Editorial
Autopsy and critical care Adrian Wong1, Michael Osborn2 and Carl Waldmann3
Journal of the Intensive Care Society 2015, Vol. 16(4) 278–281 ! The Intensive Care Society 2015 Reprints and permissions: sagepub.co.uk/ journalsPermissions.nav DOI: 10.1177/1751143715589602 jics.sagepub.com
Abstract An autopsy is a medical procedure consisting of the thorough examination of the body and internal organs after death, to evaluate disease or injury and to determine the cause and manner of a person’s death. In the intensive care setting, autopsies are usually performed to determine the cause of death or further medical knowledge. Early evidence that showed an alarmingly high rate of medical misdiagnosis found at autopsy is being called into question; the role of the procedure itself is being scrutinised. Furthermore, there has been a marked decline in the number of autopsies being performed both in the UK and across Europe. We examine the role of autopsies in modern health care for critically ill patients.
Keywords Coroner, autopsy, postmortem, intensive care
Autopsy and critical care An autopsy is a medical procedure consisting of the thorough examination of the body and internal organs after death, to evaluate disease or injury and to determine the cause and manner of a person’s death. Autopsies were, and remain a source of scientific knowledge helping to improve the understanding of pathology, and through this the treatment of disease and ultimately patient care. However, in recent decades, there has been a marked decline in the number of autopsies being performed both in the UK and across Europe.
Getting death right and improving practice Major discrepancy rates between premortem clinical diagnoses and postmortem autopsy findings continue to be reported in critically ill patients admitted to the intensive care unit (ICU). The landmark study by Goldman et al.1 found a 10% error rate in diagnosis in 100 randomly chosen autopsies. Their original classification of diagnostic errors was modified by Frohlich et al.2 in 2014 (Table 1). More recent analysis quotes a discrepancy rate of between 20% and 44%3 despite improvements in modern imaging and diagnostic processes. Furthermore, most of these discrepancies fall into Goldman class I and II. These are major unexpected findings that may have changed management, treatment, and outcome had they been made in life.
Although there are no studies that relate misdiagnosis at postmortem and actual mortality, the likelihood is that there are significant morbidity and cost implications associated with misdiagnosis. Autopsy can play a crucial role in ensuring standards of practice and forms the basis of good clinical governance. It allows assessment of the appropriateness of diagnosis, treatment, resource utilisation and the ultimate clinical outcome. Autopsy facilitates identification of weaknesses at each stage of this chain so that remedial action can be taken. Through verification of the accuracy of antemortem diagnosis and death certificates, autopsy allows assessment of the suitability of medical treatment and investigation into perioperative deaths. In the broader public health context, medical death certificates can be issued without an autopsy in the majority of cases. Error rates in death certificates in the UK are estimated at around 30%.4 This can lead to significant errors in epidemiological studies 1
Adult Intensive Care Unit, John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Oxford, UK 2 Department of Cellular Pathology, Imperial College Healthcare NHS Trust (St Mary’s Campus), St Mary’s Hospital, London, UK 3 Intensive Care Department, Royal Berkshire Hospital, Reading, UK Corresponding author: Adrian Wong, Adult Intensive Care Unit, John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Headley Way, Oxford OX3 9DU, UK. Email: [email protected]
Wong et al.
279
Table 1. Classification of autopsy diagnostic discrepancies. Class I
Class II
Class III
Class IV
Class V
Undiagnosed major unexpected findings before death that probably would have changed therapy, improved survival or significantly altered treatment of a terminally ill patient Undiagnosed major unexpected findings before death that would have not have changed the treatment or the patient’s survival Undiagnosed findings that were not immediate or primary cause of death but could, if known, have altered treatment decisions or prognosis Other unexpected clinically relevant findings unrelated to cause of death or findings that are of possible epidemiological or genetic interest Complete agreement between premortem clinical diagnosis and autopsy findings
Source: Adapted from Frohlich et al.2
resource allocation, thereby affecting public health policies and limiting the ability to monitor outbreaks and detect environmental–occupational hazards. The 2006 National Confidential Enquiry into Patient Outcome and Death (NCEPOD)4 report commented on the importance of autopsy in investigating perioperative death, recommending greater consultant involvement and increased autopsy rates.
