Historical Perspective

Declining Clinical Autopsy Rates Versus Increasing Medicolegal Autopsy Rates in Halifax, Nova Scotia
Why the Difference? A Historical Perspective
Marnie J. Wood, BSc; Ashim K. Guha, MD, PhD

The downward trend in the rate of clinical autopsies has been extensively documented in the literature. This decline is of concern when the benets of the clinical autopsy are considered. In contrast, the rate of medicolegal autopsies has not been studied in such detail. What little reference there is to medicolegal autopsy rates suggests an absence of the same downward trend. A retrospective review of autopsy data over a 13-year period from the Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia, and from the Ofce of the Chief Medical Examiner of Nova Scotia was conducted. This review showed a difference between the rates of clinical and medicolegal autopsies for the metro Halifax area. The clinical autopsy rate was consistently less than 30% and declined to 15% in 1999, while the medicolegal autopsy rate was consistently greater than 40% and rose to 62% in 1999. The literature proposes many reasons for the decline in the clinical autopsy rate, but none for this difference between rates. The explanation proposed here is the changing and currently uncertain purpose of the clinical autopsy versus the clear, and consistent over time, purpose of the medicolegal autopsy. (Arch Pathol Lab Med. 2001;125:924930) he word autopsy is from the Greek, meaning to see for oneself. Along with the terms necropsy (literally, to look at the dead) and postmortem examination (which may include external and/or internal examination), autopsy has come to refer to the systematic examination of a dead person for medical, legal, and/or scientic purposes.1 The autopsy is the ultimate medical consultation. 2 Historically, the autopsy played a very important role in the development of medicine. Early clinical autopsies were done by researchers and clinicians who had followed their patients throughout life and the course of their diseases. Autopsies were more limited several centuries ago, when the postmortem examination was carried out only until the suspected cause of death had been demonstrated, and
Accepted for publication February 13, 2001. From the Faculty of Medicine, Dalhousie University, Halifax, Nova Scotia (Ms Wood and Dr Guha); and the Department of Pathology, Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia (Dr Guha). Reprints: Ashim K. Guha, MD, PhD, Division of Anatomical Pathology, Mackenzie Bldg, Room 720, Queen Elizabeth II Health Sciences Centre, Victoria General Site, Halifax, Nova Scotia, Canada B3H 1V8 (e-mail: akguha@is.dal.ca). 924 Arch Pathol Lab MedVol 125, July 2001

may only have included examination of 1 organ or a small area of the body.3 In those cases, the recorded cause of death may have been unrelated to the actual cause or may have been merely a complication of the unrecognized fatal disease. Teachers of anatomy also used the autopsy as a teaching tool. Gradually, pathology became a separate specialty, dened by Virchow as the science that studies the causes, mechanisms, and consequences of diseases. 4 The task of autopsy performance has fallen to pathologists, and the autopsy has become a more complete and consistent procedure. The procedure is done in an autopsy suite equipped with all the necessary tools, which may be located at a variety of sites, such as the hospital or medical examiners building, but which is usually away from the general hustle and bustle. Modern techniques, such as microscopy, immunouorescence, and immunoperoxidase staining have added new dimensions to this very old procedure. Two basic types of autopsy exist. The clinical or academic autopsy (also referred to as a hospital autopsy5) is done at the request of the family or the physician (with permission from the family) of the deceased.6 In this case, the cause of death is usually known. The purposes of this type of autopsy include determining the cause of death (if unknown), providing correlations between clinical diagnoses and symptoms, determining the effectiveness of treatment, studying the course and extent of disease processes, and educating medical personnel.7 The second type, a forensic or medicolegal autopsy, is performed when the cause of death is not certain, usually in the case of unexpected, unexplained, or unnatural deaths.8 These medicolegal cases may be further divided into criminal cases, which are suspicious in nature, and noncriminal cases, in which no foul play is suspected.6 A medicolegal autopsy may be ordered by an ofcial death investigator to assist in determining the cause and manner of death.7 This information must be determined accurately for the death certicate. In all medicolegal cases, the legal authority (such as the provincial medical examiner in Nova Scotia) has the sole power to order an autopsy, and no postmortem procedures may be carried out without his or her permission.9 The medical examiner, if also a pathologist, may perform an autopsy or may refer the autopsy to be done by a hospital pathologist.8,10 Even the general public has a vague idea of what the performance of an autopsy entails, although they may not
Autopsy Rates in Halifax, Nova ScotiaWood & Guha

