Six-year analysis of postmortem examination records at a teaching hospital--demographic profile and causes of deaths.
Analyzing the ultimate, underlying and contributing causes of death is mandatory under Section 10(3) and 30 of the Birth and Death Registration Act, 1969.  In a substantial number of cases, the cause of death cannot be easily ascertained by the attending medical practitioner. This requires a special inquest into the cause of death. Medicolegal postmortem examinations are performed under the Section 174 of the Criminal Procedure Code, 1973  for establishing the cause of death in cases where persons die in possible unnatural circumstances.
Autopsy (also known as postmortem examination or necropsy) is performed to--(1) determine the identity of the deceased, when not known; (2) ascertain the time since death and the cause of death; and (3) find out whether the cause of death was natural or unnatural and if unnatural, whether it was homicidal, suicidal, or accidental. 
A "medicolegal" postmortem examination is performed when ordered by a legal authority responsible for the investigation of sudden, suspicious, obscure, unnatural, litigious, or criminal deaths.  These medicolegal autopsies do not require the consent of relatives of the deceased. A "clinical" or "academic" autopsy is carried out to reveal the medical cause of death in cases of unknown or uncertain death with the primary objective of determining the cause of death, antemortem health status and appropriateness of antemortem diagnosis and treatment.  Autopsies are also performed to ascertain whether deaths were due to iatrogenic errors. When autopsies are carried out for academic reasons, consent of relatives of the deceased is required. Causes of death ascertained by postmortem examination are a subset of the total number of deaths registered in an area.
This study aims to describe the demographic profile and analyzes the causes of deaths certified after postmortem examinations, in order to facilitate improved and more reliable certification of the cause of death.
Materials and Methods
This study is a complete enumeration record-based analysis of data recorded from January 1, 2009 to December 31, 2014 (6 years) in the postmortem registers of the Forensic Medicine and Toxicology Department of a municipal medical college in Kalwa, Thane, located about 30 km from Mumbai city in the state of Maharashtra in Western India.
After obtaining permissions from the Institutional Ethics Committee (IEC) and institutional authorities for conducting the study, a total of 3137 postmortem examinations conducted during a 6-year period (January 1, 2009 to December 31, 2014) were analyzed for demographic details (age, gender, religion, place of residence), medicolegal categorization and cause of death.
The Kalwa creek is a natural geographical barrier that divides Thane Municipal Corporation area into Thane City and outlying areas of Kalwa, Mumbra, and Diwa (KMD). This natural division was also used for analyzing the geographical distribution in this study.
Diagnoses provided by postmortem examination were categorized.  The cause of death selected for categorization was the underlying cause of death agreed upon at the Sixth Decennial International Revision Conference. [R] The underlying cause of death was selected in accordance with recommended procedure. [R]
Categorical data were presented as frequencies and continuous data as mean and standard deviation (SD). Significance of difference in parameters was calculated by Karl Pearson's [chi square]-test at 95% confidence interval (p < 0.05). Epilnfo Version 7.0 (public domain software package from Centers for Disease Control and Prevention, Atlanta, GA, USA) was used for calculating the [chi square]-test (with Yates' correction where applicable) and odds ratio (OR).
The mean number of postmortem examinations conducted during the 6-year period of study (January 1, 2009 to December 31, 2014) was 523 per year with a SD of 67.87. The solitary postmortem case excluded from study was that of a 32-year-old Hindu, phenotypically a true hermaphrodite, resident of Kalwa, who had died due to pulmonary tuberculosis. This case was excluded from analysis due to lack of confirmation of gender by genotypic studies.
Of the 3137 postmortem examination records analyzed, 27.41% were females (mean age = 32.95 years; SD = 20.43 years) and 72.59% were males (mean age = 37.42 years; SD = 16.78 years). 80.3% cases were from the working-age population (15-60 years of age). Age, religion, and place of residence could not be determined in 350 (11.16%) cases whereas the religion and age could not be ascertained in 337 (10.74%) and 11 (0.35%) cases, respectively [Table 1].
Major causes of death were trauma (27.26%; Table 2 and Figure 1), asphyxia and hanging (15.3%), coronary artery disease (11.59%), tuberculosis (11.04%), alcoholic liver disease (6.41%), lower respiratory tract infections (6.64%; Table 2 and Figure 2), and drowning (6.18%). Other causes of death included septicaemia, peritonitis, malaria, jaundice, and pulmonary thromboembolism [Table 2]. Obstetric or gynecological causes accounted for 2.21% deaths in females [Table 2 and Figure 3].
