Chapter 17: General Practice

Judy Shakespeare


Although general practitioners (GPs) have been involved in the care of pregnant women for the fifty or more years of this Report, this is the first time it has been possible for a Chapter to be written by a GP for other GPs and staff working in family practice. Although this is rather late in comparison with most other specialties who care for pregnant women, this is to be welcomed because, despite the recent changes in maternity service provision, most mothers will still have contact with their GP at some point during and after their pregnancy. It is also the first time that it has proved possible for a practising GP to be assisted to find the time to review some of the maternal deaths considered in this Report in order to distil and promulgate those general lessons which are of particular relevance to primary care.

The aim of this new Chapter is to improve the care pregnant women or new mothers receive from primary care by drawing attention to the key messages contained within the overall Report that are of relevance to GPs and to highlight the key issues that affect all other staff working in community based family practice. It cannot provide an exhaustive overview of all the key findings and recommendations in the entire Report, but GPs should be aware of the major recommendations and act on the findings in this Chapter and the overarching risk factors, and key recommendations which are highlighted in Chapter 1.

In all, 66 report forms out of the 295 Direct and Indirect maternal deaths which occurred during this triennium were assessed by the GP assessor. These were cases selected as having particular relevance for general practice. As with all the other professional groups who care for pregnant women there were many examples of exemplary care and kindness provided by GPs for some very unfortunate women and their families. However, by the nature of things, the recommendations for improving practice contained here are developed largely from those cases when opportunities for excellent care were missed. Box 17.1 summaries the major issues that have emerged in relation to general practice for 2003-05.

Box 17.1

Summary of the key issues arising from maternal deaths in relation to general practice; United Kingdom: 2003-05.

Clinical issues

Communication issues

Maternity Services reconfiguration

Clinical issues

Identifying seriously ill women

In many of the cases and conditions mentioned in this chapter there was a consistent issue with failure to recognise the signs and symptoms of seriously ill women. This does not just apply to GPs, but to all other medical, midwifery and nursing staff alike. As a result a key recommendation of this Report is the introduction and use of a modified early warning system for hospitalised women as described in Chapter 19, on Critical Care. This is not appropriate for primary care as women are not under constant observation. However, recognising sick pregnant or recently delivered women in the community is extremely important as the speed of referral and subsequent treatment can affect both their and their babies’ lives.

In contrast to hospital doctors, family doctors often have the advantage of knowing the woman’s normal health and appearance. However pallor, cold extremities or looking unwell alone may not alert the “sixth sense” that something is seriously wrong. All staff, including family doctors, need to rely on abnormal clinical findings and measurements: the more a measurement deviates from normality the more the doctor should worry and act on it. In these situations a GP should arrange for a woman’s emergency admission to hospital, irrespective of the possible diagnosis. Some of the key physical signs and symptoms indicating a possible life threatening physical illness, which every health professional caring for pregnant women should be able to recite in their sleep, are given in Box 17.2.

Box 17.2  

Key signs and symptoms of possible serious illness in pregnant women or recently delivered mothers.

The following signs should alert all health professionals including midwives, GPs, junior doctors and obstetric and other consultants that serious illness is a possibility:

Recognising “red flags” when pregnant women need emergency hospital admission

This section highlights the key clinical issues that have emerged in reviewing many of the maternal deaths discussed in this Report where there was scope for improvement in GP management, due to of a lack of either knowledge or skills. Further and fuller details are contained within the individual chapters relating specifically to each cause of death and these are can also be readily downloaded from www.cemach.org.uk.

Breathlessness may be due to pulmonary embolus (PE)

The number of women who died from thromboembolism, particularly in early pregnancy, has increased in this triennium despite the continuing decline in those following caesarean section which have continued to fall as thromboprophylaxis becomes routine. Most of the women who died of pulmonary embolus (PE) had the well-known risk factors for PE, especially obesity, and could have been identified as being at higher risk in pregnancy. The risk factors are given in Box 17.3. 

Box 17.3

Risk factors for venous thromboembolisma in pregnancy and the puerperium1.

