Letter: Enhancing the Diagnostic Performance of the Oral Cholate Challenge Test: Implications for Avoidance of Potentially Unnecessary Endoscopy

Abstract

In our previous publications in Alimentary Pharmacology & Therapeutics [1, 2], we highlighted the diagnostic performance of the disease severity index (DSI) derived from the oral cholate challenge test [3-5] (HepQuant DuO) to rule out large oesophageal varices (LEV) and other significant lesions of portal hypertension. In these publications, two criteria were evaluated: (1) using DSI ≤ 18.3 alone and (2) using DSI ≤ 24 or < 25 with platelet count > 135,000 μL⁻¹. The sensitivity cut-off for rule-out of LEV was based on Baveno VII criteria for miss rate < 5% (sensitivity > 95%) [6]. The diagnostic performance for DSI ≤ 18.3 was previously reported [1, 2], but the diagnostic performance of DSI with platelet count was not. Table 1 shows the combination of DSI ≤ 18.3 or DSI ≤ 24 with PLT > 135,000 μL⁻¹ for the HepQuant DuO test. The HepQuant DuO test had miss rates < 5% (sensitivity > 95%), negative predictive values (NPV) > 99% and negative likelihood ratios (NLR) ≤ 0.08. Specificity was 63.8% for LEV and 58.3% for all significant lesions of portal hypertension.

TABLE 1. Diagnostic performance (95% CI) of the combination of DSI ≤ 18.3 with DSI ≤ 24 and platelet count > 135,000 μL⁻¹ from the oral cholate challenge test (HepQuant DuO) to rule out of large oesophageal varices (LEV) and all portal hypertensive lesions (defined as oesophageal varices needing treatment, LEV, large gastric varices or severe portal hypertensive gastropathy).

	Large oesophageal varices	All portal hypertensive lesions
Sensitivity (%)	95.9 (86.0, 99.5)	97.2 (85.5, 99.9)
Specificity (%)	63.8 (58.9, 68.5)	58.3 (51.1, 65.2)
Positive likelihood ratio	2.65 (2.30, 3.05)	2.33 (1.96, 2.77)
Negative likelihood ratio	0.06 (0.02, 0.25)	0.05 (0.01, 0.33)
Prevalence (%)	10.8 (8.1, 14.1)	15.3 (11.0, 20.6)
Positive predictive value (%)	24.4 (21.8, 27.1)	29.7 (26.2, 33.4)
Negative predictive value (%)	99.2 (97.1, 99.8)	99.1 (94.4, 99.9)
Miss rate (%)	4.1	2.8
EGDs avoided (%)	57.3	49.8

From this analysis, we conclude that the combined criteria of DSI ≤ 18.3 with DSI ≤ 24 plus platelet count > 135,000 μL⁻¹ may be sufficient to rule out LEV and aid the decision to avoid endoscopy (EGD) (Table 1). When used in combination, the diagnostic performance for LEV is maintained with a low miss rate (4.1%) but a much higher likelihood to avoid endoscopy (57.3%). The same criteria when applied to clinically significant lesions of portal hypertension showed similar improvements in diagnostic performance.

Although additional studies and further validation are needed, using these criteria may enhance the utility of HepQuant DuO testing for informing the decision to avoid or proceed with endoscopy to screen for LEV when used in conjunction with other clinical, laboratory and radiologic information.