Abstract

A study of the aetiopathogenetic mechanisms of still births is a subject of prime importance and the understanding of which will greatly help future generations and this can be achieved only through the medium of well conducted autopsies. With this aim in mind, we analyzed our stillbirth autopsy data, to recognize our leading cause of mortality, for this we utilized three major classifications of perinatal deaths namely, Wigglesworth’s classification, Joshi’s classification and the obstetrics Aberdeen’s classification. We also correlated the foetal lesions with placental lesions wherever available and with maternal risk factors. In the two year period from 1994 -95 our stillbirth rate was 33.75% and the stillbirth autopsy rate was 73.28%, stillbirths of low gestational age and low birth weight predominated, 92.7% and 78.12% respectively. 70.8%cases had associated maternal risk factors, out of which anaemia and antepartum haemorrhage predominated. Placentae were available for study in only 57% cases and pathological lesions were encountered in the majority 96.36% the commonest being infarction and infection. The commonest cause of stillbirth, in all the three classifications was asphyxial lesions, followed by infection. Congenital malformations were few, indicating that unlike the west we are still battling with preventable causes of death.

Introduction

“To lose an offspring is hard and harder still to lose him by a disease not yet fully understood by doctors. But for the sake of other children; I think that to have seen his organs is of greatest utility.”
– Lester King1

The loss of a viable foetus has multidimensional repercussions on society. To the mother it is a great psychological trauma and a tremendous strain on health. To the family, it is a psychological trauma, coupled with financial burden. What holds true for the family as far as the financial burden is concerned, holds true for the community. In the kind of orthodox society that we live in, giving birth to a stillborn foetus is a social stigma. In the earlier days, a stillbirth was considered to be the result of the sin of the woman or the wrath of God. Advances in the field of medicine have cleared many misconceptions about the aetiology of stillbirths. It is now realized that stillbirths are influenced by environmental factors. Factors pertaining to the mother, placenta and foetus have been recently brought to light. But we are still in the dark regarding many factors responsible for stillbirths. Hence, well conducted autopsies on stillborn foetuses are of prime importance. They throw light on the aetiopathogenetic mechanisms of stillbirths, an understanding of which will greatly help future generations.

Material and Methods

This is a two year autopsy study. The total number of deliveries that took place in our institute during this period were recorded. All the cases of macerated and fresh stillbirths were included in this study. The criteria utilized to identify these stillbirths were: 20 weeks gestation with an Apgar score of 0 immediately on birth with the same score continued over a period of 10 minutes.

The relevant maternal history was recorded in the proforma prepared for this study and the gross and microscopic findings of the placenta and cord that accompanied some of these stillbirths were also recorded in another proforma prepared for this purpose. All the stillbirths were autopsied and autopsy findings were recorded systematically in the same proforma.

The autopsy included microscopic examination of all the organs. After gross dissection and examination of all the organs, tissue sections were carefully selected from them, particularly, those sites which were grossly pathological. When no gross lesions were identifiable on gross examination, the maximum number of sites were chosen for histology.

The placenta, membranes and cord were studied grossly and adequately sectioned for histopathology; 1 mm thin sections were fixed in 10% formalin (20 times the volume of the tissue) for a minimum period of 4 hours. Subsequently, they were processed in the histokinette. 5 micron thin sections were cut from the paraffin blocks and these sections were routinely stained with hematoxylin and eosin (H and E stains). Special stains like reticulin, Massons, PAS, Van Gieson’s and Verhoeff’s Van Gieson’s stain were done wherever necessary.

Composite clinicopathological conclusions were drawn regarding the cause of the stillbirth on the basis of the maternal history, gross and microscopic pathology of the stillbirth, placenta and cord. These individual parameters were correlated and the conclusions drawn from each case were analyzed so as to fulfill the aims of this study. The following table gives the details of the material studied.

Average number of histological sections from each case 20.

Results

The total number of hospital live births during the two years was 3750, the total number of still births during the same period was 131 i.e. approximately 1/29th the number of live births. The still birth rate (SBR) calculated per 1000 live births was 33.75 i.e.= Stillbirths x 1000 / Stillbirths + Live births.

The total number of still births autopsied during this period was 96, giving an autopsy rate of 73.28%. Out of the 96 still births 52 were ante-partum deaths (MSB’s) and 44 were intrapartum deaths (FSB’s). The ratio of MSB’s to FSB’s being 13: 11. the male/female ratio of the whole group was 25: 23 i.e. males predominated slightly. However in antepartum deaths females predominated.

The maximum number of stillbirths were seen in mothers in the age range of 21-25 years, followed by the 26-30 years age range. In fact, the 20-30 years age range has the bulk of the cases i.e. 74/96. The minimum number of stillbirths are in the age range of 31-35 years. There were no cases in the age range of 36 or more.

As far as parity is concerned, the group of para 0-4 had the maximum number of stillbirths i.e. 60, and the group of para 5 or more, surprisingly had only 1 case. In 35 cases, i.e. approximately half the leading group of para 0-4, the mothers were primiparous.

When the stillbirths were split up into the various gestational age categories, the stillbirths who were preterm by gestational age i.e. < 37 weeks predominated: 92.7% out of which 15.73% were < 28 weeks. Among the individual gestational age categories the maximum number of stillbirths were in the 28-33 weeks category: 42.7%. The > 37 weeks category formed the smallest percentage: 4.16%.

When the stillbirths were split up into the different birth weight categories, the maximum number of cases were in the low birth weight category (LBW) i.e. < 2500 gms: 78.12% out of which 26.67% were of very low birth weight (VLBW) : < 1000 gms. The normal birth weight (NBW) stillbirths > 2500 gms were relatively fewer in number : 20.83% i.e. approx. 1/4th the LBW category. Among the individual birth weight categories, the maximum number of cases were in the 1001-1500 gms category 40%. This was followed by the 1000 and > 2500 gms categories 20.83%. Stillborns weighing between 1501- 2500 gms fared better and mortality was lowest in the 1501-2000 gms. and 2001-2500 gms categories i.e. 12.5% and 13.54% respectively.

The total number of cases with maternal risk factors was 70. 37 cases had general risk factors and 40 had pregnancy related risk factors. Anaemia predominated in the general risk factor category; antepartum haemorrhage and pregnancy induced hypertension in the pregnancy related risk factors. In 40 cases, to the disadvantage of the foetus the risk factors were more than 1. When the risk factors were multiple the process of categorizing the exact clinico-pathological cause of stillbirth was complex and needed proper assessment of all the factors i.e. maternal age, parity, general, pregnancy related and labour related risk factors. However, the risk factors were not always directly the causes of stillbirths

The total number of cases with labour related risk factors were 45. Of these, 9 cases had more than 1 risk factor. The labour related risk factor most commonly present was breech presentation - 15/45 cases-33.33% closely followed by abruptio-placentae - 13/45 cases – 28.88%.

Out of the 96 cases studied, the placenta was available for study in 55 cases, i.e. in more than 50% cases. Lesions were detected in 53 cases and no lesions found in only 2 cases. Lesions were detected in 57.7% MSB’s and 53.3% FSB’s indicating the importance of examining the placenta in all stillbirths but especially in cases of MSBs where usually the cause of death may be apparent in the placenta, particularly, if there is no significant positive maternal risk factor.

