A red blood cell agglutination D-dimer test to exclude deep venous thrombosis in pregnancy

Background: D-Dimer testing is often used with compression ultrasonography for the diagnosis of deep venous thrombosis (DVT) in nonpregnant patients. The D-dimer test is highly sensitive, and a negative result can obviate the need for further testing for DVT. This test has not been studied for DVT diagnosis in pregnancy because its specificity was deemed too poor. Objective: To determine the sensitivity and specificity and assess the utility of the SimpliRED assay (Agen Biomedical, Brisbane, Australia) for the diagnosis of DVT in pregnant women. Design: Prospective cohort study. Setting: 5 tertiary university-affiliated Canadian hospitals providing care to obstetric patients. Patients: 149 consecutive pregnant women with suspected DVT who presented to 1 of the participating centers over 5 years. Intervention: Participating women were tested with compression leg ultrasonography (single or serially on days 0, 3, and 7) and received 3 months' clinical follow-up for the presence or absence of DVT. Whole blood was tested with the SimpliRED assay at initial presentation, and results were correlated with ultrasonography and clinical findings for the presence or absence of DVT. Measurements: The sensitivity, specificity, and negative predictive value of the SimpliRED assay were calculated, along with the prevalence of false-positive SimpliRED assay results (with 95% CIs). Results: The prevalence of DVT in the cohort was 8.7% (95% CI, 5.2% to 14.4%). The sensitivity of the SimpliRED assay was 100% (CI, 77% to 100% [13 of 13 patients]), the specificity was 60% (CI, 52% to 68% [81 of 135]), and the negative predictive value was 100% (CI, 95% to 100% [81 of 81]). The SimpliRED assay was positive in 0% (CI, 0% to 60%), 24% (CI, 14% to 37%), and 51% (CI, 40% to 61%) of women in the first, second, and third trimesters, respectively, among pregnant patients in whom DVT was not diagnosed. Limitations: The prevalence of DVT in the cohort was low, resulting in wide CIs. The clinician's previous impression was used to determine pretest probability in the absence of a validated clinical prediction rule for pregnant women. Conclusion: The SimpliRED assay may be useful in pregnancy because a normal result excludes DVT and occurs frequently enough to be clinically helpful.