New onset or transientb
  • Previous VTE
  • Thrombophilia
    • Congenital
      • Antithrombin deficiency
      • Protein C deficiency
      • Protein S deficiency
      • Factor V Leiden
      • Prothrombingene variant
    • Acquired
      • antiphospholipid syndrome
  • Lupus anticoagulant
  • Anticardiolipin antibodies
  • Age over 35 years
  • Obesity (BMI over 30kg/m2) either pre-pregnancy or in early pregnancy
  • Parity over 4
  • Gross varicose veins
  • Paraplegia
  • Sickle cell disease
  • Inflammatory disorders e.g. inflammatory bowel disease
  • Some medical disorders e.g. nephritic syndrome, certain cardiac diseases
  • Myeloproliferative disorders, e.g. essential thrombocythaemia, polycthaemia rubra vera
  • Surgical procedure in pregnancy or puerperium
  • e.g. evacuation of retained products of conception, postpartum sterilization,
  • Hyperemesis
  • Dehydration
  • Ovarian hyperstimulation syndrome
  • Severe infection, e.g. pyelonephritis
  • Immobility (over 4 days bed rest)
  • Pre-eclampsia
  • Excessive blood loss
  • Prolonged labourc
  • Mid-cavity instrumental deliveryc
  • Immobility after deliveryc

aAlthough these are all accepted as thromboembolic risk factors, there are few data to support the degree of increased risk associated with many of them.

bThese risk factors are potentially reversible and may develop at later stages in gestation than the initial risk assessment or may resolve; an ongoing individual risk assessment is important.

cRisk factors specific to postpartum VTE only.

The following vignette is typical of a death from PE considered in this Report:

An obese, parous woman had a home booking carried out by a midwife early in pregnancy. A few weeks later she telephoned her GP because of breathlessness and was told to contact the practice again if it got worse.  The next day her husband telephoned the practice requesting a home visit as she had become worse and was told that the doctor would attend after morning surgery. The woman died a few minutes later.

On reviewing this case it was clear that the woman’s GP had failed to take an adequately detailed history of her symptoms; this is substandard care. The GP later commented that the woman’s husband did not convey a sense of urgency, but all GPs should know that sudden and continuing breathlessness in an obese pregnant woman is a medical emergency. The risks of such telephone consultations are discussed later in this Chapter. This was not the only case where a GP’s judgement was open to question:

Another multiparous woman, of normal weight, had known thrombophilia. She was managed in a joint obstetric / haematology clinic throughout her pregnancy and received dalteparin prophylaxis both during pregnancy and for six weeks postnatally. A few weeks after she stopped dalteparin she presented with an extensive deep vein thrombosis (DVT) for which she was anticoagulated with warfarin. Some weeks later she saw her GP with a cough and shortness of breath: a chest infection was diagnosed and cough medicine prescribed. She had been poor in attending for her international normalised ratio (INR) tests and admitted to forgetting to take her warfarin. On the day before her death she attended the practice nurse for a blood test: her INR result was 1.3. The next day she collapsed and died of a massive pulmonary embolus.  

This woman was known to be at high risk and was already on treatment for a DVT; her GP should have noticed that her compliance with blood tests and medication was poor. A diagnosis of a chest infection in this situation should only have been made once pulmonary embolism had been excluded by hospital investigations. This woman might have survived had she been referred promptly by her GP when she attended with a cough and shortness of breath.

Box 17.4

GP learning points: pulmonary embolism

A sudden onset of breathlessness in a pregnant or postpartum woman, in the absence of a clear cause, such as asthma, should raise the suspicion of pulmonary embolus, especially if the woman has risk factors.

Women with suspected pulmonary embolus should be referred as an emergency to hospital as the diagnosis of pulmonary embolus can only be made or excluded by secondary care investigations.

Severe headaches may be suggestive of pre-eclampsia or cerebral haemorrhage

Although, fortunately, in this triennium no women died from eclampsia because their GP had failed to identify headaches or hypertension indicative of pre-eclampsia or eclampsia, this has not always been the case. Severe headaches in pregnancy can also be indicative of intracerebral bleeding. During 2003-05, 21 pregnant or recently delivered women died from intracerebral haemorrhage, 12 of which were due to subarachnoid haemorrhage, a condition for which pregnancy, and hypertension in pregnancy, are risk factors. The following is a typical case:

After a normal pregnancy and birth, a mother developed a severe headache with new onset hypertension early in her puerperium. Her headache was not relieved by analgesics and was described as very severe. The midwife reassured the mother but she still had a very painful headache two days later: no action was taken. Her midwife had planned to review her again four days later but, before that, she was admitted to the Emergency Department (ED) with a fatal subarachnoid haemorrhage.

