The annual incidence of sepsis is 750,000 cases in the United States, and this number increases by 9% each year.1,2 Although the mortality rate has declined to 18%,3 the absolute number of sepsis-related deaths is increasing. Accordingly, sepsis places a major burden on the United States health care system, with annual costs of $16.7 billion.4 Based on the Surviving Sepsis Campaign recommendations, the treatment of sepsis patients must concentrate on rapid control of the infectious focus, early administration of broad-spectrum antibiotics, as well as supportive and adjunctive therapies.5
Although the autopsy is one of the most important tools of quality assurance in critical care medicine,6 postmortem examination rates have continued to decrease during recent years.7 All medical specialities are involved in this decline, especially internal medicine, surgery and intensive care, all with current autopsy rates of <7%.8 While detailed analyses of the postmortem findings of critically ill cancer,9 trauma,10 cardiologic,11 and pediatric12 patients have been published, no such study has been performed in sepsis patients. Detailed knowledge of postmortem findings in these patients could not only improve our understanding and treatment of sepsis, but also provide direction for future research strategies.
This retrospective cohort study reviews the macroscopic postmortem examinations of 235 surgical intensive care unit (ICU) patients who died from sepsis or septic shock. Since this is an uncontrolled study, we did not intend to define macroscopic pathologies specific for patients dying from sepsis.
The medical database of a 12-bed general and surgical ICU in a tertiary university hospital was reviewed for patients who were admitted to the ICU between January 1, 1997, and September 30, 2006, because of sepsis/septic shock, or who developed sepsis/septic shock during their ICU stay and subsequently died from sepsis/septic shock. The study ICU receives patients after elective or emergency surgery, but occasionally treats surgical and nonsurgical patients with internal medical diseases.
According to the American College of Chest Physicians and the Society of Critical Care Medicine criteria,13 sepsis was defined as the presence of two or more signs of systemic inflammation in association with an infectious focus. Septic shock was defined as persistent hypotension (mean arterial blood pressure <70 mm Hg) which did not respond to fluid resuscitation and subsequently required either inotropic or vasopressor drugs in patients with sepsis. The clinical records of all patients were checked to verify that sepsis/septic shock persisted until death, and the results of the microbiological specimens of each patient were retrieved to confirm infection.
After identification of the patients from the institutional database, the autopsy registers of the Institute of Pathology and the Institute of Forensic Medicine, both at the Innsbruck Medical University, were searched for the final reports of the postmortem examinations. According to Austrian law,14 an autopsy should be conducted in all patients dying in the ICU. The retrospective study protocol was approved by the Ethics Committee of the Innsbruck Medical University.
As part of the clinical routine, patients with sepsis and septic shock were treated as follows: diagnosis of the septic focus was based on the patient’s history, clinical examination, microbiological cultures, and imaging studies. If a septic focus was amenable to mechanical evacuation, either surgical and/or percutaneous/endoscopic interventional techniques were applied. Broad-spectrum antibiotics (mostly ureidopenicillins, chinolones or carbapenems) were empirically administered and de-escalated after reception of microbiological culture results. Hemodynamic stabilization was attempted as early as possible using combined crystalloid/colloid-based fluid resuscitation as well as norepinephrine and milrinone as the first-line vasopressor and inotropic drugs. In cases of advanced shock states, epinephrine, hydrocortisone (up to 300 mg/d; since 1999) and/or arginine vasopressin (since 1998) were additionally administered. Mechanical ventilation was based on pressure-control and pressure-support modes, with the goal of maintaining peak inspiratory pressures <30 mbar. Continuous veno-venous hemofiltration was started early in the course of sepsis-associated renal failure. Since 2001, whenever technically possible, filtration rates of >35 mL/min were used. A balanced enteral/parenteral nutrition plan targeting early enteral feeding as well as stress ulcer and deep vein thrombosis prophylaxis was applied in all patients. Beginning in 2001, parenteral vitamin C, selenium and glutamine were routinely administered. Tight glucose control aiming at serum levels between 80 and 150 mg/dL was implemented in 2003. Patients were analgosedated using a continuous midazolam and/or sufentanil infusion as clinically indicated.
