Key Health Data for the West Midlands 2001
Chapter 8: Perinatal Mortality
List
of Tables
List
of Figures
Main Body
8: Perinatal Mortality
9: Crime
10: Sports Facilities
11: Housing Quality
12: Inequalities, Focusing on the early years
8.1 Introduction
A primary aim of the Perinatal Institute is to seek to understand the causes and to develop strategies for the prevention of adverse perinatal outcome. In this chapter, we will present recent perinatal statistics and outline some of our current initiatives aimed to address these problems.
The number of births in the West Midlands has fallen by 2,012 from 63,857 to 61,845 in 2000, a 3.2% reduction on the previous year compared with a 2.8% drop for England & Wales.
The perinatal mortality rate is 9.7/1,000
births (597 cases) in the West Midlands. This is higher than the rate
for England & Wales (8.1/1,000 births), and of all other Health Regions.
|
Figure 8.1. National mortality rates by Region, 2000 |
The stillbirth rate in 2000 is 5.6/1,000
births (348 cases) in the West Midlands. This compares with a rate of
6.1/1,000 births in 1999. The stillbirth rate in the West Midlands exceeds
that for England & Wales, which is 5.2/1,000 births. The regional
rate is the second highest of all Health Regions.
The early neonatal death rate in 2000 for the West Midlands is 4.0/1,000
live births (3.8/1,000 live births in 1999). This remains higher than
the early neonatal death rate for England and Wales (2.9/1,000 live births),
and is the highest of any region.
Within the West Midlands, there is wide variation in perinatal mortality
rates.
Figure 8.2. shows the rates for the last
5 years. Further details in tabular form are available on www.wmpi.net/pnm/trends.htm.
|
Figure 8.2. West Midlands mortality rates by Health Authority, 1996-2000 |
Interpretation of this variation is difficult because of a lack of denominator data. The Perinatal Institute is currently seeking funding to establish region-wide collection of a minimum maternity dataset, and has developed data collection software that is ready for installation in all units.
8.2. Perinatal Mortality and Poverty
We carried out an analysis of perinatal mortality rates against socio-economic factors, using mapping from the Child Poverty Index. For these analyses, the measure of deprivation used was the DETR Child Poverty Index. Other DETR domains, such as that for Health Deprivation and Disability, were not used as they contain measures such as proportion of low birthweight deliveries, making them dependent variables. The Child Poverty Index is calculated at ward level and is the proportion of children under 16 living in means-tested, benefit reliant families. For the West Midlands, the median category of deprivation for all births is 30-40% (range 1.9 to 79.8%).
Although data are grouped over five years (1995-1999), adverse perinatal events are rare for populations the size of electoral wards. Therefore, each event (birth or death) was categorised into one of eight groups of deprivation scores, determined by their electoral ward. Rates were then calculated for each category. c2 values were calculated for linear trend.
 |
Figure 8.3. West Midlands mortality rates - electoral wards grouped by CPI group, 1995-1999 |
The graph shows the association between Child Poverty Index and perinatal mortality in the West Midlands. It is highly significant both for stillbirths, (χ2= 52.7, p < 0.001) and for early neonatal deaths (χ2= 37.0, p < 0.001).
The association between the Child Poverty Index and perinatal mortality extends to all causes of death according to the Wigglesworth classification, except deaths from intrapartum causes and unclassified cases. There are significant associations between the Child Poverty Index and perinatal deaths from congenital anomaly (χ2= 33.1, p < 0.05), unexplained antepartum death (χ2= 40.3, p < 0.001), prematurity (χ2= 13.7, p < 0.01) and other specific causes (χ2= 6.1, p = 0.05).
8.3. Stillbirths
Analysis of cause of death by the Wigglesworth classification (Figure 8.4 shows that the largest component of stillbirths is in the 'unexplained' category: 69% (up from 66% in 1999).
 |
Figure 8.4. West Midlands stillbirths 2000 by Wigglesworth |
Further analysis of the 'unexplained' group suggests that in many of these deaths the baby was small-for-gestational age before demise.
