Using IT to Improve Quality:Past Results and Future PotentialDavid W. Bates, MD, MScMedical Director of Clinical and Quality Analysis, Partners HealthcareChief, Division of General Medicine, Brigham and Women’s HospitalGoals• Major gaps between evidence, practice• Costs high• Problems with errors• Computerized decision support• Current Partners system• Errors• Costs• Guidelines• Next 5 years at Partners IS• ConclusionsLeadership and ITLeadership is the capacity to hold a shared vision of that we wish to create.– Peter SengeThe best way to predict the future is to invent it. – Peter DruckerOld Paradigm• Authorities are infallible• Heuristics work well• If in doubt, do it• Clinical judgement and the “art of medicine” get you to the right answers• Community standards are correctDavid Eddy, Aetna Quality Forum 1999New Paradigm• Authorities vary substantially• Heuristics don’t work• Clinical judgement is insufficient• Huge variation by communityTherefore• Need to begin to practice evidence-based medicineDavid Eddy, Aetna Quality Forum 1999The IOM Report• Report targets hospital errors: Mistakes killing thousands every year 11/30/99• Medical errors kill 44,000-98,000 people per year• “More people die from medical errors each year than from suicides, highway accidents, breast cancer, or AIDS”• “These stunningly high rates of medical errors -resulting in deaths, permanent disability, and unnecessary suffering - are simply unacceptable in a system that promises to first ‘do no harm.’” William RichardsonReengineering Medicine:The Role of IS• Could be changed by providing external aids• Linking medical knowledge and patient-specific data• Identifying options• Without such tools, experts• Make errors• Overlook available knowledge• Don’t sufficiently account for uniqueness• Patients could participate in decision-makingWeed LL, Weed L, Federation Bulletin, 1994Development and Implementation of POE• Physician involvement and leadership• Decision to automate existing systems as is• Constant focus on speed• Strong support from hospital administration• Willingness to be flexible, modify systemEvent monitor architectureRule editorKnowledge baseInference engine(decisions)Applications (new data)Applications (new data)Applications (new data)Patient databaseAnnun-ciatorspage, email, write to file, [real time message]Coverage ListPhysician Coverage List• Functions• Identifies first and second-call physicians• Manages physician rotation• Handles evening coverage and signing out• Facilitates delivery of computer-generated messages• Computer-page interface allows automated pagingPharmacy Computer System Field Test of Unsafe OrdersUnsafe OrderNot DetectedCephradine oral suspension IV 61%Vincristine 3 mg IV x 1 dose 62%(2-year-old)Colchicine 10 mg IV for one dose 66%(adult)Cisplatin 204 mg IV x 1 dose 63%Source: ISMP Medication Safety Alert! Feb 10, 1999Handwriting exampleMedication Error Frequency and Potential for HarmIn 10,070 Orders530 Medication Errors 1.4 per admission35 Potential ADEs5 Preventable ADEs• 1 in 100 medication errors results in an ADE• 7 in 100 represent potential ADEsADE Prevention Study: Key Results• 6.5 ADEs/100 admissions• 28% preventable• 3 potential ADEs for every preventable ADE• 62% of errors at ordering and transcription stages• Systems analysis• No individual responsible for repeated errors• Systems should be designed to:• Make errors less likely• Catch those that do occurJAMA 1995;274:29-43Costs of ADEs• ADEs are expensive• $2461 per ADE, $4555 per preventable ADE• Annual BWH costs:• $5.6 million for all ADEs• $2.8 million for preventable ADEs• These figures exclude costs of:• Injuries to patients• Malpractice costs• Costs of admissions due to ADEs• Justifies investment in prevention effortsJAMA 1997;277:307-311• Streamline, structure process• Doses from menus• Decreased transcription• Complete orders required• Give information at the time needed• Show relevant laboratories• Guidelines• Guided dose algorithms• Perform checks in backgroundDrug-allergy Dose ceiling Drug-labDrug-drug Drug-patient Improving the Quality of Drug Ordering with Order EntryAllergy to MedicationChemotherapy Order:Patient CharacteristicsHigh Chemotherapy Dose WarningHigh Chemotherapy Dose:Requires Attending ApprovalSerious Medication Error Rates Before and After OE024681012Serious Medication ErrorsEvents/1000 Patient-daysPhase IPhase IIDelta = -55%p < .01Bates et. al. JAMA 1998Impact of BWH Inpatient Provider Order Entry•Nizatidine use, for all oral H2 blocker orders, increased from 12% to 81%• The percent of doses over the suggested maximum decreased from 2% to .6%• The percent of orders for ondansetron, with a frequency of 3 times daily, increased from 6% to 75%• The percent of bed rest orders with a consequent order of heparin increased from 24% to 54%Teich, Arch Int Med 2000“Panic” Laboratory Study• For markedly abnormal results (K, Na, glucose, Hct)• Allows consideration of other factors • Direct interface with paging system• “Before” data• Median time to rx 2.5 hours• For 25% > 5.3 hours• RCT results• Mean time to rx 11% shorter (p<.0003)• Mean time to resolution 29% shorter (p=.11)• 95% physicians pleased to be pagedKuperman, JAMIA 1999Reducing Drug Costs with Order Entry• Types of useful suggestions• Drug interchange• Lower dose• Different route (IV-PO switches)• Guidelines for useEffect of Changing Default Dosing Frequency for Ceftriaxone01020304050601 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17WeekOrders/weekBIDQDSelected Laboratory Interventions• Charge display RCT• No statistically significant effect• BUT $1.7 million lower lab charges in intervention group• Redundant labs• 67% reminders followed• Annual charge savings $31,000, vs. estimate of $376,000• Only 44% tests performed had computer order• Substantial improvement possible if loop closed with laboratory “back end”Other Laboratory Evaluations• Antiepileptic drug levels• Only 28% of BWH inpatient levels appropriate• RCT of structured ordering showed improvement• Digoxin levels• Only 16% of BWH inpatient levels appropriate• Potential charge savings $388,000• PSA levels• 19% inappropriate (age, frequency issues)• Thyroid studies• Initial testing TSH alone in only 73% of