Table 2.

Thromboembolic outcomes in accuracy studies

Potential testing schemes	Study, year	d-dimer test	No of patients (No of venous thromboembolism events)	Three month cumulative incidence of venous thromboembolism in% (95% CI)
Low clinical probability and a normal d-dimer result
	Bucek et al., 200297	STA-LIA	48 (0)	0.0 (0 to 7.4)
	Kraaijenhagen et al., 200298	SimpliRED	561 (10)	1.8 (0.9 to 3.2)
	Shields et al., 200299	SimpliRED	32 (0)	0.0 (0.0 to 10.1)
	Wells et al., 1998100	SimpliRED	206 (1)	0.5 (0.0 to 2.7)
	Ginsberg et al., 1997101	SimpliRED	178 (1)	0.6 (0.0 to 3.1)
	Anderson et al., 20009	SimpliRED	97 (0)	0.0 (0.0 to 3.7)
Pooled.*
Included patients with history of deep vein thrombosis				1.3 (0.5 to 2.3)
Excluded patients with history of deep vein thrombosis				0.3 (0.01 to 1.1)
Moderate clinical probability and a normal d-dimer result
	Shields et al, 200299	SimpliRED	20 (0)	0.0 (0.0 to 16.8)
	Wells et al, 1998100	SimpliRED	87 (3)	3.4 (0.7 to 9.8)
	Ginsberg et al, 1997101	SimpliRED	97 (3)	3.1 (0.6 to 8.8)
	Anderson et al, 20009	SimpliRED	51 (3)	5.9 (1.2 to 16.2)
Pooled†
High probability and a normal d-dimer result
	Shields et al, 200299	SimpliRED	7 (2)	29 (3.7 to 71.0)
	Wells et al, 1998100	SimpliRED	7 (1)	14.3 (0.4 to 57.9)
	Ginsberg et al, 1997101	SimpliRED	5 (2)	40.0 (5.3 to 85.3)
	Anderson et al, 20009	SimpliRED	15 (2)	13.3 (1.7 to 40.5)
Pooled†				21.0 (8.0 to 37.0)

^*P value comparing studies that excluded and included patients with history of deep vein thrombosis was significant, P=0.04.

^†P value comparing studies that excluded and included patients with history of deep vein thrombosis were not significant (P=0.5 for moderate probability and P=0.3 for high probability).