infranet Project, University of Waterloo

Transcription

1 infranet Project, University of Waterloo Smarter Health: The Value of IT in the Health System Feb Bob Bernstein Gary Hollingworth Gary Viner

2

3 University teacher of family medicine GP Omnipracticien Cradle to grave URI s to schizophrenia, Well babies to heart attacks 30 Patients per day NOT a GST collector

4 If we make an adequate investment in the technologies and the human resources to deploy them, what shall we reap? If we keep on the way we have been, on the other hand, what will we lose? What can the system be tomorrow that it cannot be today (examples of new capabilities and deliverables of the health system)? What would the beneficial impacts be on the cost of health care, access to the system, the quality of the care provided, etc.? How would the lot of the health planners, the provincial ministries, the regional health organizations, the administrators, the providers, and the patients be improved? How will we know if we have succeeded?

5 By the end of this presentation I will have answered all of Dom s questions, BUT Indirectly With a focus on clinical care Technology serves patient care not the reverse It s not about cost or management It s about quality

6

7 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

8 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

9 In ONE word, what distinguishes general practice from specialty medicine?

10 The water we swim in: What is the diagnosis seen most frequently in primary care, what percent of encounters?

11 Primary care issues -Prevalence &Signal to noise Hypertension The content of family practice Proportion of visits Anxiety URI Bronchitis Prenatal CHF Ischemic H.D. COPD Annual Exam Osteoarthritis Acute O.M Common conditions Conditions <1.5%

12 Complexity of encounter 3-6 health problems per encounter Course of disease Early in course Undifferentiated complaints Time Longitudinal person oriented information Evolution of Health Problems over Time Information overload No institutional support

13 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

14 In the beginning: 3 x 5 inch index cards Then came Clem MacDonald and there was light Protocol-based computer reminders, the quality of care and the non-perfectability of man. N Engl J Med Dec 9;295(24):

15 The computer powerful ally in managing vast amounts of health information. Computers organize, process knowledge & information we would otherwise forget, overlook or, never have access to. Clinical Decision Support Systems (CDS) needed to allow physicians to actually make use of the electronic medical record.

16 May or may not be used at point of service Can take many forms, such as: warning of a potential drug-drug interaction providing information about costs of diagnostic procedures clinical practice guidelines reminders to follow up on routine procedures suggesting an entire plan of care that has been pre-established for a given clinical situation.

17 A fully integrated CDS in an EMR analyzes the incoming stream of clinical data searches for unusual patterns potential adverse events, makes recommendations of possible actions to remedy the situation.

18 A CDS takes the computer from simply being a passive reporter of facts to being an active participant in physician practice. It provides the professional with important information at the right time so that s/he can make the right decision about a patient s care.

19 Ever increasing body of medical knowledge, makes it impractical and expensive to perpetuate our current memory-based system of practice and education results in greater cost to patients owing to poorer practitioner decisions.

20

21 Feedback: Responds to an action taken by the physician or to new data filed in the system. Data Organization: Gathers disparate forms of data and presents it in a logical fashion that is just right for a given scenario.

22 Intelligent Action: Can automatically complete a review of test results, group them, find abnormals and generate letters. Communication: Alerts physician to unusual elements of data; communicates physician s response to others who need to know the information.

23 Substitute Therapy Alerts: Suggests alternatives to therapeutic or diagnostic orders placed into the system by the physician. Drug Checks: Provides drug allergy warnings and drug interaction checks.

24 Parameter Checks: Looks for overdoses, underdoses, and other discrepancies i.e., recommending a different dosage for a patient with renal failure.

25 Redundant Utilization Checks: Checks to see if a test or therapy has been ordered more frequently than normally required in a given interval. Often, redundant orders are due to oversight. In one study, 65% of such alerts resulted in the clinician retracting the order. Conversly, 35% of alerts are ignored We still need clinical judgement

26 Relevant Information Display: Shows other information that is important to know before the current transaction is completed; i.e., displaying relevant past lab results when a new medication is prescribed.

27 Time-based Checks: Reminders for potential errors of omission when an expected transaction has not occurred i.e., check to see if medication refills are due. Templates: A standard set of therapeutic and orders appropriate for a given situation.