The decline There are two types of autopsies – coroner’s and hospital’s.5 Coroner’s autopsies are performed when the cause of death is unknown and help decide if an inquest is needed. Hospital autopsies are requested to provide more information about an illness or the cause of death, or to further medical research. Since the 1970s, there has been a sharp decline in the number of autopsies performed. In the UK, almost all autopsies are now performed on the instruction of a medicolegal authority i.e. coroner rather than at the request of relatives or clinicians. The law in England and Wales provides that unless an autopsy has been ordered by a coroner, it cannot be performed for any other reasons (including research) without family consent. Almost 230,000 deaths were reported to the coroner in 2013, with autopsies performed in 41% of cases.6 While there is significant variation between coroners’ jurisdictions, the continual downward trend in numbers is fairly consistent across the country. The decline in autopsy numbers has important implications in research as well as medical education. Other than establishing the cause of death, neglected aspects of autopsy include the monitoring of the quality of care provided, identification of new diseases or
new manifestations of already known diseases and evaluation of therapy effectiveness. By far the most common reason for autopsy in the UK is an unexpected death where the cause is not apparent (i.e. coroner’s autopsy), and this has been declining in number. Hospital autopsy numbers have also fallen, both in the UK and across Europe. A recent study quantified the UK-wide hospital autopsy rate for each of the constituent countries as: 0.69% UK, 0.51% England, 0.46% Northern Ireland, 0.65% Wales and 2.13% Scotland. Factors that have contributed to the decline in the number of autopsies include . Distaste of procedure by both physicians and relatives . Lack of financial incentives . Increased faith in imaging technology . Fear of litigation . Lack of emphasis on the importance of autopsy in the teaching curriculum . Poor communication skills on the part of requesting physicians. Organ retention scandals at Alder Hey Hospital and the Bristol Royal Infirmary undoubtedly play on the minds of the public and are also likely to have contributed to this decline.
Autopsies and ICU In the intensive care setting, the majority of autopsies in the UK are under coronial jurisdiction. The ICU is a highly complex and inherently dangerous environment; diagnostic and interventional processes are complex and hence prone to error. Patients on the ICU are extensively investigated during their admission, undergoing numerous invasive procedures, both diagnostic and therapeutic, such that at the point of death, the initial diagnosis may have become irrelevant. Most critically ill patients have complex interconnected pathophysiology. Death due to ‘multiple organ failure’ is vague at best but probably the most accurate diagnosis that can be provided at our current level of comprehension of severe illness. Prolonged ICU admissions and complications alter patients’ physiology and pathology; it should be acknowledged that autopsy may detect and diagnosis the sequelae of ICU treatment rather than the admitting diagnosis per se. Hence the common complaint of clinicians that autopsy results bear no resemblance to a patient’s clinical illness before death. Even when the cause of death is unclear and consent has been granted, an autopsy may not be able to provide a conclusive cause of death for a variety of reasons. These include evolution of the initial disease, opportunistic infections and the impact of medical treatment and procedures. The longer the patient has been on the ICU, the
280 less likely it is that a conclusive diagnosis will be reached. Early studies on the ICU population quote an alarmingly high diagnostic error rate. One systematic review showed that 28% of autopsied ICU patients had at least one misdiagnosis.7 The authors concluded that up to 40,500 adult patients in American ICUs may die because of a misdiagnosis annually. The literature shows an extremely variable rate of between 2.3% and 26.8% in ICU-based studies. However, more recent studies have questioned this high rate. Indeed, Frohlich et al.2 looked at autopsies performed between 2006 and 2011 and found a class I and II error rate of 2.4% and 5.4%, respectively. Class I errors are major findings at autopsy that if known in life would have significantly altered management. Class II errors are major findings that would not have changed medical management or patient outcome. Very early studies looking at diagnostic errors on ICU found that clinicians were likely to miss the diagnoses of infections and vascular diseases e.g. pulmonary embolism.8 Critical illness subjects ICU patients to a higher risk of thromboembolic disease. Improvements in diagnostic and treatment pathways along with public attention on sepsis and thromboembolic prevention are potential reasons for overall decline in their incidence. Improved communication and working practice between intensivists and pathologists would contribute to greater understanding of the pathological processes and causes of death in these complex patients and by extrapolation, better patient care. Both specialties need to work together – intensivists should be encouraged to attend postmortem examinations as part of the clinical governance role. Equally, pathologists can be asked to present their findings to the ICU team such as at MultiDisciplinary Team meetings (MDT).11
Future autopsies In light of falling autopsy rates across Europe, alternatives to conventional postmortem examination are being explored. Such techniques tend to involve postmortem imaging of some form although a major drawback of this noninvasive autopsy approach is the lack of tissue for histopathological and microbiological examination. The Department of Health UK-funded trial sought to identify the accuracy of postmortem CT and MRI compared with full autopsy in a large series of adult deaths.9 They concluded that CT was more accurate than MRI for providing a cause of death compared with traditional autopsy. However, this technique had its limitations when the cause of death was due to ischaemic heart disease, pulmonary embolism, pneumonia and intra-abdominal lesions. However, a combination of postmortem imaging with other special techniques including, if necessary,
Journal of the Intensive Care Society 16(4) targeted biopsy can create a ‘minimally invasive autopsy’ (MIA), a useful compromise that can provide an accurate cause of death in approximately 70% of cases.10 MIA has the advantage of maintaining the integrity of the body and is therefore often more acceptable to families of the deceased. In addition, MIA retains an accurate record of the postmortem findings, in the form of the scans, for further assessment and review at a later stage should that be necessary, something which standard autopsy can never do.
