The Case
So the radiology resident says to me in disbelief, “you want another lower extremity Doppler when this patient had a negative study 2 weeks ago?” I answered, “Yes. A single one does not rule out DVT.” He asked me where I got that fact, and unfortunately, I could not honestly/smugly answer, “the literature.” All I could think was, “yep. That’s what Cindy Benson told me my first month of intern year.” For those of you who know her, the “Cindy told me” response is convincing enough. My problem is that this radiologist, along with many other deprived people, do not know her – meaning I should probably read up on why she is right.
Questions: At a time when health care costs are “spiraling out of control” why are we repeating an already negative US to rule out DVT? Is there a better one-time study?
The Literature
A 1998 review in Annals of Internal Medicine compiled statistics from numerous papers. In short, in first time symptomatic patients, US is 95% sensitive, and 96% specific for DVT. Of people with an initially negative US, most studies find a proximal DVT on repeat US 1 week later an average of 2 % of the time.
The literature suggests that about 5% of those with a negative initial study have distal (calf) DVTs. Out of known distal DVTs in untreated patients, 25% propagate proximally. These are generally expected to propagate within 7 days. Keep in mind that usual US only looks down to the popliteal and does not look into the lower leg.
According to a 2004 Annals of Internal Medicine paper, 500,000 patients are evaluated yearly for DVT with 80% being normal on the first time and (once again) 2% of negative studies found to be positive on the second study. That’s over 390,000 negative second studies to find 8,000 second positives! All the papers recommend getting (and it is standard of care to get) the second study, but that is a lot of extra negative tests.
Are there any better tests?
Though early reports (from the 1980s) of full leg US showed low sensitivities (70%) for calf thrombosis, some more recent studies report there may be some utility to full lower extremity Doppler. There have actually been approximately 2,253.15 of these, and I will try to summarize each concisely.
Paper 1: Patients with clinically suspected DVT received full leg US. Out of 445 patients, 61 had DVTs. Of the 384 patients with negative US, 375 followed up with 22 getting repeat studies for persistent/worsening symptoms. Three patients were found to have DVTs with 1 of the 3 having proximal DVT. They concluded that this incidence was so low (<1%), that: A full leg study improves the sensitivity of the US and may preclude the need for a second study.
Paper 2: Radiology in 2005 had similar findings. They had 423 patients with initially negative complete LE Doppler. They called all 423, got in touch with 413, and found that ONLY one patient was diagnosed with a PE during the 3 month follow-up period. Not a single patient died during follow-up, and all the repeat LE US done (out of the 16 people who had changing LE symptoms) were negative for DVT.
Paper 3: followed 401 patients with negative initial complete LE doppler. They found 2 patients with calf dvt’s at 3 months, no instances of proximal DVT, or PE. They conclude that with a single, negative, complete US, the incidence of VTE is 0.5% and that one US may be enough.
The main issue with these papers is that they didn’t repeat the complete US on everyone, instead, they focused on chart reviews along with phone or office interviews. While this strategy can miss DVTs, the data is still compelling.
You may be wondering, what about inpatients who are at higher risk for DVT?
Well, in The American Journal of Medicine, 2010 – near 2000 inpatients suspected of having DVT got complete, bilateral LE dopplers. Five hundred thirteen randomly selected patients with negative first studies were followed-up by telephone, 1 had a PE and 2 had DVT. Out of the >400 deaths in enrolled inpatients, 7 of them were deemed “possibly due to PE.” These 10 cases made the rate of VTE in the first 3 months after negative complete LE doppler 3.5%! The conclusion is that a single complete LE doppler might not be enough in inpatients or high risk patients.
So, are you convinced that complete LE dopplers once to rule out DVTs should be done? That’s great, though I don’t think our hospitals do them. If only we had another option….
I found two papers that looked at single US in conjunction with a D-Dimer.
1998 BMJ paper had 598 people with negative D-Dimer and negative initial US. At three month follow-up (an office appointment/phone call/home visit), 0.4% of patients had a positive VTE. Out of the 88 (13%) of patients with a negative US and a positive D-Dimer, the rate of DVT on repeat US 1 week later was a whopping 6%. In the interim, none of the 88 patients with positive D-dimer and negative US who were not anti-coagulated developed PE, died or had any complications.
D-dimer seems like a good tool here, as long as you are comfortable discharging someone with a negative LE doppler and a positive D-Dimer.
Lastly, a paper in 2005 (once again in annals of Internal Medicine), looked at D-dimer after negative initial LE doppler as well. They had 309 negative D-dimers after negative US. Of the 305 followed up with office visit or telephone call at three and six months, 3 people (<1%) had proximal DVT by 6 month follow-up with no instances of PE. Crazy point of interest: for this paper they used a point of care D-dimer (qualitative). Done in less than 1 minute, cheaply, at the bedside, with a drop or two of blood. Amazing!
Interestingly a positive D-Dimer with negative initial US resulted in + VTE on venography/history in almost 25% of patients! They conclude that negative D-dimer with neg US is enough to rule out DVT and that you can discharge patients with a negative US and a positive D-Dimer without anticoagulation.