28 Profile Display and Analysis: Provides periodic digests of a patient s general clinical profile (i.e., especially in multidoctor clinics) Aggregate Data Trending: Observes key indicators for a large number of patients over time to advise of possible clinical trends.

29 Currently there are CDS systems integrated into EMR s used in the States in acute and subacute centres and in some ambulatory care centres. One example is Problem Knowledge Couplers developed by Weed. Relevant clinical information entered into the computer. coupled or matched with all possible diagnoses and/or management options for that patient. The physician s judgement and the patient s values and preferences applied for a final selection of a diagnosis or management option.

30 An example of a stand-alone CDS is PROMPTOR- FM (PRObabilistic Method of Prompting for Test ORdering in Family Medicine)

31 Family Practice Residents Randomized to use PROMPTOR or not Clinical Vignettes 5 scenarios, 2 or 3 tests each 38% reduction in test ordering 12% reduction in costs

32 There is evidence to show that when physicians perceive that a CDS system facilitates their practice, they will learn to use it even if it requires changes in how they practice. (Berkowitz LL. Healthcare Informatics 1997; 14 (10): )

33 In order to ensure the acceptance and use of these systems, they must: provide immediate benefits to the physician and patient; be flexible enough to fit into the normal physician workflow; above all, be available at the point of patient care. (i.e., a computer in every room).

34 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

35 I Love My EMR!

36 1985-5% of physicians use the computer for more than billing/accounting % of physicians use the computer for more than billing/accounting New CIHI standards and initiatives for data recording in primary care

37 CFPC Library & Medline Electronic medical records Primary care 38 relevant articles 20 were some form of experimental study Clinical decision support Guidelines Patient issues EMR s

38

39 Physician and patient reminders, treatment planner and patient education were effective Improved preventive care, drug levels Show cost of drugs -> some increase others decrease

40 Compliance increases from 6 to 35% or 15.6 to 32% depending on study Elson: cholesterol guidelines followed inadvertently when flow sheet used

41 NONE!!! Patients neither more nor less satisfied No depersonalisation More doctor initiated talk, less patient initiated social talk

42 Increase preventive tasks Better BP control More complete problem & med lists Computer and chart info not always the same More complete documentation for decision making Better communication with consultants Able to predict heart disease from patient info

43

44 Data entry problem Add secs/visit Patient volume stays the same Costs killed one outpatient system Estimated costs are 7.5 to 13.5% of enterprise budget!! Societal costs exceed benefits!!

45 Electronic Medical Records in Primary Care

46 Why EMR s in Primary Care Fail (COACH proceedings 1997) User input into design Data entry Complexity Modularity Standards Conflicting uses of data Vendor stability Technical and conceptual compatibility between systems Error Fear of failure, Cost of redundancy Stability/Functionality All or none Networking Cost: Penalised for being smart Hardware Accessibility / availability, Upgrades Redundancy Networking

47 Coded data: Researchers Managers Free text: Clinicians

48 PARTS of EMR improve care WHOLE EMR is more costly Full EMR Opinions are positive Studies are negative

49 INTEGRATION OF HEALTH INFO

50 The PATIENT is seen at many disconnected points & times in the health care system: Primary care Tertiary care Respite care Secondary care Laboratory Home care Long term care Radiology

51 Tertiary care Secondary care Primary care Home care Long term care Respite care Laboratory Radiology

52 In-hospital care Sub-Specialist care Long term care Specialist care Ambulatory care Primary Care Community care Rostered Population

53 INFORMATION SYSTEM IHS Systems Goals & Policies

54 The Patient is my specialty Primary care Secondary care Some tertiary care Community care Respite care Long term care Office care Ambulatory care

55 COMMUNITY Laboratory Primary care Tertiary care Continuity of Care Secondary care Long term care Respite care Pharmacy Home care

56 COMMUNITY Laboratory Primary care Tertiary care Respite care Personal Physician Low Tech Continuity of Information Hi Tech Pharmacy Secondary care Long term care Home care

