Abstract
Antithrombin deficiency is a strong risk factor for venous thromboembolism (VTE),
but the absolute risk of the first and recurrent VTE is unclear. The objective of
this paper is to establish the absolute risks of the first and recurrent VTE and mortality
in individuals with antithrombin deficiency. The databases Embase, Medline Ovid, Web
of Science, the Cochrane Library, and Google Scholar were systematically searched
for case–control and cohort studies. Bayesian random-effects meta-analysis was used
to calculate odds ratios (ORs), absolute risks, and probabilities of ORs being above
thresholds. Thirty-five publications were included in the systematic review and meta-analysis.
Based on 19 studies, OR estimates for the first VTE showed a strongly increased risk
for antithrombin deficient individuals, OR 14.0; 95% credible interval (CrI), 5.5
to 29.0. Based on 10 studies, meta-analysis showed that the annual VTE risk was significantly
higher in antithrombin-deficient than in non-antithrombin-deficient individuals: 1.2%
(95% CrI, 0.8–1.7) versus 0.07% (95% CrI, 0.01–0.14). In prospective studies, the
annual VTE risk in antithrombin deficient individuals was as high as 2.3%; 95% CrI,
0.2–6.5%. Data on antithrombin deficiency subtypes are very limited for reliable risk-differentiation.
The OR for recurrent VTE based on 10 studies was 2.1; 95% CrI, 0.2 to 4.0. The annual
recurrence risk without long-term anticoagulant therapy based on 4 studies was 8.8%
(95% CrI, 4.6–14.1) for antithrombin-deficient and 4.3% (95% CrI, 1.5–7.9) for non-antithrombin-deficient
VTE patients. The probability of the recurrence risk being higher in antithrombin-deficient
patients was 95%. The authors conclude that antithrombin deficient individuals have
a high annual VTE risk, and a high annual recurrence risk. Antithrombin deficient
patients with VTE require long-term anticoagulant therapy.
Keywords
anticoagulants - antithrombin deficiency - venous thromboembolism - meta-analysis
- Bayes theorem