Conclusion Economic drivers and technological change will continue to reshape the pathology landscape. The number of consented autopsies being conducted in UK hospitals is diminishing year on year leading to a continuous downward spiral in autopsy numbers. Not only does this have implications for those whose research depends on the availability of autopsy data, it also reduces our awareness and understanding of diagnostic errors, which autopsy can highlight. NCEPOD recommends increased communication between clinicians and pathologists. This, together with other simple measures such as formally training doctors and nurses to obtain consent for autopsies should be considered as these have been shown to increase autopsy numbers. Is there still a role for autopsy on ICU patients in modern health care? Undoubtedly there is, but like most things in health care, one must consider the benefits and costs. On the one hand, modern studies have demonstrated that misdiagnosis rates are lower than previously thought. With the decline in autopsy numbers, there is now a reduction in the number of hospitals and indeed the number of pathologists who have the experience and expertise to perform an autopsy on ICU patients. The value of the autopsy as a teaching and audit tool for ITU however seems unquestionable, even today and therefore the possibility of referring patients to specialist hospitals for autopsy should be considered if it cannot be conducted locally. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding The authors received no financial support for the research, authorship and/or publication of this article.
Acknowledgement This work came as a result from the European Critical Care Foundation organised study day – ‘The decline of autopsy and its implication for critically ill patients’ held on the 6th March 2015.
Wong et al.
281
References 1. Goldman L, Sayson R, Robbins S, et al. The value of the autopsy in three medical eras. NEJM 1983; 308: 1000–1005. 2. Frohlich S, Ryan O, Murphy N, et al. Are autopsy findings still relevant to the management of critically ill patients in the modern era? Crit Care Med 2014; 42: 336–343. 3. Shojania KG, Burton EC, McDonald K, et al. Changes in rates of autopsy detected diagnostic errors over time: a systematic review. JAMA 2003; 289: 2849–2856. 4. The Coroner’s Autopsy. Do we deserve better? http:// www.ncepod.org.uk/2006Report/Downloads/ Coronial%20Autopsy%20Report%202006.pdf (accessed 15 April 2015). 5. http://www.nhs.uk/conditions/Post-mortem/Pages/ Introduction.aspx (accessed 15 April 2015). 6. Office for National Statistics. (2003). Mortality statistics series DH2 no 29. Review of the registrar general on
7.
8.
9.
10.
11.
deaths by cause, sex and age in England and Wales, 2002. London: HMSO, 2003. Winters B, Custer J, Galvagno SM, et al. Diagnostic errors in intensive care units: a systematic review of autopsy studies. BMJ Qual Saf 2012; 21: 894–902. Cullen DJ, and Nemaskal AR. The autopsy incidence of acute pulmonary embolism in critically ill surgical patients. Int Care Med 1986; 12: 399–403. Roberts IS, Benamore RE, Benbow EW, et al. Postmortem imaging as an alternative to autopsy in the diagnosis of adult deaths: a validation study. Lancet 2012; 379: 136–142. Wichmann D, Obbelode F, Vogel H, et al. Virtual autopsy as an alternative to traditional medical autopsy in the intensive care unit: a prospective cohort study. Ann Int Med 2012; 156: 123–130. Dimopoulos G, Piagnerelli M, Berre J, et al. Post mortem examination in the intensive care unit: still useful? Int Care Med 2004; 30: 2080–2085.
