Outcome of patients with suspected lower limb symptomatic deep vein thrombosis and a normal ultrasound-based initial diagnostic workup: a prospective study


Enrico Bernardi, Emergency Department, Ospedale S. Maria dei Battuti, Via Brigata Bisagno, 4 - 31015 Conegliano (TV), Italia.
Tel.: +39 438 66 3660, fax: +39 438 66 3580
E-mail: enrico.bernardi@ulss7.it
Giuseppe Camporese, Unit of Angiology, University Hospital of Padua, Via Nicolò Giustiniani, 2 - 35128 Padova, Italia.
Tel.: +39 49 8211248, fax: +39 49 8218739
E-mail: giuseppe.camporese@sanita.padova.it


Camporese G, Bernardi E, Scarano L, Ghirarduzzi A, Imberti D and Prandoni P, on behalf of The PALIST Study Group. Outcome of patients with suspected lower limb symptomatic deep vein thrombosis and a normal ultrasound-based initial diagnostic workup: a prospective study. J Thromb Haemost 2012; 10: 2605–6.

The current referral pattern for suspected deep vein thrombosis of the legs (LDVT) in Western countries implies that only about 20–30% of the patients presenting to an emergency department or to a vascular laboratory will eventually have the disease confirmed [1,2]. As venous thromboembolism (VTE) is a severe disease, clinicians obviously tend to focus on the exclusion of VTE, rather than on identifying the cause of the patients’ leg complaints. As a consequence, patients labeled ‘free from thrombosis’ may be often discharged without a definite diagnosis. Little is known about the short–medium-term outcome of these subjects, although <1% of them [2,3] will eventually receive a diagnosis of DVT, usually within the 3 months following the initial evaluation. We performed a prospective multicenter nested observational study to assess the outcome of symptomatic patients in whom DVT is ruled out by an ultrasound-based diagnostic workup.

Consecutive symptomatic outpatients referred to the study centers for suspected LDVT between 2003 and 2007 in whom the clinical suspicion had been excluded with the use of objective tests [2] were eligible for the study. The only exclusion criterion was unwillingness to participate in the study. Symptoms, signs and risk factors for LDVT were recorded for all patients during the baseline visit. In all instances we sought to establish a likely cause for leg complaints, and gave advice about any further diagnostic procedure (e.g. X-rays, ultrasonography, MRI and so on) deemed necessary to reach a definite diagnosis. All patients were scheduled for a follow-up visit after 3 months to assess their health status; those failing to attend the final visit were to be contacted by phone.

About 779 consecutive patients with suspected DVT were scanned; 22.7% of the patients had LDVT. The remaining 602 subjects with a normal ultrasound-based workup were included in the follow-up study. The demographic details, leg signs and symptoms, along with risk factors for LDVT as assessed at presentation, are reported in Table 1.

Table 1.   Demographics and clinical features of the 602 outpatients at presentation
  1. PDD, patient–doctor delay.

  2. Unless otherwise specified numbers in parentheses indicate percentages.

 Age, mean (range)63.3 (18–96)
 PDD, mean (range)4 (1–12)
Signs and symptoms n
 Pain479 (79.6)
 Pretibial edema424 (70.4)
 Tension299 (49.7)
 Redness206 (34.3)
 Paresthesias104 (17.3)
 Erythema102 (17.0)
 Cramps91 (15.2)
 Venous induration46 (7.6)
 Fever44 (7.3)
 Superficial phlebitis44 (7.3)
 Whole leg swollen43 (7.1)
 Cyanosis30 (5.0)
 Collateral veins16 (2.6)
Risk factors for DVT
 Age > 65 years315 (52.4)
 Varicose veins145 (24.1)
 Obesity123 (20.4)
 Trauma69 (11.5)
 Cancer57 (9.4)
 Surgical intervention52 (8.6)
 Immobilization41 (6.8)
 Diabetes41 (6.8)
 Fracture28 (4.7)
 Chronic heart failure27 (4.5)
 Hormone replacement treatment25 (4.2)
 Lung disease20 (3.4)
 Oral contraceptives17 (2.9)
 History of DVT17 (2.9)
 Peripheral arterial disease14 (1.4)
 Cirrhosis11 (1.8)
 Rheumatologic disease9 (1.6)
 Kidney disease5 (0.8)
 Puerperium3 (0.5)
 Paralysis3 (0.5)
 Pregnancy3 (0.5)