The commonest gross lesion encountered was infarction (Fig. 1), seen in 17/53 cases- 32.07%. This was encountered slightly more often in MSBs - 9/17 cases-52.94%, where the cause of death was attributable to chronic utero-placental insufficiency, especially maternal pregnancy induced hypertension - 4/17 cases- 23.52%. Infarction is known to be associated with IUGR. However, we encountered IUGR in only 1/17 case- 5.88%. Infarction was closely followed by retroplacental clots - 15/53 cases- 28.30%, mostly seen in the cases of abruptio-placentae - 10/15 cases- 40%.

Similarly the commonest microscopic lesion encountered was inflammation - 25/53

Fig. 1 : Photograph showing extensive placental infarction in an MSBs of 32weeks gestation associated with maternal PIH and h/o stillbirths due to PIH.

cases- 47.16% seen more often in cases of MSBs. Death was attributable to infection in 10/25 cases- 40%. In the rest, inflammation was only an incidental finding or an associated lesion that may have contributed to mortality. Of all the types of inflammation, chorioamnionitis was the commonest -21/25 cases- 84% and in only one case was it picked up on gross examination because of the presence of well formed exudates. Out of the 21 case of chorioamnionitis, none concomitantly showed evidence of perinatal infection i.e. pneumonia indicating that it is not such an important risk factor.

The microscopic lesion next commonly encountered was infarction - 24/53 cases- 45.28%. This figure was higher than the 17 cases 32.07% of infarction encountered on gross, indicating the utility of microscopic examination in recognizing particularly fresh infarcts which may be overlooked on gross examination. We encountered fresh infarcts in 10 cases, healed infarcts in 8 cases and maternal floor infarcts in 8 cases. Some of the cases showed 2 or even all 3 of the above kinds of infarcts. Though retroplacental haemorrhage was seen in 15 cases 28.30% on gross examination, it was confirmed in only 10 cases (18.86%) microscopically. In the rest, the clot must have got dislodged during processing.

Neither the effects, nor the aetiopathogenesis of fibrin deposition in the placenta has been clearly understood, though we found it in a significant number of cases - 21/53 cases (39.62%) especially on microscopy. Both villous edema and cytotrophoblastic proliferation, encountered in 12/53 cases 22.64% and 7/53 cases 13.20% cases respectively, are nonspecific indicators of foetal hypoxia and were seen when there was evidence of foetal hypoxia. The presence of calcification showed a good correlation on gross and microscopic examination 7/53 cases 13.20% each. 6/7 cases 85.71% were seen at or after 32 weeks gestation and only one, an anencephaly, at 28 weeks. Most were FSBs and none were associated with post maturity.

In the 52 cases of MSBs, 30 placentae were studied i.e. > 50%. We encountered the following secondary changes in the placenta following antepartum death; hypovascularity and stromal fibrosis of terminal villi in 15 cases 50% and 14 cases 46.66% cases respectively and sclerosis of stem vessels in 9 Cases 30%. The placental lesions that were of significance and in which the placental pathology contributed significantly 50% cases to death of the foetus were, retroplacental clots, calcification and exudate amongst the gross lesions and infarction, haemorrhage, cytotrophoblastic proliferation and calcification amongst the microscopic lesions.

Among the 55 placentas studied, the total number of cases where either clinical, gross or microscopic cord lesions were detected, was 24 i.e. approx. 50%. Out of these, death was directly attributable to the cord lesion in 11 cases i.e. approx. 1/2. In the rest, death was ascribable to other associated factors and the cord lesion was only incidental. Of the two cases of cord prolapse, in 1 case death was attributable to cord prolapse itself and in the other, to abruptio-placentae in a breech presentation. Among the 8 cases of cord loops around the neck, in 6, death was attributable to loops around the neck, this turned out to be one of the higher risk factors among the cord lesions. In the other two cases death was attributable to disseminated intravascular coagulation (DIC) and pregnancy induced hypertension (PIH) in the mother. In both the cases of velamentous insertion, death was attributable to other causes. In the 5 cases of marginal insertion, death was due to maternal heart disease, abruptio-placentae, breech presentation, anencephaly and hydrocephalus, one case each. In the 10 cases of cord haemorrhage, death was most often attributed to other causes. In 2 cases however, the cause of death was mechanical unknown and in these two, the cord haemorrhage could have contributed to death, as it was quite extensive. Both the cases of funisitis were seen in association with maternal syphilitic infection and in both these cases death was attributable to maternal infection transmitted to the foetus. In the one case of vasculitis, the cause of death remained unexplained, as there was no evidence of infection either in the placenta, membranes or foetus.

The majority of stillbirths in which hypoxic lesions were encountered were associated with extrinsic causes. The commonest were maternal causes where PIH was the leading cause. This was followed by placental causes in which abruptio placentae predominated. In the 7 cord causes the majority were cord loops around the neck. The labour related causes were the least. Out of which 50% were labeled ‘extrinsic perinatal hypoxia – EPH, a term coined by V.V. Joshi.

In the 34 cases where the hypoxic lesions were due to intrinsic causes, the majority were due to infection followed by congenital anomalies 11/34cases 32.35% .When the causes of death were classified according to the Obstetric Aberdeen Classification proposed by Baird and Thomson (1969) and modified by Cole et al (1986) (Table 1), the leading cause of death was “Mechanical” - 18/96 cases (18.3 %). Cord around the neck led this category - 6/18 cases (33.33%) followed by abnormal presentation - 4/18 cases (22.22%). There were also 4/18 cases (22.22%) where the cause of death was mechanical but it was difficult to pinpoint the exact cause, probably because the history was insufficient. The “mechanical” category was followed by the unexplained category in which there were 16/96 cases (16.6 %). This was again the category where sufficient history was not available. The majority of these cases fall into the < 2.5 kg. weight group - 14/16 cases (87.50%).

Pre-eclampsia was the next leading cause of death - 14/96 cases (14.58%), the majority of which were not associated with antepartum haemorrhage (APH) - 11/14 cases (78.57%). Pre-eclampsia was closely followed by APH - 13/96 cases (13.5%). Most of these cases were abruptio-placentae -10/13 cases (76.92%).

Maternal disease was responsible for death in only 12/96 cases (12.5%). Here, maternal infection predominated - 10/12 cases (83.33%). Out of these, 2 were maternal syphilis, 2 malaria and 6 were ascending infection. The other maternal diseases were : 1 case of DIC and 1 case of heart disease.

Congenital anomalies was low on the list of the causes of death - 11/96 cases (11.43%). 4 cases were neural tube defects i.e. anencephaly (Fig. 2) and 7 were other anomalies : 2 diaphragmatic hernias, 1 exomphalos (Fig. 3), 1 Dandy Walker

Table 1 : Classification of the causes of death according to the obstetric Aberdeen classification of perinatal deaths Proposed by Baird and Thomson (1969), Modified by Cole et al (1986)
	;
Fig. 2 :Photograph showing anencephaly (neural tube defect) in an FSB of 28 weeks gestation associated with polyhydramnios and placental infarction in a placenta previa with retroplacental clots.	Fig. 3 :Photograph showing Exomphalos with lower limb skeletal deformity – one of a pair of twins with h/o maternal diabetes. The other twin was a normally formed live birth.

syndrome, 1 hydrocephalus, 1 hydranencephaly and 1 case of single ventricle. Thus most of our cases were related to the central nervous system - 7/11 cases (63.63%).