Although she had not been seen by a GP, it is worth emphasising that a severe new onset headache, the worst a patient has ever described, must be taken seriously. This is especially the case for pregnant or recently delivered women. Patients with fatal subarachnoid haemorrhage often have preceding warning headaches. The indications for emergency referral of women with headaches given in the learning point box below are taken from the PRECOG guidelines2 and are included as a brief reminder of good practice. The other recommendations are taken from the findings in the relevant Chapters, 3 and 10, of this Report.

Box 17.5

GP learning points: severe headaches in or after pregnancy/ pre-eclampsia and cerebral bleeds

Women with a headache severe enough to seek medical advice or with new epigastric pain should have their blood pressure taken and urine checked for protein as a minimum.

Women with severe incapacitating headaches described as the worst they have ever had should have an emergency neurological referral for brain imaging in the absence of other signs of pre-eclampsia.

The threshold for same day referral to an obstetrician is hypertension ≥ 160mmHg systolic and or ≥ 90 mm Hg diastolic or proteinuria ≥1+ on dipstick. The systolic BP is as significant as the diastolic.

Automated blood pressure machines can seriously underestimate blood pressure in pre-eclampsia. Blood pressure values should be compared with those obtained by auscultation (an anaeroid sphygmomanometer is acceptable).

Ectopic pregnancies continue to be missed, and can mimic gastroenteritis

Ten women died from ectopic pregnancy; care was sub standard for seven. The two previous Reports have highlighted the lesser-known presentations of ectopic pregnancy, especially with vomiting and diarrhoea in the absence of vaginal bleeding3 4. It is important that GPs are aware of the possibility of ectopic pregnancy in all women of childbearing age. Avoidable deaths continue:

A young woman developed abdominal pains, diarrhoea and vomiting (D&V) with one episode of fainting and a history of amenorrhoea. She had telephone contact with an out of hours GP but was not visited. The next day she had a home visit as her D&V persisted but no action was taken and the possibility of an ectopic pregnancy was not considered. The following day a relative asked for another home visit as she had been “restless” overnight, although her diarrhoea had settled. She was found dead in bed by the GP when he visited later in the day. At postmortem she had a ruptured ectopic pregnancy.

Gastroenteritis is common in the community, but clinicians should be aware that in women of child bearing age diarrhoea should raise suspicion of an ectopic pregnancy, especially if there has been abdominal pain, amenorrhoea or episodes of fainting. The absence of vaginal bleeding does not exclude an ectopic pregnancy. Once again, this case highlights the potential perils of telephone consultations.

Box 17.6

GP learning points: ectopic pregnancy

Clinicians in primary care need to be aware of atypical clinical presentations of ectopic pregnancy and especially of the way in which it may mimic gastrointestinal disease.

Fainting in early pregnancy may indicate an ectopic pregnancy.

Puerperal fever is not a disease of the past

For many years puerperal sepsis, “childbed fever”, was a leading cause of maternal death and its signs and symptoms were widely known. The advent of antibiotics means that this illness has largely disappeared from the collective memory of health professionals and patients. However, women continue to die from puerperal sepsis in this, as in previous Reports. For example:

A woman had early discharge from hospital after normal delivery despite having pyrexia and tachycardia. Her community midwife visited daily for a few days but failed to recognise the signs and symptoms of developing sepsis. Despite the mother’s continuing pyrexia, a tachycardia of 140 bpm, abdominal pain and diarrhoea, the midwife took no action and advised that she would phone the following day and visit again in a few days time. An emergency GP was called who advised analgesia and fluids. The woman was admitted to hospital a few hours later and died rapidly from septic shock.

In this case both the midwife and GP provided substandard care. In the early stages of sepsis symptoms can be insidious in onset and non-specific. Fever and offensive lochia are not always present; diarrhoea is common. By the time sepsis is obvious clinically, infection is already well established and clinical deterioration into widespread septicaemia, metabolic acidosis, coagulopathy and multi-organ failure is very rapid and often irreversible. The best defence against this situation is awareness of the early signs of sepsis by routine basic clinical observations for several days after delivery; these should include pulse, temperature, BP, respiration, and lochia. The diagnosis of sepsis should always be considered in recently delivered women who feel non-specifically unwell. A white cell count, urine and vaginal cultures should be taken and the woman should be treated with antibiotics proactively while awaiting the results of investigations. There should be a low threshold for emergency admission for intravenous antibiotics if a woman fails to respond to treatment or has symptoms such as abdominal pain and fever or tachycardia. Good communication is essential between hospital and community carers.