A likely or possible explanation for leg symptoms was identified in 377 (62.6%) and 93 (15.4%) patients, respectively, while in 132 (22%) patients no differential diagnoses could be hypothesized. Of 470 patients with a possible or probable alternate reason for their leg complaints, 115 (24.5%) had either a leg trauma or a suspected muscle tear, 80 (17.1%) had suspected erysipelas, 68 (14.4%) a superficial phlebitis, 49 (10.4%) hip or knee arthrosis/arthritis, 46 (9.7%) chronic venous insufficiency, 36 (7.7%) a lymphatic stasis, 36 (7.7%) a sciatic nerve pain, and the remaining had either a suspected Baker’s (popliteal) cyst (21; 4.4%), a chronic heart failure (13; 2.7%) or a post-thrombotic syndrome (6; 1.3%).

Of the 602 included patients, only 13 (2.1%) did not attend the required 3-month follow-up visit and were contacted by phone, while the remaining 589 were evaluated at the study centers. No patient was lost to follow-up. Overall, after 3 months, 317 (52.6%) patients were completely asymptomatic while 285 (47.4%) still complained of leg symptoms. Of the latter group, 214 (75.1%) patients reported an improved status, 63 (22.1%) had a stable clinical picture, while 8 (2.8%) were worse. These eight patients had chronic lymphatic edema (n = 2), arthrosis (n = 3), sciatic nerve pain (n = 1) or advanced-stage chronic venous insufficiency (n = 2).

The initially hypothesized cause of leg complaints was confirmed in 418 (69.4%) and excluded in 52 (8.6%) patients. Vascular and soft-tissue ultrasonography represented the definitive diagnostic tool for more than two-thirds of the patients with confirmed alternate diagnosis (279/418, 66.7%). The actual cause of leg complaints in the 52 patients in whom the initial diagnosis was ruled out was: muscle hematoma (n = 26), ruptured knee ligament or meniscus (n = 12), metastatic pelvic or abdominal lymphadenopathy (n = 6), inguinal hernia (n = 5) or pseudo-aneurysm of the femoral artery in patients with history of percutaneous coronary interventions (n = 3). Out of the 132 (22%) patients in whom no alternative hypotheses could be devised at presentation, a new diagnosis was established in 48 (8.0%), including isolated distal DVT in six (1%). In all these six patients distal DVT was diagnosed by whole-leg ultrasonography, performed because of worsening leg symptoms at 8, 12, 19, 22, 30 and 43 days from inclusion.

The 3-month mortality rate was 1.8% (11/602), mainly due to end-stage cancer, without fatal thromboembolic events.

Although the sample size may not be high enough to draw definitive conclusions, based on our experience a differential diagnosis may be established on clinical grounds in up to 70% of outpatients with suspected LDVT in whom the vascular disorder is ruled out by the initial ultrasound-based diagnostic workup. Interestingly, ultrasonography may be used to achieve a definite diagnosis in a significant number of subjects. As an example, a muscle hematoma was diagnosed by means of ultrasound in 34 of the 115 patients with a leg trauma, and similarly in the 21 patients with a ruptured Baker’s cyst in our series.

Although a certain number of ultrasound scans requested by either the general practitioners or the emergency physicians could have been avoided, it should be emphasized that ultrasonography plays a pivotal role both in supporting the differential diagnosis in patients in whom LDVT has been objectively ruled out and in excluding a concomitant LDVT in patients with another condition clearly evident on clinical grounds.

The 3-month prognosis of these patients is usually benign, as only around 3% have worsening symptoms, with a <5% 3-month LDVT incidence and a <2% mortality rate.

Disclosure of conflict of interests

The authors state that they have no conflict of interest.


Additional investigators of the PALIST Study

Fabio Verlato and Chiara Tonello (Unit of Angiology, University Hospital, Padova), Arrigo Berchio (Department of Emergency and Accident Medicine, S Giovanni Battista Hospital, Torino) and Alberto Cogo (Department of Vascular Medicine, Villa Berica Hospital, Vicenza), all in Italy.