Comparison of Premortem Clinical Diagnoses in Critically Ill Patients and Subsequent Autopsy Findings

doi:10.4065/75.6.562

Mayo Clinic Proceedings

Volume 75, Issue 6, June 2000, Pages 562-567

https://doi.org/10.4065/75.6.562 Get rights and content

Objective

To determine whether our practice of requesting an autopsy for patients who die in the medical intensive care unit (MICU) continues to be a valid approach to obtain clinically and educationally relevant findings.

Methods:

In this retrospective study conducted in an adult MICU population of a university hospital, the clinical diagnoses and postmortem major diagnoses of 100 patients who died in 1996 (autopsy rate of 93%) were compared.

Results:

Eighty-one percent of the clinical diagnoses were confirmed at autopsy. In 16%, autopsy findings revealed a major diagnosis that, if known before death, might have led to a change in therapy and prolonged survival (class I missed major diagnoses). The most frequent class I missed major diagnoses were fungal infection, cardiac tamponade, abdominal hemorrhage, and myocardial infarction. Another 10% of autopsies revealed a diagnosis that, if known before death, would probably not have led to a change in therapy (class II error).

Conclusions:

Autopsy remains an important tool for education and quality control. In contrast with historical series of 1 to 2 decades ago, there is a clear shift in the type of class I missed major diagnoses toward opportunistic infections. Bedside-applicable techniques such as electrocardiography with supplemental posterior leads, echocardiography, and meticulous abdominal ultrasonography might improve the outcome in selected MICU patients.

Section snippets

Patients

The patients included in this retrospective study were hospitalized in the 17-bed MICU of the 1500-bed University Hospital of the Catholic University of Leuven, Belgium. Approximately 95% of the patients admitted to this unit are medical patients, and 5% are surgical (postoperative) patients. The unit serves the need for intensive care of patients hospitalized mainly in internal medicine beds, including 3 organ transplant units, and of nonsurgical patients admitted via the emergency department.

Patients

Of the 100 patients included in the study, 40% were women and 60% were men, 98% were medical patients, 14% were considered to be immunosuppressed, and 40% were admitted from hospital wards, 48% from the emergency department, 10% from critical care units in other hospitals, and 2% from the operating room. The mean age was 63.6 years (range, 17–90 years). The mean ± SD APACHE II score was 25±9. The average length of stay in the MICU was 11.8 days (range, 1–99 days); 46 patients died during the

Discussion

Despite decreasing autopsy rates, postmortem examination is still considered an important tool in quality control and in medical education.4, 5, 6 This study aimed to validate our continued practice of requesting autopsies after the deaths of ICU patients. The numbers of missed major diagnoses in our study are comparable to those published in previous studies.3, 8, 9, 11, 15, 16, 17 Several factors can influence the type and frequency of missed diagnoses.

Selection of patients may explain

References (27)

TS Welsh et al.
The role of postmortem examination in medical education
Mayo Clin Proc
(1998)
DK Kleiner et al.
Necropsy as a research method in the age of molecular pathologv
Lancet
(1995)
G Stevanovic et al.
Correlation of clinical diagnoses with autopsy findings: a retrospective study of 2.145 consecutive autopsies
Hum Pathol
(1986)
B Veress et al.
A retrospective analysis of clinical diagnoses and autopsy findings in 3.042 cases during two different time periods
Hum Palhol
(1994)
SJ McPhee et al.
Autopsy: moribund art or vital science?
Am J Med
(1985)
RC Cabot
Diagnostic pitfalls identified during a studv of three thousand autopsies
JAMA
(1912)
WC Roberts
The autopsy: its decline and a suggestion for its revival
N Engl J Med
(1978)
L Goldman et al.
Hie value ofthe autopsy in three medical eras
N Engl J Med
(1983)
AG Scottolini et al.
The autopsy in clinical quality control
JAMA
(1983)
HM Cameron et al.
A prospective study of 1152 hospital autopsies. II: analysis of inaccuracies in clinical diagnoses and their significance
J Pathol
(1981)

RM Battle et al.

Factors influencing discrepancies between premortem and postmortem diagnoses

JAMA

(1987)

T Kireher et al.

