Bioterrorism: The Use of Information Technologies and Decision Support Systems Dena M. Bravata, M.D., M.S. Policy Analysis MS&E 290 Case Presentation: Day 1 (11/16) Ottilie Lundgren, a 94 year old widow, living in Oxford, Connecticut presented with a 3-day history of fever, cough, and myalgias. Her physical examination showed an elderly woman with a fever (102.3F), elevated heart rate (118). Her chest x-ray and initial laboratory tests were all negative. She was admitted to the hospital for dehydration and an infection of the kidney tract. Case Presentation: Days 2-4 Hospital Day 2-3 (11/17-11/18) Labs: Bacteria found in her blood and urine but not yet identified Mrs. Lundgren had progressive respiratory distress and confusion. Repeat CXR: fluid around the left lung with the suggestion of pneumonia Hospital Day 4 (11/19) A sample of fluid from around lung taken for culture. The Connecticut Department of Public Health was notified by the hospital laboratory of blood culture results Case Presentation: Day 5-6 Hospital Day 5 (11/19) She was transferred to ICU on a ventilator Blood culture isolates identified as Bacillus anthracis at the State public health laboratory Hospital Day 6 (11/21) CDC confirms blood culture isolates as B. anthracis, molecular typing and susceptibility patterns identical to recent cases Ottilie Lundgren’s condition continued to deteriorate and she died. Current Cases of Bioterrorism-related Anthrax Confirmed Suspected Cutaneous 7 5 Inhalational 10* 0 Total 17 5 * 4 Deaths associated with inhalational anthrax Biothreat Agents • • • • • • Variola major (smallpox) Bacillus anthracis (anthrax) Yersinia pestis (plague) Clostridium botulinum toxin (botulism) Francisella tularensis (tularemia) Filoviruses (Ebola hemorrhagic fever and Marburg hemorrhagic fever) • Arenaviruses (Lassa fever, Junin/Argentine hemorrhagic fever) and related viruses The Project The objective of our project was to synthesize the evidence on information technologies and decision support systems (IT/DSSs) that may serve the information needs of clinicians and public health officials in the event of a bioterrorist attack. Methodologic Challenges of this Project • Conceptualizing the problem – Complex topic, no single effect size • Finding the literature – Not limited to peer-reviewed medical literature – Government documents – Web-based sources • Evaluating the quality of the evidence – No single, published quality scale relevant The Methodologic Approach • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence The Method • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence 5 Key Questions • What are the information needs of clinicians and public health officials in the event of a bioterrorist attack? • What are the criteria by which IT/DSS should be evaluated for usefulness in a bioterrorist event? • What IT/DSSs are available for detection and diagnosis, management, surveillance, and communication? 5 Key Questions Continued • What is the quality of the evidence about IT/DSS usefulness? • If no existing IT/DSSs meet the information needs of clinicians and public health officials, what considerations are important in the design of future IT/DSSs to support response to bioterrorism events? The Method • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence Structure of the Conceptual Framework Time period: 1 Time period: 3 Decision maker: Clinicians Decision maker: Clinicians Events: associated with the initial cases Time period: 2 Decision maker: Public health officials Events: associated with the initial cases Events: associated with subsequent cases Diagram of the Conceptual Framework Exposure Clinical Syndrome Exposure Infection Status (unobserved) Surveillance Alert Susceptible Diagnosis Susceptible Diagnosis Test Result Lives Saved Management Clinical Syndrome Infection Status (unobserved) Test Result Lives Saved Management Prevention Prevention Report Report Surveillance Report Outbreak Investigation Epidemiologic Control Surveillance Alert Time period 1 Time period 2 Time period 3 Time period 2: Surveillance Systems Detect Potential Events Healthcare provider Reports Coroner Reports Foodborne Disease Reports Veterinarian Reports Laboratory Reports Surveillance Report Detector Data Outbreak Investigation School/Work Absenteeism Data EMT Data Pharmacy Sales Data Hospital Discharge Data Surveillance systems receive data from a variety of sources and provide reports to public health authorities. Epidemiologic Control Surveillance Alert Lives Saved Role for Information Technologies and Decision Support Exposure Clinical Syndrome Exposure Infection Status (unobserved) Surveillance Alert Susceptible Diagnosis Susceptible Diagnosis Test Result Lives