| Why Mothers Die | ||||||
|
CHAPTER 2 THROMBOSIS AND THROMBOEMBOLISM Summary
Forty-eight deaths from thrombosis or thromboembolism are counted in this Chapter. Forty-six were from pulmonary embolism and two from cerebral thrombosis secondary to a deep vein thrombosis (DVT). In addition there were two Late deaths from pulmonary embolism which are counted in Chapter 14.
Of the 46 deaths from pulmonary embolism, three occurred after operative procedures in early pregnancy. There were 15 other antenatal deaths, mainly in the first trimester of pregnancy. Of the 25 postpartum deaths, 15 occurred after caesarean section and 10 followed vaginal delivery. In three cases the only information available was from the death certificate, and the mode of delivery was not stated.
The total of 46 deaths from pulmonary embolism represents a significant increase from the 30 cases in 1991-93. There were increases in all categories, the largest increases being in deaths after vaginal delivery and after ectopic pregnancy or abortion. Although part of this increase could be due to increased case ascertainment, this would only account for 9 of the extra 16 cases by comparison with 1991-93.
Substandard care was present in many cases, although it is impossible to quantify this with accuracy due to the incomplete details available on some report forms. Often there was a failure to appreciate the importance of risk factors such as obesity, or of symptoms such as calf pain. There is a need to educate all doctors, not just obstetricians, that the risk of thromboembolism is increased from early pregnancy until the late puerperium.
There were at most four deaths because of failure of treatment. Once patients are treated for venous thromboembolism in pregnancy they usually survive. The problems are failure of diagnosis and, in particular, failure even to consider the possibility of venous thromboembolism and failure to give adequate prophylaxsis.
Wider use of thromboprophylaxis (not only after caesarean section) and better investigation of classic symptoms (particularly in high-risk women) are urgently recommended.
Thrombosis and thromboembolism: key recommendations
Wider use of thromboprophylaxis (not only after caesarean section) and better investigation of classic symptoms (particularly in high risk women) are urgently recommended.
Obstetricians and gynaecologists are reminded that even in the first trimester, pregnancy carries a risk of thrombosis and additional risk factors such as bed rest and dehydration may indicate thromboprophylaxis, as should a family history or known thrombophilia. Both unfractionated heparin and low molecular weight heparins are safe in early pregnancy as they do not cross the placenta.
Close attention should be paid to any pregnant woman with chest or leg symptoms to exclude the presence of DVT or PE by Duplex ultrasound and ventilation/perfusion lung scanning respectively, both of which do not carry any significant risk to mother or fetus.
All women undergoing caesarean section should be assessed for prophylaxis against VTE. Multiple risk factors are often present and in these cases the most effective method of prophylaxis, heparin at appropriate doses, should be used.
Midwives, general practitioners and other medical staff should take particular attention of women in the puerperium with chest or leg symptoms after vaginal delivery, to exclude the presence of DVT or potential PE.
Women with risk factors for DVT (bed rest, pre-eclampsia, other medical disorders, family history) should be carefully screened and consideration given to thromboprophylaxis.
Pulmonary embolism
Pulmonary embolism (PE) remains the single major direct cause of maternal death in the United Kingdom. The total of 46 deaths (excluding Late deaths) equates to a rate of 2.1 per 100,000 maternities compared to 1.3 in the previous Report. The comparison with previous triennia is shown in Table 2.1.
Table 2.1- Deaths from Pulmonary Embolism (excluding Late deaths) and rates per 100,000 maternities; United Kingdom 1985-96.
There are increases in all categories, though the increase in those after vaginal delivery is particularly striking.
Many of the women had risk factors for venous thromboembolism (VTE) and appropriate prophylaxis might have altered the outcome. There is still a lack of awareness about the importance of risk factors and the need for thromboprophylaxis. Although prophylactic measures against VTE can result in complications, these need to be balanced against their potential to prevent fatalities. The recommendations of the RCOG Working Party on Prophylaxis against Thromboembolism in Gynaecology and Obstetrics 1 were published in March 1995 and we have not yet had a full triennium in which their effect on PE after caesarean section can be assessed.
