Definition
Perinatal mortality defined by the Ninth Revision of the International Classification of Diseases (leD) com-prises late fetal deaths or stillbirths and deaths in the first week of life (early neonatal deaths). The Ninth Revision recommends that babies chosen for inclusion in perinatal mortality statistics should be those above a
J~~:;i'lfi'
Fiji.kiPi·"· ~
Mexico ~
Dominican Rep ... , '.P "" ·.X·' ,~ x:1.* . : :.;, Costa Rica )tw:, :~',il:':l:W ,.x'l!MWl:' : :![:':!: ~ :~,,: : ~*: :"::'w :' :~'
Thailand
Bangladesh . :§~',. . ". ,~:
Sri Lanka 1l$,:
30 40 50 60 70 80 Percentage
Fig. 2.1.1 Percentages of women with three living children who want no further pregnancies. (Adapted from Maine D.
Family Planning: Its Impact on the Health of Women and Children. New York, The Centre for Population and Family Health, Columbia University, 1981.)
mmlmum birth weight, and if this is not available, gestation or body length should be used. The minimum values for inclusion in national and international perinatal statistics are shown in Table 2.1.1.
The duration of gestation is measured from the first day of the last normal menstrual period. Gestational age is expressed in completed days or completed weeks - events occurring from 280 to 293 days after the onset of the last normal menstrual period are considered to have occurred at 40 weeks of gestation.
(See Fig. 2.1.2).
The International Classification of Diseases is revised every 10 years and the Tenth Revision is due soon. Whatever new recommendations are made for perinatal statistics, it would be difficult to improve on the proposals for international perinatal statistics in the Ninth Revision, particularly in developing countries where perinatal mortality statistics are difficult to obtain, and, those that are available are collected in hospitals. Data on perinatal deaths in the community where most babies are born, are generally not available in developing countries.
Why consider perinatal mortality?
While infant mortality rate (IMR) is a measure of socio-economic development of a country, indicators such as maternal deaths and perinatal mortality reflect the health status of women and the quality of health care received during pregnancy, labour, delivery and in the early days oflife. This has been seen in the development of health care for mothers and newborn infants in the United Kingdom from the second decade of the twentieth century until now. Four phases of develop-ment have been identified:
1. 1910-30 both maternal mortality and perinatal mortality rates were high, and this led to the licensing of midwives.
2. 1931-40, maternal mortality gradually decreased but perinatal mortality remained high -
sulphon-Table 2.1.1 Definition of babies for inclusion in national and international perinatal statistics
Minimum value Birth weight Gestational age Body length
(crown-heel)
National perinatal statistics 500 g 22 weeks 25 cm
International perinatal statistics 1000 g 28 weeks 35 cm
First day of last menstrual
period
Conception
2
Preterm less than 37
completed weeks (259 days)
Term (259-293
days)
37
amides were used to treat infections during this decade.
3. 1941-70 saw a rapid decline in maternal mortality to its present low level associated with the liberal use of Caesarean section, blood transfusion and anti-biotics, but the perinatal mortality rate para-doxically increased.
4. 1970-89, maternal mortality remained low and perinatal mortality declined substantially to its present low level in association with introduction of neonatal care units and technological Improve-ments.
Experience m developed countries including the United Kingdom has confirmed the value of birth weight specific mortality, after excluding lethal mal-formations, as a sensitive indicator of health care quality. 1 Where health care is rudimentary, the death rate among infants weighing more than 2500 g at birth is high; this perinatal mortality rate improves when an effective system of health care for mothers and babies is provided in the community. Perinatal deaths and prevalence oflow birth weight are two simple measures to monitor health care provided for women and their offspring. As the health status of a community improves, the perinatal mortality rate decreases.
A population with a higher than average proportion of low-birth-weight infants will have a higher crude perinatal mortality rate. Fifty-seven per cent of perinatal deaths occurred among babies with a birth weight of less than 2500 g although they constituted only 10.8 per cent of births in Cuba in 1973.2 To determine how well newborn infants are being cared for, it is necessary to calculate the risks of death within different birth weight bands. Comparison of birth weight specific mortality rates indicate trends in mortality within the same population or in different populations with similar mean birth weights and birth weight distributions.
