A Murage*, M Gaudoin
(*Overseas Fellowship Training Scheme)
Department of Obstetrics and Gynaecology, Southern General Hospital, 1345 Govan Rd, Glasgow
Corresponding author: Dr Marco Gaudoin, Consultant Obstetrician & Gynaecologist, Dept of Obstetrics & Gynaecology, Southern General Hospital, 1345 Govan Road, Glasgow G51 4TF E-mail: Marco.Gaudoin@sgh.scot.nhs.uk
SMJ 2004 49(4): 150-151
Antenatal coagulation testing is commonly performed and yet there is no evidence on which to base practice. This retrospective review of testing and the results obtained in a one-year period at the Southern General Hospital, Glasgow, UK, showed that the majority of the tests were normal regardless of the clinical indication for testing. However, there was a significant association between thrombocytopaenia and an abnormal coagulation result (p<0.001). This analysis showed that the platelet count might be used to select patients who require subsequent coagulation studies without compromising maternal or fetal safety and with attendant savings in time and money.
Key words: Coagulation tests, obstetric, coagulopathy, thrombocytopenia
Antenatal patients often undergo coagulation testing for various clinical indications including hypertensive disease, antepartum haemorrhage and intra-uterine death. Each test costs approximately £18 and takes up to 20 minutes to perform.1 However, there are no formally published evidence-based guidelines on which to base practise. In the absence of any guidance, the aim of this analysis was to determine the indications for testing, correlate this with test results and use the findings to formulate guidelines for future testing in our unit.
An initial Medline search for relevant comparative studies was conducted but produced no formal studies specifically addressing this issue. The haematology department in our hospital provided a computer print out of all antenatal coagulation tests performed (patients’ names, hospital number, date(s) tests requested) for the year 2001. The medical records were retrieved and manually reviewed. Basic obstetric demographics (age, parity, gestation etc), indication for testing, full blood count (FBC), coagulation test results, liver function tests and urea and electrolyte assays were determined and entered directly in to a Microsoft Access database. Thrombocytopaenia was defined as a platelet count less than 150'109/L. We used our laboratory ’s reference ranges for the coagulation screen test values: prothrombin time, PT (normal range 10-13 seconds), activated partial thromboplastin time, APTT (normal range 23-35 seconds), and fibrinogen concentration (normal range 2-4 g/L). Hypertensive disease was defined as systolic or diastolic blood pressure more than 20mmHg from the booking blood pressure. Mild-moderate hypertension was defined as diastolic blood pressure of 90 – 109 mmHg ± proteinuria. Mean values and standard deviations were determined using the SPSS (Version 8.0) statistical programme.
A total of 200 women (7% of the 2896 annual deliveries) had at least one test performed in the study period. Twentyfive of these women had repeated testing (ranging from two to five tests). The mean age was 30.8 ± 5.8 years (compared with 28.9 ± 6.1 years for the general antenatal population, P<0.001). Sixty-two percent were primigravidae compared with 62.7% for the general antenatal population (P=not significant). Seventy percent of women tested were more than 33 weeks’ gestation and 11 women (5.5%) had a twin pregnancy (compared to 1.7% in the general hospital population, c2=14.8, P<0.001). The primary indications for coagulation testing were hypertensive disorders (61%), antepartum haemorrhage (15.5%), non-specific abdominal pain (2.5%), intrauterine death (6.5%) and various miscellaneous indications (14.5%). Of women with hypertensive disorders, 75% had only mildmoderate disease and none of these had an abnormal coagulation test result.
Miscellaneous indications for testing included unexplained maternal collapse (1 woman), idiopathic thrombocytopaenia (4), suspected thrombo-embolism (12), obstetric cholestasis (1) and epistaxis (1). In 10 women there was no apparent reason for testing discernible from the case notes.
The platelet count was normal in 188 tests (176 women) and none had an abnormal coagulation test result. The remaining 37 tests showed thrombocytopenia (table 1) but only two of these women had an abnormal coagulation test result (hypofibrinogenaemia; c2=17.3, p<0.001). Of the two women who had abnormal coagulation tests, one had severe pre-eclampsia and the other had an abruption. This latter woman had a normal platelet count and normal coagulation test on admission but as the abruption progressed, she developed thrombocytopaenia and subsequently a coagulopathy.
The women in the study population were older than the general antenatal population which may be related by the large proportion of women with hypertensive disorders. The majority of women tested had pre-eclampsia but three-quarters of these women had only mild-to-moderate disease. In addition, it was disappointing that in ten women there was no apparent reason for a coagulation test and no documentation regarding the rationale behind requesting the test.
Ninety-nine percent of all the coagulation tests were normal regardless of the clinical indication for testing. Only 1% of the women tested had a coagulopathy (hypofibrinogenaemia) and both had thrombocytopaenia (p<0.001). These data show that in the absence of thrombocytopaenia, a coagulation test is unlikely to be abnormal.
At the time of conducting this analysis, the British Committee for Standards in Haematology (BCSH) had convened a committee to investigate the “Use and misuse of the coagulation screen” and to formulate guidelines for testing.2 To our knowledge, this committee has yet to publish its recommendations. This was a small study and probably needs to be repeated on a larger scale but in the absence of available guidelines, we believe that the platelet count might be used to direct subsequent requests for coagulation testing. Frequently there are delays in siting a spinal or epidural anaesthetic because the results of a coagulation test are awaited (which takes approximately 20 minutes). Processing the FBC only takes about five minutes1 and we propose that a FBC and clotting screen may be taken simultaneously in clinical situations where a coagulopathy may develop. However, the clotting screen should be processed only if the platelet count is less than 150x109/L. This would save time and, with such selective testing, for the study period we estimate that we would have saved approximately £4000 without compromising maternal or fetal safety.
ACKNOWLEDGEMENTS: The authors wish to thank Medical Records staff and the staff at the Department of Haematology, Southern General Hospital for their help and advice.
REFERENCES
1 Haematology laboratory, Southern General Hospital, Glasgow, UK. Verbal
communication.
2 British Society for Haematology: www.b-s-h.org.uk or www.bcshguidelines.com.