Saved Management Clinical Syndrome Infection Status (unobserved) Test Result Lives Saved Management Prevention Prevention Report Report Surveillance Report Outbreak Investigation Epidemiologic Control Surveillance Alert Decisions marked in blue indicate those that can be affected by decision support systems and arrows marked in blue indicate processes in which information technologies could play a role. The Method • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence Task Decomposition • Formal framework for specifying, documenting, and evaluating the data that should be contained within the knowledge base of a DSS in order for it to serve its purpose. • Literature review to define tasks: – – – – U.S. Offensive weapons program Known bioterrorist events Tabletop exercises Infectious disease outbreaks • 5 Tasks – – – – – Surveillance Diagnosis and detection Management Prevention Communication Example: Task Decomposition: Surveillance Subtask Key Concepts Data Requirement Collect surveillance data Continual, timely collection of sensitive and specific data from multiple sources for early detection of a bioterrorist attack •Environmental detectors •Pharmacy sales data •School & work absenteeism data •Clinicians’ reports •Laboratory reports Analyze surveillance data Timely analyses and presentation to public health decision makers •Baseline information for each data source to calculate expected trends over time •Threshold information for each data source to know when an outbreak has occurred The Method • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence Literature Searches • Literature Sources – Peer-reviewed articles and government documents from databases • Medline, GrayLit, National Technical Information Service, Catalog of U.S. Government Publications – Government documents from websites – Web-based information • Search Strategies – Professional librarians for peer-reviewed and government documents – Copernic 2001 metasearch engine for web-based information • Search Terms: same for our Government document and Copernic searches The Method • Identify key research questions • Develop a conceptual model • Define the tasks IT/DSSs must perform to meet the information needs of clinicians and public health officials • Perform literature searches • Evaluate the evidence Results of Literature Searches • Reviewed 16,751 citations and 8,620 websites • 251 articles and 41 websites met inclusion criteria • Double abstractions (blinded to study author) for all peer-reviewed articles and single abstractions for web-based information • Total of 204 systems – – – – – – 52 detection systems 23 diagnostic systems 14 management systems 88 surveillance systems 27 communication systems 7 systems that integrate surveillance, communication, and command and control functions. Results • Generally: few clinically evaluated systems • No IT/DSSs for Diagnosis or Management have been developed specifically for bioterrorism • Systems for Detection, Surveillance, and Communication have been developed for bioterrorism – Some show considerable promise but almost none has been evaluated for its sensitivity, specificity, or timeliness Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems Detection Systems: Overview • Developed for Military need to be adapted to civilian use • Promising projects include: – PROTECT (Program for the Response Options and Technology Enhancements for Chemical/Biological Terrorism) • Detectors set up in subways, airports, government buildings • Developing methods for signal-noise-processing – LEADERS (Lightweight Epidemiology and Advanced Detection and Emergency Response System) • Integrates detector data with other surveillance data • Particularly for event-based surveillance Detection Systems: collection and identification BioCaptureTM • Only collection system clinically evaluated • 50-125% collection efficiency of other devices • Used by first responders BioThreatAlert (BTA) Strips • Antigen/Antibody system available for a limited number of agents • Sensitivity and Specificity not evaluated • Used by first responders or clinicians Conclusions: Systems have not been clinically evaluated, sensitivity and specificity poorly characterized, can only test one sample at a time, tests not available for many worrisome agents (e.g., smallpox) Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems General Diagnostic Systems: Overview • Purpose: Clinician enters patient information (usually manually) and system provides a differential diagnosis • Examples: DXplain, Iliad, QMR (newer handheld versions now available) • Clinical evaluations: the differential diagnoses provided are highly dependent on descriptors entered Conclusions: Have rarely been