Comparison of clinical and medicolegal autopsy rates for Halifax, Nova Scotia, from 1987 to 1999. Each point represents the autopsy rate for a particular calendar year. The clinical autopsy rate was calculated as the number of clinical autopsies performed divided by the number of in-hospital deaths (excluding medical examiner cases) for the same year multiplied by 100. The medicolegal autopsy rate was calculated as the number of medicolegal autopsies performed divided by the number of deaths reported to the medical examiner for the same year multiplied by 100.

fully understand why autopsies are done. Indeed, this lack of understanding may also exist within the medical community itself. In the last half of the century, the rate of clinical autopsy performance has dropped.2,1113 The literature suggests medicolegal autopsies are not experiencing the same trend. Rates of autopsy are now so low that postmortem examination other than for medicolegal purposes is in jeopardy. 14 This article demonstrates a difference in rate between clinical and medicolegal autopsies using local data and then considers reasons this difference might exist. CURRENT STATUS OF THE AUTOPSY Clinical autopsy rates have been in decline for the last half of the century. The gures cited are American, but similar traits have been noted in Canada.13 As early as the 1950s, concerned persons were expressing alarm over the falling rates.14 In the early 1940s, the autopsy rate for inhospital deaths was approximately 50%.2,13 In the 1950s, the required rate for accreditation was 25% in teaching hospitals and 20% in nonteaching hospitals.2 In the early 1970s, the required rate was dropped and the numbers plummeted.2,13 By 1973, rates had fallen to 22%, and by 1984, to 13.2%.2 A national study in 1994 by the College of American Pathologists found 75% of hospitals had an autopsy rate less than 13.5%, and 50% had rates less than 8.5%.15 In 1995, some North American hospitals had rates as low as 7%.13 A 1996 survey of 2 major teaching hospitals found rates of 5% and 15%.2 In contrast, there is less literature on the rate of medicolegal autopsies. It has, however, been suggested that medicolegal autopsy rates are not falling as steeply as clinical autopsy rates.14 A retrospective review of data from the Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia, and the Ofce of the Nova Scotia Medical Examiner conrms a difference between clinical and medicolegal autopsy rates in the metro Halifax area. The Figure illustrates these ndings. The rate of clinical autopsies is consistently less than 30%, falling to less than 20% in the nal 4 years examined. The medicolegal rate is consistently greater than 40%, rising to 62% in the last 2 years examined. Although the clinical autopsy rates are higher than rates recorded in the literature, the clinical autopsy rate is far less than the medicolegal autopsy rate in each of the 13 years
Arch Pathol Lab MedVol 125, July 2001

examined. The data also demonstrate that the clinical autopsy rate has decreased over time, while the medicolegal autopsy rate has increased. The literature suggests a variety of reasons for the decrease in the number of clinical autopsies. The increasing cost of the autopsy, which fewer people seem willing to pay, is almost certainly a factor. A single autopsy, depending on the extent and number of extra procedures used, may cost between US $1200 and $3000.14 In Canada, the cost of clinical autopsies is covered by the hospital, and many administrators want to spend only a small portion of their budget on autopsies.13 The budget for medicolegal autopsies is covered by the Department of Justice. A minimum autopsy rate is no longer required for hospital accreditation, so there is no push to maintain a certain standard rate.14 Also, physicians do not realize the importance of the autopsy because medical education no longer emphasizes this procedure.14 Physicians graduating in the past 10 years from two thirds of medical schools have not been exposed to an autopsy. Why should doctors care about having autopsies performed when they dont even understand what the procedure involves? 13 Physicians who do not realize the importance of clinical autopsies are probably less likely to request permission to perform autopsies. Many physicians take comfort in the perceived certainty that new clinical diagnostic tools can replace the autopsy, making it out-of-date and unnecessary. This opinion has led to a reduction in the clinical autopsy rate. The clinical autopsy might, however, still be useful in elucidating the limits and weak areas of new diagnostic tools. The fear physicians have that an autopsy will reveal a mistake in treatment and lead to litigation prevents many from requesting an autopsy. No one wants to know about errors in this time of litigation. 16 There is evidence to suggest these fears are ungrounded. A study by Haque et al17 at the University of Texas Medical Branch at Galveston found only 2 of 6168 clinical autopsies provided results leading to malpractice suits. Many more autopsies provide evidence that exonerates hospital staff of wrong-doing.10,14,17 The clinical autopsy is also losing prestige among pathologists and is competing for time with other laboratory duties. In addition, family objections based on misconceptions may result in permission for a clinical autopsy not being granted. It is also possible that physicians
Autopsy Rates in Halifax, Nova ScotiaWood & Guha 925