Box 17.7

GP learning points: sepsis

Puerperal infection is not a disease of the past and health professionals are still failing to recognise its classic early symptoms and signs.

Puerperal sepsis should be considered in all recently delivered women who feel unwell and have pyrexia.

Women with sepsis can deteriorate rapidly, with the potentially lethal consequences of severe sepsis and septic shock. Abdominal pain, fever and tachycardia are indications for emergency admission for intravenous antibiotics.

Heartburn may be ischaemic heart disease

Sixteen women died from ischaemic heart disease, representing a continuing rise in the number of cases. This reflects both the trend towards older motherhood as well as public health risk factors such as smoking and the rising incidence of obesity. Therefore, although cardiac disease is rare in pregnant women, GPs need to consider the possibility in women with risk factors and typical symptoms. For example:

A parous woman who smoked and who had a family history of ischaemic heart disease presented late in the second trimester with a history of retrosternal pain radiating to the jaw. Her GP attributed the symptoms to heartburn. She collapsed and died the next day from a myocardial infarction.

This woman’s care was substandard: she had risk factors and classical angina pain. Her GP should have arranged for an emergency cardiological referral. Although heartburn is a common symptom of pregnancy it can mask serious pathology. Radiation of retrosternal pain to the jaw is rare in heartburn and should alert clinicians to the possibility of ischaemic pain.

Box 17.8

GP learning points: Heartburn and ischaemic heart disease

The prevalence of ischaemic heart disease in pregnancy and the puerperium is increasing.

The possibility of ischaemic pain should be considered in women who have risk factors and atypical heartburn.

If a woman is suspected of having cardiac chest pain she should be admitted as an emergency.

Recognising urgent conditions when pregnant women need “fast track” referral to secondary care

Apart from the “red flag” conditions and signs and symptoms discussed above, there were other instances when women with serious medical conditions were not referred urgently by their GP and where such a referral may have changed the outcome.

Congenital cardiac disease

More women who have had surgery for significant congenital heart disease are reaching an age when they become pregnant. Fallot’s tetralogy is the commonest form of cyanotic heart disease (1:3600) and surgically repaired tetralogy of Fallot is probably the commonest condition likely to be seen in general practice. A woman and her GP may not appreciate the cardiovascular risks of pregnancy:

A young woman had a repair for a congenital heart condition and was under paediatric cardiac follow up every few years. She was well and asymptomatic and due for transfer to adult congenital heart disease services. When she became pregnant she saw her GP early in pregnancy and was referred to the local specialist “teenage” midwifery service. She collapsed and died in the second trimester, probably from an arrhythmia.

In this case the GP should have referred her urgently for a cardiological opinion at the start of pregnancy. And, although it would be helpful for her to have the specialist support that a teenage pregnancy service could provide, she clearly needed more specialist medical care. Paediatric cardiologists and GPs have a responsibility to counsel their teenage patients at any opportunity about the risks of pregnancy and to seek specialist care as early as possible should the patient wish to become pregnant. They should also be proactive in discussing the need for adequate contraception. A consensus view from the 51st RCOG Study Group on heart disease and pregnancy reads:

“A proactive approach to preconception counselling should be started in adolescence and this should include advice on safe and effective contraception. Proper advice should be given at the appropriate age and not delayed until transfer to the adult cardiological services5”.

Eleven women died from epilepsy; six of the deaths were from sudden unexpected death in epilepsy (SUDEP). Previous Reports have emphasised that pregnant women with epilepsy should have prompt specialist care from a consultant obstetrician and a neurologist or specialist physician with an interest in epilepsy and pregnancy. Pregnant woman should also be seen as soon as possible after such a referral is made. In this triennium there were examples where urgent referrals were either not made by GPs or not responded to by neurologists:

A woman with a history of epilepsy since childhood had experienced intolerable side effects from anticonvulsants so stopped treatment. When she became pregnant she did not want to try newer drugs because she was concerned about teratogenicity. She saw her GP after she had a single fit, then again at the start of her second trimester, by which time she had had a few more fits. Her GP did not refer her back to her neurologist. She was referred urgently, by letter, by the obstetric registrar. He had found out about her fits because she reported them to the radiologist undertaking her routine fetal anomaly scan. The neurology appointment was given for a month later, by which time she had already died from SUDEP.

