SJ Kotecha, Cardiff University; P Banfield, Consultant Obstetrician, BCUHB; Chairman, Welsh Initiative on Stillbirth Reduction
- Key messages
- Risk factors for stillbirth
- Cause of death in stillbirths
- Stillbirths and intrauterine growth retardation
- Table 1 Rate of IUGR in stillborn fetuses classified as unexplained death prior to onset of labour or unclassifiable by CP classification born to Welsh residents, 1993-2012 for LMSgrowth and 1996-2012 for GROW
- Risk – high, low, but never ‘no’ risk
- A note on intrapartum electronic fetal monitoring
- The stillbirth rate in Wales in 2012, excluding late terminations, was 4.5 / 1000, but this rate is not significantly different from 10 years ago and is twice as high as some European countries.
- There are a number of risk factors identified that can lead to increased surveillance and intervention, such as previous stillbirth, obesity and smoking. Some are more amenable to change than others, but we need a specific strategy to lower and / or mitigate risk where possible.
- The current way that deaths are classified is unsatisfactory, because a large number of cases (43%) have no cause of death attributed, which makes advice about future pregnancy less certain. A post-mortem can help identify what happened this time and sometimes adds information to help with any future pregnancy.
- It is possible that we are not diagnosing placental problems that lead to a reduction in fetal reserve and fetal growth restriction, where knowing this antenatally would result in a different management plan.
- There is a requirement to make sure that women, without any specific identified risk, are listened to – particularly with regards to reduced fetal movement.
- The Welsh Initiative on Stillbirth Reduction (WISR) is the national programme to implement evidenced based interventions via the 1000 Lives Plus Transforming Maternity Services mini-collaborative.
In this year’s report we present and discuss up to date data on stillbirths, and analyse in more detail one aspect of stillbirths – intrauterine growth retardation in stillbirths classified as unexplained death prior to onset of labour or unclassified by CP classification code. All the Welsh data presented and discussed here is given in the data section of the report. In the UK stillbirth is defined as late fetal death from 24 weeks gestation. The stillbirth rate in Wales in 2012, including late terminations, was 5.1 per 1,000 registrable births, which is higher than the 2011 rate of 4.7 per 1,000 births and the annual rate for the combined three years 2009-2011 (5.0 per 1,000 births). Data on stillbirths in other parts of the UK are at present unavailable for 2012. The stillbirth rate in 2011 was 5.2 per 1,000 registrable births in England (ONS), 5.1 per 1,000 registrable births in Scotland, 3.6 per 1,000 registrable births in Northern Ireland1 and 4.7 per 1,000 in Wales. All these rates include late terminations.
The stillbirth rate in Wales excluding late terminations in 2012 was 4.5 per 1,000 registrable births, which is not significantly different to the previous year (4.2 per 1,000 registrable births in 2011). The slowly declining trend of recent years seems to have reached a plateau. Similar trends are observed for stillbirth rates in Health Boards and regions in Wales although there are variations in rates within regions. Stillbirth rates between maternity units are difficult to compare because of the wide confidence intervals arising from this relatively infrequent occurrence (1:200 pregnancies). However, stillbirth is 10 times more frequent than cot death and 3 times more common than Down Syndrome and it is now unacceptable for us not to do more to prevent it.
