Written by Clay Smith
Using a diagnostic algorithm of revised Geneva score (rGS), D-dimer, leg ultrasound, CTPA (and V/Q if inconclusive) safely ruled out PE in pregnant women.
Why does this matter?
Much has been written on how to work up PE. In all these studies, pregnant women are excluded. We know pregnancy increases risk for venous thromboembolism (VTE), and D-dimer increases normally during pregnancy. But might we still be able to risk stratify pregnant patients to determine who can be ruled out without CT?
The CT-PE-Pregnancy Group brings you…
This was a multi-center prospective study of 395 pregnant women presenting with clinical suspicion for PE. PE prevalence was 7.1%. See the algorithm.
The primary outcome was any VTE event over the next 3-months, “in women who did not receive anticoagulant therapy on the basis of negative results on the initial work-up.” Pretest probability on rGS was high in just 3 women (0.8%). Of the low to intermediate pretest probability patients, 341/395 (87%) had a positive D-dimer. Of the 11.7% with a negative D-dimer, none had VTE in 3-month follow up. Following the algorithm, those with a positive D-dimer had leg ultrasound, of which only 2% were positive for DVT. They then went on to CTPA or V/Q if CTPA was inconclusive. Of those with a completely negative workup on an intention-to-treat basis, 0% (95%CI, 0.0%-1.0%) had VTE in 3-month follow up. Some of these, 22 patients, were started on prophylactic anticoagulation due to prior VTE (n=17), pre-eclampsia (n=3), calf vein clot (n=1), or upper extremity clot (n=1). If these had been excluded, results would have remained with zero patients having VTE at 3 months, though with a slightly higher upper 95%CI.
Take home points are:
The revised Geneva wasn’t meant to be used in pregnant women but worked well in this study.
D-dimer was negative more often in the 1st > 2nd > 3rd trimester but was still clinically useful, ruling out disease without imaging in 11.7% of the patients - not great but not nothing.
Ultrasound before CTPA ruled out only a small number of patients. Given that it has zero radiation, it’s still seems worth it. The downsides are time and cost.
CTPA was a good test and did not have too many inconclusive results. It is probably better and delivers less radiation than V/Q.
Where to go from here? The authors will likely revise the revised Geneva and tailor it for pregnant women. They also may look at raising the D-dimer threshold based on trimester. I plan to use this algorithm in practice, as it is the best quality evidence to date in pregnant patients.
Diagnosis of Pulmonary Embolism During Pregnancy: A Multicenter Prospective Management Outcome Study. Ann Intern Med. 2018 Oct 23. doi: 10.7326/M18-1670. [Epub ahead of print]
Open in Read by QxMD