Autopsy and critical care Adrian Wong1, Michael Osborn2 and Carl Waldmann3
Journal of the Intensive Care Society 2015, Vol. 16(4) 278–281 ! The Intensive Care Society 2015 Reprints and permissions: sagepub.co.uk/ journalsPermissions.nav DOI: 10.1177/1751143715589602 jics.sagepub.com
Abstract An autopsy is a medical procedure consisting of the thorough examination of the body and internal organs after death, to evaluate disease or injury and to determine the cause and manner of a person’s death. In the intensive care setting, autopsies are usually performed to determine the cause of death or further medical knowledge. Early evidence that showed an alarmingly high rate of medical misdiagnosis found at autopsy is being called into question; the role of the procedure itself is being scrutinised. Furthermore, there has been a marked decline in the number of autopsies being performed both in the UK and across Europe. We examine the role of autopsies in modern health care for critically ill patients.
Keywords Coroner, autopsy, postmortem, intensive care
Autopsy and critical care An autopsy is a medical procedure consisting of the thorough examination of the body and internal organs after death, to evaluate disease or injury and to determine the cause and manner of a person’s death. Autopsies were, and remain a source of scientific knowledge helping to improve the understanding of pathology, and through this the treatment of disease and ultimately patient care. However, in recent decades, there has been a marked decline in the number of autopsies being performed both in the UK and across Europe.
Getting death right and improving practice Major discrepancy rates between premortem clinical diagnoses and postmortem autopsy findings continue to be reported in critically ill patients admitted to the intensive care unit (ICU). The landmark study by Goldman et al.1 found a 10% error rate in diagnosis in 100 randomly chosen autopsies. Their original classification of diagnostic errors was modified by Frohlich et al.2 in 2014 (Table 1). More recent analysis quotes a discrepancy rate of between 20% and 44%3 despite improvements in modern imaging and diagnostic processes. Furthermore, most of these discrepancies fall into Goldman class I and II. These are major unexpected findings that may have changed management, treatment, and outcome had they been made in life.
Although there are no studies that relate misdiagnosis at postmortem and actual mortality, the likelihood is that there are significant morbidity and cost implications associated with misdiagnosis. Autopsy can play a crucial role in ensuring standards of practice and forms the basis of good clinical governance. It allows assessment of the appropriateness of diagnosis, treatment, resource utilisation and the ultimate clinical outcome. Autopsy facilitates identification of weaknesses at each stage of this chain so that remedial action can be taken. Through verification of the accuracy of antemortem diagnosis and death certificates, autopsy allows assessment of the suitability of medical treatment and investigation into perioperative deaths. In the broader public health context, medical death certificates can be issued without an autopsy in the majority of cases. Error rates in death certificates in the UK are estimated at around 30%.4 This can lead to significant errors in epidemiological studies 1
Adult Intensive Care Unit, John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Oxford, UK 2 Department of Cellular Pathology, Imperial College Healthcare NHS Trust (St Mary’s Campus), St Mary’s Hospital, London, UK 3 Intensive Care Department, Royal Berkshire Hospital, Reading, UK Corresponding author: Adrian Wong, Adult Intensive Care Unit, John Radcliffe Hospital, Oxford University Hospitals NHS Trust, Headley Way, Oxford OX3 9DU, UK. Email: [email protected]
Wong et al.
279
Table 1. Classification of autopsy diagnostic discrepancies. Class I
Class II
Class III
Class IV
Class V
Undiagnosed major unexpected findings before death that probably would have changed therapy, improved survival or significantly altered treatment of a terminally ill patient Undiagnosed major unexpected findings before death that would have not have changed the treatment or the patient’s survival Undiagnosed findings that were not immediate or primary cause of death but could, if known, have altered treatment decisions or prognosis Other unexpected clinically relevant findings unrelated to cause of death or findings that are of possible epidemiological or genetic interest Complete agreement between premortem clinical diagnosis and autopsy findings
Source: Adapted from Frohlich et al.2
resource allocation, thereby affecting public health policies and limiting the ability to monitor outbreaks and detect environmental–occupational hazards. The 2006 National Confidential Enquiry into Patient Outcome and Death (NCEPOD)4 report commented on the importance of autopsy in investigating perioperative death, recommending greater consultant involvement and increased autopsy rates.
