Forensic Sci Med Pathol (2008) 4:9–14 DOI 10.1007/s12024-007-0025-2
ORIGINAL PAPER
Diagnostic yield and characteristic features in a series of decomposed bodies subject to coronial autopsy Roger W. Byard Æ Emily Farrell Æ Ellie Simpson
Accepted: 10 March 2007 / Published online: 7 August 2007 Ó Humana Press Inc. 2007
Abstract A study was undertaken of 629 cases autopsied at Forensic Science SA, Adelaide, Australia over a 10-year period from 1994 to 2004 where significant decomposition had been documented. The age range of victims was 10 months to 92 years (mean 51.5 ± 18.1 years). There were 498 males (79%) and 131 females (21%) (M:F = 3.8:1). Deaths in 289 of the 629 cases (46%) were due to natural causes, with 179 suicides (28%), 83 accidents (13%), and 36 homicides (6%). The cause of death was ascertained in 89% of cases and the manner of death in 93% of cases. In 35 cases (6%) both the cause and manner could not be determined. Numbers of suicides were higher in younger age groups while deaths from organic illnesses predominated in later life. The number of cases in summer was significantly greater than in winter. Despite technical difficulties in handling decomposed bodies, and artifactual alteration of tissue structure and microscopic features, the autopsy was still a useful diagnostic exercise. While it is likely that more information may be gleaned from fresh bodies in perfectly preserved states, decomposed bodies may reveal significant anatomical and pathological features that enable both the cause and manner of death to be established.
Introduction
Keywords Forensic pathology Decomposition Putrefaction Autopsy Postmortem Cause of death Manner of death
Materials and methods
R. W. Byard (&) Discipline of Pathology, The University of Adelaide, Level 3 Medical School North Building, Frome Road, Adelaide, SA 5005, Australia e-mail:
[email protected] E. Farrell E. Simpson Forensic Science SA, Adelaide, SA, Australia
Once death has occurred, a body is subject to the inevitable consequences of autolysis and putrefaction, both of which may result in marked changes in the tissues [1] with loss of cellular detail on subsequent microscopy. Putrefaction is also often associated with postmortem insect depredation with further modification and alteration of tissues and organs [2]. The changes in advanced decomposition may result in complete loss of tissues, culminating in skeletonization [3]. Obviously once there has been significant loss of soft tissues the yield in the search for organic diseases such as ischemic heart disease, or soft-tissue injuries such a stab wounds, is at best greatly reduced. For this reason it is sometimes perceived that the autopsy examination is of limited value in cases where putrefactive changes have occurred. The following study was undertaken to evaluate the characteristic features of cases where decomposition had occurred, to review predisposing factors, and also to examine the diagnostic yield for cause and manner of death.
The files of Forensic Science SA in Adelaide, South Australia, Australia, were examined over a 10-year period from July 1st 1994 to June 30th 2004 for cases where ‘‘decomposed’’ had been recorded on admission files. Bodies were routinely assessed by experienced mortuary technicians upon admission to FSSA as to their state of preservation, as cases with significant decomposition are handled differently to fresh bodies. For example, decomposed cases may be placed in a second body bag, stored in a cooler refrigerator than is usual, and autopsied last on the
10
daily list. In recent years such cases have been autopsied in a special dissection room that has additional ventilation and that can be closed off from the main mortuary. Cases were recorded as decomposed if there was significant discoloration, smell, venous marbling, purging of putrefactive fluids, insect activity, and/or bloating. The degree of putrefaction was not graded. Autopsy and case records were reviewed for age, sex, date of admission, and cause and manner of death. All cases had been the subject of complete police and coronial investigations with full autopsy examinations. The cause of death was based on the examining pathologist’s assessment. The manner of death was assigned once police and autopsy information had been integrated. Comparison was made with the manner of death in non-decomposed bodies over the same time period. Statistical analyses were performed using a z-test for differences in proportion. Differences were considered significant at a minimum of P < 0.05. Detailed analyses of toxicological findings were considered beyond the scope of this study.