Figure 8.5. baby weights are plotted for the 232 (69.5%) of 334 West Midlands stillbirths in 2000 that had the classification 'unexplained'. The gestational ages are adjusted for an average 2 day delay between fetal demise and delivery. The 10th, 50th and 90th weight-for-gestational age centiles are also shown.
 |
Figure 8.5. Weight for gestational age of West Midlands stillbirths, 2000 |
| Table 8.1. Number of stillbirths for 1996-2000 for each of the 13 West Midlands HA's | |
The proportion of 'unexplained' stillbirths that are small-for-gestational age is high in all health authorities of the West Midlands, regardless of the actual level of perinatal mortality in that area. Overall, 63.6% of unexplained stillbirths were SGA (range 43 to 72%).
New Classification for Stillbirths
The traditional classification systems for perinatal mortality result in a high proportion of stillbirths in the 'unexplained' category, a fact that is not conducive to understanding and working to avoid perinatal compromise (CESDI, 2001). Neither of the three classification systems in use (Wigglesworth, Fetal Neonatal, Aberdeen/Obstetric) have intrauterine growth restriction (IUGR) as a category, even if a postmortem has shown that the fetus was small. Strictly speaking, IUGR may not be the actual cause of death, although it may be a condition that existed before demise, and hence a relevant condition when trying to seek clinical relevance in the statistics.
A new classification system has been developed in the West Midlands that includes IUGR as a category. IUGR is considered present if the weight-for-gestation is below the 10th customised centile, after adjustment for the known physiological variables affecting weight (maternal height, weight in early pregnancy, parity and ethnic group) (Clausson et al, 2001). This method has been shown to allow the distinction between physiological and pathological causes of low weight for gestation in the study of perinatal mortality (Gardosi et al, 2000). The centile programme is freely available on www.gestation.net/centiles.
The new classification system (Table 8.2) seeks to identify the relevant condition rather than the actual last cause of death, and is hierarchical in that the first category into which a death 'fits' is identified as the most relevant condition. Thus if two conditions exist - for example Congenital Anomaly and IUGR - the category given is that of 'Congenital Anomaly', as it is the one higher up the list.
| Table 8.2. The ReCoDe System: Relevant Condition(s) at Death | |
By the new ReCoDe system, 92 (46%) of all stillbirths fell into the IUGR category. Only 33 (16.5%) remained unexplained. In contrast, 147 (73%) fell into the 'unexplained' category by the Wigglesworth classification, and 107 (54%) were 'unexplained' according to the Obstetric/Aberdeen classification. The ReCoDe system gives us a better understanding of the antecedents of stillbirth and the clinical priorities, which need to be addressed to reduce their incidence.
Avoiding stillbirths due to intrauterine growth restriction
IUGR can be suspected antenatally by screening with fundal height measurements and diagnosed by ultrasound biometry and Doppler. Yet the screening strategy is failing, as approximately three quarters of all babies born SGA are not detected as such during standard antenatal care (Hepburn and Rosenberg, 1986) Better screening and improved detection of those pathologically affected would allow earlier delivery from an unfavourable intrauterine environment. Although many of these babies would be delivered at gestations where they would be considered sufficiently mature, there would be a slight increase in the prematurity rate. However, the majority of unexplained stillbirths (158 of the 232; 68.1%) occur from 30+ weeks - when survival of a neonate delivered in good condition exceeds 95% (Draper et al, 1999).
The Perinatal Institute is due to commence a regional initiative aimed at improving the detection of SGA. The project has the support of many perinatal professionals as well as the regional heads of midwifery and directors of public health. Funding is being sought. The protocol entails the training of a standardised method of regular measurement of fundal height and plotting on customised growth charts, with agreed care pathways to refer for ultrasound biometry and further investigations, as indicated. A controlled pilot study has shown that this method results in increased detection of growth abnormalities while decreasing the number of unnecessary investigations and admissions to hospital (Gardosi and Francis, 1999) study.
8.4. Neonatal Deaths
Analysis of cause of death by the Wigglesworth classification (Figure 8.6) shows that the largest component of early neonatal deaths is prematurity (57%).
 |
Figure 8.6. West Midlands early neonatal deaths by Wigglesworth, 2000 |
The main contributing factors to prematurity are maternal smoking, infection and intrauterine growth restriction. The CESDI 'Project 27/28' is a confidential enquiry of the management of babies born at these early gestations, comparing deaths and survivors. The final report is due during 2002, but analysis of first year West Midlands data by the Perinatal Institute has highlighted the importance of appropriate fluid volume management at these early gestations (Ewer, 2001).
The regional neonatologists have responded quickly to these findings and, through their regional standards group, developed up-to-date, clear guidelines which have since already been instituted in most West Midlands neonatal units.