are simply not asking for the familys permission to perform a clinical autopsy because they are not trained to deal with this uncomfortable situation.14,18 These reasons may explain the decline in number of clinical autopsies; however, they do not explain why the rate of medicolegal autopsies is not declining as well. The cost of both procedures is equally high, and neither type of autopsy is emphasized in medical education. Granted, families of the deceased are not able to refuse permission for a medicolegal autopsy to be performed, but pathologists may be equally reluctant to perform this time-consuming procedure if newer techniques exist to obtain the same information. An examination of the history of the clinical and medicolegal autopsy may provide some insight into the recent difference in performance rates. THE CLINICAL AUTOPSY The knowledge of normal versus abnormal anatomy dates as far back, and perhaps further, as 1500 BC, when animal entrails were examined for clues to the future.1 During the mummication process, the Egyptians also gathered such information about humans. As early as 400 BC, Hippocrates believed that disease resulted from natural causes,3 in contrast to the belief in supernatural, spiritual, and humor disturbance causes held by the majority of the population at that time. These mistaken beliefs would initially prevent the clinical autopsy from nding an accepted place in medicine. It took a long time for the connection to be made between disease and associated changes in the organs of the body. The autopsy did have an early place in medicolegal cases and in the further study of anatomy. Around 300 BC, Herophilus was using dissections to teach anatomy, and he wrote a treatise on human anatomy but paid no attention to abnormalities of structure.1 In this manner, empirical knowledge of anatomy was gradually gathered. In 200 BC, Erasistratus was probably the rst to carry out dissections to look for changes due to disease. However, when Galen popularized disturbances in the 4 body humors as the cause of disease,3 he sent medicine on a long-lasting wild goose chase. In the 13th century, dissections were being done during plagues in order to better understand the illness. 4 These procedures were paid for by the state and illustrate that people were beginning to see the autopsy as more than a way to learn anatomy. The rst recorded autopsy performed in the New World was in 1533. Female conjoined twins were dissected to determine if they had 1 soul or 2.3 At the end of the 15th century, the Italian physician Antonio Benivieni was the champion of the autopsy. In 1507 he published the rst book of anatomical pathology, titled Remarkable Hidden Causes of Disease.3,4 This was the very beginning of clinicopathologic correlation and set the stage for wider acknowledgement of the link between disease and body structure. During the 17th century, clinical autopsies and the recording of their results became more common. While the prevailing view was still that clinical autopsies were a waste of time, in 1679 the Sepulchretum was published. The Sepulchretum contained a collection of autopsy reports from noted physicians of the time.3 The reports were inconsistent and contained few clinicopathologic links, but this book set the stage for the clinical autopsy to assume a place of importance in medicine. By the 18th century, reference to clinical autopsy reports could be found in
926 Arch Pathol Lab MedVol 125, July 2001