Within Europe data on stillbirths are available for 2010, collated in the European Perinatal Health Report2. Stillbirth rates (from 28 weeks gestation) ranged from under 2.0 per 1000 live births and stillbirths in the Czech Republic and Iceland to 4.0 or more per 1000 in France, Lativa, the region of Brussels in Belgium, and Romania. The countries of the United Kingdom also had relatively high stillbirth rates at 28 weeks and above 3.8 per 1,000 in England and Wales, 3.6 in Scotland and 3.4 in Northern Ireland. Differences in ascertainment and registration may contribute to some of this observed variation so that direct comparisons between countries may be inaccurate3. However, it is generally accepted that the UK has high stillbirth rates compared to similar European countries. Scandinavian countries have far lower stillbirth rates (2 – 3.5/1000). The populations are not directly comparable because of, for example, the association between deprivation and increased stillbirth rates, but Wales aspires to significantly reduce the number of babies stillborn each year. 4
Risk factors for stillbirth
Previous CESDI confidential enquiries5 found that 45% of stillbirths were associated with ‘suboptimal’ care. The five most frequent areas of suboptimal care remain:
- Assessment and communication of risk by and between doctors and midwives
- A failure to take into account a previous pregnancy with intrauterine growth restriction or to suspect or detect it, or a failure to manage this appropriately
- A failure of women to appreciate the significance of reduced movements of their baby, to report this in a timely manner or of the clinical team to respond appropriately
- A failure of women to engage with advice on smoking cessation or for services to support this to be provided or for health professionals to refer to such programs
- A failure of the clinician to suggest, or gain parental consent, for post mortem. Specialist pathological services for Wales are provided in Cardiff.
Although public health initiatives are making some inroads into smoking in pregnancy, the numbers of women smoking in pregnancy are alarming 6. Wales also needs to find a way to tackle obesity and poverty if it is serious about reducing stillbirths. Together with advanced maternal age (which obviously cannot be modified) the extra risks can be mitigated with better antenatal surveillance and lower thresholds for investigation and intervention.
Within Wales, stillbirth rates are persistently higher in the most deprived quintile of social deprivation measured using the Welsh Index of Multiple Deprivation (WIMD_2008). Similar trends have been observed in England7. In Scotland a report published in 2009 on the trends in perinatal mortality in Scotland over 30 years reported the stillbirth rate for women from areas of high deprivation (Depcat 5) was 7.22 per 1,000 total births for the years 2005/7 and lower 4.67 per 1,000 total births for women from areas of low deprivation (Depcat 1).8 The difference in rates has remained constant from 1992 to 2007. ONS data for 2011 for England and Wales using father’s occupation and employment status as a proxy for socio-economic status hence excluding births which were registered by the mother alone reported father’s classified as “other”, a mixed group with the highest rates for stillbirths. The second highest rates were observed for married fathers in routine occupations (5.9 deaths per 1,000 total births) and unmarried fathers in semi-routine occupations (5.5 deaths per 1,000 total births) (ONS).
There are independent associations between increasing parity and advanced maternal age with increased risk of stillbirth. Recent decades have seen an increase in mean maternal age at childbirth in many high income countries.9 ONS data for 2011 for England and Wales showed that babies of mothers aged 40 years and over had the highest stillbirth rate at 8.0 per 1,000 total births, the second highest rates were observed in women aged less than 20 years (6.6 deaths per 1,000 total births). In Scotland a report published in 2009 on the trends in perinatal mortality in Scotland over 30 years reported that the singleton stillbirth rate for mothers over 40 years of age was considerably higher in comparison to women aged 25 to 34 years.8
A systematic review of major risk factors for stillbirth in high income countries has identified maternal overweight and obesity (body-mass index >25 kg/m2), advanced maternal age and maternal smoking as the highest ranking modifiable risk factors.
Overall in the UK smoking accounts for 7% of all stillbirths but it is estimated that in disadvantaged populations maternal smoking contributes to up to 20% of stillbirths10. Data from the Infant Feeding Survey for Wales 201011 show Wales has the highest rate in the UK for smoking during pregnancy and 16% of pregnant women smoke throughout pregnancy. Encouragingly, an ONS report on the 2011 General Lifestyle Survey reports among women aged 16 to 49 years of age pregnant women were less likely to be smokers than women who were not pregnant or unsure if they were pregnant. In addition pregnant women were more likely to be ex-smokers than women who were not pregnant or unsure if they were pregnant suggesting many women give up smoking when pregnant.
Data published in the CMACE report on maternal obesity shows that Wales has the highest prevalence of severe obesity (BMI 35+) in pregnancy in the UK at 6.5%, compared with 5.5% in Scotland, 4.9% in England and 5.3% in Northern Ireland12. The study found that the stillbirth rate in women with a BMI 35+ (8.6 per 1,000 singleton births) was twice as high as the overall national stillbirth rate (3.9/1,000 singleton births), and that the risk of stillbirth increases with increasing obesity.