Age
Age has been identified in previous Reports as a risk factor for VTE. In this triennium, for those cases where there are sufficient data, one of the women was a teenager, 32 were aged 20-34 and 13 were aged 35 or over. Using the table of maternities by age band in Annex 1 to Chapter 1 the maternal mortality from pulmonary embolism for women under 35 years of age is 2.3 per 100,000 maternities in that age group, but rises to 5.0 per 100,000 for women aged 35 or older.
Antepartum deaths
Three deaths occurred after operative procedures in early pregnancy. Two were after terminations of pregnancy and one followed an operation for an ectopic pregnancy. There were no features of substandard care in these cases.
A total of 15 other patients died from PE during the antenatal period. Summaries are given in Table 2.2. The gestations were as follows:
Up to 12 weeks, 10; 13 to 24 weeks, 3; 24 weeks to term, 2.
Table 2.2 - Antepartum Deaths from Pulmonary Embolism; United Kingdom 1994-96.
Two-thirds of the antepartum deaths were in the first trimester. Obstetricians and gynaecologists are reminded that even in the first trimester pregnancy carries a risk of thrombosis and additional risk factors, such as bed rest and dehydration, may indicate thromboprophylaxis, as should a family history or known thrombophilia. Both unfractionated heparin and low molecular weight heparins are safe in early pregnancy as they do not cross the placenta.
Many of these women, however, were not under the direct care of an obstetrician at the time of their death although several had seen other specialists as outpatients and some were even inpatients at the time. Details of the cases are summarised in Table 2.2 but one case is described in more detail here:
An older multiparous woman was seen by her GP with leg pain on two consecutive days at 12 weeks of pregnancy. DVT was considered but thought unlikely. Five days later she presented with chest pain and was admitted to a general medical ward in hospital. She was seen by junior medical staff who recognised the clinical picture of DVT but did not start treatment. Her blood gases worsened but the results were thought to be an error. The next day she collapsed and died. Autopsy showed a pulmonary embolus.
There was substandard care in this case. The GP should have arranged admission for investigation when DVT was considered. The hospital medical staff should have arranged prompt investigation and treatment.
Other risk factors in addition to pregnancy were often present, and there were sometimes multiple factors in one individual. These included obesity (weight >80 kg at booking), bed rest, dehydration and in two instances a known past history of pulmonary embolus. Combinations of risk factors - such as age and operative delivery - can lead to an increase in risk greater than the additive effect of the two factors 2.
The thromboprophylactic method chosen will depend on the patient, but heparin is the most effective technique presently available. Low molecular weight heparins have not yet been proven to be more effective than unfractionated heparin in pregnancy but they have fewer side effects and only need to be given once a day. A past history of PE and/or DVT is a high risk factor for recurrence in pregnancy. Many women, particularly those with a family history, will have underlying thrombophilia, which can be found in about 50% of patients with VTE in pregnancy. Knowledge of this condition is rapidly evolving and patients with thrombophilia should be seen by specialists with particular expertise in this area. In this triennium, two women with past histories of VTE developed problems in very early pregnancy before formal booking. Previous investigation (e.g. in relation to contraception) and a warning to contact an obstetrician as soon as pregnancy is suspected would be prudent, but might not have prevented these two deaths.
It is of particular concern that five women had presented to GPs or casualty departments with symptoms of chest pain, shortness of breath or calf pain but had been discharged home. More complete investigation might have led to treatment which could have prevented the fatal PE which occurred a few days later. Most diagnoses can now be made using Duplex ultrasound scanning, which is now available in virtually all major hospitals. Its use is recommended in any case where there is suspicion, as clinical diagnosis is notoriously unreliable. Women and indeed doctors need to be reassured that nowadays only very low radiation doses are required for ventilation perfusion scanning, chest X-ray and even X-ray venography.
Close attention should be paid to any pregnant woman with chest or leg symptoms to exclude the presence of DVT or PE by Duplex ultrasound and ventilation/perfusion lung scanning respectively, neither of which carry any significant risk to mother or fetus. Deaths after caesarean section
Caesarean section, like other major surgical procedures, remains a risk factor for VTE. Fifteen deaths occurred after caesarean section in this triennium. Complete details are available for 13 of these cases, as summarised in Table 2.3, but in the other two the details and timing of death are unknown. One case is described in more detail here:
Care was substandard. As well as the risk factors of caesarean section and a low haemoglobin, the woman was a cigarette smoker. She was not seen by a consultant. When acute complications of pregnancy, such as abruption, are successfully treated, the importance of thromboprophylaxis must not be overlooked.