many journals.3 Physician Hermann Boerhaave was publishing far more detailed reports than had been seen before. Boerhaave was also the rst to begin placing importance on the complete clinical history of the patient, stating, Everything pertaining to the case must be listed: nor that least thing neglected which a critical reader might rightly seek to understand the malady.3 This attitude can be seen throughout the following century, as the clinical autopsy became the end of the complete story of an illness. The clinical autopsy was gaining a major place in medicine. Italian physician Morganis work at the University of Padua solidied the correlation between clinical symptoms and pathologic ndings, and pathology became the basis for medicine of the time.3,19 Xavier Bichat, a French physician in the 19th century, felt very strongly about the place of the clinical autopsy in medicine: You can take notes for 25 years, from morning to evening, by the patients bedside on diseases of the lung, heart and stomach, and the result will be a long list of confusing symptoms, leading to incoherent conclusions. Open a few bodies, and you will see darkness immediately recede. 4 Bichat was also the rst to consider looking farther than the organs and examine their component tissues. This was before the time of the microscope, but he still made much progress in this area. At this point, autopsies were still done by clinicians. Clinicopathologic correlations allowed diseases to be dened much more clearly and their progression over time noted. The development of the microscope allowed others to pick up where Bichat had left off, and pathology was revolutionized. The gross examination of the body was joined by the rst of many high-tech companions. In fact, the limits of gross examination alone were being recognized. Rokitansky (mid-19th century) worked at the Pathological Institute at Vienna and began the separation between pathologists and clinicians. Rokitansky performed or observed autopsies on almost every patient who died in the hospital (more than 30 000).1 He attempted to work backward, using autopsy results to determine the clinical ndings.19 Although many of his theories proved incorrect, Rokitansky provided a huge databank of autopsy information. His advances were spread throughout the world. Around the same time, a German physician, Virchow, believed that pathology was a science in its own right, separate from the clinical side of medicine. Virchow was also a leader in setting forth a regular and denite technique for the autopsy, and he emphasized a long, complete autopsy.3,4,19 At the beginning of the 20th century, most clinicians were still spending a great deal of time in the autopsy room, conrming or clarifying diagnoses. Prominent American physicians spent time in Europe learning anatomical pathology from the masters.19 Sir William Osler was a Montreal physician who worked at Johns Hopkins and turned the place upside down. 19 Osler championed the autopsy, following his patients to the autopsy room and even leaving detailed instructions for his own autopsy.19 He is quoted as writing, To investigate the causes of death, to examine carefully the condition of organs, after such changes have gone on in them as to render existence impossible, and to apply such knowledge to the prevention and treatment of disease, is one of the highest objects of the physician. 19 This expresses the use of the clinical autopsy at the time, which was to elucidate disease processes and the effects of treatment. Clinical autopsies soon
Autopsy Rates in Halifax, Nova ScotiaWood & Guha

came to have another use. In 1910, an American report by Flexner was published. The report stated that good hospitals had high autopsy rates, otherwise physicians could bury their mistakes.3 This report was followed by the Cabot Report in 1912, which revealed many diagnostic errors were occurring in American hospitals and could be found only through autopsies. Through these reports, the clinical autopsy became associated with quality control.3,19 The purpose of the clinical autopsy has shifted since its beginnings as a tool for teaching anatomy and a way to search for the soul. As information about human anatomy was gathered, the autopsy gradually became a way to explore disease processes. Pathology, and the autopsy as its tool, then became a science in its own right for gathering more information about disease processes. In the rst part of this century, the clinical autopsy became a method of monitoring the quality of care and the diagnostic abilities of physicians and their equipment. Throughout the last half of the 20th century, the importance of the autopsy to the medical community and the public had to be defended. This is a long fall from the esteemed position the autopsy held a short 100 years ago. Administrators, physicians, and the public may be falling into the same trap. There is a tendency to place much faith in new, high-tech diagnostic techniques, such as ultrasound, magnetic resonance imaging, computed tomography, and endoscopy. The belief is that these procedures illustrate all there is to be known about a particular disease and make autopsy results redundant. However, this faith in high-tech diagnostic tools over the clinical autopsy may be somewhat misplaced. As Lundberg20 pointed out, there is a gap between how well high-tech diagnostic medicine performs in theory and how well it performs when human decision making (how and when to use the technology) is added to the equation. Several studies have shown that the autopsy should still be considered the gold standard in diagnosis and that autopsies are absolutely necessary to assure the quality of care given to patients, both those treated successfully and, especially, those who die. In a study by Nichols et al21 at a major tertiary-care center, 44.9% of autopsies found at least 1 undiagnosed cause of death. For example, a patient with cirrhosis had what was interpreted radiologically to be a scar. On necropsy, a rare mixed hepatocellular and cholangiocarcinoma was discovered. In two thirds of these cases the undiagnosed cause of death was determined to have been treatable. These misdiagnoses were often related to patient inability to relate symptoms or the masking of symptoms by other treatments. In another study, Burton et al22 found a 44% discordance between the clinical and autopsy diagnosis of neoplasms at the Medical Center of Louisiana at New Orleans. Furthermore, Roosen et al23 looked at autopsies performed on patients dying in the medical intensive care unit. In spite of more modern diagnostic techniques, 16% of these autopsies found missed major diagnoses, the treatment of which might have prolonged survival. Finally, Sonderegger et al24 showed a drop in the frequency of major missed diagnoses from 30% to 14% over 20 years, mainly due to a decrease in the misdiagnosis of cardiovascular disease. The autopsy rate at this facility in Zurich, Switzerland, was approximately 90% consistently over the study period, a record that enhances the credibility of the ndings. While the purpose of the clinical autopsy may not be as clear as in the past, it can be seen from these studies that the autopsy has a large role to play in quality
Arch Pathol Lab MedVol 125, July 2001