The miscellaneous group included - 10/96 cases (10.41%). Of these 6/10 cases (60%) were perinatal infections and 4/10 cases (40%) were specific foetal conditions which included 3 cases of non-immune hydrops and 1 case of disseminated intravascular coagulation (DIC).The least number of cases - 2/96 (2.08%) were due to isoimmunization and both were due to Rh isoimmunization.

When the primary causes of death were classified according to V.V. Joshi's Classification (1976), a modification of the classification utilized by the Perinatal Mortality Committee of Quebec Province, Canada (1967) (Table 2), the leading cause of death was attributable to Extrinsic Perinatal hypoxia (EPH) - 44/96 cases (45.83%). This compares well with the leading causes of death in the Aberdeen classification which correspond to the category of EPH, i.e. mechanical - 18, pre- eclampsia - 14, APH - 13 and maternal disease - 12 cases, which total up to a figure of 47. In fact, if the 3 cases of birth trauma which are closely related to EPH, are added to the 44 cases of EPH, it totals to exactly 47.

In the EPH category, the majority were due to maternal factors - 18/44 cases (40.90%). This figure also corresponds to the total figure of 14 cases of pre-eclampsia plus 2 cases of other maternal diseases i.e. 16 cases in the Aberdeen classification, since the cases of maternal infection fall into a separate category in Joshi’s classification.

Table 2 : Classification of the causes of death according to primary causes of perinatal mortality The Perinatal Mortality Committee of Quebec Province, Canada (1967) Modified by V.V. Joshi (1976)

The maternal causes were followed by placental causes - 13/44 cases (29.45%), which again correspond to the 10 + 2 cases of abruptio-placentae and placenta previa respectively of the Aberdeen classification. The next broad category was of infection - 17/96 cases (17.70%). Here again, we have a correlation with the Aberdeen classification in which there were 10 cases of maternal infection and 6 cases of perinatal infection. The unexplained category harboured 14 cases vs. the 16 cases of the Aberdeen classification. This was followed by the category of congenital anomalies - 11 cases, which give the best correlation with the Aberdeen classification of 11 cases as is to be expected. Other specific causes and immunization also give exact correlation of 4 + 2 cases respectively. IUGR was introduced for the first time as a separate category in Joshi’s classification and we had only I case where death was attributable to IUGR. In most of the categories, except for birth trauma and cord lesions, death was higher in the low birth weight (LBW) category than in the normal birth weight (NBW) category. Larger babies are naturally more prone to trauma compared to smaller babies.

The outstanding characteristic of the Wigglesworth classification is simplicity. It is designed to provide elementary but valid data on which to base decisions on alteration of management or request for improved facilities. This 2-stage method of analysis thus combines simplicity with implications for clinical management and in fact can be utilized even in cases where necropsy is not undertaken, though autopsy permits the most accurate categorization. Its value is also enhanced when the pathological subgroups are considered as subsets of primary birth weight categories.

In our series, (Table 3) the leading category is that of birth asphyxia - 36.46% , which has been the leading cause in both the previous classifications. A high percentage in this category indicates need for improvement of intrapartum monitoring and resuscitation. The next common group is the waste basket of “Others”, accounting for 27%. Here the corrective measure is cause-dependent. This is followed by the category of normally formed MSB – 25%, indicating a need for improvement in antenatal care. The category of congenital malformations was the least frequent - 11.46%.

When the second stage of the analysis is applied, i.e. the birth weight categorization, the general trend is that the perinates who died of birth asphyxia were of the higher birth weight categories, since larger babies are more prone to birth asphyxia. The perinates who died of lethal congenital malformations were of the lower birth weight categories.

Table 3 : Classification of causes of death with birth weight split up according to the simplistic Wigglesworth Classification

The normally formed MSBs remained in the middle category of 1001-2500gms.

All the three classifications do show a fair amount of good correlation. The disagreement if at all, is mainly due to differences in interpretation of the sequence of death, minimum information available in order to classify the cause of death, and misunderstanding of the existing information. To a lesser extent, the disagreement was caused by a failure to comply with the rules laid down for classification. The Aberdeen classification is the classification that the obstetricians are comfortable with since it mainly lists the maternal causes. We pathologists are happier with Joshi’s classification since it has the foetal approach and deals with pathological lesions found in the foetus. But the classification that all are content with is the Wigglesworth classification since it is simple and is practically oriented.

Discussion

The still birth rates of communities/regions/nations are known to be very good indices of the efficacy of their health programmes. A quick comparative look at the stillbirth rates of foreign countries and India indicates that we have a long way to go to achieve foreign standards and we are indeed lagging far behind the Western World. This could be attributed to higher standards of health and nutrition, better antenatal care, advances in the management of labour and delivery, better diagnosis and management of antenatal problems in the western world compared to a developing country like ours.

A closer look at the foreign stillbirth rates (SBRs), indicates a significant decline in the SBR in the late 60s and early 70s. This decline has been steadily maintained except perhaps for the 1983 figures of Parsons et al (1990)1. The high SBR noted was in the study of Evans and Smith (1946)2 i.e. 41.8, but this is nowhere near the highest indian figure of 89.1 in the New Delhi study of Dass and Bhargava (1961)3; in fact it is less than half of it. The lowest SBR i.e. 5.92 in the study of Widjaja et al (1988)4, is just about 1/3rd the figure of the lowest Indian SBR of 19 of Arland et al (1988)5. Indeed, the lowest Indian SBR figure of 19 is almost in the range of the higher western SBR figures with the exception of the disproportionately high SBR figure 14.18 of Evans and Smith (1946)2.

Even though readymade nationwide statistics for our country are not available, a scrutiny of the Indian SBR figures indicate some improvement from the mid 60’s onwards except for the high figure of 86.9 in the series of Sultana et al (1975)6 of Aurangabad.

The SBR figures from our institute in Bombay7 have shown a decline in the late 70’s and very early 80’s upto 1981 and then again a rise in 1983, with a small decline in 1984. In the 10 year period from 1984 to 1994, however there has been no significant decline in the SBRs, the figures being 1984 : 32.68 and 1994-95 : 33.75. However our current figures are certainly lower than the relatively higher figures of Sultana et al (1975)6 of Aurangabad i.e. 86.9, Singhal et al (1986)8 of Gorakhpur i.e. 44.65 and Chowdhary et al (1989)9 i.e. 39, but are higher than the lowest SBR figure of 19 of Anand et al (1988)5 of Jamnagar and are significantly higher than the low rates of western literature which approximately lie in the range of 6 to 13. So we seem to be faring better than our Indian counterparts but are certainly lagging far behind our foreign counterparts, and have a long way to go to achieve the standards they have attained.

The value of perinatal autopsies have been questioned time and again not only by clinicians but also by some general pathologists, who feel that the performance of a perinatal autopsy tends to be a routine formality, with repetitive non-specific findings that do not contribute much additional knowledge.10,11,12

Hospitals general autopsy rates in the U.S. have unfortunately shown a dramatic decline from 50% in 1952 to 15% in Craft and Brazy 198613 and Beckwith 198714 13% in 1989. Fortunately the decline in perinatal, neonatal and paediatric autopsies was not as steep as adult autopsies. In fact in the study of Widjaja et al (1988)4 in Auckland, New Zealand the stillbirth autopsy rate was as high as 74.7%. Our autopsy rate of 73.3% compares well with this high figure.

