The authors discuss the use of low-dose computed tomography (LD-CT) in collectives with low COVID-19 prevalence (1). They note that the predictive values (positive predictive value [PPV] and negative predictive value [NPV]) of diagnostic test procedures decrease with the prevalence, such that a direct arithmetic scaling of the predictive values for scenarios of low prevalence is not possible. They justify this with the fact that prevalence is taken into account when interpreting the findings. Aside from the fact that a lower prevalence increases rather than decreases the NPV (2), the diagnostic precision of radiological findings should be independent of prevalence (= pre-test probability) if done according to standardized criteria (CoV-RADS classification).
In East Thuringia, only a few patients were infected with COVID-19, which is why the prevalence from the Jena University Hospital was significantly lower (at 8%) than that from the Aachen study (39%) in the same period. Nevertheless, we observed a similar test quality for the COVID-19 CT findings: positive likelihood ratio (LR +) 16.1, and negative likelihood ratio (LR-) 0.16 (Aachen: LH + 11.0 and LH- 0.06). Thus, the test quality of radiological examinations does not depend significantly on prevalence, and a direct arithmetic scaling to scenarios of low prevalence is entirely possible. The Fagan nomogram (3) is used to calculate the post-test probabilities for various pre-test probabilities and for constant test quality.
Due to low prevalence and the resulting high NPV of 98.6%, the combination of LD-CT and SARS-CoV-2-PCR (PCR, polymerase chain reaction) has been successfully used in Jena to exclude COVID-19 (4). The limitations of the combined diagnostics are reduced by a strict internal containment concept in the event of suspected COVID-19, with further diagnostic tests.
References
- 1.Schulze-Hagen M, Hübel C, Meier-Schroers M, et al. Low-dose chest CT for the diagnosis of COVID-19—a systematic, prospective comparison with PCR. Dtsch Arztebl Int. 2020;117:389–395. doi: 10.3238/arztebl.2020.0389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mausner JS, Kramer S. Mausner and Bahn Epidemiology: an introductory text. Philadelphia: WB Saunders. 1985 [Google Scholar]
- 3.Fagan TJ. Letter: Nomogram for Bayes theorem. N Engl J Med. 1975;293 doi: 10.1056/NEJM197507312930513. [DOI] [PubMed] [Google Scholar]
- 4.Baeyens JP, Serrien B, Goossens M, Clijsen R. Questioning the „SPIN and SNOUT“ rule in clinical testing. Arch Physiother. 2019;9 doi: 10.1186/s40945-019-0056-5. [DOI] [PMC free article] [PubMed] [Google Scholar]