The immediate cause of death was mentioned as "shock and haemorrhage" in 310 (9.88%) of the analysed autopsy records. The contributing cause of death was not mentioned in 2,967 (94.58%) cases. The contributing cause of death was alcoholic liver disease in 33 (0.9%) cases and trauma in 85 (2.74%) cases. All the five persons who died of firearm injuries were aged between 25-45 years. Opinion about the cause of death was reserved and the viscera preserved for 954 (30.41%) cases and 41% of infant deaths [Table 3]. There was significant (p = 0.00005; OR = 0.327) difference in proportion of deaths due to electrocution in various areas of Thane city [Table 4].
In this study, the annual number of postmortem examinations revealed an increasing trend from 2009 to 2014. Females were substantially fewer than males in the adult age categories [Table 4] but the gender difference was not significant (p = 0.707; OR = 1.73). Majority of the cases were Hindus followed by Muslims and then by other minority religions. The gender difference in the religion-based distribution was highly significant [Table 4]. The religion of 337 (10.74%) postmortem subjects could not be determined. The proportion of deaths due to trauma was significantly higher (p = 0.0007; OR = 1.47) in males (29.16%) as compared to that in females (21.83%). Deaths due to alcoholic liver disease were higher in males (7.86%) than in females (2.29%) with a highly significant (p = 0.000006; OR = 3.64) gender difference. Coronary heart disease was reported as the cause of death in 12.32% males and 9.5% females but the gender difference was not significant (p = 0.071; OR=1.33). Pulmonary tuberculosis was reported as the cause of death in 10.96% males and 11.26% females. Deaths due to drowning were significantly higher among males (p = 0.001; OR = 2.236) but the difference in geographical distribution was not significant (p = 0.137; OR = 1.30). Of the total deaths due to drowning, infants and children aged 1-5 years comprised 4.4% and 7.40%, respectively. The gender difference in deaths due to electrocution was not significant (p = 0.124; OR = 1.63). The cause of death could not be established in 41% infants. Of the infants subjected to autopsy, 45% belonged to KMD area. In all the areas, the gender difference among infant deaths was not significant. Lower respiratory tract infections (15%), still births (10%), drowning (6%), head injuries (6%) were leading causes of infant deaths. Non-viable fetus/abortus (7%) was a major cause of perinatal mortality. There were 19 maternal deaths.
Unnatural deaths and medicolegal autopsies also showed an increasing trend. In an Indonesian study, though the number of unnatural deaths showed a decreasing trend, the number of medicolegal autopsies increased over the study period (2007-2011).  Children below 5 years constituted 180 (5.74%) of the total postmortem examinations whereas they constitute 10.7% of the total Indian population.  This could be due to the fact that major causes of death in that age group would have been certified without requiring a postmortem examination. 151 (4.81%) of the postmortem examinations were performed on senior citizens (65 years and above) while they comprise 4.8% of the Indian population.  This could be due to the fact that the fraction of the population brought for postmortem examination and the deaths certified would be comparable. Working-age population (15-64 years) constitutes 59.5% of the Indian population  while those belonging to the same age group brought for postmortem examination in this study constituted 85.08% of the total cases. This could be due to the fact that trauma, asphyxia, and drowning were major causes of death [Table 21 which required postmortem examination to confirm diagnosis and refute any other cause. The relatively higher number of female subjects among infants and neonates suggests a higher infant and perinatal mortality, as compared to males. It is also possible that if dubious circumstances of death are suspected, medical practitioners may refrain from certifying the death of a female infant and ask for a postmortem examination. The numbers of male and female subjects in the 1-15 year age group were comparable [Table 1]. The proportion of deceased with undetermined religion was significantly higher among males probably because as compared to males, more females wear identifiable symbols of religion and religious circumcision is a norm among Muslims and Jews only. Majority of the cases are from KMD area due to the jurisdiction of local police station, proximity of the tertiary care teaching hospital to these places and availability of expertise for conducting autopsies.