shown to improve patient outcomes in general, none has been evaluated for diagnostic capability for bioterrorism-related diseases Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems Management Systems: Overview • Generally: Depend on electronic medical record to derive patient-specific recommendations • Most commonly for antibiotic recommendations for hospitalized patients Management Systems Example: HELP system at LDS hospital in Salt Lake City –When patients present to the ED, HELP’s databases are queried every 10 minutes for any new clinical information on the patient. Uses this information to calculate the probability of pneumonia. –Specificity 92%, PPV 15.1%, NPV 99.9% Conclusions: No evidence for their usefulness in a bioterrorist event; would have to incorporate new guidelines for biothreat agent diagnosis and management; typically require EMR. Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems Data Sources for Biosurveillance Earlier Detection Data Later Detection Data School & Work Absenteeism Phone Triage Nurses Sentinel Physicians Pharmacies (Rx) Pharmacies (OTC) Environmental detectors Emergency Department 911 Calls Laboratory Hospital admissions & discharge Surveillance Systems • • • • • • • • Syndromal surveillance reports (7) Reports from clinicians (6) Influenza-related data (10) Systems for laboratory and antimicrobial resistance data (23) Systems for nosocomial infections (15) Food-borne illnesses (10) Zoonotic illnesses (5) Other types of surveillance data (12) Syndromal Surveillance • Syndromes associated with biothreat agents – – – – – – Flu-like illness Acute respiratory distress Gastrointestinal symptoms Febrile, hemorrhagic syndromes Fever and rash Fever and mental status change ® HealthBuddy ESSENCE Integrated Systems Data Analysis and Presentation Results • • • • • Detection Systems General Diagnostic Systems Management Systems Surveillance Systems Communication Systems Communication Systems: Overview • Email systems for communication between patients and clinicians • Web-based secure networks linking branches of the public health • Radio/Microwave-based systems for linking emergency personnel in the field with EDs • Alert systems to notify clinicians of abnormal laboratory tests Communication Systems for Bioterrorism Local Public Health Communication Systems for Bioterrorism WHO CDC State Public Health Local Public Health Communication Systems for Bioterrorism WHO CDC FBI Zoo State Public Health Food Inspectors Schools HazMat Local Public Health Pharmacies Police Laboratories Fire Hospitals & Nursing Facilities The Public Clinicians Coroner & Medical Examiner The Media Communication Systems for Bioterrorism WHO CDC FBI * Zoo State Public Health Food Inspectors Schools * HazMat Pharmacies Local Public Health * Police Laboratories Fire * * The Public * * Hospitals & Nursing Facilities Clinicians Coroner & Medical Examiner The Media * * Communication Systems for Bioterrorism WHO CDC FBI * Zoo State Public Health Food Inspectors S * HazMat Schools S Pharmacies Local Public Health Police * S Laboratories Fire S * * The Public * * Hospitals & Nursing Facilities Clinicians Coroner & Medical Examiner The Media * * Communication Systems for Bioterrorism WHO CDC FBI * Zoo State Public Health Food Inspectors S * HazMat Schools S Pharmacies Local Public Health Police * S Laboratories Fire S * * The Public * * Hospitals & Nursing Facilities Clinicians Coroner & Medical Examiner The Media * * Conclusions • Overall: IT/DSSs have not be subjected to critical evaluations • Detection Systems: – – – – Rapid Portable Sensitive and Specific Can test for more of the worrisome biothreat agents • General Diagnostic Systems: – Link to EMR • Management systems: – Utilize available bioterrorism guidelines • Surveillance systems: – Greater integration – Early warning data – Methods for the determination for what constitutes an outbreak • Communication systems: – Single system for public health officers at various levels – Public health media Future analysis • Models of natural history of disease processes of biothreat agents • Cost-effectiveness analyses of surveillance systems using different sources of data • Cost-effectiveness of prevention and management strategies for the most worrisome biothreat agents • Analyses of the adequacies of current capacity of hospitals, law-enforcement, and public health to respond to a bioterrorist attack. Collaborators • • • • • • • • Douglas K. Owens Kathryn McDonald Wendy Smith Chara Rydzak Herbert Szeto Corinna Haberland David Buckeridge Mark Schleinitz • • • • • • • • • Dean Wilkening Mark Musen Bahman Nouri Bradford W. Duncan Mariana B. Dangiolo Hau Liu Scott Shofer Justin Graham Sheryl Davies