Epilepsy is dangerous in pregnancy because it may be more difficult to control. This may be because therapeutic drug levels decline or because women may be reluctant to take their medication. Women who stop anticonvulsant therapy in pregnancy must be made aware of the risk of SUDEP. Here the GP missed the opportunity to refer her both at the start of pregnancy and urgently when her fits started. The obstetric registrar made an appropriate referral when he saw the woman, but a phone call would have been more effective than a letter. The neurologist did not recognise the urgency of the situation and responded too slowly.

Mental health problems during pregnancy and /or after delivery

Ninety-eight pregnant women or new mothers who died within a year of delivery, from whatever cause, were affected by or died as a result of a psychiatric disorder. These include deaths from suicide and drug overdose and violent deaths from murder, accidents etc. Although many of these deaths were not directly related to pregnancy these figures highlight the impact that ongoing mental health and substance abuse problems, particularly amongst the more vulnerable, have on maternal health. This Report has been recommending for the last few years that women are routinely asked in early pregnancy about significant severe past mental health problems, and, if necessary, referred to the local perinatal psychiatric services to develop pre-emptive care and management plans. These policies have been widely adopted and it is gratifying to see that the number of women who died from suicide from a perinatal mental illness is significantly reduced in this triennium.

It needs to be more widely known that women who have had a previous episode of a serious mental illness either following childbirth or at other times, are at an increased risk of developing a postpartum onset illness even if they have been well during pregnancy and for many years previously. This risk of recurrence is estimated at least one in two for women with previous puerperal psychosis, as discussed in detail in Chapter 12. It is also known that a family history of bipolar disorder increases the risk of a woman developing puerperal psychosis following childbirth. The last two Reports 3, 4 found that over half of those women who died from suicide had a previous history of serious mental illness. Even if this history is known it may be communicated poorly to other members of the team, including the midwife. For example:

A young woman with existing children killed herself by violent means some weeks after the birth of her next child. She had a previous history of bipolar disorder with a number of hospital admissions for mania. She was last seen by psychiatric services shortly before the beginning of pregnancy and stopped her mood stabiliser medication at about this time. Her GP had mentioned her previous psychiatric history in his referral letter to the obstetrician, but the midwife was unaware of this history at the booking clinic. The woman was well during pregnancy, but no steps were taken to monitor her mental health following delivery.

This woman faced a risk of recurrence of at least 50% in the first three months after delivery. Her psychiatric team should have made the woman, her family and her GP aware of this. When she became pregnant again her GP should have told her midwife and should have referred her back to the psychiatric team so that a management plan could have been put into place. She should have restarted her medication after delivery and been kept under close psychiatric supervision. If these interventions had been made, her death might have been avoided.

In other situations a pattern of escalating self harm may develop prior to suicide. Health professionals may be slow in offering interventions, they may not appear to take women’s threats seriously or hospital based services may be inaccessible for a woman who is distressed:

A single mother with a number of children had a history of self harm and alcohol abuse. She had a miscarriage following which she repeatedly attended the Emergency Department (ED) with overdoses, alcohol intoxication and deliberate self harm. Her GP had received calls from her family expressing concern about her state. The severity of her self-harm escalated before her death, some months after her miscarriage. Shortly before she died her children were put on the child protection register after a case conference. Family members were so concerned about her that they rang the GP and he arranged for a psychiatric review the next day. She failed to attend and killed herself some days later. A subsequent psychiatric report said she found it difficult to engage with services, “maybe because of associated alcohol problems”.

Her care was substandard because the GP failed to recognise and act on her escalating self-harm despite concerns expressed by family and multiple attendances at hospital. This case is complicated by her dual diagnosis of psychiatric illness and alcohol misuse. The GP did not appear to involve other members of the team such as the health visitor. Whilst this may have made no difference to the outcome there was a sense from the case report that everyone thought she was a “hopeless” case in any event.

Another woman had poor care:

A vulnerable woman had a family history of suicide, was the victim of domestic abuse and found it hard to engage with services. Antenatally she only saw her midwife because she insisted on home visits. The community mental health team (CMHT) closed her case because she did not attend appointments. Her GP saw her once after delivery, made a diagnosis of postnatal depression and prescribed citalopram but did not arrange any follow up. The health visitor carried the burden of care, appeared to be working in isolation and was unsupported. The woman killed herself some weeks later.