In a tertiary referral unit in Northern Ireland in singleton pregnancies over an 8 year period, 2004-2011, women with a BMI ≥40 were at increased risk of stillbirth (OR 3.0, 99% CI 1.0-9.3)13 compared to women of a normal weight.
These findings highlight the importance of public health initiatives to tackle smoking and obesity in women of reproductive age.
Cause of death in stillbirths
The two most common classification systems for stillbirth in the UK attempt to provide a clinical correlation with pathological findings at post-mortem. Modification of these approaches led to the CMACE classification used in this current AWPS report. That there is a failure to identify a specific cause in 43% of cases reported last year in Wales suggests that we need a system that is more directly meaningful to both parents and clinicians.
Other classification systems have been developed in Scandinavia and Australia / New Zealand (CODEC), but there is also great interest in the work from the West Midlands Perinatal Institute, who developed the Re/Co/De classification that searches for and recognises abnormal fetal growth from dysfunction of the utero-placental unit, through the use of ‘customised fetal growth charts’ and detailed pathological examination of the placenta and fetal organs to look for specific evidence of this. When using this approach they report the number of unexplained or unclassified stillbirths falls to 15%. It is likely that a combination of these latter systems will form the basis the new UK MBRRACE reporting system.
The findings of previous CESDI reports in relation to possible alternative outcomes to known interventions and improvements in classification would suggest that ‘unexplained’ does not equate to ‘unavoidable’.
In 43% of stillbirths in Wales in 2012 no antecedent or associated obstetric causes were identified using CMACE classification. The low autopsy rate and low rate of pathological examination of the placenta contributes to the high proportion of stillbirths that are classified as ‘unexplained’. In Wales in 2012 data regarding whether or not consent for an autopsy was requested from parents was available in 95% (173/182) of stillbirths. Of these 173 cases autopsy was requested by clinicians in 99% of cases. Parents gave consent in 47% of these cases.
Where it was possible to identify a cause, antepartum haemorrhage and congenital anomalies were leading causes of stillbirth. Screening and monitoring in pregnancy are used to identify high risk pregnancies to provide appropriate clinical management. However a recent systematic review of screening and monitoring interventions in pregnancy has reported there is limited evidence for the impact of these interventions on stillbirth14. Screening and interventions to reduce antepartum stillbirth as a result of placental dysfunction has been identified as a priority for future research15-17.
Stillbirths and intrauterine growth retardation
This year we focus on one aspect of stillbirths. Stillbirth rates have remained static this year. Gardosi et al. have found that many stillborn fetuses failed to reach their growth potential18. We aimed to discover how many of the stillborn fetuses classified as unexplained death prior to onset of labour or unclassifiable by CP classification code could be classified as having intrauterine growth retardation (IUGR), below the 10th centile weight for gestational age centile, when IUGR was defined in two different ways:-
- Using LMSgrowth19 which takes into consideration gender and gestation.
- Using the gestation related optimal weight standard (GROW)20 which defines the fetal growth potential by excluding factors such as maternal smoking and diabetes and takes into consideration gender, gestation, maternal height, weight, ethnic origin, and parity.
Table 1 Rate of IUGR in stillborn fetuses classified as unexplained death prior to onset of labour or unclassifiable by CP classification born to Welsh residents, 1993-2012 for LMSgrowth and 1996-2012 for GROW.
|LMSgrowth||GROW by gestation at delivery with multiples included||GROW by gestation at delivery minus 2 days with multiples included|
|Number (%) IUGR||750 (37.8%)||602 (52.3%)||563 (48.9%)|
|Number (%) not IUGR||1234 (62.2%)||550 (47.7%)||589 (51.1%)|
We excluded from the analysis birthweights <300g and stillborn fetuses that died more than 5 days prior to delivery. The rate of IUGR was very similar when IUGR was classified using LMSgrowth using only the stillborn fetuses included when the rates of IUGR were calculated using GROW. In addition excluding or including stillborn fetuses from multiple births made little difference to the rates of IUGR by either method. Professor Gardosi advises that 2 days is taken off the gestational age at delivery when using GROW, the data for this method is shown. Using the date of death variable included in the AWPS database for some stillbirths made little difference to the rates of IUGR.