After death, bodies were transferred to a cooling chamber (temperature 3.8–4°C, relative humidity 85%) which is close to the department in the basement of the hospital. Only in rare exceptions did the time delay between death and transfer to this cooling chamber exceed 1 h. On the morning of the next working day, the corpses were transported either to the Institute of Pathology or the Institute of Forensic Medicine, where the autopsy subsequently took place. As part of the institutional routine, the postmortem examination was performed in all study patients as follows: A vertical incision from suprasternal notch to the symphysis was used to expose internal organs. The organs were removed in four blocks: 1) heart and lungs, 2) liver and gastrointestinal tract, 3) urogenital system, 4) brain. Afterwards, all organs were systematically examined for macroscopic pathologies. Internal examination of the extremities, spinal column and facial skull was performed only in the case of clinically suspected pathologies. The definitions of specific postmortem findings are presented in Table 1. Whenever macroscopically unclear lesions were detected, tissue samples were taken for histological analyses to either confirm or refute the macroscopic diagnosis. Therefore, only clear or histologically confirmed diagnoses were entered into the final autopsy record, which served as the basis for data documentation in this analysis.
The following variables of all study patients were extracted from the institutional database: gender, age, chronic disease status, classification of the American Society of Anesthesiologists,15 the Simplified Acute Physiology Score II,16 time of onset of sepsis, the source of infection, the pathogen type cultured, presence of sepsis or septic shock13; the number of failing organs, defined according to the Goris multiple organ dysfunction syndrome score (data supplement available at www.anesthesia-analgesia.org)17; need for continuous veno-venous hemofiltration; presence of acute respiratory distress syndrome, defined according to the American-European consensus conference on acute respiratory distress syndrome18; length of ICU stay, and the clinical cause of death as documented by a senior intensivist.
Autopsy reports of all study patients were searched for pathologies of the following organ systems: cardiovascular system, lungs, liver, kidneys and urinary tract, gastrointestinal tract, spleen, pancreas, and central nervous system. Pathologies of other organ systems were recorded separately. Postmortem findings of each organ system were entered into the database in a binary fashion. Additionally, the cause of death as documented by the pathologist was recorded for all study patients.
The SPSS software program (SPSS 12.0.1.; SPSS, Chicago, IL) was used for statistical analysis. Descriptive statistical methods were applied to evaluate the frequency of pathologies of single organ systems. In order to compare the frequency of organ pathologies between groups, a χ2 or Fisher’s exact test was used, as appropriate. P values <0.05 were considered as indicating statistical significance. Data are given as mean values ± sd, if not otherwise indicated.
During the study period, 5226 patients were admitted to the ICU. Four-hundred-fifteen patients suffered from sepsis and 442 from septic shock. The ICU morality was 32.3% (n = 277) for all sepsis/septic shock patients (sepsis, 11.8%, n = 49; septic shock, 51.6%, n = 228). An autopsy was performed in 256 of the 277 patients (92.4%). Clinical records, autopsy results, or results of microbiological specimens could be retrieved in 235 of the 277 patients (84.8%) (Tables 2 and 3). The mean storage time of corpses in the cooling chamber before the postmortem examination was 24 ± 20 h.
Pathologies of the cardiovascular system were detected in 126 patients (53.6%) (Table 4). Two of the five patients with endocarditis found on autopsy were admitted to the ICU because of this condition. In 211 patients (89.8%), pathologies of the lungs were observed (Table 5). Fifty-two of the 97 patients in whom pneumonia was found at autopsy were diagnosed as having pneumonia during their ICU stay. One-hundred-twenty-seven (54%), 112 (47.7%), 78 (33.2%), and 20 (8.5%) showed pathologies of the gastrointestinal tract, liver, spleen and pancreas, respectively (Table 6). Patients with biochemical evidence of hepatic dysfunction (increased bilirubin and/or transaminase plasma levels) showed a higher incidence of liver pathologies (27/79 [34.2%] vs 85/156 [54.5%], P = 0.004). Seven of the 20 patients with necrotizing pancreatitis at autopsy were admitted to the ICU because of this condition. In 141 patients (60%), pathologies of the kidneys and the urinary tract were described (Table 7). There was no difference in the incidence of renal pathologies between patients with and without continuous veno-venous hemofiltration (103/170 [60.6%] vs 38/65 [58.8%], P = 0.77). Pathologies of the central nervous system were detected in 44 patients (18.7%) (Table 8). Of the patients who were admitted to the ICU after cardiopulmonary resuscitation or with a primary cerebral pathology (n = 26), 2 were found to have brain edema in the autopsy.