Here the CMHT did not recognise that the woman’s failure to attend appointments might mean that she was so depressed that she was unable to leave the house. Failure to attend should be a “red flag” of severity requiring prompt outreach care rather than discharge. It is also possible that a specialist perinatal team may have managed this woman more actively. Her GP failed to recognise the severity of her depression when he saw her postnatally and failed to arrange follow up after starting antidepressants for her depression. Her health visitor was unsupported. Better team work may have built a support network around this woman and allowed her to access appropriate care that may have prevented her death.

Substance misuse and pregnancy

Around 11% of all the women whose deaths were assessed during this triennium, from any cause, had problems with substance misuse. They did not all die from the drug misuse itself, but also from associated physical problems and from accidents and murder. These women are vulnerable because they often have complex and chaotic social lives; they often suffer domestic abuse, are hard to engage in antenatal care and usually fail to develop trusting relationships with health professionals. There was also much evidence of good relationships and excellent care by GPs. However, there were also instances of GPs apparently working beyond their level of expertise which may have contributed to the outcome:

A young single parent with a history of polydrug misuse and psychiatric problems attended infrequently during her pregnancy but claimed to have stopped all her drugs by the third trimester. Although her baby was delivered prematurely and admitted to special care, the mother discharged herself immediately after the delivery. The GP then supervised her postnatal methadone prescribing, unsupported by a specialist in substance abuse. The administration of methadone was not supervised by the pharmacist. She died from a methadone overdose some weeks after delivery, shortly after a case conference at which the baby was placed on the “at risk” register.

This GP was working beyond his/her level of expertise and should have referred her to a specialist drug team. It may be that the woman refused such care, but their advice should have been sought at the very least. She may have left hospital immediately after delivery to re-establish her drug supply. Lack of supervised consumption postnatally may have contributed to an unreliable supply of opiates and diversion to the black market. In this case the involvement of social services may also have been a factor in her death and this risk factor for suicide is discussed in more detail in Chapter 12.

Box 17.9

GP Learning points: mental health and substance misuse

NICE have recently produced guidance on the management of antenatal and postnatal mental health and all GPs should be encouraged to read, and follow them6.

It may be safer for women who misuse drugs to continue their maintenance treatment during and after pregnancy to prevent inadvertent overdose from street drugs.

Pregnant drug users may be difficult to engage in treatment. This may be because they do not recognise pregnancy, have chaotic lifestyles or have fears about the consequences of social services involvement.

Pregnancy in refugees and asylum seekers

Women who have recently arrived from countries around the world, particularly those from Africa and the Indian sub-continent, but increasingly from central Europe, tend to have poorer overall general health and are at risk from illnesses that have largely disappeared from the UK, such as TB and rheumatic heart disease. They are also more likely to be at risk of HIV infection. Some may also have suffered female genital mutilation/cutting (FGM/FGC). All of these conditions, alone or in combination, contributed to a number of the maternal deaths considered in this Report. If newly arrived women are unwell, GPs may need to consider unfamiliar possibilities in their differential diagnoses.

Rheumatic heart disease

No pregnant women have died from the consequences of rheumatic heart disease since the 1991-1993 Enquiry7. However, in this current triennium, two immigrant women died from mitral stenosis as a result of rheumatic heart disease. Rheumatic heart disease is likely to become even more common with increasing numbers of women who are asylum seekers or refugees, who may never have had a cardiac assessment.8

A previously well young immigrant woman with poor English booked for midwifery-led antenatal care and was seen only by her midwife and GP. Her GP did not examine her heart. She was admitted to an ED at the end of her second trimester with cough, breathlessness and chest pain. Even though she was admitted, the diagnosis of mitral stenosis was not considered until she was moribund.

If she had been examined by her GP the murmur of mitral stenosis may have been picked up. It can be hard to hear the soft diastolic murmur of mitral stenosis if a woman is already sick and has a tachycardia. Women with mitral stenosis may seem well in early pregnancy but they commonly decompensate at the end of the second trimester. Referring women with unexplained murmurs to a cardiologist early in pregnancy would enable earlier diagnosis of rheumatic heart disease and catastrophic deterioration may be prevented with proper assessment, monitoring and intervention.


Two cases of tuberculous meningitis in pregnancy were diagnosed late, as often happens, and both occurred in women whose families were from the Asian sub-continent. In one case the diagnosis was delayed as the husband was acting as the interpreter, a recurrent feature amongst other deaths in women who could not speak English. In one case her GP commented that this was a particular problem in his practice because there was no agreed source of funding for interpreters.