By contrast and fortunately, the utility of perinatal autopsies has also been stressed in literature -Beckwith (1989)17 and is supported by the experience of other workers.13,15,16

So weighing the pros and cons, it is clearly evident that the pros outweigh the cons and therefore it is heartening to know that our clinicians have realized it, as indicated by our high autopsy rates.

As far as the sex ratio is concerned, we had a male-female ratio of 25:23, males being 50 and females 46 in number. Our findings compare well with Western literature. Jakobovits et al (1987)17 reported that in almost all weight groups and all pathogenetic categories, the number of males was higher than that of females. In fact, in Japan, Fujita et al (1994)18 have clearly implicated the male infant to be a risk factor for “congenital anomalies and certain conditions originating in the perinatal period”. However in the Indian studies of Singhal et al (1986)8 in Uttar Pradesh, female babies had not only a higher SBR but also a higher perinatal mortality rate (PNMR) and neonatal mortality rates (NMR). Could this be an indirect reflection of the Indian attitude towards the female infant?

The ratio of antepartum deaths (APDs) to intrapartum deaths (IPDs) was 13:11, which was slightly in favor of antepartum deaths. Similar findings have been observed by various authors. In fact in most of the studies the ratio of APDs:IPDs was disproportionately high in favour of APDs. Magani et al (1990)19 - Galway, also found a much lower percentage of IPDs i.e. 54/325 stillbirths - 16.6% and in the Australian studies of Fleigner (1992)20 similar low percentages of IPDs were noted in Melbourne - 7.7% and Victoria - 27%. These findings seem to absolve the obstetrician at least as far as the conduct of labour and delivery is concerned.

As far as maternal age at conception is concerned, the maximum numbers of stillbirths were seen in mothers, in the age range of 21-25 years: 45.8% cases. This was followed by the age range of 26-30 years : 31.25% cases. Thus the bulk of the cases fall in the age range of 20-30 years 77% cases with a small number 16.66% cases falling in the below 20 year age range and a still smaller number (approx. 1/3rd) 6.25% cases in the above 30 year age range. This is at variance with the commonly acknowledged fact that stillbirths are more common in women below 20 years of age, particularly if they are primiparous and in women above 30 years of age, if they are multiparous. In our study only 13/35 primiparas i.e. approx. 1/3rd mothers were below 20 years of age and 5/61 multiparas i.e. approx 1/12th were above 30 years.

In the case of parity primiparas were approximately half the number of women who were in the group of para 0-4 i.e. 35/60 cases. Other Indian authors, Daga AS et al (1990)7 have found that primigravidity carried a higher risk, but this was only in cases of birth asphyxia. What was surprising was that we had only 1 case of a grand multipara which goes against the findings of Sipila et al (1990)21 of Northern Finland who found the stillbirth death rate in grand multipara to be higher than in women of low parity (1.9% vs 0.9%). He attributes it to a higher incidence of major congenital anomalies in this group. The cause of stillbirth in our one case of grand multipara by contrast was infection.

When our stillbirths were split up into the various gestational age categories, the stillbirths who were pre term by gestational age i.e. < 37 weeks predominated-92.7% cases, out of which 15.7% were < 28 weeks. Among the individual gestational age categories, the maximum number of stillbirths were in the 28-33 weeks category 42.7% cases. The > 37 weeks category formed the smallest percentage i.e. 4/96 cases (4.16%). Our findings are consistent with workers both from the East and from the West. Widjaja et al (1988)4 in his study on stillbirth in Auckland, New Zealand found that stillbirths occurred more commonly in small-for-date infants, at earlier gestation. This trend disappeared by term, but when the risk of stilloirth following term, during each subsequent week of pregnancy was calculated, it was found that this risk increased sharply after 42 weeks of gestation. In our category of > 37 weeks we did not have a single case of > 42 weeks gestation to support or refute this finding.

In the study of Tan KC (1989)22, among the 287 stillbirths at the Toa Payoh Hospital in Singapore, in a 10 year period, 54% were premature, this is lower than our figure of 92.7%. In the Australian series of Fliegner et al (1992)20, 74% perinatal losses occurred before 33 weeks and only 10% after 37 weeks in Melbourne, whereas in Victoria 47% occurred before 33 weeks and a high percentage of more than 35% after 37 weeks of gestation.

Our figures of 57.28% before 33 weeks are closer to the Victorian figure of 47% and 4% of after 37 weeks of gestation to the Melbourne figure of 10%. Harms et al (1994)23 in his study of mortality of premature infants at Gottigen also found that an increase in gestational age of 1 week resulted in a lowered risk of mortality.

When the stillbirths were split up into the different birth weight categories, the maximum numbers of cases were in the low birth weight (LBW) category i.e. < 2500 gms 78.12% cases, out of which 26.67% cases were of Very Low Birth Weight (VLBW) 1000 gms. The normal birth weight stillbirths (NBW) > 2500 gms were relatively fewer in number 20.83% i.e. approximately 1/4th the LBW category. Our findings are in concordance, with most of those reported both in western and Indian literature.

In an autopsy study of 100 cases at Virginia, U.S.A., V.V. Joshi (1976)24 found a slightly lower percentage of 65% of LBW perinates and a slightly higher percentage of 35% of NBW perinates. In Hammersmith hospital,25 the total number of deaths in the LBW category of < 2500 gms is 86.5% which is even higher than our figure of 78.12% and the total number of deaths in the NBW category of > 2500 gms is 13.5%, which is lower than our figure of 20.83%. The Hammersmith series shows a figure of 31 in the < 1000 gms category which is almost 3 times the figure of 10 in the > 2500 gms category. This is in sharp contrast to our findings, which show an almost equal incidence of cases in both the categories i.e. 20 each (20.83%).

The findings of Barson et al (1982)26 from Manchester are very similar to those of ours They found 77% in the LBW category and 23% in the NBW category, and they too got almost equal percentages in the < 1 Kg. (24.5%) and in the > 2.5 kg. (23%) categories. However their percentage of cases in the 1-1.5 kg Category (22%) was lower than the former two, unlike our series which shows a higher incidence in this category.

Even in, eastern literature, the findings encountered are similar. In the study of Tan KC et al (I989)22 of 287 stillborn infants at Toa Payoh Hospital (Singapore), in a 10 year period, 64.2% were of LBW i.e. had birth weights of < 2.5 kgs. If the birth weight categories are split up into < 1500 gms and > 1500 gms, we had a slightly higher per cent (51.79%) in the former category compared to 48.21% in the latter category. These findings are similar to those of the Melbourne figures of Fliegner JR (1992)20, who also found a lower per cent (36%) in the > 1500 gms. category compared to the higher (64%) in the < 1500 gms. category. This contrasts with the figures of the same author, from Victoria in which a higher per cent (55%) was found in the > 1500 gms. compared to the lower 45% in the < 1500 gms. category Magani et al (1990)19 also found a higher per cent (68.6%) in stillbirths weighing > 1500 gms, in his 10-year study of 325 stillbirths.

