Although clinical diagnosis of deep-vein thrombosis (DVT) has been regarded as equivocal, the use of symptoms scores may increase the probability of instrumental confirmation and help the clinician in therapeutic decisions, expecially when a compression ultrasound scanning is not immediately available. We propose a new and simple score, based on a retrospective analysis of our inpatients.
154 patients, out of 1650 consecutively admitted in our emergency medicine department, were submitted to a doppler -ultrasound evaluation for suspected DVT. We revised hystories, clinical examinations, and laboratory data of all patients. Odds ratios (with confidence intervals) were calculated for all rilevant variables (age, risk factors, physical examination, alternative diagnosis, D-dimer). Statistical analysis was performed employing the X2. A new score, named HOOD (Hystory, Objective, Other diagnosis and Dimer), was then developed.
A DVT was confirmed in 96/154 patients (62.3%). Sex and age over 70 years (OR 0.69) were not predictive of DVT. The presence of two o more risk factors (OR 3.15), as one or more signs of DVT (OR 6.2) and Dimer value greater than cut-off (OR 4.2) were all strong predictors of the ultrasound results. A HOOD score of 1 or less was not associated with thrombosis, whereas a score of 2 or more was highly predictive of DVT (OR 6.41; sensivity 0.96, specificity 0.52; positive and negative predictive value 0.75 and 0.88 respectively).
DVT can been predicted by a simple, clinical score. When a CUS is not immediately avalaible, a treatment with eparin should not delayed in patients with high clinical probability of DVT.
The clinical suspicion of DVT is made by gestalt or by use of scores that divide cases in three groups of risk (i.e., low, intermediate and high risk). We propose a clinical method that segregate patients in two clearly distinct groups of risk, in one of which initial treatment could be probably safely omitted.
N. Mumoli, None.