The ‘Letter in Focus’ by Brenner et al.  and the commentaries by Lindqvist  and Gris  raise important questions in this age when women increasingly expect pregnancy to be safe and uncomplicated. The results of the many studies which have examined possible relationships between maternal heritable thrombophilia and pregnancy complications such as fetal loss, intrauterine growth restriction, and pre-eclampsia have been inconsistent. However, a meta-analysis of 31 studies reported significant associations between thrombophilia and fetal loss  and the results of two further meta-analyses suggest that inherited thrombophilia is associated with an increased risk of pre-eclampsia  or at least linked to the severity of disease expression .
It has been suggested that aspirin and/or heparin may be beneficial in preventing miscarriage or fetal loss associated with antiphospholipids [7–10], but these studies did not include a control group, where neither aspirin nor heparin was administered. Observational studies, including that reported by Brenner et al. [1,11,12] suggesting a role for thromboprophylaxis in women with previous fetal loss and inherited thrombophilia are small and not randomized. Sanson et al. in their review concluded that low-molecular-weight heparins were suitable for use during pregnancy . However, the studies included were not randomized, controlled trials and it is not possible to estimate accurately the balance of risks and benefits.
The recently published guideline from the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy  suggests ‘…women with recurrent pregnancy loss, a second trimester miscarriage, or a history of intrauterine death or severe or recurrent pre-eclampsia should be screened for underlying congenital thrombophilias’. In the next sentence the authors concede ‘…there are no convincing data to indicate whether antithrombotic… therapy is beneficial in women with congenital thrombophilia and pregnancy complications’. It is perhaps regrettable therefore that, in this evidence-based guideline, the authors go on to recommend ‘…antithrombotic therapy should be considered in this population’.
Gris , accepting Brenner's finding that LMWHs are effective in reducing the risk of pregnancy loss, suggests that choosing an appropriate dose may depend on stratifying the patients ‘…accordingly to the level of risk associated with their underlying thrombophilia marker’– but how? What evidence do we have on which to base this stratification? The evidence linking heritable thrombophilias with a risk of pregnancy complication is weak . Furthermore, we seem to have forgotten that ‘routine’ thrombophilia screening – even if we limit our testing to the widely accepted thrombophilias – deficiencies of the natural anticoagulants, factor V Leiden, the Prothrombin G20210A polymorphism and antiphospholipid syndrome – will reveal an ‘abnormality’ in around a quarter of women attending an antenatal clinic.
If we postulate that identifiable thrombophilias provoke pregnancy problems because they are prothrombotic, we must expect that hypercoagulability resulting from other, as yet unidentifiable defects or even the acquired haemostatic response to pregnancy itself, may also increase the risk of pregnancy complication. Pursuing this line of argument must lead to the suggestion that intervention with an antithrombotic such as a LMWH may be indicated in women with a history of recurrent or late pregnancy loss even if there is no evidence of an underlying identifiable thrombophilia.
I broadly agree with the many criticisms which Lindqvist  has of the Brenner study , but I am not convinced that it will ever be possible to study the effect of individual heritable thrombophilias on pregnancy outcome. Should we not adopt a more pragmatic approach and investigate the hypothesis that hypercoagulability may have caused, or contributed to, recurrent or late pregnancy loss in women with no other identifiable reason for their history and LMWH use in these women may improve pregnancy outcome – irrespective of whether or not an underlying thrombophilia can be identified?
Recurrent or late pregnancy losses are not only emotionally devastating for the woman and her partner but have serious social implications. While the vast majority of women are very grateful for the reassurance afforded by regular clinical examination during pregnancy with early identification and management of problems, we should not lose sight of the fact that most pregnancies are uncomplicated. We are teetering on the verge of implying that all women should be screened for thrombophilia prior to their first pregnancy so that those found to have a thrombophilic defect could be treated as ‘at special risk’ and some, or many, ‘routinely’ prescribed prophylactic thromboprophylaxis. It cannot be scientifically or ethically justifiable to even hint that such massive intrusion into pregnant women's management is appropriate based on the current very slim evidence. We are not too late to act more scientifically. We urgently need to agree which hypotheses are clinically testable and then to establish adequately powered, randomized, controlled studies to examine them.