Opinion about the cause of death was reserved and the viscera preserved for 954 (30.41%) cases. In a cohort analysis of 400 consecutive perinatal and infant postmortem examinations in Wales, UK, the cause of death was disclosed only in 18% cases.  Trauma, a leading cause of death, is often obscure on external inspection of the body and is revealed only by a postmortem examination. The number of deaths due to trauma increased gradually with age to peak between 15 and 30 years and then gradually decreased as age increased. Similar findings were reported in a study conducted in Manipal, South India.  As bread winners in the family, more males venture out of their houses as compared to females and consequently, males are more at risk for trauma. In this study, firearm injuries comprised 0.84% while a Bangkok-based study  reported firearm injuries in 2.09% of all autopsy cases. Alcoholic liver disease was also another leading cause of death confirmed by postmortem examination. Alcohol consumption has been incriminated in more than one-fifth of traumatic brain injuries  and 60% of all injuries  and is associated with intentional self-harm.  In both sexes, frequent and heavy drinking was associated with 8 or less years of formal education and income below the 75th percentile.  The youngest subject dying of acute myocardial infarction was a 26-year old female and there were three persons aged less than 30 years of age. The deaths due to coronary artery disease increased with age, peaking between 50 and 60 years. Harveit  has emphasized that the changing spectrum of diseases in aging populations should be recognized and that deaths due to coronary artery disease have remained unchanged. Ischaemic heart disease accounted for 63% of sudden adult deaths in an Irish hospital and this rate remained constant over one decade.  Acute myocardial infarction constituted 63% of all natural deaths in rural Australia in a 7-year study.  In this study, the gender difference in deaths due to pulmonary tuberculosis was not statistically significant (p = 0.840; OR = 0.969) and only one person (female) had died due to extra-pulmonary tuberculosis. In a Croatian hospital, clinically undiagnosed active tuberculosis was diagnosed on autopsy in 33 subjects (54%), of which 70% were male.  In a meta-analysis, about one-third of death certificates were found to be erroneous and about half of the autopsies revealed findings that were not clinically suspected when the subjects were alive.  The proportion of deaths due to electrocution in KMD area was significantly higher than that in other areas (p = 0.00005; OR = 0.327). This could be ascribed to inadequate domestic electrical safety in these areas. Postpartum hemorrhage (52.63%) and ruptured ectopic gestation (15.79%) were major causes of maternal deaths. Shock due to hemorrhage was the immediate cause of death in both these conditions. In a Ghana-based study, 12.1% of all female deaths were due to pregnancy-related causes with hemorrhage, abortion, and hypertensive disorders as major causes of death.  In this study, the contributing cause of death was not mentioned in 94.58% of postmortem reports. Ravakhah  has reported inconsistencies in the causes of death mentioned in the death certificates and those determined by autopsies and have stated that death certificate-based vital statistics ought to be rectified in the light of autopsy findings. A systematic review  reported that autopsies enable detection of diagnostic errors; however, deaths due to iatrogenic errors are seldom investigated and thus corrective measures are not taken to prevent repetition of errors.
The strength of this record-based complete-enumeration study is that it portrays the profile of subjects in postmortem examinations, which is a specific subset of the mortality picture of the general urban population at large. The limitation of this study is that it lacks analysis of incidental factors leading to the underlying causes of deaths, owing to its record-based design.
Though the identity of nearly one-tenth of the deceased could not be established, the postmortem subjects were predominantly males; most were aged between 15 and 60 years. Major causes of death were trauma, asphyxia and hanging, coronary artery disease, tuberculosis, alcoholic liver disease, lower respiratory tract infections, and drowning. Postpartum hemorrhage and ruptured ectopic gestation were major causes of maternal deaths. Infant deaths occurred due to completely avoidable drowning and electrocution. A high percentage of postmortem examinations needed laboratory confirmation for establishing the cause of death. The contributing cause of death was not mentioned in considerable number of postmortem reports.
[1.] Birth and Death Registration Act 1969. Available at: http://www. pbnrhm.org/docs/b&d_reg.pdf (last accessed on January 17, 2016).
[2.] Section 174 of the Code of Criminal Procedure 1973. Available at: http://www.hyderabadpolice.gov.in/acts/CrPC.pdf (last accessed on January 17, 2016).
[3.] Mathiharan K, Patnaik AK (Eds.). In: Modi's Medical Jurisprudence and Toxicology, 23rd edn. New Delhi: LexisNexis Butterworths, 2006. pp. 357-8.
[4.] World Health Organization. International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10)-WHO Version for 2016. Available at: http://apps.who. int/classifications/icd10/browse/2016/en#/XX (last accessed on January 17, 2016).
[5.] World Health Organization. International Statistical Classification of Diseases and Related Health Problems. 10th Revision (ICD-10), Volume-2, Instruction Manual. 2010: 31-40. Available at: http:// www.who.int/classifications/icd/ICD-10_2nd_ed_volume2.pdf (last accessed on January 17, 2016).
[6.] Afandi D. Profile of medicolegal autopsies in Pekanbaru, Indonesia 2007-2011. Malays J Pathol 2012;34(2):123-6.
