Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with the interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.
Published evidence
Six studies are summarised in this briefing including 3,134 patients who had either come to hospital with acute abdominal pain or who have had endoscopic retrograde cholangiopancreatography (ERCP).
The clinical evidence and its strengths and limitations are summarised in the overall assessment of the evidence.
Overall assessment of the evidence
The evidence suggests that the Actim Pancreatitis rapid test may be a reliable method for diagnosing acute pancreatitis. Included studies consist of 1 meta-analysis and 5 prospective observational studies. Most of the evidence base for the technology was on patients coming to hospital with acute abdominal pain (patients with susceptive acute pancreatitis). Three prospective single centre studies investigated the diagnostic accuracy in patients after ERCP.
Results from the meta-analysis, which involved 13 studies on patients with suspected acute pancreatitis, showed a pooled sensitivity of 82.3% and specificity of 93.5% in these patients. This indicated that the test may result in 17.7% false negative and 6.5% false positive results. Two multicentre prospective observational studies also assessed diagnostic accuracy for this clinical situation and showed sensitivities of 73.1% and 68.6%, and specificities of 62.5% and 87.1%. The sensitivity and specificity for detecting post-ERCP ranged from 81% to 100% and 96% to 97.1%, respectively.
All studies had a urinary trypsinogen-2 cut-off value of 50 micrograms/L. Of the studies that reported on the severity of disease, the proportion of severe pancreatitis patients was between 13% and 40%, which may be reflective of the patient population (approximately 25% of acute pancreatitis cases are severe). Most of the individual studies compared the test with serum or urine amylase measurements; 4 studies used serum lipase as a reference standard. Some of the studies may have been underpowered to detect diagnostic accuracy because of their small sample size. None of the studies were done in the UK, limiting the generalisability of results to the NHS. Available evidence reports on the diagnostic accuracy of the test only. There are no data on the effect of the test on clinical outcomes or healthcare resource use.
Yasuda et al. (2019)
Study size, design and location
Prospective multicentre observational study on 94 patients with acute abdominal pain, from 17 centres in Japan between April 2009 and December 2012.
Key outcomes
Of the 78 patients with acute pancreatitis, 57 had a positive trypsinogen-2 dipstick test result. The test had a sensitivity of 73.1% and specificity of 62.5%. The positive and negative predictive values of test for diagnosing acute pancreatitis were 90.5% and 32.3%, respectively. The median levels of urinary trypsinogen-2 were 2.87 mg/dL and 6.49 mg/dL in patients with mild and severe pancreatitis and the area under the curve (AUC) score was 0.704. This was numerically higher than that of other pancreatic enzymes tested which included urinary and serum amylase, creatine and lipase, and urinary trypsinogen activation peptide.
Mayumi et al. (2012)
Study size, design and location
Prospective multicentre observational study on 412 patients with acute abdominal pain, from 21 centres in Japan between September 2008 and April 2009.
Intervention and comparators
Urinary trypsinogen-2 dipstick test and quantitative trypsinogen-2 assay (Actim Pancreatitis) compared with serum amylase and lipase tests.
Key outcomes
The trypsinogen-2 dipstick test had a sensitivity of 68.6% and a specificity of 87.1%. The sensitivity of the dipstick test for pancreatitis caused by alcohol and gallstones was 72.2% and 81.8%, respectively, which was much higher compared with amylase testing. Changing the cut-off point to include positive (+) and very positive (++) results only, increased the specificity to 92.2%, and the positive likelihood ratio was 7.63.
Strengths and limitations
Useful real-world data on patients presenting as an emergency with acute abdominal pain. The study was done in Japan, so the relevance to the NHS is limited. Study enrolment was done by gastroenterologists and surgeons and it involved a high proportion of patients with mild pancreatitis, which may not reflect standard clinical practice or typical patient population in the UK.
Chang et al. (2012)
Key outcomes
The pooled sensitivity was 82.3% and the specificity was 93.5%. The diagnostic odds ratios for the test was 85.23 and the AUC was 0.9673.
Strengths and limitations
Included 13 studies that the authors judged as generally high quality. In total, involved a large number of patients, but some of the included studies had a small sample size and may not have been adequately powered to estimate diagnostic accuracy. There could have been publication bias. Only 1 of the studies used a serum lipase test as a reference standard. None of the included studies were done in the UK.
Tseng et al. (2011)
Intervention and comparators
Urinary trypsinogen-2 dipstick test (Actim Pancreatitis) compared with serum amylase and lipase tests.
Key outcomes
Of the 13 patients with post-ERCP pancreatitis, 11 (84.6%) had a positive dipstick test result 3 hours after ERCP. All 13 patients with post-ERCP pancreatitis showed a positive test result 24 hours after ERCP. Three hours after ERCP, the dipstick test had a sensitivity for diagnosing post-ERCP pancreatitis of 84.6% and a specificity of 97.1%. The positive and negative predictive values for the test were 73.3% and 98.5%, respectively. The test showed numerically higher positive predictive values compared with the serum amylase test (42.9%) and lipase tests (36.4% and 42.3% at the cut-off level of 3 and 5 times the upper reference, respectively) 3 hours after ERCP.
Strengths and limitations
The study was done in Taiwan, so the relevance to the NHS is limited. The definition of post-ERCP pancreatitis is not consistent in the literature, the study used a definition from a consensus in 1991. The study excluded patients who had a positive urinary trypsinogen-2 dipstick test result before ERCP.
Sankaralingam et al. (2007)
Study size, design and location
Prospective single centre observational study on 30 patients having ERCP in the US.
Key outcomes
One patient with a history of pancreatic adenocarcinoma was excluded because the pre-ERCP test results were positive. One of the patients was unable to give a urine sample for the 1-hour test. Five of the 29 patients developed post-ERCP pancreatitis, diagnosed by the gastroenterologist. Six out of 28 patients had positive results in 1 hour and 6 of 29 patients had positive results in 4 hours. One hour after ERCP, the dipstick test had a sensitivity for diagnosing post-ERCP pancreatitis of 1.0 and a specificity of 0.91. The positive predictive value was 0.66, and the negative predictive value was 1.0. Four hours after ERCP, the test had a sensitivity of 1.0 and a specificity of 0.96. The positive predictive value was 0.8, and negative predictive value was 1.0.
Strengths and limitations
All ERCPs were done by 1 gastroenterologist using the same preoperative preparation. The treating physician was blinded to the urinary trypsinogen-2 results. The study was a pilot study with a small sample size. It was done in the US, so the relevance to the NHS is limited.
Kemppainen et al. (1997)
Intervention and comparators
Urinary trypsinogen-2 dipstick test (Actim Pancreatitis) compared with serum and urine amylase measurements.
Key outcomes
Post-ERCP pancreatitis developed in 11 of the 106 patients studied. At 6 hours, the test was positive in 9 of these patients and in 9 of the 97 patients without post-ERCP pancreatitis. The sensitivity of the dipstick test for diagnosing post-ERCP pancreatitis was 81% and the specificity was 90%. When asymptomatic patients were excluded the specificity was 97%. The dipstick test showed a good correlation with quantitative trypsinogen-2 assays (0.75). The sensitivities of serum and urine amylase measurements were 91% and 81%, respectively. The specificities were 96% and 95%, respectively.
Recent and ongoing studies
Urine trypsinogen 2 dipstick for the early detection of post-ERCP pancreatitis. ClinicalTrials.gov identifier: NCT03098082. Status: recruiting. Indication: post-ERCP acute pancreatitis. Devices: Actim Pancreatitis. Study completion date: August 2021. US.