Both these papers had the same issue as the previous ones where repeat studies were not done on all patients, just H&Ps at 3 months and 6 months.
Who out there is drinking the Kool-Aid? We could technically do D-Dimer with LE doppler at our shop and, if both are negative, eliminate the need for a repeat US 1 week later. What do you guys think? It’s cheaper, faster, the patient doesn’t have to follow-up in one week. Also, if the D-dimer is positive then you have even more convincing data to convince the patient to follow-up.
Conclusion
1. If you suspect DVT with a negative US, there is around a 2% chance the person has a DVT. They can be sent home without anticoagulation BUT they need a repeat US in 1 week.
2. A negative single Complete LE US seems to be enough to rule out DVT for outpatients not at high risk.
3. Negative US with negative D-dimer is enough to rule out DVT. Supposedly, negative US with positive D-dimer is safe to follow-up for repeat sono in 1 week.
I think the better conclusion is that Cindy Benson is always right, even if you don’t know the literature behind why.
By Dr. Andrew Grock
Read more articles on Ultrasound use in the Emergency Department here.
References
Kearon et al. A Randomized Trial of Diagnostic Strategies After Normal Proximal Vein Ultrasonography for Suspected DVT: D-Dimer Testing Compared with Repeat US. Ann Intern Med;2005:142:7, 490-497
Bernardi et al. D-dimer Testing as an adjunct to US in patients with clinically suspected DVT: prospective cohort study. BMJ;1998:317,1037-1040 Ann Intern Med. 1998 Dec 15;129(12):1044-9.
Stevens et al. Withholding Anticoagulation after a negative result on duplexUS for suspected symptomatic DVT. Ann Intern Med. 2004;140:12, 985-992
Cogo et al. Compression US for diagnostic management of patients with clinically suspected DVT: prospective cohort study. BMJ. 1998;316,17-20
Kearon et al The role of venous ultrasonography in the diagnosis of suspected deep venous thrombosis and pulmonary embolism. Thromb Res. 2013 Jun;131(6):487-92. doi: 10.1016/j.thromres.2013.04.022. Epub 2013 May 9.
Johnson et al. Risk of DVT Following a Single Negative Whole-Leg Compression US. JAMA. 2010;303-5, 438-445
Elias et al. A single complete US investigation of the venous network for the diagnostic management of patients with a clinically suspected first episode of DVT of the lower limbs. Thromb Haemost. 2003:221-7
Schellong et al. Complete Compression US of the leg Veins as a singel test for the diagnosis of DVT. Thromb Haemost. 2003:228-34
Elias et al. Diagnostic Performance of complete lower limb venous ultrasound in patients with clinically suspected acute PE. Thromb Haemost. 2004;91:635
Subramaniam et al. DVT withholding anticoagulation therapy after negative complete lower limb US findings. Radiology. 2005: 237:348-352
andygrock
Latest posts by andygrock (see all)
- Levamisole Toxicity - May 26, 2015
- Of Course, US Only for Kidney Stones… - May 18, 2015
- Case of the Month 11: Answer - May 12, 2015
3 Comments
Ian deSouza · March 8, 2014 at 9:55 am
Nice one Grock. All of this additional testing (and likely repeat ED visits in our system) to find small % of DVTs that were not evident initially seems unnecessary. Despite widespread use of anticoagulation for DVT and acceptance as “standard of care”, bear in mind that there is not much evidence out there to support this practice.
Cundiff DK et al. Anticoagulants vs non-steroidal anti-inflammatories or placebo for treatment of venous thromboembolism. Cochrane Database Syst Rev. 2006 Jan 25;(1):CD003746.
Robin Fordham · October 31, 2016 at 10:04 pm
My patient story. DVT three years ago, only symptom was pain in calf. No risk factors other than three days car travel. Stopped warfarin at 3 months.
3 years later, calf pain again same side. Positive d-dimer, negative US. Told by two physicians (one my dr of 20 years) likely muscle pain, despite positive d-dimer and previous DVT in same leg. 5 days later chest pain and continued calf pain brought me to ER where PE was diagnosed with two small clots by CT. This blog post kept me vigilant given the positive d-dimer.
Should physicians have been more concerned given positive d-dimer in combination with previous DVT? D-dimer results were not available at time of ultrasound – might knowing I had a positive d-dimer directed the ultrasound in any way?
Ian deSouza · November 1, 2016 at 1:44 am
Thanks for sharing your story. I suppose that if you knew the positive d-dimer result, it may have been recommended that you repeat the sonogram in 2-7 days. Now with more available, quantitative d-dimer assays and direct oral anticoagulants, a case such as yours is much easier to diagnose and treat.
Here’s a somewhat relevant, more recent study published in BMJ:
Should we be looking for and treating isolated calf vein thrombosis?
Horner D, et al. Emerg Med J 2016;33:431–437. doi:10.1136/emermed-2014-204230