Results In the 10-year period from July 1st 1994 to June 30th 2004 there were 629 autopsies performed at Forensic Science SA on decomposed bodies, representing 6.3% of the total of 9,963 autopsies. The average number of autopsies with significant decomposition per year was 62.9 (range 45–81). The age range of victims was 10 months to 92 years (mean 51.5 ± 18.1 years) with 498 males (79%) and 131 females (21%) (M:F = 3.8:1) (P < 0.0001). Males were aged between 1 and 91 years (mean 50.8 ± 17.7 years), and females between 10 months and 92 years (mean 53.9 ± 20.0 years). The peak incidence was in the 40– 49 year age group for both males and females (101 males; 27 females; M:F = 3.7:1) (Fig. 1). Deaths in 289 of the 629 cases (46%) were due to natural causes (230 males: age range 18–91 years, mean 62 years; 59 females: age range 20–92 years, mean 63.4 years; M:F = 3.9:1), with 179 suicides (28%) (142 males: age range 17–86 years, mean 40.8 years; 37 females: age range 24–83 years, mean 44.5 years; M:F = 3.8:1), 83 accidents (13%) (65 males: age range 1–81 years, mean 38 years; 18 females: age range 16–88 years, mean 57.5 years; M:F = 3.6:1) and 36 homicides (6%) (25 males: age range 21–72 years, mean 40.9 years; 11 females: age range 10 months-74 years, mean 33.2 years; M:F = 2.3:1). The manner of death could not be established in 42 cases (7%) (36 males: age range 16–79 years, mean 48.5 years; 6 females: age range 33–60 years, mean 45.7 years; M:F = 6:1) (Fig. 2). This compared to the manner of death in 9,334 non-decomposed bodies of 50% natural causes,
Forensic Sci Med Pathol (2008) 4:9–14 140 Total NATURAL SUICIDE 120
100
80
60
40
20
0 0-9
10-19
20-29
30-39
40-49
50-59
60-69
70-79
80-89
90-99
Fig. 1 Numbers of suicides, natural deaths, and total deaths in 629 coronial autopsy cases with decomposition related to age
28% accidents, 19% suicides, 3% homicides, and 1% undetermined (Fig. 3). There were no significant differences in the proportion of males and females in each category of the decomposed cases. Suicides accounted for the largest category of deaths in the 20–59 year age group (N = 157 out of a total of 394; 40%), whereas natural disease was responsible for death in most of the victims aged between 60 and 99 years (N = 170 out of a total of 219; 78%) (Fig. 1). The majority of the 289 victims who died of organic diseases (174; 60%) had lethal cardiovascular conditions, in particular ischemic heart disease. Other cardiovascular causes of death included aortic aneurysms (dissected/ruptured), congestive cardiac failure, hypertensive heart disease, and left ventricular hypertrophy. Respiratory causes accounted for 21 cases (7%), cerebrovascular causes of ACCIDENT 13%
SUICIDE 28%
HOMICIDE 6%
UNDETERMINED 7%
NATURAL 46%
Fig. 2 Manner of death in 629 coronial autopsy cases with decomposition
Forensic Sci Med Pathol (2008) 4:9–14
11
60 % non-decomp % decomp 50
40
30
20
10
0 NATURAL
SUICIDE
ACCIDENT
HOMICIDE
UNDETERMINED
Fig. 3 Manner of death in 629 coronial autopsy cases with decomposition compared to 9,334 non-decomposed controls
death, including intracerebral hemorrhage, infarcts or ruptured berry aneurysms, accounted for 20 cases (7%); and other conditions such as epilepsy, gastrointestinal bleeding and malignancies accounted for 48 cases (17%). Although death was attributed to ‘natural disease’ in 26 cases (9%), based on the absence of injuries and negative toxicological screening, the exact causes of death were not determinable. Carbon monoxide toxicity accounted for 80 of the 179 cases of suicide (45%; 67 males, 13 females; M:F = 5.2:1). Intentional drug toxicity was the next most common category with 36 cases (20%), followed by hanging (29 cases, 16%) and gun shot wounds (13 cases, 7%). The remaining 20 cases (11%) included drowning and self-inflicted incised wounds. In one case (1%) the exact cause of death could not be determined, although the police investigation indicated that the death had been suicidal (Fig. 4). Accidental deaths were mainly due to drug and/or alcohol toxicity in 41 of 83 cases (50%). The next most
common categories consisted of drowning in 16 cases (19%), exposure in 11 (13%), and trauma (including vehicle accidents) in 10 (12%). The remainder included dehydration and hanging in 5%. There was one case (1%) in which the cause of death could not be determined, although again the police investigation indicated that the manner of death was accidental. The 36 cases of homicide included seven cases each (19%) of blunt trauma, gun shot wounds, stab wounds, and strangulation. Two cases (6%) were attributed to asphyxia and one (3%) to drowning with head injuries. In five cases (14%) the exact cause of death could not be determined; the bodies involved had been stored for some time in barrels. In the 42 cases (7%) where the manner of death could not be determined, the cause of death was identified in seven, all being due to drug(s) and/or alcohol toxicity. In the 68 cases (11%) in which the cause of death could not be determined, the manner of death was identified in 33. Twenty-six of these deaths were from presumed natural causes. There were five homicides, one accidental death, and one suicide. Thus, combining these two groups, there were 35 cases (6%) in which both the manner and cause of death were not able to be elucidated. Seasonal variation was found with 202 cases in summer (December–February), 158 in spring (September–November), 149 in autumn (March–May), and 120 in winter (June–August) (Fig. 5). There was a significant difference in the number of decomposed cases in summer compared to winter (P < 0.0001).