The results show that a larger percentage of stillbirths were classified as having IUGR by either definition than the 10% you would on average expect to find. In addition the results show that a greater percentage of stillbirths were classified as IUGR when GROW was used to define IUGR compared to when LMSgrowth was used to define IUGR. This is in line with other research which found that many stillborn fetuses had failed to reach their growth potential18. Gardosi et al. found that in a cohort of stillbirths in the West Midlands when multiple pregnancies and congenital anomalies were excluded the overall stillbirth rate was 4.2 per 1000 births but 2.4 in pregnancies without fetal growth restriction, rising to 9.7 per 1000 births with antenatally detected fetal growth restriction but more worryingly was 19.8 per 1000 births when fetal growth restriction was not detected18
Risk – high, low, but never ‘no’ risk
The stillbirth rates for women identified with risk factors have fallen – and the management of ‘high risk’ pregnancy has slowly and consistently led to better mortality figures in this group. This is partly because of improved therapies and surveillance, but also because the obstetrician can deliver the mother of her baby before it dies. This can, of course, mean that mortality is shifted from the antenatal or intrapartum to the neonatal period – early delivery may expose the newborn to the risks of prematurity, for example, but perinatal mortality rates are slowly and consistently falling.
Deterioration in placental function alters the fetal metabolism. Anaerobic glycolysis necessary during fetal hypoxaemia, leads to depletion of glycogen stores in the fetal liver. Because the liver makes up so much of the abdominal content in-utero, restrictions in the abdominal circumference on ultrasound correlate reasonably well with the resulting acidaemia. The protective re-distribution of blood to the fetal head and the consequence to fetal right ventricular function can also be tracked using Doppler scanning.
Clinical assessment of fetal growth is relatively poor in predicting growth restriction reliably, but this usually heralds referral for more formalised fetal growth assessment using ultrasound. 50% of babies who do not reach their growth potential are missed by this approach and wrongly classified as being ‘low risk’. As discussed in this current AWPS report, stillbirths contain a disproportionate number of growth restricted babies and even more so when one considers more accurate means of ascertaining growth potential. Growth charts customised for factors known to affect growth potential (such as fetal sex and parity) appear not only to highlight growth problems in-utero, but seem to significantly reduce the number of ‘unexplained’ stillbirths when applied to such babies. Although the use of customised growth charts is unproven, they show huge potential and there is a need to fund implementation and further research in this area in Wales.
These changes in placental (dys)function can occur before maternal perception of reduced fetal movements, which in turn occurs often earlier in the process of fetal deterioration than abnormalities of the fetal heart rate with the consequence that reduced fetal movements are not an accurate predictor of fetal well-being; only 50% of women complain of reduced fetal movements prior to presenting with an antepartum stillbirth. However, there may be an association between placental abnormalities and reduced movements when the placenta is looked at in meticulous detail 21.
Randomised trials of formal fetal movement ‘kick charts’ did not have the anticipated effect – there was no improvement in neonatal outcomes and maternal anxiety was increased. This is reinforced by the Cochrane Review and NICE guidance on antepartum care. However, a change in character or relative number of movements may indeed be clinically significant, and there are studies underway to investigate this further.
There are also limitations in protocols that rely on the fetal cardiotocograph as an assessment of longer term fetal well-being. Alteration in fetal heart rate occurs relatively late in the process of placental dysfunction, even when it forms part of a biophysical score that includes liquor volume. The fetal heart can show reductions in variability that cannot be picked up easily on auscultation and computerised analysis appears to be more sensitive in this respect. By the time there is a pathological CTG, there may be a maximum of 72 hours before a baby dies.