In 180 patients (76.6%), the autopsy revealed a continuous septic focus. Eighty patients (34%) had 1 focus, 72 patients (30.6%) 2, 21 patients (8.9%) 3, and 7 patients (3%) ≥4 foci. The most frequent foci were pneumonia (41.3%), tracheobronchitis (28.9%), peritonitis (23.4%), uterine/ovarial necrosis (9.8% of female patients), intraabdominal abscesses (9.1%) and pyelonephritis (6%). Of the 71 patients admitted to the ICU because of sepsis/septic shock and treated ≥7 days, a continuous septic focus was observed in 63 (88.7%).
There was no difference in the frequency and pattern of postmortem findings between patients with refractory multiple organ failure and uncontrollable shock as the clinical cause of death. In view of the small sample size in the sepsis group (of 40 patients, 16 died without developing shock), no comparison of the postmortem diagnoses between patients with sepsis and septic shock was performed.
The most striking finding of this analysis was that approximately 80% of all study patients still had a septic focus at autopsy. It can be argued that detection of an infectious focus can be expected in patients who die early in the course of sepsis. However, even in patients who were admitted to the ICU because of sepsis and treated for ≥7 days, a sufficient time to control the source of infection, the incidence of a continuous septic focus in the autopsy was approximately 90%. In spite of the fact that immediate removal of the septic focus combined with antibiosis is the cornerstone of sepsis therapy,5,19 it appeared impossible to control the focus in the vast majority of our study patients, and this seems to have been the main cause of death.
A continuous septic focus continues to stimulate the immune system, finally leading to multiple organ failure and death.2 Although it sounds simple, and clinicians tend to look for more complex treatment strategies (e.g., adjuvant therapies), our results highlight that source control must be the key component in sepsis management.5,19 Accordingly, future research should concentrate on improving diagnostic and therapeutic strategies to eliminate the infectious source.
In our patient population, most continuous septic foci detected were located in the lungs, abdomen, female urogenital tract and kidneys. However, it is possible that other potential septic foci, such as intravascular devices or spontaneous bacteremia, cannot be detected on postmortem examination and were thus missed in this analysis. Moreover, some foci, such as the sinuses or the vertebral column, were only examined in the autopsy if there had been clinical evidence of an infection. Thus, the frequency of continuous septic foci may even have been under-estimated by this analysis. Our data also did not show whether the focus was unrecognized or the therapy was insufficient.
Of all organ pathologies, the most relevant for patient mortality seems to have been pathology of the cardiovascular system, since uncontrollable shock was the clinical cause of death in 25% of the study population and almost all patients had shock. Even though pulmonary pathologies were observed in approximately 90% of patients, pulmonary failure was a rare clinical cause of death. In our study, the most frequent cardiac pathology was myocardial ischemia. One-third of all study patients had a nonocclusive myocardial ischemia. In contrast, occlusive ischemia was observed only in a minority of patients (4.3%). A comparably low incidence of occlusive ischemia has been reported in perioperative myocardial infarction, whose predominant pathophysiologic feature is an imbalance of myocardial oxygen demand and supply.20 Furthermore, other myocardial pathologies occurring during the last hours before death, such as myocardial stunning or myocardial ischemia, cannot be reliably detected in the postmortem examination. Therefore, the actual frequency of myocardial pathologies may have been higher than suggested by our results.
