Septic shock : providing early warnings through multivariate logistic regression models

Early goal-directed therapy (EGDT) in severe sepsis and septic shock has shown to provide substantial benefits in patient outcomes. However, these preventive therapeutic interventions are contingent upon an early detection or suspicion of the underlying septic etiology. Detection of sepsis in the early stages can be difficult, as the initial pathogenesis can occur while the patient is still displaying normal vital signs. This study focuses on developing an early warning system (EWS) to provide clinicians with a forewarning of an impending hypotensive crisis-thus allowing for EGDT intervention. Research was completed in three main stages: (1) generating an annotated septic shock dataset, (2) constructing multivariate logistic regression EWS models using the annotated dataset, and (3) testing the EWS models in a forward, causal manner on a random cohort of patients to simulate performance in a real-life ICU setting. The annotated septic shock dataset was created using the Multi-parameter Intelligent Monitoring for Intensive Care II (MIMIC II) database. Automated pre-annotations were generated using search criteria designed to identify two patient types: (1) sepsis patients who do not progress to septic shock, and (2) sepsis patient who progress to septic shock. Currently, manual review by expert clinicians to verify the pre-annotations has not been completed. Six separate EWS models were constructed using the annotated septic shock dataset. The multivariate logistic regression EWS models were trained to differentiate between 107 high-risk sepsis patients of whom 39 experienced a hypotensive crisis and 68 who remained stable. The models were tested using 7-fold cross validation; the mean area under the receiver operating characteristic (ROC) curve for the best model was 0.940 ± 0.038. The EWS models were then tested in a forward, casual manner on a random cohort of 500 ICU patients to mimic the patients' stay in the unit. The model with the highest performance achieved a sensitivity of 0.85 and a positive predictive value (PPV) of 0.70. Of the 35 episodes of hypotension despite fluid resuscitation present in the random patient dataset, the model provided early warnings for 29 episodes with a mean early warning time of 582 ± 355 minutes. Thesis Supervisor: Roger G. Mark, M.D., Ph.D. Title: Distinguished Professor in Health Sciences and Technology Professor of Electrical Engineering

[1]  K. Wood,et al.  Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock* , 2006, Critical care medicine.

[2]  Douglas E. Lake,et al.  Heart rate characteristics monitoring for neonatal sepsis , 2006, IEEE Transactions on Biomedical Engineering.

[3]  G. Sakorafas,et al.  Septic shock; current pathogenetic concepts from a clinical perspective. , 2005, Medical science monitor : international medical journal of experimental and clinical research.

[4]  M. Langer,et al.  The Italian SEPSIS study: Preliminary results on the incidence and evolution of SIRS, sepsis, severe sepsis and septic shock , 1995, Intensive Care Medicine.

[5]  G.D. Clifford,et al.  The annotation station: an open-source technology for annotating large biomedical databases , 2004, Computers in Cardiology, 2004.

[6]  G.D. Clifford,et al.  An open-source, interactive Java-based system for rapid encoding of significant events in the ICU using the unified medical language system , 2004, Computers in Cardiology, 2004.

[7]  Mitchell M. Levy,et al.  2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference , 2003, Intensive Care Medicine.

[8]  R. Hotchkiss,et al.  The pathophysiology and treatment of sepsis. , 2003, The New England journal of medicine.

[9]  Anand Kumar,et al.  Clinical review: Myocardial depression in sepsis and septic shock , 2002, Critical care.

[10]  R G Mark,et al.  MIMIC II: a massive temporal ICU patient database to support research in intelligent patient monitoring , 2002, Computers in Cardiology.

[11]  S. Zeerleder,et al.  The endothelium in sepsis: Source of and a target for inflammation , 2001, Critical care medicine.

[12]  Derek C. Angus,et al.  Epidemiology of sepsis: An update , 2001, Critical care medicine.

[13]  J. Pugin,et al.  Normal responses to injury prevent systemic inflammation and can be immunosuppressive. , 2001, American journal of respiratory and critical care medicine.

[14]  M. P. Griffin,et al.  Toward the early diagnosis of neonatal sepsis and sepsis-like illness using novel heart rate analysis. , 2001, Pediatrics.

[15]  B. Holzmann,et al.  Sepsis after major visceral surgery is associated with sustained and interferon-gamma-resistant defects of monocyte cytokine production. , 2000, Surgery.

[16]  K. Reinhart,et al.  Treating patients with severe sepsis. , 1999, The New England journal of medicine.

[17]  T. Yamashita,et al.  Change in the ratio of interleukin-6 to interleukin-10 predicts a poor outcome in patients with systemic inflammatory response syndrome. , 1999, Critical care medicine.

[18]  N. Minamino,et al.  Increased plasma levels of adrenomedullin in patients with systemic inflammatory response syndrome. , 1999, American journal of respiratory and critical care medicine.

[19]  P. Puolakkainen,et al.  Extracellular phospholipases A2 in relation to systemic inflammatory response syndrome (SIRS) and systemic complications in severe acute pancreatitis. , 1999, Pancreas.

[20]  J. Vincent,et al.  Has the mortality of septic shock changed with time. , 1998, Critical care medicine.

[21]  T. Staudinger,et al.  Time course of immunological markers in patients with the systemic inflammatory response syndrome: evaluation of sCD14, sVCAM-1, sELAM-1, MIP-1 alpha and TGF-beta 2. , 1998, European journal of clinical investigation.

[22]  D. Bates,et al.  Epidemiology of sepsis syndrome in 8 academic medical centers. , 1997, JAMA.

[23]  J. Vincent,et al.  Dear SIRS, I'm sorry to say that I don't like you... , 1997, Critical care medicine.

[24]  M. Newport,et al.  Variation in the tumor necrosis factor-alpha gene promoter region may be associated with death from meningococcal disease. , 1996, The Journal of infectious diseases.

[25]  T. Evans,et al.  The role of macrophages in septic shock. , 1996, Immunobiology.

[26]  D. Pittet,et al.  The natural history of the systemic inflammatory response syndrome (SIRS). A prospective study. , 1995, JAMA.

[27]  S. Lemeshow,et al.  A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. , 1993, JAMA.

[28]  W. Knaus,et al.  Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. , 1992, Chest.

[29]  Increase in National Hospital Discharge Survey rates for septicemia--United States, 1979-1987. , 1990, MMWR. Morbidity and mortality weekly report.

[30]  D. Hosmer,et al.  Applied Logistic Regression , 1991 .

[31]  E. Draper,et al.  APACHE II: A severity of disease classification system , 1985, Critical care medicine.

[32]  William R. Swartout,et al.  Rule-based expert systems: The mycin experiments of the stanford heuristic programming project , 1985 .

[33]  G. Liljestrand,et al.  Vergleichende Bestimmungen des Minutenvolumens des Herzens beim Menschen mittels der Stickoxydulmethode und durch Blutdruckmessung , 1928 .