It is best to base ᴅ-dimer testing on a patient's clinical pretest probability (C-PTP) for deep vein thrombosis (DVT), rather than testing all patients who present with symptoms of first DVT episode. This strategy can exclude DVT in more patients without increasing missed diagnoses, according to a randomized, multicenter, controlled trial in 1723 patients at 5 medical centers in Canada.
Lori-Ann Linkins, MD, an assistant professor in the Division of Hematology and Thromboembolism, Department of Medicine, McMaster University in Hamilton, Ontario, Canada, and colleagues published their findings in the January 15 issue of the Annals of Internal Medicine.
ᴅ-dimer testing is sensitive but not specific for identifying DVT. Selectively testing ᴅ-dimer levels lowered the proportion of patients who needed ultrasonography and decreased the percentage of patients who required ᴅ-dimer testing by 21.8% (95% confidence interval [CI], 19.1% - 24.8%).
"In this trial comparing uniform with selective ᴅ-dimer testing in patients with suspected first DVT, a selective strategy — which used a higher ᴅ-dimer threshold to exclude first acute DVT in outpatients with low C-PTP and omitted ᴅ-dimer testing in outpatients with high C-PTP and all inpatients — was as safe as and more efficient than the uniform testing strategy, which used the same threshold to exclude DVT in all patients," the authors write.
Patients were randomly assigned to the selective testing (n = 860) or uniform testing (n = 863) groups on presentation for suspected first DVT episode. Of the study participants, 1542 (89%) were outpatients and 181 (11%) were inpatients.
All patients in the uniform testing group underwent ᴅ-dimer testing. Levels less than 0.5 μg/mL were considered negative, and levels of 0.5 μg/mL or higher were considered positive. For patients with positive results, ultrasonography of the proximal veins in the symptomatic legs was conducted; patients with normal ultrasonogram and high C-PTP had ultrasonography repeated on the same legs 6 to 8 days later.
Patients in the selective testing group only underwent ᴅ-dimer testing if they were outpatients and had low or moderate C-PTP. Outpatients with high C-PTP and all inpatients underwent ultrasonography only. ᴅ-dimer levels in the low C-PTP group were considered negative if they were below 1.0 μg/mL and positive if they were 1.0 μg/mL or above.
For patients in the moderate C-PTP group, ᴅ-dimer levels were considered negative if they were below 0.5 μg/mL and positive if they were 0.5 μg/mL or above. Patients with positive results had ultrasonography, and patients with normal ultrasonogram and moderate or high C-PTP had ultrasonography repeated 6 to 8 days later.
Of the patients in the uniform testing group, 859 (99.5%) had ᴅ-dimer testing, 505 (58.5%) had initial ultrasonography, and 334 (38.7%) had ultrasonography repeated after 6 to 8 days.
Positive ᴅ-dimer results were found in 506 patients (418 outpatients and 88 inpatients), and negative results were found in 353 patients (351 outpatients and 2 inpatients). Four patients had no test. DVT was diagnosed by initial ultrasonography in 56 ᴅ-dimer-positive patients (11.1% of the 506 ᴅ-dimer-positive patients and 6.5% of 863 patients in the uniform testing group). None of the 81 patients with low C-PTP and a ᴅ-dimer level between 0.5 and 1.0 µg/mL had DVT on ultrasonography.
"You're Not Missing Cases"
Venous thromboembolism (VTE) was diagnosed during follow-up in 4 patients: 0.8% (95% CI, 0.2% - 2.0%) of the 506 ᴅ-dimer-positive patients with normal initial ultrasonogram and 0.5% (95% CI, 0.1% - 1.3%) of the 798 patients without DVT on initial testing who were still in the study at 3-month follow-up. No outpatients with low C-PTP and ᴅ-dimer levels between 0.5 and 1.0 μg/mL were diagnosed with VTE during follow-up. No VTE was diagnosed during follow-up in any ᴅ-dimer-negative patient (0.0%; 95% CI, 0.0% - 1.1%).