assurance and in elucidating the limitations of diagnostic techniques. It can therefore be concluded that the autopsy is still a cost-effective and useful procedure. THE MEDICOLEGAL AUTOPSY Forensic pathology is generally considered a recently developed subspecialty of pathology dealing with the examination of living or dead persons to provide expert opinion regarding the cause, mechanism, and manner of disease, injury, or death. This discipline is also concerned with identifying persons, determining the signicance of biological and physical evidence, correlating and reconstructing wounds, performing comprehensive medicolegal death investigations,7 and applying medical knowledge to legal issues. However, the history of forensic medicine and the postmortem examination actually dates back to antiquity, when bodies were most often examined to determine if death occurred as a result of suicide or homicide. The ancient Greeks felt suicide to be a rebellion against the gods and the Romans saw it as a crime against the state. As such, there was great stigma attached to the families of those who committed suicide. In 1184, the Roman Catholic Church added suicide to canon law, further adding to the importance of distinguishing it from other manners of death.16 During the 6th century, Justinian law called medicolegal experts to testify in cases of rape, criminal abortion, and murder. Attitudes at the time were still against opening the bodies of the deceased, but external postmortem examinations were done and the evidence obtained was used in court.16 Subsequently, during the medieval period, dissections of corpses were done in Italy at the University of Bologna under the control of the faculty of law to solve legal matters. There is also a report dating from 1302 of a court-ordered autopsy done at the University of Bologna to investigate the suspicious death of a nobleman.3 Furthermore, there was a Chinese publication in the 13th century that was much ahead of its time. Titled His Yuan Lu or Instructions to the Coroner, it dealt with ndings in cases of infanticide, drowning, hanging, poisoning, and assault. In Germany during the 16th century, the code of Bamburg, and later the broader penal codes, brought about a requirement for medical testimony in putative forensic cases. These codes also allowed the opening of bodies to examine the depth of and damage caused by wounds.16 Today, a murder case would be thrown out of court without complete autopsy results. Throughout history, there have always been religious barriers to the performance of autopsies. Clinical autopsies in some cultures were forbidden. For example, Jewish laws stated that an autopsy was a disgrace to the body, and prior to the 18th century, clinical autopsies were not permitted. However, there is a passage in the Talmud allowing an autopsy if the evidence provided might save the life of an accused murderer.16 This illustrates that even religious authorities at times have recognized the importance of the forensic autopsy and its difference from a clinical autopsy. The eld of legal medicine continued to grow, and in the 18th century the rst series of lectures on the topic was given at the University of Edinburgh by Sir Andrew Duncan. As the eld expanded, Europeans developed 2 different ways of providing medicolegal professionals for society to use. On the continent and in Scotland, the universities and associated hospitals took over the provision
Autopsy Rates in Halifax, Nova ScotiaWood & Guha 927