Only the study of Parsons et al (1990)1 on the socio-economic and ethnic factors in stillbirth and neonatal mortality in the NE Thames region presents conflicting results. He found that in Asian women the mortality was higher in the higher birth weight groups compared to the lower birth weight groups. Indian literature is also not contradictory to our findings. In fact, the figures are almost similar in the autopsy study of 134 cases at Chandigarh by Joshi et al (1979).27

A total of 70 cases had maternal risk factors. 37 cases had general risk factors and 40 cases had pregnancy related risk factors. Anaemia predominated in the general risk factors 21/37 cases (56.75%); APH and PIH predominated in the pregnancy related risk factors 21/40 (52.50%) and 18/40 (45%) cases respectively. In 40 cases, to the disadvantage of the foetus the risk factors were more than one.

In the general factors we found a VDRL positivity in only 2 cases, whereas Aiken CG (1992)28 from central Africa has reported that 11% of mothers booking in the antenatal clinics had a positive syphilis serology and they delivered 101/466 (21.67%) stillbirths most of whom died prematurely before labour and often had abdominal distension. We had only 3 cases of malaria in the mother but Taha et al (1993)29 from Central Sudan have reported that the risk of stillbirth (particularly MSB) was significantly higher among women who reported malaria attacks in the first and second trimester of pregnancy and a reduced risk was associated with attacks in the third trimester. However, he has mentioned that there was no overall association between perinatal mortality and malaria. In our study though all the three cases were MSBs, all occurred in the third trimester. In Indian literature, in the series of Singhal et al (1986)8, maternal diseases figured only in 20.93% of the causes of stillbirths.

As far as pregnancy related risk factors are concerned, in the series of Tan KC et al (1989)22 from Singapore, 8.4% mothers had a past history of 1 to 3 abortions and 5.6% a past history of one or more stillbirths. In our series also there were more cases with a past history of abortion (9 cases) rather than stillbirths (5 cases).

In Indian literature Singhal et al (1986)8 also found a high percent of toxaemia (25.58%) and APH (20.93%). Daga et al (1990)7 found that PIH and APH carried a higher risk but only in cases of birth asphyxia. In our series in 8/21 cases (38.09%) of APH and 5/18 cases (27.77%) of PIH i.e. in approx. 1/3rd cases, death was attributed to birth asphyxia.

As far as labour related risk factors are concerned, the total number of cases with labour related risk factors were 45. Out of these 9 cases had more than one risk factors. The risk factor most commonly present was breech presentation - 15 cases, closely followed by abruptio-placentae - 13 cases. In the study of Aikea C.O. (1992)28 from Central Africa, one quarter of the deaths from asphyxia were caused by stuck head in breech delivery, prolapsed cord and retained second twin. Though our leading labour related risk factor was breech presentation, we had only 2 cases of cord prolapse and none of retained second twin. In our cases, however only 2/15 cases of breech presentation but 1/2 (50%) cases of cord prolapse, had perinatal death attributable to birth asphyxia.

Out of the 96 cases studied, the placenta was available for study in 55 cases i.e. 50%. Some sort of a lesion, either gross, microscopic or secondary was detected in most of the cases i.e. 53/55 cases (96.36%), indicating the importance of examining the placentae in all cases of stillbirths. Though we would expect lesions to be most commonly encountered in MSBs, we found an almost equal frequency of lesions detected in placentas of MSBs (57.7%) and FSBs (53.3%).

The commonest gross lesion encountered was infarction 17/53 cases(32.07%) (Fig. 1) seen more often in MSBs 9/17 cases (52.94%). The cause of death was attributable to chronic utero-placental insufficiency, especially PIH in 4/17 cases (23.52%). Infarction is known to associated with IUGR, however we encountered IUGR in only 1/17 cases (5.88%). Infarction was closely followed by retro-placental clots 15/53 cases (28.30%) most seen in cases of abrubtio-placentae 10/15 cases (66.66%). Our findings are similar to those of Tingle and Wigglesworth (1986)30 who also found infarction and retro-placental haematomas as most frequent contributors to foetal deaths. Hafalto et al31 found placental failure in 50% cases. Raylam et al32 found significant histologic alteration in 98% cases important amongst these being vascular insufficiency, retroplacental haematomas and acute chorioamnionitis with foetal involvement . Han LC et al33, found placental and cord abnormalities in 75.5% cases and Incarpi et al34 in 36% cases. Tetley et al35 found large placental infarcts, villous oedema, increased intervillous fibrin deposition subchorionic haemorrhage, maternal floor infarction and placental infection in their study.

In the 52 cases of MSBs 30 placentae were studied i.e. more than 50%. We encountered the following secondary changes in the placenta following antepartum deaths: hypovascularity and stromal fibrosis of terminal villi in 15 and 14 cases respectively and sclerosis of stem vessels in 9 cases. Agapitos et al36 found vascular insufficiencies in cases of IUD in 16.47%

Among the 55 placentas studied cord lesions were detected in 24 cases i.e. approx. 50%. The majority of the lesions were detected on gross examination i.e. in 17/24 cases (70.83%). Microscopic pathology was seen in 15/24 cases (62.5%) and clinical pathology in 10/24 cases (41.66%). Out of the 24 cases of cord lesions, death was attributable to the lesion in about half the cases (11/24). This was more commonly encountered in the clinical lesions of cord prolapse and loops around the neck - 7/10 cases (70%) followed by microscopic lesions 4/15 cases (26.66%) in which 2 cases were of syphilitic infection and two unknown mechanical factors, in which extensive cord haemorrhage may have contributed to death. As far as the gross lesions are concerned, they did not seem significant, especially the abnormal insertion.

Calcification of the umbilical artery was seen in 2 cases and in both death was attributable to other factors (maternal PIH and cord loop around the neck). In a study by Khong et al (1989)37 the clinical and pathological features of 5 cases of calcification of the umbilical arteries were studied and they have described 2 distinct lesions produced by the calcification. One was sclerosis of the wall (3 cases) and the other, obliteration of the lumen (2 cases). We did not find evidence of sclerosis or obliteration in both our cases. In the sclerotic variant the aetiology has been attributed to inflammation in the cord, vessels, membranes and decidua and in the obliterative variant to thrombosis. Both these findings were not detected in our cases. In the study of Widjaja et al (1988)4 on stillbirths in Auckland, cord problems were rare and occurred mainly in babies > 2500 gins. The absence of postmortem evidence of anoxia in the majority of these babies suggested that the cord complications occurred as a postmortem phenomenon. In our series only 8/24 i.e. 1/3rd cases with cord lesions were > 2500 gms. However in 13/24 i.e. approx. 50% cases death was attributable to birth asphyxia. Hovalto et al31 found cord complication in 12% cases

As far as the Aberdeen Classification (Table 1) is concerned, the largest number of cases fall into the category of mechanical causes - 18/96 cases (18.75%). This coupled with the categories of pre-eclampsia – 14/96 cases (14.58%) and APH – 13/96 cases (13.54%) form a sizably large leading category - 45/96cases (approx. 50%). These conditions lead to foetal hypoxia either in an acute or chronic manner. Unexplained cases, figured next – 16/96 cases (16.66%) whereas maternal disease – 12/96 cases(12.50%), congenital anomalies – 11/96 cases (11.45%) (Figs. 2 and 3) and the miscellaneous group which included foetal infection were not so frequent. The rarest cause of death encountered was isoimmunization - 2 cases, both of which were of Rh factor. The studies of Hovalto et al 31 and Windu et al38 also found birth asphyxia to be high 38% and 56% respectively. Congenital malformation to be infrequent 17% and 32% respectively but Han et al33 found congenital malformation in a high percentage of 50%.

