An autopsy study of maternal mortality: A tertiary healthcare perspectiveTS Panchabhai, PD Patil, DR Shah, AS Joshi
Department of Pathology, Seth GS Medical College and KEM Hospital, Parel, Mumbai - 400 012, India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.48434
Source of Support: None, Conflict of Interest: None
Background: An audit of autopsies of maternal deaths is important for the establishment of accurate cause of maternal deaths and to determine the contribution of various etiologies responsible in a given community. Aim: To study the causes of maternal deaths as determined by a pathological autopsy. Settings and Design: A retrospective study of all the cases of maternal deaths that underwent a pathological autopsy in a tertiary healthcare center from January 1998 to December 2006. Materials and Methods: The autopsy records with clinical notes were retrieved; gross and histopathology specimens and slides were studied to establish the accurate cause of maternal deaths. The variables like age (years), stay in the hospital, gravidity, trimester of pregnancy and method of delivery were used to classify and analyze the data from the autopsies. The causes of maternal deaths were divided in to direct and indirect; each being classified into subgroups based on the most evident pathology on autopsy. Results: The Maternal Mortality Rate (MMR) over a nine-year period (1998-2006) was 827/100000 live births (471 maternal deaths against 56944 live births). An autopsy was performed in 277 cases (58.8%). In the autopsy group, the most common causes of maternal mortality were pre-ecclampsia/ecclampsia (40 of 277, 14.44%) and hemorrhage (32 of 277; 11.55%); However, indirect causes like infectious diseases (27 of 277; 9.75%) and cardiac (27 of 277; 9.75%) disease also contributed to maternal deaths. Conclusion: Indirect causes like rheumatic heart disease and infections like tuberculosis, malaria or leptospirosis and nutritional anemia are still major causes of maternal mortality in developing countries like India. Intensive efforts need to be taken in these areas to reduce the maternal mortality in developing countries like India.
Keywords: Direct and indirect causes, etiology of, maternal mortality
A maternal death is defined as death of a woman occurring while pregnant or within 42 days of termination of pregnancy, irrespective of the duration or its management but not from accidental or incidental causes.  Maternal mortality rate (MMR) is dependent upon the general socioeconomic status, nutrition level and the level of maternal healthcare in the community. It is recognized as a social indicator, and there is a large gap between the MMR in developed countries and that in developing nations. The World Health Organization reports that the MMR in the South Asian region ranks second only to sub-Saharan Africa.  Together, these two regions account for 86% of global maternal deaths. The causes of maternal deaths have been classified as direct (resulting from obstetric complications of pregnancy, labor or puerperium) or indirect (resulting from preexisting disease or disease aggravated by the physiological effects of pregnancy) depending upon their relationship with pregnancy.  This classification, however, depends on accurate establishment of the cause of death by a pathological autopsy correlated with the clinical diagnosis. Our center, King Edward Memorial Hospital, (1800-bed tertiary care teaching Institute in Mumbai) serves as a referral center for patients from all over India. Being an apex referral center, the complications as well as the time of referral may have a significant effect on the final pregnancy outcomes. Pathological autopsy is an important tool in the investigation of maternal death. The autopsy provides valuable information about the pathophysiological changes in various organs which may be important in the delineation of the sequence of events leading to death. However, without accompanying clinical data, its utility cannot be well defined. The present retrospective study was designed to determine the MMR and to study the cases that were subjected to pathological autopsy.
The study was conducted after approval from the Institutional Review Board. The number of maternal mortalities and deliveries from January 1998 to December 2006 were recorded. The MMR was then calculated for each year [Table 1]. All the cases of maternal deaths autopsied by the pathology department were included in the study. The cases which were not autopsied and those which were subjected to medico-legal autopsy (caused by burns, trauma, suicide, homicide etc) were excluded from the study. The need for pathological autopsy in each case was determined by the treating obstetrician. In all maternal deaths, autopsies were requested and were definitely performed when death was sudden, unexpected and not in line with the clinical workup.
Whereas the protocol followed during autopsy in a maternal mortality is similar to other pathological autopsies, other specific changes that were sought were frothy bubbles in the right atrium in search of air embolism, detection of acute fatty liver of pregnancy or evidence suggestive of pulmonary or amniotic fluid embolism. All cases underwent an external and in situ examination followed by dissection and preservation of the organs in 10% formalin. Blood culture and culture of other specimens was done when indicated. Multiple cultures were not obtained due to resource constraints. A gross examination of the organs and histopathological examination of at least one block each from the cerebrum, cerebellum, meninges, heart, liver, spleen, kidney, stomach, intestines, pituitary and adrenal glands was carried out. Multiple sections were studied from the lungs in an attempt to identify pulmonary emboli. Whenever indicated, more sections from these organs were studied. Paraffin sections were then stained and examined (Hematoxylin and Eosin; special stains like Ziehl Nielsen, Periodic Acid Schiff and Oil Red O when indicated). Autopsy findings were correlated with clinical details and investigation in each case to establish an accurate cause of death. Age, gravidity, hospital stay, trimester of pregnancy and method of delivery were noted [Table 2].
All cases were then classified into direct or indirect causes which included subgroups based on the primary pathology evident on autopsy [Table 3]. The direct causes of maternal mortality included the causes which were classified into: hemorrhagic disorders of pregnancy, pre-ecclampsia or ecclampsia, hepatic disorders due to pregnancy, amniotic fluid embolism, pulmonary embolism, abortion-related causes, puerperal sepsis, and intra-uterine fetal death (IUFD)-induced maternal deaths. The indirect causes of death were classified further into hepatic, pulmonary (excluding tuberculosis), neurological, cardiovascular, renal, hematological, gastrointestinal, malignancy and infectious disease.