Table 2.3 - Deaths from Pulmonary Embolism following caesarean section; United Kingdom 1994-96.
Only two of the 13 deaths for which full details were available occurred in women aged 35 or over. Guidelines currently emphasise that such women should receive thromboprophylaxis and the relatively low number of deaths in this age group might suggest that such prophylaxis is effective. Eight of the 13 women received some form of prophylaxis - five subcutaneous heparin, and three TED stockings alone. According to the RCOG Guidelines, however, most of these cases would be classified as high risk, with three or more moderate risk factors (see the Annex to this Chapter). In such cases, the Guidelines recommend both heparin prophylaxis and leg stockings.
However, the dose of herapin recommended by the RCOG for caesarean section is relatively low. In practice it might be better to give high risk patients higher dose prophylaxsis such as unfractionated herapin 7,500 units 12 hourly, enoxaparin 40mg 24 hourly, dalteparin 5,000 units 24 hourly or other low molecular weight herapin in equivalent dose. Such doses of herapin are recommended by the RCOG for thromboprophylaxsis in the antenatal period in those at risk because of previous thromboembolism.
The timing of the PE after caesarean section shows a change from 1991-93. Most deaths occurred between 15 and 42 days, mainly in the 21-28 day interval (see Table 2.5).This represents a shift to later presentation and may reflect the effects of some form of PTE prophylaxis delaying but not fully preventing the formation of thrombus. In patients with significant risk factors, such as very obese women, consideration should be given to prolonged thromboprophylaxis, for example for six weeks after delivery.
All women undergoing caesarean section should be assessed for prophylaxis against VTE. Multiple risk factors are often present and in such cases the most effective method of prophylaxis, heparin at appropriate doses, should be used.
Deaths after vaginal delivery
There were ten deaths from PE after vaginal delivery, a marked increase on the four in 1991-93. The details of these cases are summarised in Table 2.4. No deaths from PE occurred in the first week after vaginal delivery, and most occurred between days 15 and 28 as shown in Table 2.5.
Table 2.4 - Deaths from Pulmonary Embolism following vaginal delivery: United Kingdom 1994-96.
Table 2.5 - Interval between delivery and Pulmonary Embolism; United Kingdom 1994-96.
In half the cases the patients were obese (>80kg) and most of these women weighed more than 100kg. Several had had bed rest before delivery (because of hypertension or other medical problems), or after delivery, on an ICU or in the postnatal ward. Some of the patients had complained to midwifery or medical staff of chest symptoms, shortness of breath or leg/calf pain - symptoms suggestive of developing or actual PE or DVT.
In two cases there was a family history of DVT or PE as well as signs and symptoms. These women may have had an underlying thrombophilia. This emphasises the need to take a family history at booking and to seek specialist guidance when this is positive.
Most of the deaths after vaginal delivery occurred after spontaneous, not instrumental, delivery. There is a need for the development of guidelines on thromboprophylaxis after normal delivery. The RCOG Guideline focuses largely on caesarean section. Many maternity units have now drawn up their own guidelines for vaginal delivery, whereby women, such as those over 35 and the obese, receive thromboprophylaxis.
Two deaths were probably unavoidable - one in a patient already fully anticoagulated, with a cardiomyopathy, and the other in a woman who had been fully investigated with a negative venogram two days prior to death from PE.
Midwives, GPs and other medical staff should take particular attention of women with chest or leg symptoms after vaginal delivery, to exclude the presence of DVT or potential PE.
Women with risk factors for DVT (bed rest, pre-eclampsia, other medical disorders, family history) should be carefully screened and consideration given to thromboprophylaxis.
Late deaths
Two Late deaths related directly to PE are recorded in this triennium. They are described here but counted in Chapter 14:
A teenager had a caesarean section for a failed induction with pre-eclampsia. She had prolonged bed rest and wore TED stockings after delivery. She was admitted two months after delivery with a stroke and hemiplegia, recovered and was sent home. She was readmitted two weeks later with chest pain and a PE. She was anticoagulated and was readmitted four weeks later and died. Autopsy revealed a dilated cardiomyopathy with systemic and pulmonary emboli.