In the study of Tingle and Wigglesworth,30 the causes of death were divided into maternal causes, placental causes and foetal causes. In the maternal causes, eclampsia or albuminuria featured the most. In our series, it was third in the list. In placental causes, placenta previa, retroplacental haematoma and infarction were common. Though we had only 2 cases of placenta previa, retroplacental haematoma were seen more frequently - due to abruptio-placentae without pre-eclampsia (10 cases) and with pre-eclampsia (3 cases). Infarction however was one of our most frequent placental lesions, both grossly and microscopically. The foetal causes were trauma, infection and anomalies, all 3 of which were infrequently encountered in our study.

In the study of the classification of perinatal deaths by Jean Keeling (1989).39 The largest number of cases fell into the category of “Unexplained” - 108/239 cases (45%), of which approximately 80% were premature < 2.5 kg. and 20% were mature 2.5 kgs. In our series, the unexplained category was the second most frequent - 16/96 cases (16.66%) and about 88% were premature i.e. < 2.5 kgs 12% were mature i.e. 2.5 kgs. The next common individual categories was malformation - 46/239 cases (19.2%). In our series, congenital malformation was low on the list. This was followed by APH - 26/239 cases (10.875), toxaemia - 18/239 cases (7.53%) and mechanical causes - 13/239 cases (5.43%). In our series the order is just the reverse - mechanical followed by pre-eclampsia followed by APH however, when considered as a combined group causing foetal hypoxia, it features second - 57/239 cases (23.84%) in Keeling’s series and first in our series. Maternal disease, miscellaneous causes and Rh isoimmunization did not occur so frequently in Keeling’s series as well as in our series.

In a five decade study of international trends in relation of perinatal mortality and congenital malformations, Kalter et al (1991)40, compared European, American and Canadian data, from 1950, he found that congenital malformation was among the prominent causes of death which accounted for 30% of perinatal deaths. In our series, congenital malformation accounted for only 11% deaths, which is even lower than the series of Fleigner JR (1992)20 which had 19.1% cases of congenital anomalies. In his series, APH accounted for 14% of deaths, which is approximately the same as our figure of 13%. Hypertensive disorders were 9.7%, whereas we had no case of essential hypertension, but approx. 14% cases of pre-eclampsia. Again the series of Pauli et al (l994)41 showed a rather high percent (44% cases) of congenital malformation.

Japanese literature also shows congenital malformation to be a leading cause of death, along with anoxia and maternal causes, Nakamura et al (1982)42. However, Indian literature reveals figures which are more compatible with ours as far as congenital malformations are concerned. In the series of Singhal PK et al (1986)8, congenital malformations contributed to 11.63% of the causes of death, however his figures for toxaemia (25.58%), APH (20.93%) and maternal disease (20.93%) are higher than our figures of 14%, 13% and 12% respectively.

When the causes of death were classified according to V.V. Joshi’s classification (Table 2), the leading cause of death was Extrinsic Perinatal Hypoxia which accounted for 44/96 cases (45.8%), this was followed by infection in 17/96 cases (17.7%). The unexplained category accounted for 14/96 cases (14.6 %), and congenital anomalies - 11/96 cases (11.45%). Our findings are at variance with the Western series of Joshi (1976)43 from Virginia, U.S.A., but in concordance with his Indian series from P.G.I., Chandigarh (1979)27. In his Western series congenital anomalies was second among the causes of death (28%), followed by EPH (17%), whereas infection and unexplained causes featured rarely - 2% each. In his Indian series, however EPH headed the list (43.3%) similar to our figure of 44%. However, in the further split up, placental causes were the highest (14.2%) followed by the unexplained causes (11.2%) and labour causes (9%). In our series, maternal causes headed the list (18%), followed by placental causes (13%) and cord lesions (7%). In Joshi’s series, infection also featured second (17.9%) and congenital anomalies occurred (i.e. 10.5% cases), similar to our series.

In the rest of the western literature, congenital anomalies feature more prominently compared to our series Widlaja et al (1988)4 and Kalter et al (1991)40 - 30%. However, in Central Africa (Aiken -I992)28, infection headed the list, with congenital syphilis figuring prominently (20.5%). This was followed by birth asphyxia (18.8%). In our series, though these were also the leading causes of death, the order was reversed. Unexplained causes featured third in both series, and their rate of congenital malformation were lower (7.7%) than our series which was 11%. In the East i.e. Singapore (Tan K.C. et al -1989)22, intrauterine anoxia and congenital malformation were the commonest diagnosis, but they had a high (48.1%) of unexplained causes. compared to our figure of 14% and they attributed it to maceration. When the causes of death were classified according to the Wigglesworth Classification (Table-3), the leading cause of death was asphyxial conditions - 35/96 cases (36.45%) followed by the “Others” category - 26/96 cases (27%) closely followed by normally formed macerated stillbirths - 24/96 cases (25%). The least frequent was the category of lethal congenital malformations - 11/96 cases (11.45%).

Our findings, when classification was attempted according to this method was again not in concordance with western literature but was similar to Indian figures. In the study of Wiggles-worth (1980)25 from London the percent of MSBs - 17/74 cases (22.9%) was higher than asphyxial cases -12/74 cases (16.2%). The latter was similar to congenital malformation -10/74 cases (13.5%) but their figure of “others” category was the lowest - 3/74 cases (4%) unlike ours which was second in order. When the birth weight was considered, the largest group was 1001-2500 gms - 55/96 cases (57.3%). This was followed by an equal number of > 2500 gms and d”1000 gms - 20/96 cases (20.83%) each. In the series of Wigglesworth (1980)25 the 1001-2500 gms category was again the highest - 33/74 cases (14.59%) but this was closely’ followed by the < 1000 gms. category - 31/74 cases (41.9%) and the > 2500 gins. category was less frequent, almost 1/3rd this figure - 10/74 cases (13.5%).A most resent classification of Gandori et al44 is the Re Co De (relevant condition at death) in this classification unexplained cases are only 15.2% compared to the 66.2% in Wigglesworth classification.

In the series of Barson et al (1984)26 from Manchester, normally formed MSBs topped the list again - 127/440 cases (28.86%). However their rate of malformation was high - 109/440 cases (24.7%), even higher than the figures of Wigglesworth. Their per cent of asphyxia was lower - 72/440 cases (16.4%) and the ‘others’ category was the lowest - 25/440 cases (5.7%). However their birth weight split up was similar to ours, the highest group was 1 - 2.5 kgs - 231/440 cases (52.5%) followed by an almost equal percent of < 1 kg and > 2.5 kg. - 108/440cases (24.5%) and 101/440 cases (22.9%) respectively. In the series of Keeling et al (1989)39, also the category of normally formed MSB headed the list - 53/239 cases (22.2 %), and this was closely followed by congenital malformation - 51/239 cases (21.3%), whereas birth asphyxia was the lowest figure - 30/239 cases (12.5%); along with the “others” category.

Thus western literature appears to have a higher rate of congenital malformation and a lower rate of birth asphyxia. In eastern literature Tan KC – 1989 22, however both the rates of congenital malformation and intrauterine anoxia are high, whereas the figures of Central Africa - Aiken CC (1992)28 are closer to ours except that infection (33.5%) predominates especially congenital syphilis. This is followed by birth asphyxia (18.8%) and the rate of congenital malformations is low.