The decline There are two types of autopsies – coroner’s and hospital’s.5 Coroner’s autopsies are performed when the cause of death is unknown and help decide if an inquest is needed. Hospital autopsies are requested to provide more information about an illness or the cause of death, or to further medical research. Since the 1970s, there has been a sharp decline in the number of autopsies performed. In the UK, almost all autopsies are now performed on the instruction of a medicolegal authority i.e. coroner rather than at the request of relatives or clinicians. The law in England and Wales provides that unless an autopsy has been ordered by a coroner, it cannot be performed for any other reasons (including research) without family consent. Almost 230,000 deaths were reported to the coroner in 2013, with autopsies performed in 41% of cases.6 While there is significant variation between coroners’ jurisdictions, the continual downward trend in numbers is fairly consistent across the country. The decline in autopsy numbers has important implications in research as well as medical education. Other than establishing the cause of death, neglected aspects of autopsy include the monitoring of the quality of care provided, identification of new diseases or
new manifestations of already known diseases and evaluation of therapy effectiveness. By far the most common reason for autopsy in the UK is an unexpected death where the cause is not apparent (i.e. coroner’s autopsy), and this has been declining in number. Hospital autopsy numbers have also fallen, both in the UK and across Europe. A recent study quantified the UK-wide hospital autopsy rate for each of the constituent countries as: 0.69% UK, 0.51% England, 0.46% Northern Ireland, 0.65% Wales and 2.13% Scotland. Factors that have contributed to the decline in the number of autopsies include . Distaste of procedure by both physicians and relatives . Lack of financial incentives . Increased faith in imaging technology . Fear of litigation . Lack of emphasis on the importance of autopsy in the teaching curriculum . Poor communication skills on the part of requesting physicians. Organ retention scandals at Alder Hey Hospital and the Bristol Royal Infirmary undoubtedly play on the minds of the public and are also likely to have contributed to this decline.
Autopsies and ICU In the intensive care setting, the majority of autopsies in the UK are under coronial jurisdiction. The ICU is a highly complex and inherently dangerous environment; diagnostic and interventional processes are complex and hence prone to error. Patients on the ICU are extensively investigated during their admission, undergoing numerous invasive procedures, both diagnostic and therapeutic, such that at the point of death, the initial diagnosis may have become irrelevant. Most critically ill patients have complex interconnected pathophysiology. Death due to ‘multiple organ failure’ is vague at best but probably the most accurate diagnosis that can be provided at our current level of comprehension of severe illness. Prolonged ICU admissions and complications alter patients’ physiology and pathology; it should be acknowledged that autopsy may detect and diagnosis the sequelae of ICU treatment rather than the admitting diagnosis per se. Hence the common complaint of clinicians that autopsy results bear no resemblance to a patient’s clinical illness before death. Even when the cause of death is unclear and consent has been granted, an autopsy may not be able to provide a conclusive cause of death for a variety of reasons. These include evolution of the initial disease, opportunistic infections and the impact of medical treatment and procedures. The longer the patient has been on the ICU, the
280 less likely it is that a conclusive diagnosis will be reached. Early studies on the ICU population quote an alarmingly high diagnostic error rate. One systematic review showed that 28% of autopsied ICU patients had at least one misdiagnosis.7 The authors concluded that up to 40,500 adult patients in American ICUs may die because of a misdiagnosis annually. The literature shows an extremely variable rate of between 2.3% and 26.8% in ICU-based studies. However, more recent studies have questioned this high rate. Indeed, Frohlich et al.2 looked at autopsies performed between 2006 and 2011 and found a class I and II error rate of 2.4% and 5.4%, respectively. Class I errors are major findings at autopsy that if known in life would have significantly altered management. Class II errors are major findings that would not have changed medical management or patient outcome. Very early studies looking at diagnostic errors on ICU found that clinicians were likely to miss the diagnoses of infections and vascular diseases e.g. pulmonary embolism.8 Critical illness subjects ICU patients to a higher risk of thromboembolic disease. Improvements in diagnostic and treatment pathways along with public attention on sepsis and thromboembolic prevention are potential reasons for overall decline in their incidence. Improved communication and working practice between intensivists and pathologists would contribute to greater understanding of the pathological processes and causes of death in these complex patients and by extrapolation, better patient care. Both specialties need to work together – intensivists should be encouraged to attend postmortem examinations as part of the clinical governance role. Equally, pathologists can be asked to present their findings to the ICU team such as at MultiDisciplinary Team meetings (MDT).11
Future autopsies In light of falling autopsy rates across Europe, alternatives to conventional postmortem examination are being explored. Such techniques tend to involve postmortem imaging of some form although a major drawback of this noninvasive autopsy approach is the lack of tissue for histopathological and microbiological examination. The Department of Health UK-funded trial sought to identify the accuracy of postmortem CT and MRI compared with full autopsy in a large series of adult deaths.9 They concluded that CT was more accurate than MRI for providing a cause of death compared with traditional autopsy. However, this technique had its limitations when the cause of death was due to ischaemic heart disease, pulmonary embolism, pneumonia and intra-abdominal lesions. However, a combination of postmortem imaging with other special techniques including, if necessary,
Journal of the Intensive Care Society 16(4) targeted biopsy can create a ‘minimally invasive autopsy’ (MIA), a useful compromise that can provide an accurate cause of death in approximately 70% of cases.10 MIA has the advantage of maintaining the integrity of the body and is therefore often more acceptable to families of the deceased. In addition, MIA retains an accurate record of the postmortem findings, in the form of the scans, for further assessment and review at a later stage should that be necessary, something which standard autopsy can never do.