The autopsy as a measure of accuracy of the death certificate

N Eingl J Med

(1985)

IJ Schatz

Internal medicine residency program violations of the minimum autopsy rate [letter]

JAMA

(1995)

Cited by (139)

Analysis of the major clinical-pathological discrepancies in autopsies of adults in a tertiary referral hospital between 2008-2017
2021, Revista Espanola de Patologia
Analizar el diagnóstico clínico ante mortem y el resultado de la autopsia clínica para identificar hallazgos inesperados de alta relevancia clínica.
Estudio retrospectivo observacional realizado en el Hospital Universitario Central de Asturias en el que se examinaron las autopsias de adulto y sus respectivas historias clínicas en la década 2008-2017 y se clasificaron siguiendo la Clasificación de Goldman y por servicios de procedencia.
De las 1320 autopsias se incluyeron finalmente 694 (52,6%). De ellas, hasta en un 57,6% se observaron discrepancias, la mayoría menores (39,3%). En 63 autopsias se identificaron discrepancias de tipo I (9,1%) siendo las neoplasias malignas la principal patología observada (57,1%), especialmente de origen gastrointestinal con cerca del 28% de los casos. Tras el cáncer, fueron las enfermedades infecciosas (23,8%) y después el embolismo pulmonar (15,9%). Por otra parte, 64 autopsias fueron clasificadas como discrepancias tipo II (9,2%) identificando mayoritariamente al infarto como principal causante (37,5%), sobre todo el infarto agudo de miocardio (18 casos), seguido de broncoaspiraciones (18,7%), CID (15,6%), hemorragias masivas (9,4%) y otros procesos. Se observó que tanto la UVI como el Servicio de Medicina Interna fueron los responsables del mayor número de discrepancias mayores (tipo I y II), representando cerca del 45% para discrepancias tipo I y algo más del 56% para discrepancias tipo II.
Las autopsias son una herramienta esencial de calidad hospitalaria para vislumbrar errores ante mortem significativos. El Hospital Universitario Central de Asturias muestra una proporción de discrepancias mayores (18,3%) en rangos similares a los mejores hospitales de referencia mundial.
To compare and contrast clinical diagnoses with autopsy findings in order to identify unexpected, relevant discrepancies.
A retrospective observational study of the revision of autopsies of adults and their respective medical records in order to classify them according to referral department and Goldman's classification was carried out at the Central University Hospital of Asturias between 2008-2017.
694 (52.6%) of 1320 autopsies were included in the study. Discrepancies were observed in 57.6% of cases, although the majority (39.3%) were minor. Type I discrepancies were identified in 63 autopsies (9.1%); malignant neoplasms being the main pathology observed (57.1%), mainly of gastrointestinal origin (about 28%). The second most common discrepancy was found in cases of infectious diseases (23.8%) followed by pulmonary embolism (15.9%). 64 autopsies were classified as type II discrepancies (9.2%), with myocardial infarct the most common (37.5%), especially acute myocardial infarction (18 cases), followed by bronchoaspirations (18.7%), DIC (15.6%), massive haemorrhages (9.4%) and other conditions. It was considered that both the ICU and the Internal Medicine Service were responsible for the largest number of major discrepancies (type I and II), accounting for about 45% of type I and slightly more than 56% for type II.
Autopsies are an essential means of identifying ante-mortem clinical errors. The incidence of major discrepancies in the Central University Hospital of Asturias (18.3%) is comparable to that of leading hospitals worldwide.
Invasive aspergillosis in patients admitted to the intensive care unit with severe influenza: a retrospective cohort study
2018, The Lancet Respiratory Medicine
Invasive pulmonary aspergillosis typically occurs in an immunocompromised host. For almost a century, influenza has been known to set up for bacterial superinfections, but recently patients with severe influenza were also reported to develop invasive pulmonary aspergillosis. We aimed to measure the incidence of invasive pulmonary aspergillosis over several seasons in patients with influenza pneumonia in the intensive care unit (ICU) and to assess whether influenza was an independent risk factor for invasive pulmonary aspergillosis.
We did a retrospective multicentre cohort study. Data were collected from adult patients with severe influenza admitted to seven ICUs across Belgium and The Netherlands during seven influenza seasons. Patients were older than 18 years, were admitted to the ICU for more than 24 h with acute respiratory failure, had pulmonary infiltrates on imaging, and a confirmed influenza infection based on a positive airway PCR test (influenza cohort). We used logistic regression analyses to determine if influenza was independently associated with invasive pulmonary aspergillosis in non-immunocompromised (ie, no European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group [EORTC/MSG] host factor) influenza-positive patients (influenza case group) compared with non-immunocompromised patients with severe community-acquired pneumonia who had a negative airway influenza PCR test (control group).