Box 17.10

GP learning points: refugees and asylum seekers

Newly arrived women, especially refugees and asylum seekers, are at higher risk of illnesses that are no longer familiar in the UK and this needs to be born in mind when caring for sick women from these communities.

A medical assessment of general health before booking of immigrant women may prevent death later in pregnancy. This should include a cardiovascular examination, performed by an appropriately trained doctor, who could be their usual GP.

The risks of obesity in pregnancy

Obesity represents one of the greatest and growing overall threats to the childbearing population of the UK. Fifteen percent of women who died from Direct or Indirect causes and who had a BMI recorded had BMIs of 35 or over, with half of these having BMIs exceeding 40. A further 12% of women had BMIs in the range 30-34 and and 24% had BMIs of 25-29. Obese women predominated among those who died from thromboembolism, sepsis and cardiac disease. There are many other aspects of the care of overweight women in pregnancy that cause concern beyond maternal risks, including the difficulties of prenatal diagnosis, the enhanced risk of gestational diabetes, the increased chance of caesarean section, and the challenges of analgesia and anaesthesia. The risks of obesity are discussed in more detail in Chapter 1.


Problems in communication are at the heart of many of the cases discussed in this Report and this Chapter has already raised issues about communication with patients, within the primary health care team and between GPs, midwives and specialists.

Telephone consultations

Telephone consultations are increasingly being used in medical contacts, including the triage of acute illnesses. They are acceptable to patients, and clinicians also value them but have anxieties about missing serious conditions9. There are cases described in this Report, and earlier in this Chapter, which underline this concern. The case of a woman who died from pulmonary embolus, described earlier, is an example where the quality of telephone assessment was poor and may have contributed to her death. There is evidence that telephone consultations are shorter than face-to-face consultations10 but there is little evidence about the quality of care. Telephone consultations require an additional range of skills since the importance of verbal cues and focussed history-taking need to compensate for the inability to examine the patient. The BMA recommends that:

“consulting over the telephone should normally be modified to allow the patient greater time to explain their problem. The doctor should also take a detailed history and seek the answers to all the relevant direct questions. There should be a summation and agreement with the caller/patient as to what exactly the problem is that the doctor is attempting to solve. The doctor should explain their assessment and detail the action s/he intends to take. If it is not possible to safely manage the patient over the telephone, the doctor should arrange a face to face consultation and make an appropriate referral.11” 

Doctors may need specific training in telephone consultations, an area that is currently neglected in the training and professional development of GPs12.

Referral letters; providing complete information

General practitioners are the only professionals who have access to a woman’s complete medical history and as such are the only health professionals able to provide a complete medical, psychiatric and social history. It is therefore crucial that all relevant information is included in referral letters to enable appropriate and planned care. These Reports have regularly highlighted examples of where inadequate information in referral letters led to adverse consequences for pregnant women and this triennium is no exception. A GP has a responsibility to ensure that any relevant history is conveyed in as much detail as possible to the midwife and/or obstetric team who will be caring for the woman during pregnancy.

Strategic changes in delivery of care
Increasing midwifery-led care

There have been several changes in service delivery over the period of the Enquiry which provide challenges in caring for pregnant women. The recent implementation strategy for the National Service Framework for Maternity Services13, “Maternity Matters”14 in England will result in all low risk women being offered a choice of midwifery-led care before, during and after childbirth. The lack of financial incentive for GP involvement in obstetric care under the 2004 GP contract has also led to many GPs becoming more distanced and less involved in maternity care. Maternity care has traditionally been a valued part of a GP’s work, so they have often been unhappy about this change. For the period of this Enquiry only 3% of the women who died were reported to be receiving care “shared between midwife and GP” and this direction looks set to continue. This therefore raises some crucial issues for GPs and midwives in providing maternity care.

Booking low-risk women

One challenge is how a woman can be judged to be “low risk” at booking. Following a firm recommendation in the last Report about this, the National Institute for Clinical Excellence is preparing generic consensus guidance on this, which is due to be published in early 200815. There may be medical, mental health or other problems that a woman may not appreciate, whose importance she does not understand, or that she fails to disclose. The GP is the only professional who has access to a woman’s complete medical history. In addition a GP has particular skills in understanding and managing risk, handling uncertainty and recognising the early stages of disease. In order to undertake a proper risk assessment midwives need access to the electronic and paper GP record. If this is not possible GPs should be willing to give a copy of the medical summary to either the woman or the midwife and to discuss any issues that may be of significance in the pregnancy. It would be good practice for the GP and midwife to explain the reason for this in their discussions with the woman.