Discussion This study has shown that bodies with significant decompositional changes make up only a small percentage of coronial autopsies in South Australia (6.3%). Of these
90 F M
250
80 70
200
60 50
150 40 30
100
20
50
10 0 CO POISONING DRUG &/OR ALCOHOL
HANGING
OTHER
GSW
UNDETERMINED
0 SPRING
Fig. 4 Deaths from suicide in 629 coronial autopsy cases with decomposition demonstrating a predominance of cases of carbon monoxide toxicity
SUMMER
AUTUMN
WINTER
Fig. 5 Number of deaths per season in 629 coronial autopsy cases with decomposition demonstrating summer predominance
12
cases, the majority of deaths are due to natural causes (46%), with 28% being suicides, 13% being accidents, and 6% representing homicides. Males were much more likely to present in a decomposed state than females (3.8:1) (P < 0.0001), although the gender/age distribution was similar. This compares to the male-to-female ratio of all non-decomposed coronial cases over the same time (2.3:1). The male predominance may have been influenced by the greater number of males in the suicide category and possibly by a greater degree of social isolation and failure to seek medical attention. Male predominance has also been found in other studies of decomposition, with increased unemployment and divorce rates being cited as possible reasons for underlying social isolation [4]. These factors may account for the highest number of cases occurring in males aged between 40 and 49 years (Fig. 1). Decomposition results from the dual effects of tissue autolysis and putrefaction and is most likely to occur when there has been a prolonged interval between death and a body being found or presented to autopsy, or when a body is exposed to high environmental temperatures. Delay in discovery of a body may have been the intention of the perpetrator in a homicide, or may be the inadvertent effect of a victim choosing an isolated place to commit suicide where discovery is unlikely. This is reflected in the slightly increased percentages of homicides and suicides in the decomposed group compared to the non-decomposed controls, i.e., 6% versus 3% and 28% versus 19%, respectively. Elevated temperatures will obviously hasten decomposition and this may be due to external or internal heat. This may be augmented when there is associated rainfall [5]. Decomposition occurs more often and rapidly over hot summer months, as was shown in our study with 1.7 times the number of cases occurring in summer than in winter. It is also enhanced by factors that increase local temperatures inside a dwelling such as central heating, hot baths/showers, saunas, excess bed clothing, and electric blankets. Obese individuals retain heat and decomposition will be further accelerated if bodies are not rapidly transported to storage facilities and adequately refrigerated. Finally, certain drugs such as the ring-derivative amphetamines may be associated with endogenous hyperthermia [6]. These features are summarized in Table 1. Conversely, cool temperatures may cause surprising preservation of bodies, sometimes with the development of adipocere or so-called grave wax [7, 8]. Decomposition rates of buried bodies may be influenced by the depth of burial [9] and preservation of tissues may also occur with mummification if a corpse dessicates, or tanning if a body has been immersed in water in a peat bog, as occurred with Graubelle man and Elling woman in Jutland [10]. In Casper’s experience the degree of putrefaction of a body lying in the open after one week (month) was the same as that of
Forensic Sci Med Pathol (2008) 4:9–14 Table 1 Reasons for decomposition (1) Prolonged postmortem interval (i) Concealment of a body by a perpetrator (homicide) (ii) Concealment by a victim (suicide) (iii) Failure to find a body (outdoors) (a) isolated/rural/aquatic environments (iv) Failure to find a body (indoors) (a) social isolation (v) Mass disasters (a) delay in finding/transporting victims (b) lack of resources/staff to process cases (2) High environmental temperature (i) Outdoors (a) high ambient temperatures (b) high vehicular temperatures—sunlight/exhaust fumes (c) fire (ii) Indoors (a) central heating (b) excess bedding/electric blanket (c) warm/hot shower/bath/sauna (d) fire (3) Exacerbating factors (i) Failure to refrigerate/store/transport a body adequately (ii) Obesity (iii) Drugs
