Dr. JOSEPH BRITTO MD. MRCPCH. www.isabelhealthcare.com
Hurricane Isabel September 2004
How to Heal Healthcare Recently the DHHS announced a 10- year plan to build a new health information infrastructure. We both agree that in a new system, innovations stimulated by information technology, will improve care, lower costs, improve quality and empower patients Bill Frist & Hiliary Clinton Washington Post, 25th August 2004
Time is critical Consequences of Error Knowledge is not easily accessible
Case History June 1999. 3 year old girl. TUESDAY: Five day history of chickenpox Fever for 3 days Diarrhea and vomiting for 3 days Bluish discoloration and tenderness around the groin Seen in ER Reassured and sent home WEDNESDAY: Symptoms persist. Reassured by ER and GP over telephone THURSDAY: Seen in A&E multi-system failure, resuscitated, Stabilized and transferred to St. Mary s PICU
Complications of chickenpox Secondary bacterial infection Necrotizing fasciitis Varicella encephalitis Cerebellar ataxia Reye syndrome Thrombocytopenia Purpura Fulminans (associated with protein C and S deficiency) Data sets in the mind!
Knowledge in the clinical workflow?
KNOWLEDGE IN THE CLINICAL WORKFLOW?
THE INVERTED KNOWLEDGE PYRAMID CONSULTANT MIDDLE GRADE DOCTOR JUNIOR DOCTOR GP PATIENT
WHY IS CLINICAL DECISION SUPPORT NEEDED IN THE NHS? 50 million clinical decisions for every million of the population * Physicians use 2 million pieces of information We are a knowledge intensive sector * Making Amends: Reforming the Approach to Clinical Negligence. CMO. June 2003
Knowledge in the clinical workflow? Little use of IT & knowledge management in our cognitive decisions
3 weeks on a pediatric intensive care unit + 4 weeks on a high dependency unit = $ 250,000
Patient Safety In American Hospital July 2004
Patient Safety In American Hospital July 2004
Medical Error Failure to diagnose and treat in time accounted for almost 75% of all mortality attributable to patient safety incidents. Healthgardes July 2004. Total national costs of preventable adverse events estimated at $19bn Zahn, Miller JAMA 2003/ Healthgardes July 2004. Between 18 to 28 % of patients in UK, USA, Canada, Australia and New Zealand reported medication / medical error over the last 2 years. Medical & Medication Errors in Sicker Adults. Health Affairs. June 2003 Diagnostic error comprise a substantial and costly fraction of all medical errors and are the second leading cause for malpractice suits against US hospitals. Dr Mark Graber. Academic Medicine Oct 2002
What Is ISABEL? Isabel is a reminder system and a safety net Isabel is a unique diagnostic tool, designed by clinicians Isabel works by applying Autonomy s software algorithms to a proprietary rules-based medical database Isabel is Web based and can be integrated into a customer s Electronic Medical Records (EMR) implementation
STAGES IN THE DIAGNOSTIC PROCESS I : Information gathering / eliciting data II : Pattern recognition : integration of patient data with knowledge/experience III : Decision making (working diagnosis / investigations / treatment)
Knowledge in the clinical workflow? KNOWLEDGE IN THE CLINICAL WORKFLOW?
Knowledge in the clinical workflow? KNOWLEDGE IN THE CLINICAL WORKFLOW?
Knowledge in the clinical workflow?
Knowledge in the clinical workflow?
Differential Diagnosis ISABEL knowledge in the clinical workflow
Text at Bedside ISABEL corroboration with selected text at the bedside
Forrester Report September 2004
Testimonials & Awards Since discovering this website, I have listed the main clinical characteristics of every medical admission to our unit for the past month, and Isabel has included the diagnosis in every case. Neal J Thomas MD. Asst. Professor of Pediatric Critical Care Medicine, Pennsylvania State University College of Medicine, Penn State Children s Hospital, Hershey, Pennsylvania, USA Quoted in Critical Care, 2000
Validation Process I Stage 1: 80 junior doctors involved, completed end 2000 Stage 2: completed at 4 NHS hospitals, Aug 2000 Feb 2001 91-95 % accuracy
Isabel : VALIDATION PROCESS Simulated Field Trial: Online with simulated cases + clinicians Completed August 2002 Multi-centre Clinical trial: Real cases + junior doctors, near patient & on-line, 4 NHS hospitals Completed April 2003 Presented at Royal College of Paediatrics and Child Health, Annual Conference, York, April 2004
OUTCOMES change in mean quality scores 1.2 1 Differential Diagnoses Investigations Management 0.8 0.6 0.4 0.2 0-0.2-0.4-0.6 Mean 37.9 42.7 35.0 36.9 17.33 18.23 Before Isabel After Isabel Before Isabel After Isabel Before Isabel After Isabel In 1 in 7 patients, a clinically significant diagnosis was added by the user only after consultation with Isabel
Multi-centre, USA-UK PICU study 206 medical patients 5 PICUs (3 USA, 2 UK) 3 month period in 2003 UTILITY: At least one additional clinically important diagnostic reminder in 40% cases 2 such reminders final discharge diagnoses Presented at Society of Critical Care Medicine, Annual Conference, Orlando, February 2004
Isabel : VALIDATION PROCESS Family Physicians: Pediatric Negligence Claims 1000 consecutive alleged clinical negligence claims 27 month period : July 1996 to October 1998 151 were pediatric cases 88 cases with a code of failed / delayed diagnosis were analysed Symptom lists produced and inputed into Isabel 8 specific conditions provided 59% of the cases. 2 demographic peaks : < 1 year and 17 years of age Average number of FP consultations : 5 Average number of FP involved in each case : 3
Isabel : VALIDATION PROCESS Family Physicians: Pediatric Negligence Claims 69% cases Isabel provided the correct diagnosis 17% cases the FP was thought to have acted appropriately 52% cases Isabel might have altered the outcome Presented at Royal College of Pediatrics and Child Health, Annual Conference, York, April 2004
Improve access to Knowledge / CDSS/ EMR Wireless PDA: www.isabelhealthcare.com/pda
Isabel Embedded in EMR Electronic Medical Record (EMR)
Isabel Embedded in EMR Getting it all together, in one place
Isabel Embedded in EMR
Web based EMR interface
Web based EMR interface
Web based EMR interface
Web based EMR interface
Web based EMR interface
Web based EMR interface
Isabel Embedded in EMR Isabel : Real time epidemiology and clinical paradigm recognition
Pre-subscription: Users Number of registered users:16,500 With UK NHS auto-login: 25,000 User countries: 120 Number of regular users: 2,500 (> 5 times) Average monthly hits to site: 500,000 Average monthly page requests: 200,000
Institutional Clients/Users - USA Subscribers Washington State University at St. Louis Children's Hospital, Missouri Valley Baptist Medical Centre, Harlingen, Texas Ongoing Evaluation Texas Children's Hospital, Houston Memorial Hermann Children's Hospital, Houston, Texas Cooks Children's Hospital, Dallas, Texas Children s Medical Centre, Dallas, Texas Lucille Packard in Stanford, California Children's Hospital, Los Angeles, California
Isabel : Return on Investment Reduction in diagnostic/decision errors Reduction in time spent/costs to make correct diagnosis Reduction in insurance premiums Bridging the skills gap Continuing education in the workflow