We greatly appreciate the interest that Dr. Teruhiko Imamura has shown in our article “A new pretest probability score for diagnosis of deep vein thrombosis in patients before surgery” [
- Hamamoto Y.
- Tokushige A.
- Toshinori Y.
- Ikeda Y.
- Horizoe Y.
- Yasuda H.
- et al.
A new pre-test probability score for diagnosis of deep vein thrombosis in patients before surgery.
We welcome the opportunity to continue the discussion and to clarify the issues you pointed out in your letter. We will respond to each of your points point-by-point.
First, we recognized the selection bias you mentioned. As described in the “Limitations”, we conducted on pre-operative subjects with D-dimer ≥1.0 μg/ml for developing Kagoshima-deep vein thrombosis (DVT) score. We think that further validation evidence is needed to apply the Kagoshima-DVT score to the general population. In several facilities, we started an observational study to validate the usefulness of the Kagoshima-DVT score in pre-operative subjects included with D-dimer <1.0 μg/ml. The validation findings will be shared at a later date.
Second, we were concerned that if a cut-off value is statistically calculated using our data set in this study, its internal validity will be improved but its outward validity would be diminished. We adopted the cut-off value set from a previous study or the upper limit of the standard reference value to apply Kagoshima-DVT score broadly at other facilities and boost external validity.
Third, as you pointed out, each of the patient groups excluded in our study affected D-dimer values, so it was difficult to predict DVT before surgery. Therefore, as Wells reported [
- Wells P.S.
- Owen C.
- Doucette S.
- et al.
Does this patient have deep vein thrombosis?.
], we consider it is necessary to stratify the probability with the Wells score for DVT to detect symptomatic DVT.
In fact, the incidence of 6–7 % DVT is observed in the low-probability group (Kagoshima-DVT score 0–2 points). However, our study cohort consisted of pre-operative subjects with D-dimer ≥1 μg/ml, and the prevalence of DVT was considered higher than in the pre-operative general population.
Finally, in the pre-operative general population including patients with D-dimer <1 μg/ml in a real clinical setting, we expect that the incidence of DVT in the low-probability group would be very low. We will plan to verify the points in further studies. In addition, the rate of fresh DVT in the low-probability participants in Kagoshima DVT score was 1–3 %, and quite low. In the results, we expected to exclude DVT that needed to be treated clinically.
Published online: September 12, 2022
© 2019 Published by Elsevier Ltd on behalf of Japanese College of Cardiology.