The national CESDI project has reported on survival figures of babies born at these gestations (CESDI, 2001). See figure 8.7.
 |
Figure 8.7. CESDI Regions - survival rates (excluding lethal anomalies), 27-28 week gestation live births Sep 1998-Aug 2000 |
The survival rate of babies born at 27-28 weeks is better than previously expected. The West Midlands rate is 89.2%, which corresponds to the average for England, Wales and Northern Ireland.
Regional and national analysis of the results of Project 27/28 is expected to be completed during 2002. Nationally, CESDI has decided to distribute case notes for panel assessments to be carried out in different regions during the second year of the project. This may however introduce confounding variables. As the largest CESDI region and the one with the highest neonatal mortality rate, the West Midlands has decided to put in the additional work to fulfil the obligation to the national project, but also to continue our own methodology unchanged for the full duration of the project. We will hence be able to report on two years of data assessed by our own region's panels.
The Perinatal Institute is currently in discussion
with the Specialist Services Team and the Neonatal Forum to establish
and maintain a regional neonatal register.
8.5. Congenital Anomalies
The West Midlands Congenital Anomalies Register monitors the incidence of all congenital anomalies in the region (see www.wmpi.net/car). A major recent focus has been Down's syndrome, which is also the first priority of the National Screening Committee. This is a genetic condition associated with the presence of an extra number 21 chromosome, caused by a genetic accident occurring around conception. It is the most common cause of learning disability with an incidence at birth of about 1 per 1,000. People with Down's syndrome have varying degrees of learning disability and medical problems, with about half the babies born having cardiac defects. The incidence of Down's syndrome is slowly increasing, which is consistent with the increasing age of the maternity population.
 |
Figure 8.8. Down's syndrome - West Midlands 1995-2000 95% confidence intervals and 3 year time trend |
The Perinatal Institute's report on Down's syndrome was published in 1999 (McKeown, 1999) executive summary - see http://www.wmpi.net/car/downs/. It found that the uptake of screening varied in the Region and that there was a clear link between this and detection rates for Down's. There was a need for standardisation of services, and the making available of all relevant information and counselling. Dating by ultrasound improved the accuracy of the serum test and, in liaison with the Regional Ultrasound Group, the Perinatal Institute issued region-wide recommendations on routine scan dating in each pregnancy.
With DoH funding, Regional Co-ordinators
for Antenatal Screening have been put in place during 2001 in all English
Regions to co-ordinate local service provision and training. In the West
Midlands, this initiative is being managed by the Perinatal Institute.
A current priority is to identify resources to put in place antenatal
screening co-ordinators for each unit, who will be instrumental in implementing
and running the agreed standardised screening programme locally. We have
recently undertaken an assessment of current screening practices in the
Region as part of a national Down's survey, which will be published in
spring 2002.
References
CESDI - Confidential Enquiries into Stillbirths and Deaths in Infancy - 8th Annual Report, 2001
Clausson B, Gardosi J, Francis A, Cnattingius S. Perinatal outcome in SGA births defined by customised versus population based birthweight standards. Br J Obstet Gynaecol 2001;108:830-4.
Draper E S, Manktelow B, Field D J, James D. Prediction of survival for preterm births by weight and gestational age: retrospective population based study BMJ 1999;319:1093-1097.
Ewer A. The significance of Volume Expansion in Potentially Avoidable Deaths. In: Carrera JM, Cabero L, Baraibar R. The Perinatal Medicine of the Millennium (Proceedings of the 5th World Congress of Perinatal Medicine) Monduzzi Editore, Bologna, 2001
Gardosi J, Francis A. Controlled trial of fundal height measurement plotted on customised antenatal growth charts. Br J Obstet Gynaecol 1999;106:309-17.
Gardosi J, Mongelli M, Wilcox M, Chang A, Sahota D, Francis A. Gestation Related Optimal Weight program (GROW). Software v 3, 2000. Gestation Network - www.gestation.net.
Hepburn M, Rosenberg K. An audit of the detection and management of small-for-gestational age babies. British Journal Obstetrics & Gynaecology 1986;93:212-216.
McKeown C, Tonks A, Wyldes M. Down's Syndrome
1995-1997. West Midlands Perinatal Institute, 1999.
For more information please
contact Carol Richards on
0121 414 3368
© Department of Public Health
and Epidemiology, University of Birmingham