Endoscopy 2020; 52(09): 819
DOI: 10.1055/a-1185-9495
Letter to the editor

A very tall order for single-operator cholangioscopy

Theodor Voiosu
1  Internal Medicine Department, Carol Davila School of Medicine, Bucharest, Romania
2  Gastroenterology Department, Colentina Clinical Hospital, Bucharest, Romania
,
Cristian Băicuș
1  Internal Medicine Department, Carol Davila School of Medicine, Bucharest, Romania
3  Internal Medicine Department, Colentina Clinical Hospital, Bucharest, Romania
,
Andrei Voiosu
2  Gastroenterology Department, Colentina Clinical Hospital, Bucharest, Romania
› Author Affiliations

We read with great interest the study by de Vries et al. [1] reporting on the diagnostic accuracy of single-operator peroral cholangioscopy (sPOCS), which seems to run contrary to the trend of recent findings [2].

We believe the underlying issue for most studies in this field is conceptual: overlooking Bayes’ theory and its practical implications [3]. The “one-size-fits-all” approach places sPOCS in the untenable position of having to simultaneously rule in and rule out cancer with high accuracy ([Fig. 1]). The authors acknowledge the high percentage of primary sclerosing cholangitis (PSC) patients in their cohort (40 %) as a limitation [4] but underestimate its practical importance.

Zoom Image
Fig. 1 A visual representation of the challenge of applying the same diagnostic test (i. e. single-operator peroral cholangioscopy) in two types of population: high risk and low risk for malignancy, respectively. Highlighted bars represent the target population in each clinical scenario (constructed using the Two by Two Diagram display and calculator. http://54.221.244.58:3838 /Truth/). T + true positive; T- true negative; F + false positive; F- false negative; CCA cholangiocarcinoma; PSC primary sclerosing cholangitis.

In PSC patients, the endoscopist is called to rule out cancer, as the pretest probability of malignancy is much lower (< 10 %) [5] than that of non-PSC-related strictures. Applying the positive and negative likelihood ratios reported in this study for sPOCS biopsy and brush cytology, we calculated that they could increase the post-test probability of malignancy from 50 % to 86 % and 99 % respectively in a high-risk population (50 % incidence of cancer) while, in a low-risk setting, sPOCS biopsy could not significantly decrease the post-test probability of malignancy. This underscores the caveats of conducting diagnostic tests without focusing on the relevant clinical end points.



Publication History

Publication Date:
26 August 2020 (online)

© Georg Thieme Verlag KG
Stuttgart · New York