basis due to increased risk of mortality; Spitzer et al (reference below) demonstrated that macrovesicular steatosis >30% was an independent predictor of reduced 1-year graft survival.  In addition, a previous report has indicated that both macrosteatosis and microsteatosis had similar impacts on postoperative  liver function.


  • 3 (0.6%) had severe total steatosis, moderate or greater steatosis was diagnosed in 4 (0.8%) for macrosteatosis, in 26 (5.3%) for microsteatosis, and 56 (11.4%) for total steatosis.
  • There were two identified risk factors BMI >23 kg/meter-squared and triglycerides >88 mg/dL.  Individuals with both risk factors had a 28.6% prevalence of moderate or greater degree of total steatosis compared with 6.6% with no risk factors.  In these individuals, a liver biopsy may be worthwhile.

Why this study matters for the non-transplant physician:  This study provides additional data that ultrasonography is not adequate to exclude significant degrees of fatty liver.

Study limitations included the retrospective analysis which relied on medical record accuracy, degree of steatosis was not based on a single pathologist, ultrasonography was not based on not based on a single radiologist, both BMI and triglycerides may vary based on age, gender, ethnicity and other factors.

Related blog entries:

Related references:

  • -Spitzer AL et al. Liver Transpl 2010; 16: 874-84.
  • -Liver Transpl 2013; 19: 437-49. Difficulty with precisely determining steatosis
  • -Hepatology 2011; 54: 1082.  U/S w ~85% sensitivity in detecting fatty liver.
  • -Gastroenterol 2008; 135: 1961.  Liver biopsy (in pediatrics) still needed as surrogates not accurate for correlating degree of fibrosis/injury.
  • -J Pediatr 2009; 155: 469.  Review.  No evidence-based guidelines for treating in pediatrics –main Rx wt loss/exercise.  Consider obtaining ultrasound.