In contrast to 13 patients with acute left ventricular dilation, 24 patients presented with acute dilation of the right ventricle, suggesting significant right ventricular dysfunction. However, only seven of them sustained pulmonary embolism, suggesting that in these patients either direct right ventricular dysfunction or pulmonary arterial hypertension played the dominant pathophysiologic role. A recent prospective echocardiographic survey of patients with septic shock indicated that ventricular dilation is only a rare (11%) and severe form of septic cardiomyopathy. Isolated or combined systolic and diastolic dysfunction was observed more frequently.21 Although functional variables by nature cannot be evaluated in the postmortem examination, our findings support current evidence that right ventricular dysfunction is a clinically under-estimated problem in critically ill patients.22
Another interesting finding of this study is that brain edema was observed in nearly 15% of patients dying from sepsis. Only two of these patients were admitted to the ICU because of a primary cerebral pathology or after cardiopulmonary resuscitation. Although severe hypotension may cause cerebral hypoperfusion and thus induce brain edema, recent evidence suggests that the pathophysiology of septic delirium may include not only functional but also anatomic components.23,24 Nonetheless, we cannot exclude the influence of terminal hypoxia and hypotension developing after withdrawal of life-sustaining therapy on the formation of brain edema in our study population.
Due to this study’s uncontrolled, retrospective approach, our findings cannot be considered to be specific for patients with sepsis, but may also be seen in critically ill patients dying from other pathologies. Whether these findings are specific for sepsis must be examined in a prospective, controlled trial. However, the aim of this analysis was to offer a cross-sectional overview of macroscopic postmortem findings observed in patients succumbing to sepsis.
There are clearly more limitations deserving consideration when interpreting the results of this analysis. Although the bodies had been cooled rapidly after death, it cannot be excluded that autolysis before autopsy significantly influenced the postmortem findings of our study. Another innate insufficiency of macroscopic postmortem studies is that diagnoses are based on macroscopic and not histological examinations. Even though histological analyses were performed whenever macroscopically unclear lesions were detected, the lack of universal histological examinations is a clear limitation of this analysis. It is possible that histological examinations in all study patients would have revealed more pathologies in certain organ systems, e.g., the kidneys particularly in patients on renal replacement therapy. Similarly, since postmortem diagnoses like pneumonia or tracheobronchitis were based on macroscopic instead of microbiological examinations, the incidence of pulmonary infections may have been over-estimated in our analysis. Furthermore, patients who actually had sepsis but were clinically misdiagnosed as not having sepsis were not included in our study. Omission of these patients could have had a relevant influence on the actual incidence of organ pathologies in patients dying from sepsis or septic shock. Functional problems (e.g., liver, coagulation, etc.) cannot be detected by the pathologist and were equally missed by this analysis. In particular, acute renal injury requiring renal replacement therapy did not have a clear macroscopic correlate. Moreover, since this study included critically ill patients mainly suffering from surgical pathologies, the results cannot be extrapolated to other ICU populations, such as internal or neurological ICU patients.
Finally, during the long study period, new therapeutic interventions were introduced and mortality decreased. However, it is unlikely that these changes influenced the results of our analysis, because no difference in the frequency and pattern of organ pathologies could be observed during the study period (data supplement available at www.anesthesia-analgesia.org).
The main clinical and postmortem causes of death in critically ill surgical patients succumbing to sepsis and septic shock were refractory multiple organ dysfunction syndrome and uncontrollable cardiovascular failure. Relevant postmortem findings explaining these results were a continuous septic focus in 80% and cardiac pathologies in 50% of patients. The most frequently affected organs were the lungs, abdomen, and urogenital tract. More diagnostic, therapeutic and scientific efforts should be launched to identify and control the infectious focus in patients with sepsis and septic shock.