Other factors influencing perinatal mortality in developing countries are maternal age and parity. A
Perinatal mortality in developing countries 133
Post-term completed 42 weeks (294 days or more)
42 weeks Fig. 2.1.2 Gestational age.
population-based study in rural Kenya3 showed that maternal age of 35 years or more and parity of seven or more were associated with increased perinatal mortality. A history of previous perinatal death and breech delivery were also associated with higher perinatal mortality. Studies have shown that for a given maternal weight, mothers who are taller give birth to heavier babies, whereas mothers who are fatter give birth to lighter babies. However, a study in an urban hospital in Uganda indicated that maternal weight of less than 55 kg was associated with increased perinatal deaths and low-birth-weight babies.4 It is clear that more community-based controlled studies are needed in developing countries to identify risk factors asso-ciated with perinatal deaths, in order that interventions may be introduced for better survival.
Causes of perinatal deaths
Ideally, causes of perinatal deaths should be based on data collected in the community. But most studies on perinatal mortality have been done on deliveries in hospitals. Classification of perinatal deaths into ante-partum, intraante-partum, and early neonatal deaths provides a helpful framework for analysis. Intrapartum deaths occur more often among normal babies than low-birth-weight babies and are associated with inadequate supervision of labour. A survey carried out in Indian hospitals in 1977-95 reported a perinatal mortality rate of66.3 per 1000 total births, an incidence of fresh stillbirths of 26.2 per 1000 total births and a ratio of fresh stillbirths to macerated stillbirths of 1. 5. In aN airobi birth survey the stillbirth rate was 23 per 1000 total births, the incidence of fresh stillbirth was 14.4 per 1000 total births and the ratio of fresh stillbirths to macerated stillbirths was 1.7.6 In European countries where the perinatal mortality rate is around 10 per 1000, the incidence of fresh stillbirth is about 1 per 1000, and the ratio of fresh stillbi?ths to macerated still-births is around 0.2. The above figures show that
134 Introduction
improvement in care of women in pregnancy, labour and delivery produces a sharp decrease in the incidence of fresh compared to macerated stillbirth.
Clinical classification of causes of perinatal deaths This classification assists the identification of prevent-able deaths where intervention is required. The Aberdeen classification7 proposed in 1954 promoted the use of eight categories: cause unknown (premature), cause unknown (mature), trauma, pre-eclampsia, ante-partum haemorrhage, congenital malformations, maternal disease, and miscellaneous. Wigglesworth, in 1980, suggested a classification based on five patho-logical groups analysed by birth weight. 8 This classi-fication is simple and can be used where perinatal pathology facilities are not available (Table 2.1. 2).
If deaths due to asphyxia occurring in mature infants (over 2500 g) were high, this indicates a need to improve care during labour and delivery by proper management of obstructed labour and hypertensive disorders, and preventing birth injury. In the rural Kenya ctudy, half of all perinatal deaths were caused by eiUler preterm delivery or birth trauma3.
Perinatal audit
Perinatal audit should arise from the clinical classi-fication of perinatal deaths, when health workers learn from discussion of the circumstances leading to deaths, so that future clinical management may be improved.
Perinatal audit, initiated by hospital departments of obstetrics and paediatrics would involve the active par-ticipation of community-based personnel in maternal and child health services who can identify and refer high-risk mothers to appropriate levels of care.
Morbidity
In the perinatal period, morbidity may have long-term effects of mental and physical handicap. Morbidity affecting the fetus or neonate can originate at five periods of development: preconception (genetically-determined diseases), embryonic period (neural tube defects), prenatal period (intrauterine growth retarda-tion), intrapartum period (trauma), and neonatal period (ineffective resuscitation at birth). Data collected during the perinatal period would also identify causes of morbidity.
Perinatal data collection
Perinatal mortality is the most reliable index of the quality of antenatal and obstetric care. Most deaths in the perinatal period (excluding congenital defects) are the result of complications in pregnancy and childbirth.
Therefore, the following recommendations have been made on perinatal data collection in developing countries by the World Health Organization2:
• Maximal use should be made of available data while taking into consideration their quality.
• Linkage of data held in different locations such as hospitals, community health services and civil registers should be explored to provide a more reliable information system for routine surveillance.
• The epidemiology of perinatal mortality is required if effective policies and interventions are to be for-mulated and put into practice, especially outside the hospital where the majority of births take place.