[7.] Census of India 2011, Government of India, Ministry of Home Affairs data. Available at: http://censusindia.gov.in/Census_And_You/age_structure_and_marital_status.aspx (last accessed on January 17, 2016).
[8.] Cartlidge PH, Dawson AT, Stewart JH, Vujanic GM. Value and quality of perinatal and infant postmortem examinations: cohort analysis of 400 consecutive deaths. BMJ 1995;310 (6973):155-158.
[9.] Kanchan T, Menezes RG, Bakkannavar SM. Age and gender variations in trend of road traffic fatalities in Manipal, India. Med Sci Law 2010;50(4):192-6.
[10.] Myint S, Rerkamnuaychoke B, Peonim V, Riengrojpitak S, Worasuwannarak W. Fatal firearm injuries in autopsy cases at central Bangkok, Thailand: a 10-year retrospective study. J Forensic Leg Med 2014;28:5-10.
[11.] Gururaj G. Epidemiology of traumatic brain injuries: Indian scenario. Neurology Res 2002;24:24-8.
[12.] Benegal V, Gururaj G, Murthy P. Project report on a WHO multi-centre collaborative project on Establishing and Monitoring Alcohol's Involvement in Casualties: 2000-2001. Available at: http://www.nimhans.kar.nic.in/Deaddiction/lit/Alcohol%20and% 20%20Injuries_WHO%20Collab.pdf 2002. (last accessed on January 17, 2016).
[13.] Gururaj G, Isaac MK. Epidemiology of suicides in Bangalore. Bangalore: National Institute of Mental Health and Neuro Sciences. NIMHANS Publication No. 43. 2001, p. 34.
[14.] Benegal V, Nayak M, Murthy P, Chandra P, Gururaj G. Women and alcohol in India. Chapter 5. In: Alcohol, Gender and Drinking Problems: Perspectives from Low and Middle Income Countries, Obot IS, Room R (Eds.). Geneva: World Health Organization, 2005. pp. 89-124.
[15.] Hartveit F, Karwinski B, Giertsen JC. Changes in autopsy profile--1975 and 1984. J Pathol 1987;153(2):91-8.
[16.] Downes MR, Thorne J, Tengku Khalid TN, Hassan HA, Leader M. Profile of sudden death in an adult population (1999-2008). Ir Med J 2010;103(6):183-4.
[17.] Zardawi IM. Coronial autopsy in a rural setting. J Forensic Leg Med 2013;20(7):848-51.
[18.] Pavic I, Radulovic P, Bujas T, Balja MP, Ostojic J, Balicevic D. Frequency of tuberculosis at autopsies in a large hospital in Zagreb, Croatia: a 10-year retrospective study. Croat Med J 2012;53(1):48-52.
[19.] Roulson J, Benbow EW, Hasleton PS. Discrepancies between clinical and autopsy diagnosis and the value of post mortem histology; a meta-analysis and review. Histopathology 2005; 47(6):551-9.
[20.] Der E, Moyer C, Gyasi RK, Akosa AB, Tettey Y, Akakpo P, et al. Pregnancy related causes of deaths in Ghana: a five-year (2004-2008) retrospective study. Ghana Med J 2013;47(4):158-163.
[21.] Ravakhah K. Death certificates are not reliable: revivification of the autopsy. South Med J 2006;99(7):728-33.
[22.] Shojania KG, Burton EC, McDonald KM, Goldman L. Changes in rates of autopsy-detected diagnostic errors over time: a systematic review. JAMA 2003;289(21):2849-56.
Source of Support: Nil, Conflict of Interest: None declared.
Sundaram Kartikeyan (1), Aniruddha A Malgaonkar (1), Mangesh Ghadge (2)
(1) Department of Community Medicine, Rajiv Gandhi Medical College, Kalwa, Thane, Maharashtra, India.
(2) Department of Forensic Medicine and Toxicology, Rajiv Gandhi Medical College, Kalwa, Thane, Maharashtra, India.