1. Russell JA. Management of sepsis. N Engl J Med 2006;355: 1699–713
2. Annane D, Bellissant E, Cavaillon JM. Septic shock. Lancet 2005;365:63–78
3. Martin GS, Mannino DM, Eaton S, Moss M. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003;348:1546–54
4. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001;29:1303–10
5. Dellinger RP, Levy MM, Carlet JM, Bion J, Parker MM, Jaeschke R, Reinhart K, Angus DC, Brun-Buisson C, Beale R, Calandra T, Dhainaut JF, Gerlach H, Harvey M, Marini JJ, Marshall J, Ranieri M, Ramsay G, Sevransky J, Thompson BT, Townsend S, Vender JS, Zimmerman JL, Vincent JL; for the International Surviving Sepsis Campaign Guidelines Committee. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock: 2008. Crit Care Med 2008;36:296–327
6. Perkins GD, McAuley DF, Davies S, Gao F. Discrepancies between clinical and postmortem diagnoses in critically ill patients: an observational study. Crit Care 2003;7:129–32
7. Dimopoulos G, Piagnerelli M, Berré J, Salmon I, Vincent JL. Post mortem examination in the intensive care unit: still useful? Intensive Care Med 2004;30:2080–5
8. Loughrey MB, McCluggage WG, Toner PG. The declining autopsy rate and clinicians’ attitudes. Ulster Med J 2000;69:83–9
9. Pastores SM, Dulu A, Voigt L, Raoof N, Alicea M, Halpern NA. Premortem clinical diagnoses and postmortem autopsy findings: discrepancies in critically ill cancer patients. Crit Care 2007;11:R48
10. Ong AW, Cohn SM, Cohn KA, Jaramillo DH, Parbhu R, McKenney MG, Barquist ES, Bell MD. Unexpected findings in trauma patients dying in the intensive care unit: results of 153 consecutive autopsies. J Am Coll Surg 2002;194:401–6
11. Saad R, Yamada AT, Pereira da Rosa FH, Gutierrez PS, Mansur AJ. Comparison between clinical and autopsy diagnoses in a cardiology hospital. Heart 2007;93:1414–9
12. Cardoso MP, Bourguignon DC, Gomes MM, Saldiva PH, Pereira CR, Troster EJ. Comparison between clinical diagnoses and autopsy findings in a pediatric intensive care unit in Sao Paulo, Brazil. Pediatr Crit Care Med 2006;7:423–7
13. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, Cohen J, Opal SM, Vincent JL, Ramsay G; SCCM/ESICM/ACCP/ATS/SIS. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003;31: 1250–6
14. Tiroler Krankenananstaltsgesetz–Tir KAG. §37–Leichenöffnung (Obduktion). www.ris.bka.gv.at
Accessed January 17, 2008
15. American Society of Anesthesiologists (ASA). New classification of physical status. Anesthesiology 1963;24:111
16. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993;270:2957–63
17. Goris RJ, te Boekhorst TP, Nuytinck JK, Gimbrere JS. Multiple-organ failure. Generalized autodestructive inflammation? Arch Surg 1985;120:1109–15
18. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, Hudson L, Lamy M, LeGall JR, Morris A, Spragg R. Report of the American-European consensus conference on ARDS: definitions, mechanisms, relevant outcomes and clinical trial coordination. The Consensus Committee. Intensive Care Med 1994;20:225–32
19. Jimenez MF, Marshall JC. Source control in the management of sepsis. Intensive Care Med 2001;27:S49–S62
20. Priebe J. Perioperative myocardial infarction–aetiology and prevention. Br J Anaesth 2005;95:3–19
21. Etchecopar-Chevreuil C, Francois B, Clavel M, Pichon N, Gastinne H, Vignon P. Cardiac morphological and functional changes during early septic shock: a transesophageal echocardiographic study. Intensive Care Med 2008;34:250–6
22. Cecconi M, Johnston E, Rhodes A. What role does the right side of the heart play in circulation? Crit Care 2006;10 (Suppl 3):S5
23. Ebersoldt M, Sharshar T, Annane D. Sepsis-associated delirium. Intensive Care Med 2007;33:941–50
24. Sharshar T, Annane D, de la Grandmaison GL, Brouland JP, Hopkinson NS, Françoise G. The neuropathology of septic shock. Brain Pathol 2004;14:21–33