Data were collected from patients admitted to the ICU between Jan 1, 2009, and June 30, 2016. Invasive pulmonary aspergillosis was diagnosed in 83 (19%) of 432 patients admitted with influenza (influenza cohort), a median of 3 days after admission to the ICU. The incidence was similar for influenza A and B. For patients with influenza who were immunocompromised, incidence of invasive pulmonary aspergillosis was as high as 32% (38 of 117 patients), whereas in the non-immunocompromised influenza case group, incidence was 14% (45 of 315 patients). Conversely, only 16 (5%) of 315 patients in the control group developed invasive pulmonary aspergillosis. The 90-day mortality was 51% in patients in the influenza cohort with invasive pulmonary aspergillosis and 28% in the influenza cohort without invasive pulmonary aspergillosis (p=0·0001). In this study, influenza was found to be independently associated with invasive pulmonary aspergillosis (adjusted odds ratio 5·19; 95% CI 2·63–10·26; p<0·0001), along with a higher APACHE II score, male sex, and use of corticosteroids.
Influenza was identified as an independent risk factor for invasive pulmonary aspergillosis and is associated with high mortality. Future studies should assess whether a faster diagnosis or antifungal prophylaxis could improve the outcome of influenza-associated aspergillosis.
None.
Autopsy-detected diagnostic errors over time in the intensive care unit
2018, Human Pathology
We evaluate the evolution over time of discrepancies between clinical diagnoses and postmortem findings in critically ill patients and assess the factors associated with these discrepancies. We conducted a prospective study of all consecutive patients who underwent autopsy in a medical-surgical intensive care unit (ICU) between January 2008 and December 2015. Among 7655 patients admitted to our ICU, 671 (8.8%) died. Clinical autopsy was performed in 215 (32%) patients. Major missed diagnoses were noted in 38 patients (17.7%). Eighteen patients (8.4%) had class I discrepancies, and 20 patients (9.3%) had class II discrepancies. The most frequently missed diagnoses were invasive aspergillosis, intestinal ischemia, myocardial infarction, cancer, and intra-abdominal abscesses. We did not find a statistically significant correlation between any premortem factor, including age, sex, severity of illness, length of hospital stay before ICU admission, length of ICU stay before death, duration of mechanical ventilation, or admitting unit, and the level of agreement between clinical and pathological diagnosis. In the last decades, the discrepancies between clinical and autopsy diagnoses persisted despite advances in medical skills and technology. Specific clinical entities such as invasive aspergillosis, mesenteric ischemia, myocardial infarction, intra-abdominal abscesses, and neoplastic diseases remain a diagnostic challenge in critically ill patients. Clinical level of diagnostic certainty does not increase with specific premortem characteristics.
Recovery from serious fungal infections should be realisable for everyone
2017, The Lancet Infectious Diseases
Comparison of Clinical and Pathological Diagnoses in Cats and Dogs
2017, Journal of Comparative Pathology
Citation Excerpt :
The scheme for comparison of diagnoses in the present study, focusing on the gain of information from necropsy examination, differs slightly from that commonly used in human medicine. Goldman proposed a classification for the comparison of diagnoses in 1983 that was modified by Battle et al. (1987) and used in numerous studies for the comparison of diagnoses in human medicine (Mort and Yeston, 1999; Roosen et al., 2000; Sonderegger-Iseli et al., 2000; Twigg et al., 2001; Perkins et al., 2003; Maris et al., 2007; Thurnheer et al., 2009; Tejerina et al., 2012). Goldman's classification is based on the question of whether knowledge of the discrepancy would have led to a change in patient management or survival.