a body immersed in water for two weeks (months), or after burial for eight weeks (months) [11]. It is not surprising that the majority of victims in this study succumbed from cardiovascular disease, as this is the most common cause of natural adult death in Australia. The nature of fatal natural diseases other than sepsis or hyperthermia should, however, have minimal relationship to the presence or absence of decomposition, as putrefactive changes relate more to the factors summarized in Table 1 than to underlying disease states. Conversely, suicide methods may have a distinct relationship to decomposition, for example with inhalation of exhaust fumes. Many of the diagnoses of natural diseases such as cardiovascular conditions were based on macroscopic observations that included significant stenoses of coronary arteries, myocardial scarring and hypertrophy and aneurysm rupture, rather than on histology, as microscopic tissue preservation was usually suboptimal. In cases such as epilepsy the medical history provided guidance as to the possible cause of death and the autopsy provided useful negative information with no evidence of other natural or unnatural conditions. Very occasionally histologic examinations revealed areas of myocardial fibrosis or aggregated ghost neutrophils, indicating an inflammatory infiltrate in
Forensic Sci Med Pathol (2008) 4:9–14
an acute myocardial infarct or pneumonia. Toxicologic evaluations were sometimes difficult to interpret with advanced putrefaction and meaningful microbiologic studies were not possible. Although immunohistochemistry has been utilized in these circumstances [12], it is not routine and was not used in any of the cases in this series. The percentage of suicides due to intentional carbon monoxide inhalation in our series was 45%, compared to 16% in our overall suicide data, and 20.2% listed under ‘gas inhalation’ in aggregated data for suicides in Australia from 1986 to 1992 [13]. The markedly increased rate of decomposition in suicide victims who utilize carbon monoxide inhalation is a reflection of several factors, as those who elect to commit suicide by inhaling exhaust fumes often choose an isolated area to undertake this. This may result in the body being exposed for some time to the inside of a vehicle heated by both exhaust fumes and the sun, with delayed discovery. Suicidal deaths were more common in younger ages, while deaths from natural diseases were found more often in older age groups (Fig. 1). Drowning victims are another group who may not be found immediately despite intensive searches and so may be exposed to warm river/sea water for some time. This is a diagnosis influenced by the circumstances of the death and the absence of significant findings at autopsy. The higher percentage of accidents in the non-decomposed group was due to the number of victims of motorvehicle crashes who would have been found very soon after the event. Decomposition may also be a marker of social isolation, particularly in the elderly and may be a feature of Diogenes syndrome [4, 14]. It may also occur in children who have died at home from dehydration once their socially isolated carer has died from an underlying medical condition [15]. The diagnosis of dehydration in these circumstances is usually made on circumstantial rather than pathological or biochemical evidence. It is also possible that psychiatric illness plays a role in social isolation and therefore predisposes to decomposition if lethal illness occurs. However, as background psychiatric histories were not consistently available in our cases, no comment on this could be made using the current data. Autopsies on decomposed bodies present a number of problems. Bodies are often difficult to handle due to skin slippage and subcutaneous fluid accumulation, and insect activity may also be considerable, making dissection difficult and altering physical findings. Putrefaction itself may significantly alter physical features thus precluding visual identification and may cause postmortem artifactual disruption of recent surgical wounds and herniation of organs [16, 17]. Features such as tattoos and scars may however still be present, which can assist with identification [18]. Purging of putrefactive fluids from the nose and mouth
13
may be confused with bleeding from injury, and skin discoloration may be mistaken for bruising, thus raising suspicions of assault/foul play. Specific occupational health and safety issues arise involving the smell associated with decomposed bodies, which may be quite strong and permeate outside the mortuary. Putrefactive fluids may cover dissection surfaces and make floors dangerously slippery. Although loss of tissue structure due to putrefaction may make postmortem assessment difficult, the current study has shown that a cause of death could be determined in 561 out of 629 cases (89%) and the manner of death in 587 cases (93%). A weakness of the study, due to the retrospective nature of the investigation, was the inability to grade the degree of putrefaction, as it is likely that cases with the most putrefaction would have a lower diagnostic yield than those with less profound changes. This is reflected in the lower percentage of undetermined manner of death in the non-decomposed controls (1%) compared to the decomposed cases (7%). Despite these issues, however, decomposed bodies may still reveal significant anatomical and pathological features that enable both the cause and manner of death to be established in the majority of cases. Putrefaction is not, therefore, a reason to argue against the performance of a full autopsy.