of specially trained physicians. In England, the coroner system was established under the constabulary. A coroner was an ofcial of the court with no medical training. The coroners responsibility was to investigate deaths falling under his jurisdiction. During the investigation, the coroner often consulted a police surgeon, usually a general practitioner with some special training. A pathologist was consulted only if an autopsy was required. The term coroner was derived from crowner. A crowner was a representative of the King, who collected taxes, convened court, and distributed the property of those found guilty of murder. Because of the latter responsibility, the crowner became associated with death and its investigation.16 The coroner system began in America in the 17th century, but was gradually abolished because of the potential for political interference. This system was replaced by the district attorneys ofce, which took over the legal aspects of death investigation, and the medical examiners ofce, which took over the medical aspects of the investigation. The medical examiners ofce is staffed by trained forensic pathologists.16 The College of American Pathologists rst formed the Committee on Forensic Pathology in 1952.2 Its priority was to increase forensic training for pathologists and raise the standards in what was becoming a subspecialty. Through the 1960s and 1970s, funding became available for fellowships in this subspecialty. In Nova Scotia, Canada, the medical examiners ofce is under the jurisdiction of the Department of Justice. It is the budget of this department that pays for medicolegal autopsies. Deaths falling under the jurisdiction of the medical examiner include death from any unnatural entity, when a medical cause of death cannot be established with reasonable accuracy, death occurring while a person is in custody (jail or psychiatric), and death by undue means or when there may be medical culpability.7 All deaths meeting one of these descriptions must be reported to the medical examiner, who becomes responsible for the investigation and for the ordering of a medicolegal autopsy if deemed necessary.6,7 The medicolegal autopsy differs from the clinical autopsy in several ways, one of which is that it has not experienced the same decline in rate or prestige. If anything, the forensic autopsy has had a recent gain in popularity, especially among the general public. Autopsy results are featured in news reports of suspicious deaths. Television networks, such as A&E and TLC, are full of specials on medical examiners ofces and forensic medicine. Even popular television shows, like the X-Files, often feature scenes of an autopsy (usually done by the female character, alone, late at night in a dark and spooky room). Another difference is that the medicolegal autopsy seeks to answer a specic legal question: What is the cause and manner (natural, homicide, suicide, accidental) of death in this case?16 This question has remained constant for the medicolegal autopsy throughout time, in contrast to the clinical autopsy, which seeks to answer many questions. While the point of a medicolegal autopsy remains focused, the adequate answering of this question often requires an extensive investigation. The medical examiner has the power to determine the extent of the autopsy to be performed, but medicolegal autopsies are rarely limited.7 To contribute as much as possible to the prosecution of the guilty and the defense of the innocent, there is an extra emphasis on completeness during a forensic case. The only thing worse than no autopsy is a partial autopsy in
928 Arch Pathol Lab MedVol 125, July 2001

forensic cases.25 A partial autopsy may lead to a later exhumation of the body, as what seemed to be unimportant at the time turns out to be vital to the case. While the actual dissection is often the same as the clinical autopsy, the medicolegal postmortem examination often begins at an earlier stage, the place of death. Here the position, location, clothing, personal effects, signs of death, and injuries are correlated with the circumstances and documented.6,7 Careful observations of the surroundings must be made and meticulous notes kept.16 Also, a forensic investigation may involve trying to determine information readily available in a clinical case, such as the identity of the deceased and time of death. The external examination may also take on additional importance in a forensic case. There is often much information to be gathered from external wounds. The speed and angle of a striking automobile may be determined from the characteristics of abrasions. Bullet exit wounds can be distinguished from the entrance wounds and help distinguish a homicide from a suicide.16 All of this information is key to a medicolegal case. The objectives of the medicolegal autopsy, as listed by Knight,6 are to make a positive identication of the body and to assess the size, physique, and nourishment; to determine the cause of death; to determine the mode of dying and the time of death when necessary and possible; to demonstrate all external and internal abnormalities, malformations, and diseases; to detect, describe, and measure any external and internal injuries; to obtain samples for analysis, including microbiological and histologic examination, as well as any other necessary investigations; to retain relevant organs and tissues as evidence; to obtain photographs and video lms for evidential and teaching use; to provide a full written report of the autopsy ndings; to offer an expert interpretation of those ndings; and to restore the body to the best possible cosmetic condition before release to relatives. The autopsy is a vital tool in medicolegal investigations for which there is no substitute. The nature of a medicolegal case is that there is not an adequate clinical picture with which to determine a cause and manner of death. The information gathered through an autopsy is especially important in cases of suspected foul play. The medicolegal autopsy does not seem to be facing the same problem as the clinical autopsy, that is, a decline in the rate of performance. One reason may be that medicolegal autopsies are not funded by the hospital and therefore do not compete for limited funds with treatments and diagnostic procedures, which are often perceived to provide greater yield. The political reality is that stable funding of medicolegal cases generally exists as a result of jurisdiction lying with the Department of Justice and not with the Department of Health in most systems. This reality may play a role in causing the observed rate differences between medicolegal and clinical autopsies. Or, perhaps it is because the purpose of the medicolegal autopsy has remained clear and unchanged through the centuries. Also, there is no procedure that could provide the same amount of information with which to replace the medicolegal autopsy. Oscar Schultz in 1932 in a study for the National Academy of Sciences of the United States said, The determination of the cause of death in an exact and scientic manner requires a necropsy. 16 CONCLUSIONS The decline of clinical autopsy rates means the benets these autopsies could provide are being missed. The purAutopsy Rates in Halifax, Nova ScotiaWood & Guha

