Common inherited thrombophilic defects such as factor V Leiden and G20120A mutation of the prothrombin gene interact synergistically with oral contraceptives to increase the risk of venous thromboembolism.1,2 The best approach to identify women at higher risk of venous thromboembolism before taking oral contraceptives is controversial. Universal screening is not cost effective because 8000 women need to be screened for factor V Leiden to detect 400 mutations and prevent one episode of venous thromboembolism.1 Many authors recommend selective screening in women with a personal or family history of venous thromboembolism.1 However, the effectiveness of this approach has not been proved. The aim of our study was to evaluate the sensitivity and positive predictive value of a family history of venous thromboembolism for identifying common thrombophilic defects in women without thrombosis before taking oral contraceptives.
Participants, methods, and results
We prospectively evaluated a cohort of women (age range 15-49 years) consecutively referred to our thrombophilia unit by gynaecologists at family planning clinics in Bologna, Italy, between 1998 and 2000. The gynaecologists had established that the women were eligible to take oral contraceptives and had no history of venous thromboembolism. Before the women were screened, experienced investigators administered a modified structured questionnaire3 that was designed and validated to evaluate both personal and family history (first degree=parents and siblings, second degree=grandparents, aunts, uncles, and cousins) of venous thromboembolism (see BMJ's website for details). We considered family history positive if a thromboembolism was reported in any first or second degree relatives.
Thrombophilia screening was conducted as previously described.4 Prothrombin activity was measured by chromogenic assay5 and lupus anticoagulant by LA-test and LA-check assays (Organon Teknika, Rome, Italy). If prothrombin activity was confirmed to be above 1.10 U/ml, we analysed the DNA for the G20120A mutation according to the method of Poort et al.5 The tests were performed by staff blind to the results of the questionnaire.
We calculated sensitivity and positive predictive values according to standard methods. The 95% confidence intervals for proportions were calculated by an approximate method, and we used the χ2 test when appropriate. A two sided probability value <0.05 was considered significant. All data were analysed with the statistical package SOLO ( BMDP, Los Angeles).
We evaluated 324 women (mean age 34 years) who had a negative personal history for venous thromboembolism confirmed by our questionnaire. Thirty four women reported a positive family history (10%, 95% confidence interval 7% to 14%), of whom two were heterozygous for factor V Leiden and one had protein S deficiency. Thrombophilic defects were identified in 19 women (6%, 3% to 8%), only three of whom had a positive family history. Among the 290 women with a negative family history, thrombophilic defects were detected in 16 (6%, 3% to 8%); eight were heterozygous for factor V Leiden and eight were heterozygous for the G20120A mutation.
The table shows the sensitivity and positive predictive value of family history for identifying thrombophilic defects. The proportion of women with thrombophilia was similar among those with a positive history and those with a negative history of venous thromboembolism when first and second degree family history was considered (9% (3/34) v 5% (16/290), P=0.44) and when only first degree family history was considered (8% (2/26) v 6% (17/298), P=0.68).
Comment
Family history of venous thromboembolism has unsatisfactory sensitivity and positive predictive value for identifying carriers of common thrombophilic defects before taking oral contraceptives. A policy of selective screening may therefore miss a substantial number of women at increased risk of thromboembolism when taking oral contraceptives.
Supplementary Material
Table.
Sensitivity and positive predictive values of family history as a predictor of thrombophilic defects in 324 women with no personal history of venous thromboembolism
| Family history | All defects
|
Factor V Leiden
|
G20120A
|
|||||
|---|---|---|---|---|---|---|---|---|
| No (%) | 95% CI | No (%) | 95% CI | No (%) | 95% CI | |||
| First and second degree | ||||||||
| Sensitivity | 3/19 (16) | 0.0 to 32 | 2/10 (20) | 0.0 to 45 | 0/8 | — | ||
| Postive predictive value | 3/34 (9) | 0.0 to 18.3 | 2/34 (6) | 0.0 to 14 | 0/34 | — | ||
| First degree only | ||||||||
| Sensitivity | 2/19 (11) | 0 to 24.7 | 1/10 (10) | 0 to 28 | 0/8 | — | ||
| Postive predictive value | 2/26 (8) | 0 to 18 | 1/26 (4) | 0 to 11.2 | 0/26 | — | ||
Acknowledgments
We thank the gynaecologists at Bologna family planning clinics who referred women for screening.
Footnotes
Funding: This project was supported by a grant from the University of Bologna.
Competing interests: None declared.
The questionnaire is available on the BMJ's website
References
- 1.Vandenbroucke JP, van der Meer FJM, Helmerhorst FM, Rosendaal FR. Factor V Leiden: should we screen oral contraceptive users and pregnant women? BMJ. 1996;313:1127–1130. doi: 10.1136/bmj.313.7065.1127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Martinelli I, Taioli E, Bucciarelli P, Akvahan S, Mannucci PM. Interaction between the G20120A mutation of the prothrombin gene and oral contraceptive use in deep vein thrombosis. Arterioscler Thromb Vasc Biol. 1999;19:700–703. doi: 10.1161/01.atv.19.3.700. [DOI] [PubMed] [Google Scholar]
- 3.Frezzato M, Tosetto A, Rodeghiero F. Validated questionnaire for the identification of previous personal or familial venous thromboembolism. Am J Epidemiol. 1996;143:1257–1265. doi: 10.1093/oxfordjournals.aje.a008713. [DOI] [PubMed] [Google Scholar]
- 4.Palareti G, Legnani C, Frascaro M, Flamigni C, Gammi L, Gola G, et al. Screening of activated protein C resistance before oral contraceptive treatment: a pilot study. Contraception. 1999;59:293–299. doi: 10.1016/s0010-7824(99)00033-5. [DOI] [PubMed] [Google Scholar]
- 5.Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common genetic variation in the 3'-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood. 1996;88:3698–3703. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.