Autopsy-Based Surveillance Systems
Cause of death evaluation is an important component of the investigative process for emerging infectious diseases. When evaluating potentially infectious diseases as the causes of unexplained deaths, the use of autopsies has a number of advantages over death certificates: 1) availability of human tissues allows for enhanced diagnostic capacity and results in accurate determination of cause of death; 2) insights into pathogenesis and route of infection are gained; 3) rapid public health notification of findings is possible; and 4) recognition of additional infections not on death certificates is possible. The systematic collection and evaluation of this additional information affords an important opportunity for enhancing infectious disease surveillance. Monitoring of unexplained deaths and critical illnesses via autopsies at the state and local levels yields vital information about the actual numbers of cases of infectious diseases and provides insight into strategies for prevention.
Med-X
The New Mexico Office of the Medical Investigator created a Medical Examiner Syndromic Surveillance System (Med-X) for all fatal infectious diseases, which can be used in medical examiner jurisdictions. A basic principle of the Med-X system is to seek organism-specific diagnoses in all potential infectious disease deaths investigated as unexplained by medical examiners. Designed initially to provide the capacity to identify fatalities resulting from bioterrorism and infections of public health importance, the model is based on two types of information: symptoms (Box) and pathologic syndromes found at autopsy (Table 2). The lists of both symptoms and syndromes are derived from most known bioterrorism-related illnesses. The symptom list (Box) serves to recognize and capture potential cases and drive decisions about autopsy performance; the syndrome list is used for early reporting of cases to the New Mexico Department of Health. For example, one of the 11 autopsy-based pathologic syndromes (community-acquired pneumonia and acute respiratory distress syndrome) might indicate the decedent had plague or tularemia; however, it is much more likely the decedent had influenza, pneumococcal disease, or various other more common conditions (Table 2). This information is valuable for public health officials in their decision-making regarding implementing prevention and control measures.
In New Mexico during 2000–2002, a total of 6,104 medical examiner cases were examined. Of these, 250 met entry criteria (medical examiner autopsy case with a defined symptom or syndrome), of which 141 (56%) decedents had a target pathologic syndrome and 127 (51%) were found to have an infectious disease. Three symptom sets were found to be highly predictive of infection in an otherwise unexplained death: 1) fever and respiratory symptoms (72%), 2) influenza-like symptoms (65%), and 3) encephalopathy or new-onset seizures (50%); sudden unexpected death (19%) was found to be less likely to represent an infection. Furthermore, in 81% of infectious disease cases, an organism-specific diagnosis was determined, with 58% representing notifiable conditions in New Mexico, including Streptococcus pneumoniae(37 cases), Streptococcus pyogenes (eight cases), and Haemophilus influenzae (five cases), as well as one case each of Mycobacterium tuberculosis and botulism and two cases of human immunodeficiency virus (HIV) infection. These findings indicate the value of pathologists conducting routine microbiologic testing in cases that come under their jurisdiction and have symptoms predictive of infection.
UNEX
Another surveillance system using cause of death as a tool, Surveillance of Probable Infectious Etiology for Unexplained Death (UNEX), was initiated in 1995 as part of the CDC Emerging Infections Program in California, Connecticut, Minnesota, and Oregon. The goals of UNEX are to identify novel or newly emerging pathogens; to identify sudden, unexplained deaths attributed to known pathogens; to monitor the epidemiologic features of fatal infections; and to improve diagnostic postmortem testing. In Minnesota, cases of unexplained critical illness also are included in the UNEX surveillance system. Therefore, the cases include deaths or critical illnesses unexplained by routine testing that have premortem or postmortem findings suggestive of infectious etiology such as fever, leukocytosis, cerebrospinal fluid pleocytosis, or histopathologic evidence of an infection. Although persons who were previously healthy and aged <50 years are the focus of UNEX in Minnesota, the system is not limited to this population. UNEX cases are reported by clinical partners, including infectious disease physicians, infection preventionists, and hospital pathologists, whereas the main reporters outside of acute care facilities are medical examiners. Sources of information include autopsy and pathology reports, medical records, scene investigation findings, and biologic specimens; results are correlated with pathologic and clinical findings to determine the cause of death. During 1995–2005, respiratory cases were the most common syndrome in most years, with the number of these cases increasing over time (Figure 2).
(Enlarge Image)
Figure 2.
Unexplained deaths or critical Illnesses* — UNEX surveillance system, Minnesota, 1995–2005
Abbreviations: UNEX = Surveillance of Probable Infectious Etiology for Unexplained Death; GI = gastrointestinal.
* In Minnesota, in addition to deaths, cases of unexplained critical illness also are included in the UNEX surveillance system. Cases in Minnesota include deaths or critical illnesses unexplained by routine testing that have premortem or postmortem findings suggestive of infectious etiology such as fever, leukocytosis, cerebrospinal fluid pleocytosis, or histopathologic evidence of an infection.
Med-X Combined With UNEX
A Med-X surveillance system based on the New Mexico model also was initiated in Minnesota in 2006, enabling further description of infectious etiologies of death during 2006–2011. During this period, an average rate of 12 infectious deaths per 100,000 population was identified, encompassing 1,099 cases captured by UNEX and Med-X combined (723 UNEX cases, 908 Med-X cases, and 532 that fit the criteria for both systems). In all three groups, males predominated, and UNEX identified 70 critical illnesses and 228 deaths in persons aged <50 years who were previously healthy and had specimens available for testing (i.e., the UNEX subgroup). Overall, during 2006–2011, the etiology for 29% of cases that had a specimen available for testing was determined. Cases with a respiratory syndrome were most commonly explained and sepsis/shock was the next most commonly explained syndrome. Examining the method of diagnosis in the explained UNEX subgroup cases revealed that, whereas most pathogens were detected by PCR (including both pathogen-specific PCR and 16S-PCR), other techniques such as culture and IHC also were very useful.
Correlating laboratory findings, clinical features, and pathologic evidence to establish a causal relationship allows for the detection of organisms that otherwise would likely be missed. However, death investigation as a surveillance tool is not without its challenges. One hurdle commonly encountered was the identification of potential pathogens that are not the primary cause of a syndrome or death. Another obstacle was the resource-intensive nature of the surveillance and additional testing and materials required of medical examiners, pathologists, and public health staffs and laboratories.