poses of the clinical autopsy may have changed, but it appears there is still much it can contribute. The clinical autopsy is still valuable as the gold standard by which diagnostic tools, treatment methods, and clinical decisionmaking procedures can be measured.26 How do you assess the quality of care given to your sickest patients, the ones who die? 27 We have shown in this article that the limitations of diagnostic tools can be dened through clinical autopsies. Furthermore, there is a need to evaluate clinical decision-making procedures to prevent the same errors being made repeatedly.3 Autopsy results can also show otherwise unknown adverse effects of treatments, such as the dose-related cardiotoxicity of doxorubicin. The toxic effects of many chemotherapy drugs can be monitored through autopsy results, and future doses and treatment protocols then can be altered appropriately. Clinical autopsy performance is very important in the study, treatment, and prevention of disease. Even at the current low rates, autopsies have helped elucidate conditions such as shock, cardiovascular disease, aging, and cancer.14 In 1 case that was reported, a male infant with hypoplastic left heart syndrome died after corrective surgery. Those involved in his care attended the autopsy, which provided an opportunity to learn more about the syndrome, to evaluate the surgical technique, and to perform quality assessment.28 New environmental diseases, such as some forms of asthma, are constantly being discovered, and some may be missed due to the low clinical autopsy rate. The hazards of occupational exposures to agents like asbestos have been proven through clinical autopsies. This has resulted in major changes to working conditions. A clinical autopsy may reveal the presence of a contagious disease and allow the quick treatment of unknowingly exposed individuals.27 An epidemic may be averted through early warning. Another case illustrating the great importance of incidental autopsy ndings is one in which a breast lump found on autopsy was determined to be invasive ductal carcinoma. Informing female family members of this nding and advising them of their increased risk could potentially save a life.29 Clinical autopsies may also be used to monitor public health and ensure the accuracy of vital statistics.14,27 Vital statistics are used to determine the public health policies and to assess the effectiveness of public health services. These statistics are compiled from the information on death certicates. Kircher is cited30 as having found that there is approximately a 10% error rate in death certicate diagnoses when compared to autopsy results. The death statistics, which are used to develop public health strategies, are based on the less-representative facts on the death certicates. The performance of autopsies at a high rate may correct the misinformation and have a great impact on public health policy. For example, Kircher also found the death certicate diagnosis of heart disease is an inaccurate reection of the actual incidence, which is lower.30 This is signicant when the amount of money spent on preventive strategies for heart disease is considered. Clinical autopsies are also very important in the monitoring of levels of clinically silent diseases, such as coronary artery disease.14 Without the performance of autopsies, only guesses can be made as to their incidence. There are also extensive benets for the families of those on whom clinical autopsies are performed. Relatives often do not understand how everything happened. An autopsy can provide information about how a loved one died and ease the familys mind regarding the
Arch Pathol Lab MedVol 125, July 2001