Cerebral thrombosis
There were only two cases of cerebral thrombosis in this triennium. One woman had a past history of DVT and PE and the second was anticoagulated for a DVT in her current pregnancy:
A parous woman in her thirties developed a DVT a few weeks before term. She was anticoagulated and had a spontaneous vaginal delivery at term. Twelve hours later she became disorientated then unconscious. She was transferred to specialist neurological care but died a few days later from a sagittal sinus thrombosis and massive intracerebral haemorrhage.
There was a strong family history of DVT in close relatives subsequently shown to be due to Antithrombin III deficiency. This emphasises the need for adequate screening, in particular thrombophilia screening, and specialist advice in such cases.
Comments
Once again, the major cause of direct maternal deaths is pulmonary embolism. Although some antenatal deaths, and some deaths after vaginal delivery, occurred without warning and without apparent risk factors, many patients in these categories - and the majority after caesarean sections - had multiple risk factors or symptoms suggestive of PE and/or DVT.
The risk of VTE after any operation is further increased when surgery is performed in pregnancy. This was highlighted in the 1995 RCOG Report 1, which included a risk assessment chart shown in the Annex to this Chapter. The full effect of these RCOG guidelines is yet to be seen. It should be noted, however, that these guidelines highlight the need for more effective prophylaxis when multiple risk factors are present. Multiple factors were present in many of the deaths in this triennium. Such women need subcutaneous heparin as the preferred option at appropriate doses and for appropriate duration. As noted above, the risk of VTE is increased by surgery in pregnancy and this includes the first trimester. The deaths in this triennium also indicate the need for considering thromboprophylaxis in such cases.
Each unit should be advised to develop its own guideline, based on existing national guidelines, which can be applied within the requirements of their own unit.
Implementation of such guidelines in local health care contracts, care and audit plans has the potential to reduce significantly deaths from PE in the next triennium.
A recurrent finding in many cases was of a past personal or family history of VTE. There seems to be a lack of awareness of potential thrombophilia in such cases. A strong family history or personal history of VTE should be recognised as risk factors and properly investigated. Specialist advice should be sought if a thrombophilia screen is positive.
Prolonged bed rest antenatally or after delivery should prompt the use of TED stockings as a simple measure to try and reduce VTE. In this triennium there was a failure to appreciate the risk of thromboembolism after vaginal delivery in patients with multiple risk factors. Prolonged thromboprophylaxis may be advisable in patients with ongoing risk factors. Specific notice must be taken of any symptoms of breathlessness, chest or leg pain which might well herald the development of VTE.
There was also a failure to appreciate the risk of VTE in early pregnancy. Awareness of this risk needs to be increased among all doctors, not just obstetricians and gynaecologists.
References
1. Royal College of Obstetricians and Gynaecologists. Report of a Working Party on Prophylaxis against Thromboembolism in Gynaecology and Obstetrics. London: RCOG, 1995.
2. Greer I (ed). Thrombo-embolic disease in obstetrics and gynaecology. Bailliere's Clinical Obstetrics and Gynaecology 1997; 11: 403-615.
ANNEX TO CHAPTER 2 PROPHYLAXIS AGAINST THROMBOEMBOLISM IN CAESAREAN SECTION The following recommendations, taken from the RCOG Working Party Report on Prophylaxis against Thromboembolism, are of relevance to patients requiring caesarean section.
A risk assessment of all patients undergoing elective or emergency caesarean section should be performed and prophylaxis instituted as appropriate. See box below.
Management of different risk groups
Low-risk patients
Patients undergoing elective caesarean section with uncomplicated pregnancy and no other risk factors require only early mobilisation and attention to hydration.
Moderate risk patients
Patients assessed as of moderate risk should receive subcutaneous heparin (doses are higher during pregnancy) or mechanical methods. Dextran 70 is not recommended until after delivery of the fetus and is probably best avoided in pregnant women.
High-risk patients
Patients assessed as high risk should receive heparin prophylaxis and, in addition, leg stockings would be beneficial.
Prophylaxis until the 5th postoperative day is advised (or until fully mobilised if longer).
The use of subcutaneous heparin as prophylaxis in patients with an epidural or spinal block remains contentious. Evidence from general and orthopaedic surgery does not point to an increased risk of spinal haematoma.
Prophylaxis against thromboembolism in pregnancy
The RCOG Working Party also made recommendations for prophylaxis against thromboembolism in pregnancy, which are summarised in the box below.
| ||||||