A look at the Indian figures of the causes of perinatal mortality shows concordance with our figures. The study of AR Joshi (1988)45 from our institute in Bombay shows a high per cent of birth asphyxia 360/803 cases (40.3 %). This is followed by normally formed MSBs 181/893 cases (20.3 %) and the rate of congenital malformations is low - 32/893 cases (3.6%). In the study of Raghuveer (1992)46 from Bangalore, also, birth asphyxia was the leading cause of death and accounted for 24% of the cases. Both these authors have utilized the classification of Wigglesworth and have found it most action-oriented with clear implications for clinical management. The importance of autopsy cannot be underestimated as amply proved by the studies of Incarpi et al34 and Agapitos et al36, in the former study the causes of death remained unexplained after autopsy in 31% and in 44% without autopsy and in the later autopsy gave the diagnosis in 12.9% cases.

To quote Raghuveer (1992)46, “The Classification of perinatal deaths using Wigglesworth classification appeared to be a practical and problem-oriented system. It also carried clear implications for improving perinatal care. The adaption of this method of classification by all major hospitals is recommended, so that easy comparisons can be drawn over time and between different centers”. Our practical experience is the same and hence we fully endorse this statement.

Conclusion

In the 2-year study period, there were 3750 live births and 131 stillbirths. The stillbirth rate was 33.75 per 1000 live births. Out of the 131 stillbirths, 96 were autopsied. The autopsy rate was 73.28%. The placentae and cords were available for study in 55 out of the 96 stillbirths studied, i.e. > 50% cases. Out of the 96 cases studied, 50 were males and 46 females giving a male/female ratio of 25:23, slightly in favour of males. 52 stillbirths were antepartum deaths and 44 were intrapartum deaths, the ratio being 13:11, slightly in favour of antepartum deaths. However in males, the intrapartum deaths were slightly more than the antepartum deaths. The maximum number of deaths 45.83% occurred in the 21-25 years maternal age range and in the 0-4 parous mothers 62.5%. There were only 36.5% of primiparous mothers and only one case of a grand multipara. The maximum number of stillbirths were of low gestational age (< 37 weeks) - 02.7% out of which 15.73% were < 28 weeks. Among the individual gestational age categories, 28-33 weeks headed the list : 42.7%. The majority of stillbirths were of LBW i.e. < 2500gms - 78.12%. Out of these 26.67% were VLBW i.e. < 1000 gms. The normal birth weight category i.e. > 2500 gms. showed a low per cent of 20.83% . Among the individual birth weight categories, the maximum number of cases were in the 1001-1500 gms category. There were 71.19% cases, with maternal risk factors. 37 had general risk factors and 40 had pregnancy related risk factors. In 40 cases, more than 1 risk factor was operational. Anaemia predominated in the general risk factors – 56.75%, and APH and PIH in the pregnancy related risk factors 52.5% and 45% respectively. There were 45 cases of labour related risk f out of which 9 cases had more than 1 risk factor. The labour related risk factor most commonly present was breech presentation – 33.33% cases, closely followed by abruptio-placentae – 28.88% cases. Of the 55 cases in which the placenta was studied, lesions were detected in a high per cent of 96.36%, slightly more frequently in MSBs 57.7%, than in FSBs 53.3%, indicating the importance of examining the placenta in all cases of stillbirths. The commonest gross lesion encountered was infarction - 32%, closely followed by retroplacental clots - 28.3%. The commonest microscopic lesion encountered was inflammation - 47.2%, out of which chorioamnionitis was the most frequent - 84%. This was closely followed by infarction - 45.3%. Out of 53 placentas, 30 belonged to MSBs i.e. > 50%. In these cases, we encountered secondary changes like hypovascularity and stromal fibrosis of terminal villi and stem vessel sclerosis and obliteration in about 50% of the cases. The placental lesions which were of significance and contributed to perinatal deaths were retroplacental clots, infarction and infection. Cord lesions were detected in 43.6% placentae studied. Out of these death was directly attributable to approximately half of the cases i.e. 11/24 cases. Though cord haemorrhage was most frequently seen - 41.67%, it was responsible for death in only 20%, whereas, the second commonest finding of cord loops around the neck – 33.33% was responsible for deaths in 75%, so it turned out to be the highest cord risk factor. The majority of stillbirths in which hypoxic lesions were encountered were associated with extrinsic causes - 63%. The commonest of which were maternal causes - 31%, of which PIH was the leading cause - 22.4%. This was followed by placental causes - 20.7% in which abruptio-placentae predominated - 83.3%. In the intrinsic causes, infection predominated - 50% followed by congenital anomalies - 32.35%. When the causes of perinatal deaths were classified by the Aberdeen, Joshi and Wigglesworth classifications, we found a fairly good correlation between all the 3 classifications. However, since the first two were the split up of the causes of death more in detail, the correlation was better. But the Wigglesworth classification proved to be more action-oriented with clear implications for clinical management. The commonest cause of death was related to “asphyxial lesions”, both of acute and chronic nature. In the Aberdeen classification it accounted for 57.3%, in Joshi’s classification - 45.8%, labelled as Extrinsic Perinatal Hypoxia (EPH) and in the Wigglesworth classification - 36.4%. The next commonly encountered cause of death was infection. This was seen in 16.6% of the Aberdeen classification. Out of these 62.5% were maternal and 37.5% were perinatal. We had two cases of maternal syphilis, two of maternal malaria and five cases of vaginal candidiasis. The unexplained category came next. In the Aberdeen classification it accounted for 16.7% cases, in Joshi’s classification - 14.58%, but in Wigglesworth classification the figure of “Others” was higher since it included specific conditions in addition to unexplained causes. Congenital malformations were relatively infrequent in all these classifications - 11.45%. This category was the easiest to classify and gave the best correlation. Conditions like birth trauma, immunization and IUGR were rarely encountered. When divided into LBW and NBW we found a much higher percent of deaths in the LBW group - 75% compared to the NBW group - 25%. Our findings correlate well with indian literature, in which the leading cause of death is related to asphyxial lesions, followed by infection but are at variance with western literature were these 2 causes do not feature prominently but instead congenital malformation takes the lead in many series. This indicates that we are still battling with preventable causes like birth asphyxia and infection and still have a long way to go to achieve Western standards as far as perinatal mortality is concerned.