Four hundred and seventy-one maternal deaths were recorded in association with 56944 live births during the study period giving an MMR of 827/ 100000 live births. The 277 cases that underwent autopsy, formed the per protocol set. [Table 2] depicts the variables in the cases that underwent an autopsy. When maternal death occurred within 24 h of hospital stay, the direct causes of death predominated (51 of 87; 58.62%). Most of the deaths which occurred after 24 h of ward stay were due to indirect causes (107 of 190, 56.32%).
[Table 3] depicts the direct and indirect causes of maternal deaths in the present autopsy study. Pre-eclampsia/eclampsia (40 cases) and hemorrhage (32 cases) were the principle contributors to direct causes of maternal deaths. Of the 40 cases of pre-eclampsia/ eclampsia on autopsy, 17 had evidence of disseminated intravascular coagulation (DIC) while hepatic failure [Hemolysis Elevated Liver Enzymes and Low Platelet count (HELLP) syndrome] was seen in 10 cases.
Frozen section and fat stain Oil Red O determined acute fatty liver in seven cases. They died due to bleeding and hepatic encephalopathy. In three cases of amniotic fluid embolism, two had definitive evidence on lung histology; while, in one case, it was a presumptive clinical diagnosis with negative autopsy findings in all organs including lungs. Nine cases were diagnosed to have pulmonary embolism as the cause of death. All cases however were not noted on gross examination, four pulmonary embolism cases being detected only after lung microscopy. The diagnosis of puerperal sepsis was established only after positive culture results.
Rheumatic heart disease (21) and tuberculosis (26) were the prominent indirect causes of death. Besides tuberculosis, malaria and leptospirosis were other common infectious conditions associated with maternal death. Of the 26 autopsy-diagnosed tuberculosis cases, four had meningitis and another four had perforative peritonitis. Fifteen mothers died due to Hepatitis E infection diagnosed ante-mortem. One case each of Pneumocystis carinii pneumonia and fungal pneumonia were diagnosed, the HIV status however was not known. The exact etiology could not be determined in Acute Respiratory Distress Syndrome (ARDS - two cases), cerebral infarction, sinus thrombosis, cardiomyopathy, myocarditis, cortical necrosis and hemolytic uremic syndrome.
Maternal mortality reduction has been the topmost priority for the international community. The Millennium Development Goals  and the WHO 'Make every mother and child count' initiative  describe the importance of maternal mortality reduction as a healthcare issue. A recent systematic review of the causes of maternal mortality and its geographic distribution has shown that the Indian subcontinent has a significantly higher maternal mortality attributable to sepsis, infection and hemorrhage.  Though maternal mortality has been the subject of a number of studies in India, ,,, very few detailed autopsy studies have been reported. A comprehensive summary of the magnitude and distribution of the causes of maternal deaths is critical to reform reproductive health policies. Hence we decided to conduct this large autopsy study to analyze the causes and pathological basis of maternal deaths at a large tertiary care center in Mumbai (Bombay). The MMR in our study over a nine-year period was 827/ 100000 live births. In our study, the contribution of indirect causes of death was higher than that of direct causes. Previous studies have not provided a detailed classification of the cause of death as determined on pathological studies.
The deaths classified as puerperal sepsis (16 of 134 direct causes) are lower than previous studies from India. ,,, Evidence of clinical infection (without positive cultures) was however present in cases included in other categories. The number of deaths attributed to pulmonary embolism and amniotic fluid embolism are higher as compared to other studies. ,,, We however believe that the deaths due to pulmonary embolism may be higher as many such cases may be missed due to inadequate number of lung sections taken at autopsy.
Autopsies revealed a high number of deaths attributed to infectious diseases (27 of 277) and cardiovascular diseases (27 of 277). Such high numbers have not been reported in literature from India.  The high number of deaths due to tuberculosis, heart disease and severe anemia reflect the appalling state of maternal health in our country.
We believe that the results of the present study can be better interpreted considering both the limitations and strengths. The retrospective study design is a limitation. In the present study we did not account for details like organisms isolated from cultures in puerperal sepsis. Similarly, though blood cultures were sent when indicated in the autopsy, the limited resources and lack of follow-up did not permit retrieval of all the data. All the maternal deaths were not subjected to an autopsy, which may have introduced a bias in the study findings. Hence, it is possible that exclusion of cases which were not autopsied may have skewed the data. However, the large data set analyzed and the detailed pathological analyses provide strength to the study conclusions. The present study provides detailed pathological analysis of maternal mortality in a tertiary care center over nine years. Though every attempt was made to determine the most likely diagnosis on autopsy, many cases presented with findings suggestive of multi-organ dysfunction. In such cases, the most evident gross and histology findings correlated with the clinical presentation was considered as the cause of death. While the predominance of hemorrhage and ecclampsia/pre-ecclampsia as causes of death is as estimated by other studies, [6-9] the very high number of deaths attributable to infectious diseases and cardiovascular diseases are causes of concern. The indirect causes of maternal deaths reflect on the state of maternal health and the present healthcare system of the country. Adequate prenatal testing for these indirect causes has the potential to lower maternal mortality to a significant extent. In conclusion, we believe that more number of autopsy studies with a prospective design can help elucidate the areas of weakness in maternal care, thus providing directions for community-based interventions.
[Table 1], [Table 2], [Table 3]