The comparison of clinical ante-mortem and pathological post-mortem diagnoses is a prerequisite for quality control, but is rarely done in veterinary medicine. This study reports the occurrence and concurrence of clinical and pathological diagnoses linked to death in 1,000 cats and 1,000 dogs examined and subjected to necropsy examination at the University of Zurich, Switzerland. Potential factors influencing the correlation between diagnoses were examined retrospectively. In 5.8% of cats and 5.2% of dogs no diagnosis was made; in 2.6% and 3.8% of cases only a clinical, and in 17.8% and 11.2%, respectively, only a pathological diagnosis was available. Of the 73.8% of cats and 79.8% of dogs with both diagnoses present, 38.3% and 36.2% were in agreement, while there was disagreement in 17.9% and 16.0%, respectively. The remaining cases (43.8% and 47.8%) had different levels of further diagnostic procedures following necropsy examination. In both species, the manner of death, the clinical discipline submitting the animal for necropsy examination and the quality of the necropsy submission request, as well as the timespan between death and necropsy examination in dogs, proved to influence the concurrence between diagnoses. In contrast, the organ system affected and the type of disease entity were, for both species, the most influential factors in the concurrence of diagnoses. Therefore, in veterinary medicine, even in times of improving diagnostic abilities, necropsy examination still reveals important information for quality control and education.
Invasive aspergillosis in critically ill patients
2018, Revista Iberoamericana de Micologia
Los pacientes críticos sin una grave inmunosupresión es una población en la que la aspergilosis invasiva (AI) es una enfermedad en alza. El tratamiento crónico con corticoides, la enfermedad pulmonar obstructiva crónica y la cirrosis hepática son factores de riesgo repetidamente identificados en las series publicadas. No obstante, debido a la inespecificidad de los síntomas y signos en el paciente crítico y a la relativa baja capacidad diagnóstica de las pruebas complementarias, el diagnóstico de la AI supone un reto para el especialista en medicina intensiva. La aplicación de algoritmos diagnósticos adaptados al paciente crítico, cuya activación dependerá del aislamiento de Aspergillus en una muestra respiratoria, es la metodología diagnóstica más eficaz en esta población. Entre los elementos diagnósticos, la determinación de galactomanano en el líquido broncoalveolar es la prueba diagnóstica que ha demostrado más utilidad. Una vez establecida la sospecha el tratamiento debe iniciarse precozmente. El voriconazol, la anfotericina B y el isavuconazol son los tratamientos más eficaces. Aunque el voriconazol y la anfotericina B son los fármacos con mayores evidencias científicas, adolecen de problemas con relación a efectos adversos y dificultades farmacocinéticas. Por ello, el isavuconazol, que ha demostrado una elevada eficacia y seguridad en otras poblaciones, supone una potencial alternativa de extremado interés para el paciente crítico.
Critically ill patients without severe immunosuppression make up a population in which invasive aspergillosis (IA) has been identified as an emergent pathology. Chronic treatment with corticosteroids, chronic obstructive pulmonary disease, and liver cirrhosis are repeatedly identified risk factors. However, due to the non-specificity of the symptoms and signs in the critical patient, and the relative low diagnostic capacity of the complementary tests, the diagnosis of the IA is a challenge for the specialist in critical care medicine. The application of diagnostic algorithms adapted to critical patients, in whom activation will depend on the isolation of Aspergillus in a respiratory specimen, is the most efficient diagnostic methodology in this population. Among the diagnostic approaches, the determination of galactomannan in bronchoalveolar fluid is the most useful diagnostic test. Once the suspicion is established, treatment should be started as soon as possible. Voriconazole, amphotericin B, and isavuconazole are the most effective treatments. Although voriconazole and amphotericin B are the drugs with the most scientific evidence, they are related with adverse effects and pharmacokinetic difficulties. Therefore, isavuconazole, which has shown high efficacy and safety in other populations, is a potential alternative of great interest for critically ill patients.

View all citing articles on Scopus

View full text

Original ArticleComparison of Premortem Clinical Diagnoses in Critically Ill Patients and Subsequent Autopsy Findings

Objective

Methods:

Results:

Conclusions:

Section snippets

Patients

Patients

Discussion

Mayo Clin Proc

Lancet

Hum Pathol

Hum Palhol

Am J Med

Diagnostic pitfalls identified during a studv of three thousand autopsies

JAMA

The autopsy: its decline and a suggestion for its revival

N Engl J Med

Hie value ofthe autopsy in three medical eras

N Engl J Med

The autopsy in clinical quality control

JAMA

A prospective study of 1152 hospital autopsies. II: analysis of inaccuracies in clinical diagnoses and their significance

J Pathol

Factors influencing discrepancies between premortem and postmortem diagnoses

JAMA

The autopsy as a measure of accuracy of the death certificate

N Eingl J Med

Internal medicine residency program violations of the minimum autopsy rate [letter]

JAMA

Original Article
Comparison of Premortem Clinical Diagnoses in Critically Ill Patients and Subsequent Autopsy Findings