Conclusion Economic drivers and technological change will continue to reshape the pathology landscape. The number of consented autopsies being conducted in UK hospitals is diminishing year on year leading to a continuous downward spiral in autopsy numbers. Not only does this have implications for those whose research depends on the availability of autopsy data, it also reduces our awareness and understanding of diagnostic errors, which autopsy can highlight. NCEPOD recommends increased communication between clinicians and pathologists. This, together with other simple measures such as formally training doctors and nurses to obtain consent for autopsies should be considered as these have been shown to increase autopsy numbers. Is there still a role for autopsy on ICU patients in modern health care? Undoubtedly there is, but like most things in health care, one must consider the benefits and costs. On the one hand, modern studies have demonstrated that misdiagnosis rates are lower than previously thought. With the decline in autopsy numbers, there is now a reduction in the number of hospitals and indeed the number of pathologists who have the experience and expertise to perform an autopsy on ICU patients. The value of the autopsy as a teaching and audit tool for ITU however seems unquestionable, even today and therefore the possibility of referring patients to specialist hospitals for autopsy should be considered if it cannot be conducted locally. Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding The authors received no financial support for the research, authorship and/or publication of this article.
Acknowledgement This work came as a result from the European Critical Care Foundation organised study day – ‘The decline of autopsy and its implication for critically ill patients’ held on the 6th March 2015.
Wong et al.
281
References 1. Goldman L, Sayson R, Robbins S, et al. The value of the autopsy in three medical eras. NEJM 1983; 308: 1000–1005. 2. Frohlich S, Ryan O, Murphy N, et al. Are autopsy findings still relevant to the management of critically ill patients in the modern era? Crit Care Med 2014; 42: 336–343. 3. Shojania KG, Burton EC, McDonald K, et al. Changes in rates of autopsy detected diagnostic errors over time: a systematic review. JAMA 2003; 289: 2849–2856. 4. The Coroner’s Autopsy. Do we deserve better? http:// www.ncepod.org.uk/2006Report/Downloads/ Coronial%20Autopsy%20Report%202006.pdf (accessed 15 April 2015). 5. http://www.nhs.uk/conditions/Post-mortem/Pages/ Introduction.aspx (accessed 15 April 2015). 6. Office for National Statistics. (2003). Mortality statistics series DH2 no 29. Review of the registrar general on
7.
8.
9.
10.
11.
deaths by cause, sex and age in England and Wales, 2002. London: HMSO, 2003. Winters B, Custer J, Galvagno SM, et al. Diagnostic errors in intensive care units: a systematic review of autopsy studies. BMJ Qual Saf 2012; 21: 894–902. Cullen DJ, and Nemaskal AR. The autopsy incidence of acute pulmonary embolism in critically ill surgical patients. Int Care Med 1986; 12: 399–403. Roberts IS, Benamore RE, Benbow EW, et al. Postmortem imaging as an alternative to autopsy in the diagnosis of adult deaths: a validation study. Lancet 2012; 379: 136–142. Wichmann D, Obbelode F, Vogel H, et al. Virtual autopsy as an alternative to traditional medical autopsy in the intensive care unit: a prospective cohort study. Ann Int Med 2012; 156: 123–130. Dimopoulos G, Piagnerelli M, Berre J, et al. Post mortem examination in the intensive care unit: still useful? Int Care Med 2004; 30: 2080–2085.
