Introduction:

The deep vein thrombosis (DVT) is a common entity that mainly affects the deep venous system of the lower limbs, for which multiple clinical prediction scales have been developed, which have been constructed and validated in outpatients and inpatients.

Objetives:

We aimed to validated five clinical prediction scores for the diagnosis of lower limb DVT in patients from La Sabana de Bogota, Colombia.

Methods:

A cross-sectional study with analysis of a diagnostic test was carried out in patiens with suspected deep vein thrombosis, including those who had venous Doppler ultrasound of the lower limbs for suspected DVT. The performance of five clinical prediction scales for DVT (classic and modified Wells, Oudega, CEBI and Constans) for outpatients and inpatients was calculated in those scores who are validated in both populations and only in ambulatory or hospitalized patients for those that are specific scores.

Results:

Nine hundred seventy-four patients were entered into the analysis, of which 485 (49.7%) presented DVT. The Constans scale had a better diagnostic performance among inpatients and outpatients with an area under the ROC curve of 0.73 (95% 0.70-0.78) when compared with classic Wells, modified Wells, Oudega and CEBI. When we compared Constans performance in both groups of patients separately, we observed better performance with respect to the other scores.

Conclusion:

The Constans scale presents a better diagnostic performance compared to the other scales when applied to inpatients and outpatients.

        · resumen en Español     · texto en Español     · Español ( pdf )