References

1. Parsons L, Duley L, Alberman E. Socio-Economic and Ethnic Factors in Stillbirths and Neonatal Mortality in the N.E. Thames Regional Health Authority (NETRI IA). Br J Obstet Gynacol 1990; 97 : 237-44.
2. Evans M, Smith GC. Stillbirths and Neonatal Death : A Clinicopathological Study. J of OBGY 1946; 53 : 441-52.
3. Dass A, Bhargava VL. Statistical Review of Stillbirths. J Obstet Gynacol Ind 1961; 12 : 172.
4. Widjaja AI I, Mantel CD. Stillbirths in Auckland, New Zealand Medical Journal 1988; 101 : 199-207.
5. Anand JS, Javadekar BB, Belani M. Congenital Malformations in 2000 Consecutive Births. Ind Paed 1988; 25 : 845-52.
6. Sultana Z,Talib V.1 I, Patil SD. Perinatal Mortality -A Preliminary Report of an Autopsy Study. J Obstet Gynecol Ind 1975; 25 : 331.
7. Daga AS, Daga SR, Patole SK. Risk Assessment in Birth Asphyxia. J Trop Paediatr 1990; 36 : 34–9.
8. Singhal PK, Mathur GP, Mathur S, Singh VD. Perinatal Mortality in ICDS Urban Shun Area. Ind Paediatr 1986; 23 : 339.
9. Chaudhary N,Pal S, Roy A. Mortality Pattern in Babies Delivered by Caesarean Sections and Vaginal Delivery. Ind Paediatr 1989; 26 : 247.
10. Singh NI, Deorari AK, Murali MV, Paul VK. Are Neonatal Necropsies useful Developing Countries? Lancet 1988; 2 : 502.
11. Burrows S. The Post-Mortem Examination: Scientific Necessity or Folly? JAMA 1975; 233 : 441.
12. Roberts WC. The Autopsy: Its Decline and a Suggestion for its Revival. NEJ Med 1978; 299 : 332.
13. Craft 11, Brazy JE. Autopsy - High Yield in Neonatal Population. J Dis Child 1986; 140 : 1260.
14. Beckwith JB. The Value of Paediatric Postmortem. Paed Clin N Amer 1989; 36 : 29.
15. Gau G, The Ultimate Audit, B.M.J. 1977; 1 : 1580.
16. Meir PR, Manchester DK, Shikes R.11, Clewell WK, Shewart M., Perinatal Autopsy : Its Clinical Value. Obst and Gyn 1986; 67 : 349.
17. Jakobovits A, Jakobovits AA, Viski A. Sex Ratio of the Stillborn Foetuses and Neonates Dying in the First Week : Early Hum Dev 1987; 15 : 131–5.
18. Fujita T, Minowa M, Minra Y. A Record Linkage Study on Risk Factors for Cause Specific infant Mortality. Nippon Koshu Eisei Zasshi 1994; 41 : 114-25.
19. Magani IN1, Rafla NM, Mortimer G, Meehan FP. Stillbirths : A Clinicopathological Survey. Paediatric Pathology 1990; 10 : 363-74.
20. Fliegner JR. Perinatal Mortality in a Tertiary Obstetric Institution, Aust. N-Z-J-Obs Gyn 1992; 32 : 309-12.
21. Sipila P, Von-Wendt L, Hartikainen Sorri AL. The Grand Multipara - Still an Obstetrical Challenge? Arch Gynecol - Obstet 1990; 247 (4) : 187-95.
22. Tan KC, Ho NK, Tay BL. Stillbirths: Ten Years Experience at Toa Payoh Hospital, Singapore Med J 1989; 30 (2) : 151-4.
23. Harms K, Osmers R, Kron NI, Schill M, Kuhn W, Speer CP, Schroter W. Mortality of Premature Infants 1980-1990 : Analysis of Date from the Gottingen Perinatal Centre : Z Geburt- shiye. Perinatol 1994; 198 (4) : 126-33.
24. Joshi VV. Primary Causes of Perinatal Mortality : Autopsy Study of 100 Cases. Arch Pathol Lab Med 1976; 100 : 106.
25. Wigglesworth JS. Monitoring Perinatal Mortality a Pathophysiological approach. Lancet 1980; 2 : 684.
26. Borson AJ, Tasker, Lieberinan BA, Hillier VF, Impact of Improved Perinatal Care on the Causes of Death. Archives of Disease in Childhood 1984; 59 : 199-207.
27. Joshi VV, Bhakoo ON, Gopalan S, Gupta AN. Primary Causes of Perinatal Mortality - Autopsy Study of 134 cases. Indian J Med Res 1979; 69 : 963-71.
28. Aiken CG. The Causes of Perinatal Mortality in Bulawayo, Zimbabwe. Central African J Med 1992; 38(7) : 263-81.
29. Taha T el T and Gray RH. Malaria and Perinatal Mortality in Central Sudan. Am J Epidem 1993; 138 (8) : 563-8.
30. Wigglesworth JS, Tingle CD. A Contribution to the Study of the Causation of Foetal Death. Arch of Dis of Childhood 1986; 61 : 962.
31. hovalto O, Lipasti A,Rapole J,Kanjalairian O. causes of stillbirth, a clinicopathological study of 243 patients. Br J Obstet Gynaecol 1983; 90 :691-6.
32. Rayburn W, Sandur C, Barr M Jr, Rygiel R. The stillborn fetus : placental histologic exam in determining causes. Obstet Gyenecol 1985; 65 : 637-41.
33. Han LC, Langner AC, Stiehl P, Wiltekind C, Falor R. identification of the causes of IUD during 310 concecutive autopsies. Eur J Obstet Gynecol Reprod Biol 2004; 113 : 134-8.
34. Incarp MH, Miller DA, Samadi R,Siltlage RH, Goodwin TM. Stillbirth evaluation:what tests are needed? Am J Obstet Gynecol 1999; 180 : 1595-6.
35. Teltey Y,winedu EK. Autopsy studies on still births in korle Bn Teaching Hosppathological finding in stillbirths and their placentae. West Afr J Med 1997; 16 : 93.
36. Agapitor E, Papadoporlon C, Kawantzas N, Paporlias J, Automaki V, Davari P. Andr Anat Cytol Pathol 1996; 44 : 5-11.
37. Khong TY, Dilly SA. Calcification of Umbilical Artery : Two Distinct Lesions. Clin Path 1989; 42 (9) : 931-4.
38. Windu EK, Teltey Y. Autopsy studies on stillbirths in Kolbe Teaching Hospital II: Causes of death in 93 stillbirths. West Afr J Med 1998; 17 : 148-52.
39. Keeling JW, MacGillary I, Golding J. Classification of Perinatal Deaths. Archives of Diseases of Childhood 1989; 64 : 1345-52.
40. Kalter H. Five Decade International Trend in the Relation of Perinatal Mortality and Congenital Malformations : Stillbirths and Neonatal Death Compared, International J Epidemiol 1991; 20 (1) : 173-9.
41. Pauli RM, Reiser CA, I. Wisconsin Stillbirth Service Program,II. Analysis of Diagnoses and Diagnostic Categories in the First 1000 Referrals, Am J Med Genet 1994; 50 (2) : 135-53.
42. Nakamura Y, Yano H, Nakoshima T, Nakoshima 1-I, Kayama M. Primary Causes of Perinatal Death. Human Path 1982; 13 : 54-61.
43. Joshi VV. Primary Causes of Perinatal Mortality: Autopsy Study of 100 Cases. Arch Pathol Lab Med 1976; 100 : 106.
44. Gandori J, Kady SM, Mc Geown P, Francis A, Torks A. classification of stillbirth by relevant condition at death (Re Co De) : population bases cohort study. BMJ 2005; 331 : 1113-37.
45. Joshi AR, Daga SR, Daga AS. Perinatal audit through Wigglesworth’s classification. Indian Pediatrics 1988; 25 : 525-9.
46. Raghuveer G. Perinatal deaths: Relevance of Wigglesworth’s classification. Pediatric Perinatal Epidemiology 1992; 6 (1) : 45-60.

*Associate Professor; **Resident Pathologist, Department of Pathology, Grant Medical College, Byculla, Mumbai - 400 008.