Educational message 1. 2.
3. 4.
Significant decomposition occurs in only a minority of coronial autopsies. There is a marked predominance in males, with a peak incidence in the 40–49 year age group, and in the summer months. Decomposition is a marker for social isolation. Despite loss of organ and tissue architecture autopsies in decomposed individuals are able to establish the cause and manner of death in the majority of cases.
References 1. Tsokos M. Postmortem changes. Overview. In: Payne-James J, Byard RW, Corey T, Henderson C, editors. Encyclopedia of forensic and legal medicine, vol. 3. Amsterdam: Elsevier/Academic Press; 2005. p. 456–76. 2. Byard RW, James RA, Gilbert JD. Diagnostic problems associated with cadaveric trauma from animal activity. Am J Forensic Med Pathol 2002;23:238–44. 3. Prieto JL, Magana C, Ubelaker DH. Interpretation of postmortem change in cadavers in Spain. J Forensic Sci 2004;49:918–23. 4. Honigschnabl S, Schaden E, Stichenwirth M, Schneider B, Klupp N, Kremeier E, et al. Discovery of decomposed and mummified
14
5. 6.
7. 8.
9.
10.
11.
Forensic Sci Med Pathol (2008) 4:9–14 corpses in the domestic setting—a marker of social isolation? J Forensic Sci 2002;47:837–42. Archer MS. Rainfall and temperature effects on the decomposition rate of exposed neonatal remains. Sci Justice 2004;44:35–41. Byard RW, Gilbert J, James R, Lokan RJ. Amphetamine derivative fatalities in South Australia—is ‘‘Ecstacy’’ the culprit? Am J Forensic Med Pathol 1998;19:261–5. Dressler J, Madea B. Dumping of victims in walls. Am J Forensic Med Pathol 2005;26:250–3. Fiedler S, Graw M. Decomposition of buried corpses, with special reference to the formation of adipocere. Naturwissenschaft 2003;90:291–300. Rodriguez WC, Bass WM. Decomposition of buried bodies and methods that may aid their location. J Forensic Sci 1985;30: 836–52. Stodkilde-Jorgensen H, Jurik AG, Dalstra M, Lynnerup N, Gregersen M, Boel LW, Bindslev DA, Pedersen CB, Kruse A, Hansen ES, Jorgensen CS, Jacobsen NO, Asingh P. The Graubelle man. Medical examination of a conserved moss man’s corpse. Ugeskr Laeger 2001;163:7226–9. Casper JL. A handbook of the practice of forensic medicine based upon personal experience. London: The New Sydenham Society; 1861. Ch. 2, p. 14–55.
12. Ortmann C, Pfeiffer H, Brinkmann B. Demonstration of myocardial necrosis in the presence of advanced putrefaction. Int J Legal Med 2000;114:50–5. 13. Madden R. Suicides, Australia 1982–1992. Australian Bureau of Statistics, Catalogue No. 3309.0. Canberra: Australian Government Publishing Service; 1994. 14. Byard RW, Gilbert JD, Tsokos M. Forensic issues in cases of Diogenes syndrome. Am J Forensic Med Pathol 2007;28:177–81. 15. Byard RW. Incapacitation or death of a socially isolated parent or carer may result in the death of dependent children. J Paediatr Child Health 2002;23:417–8. 16. Byard RW, Gehl A, Andeers S, Tsokos M. Putrefaction and wound dehiscence—a potentially confusing post mortem phenomenon. Am J Forensic Med Pathol 2006;27:61–3. 17. James R, Gilbert J, Byard RW. Post-traumatic diaphragmatic hernia and death—etiological factor or putrefactive artefact? Am J Forensic Med Pathol 1999;20:66–9. 18. Cains G, Byard RW. Tattoos—historical, social and pathologic significance. In: Tsokos M, editor. Forensic pathology reviews, vol. 5. Humana Press (in press).