ᴅ-dimer testing was done in 668 of the 860 patients (77.7%), initial ultrasonography was done in 438 patients (50.9%), and ultrasound was repeated after 6 to 8 days in 383 patients (44.5%). ᴅ-dimer results were negative in 288 (80%) and positive in 72 (20%) of the 360 outpatients with low C-PTP.
None of the 288 ᴅ-dimer-negative patients (200 with ᴅ-dimer level < 0.5 μg/mL and 88 with ᴅ-dimer level 0.5 - 1.0 μg/mL) experienced VTE during follow-up (95% CI, 0.0% - 1.3%).
Of the 72 ᴅ-dimer-positive patients, 8 (11%) had DVT diagnosed by ultrasonography during initial testing. No patient with a normal ultrasonogram experienced VTE during follow-up (95% CI, 0.0% - 5.1%).
A total of 132 (43%) of the outpatients with moderate C-PTP were ᴅ-dimer-negative and 176 (57%) were ᴅ-dimer-positive. Two patients had no ᴅ-dimer testing. One ᴅ-dimer-negative patient experienced VTE during follow-up (0.8%; 95% CI, 0.0% - 4.3%), and 5 were lost to follow-up. DVT was diagnosed by ultrasonography during initial testing in 23 (13%) of the 176 ᴅ-dimer-positive patients. One of the 153 ᴅ-dimer-positive patients with normal ultrasonogram developed VTE during follow-up (0.6%; 95% CI, 0.0% - 3.2%); 4 patients were lost to follow-up.
DVT was diagnosed during initial testing in 20 (10.5%) of the 100 outpatients with high C-PTP and the 90 inpatients. VTE was identified during follow-up in 2 patients with normal ultrasonograms (1.1%; 95% CI, 0.1% - 3.8%). One patient was lost to follow-up.
DVT was diagnosed during initial testing in 51 (5.9%) of the selective testing patients. VTE developed during follow-up in 4 (0.5%; 95% CI, 0.1% - 1.3%) of the patients who had no DVT diagnosed during initial testing.
During follow-up, the difference between the groups in the number of VTE events was 0.0 percentage points (95% CI, −0.8 to 0.8 percentage points) in patients not diagnosed with DVT during initial testing and −0.3 percentage points (95% CI, −1.8 to 0.8 percentage points) in favor of selective testing in the outpatient or low C-PTP subgroup.
The difference between the groups in the proportion of those undergoing testing was −21.8 percentage points (95% CI, −24.8 to −19.1 percentage points) for ᴅ-dimer testing and −7.6 percentage points (95% CI, −12.2 to −2.9 percentage points) for ultrasonography, both in favor of selective testing.
The proportion of patients in the outpatient and low C-PTP subgroup who had ultrasonography was 20.0% in the selective testing group and 41.0% in the uniform testing group (difference, −21.0 percentage points in favor of selective testing; 95% CI, −27.6 to −14.2 percentage points).
Daniel J. Giaccio, MD, vice chair of medicine at Lutheran Medical Center in Brooklyn, New York, commented on the study in a telephone interview with Medscape Medical News. Selective testing enabled the researchers to avoid unnecessary ultrasounds, he noted. "In this day and age of cost-effectiveness, and especially with the incentives...for doctors to practice more cost-effective care, [it's good] to know that you're not missing cases — you're actually picking up more," Dr. Giaccio explained.
The study was supported by a grant from the Heart and Stroke Foundation of Ontario. Dr. Linkins has disclosed no relevant financial relationships. The other authors report a variety of relevant financial relationships including the receipt of consulting fees, presentation fees, and research grants. Complete conflict-of-interest information is available on the journal's Web site. Dr. Giaccio has disclosed no relevant financial relationships.
Ann Intern Med. 2013;158:93-100. Abstract
Medscape Medical News © 2013 WebMD, LLC
Send comments and news tips to news@medscape.net.
Cite this: Targeted ᴅ-dimer Testing Best for DVT Diagnosis - Medscape - Jan 14, 2013.