There are thus inherent flaws in the way we assess ‘normality’ in terms of being ‘low risk’. We ask non-specifically whether a woman feels her baby is moving, without an ability to provide evidence of the significance from randomised trials about what our response should be. Furthermore, we shy away from using the term ‘stillbirth’ in the vernacular so women are unsure what they are trying to prevent. Fetal heart rate is recorded merely as being present and vaguely an acceptable rate with no information on the parameter most sensitive to hypoxia – the baseline variability – and no randomised controlled trial of this as an effective manoeuvre. When we do a CTG it cannot absolutely reassure us, because it deteriorates late in the process of placental dysfunction and fetal compromise – yet this is precisely our current state of play.
Thus, for both reduced fetal movements and the identification of fetal growth restriction, the evidence for a standardised routine application for all women is incomplete. Wales needs to find a pragmatic approach to dealing with this and it is being tackled as part of the Welsh Initiative on Stillbirth Reduction, part of the 1000 Lives Plus Transforming Maternity Services mini-collaborative22.
A note on intrapartum electronic fetal monitoring
Intrapartum fetal hypoxia remains an important cause of death and permanent handicap and there are many studies reporting a significant proportion of cases with evidence of suboptimal care related to fetal surveillance. Cardiotocographic (CTG) monitoring remains the basis of fetal surveillance during labour, but its interpretation by healthcare professionals is subject to great variation between observers and between the same clinician at different times – especially where a good or poor neonatal outcome is known (hindsight bias). Thus, there is often poor agreement on the features of a CTG – the presence and significance of slowing of the heartbeat, for example and the overall classification of whether the trace is normal or needs intervention – and then what that intervention might be.
Several countries with lower stillbirth and neonatal death rates have introduced developments of the conventional CTG. ST Analysis (STAN) looks at the part of the fetal ECG that changes in the presence of hypoxia – the ST segment. These changes are much more frequent than one might expect and thus interpretation depends on the likelihood that any event is significant – which means that it is used in conjunction with the need to interpret the conventional CTG reliably. The consequence to this is a huge training commitment and some people find the technology cumbersome and invasive. The addition of fetal electrocardiogram analysis has increased the potential to avoid adverse outcomes, but CTG interpretation remains its main weakness.
The RCOG has developed and launched an e-learning tool that is freely available to all NHS staff. It is both educational and assessed and is a key potential element in improving clinical staff skills in intrapartum fetal monitoring. The Welsh Risk Pool – and thus all Health Boards – has adopted this training formally into practice in NHS Wales.
In summary the stillbirth rate in Wales in 2012 is not significantly different from 10 years ago and is twice as high as some European countries. There are a number of identified risk factors that can lead to increased surveillance and intervention but we need a specific strategy to lower and / or mitigate risk where possible. The current way that deaths are classified is unsatisfactory. A post mortem can help identify what happened and sometimes adds information to help with any future pregnancy.
Commentary prepared from BMA Cymru Wales submission by Mr Philip J Banfield and Dr Mark Temple to the WG One Day Enquiry into Stillbirths 2012 along with data provided and analysed by the All Wales Perinatal Survey.
1. Northern Ireland Statistics and Research Agency. Registrar General Annual Reports ‘Available at:’ http://www.nisra.gov.uk/archive/demography/publications/annual_reports/2011/RG2011.pdf. Accessed May 2013.
2. Euro-Peristat PROJECT. The European Perinatal Health Report ‘Available at:’ http://www.europeristat.com/reports/european-perinatal-health-report-2010.html. Accessed May 2013.
3. Draper ES. Evaluating and comparing neonatal outcomes. Archives of disease in childhood. Fetal and neonatal edition 2010;95(3):F158-9
4. Government W. One-day inquiry into still births in Wales ‘Available at:’ http://www.senedd.assemblywales.org/mgIssueHistoryHome.aspx?IId=3352. Accessed June 2013.