Woman-held records

Most maternity trusts enable women carry their own maternity records. It is important that all the correspondence relating to any woman’s pregnancy should be available to every professional who is caring for her, particularly if the pregnancy is higher risk. It is now common practice for GPs to scan any correspondence into their records but they should ensure that any relevant scanned letters are also available in the hand-held record. Likewise, it would be good practice for specialists to send a copy of any letters to the woman and her midwife for inclusion in the patient-held record.

Communication of abnormal results

Midwifery-led care may mean that the results of investigations performed by midwives are not automatically reviewed by the GP:

A young refugee woman who had just arrived in the UK had a blood count that showed a pancytopenia which her GP thought was “cultural”. She soon became pregnant, was booked and had bloods taken by the community midwife. The GP states that he never received copies of the antenatal results. At the end of her first trimester the woman developed an upper respiratory tract infection and was treated with amoxicillin. A few days later she was referred to the ED and admitted with a history of haemoptysis, fever and general aches and pains. She was found to have acute myeloid leukaemia and she died of overwhelming infection a few days later.

If her GP had realised the significance of the initial abnormal result, or had received a copy of the antenatal results, he might have been able to refer this women earlier so that she did not develop a life threatening infection. With the shift to midwifery-led care it is important that the GP’s name is automatically put on all bloods taken by the midwife.

Out of hours (OOH) care

There have been major changes in out of hours (OOH) care since the introduction of the new GP contract in 2004 when most registered GPs relinquished 24-hour responsibility for their patients. There are some indications that this could have consequences for the quality of maternity care. One problem may be communication with the woman’s usual GP practice:

A woman in early pregnancy, with no past medical history, called an OOH doctor with a history of breathlessness, anorexia and vomiting. She also had a two-week history of pain in her left leg. She was reassured during the telephone consultation but collapsed and died of a pulmonary embolus a few days later. Her usual GP had received no record of this contact with the patient from the OOH service.

The emergency GP should have considered the diagnosis of pulmonary embolus with the history obtained and the OOH organisation failed to pass information about the contact to her usual GP. This meant her usual GP had no opportunity to reconsider the diagnosis the next day.

Changes to OOH services also mean that patients may delay contacting a GP until they know their usual GP is available. This is a particular worry over weekends or Bank Holidays, when reassessment by the usual GP is impossible. The usual GP needs to ensure that s/he anticipates potential problems before long weekends:

A young pregnant woman who could not speak English was seen just before a Bank Holiday with bronchopneumonia and treated with antibiotics. Her GP thought he had “safety netted” by telling her husband to take her to the ED over the weekend if her condition worsened. The husband did not contact the OOH service and she was not seen again until four days later when she was seriously ill and admitted to hospital. She died shortly afterwards of fulminating pulmonary TB.

In this case the GP care was substandard; her usual GP, who saw her before a Bank Holiday failed to recognise how ill she was. He should have admitted her to hospital directly if early reassessment was impossible or communicated his worries about the patient directly to the OOH services. It is inappropriate to expect sick patients or their relatives to do this.

Box 17.11

GP learning points: out of hours (OOH) care

OOH services and usual GPs need to be able to communicate with each other rapidly and effectively, preferably electronically or by fax.

OOH services and usual GPs need to maintain records of all contacts with patients both within and between each service.


It is ironic that this Chapter, the first to be written specifically for GPs, comes at a time when GPs are no longer the main providers of antenatal care for women with low risk pregnancies. Nevertheless the contributions that they can make are still very significant: GPs are “experts” in managing uncertainty, the early presentation of illness and in managing and minimising risk. There is a risk that changes in midwifery care will lead to GPs becoming de-skilled, although they will still be the first to be involved if the family or midwife suspect something may be wrong. This role needs to be recognised and encouraged. They need to maintain their skills and professional development to be able to provide excellent care for all pregnant or recently delivered women, including those at higher risk or in emergency situations.

Even if they are no longer the lead carer, GPs still have a duty of care for pregnant women and should be interested in their health and well-being as they will be caring for these women, and their families, for many years to come. GPs should therefore not only make sure that the midwives or specialists caring for their pregnant women are as fully informed as possible of any past or current medical, psychological or social problems, but should also give them access to her complete case notes on request.

It is hoped that the learning points and recommendations in this Chapter will help to maintain and improve the care that GPs can provide.