• Methodologies should be developed that are simple and scientifically sound and which can be used at community level in developing countries.
Table 2.1.2 Model form for classification of perinatal deaths (after Wigglesworth, 1980)
Birth weight (g)
<1000 1001-1500 1501-2000 2001-2500 over 2500 unknown Total
Normally formed macerated SB
Congenital malformation (SB/NND)
Conditions associated with immaturity (NND)
Asphyxia in labour (fresh SB/NND)
(SB = Stillbirth; NND Neonatal death; PNMR Perinatal mortality rate)
Other specific conditions
Total
births PNMR
• Improved understanding of risk factors related to perinatal mortality makes possible timely antenatal interventions and referral of high-risk mothers to appropriate levels of care.
The collection of data on perinatal mortality and morbidity requires registration of births and deaths, and a system of records for pregnant women and the outcome of their pregnancies, including birth weights.
Few developing countries have national statistics on perinatal and maternal deaths today. However, relevant data may be gathered from some hospitals but this will not be an accurate reflection of the situation in the community. Ifhospital and teaching institution staff were to extend their interest to the community, cooperation with primary health centres and village leaders would make it possible to collect meaningful information. Interest in the community must be stimulated and harnessed to ensure collection of statistics, their analysis and interpretation.
A good example of data collection of deaths in the perinatal period has come from Curacao, the largest island in the Netherland Antilles in the Caribbean.9 This survey analysed the 223 consecutive fetal and neonatal deaths occuring in 6514 births over two years, 1984 and 1985. About 98 per cent of the deliveries took place in hospital. Full autopsies were performed on 210 deaths (94 per cent). Perinatal deaths comprised both fetal and neonatal deaths with a minimum birth weight of 500 g. Infants who died within 28 days of life were included. The crude death rate was 34.2 per 1000 total births. Death was caused by problems of preterm birth in 68 cases, 53 of whom were born before 28 weeks of gestation. Other causes of death were associated with asphyxia in 35, malformation in 28, maternal hyper-tension in 25, antepartum haemorrhage in 19 and miscellaneous (generalized infection, hydrops fetalis, kernicterus) in 14. No specific cause could be found in 34 deaths of whom 32 were macerated stillbirths. The authors of this study concluded that a substantial reduction of perinatal mortality could be achieved by
References 135
better fetal surveillance for fetal growth retardation, improved clinical judgement and better management of fetal distress, particularly in the 58 deaths (27 per cent) at or after 28 weeks of gestation due to hypertension and asphyxia, as well as many of the macerated stillbirths.
This study shows that it is possible to carry out a population-based inquiry into fetal and neonatal mortality in a region that is socially and economically disadvantaged.
References
1. Chalmers I. The search for indices. Lancet. 1979; ii:
1063-5.
2. Edouard L. The epidemiology of perinatal mortality.
World Health Statistics Quarterly. 1985; 38: 289-301.
3. Voorhoeve AM, Muller AS, W'oigo H. Agents affecting health of mother and child in a rural area of Kenya. XVI.
The outcome of pregnancy. Tropical and Geographical Medicine. 1979; 31: 607-7.
4. Holden J. Mengo hospital maternity survey 1980. Personal communication.
5. Mehta AC. Perinatal mortality. In: Menon MKK, Devi PK, Rao KB eds. Postgraduate Obstetrics and Gynaecology, 2nd Edn. Bombay, Orient Longman, 1982. pp. 195-200.
6. Mati JKG. The Nairobi birth survey - 1. The study design, the population and outline results. Journal of Obstetrics and Gynaecology
if
Eastern and Central Africa. 1982;1: 132-9.
7. Baird D, Walker J, Thomson AM. The causes and prevention of stillbirths and first week deaths. III. A classi-fication of deaths by clinical cause: the effect of age, parity and length of gestation on death rates by cause. Journal of Obstetrics and Gynaecology of the British Empire. 1954; 61:
433-48.
8. Wigglesworth J. Monitoring perinatal mortality: a patho-physiological approach. Lancet 1980; i: 684-6.
9. WildschutJIJ, Nolthenius-Puylaert MCBJET, Wiedjik Y et al. Fetal and neonatal mortality: a matter of care?
Report of a survey in Cura~ao, Netherlands Antilles.
British MedicalJournal. 1987; 295: 894-8.