Correspondence to: Aniruddha A Malgaonkar, E-mail: email@example.com
Received February 07, 2016. Accepted February 18, 2016
Table 1: Demographic profile (n = 3,137) Parameter Males Females Percent (n = 2277) (n = 860) (n = 3137) Age group 0-1 48 52 3.19 (years) 1-15 120 106 7.20 15-60 1,917 602 80.30 60+ 183 98 8.96 Unknown 9 2 0.35 Religion Hindu 1620 633 71.82 Muslim 361 166 16.80 Christian 8 6 0.45 Jain 2 0 0.06 Sikh 2 2 0.13 Unknown 284 53 10.74 Place of Thane City 223 78 9.60 residence KMD 1091 455 49.28 Neighboring 538 233 24.58 corporations Thane district 64 31 3.03 (rural + tribal) Rest of 27 5 1.02 Maharashtra Rest of India 37 5 1.34 Unknown 297 53 11.16 KMD = Kalwa, Mumbra, Diva area. Table 2: Major causes of death (n = 2183) Major cause of death Males Females Percent (n = 1615) (n = 568) (n = 2183) Asphyxia due to hanging 191 106 13.61 Other asphyxial deaths 15 22 1.69 Tuberculosis 177 64 11.04 Trauma 471 124 27.26 Coronary heart disease 199 54 11.59 Drowning 116 19 6.18 Pneumonia 71 33 4.76 LRTI other than pneumonia 21 20 1.88 Alcoholic liver disease 127 13 6.41 Electrocution 55 12 3.07 Intracranial hemorrhage 62 14 3.48 Burns 12 11 1.05 Other causes 98 76 7.97 LRTI, Lower respiratory tract infection. Table 3: Analysis of infant and foetal deaths (n = 100) Causes of death Males Females Total (n = 48) (n = 52) (n = 100) ARDS 1 1 2 Foetal distress 1 0 1 LRTI 5 10 15 Drowning 4 2 6 Electrocution 1 0 1 Head Injury 1 5 6 Decomposed newborn 1 0 1 Non-viable fetus/abortus 5 2 7 ORVP 22 19 41 Others 2 3 5 Still birth 4 6 10 Asphyxia--aspiration/choking 0 4 4 Iatrogenic 1 0 1 ARDS, Acute respiratory distress syndrome; LRTI, lower respiratory tract infection; ORVP, opinion reserved, viscera preserved (for chemical analysis). Table 4: Statistical significance of various parameters Parameters Number [chi square]-test value * Religion Male 1943 40.97 Female 807 Age Male 2268 0.121 Female 858 Pulmonary tuberculosis Male 177 0.041 Female 64 Alcoholic liver disease Male 127 21.76 Female 13 Trauma Male 471 11.39 Female 124 Coronary heart disease Male 199 3.249 Female 54 Drowning Male 116 10.666 Female 19 Other areas 77 1.23 KMD 58 Electrocution Male 55 2.361 Female 12 Other areas 18 16.52 KMD 49 Parameters p-Value Odds ratio Religion Male 0.00000014 # 0.38 Female Age Male 0.727 0.584 Female Pulmonary tuberculosis Male 0.840 0.969 Female Alcoholic liver disease Male 0.000006# 3.64 Female Trauma Male 0.0007# 1.47 Female Coronary heart disease Male 0.071 1.33 Female Drowning Male 0.001# 2.236 Female Other areas 0.266 1.24 KMD Electrocution Male 0.124 1.63 Female Other areas 0.00005# 0.327 KMD * [chi square]-test with Yate's correction where applicable; # p-value significant; KMD, Kalwa, Mumbra, Diva areas. Figure 1: Percentage distribution of asphyxia deaths [other than hanging] in males (n = 15) and females (n = 22). Males Females Traumatic asphyxia 53.33 40.91 Choking 13.33 13.64 Ligature strangulation 0.00 18.18 Aspiration 6.67 22.73 Smothering 6.67 4.54 Suffocation 20.00 0.00 Note: Table made from bar graph. Figure 2: Percentage distribution of deaths due to lower respiratory tract infection [other than pneumonia] in males (n = 21) and females (n = 20). Males Females Bronchopneumonia 38.10 40.00 Pleural effusion 33.34 35.00 Empyema 19.04 15.00 Lung abscess 9.52 10.00 Note: Table made from bar graph. Figure 3: Percentage distribution of causes of maternal deaths (n = 19). PPH 62.5 Rupture of ectopic pregnancy 18.75 Abruptio placentae 6.25 Puerperal infections 6.25 DIC + LSCS 6.25 PPH, Postpartum hemorrhage; DIC, disseminated intravascular coagulation; LSCS, lower segment cesarean section. Note: Table made from pie chart.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Research Article|
|Author:||Kartikeyan, Sundaram; Malgaonkar, Aniruddha A.; Ghadge, Mangesh|
|Publication:||International Journal of Medical Science and Public Health|
|Date:||Sep 1, 2016|
|Previous Article:||Awareness and attitude toward the use of emergency contraceptives among male and female adolescents of Jamnagar.|
|Next Article:||Comparative evaluation of effect of Olanzapine and Quetiapine on body weight and body mass index (BMI) in psychotic patients at a tertiary care...|