5. Confidential Enquiry into Stillbirths and Deaths in Infancy: 6th Annual Report. London: London: Maternal and Child Health Reasearch Consortium.; 1999 1 January-31 December 1997.
6. NHS Information Centre 2011. Infant Feeding Survey: Early Results ‘Available at:’ http://www.ic.nhs.uk/pubs/infantfeeding10 Accessed 2013.
7. Seaton SE, Smith LK, Draper ES, Mankelow BN, Springett A, Field DJ. Socio-economic inequalities in the rate of stillbirths: A population based study. Arch Dis Child Fetal Neonatal Ed 2011;96(Suppl 1):Fa17-Fa53
8. IDS Scotland. The Scottish Perinatal and Infant Mortality and Morbidity Report (SPIMMR). Trends in Perinatal Mortality in Scotland 30 year review (May 2009) ‘Available at:’ http://www.isdscotland.org/Health-Topics/Maternity-and-Births/Stillbirth-and-Infant-Deaths/. Accessed May 2013.
9. Kenny LC, Lavender T, McNamee R, O’Neill SM, Mills T, Khashan AS. Advanced maternal age and adverse pregnancy outcome: evidence from a large contemporary cohort. PloS one 2013;8(2):e56583
10. Flenady V, Koopmans L, Middleton P, Froen JF, Smith GC, Gibbons K, et al. Major risk factors for stillbirth in high-income countries: a systematic review and meta-analysis. Lancet 2011;377(9774):1331-40
11. NHS Information Centre for Health and Social Care; in partnership with Professor Mary Renfrew; Mother and Infant Research Unit; Department of Health Sciences; University of York. Infant Feeding Survey 2010: Early Results ‘Available at:’ http://data.gov.uk/data/resource_cache/7f/7f9aba90-e1aa-413f-9acd-7d3299c5e7c3/infa-seed-serv-2010-earl-resu-rep.pdf. Accessed May 2012.
12. Royal College of Obstetricians and Gynaecologists. CMACE release: National enquiry into maternal obesity – Implications for women, babies and the NHS ‘Available at:’ http://www.rcog.org.uk/news/cmace-release-national-enquiry-maternal-obesity-%E2%80%93-implications-women-babies-and-nhs. Accessed May.
13. Scott-Pillai R, Spence D, Cardwell C, Hunter A, Holmes V. The impact of body mass index on maternal and neonatal outcomes: a retrospective study in a UK obstetric population, 2004-2011. BJOG : an international journal of obstetrics and gynaecology 2013
14. Haws RA, Yakoob MY, Soomro T, Menezes EV, Darmstadt GL, Bhutta ZA. Reducing stillbirths: screening and monitoring during pregnancy and labour. BMC pregnancy and childbirth 2009;9 Suppl 1:S5
15. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet 2008;371(9606):75-84
16. Iams JD, Romero R, Culhane JF, Goldenberg RL. Primary, secondary, and tertiary interventions to reduce the morbidity and mortality of preterm birth. Lancet 2008;371(9607):164-75
17. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet 2008;371(9608):261-9
18. Gardosi J, Madurasinghe V, Williams M, Malik A, Francis A. Maternal and fetal risk factors for stillbirth: population based study. BMJ 2013;346:f108
19. Pan H, Cole TJ. LMSgrowth, a Microsoft Excel add-in to access growth references based on the LMA method. Version 2.77 ‘Available at:’ http://www.healthforallchildren.co.uk. Accessed 2012.
20. Gardosi J, Francis A. Customised Weight Centile Calculator. GROW version 6.5 (UK), Gestation Network ‘Available at:’ www.gestation.net. Accessed May 2013.
21. Warrander LK, Batra G, Bernatavicius G, Greenwood SL, Dutton P, Jones RL, et al. Maternal perception of reduced fetal movements is associated with altered placental structure and function. PloS one 2012;7(4):e34851
22. 1000 Lives Plus Transforming Maternity Services mini-collaborative ‘Available at:’ www.1000livesplus.wales.nhs.uk/maternity. Accessed September 2013.