extent of suffering. A family may also be made aware of the benets autopsy research can provide for future treatments. In this way, the family may be able to see some good come out of the death. However, for the family to reap these benets, autopsy results must be made available in an understandable format. It may be that a lack of perceived purpose for the clinical autopsy, whereas that of the medicolegal autopsy remains constant and clear, has contributed to the isolated decline in the clinical autopsy rate. The guidelines of the College of American Pathologists25 state the current purposes of the autopsy as follows: to establish a cause of death, determine the manner of death, compare premortem and postmortem diagnoses, produce vital statistics, and to monitor the health of the public. These purposes are as important to medicine today as past purposes were in their time. These purposes would be better served by an increase in the clinical autopsy rate. There is a fairly large amount of literature expressing concern over the decline in frequency and illustrating the many positives to be gained through clinical autopsies. This suggests the procedure does not suffer from a complete loss of interest but, more likely, is lagging behind in the ght for limited funds. The clinical autopsy rate at the Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia, is not as low as has been published for other hospitals, but the rate is falling. The medical community might benet from attempts to increase the rate of clinical autopsy performance. Future research into ways of increasing the clinical autopsy rate is needed.
References
1. Dada MA, Ansari NA. The postmortem examination in diagnosis. J Clin Pathol. 1996;49:965966. 2. Derman H, Wagner LR. Anatomic and consultative pathology practice. Arch Pathol Lab Med. 1997;121:12141222. 3. King LS, Meehan MC. A history of the autopsy. Am J Pathol. 1973;73:514 542. 4. Rosai J. Pathology: a historical opportunity. Am J Pathol. 1997;151:36. 5. Cotton D, Cross S. The Hospital Autopsy. Toronto, Ontario: ButterworthHeinemann Ltd; 1993. 6. Knight B. Forensic Pathology. 2nd ed. New York, NY: Arnold/Oxford University Press Inc; 1996. 7. Randall B, Fierro M, Froede R. Practice guideline for forensic pathology. Arch Pathol Lab Med. 1998;122:10561064. 8. What Deaths Are Notiable to the Medical Examiner? Advice for Doctors and Other Medical Professionals in Nova Scotia. Halifax, Nova Scotia: Department of Justice, Ofce of the Chief Medical Examiner; 2000. 9. Campbell K. Post-mortems: how and why they are carried out. Nurs Times. 1997;93(16):5254. 10. Geller SA. Autopsy. Sci Am. 1983;248:124136. 11. Nemetz PN, Leibson C, Naessens JM, Beard M, Tangalos E, Kurland LT. Determinants of the autopsy decision: a statistical analysis. Am J Clin Pathol. 1997;108:175183. 12. Goldman L, Sayson R, Robbins S, Cohn SH, Bettmann M, Weisberg M. The value of the autopsy in three medical eras. N Engl J Med. 1983;308:1000 1004. 13. Lowry F. Failure to perform autopsies means some MDs walking in a fog of misplaced optimism. CMAJ. 1995;153:811814. 14. AMA Council on Scientic Affairs. Autopsy: a comprehensive review of current issues. Arch Pathol Lab Med. 1996;120:721726. 15. Hanzlick R, Baker P. Institutional autopsy rates. Arch Intern Med. 1998; 158:11711172. 16. Hill RB, Anderson RE. The Autopsy: Medical Practice and Public Policy. Toronto, Ontario: Butterworths; 1988. 17. Haque AK, Patterson RC, Grafe MR. High autopsy rates at a university medical center: what has gone right? Arch Pathol Lab Med. 1996;120:727732. 18. McPhee SJ. Maximizing the benets of autopsy for clinicians and families: what needs to be done. Arch Pathol Lab Med. 1996;120:743748. 19. Hill RB, Anderson RE. The recent history of the autopsy. Arch Pathol Lab Med. 1996;120:702712. 20. Lundberg G. Low-tech autopsies in the era of high-tech medicine: continued value for quality assurance and patient safety. JAMA. 1998;280:12731274. 21. Nichols L, Aronica P, Babe C. Are autopsies obsolete? Am J Clin Pathol. 1998;110:210218.

Autopsy Rates in Halifax, Nova ScotiaWood & Guha 929

22. Burton E, Troxclair D, Newman W. Autopsy diagnoses of malignant neoplasms: how often are clinical diagnoses incorrect? JAMA. 1998;280:12451248. 23. Roosen J, Frans E, Wilmer A, Knockaert D, Bobbaers H. Comparison of premortem clinical diagnoses in critically ill patients and subsequent autopsy ndings. Mayo Clin Proc. 2000;7:562567. 24. Sonderegger-Iseli K, Burger S, Muntwyler J, Salomon F. Diagnostic errors in three medical eras: a necropsy study. Lancet. 2000;355:20272031. 25. Hutchins GM, ed. Autopsy Performance and Reporting. Northeld, Ill: College of American Pathologists; 1990. 26. Pellegrino ED. The autopsy: some ethical reections on the obligations of pathologists, hospitals, families, and society. Arch Pathol Lab Med. 1996;120:739742.

27. Lundberg GD. College of American Pathologists Conference XXIX on Restructuring Autopsy Practice for Health Care Reform: lets make this autopsy conference matter. Arch Pathol Lab Med. 1996;120:736738. 28. Bennett A, Collins K, Hanzlick R. Outcome analysis and quality assessment. Arch Intern Med. 1999;159:13991400. 29. Collins K, Bennett A, Hanzlick R. The autopsy and the living. Arch Intern Med. 1999;159:23912392. 30. Hill RB, Anderson RE. The evolving purposes of the autopsy: twenty-rstcentury values from an eighteenth-century procedure. Perspect Biol Med. 1989; 32:223233.

930 Arch Pathol Lab MedVol 125, July 2001

Autopsy Rates in Halifax, Nova ScotiaWood & Guha