57 COMMUNITY Laboratory Primary care Tertiary care Respite care Available Standardized Valid Reliable Concise Pharmacy Secondary care Long term care Home care

58 COMMUNITY Laboratory Primary care Tertiary care Respite care Data for Patient Care Data for Analysis Pharmacy Secondary care Long term care Home care

59 Circular integration requires: Continuity of care Personal physician Continuity of information through I.T. Conceptual compatibility A standardized terminology or terminologies

60 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

61 Patients? Doctors? Encounters? Diagnoses? Episodes? The patient with the problem over time

62 Acknowledge WONCA International Classification Committee Distinct from an encounter Refers to all care provided for a discrete health problem or disease in a particular patient A health problem or disease from its first presentation to a health care provider until the completion of the last encounter for that same health problem or disease

63 The natural way GP s think The unnatural way to record Current reporting requirements defeat episode recording attempts Cannot link current data

64 Core Elements Reasons for encounter Health problems/diagnoses = Episode Title Process of care/interventions (Outcomes) Transitions One or more encounters, including changes in their relations over time

65 1st visit RFE feeling tired Health Problem fatigue Intervention Hb 2nd visit RFE test result Health Problem anemia Intervention barium enema 3rd visit RFE test result Health Problem Ca colon Intervention referral

66 Reason for encounter (RFE) The agreed statement of the reason(s) why a patient enters the health care system, the demand for care by that person. Symptoms or complaints (headache or fear of cancer), known diseases (flu or diabetes), Requests for preventive or diagnostic services (a blood pressure check or an ECG), a request for treatment (repeat prescription), To get test results, or administrative (a medical certificate). It is the PATIENT S statement, clarified by the doctor.

67 Reason for encounter (RFE) This is a true primary care concept Primary Care is RFE driven, not diagnosis driven

68 Health problem Certainty Status Clinical findings No good terminology yet for primary care Data input: Too much coded data?

69 Processes of care Diagnostic and preventive procedures Treatment, procedures and medication Tests and results Administrative Referrals Outcome Functional status WONCA COOP Charts Severity of illness WONCA DUSOI

70 Transitions An episode may occur over many encounters Diabetes is a lifelong episode Each encounter may have more then one episode assessed Many to many relationship between episodes and encounters

71 Bleeding started in the rectal area and continued all the way to Los Angeles We hold that all men are created equal Terminology requirements Rich in RFE s/symptoms Clinical language Bilingual Synonyms and pet terms Elimination of ragbags Relevant to the discipline Maps to international and Canadian standards

72 A B D F H K L N P R S T U W X Y Z General and unspecified Blood and blood-forming organs and lymphatics (spleen, bone marrow) Digestive Eye (Focal) Ear (Hearing) Circulatory Musculoskeletal (Locomotion) Neurological Psychological Respiratory Skin Endocrine, metabolic and nutritional (Thyroid) Urological Pregnancy, child bearing, family planning (Women) Female genital (X-chromosome) Male genital (Y-chromosome) Social problems (ENCODE has community health problems, social determinants and Infancy) I

73 Components (standard for each chapter) CODES 1. Complaints and symptoms Diagnostic and preventive Treatment, procedures and medication Test results Administrative Referral and other reason for encounter Diagnoses/diseases: infectious diseases -neoplasms -injuries - congenital anomalies -other

74 1st visit RFE feeling tired Health Problem fatigue Intervention Hb 2nd visit RFE test result Health Problem anemia Intervention barium enema 3rd visit RFE test result Health Problem Ca colon Intervention referral

75 Basic Unit. The episode at an encounter RFE: Sympt. Complt. Problem RFE: Intervention Intermediate Intervention Clinical Findings Health Problem + Certainty & Status Subsequent Intervention Outcome COOP DUSOI Repeated for each episode at that encounter Repeated at each encounter for that health problem Software maintains the link.

76 Different levels of complexity of recording 1 to 4 RFE s, including process RFE s 1-2 History elements 1-2 History negations 1-3 Intermediate interventions 1-3 Subsequent interventions (Balance of coded and non coded data)

77 The only way to go Terminology must be rich, RFE oriented ICPC is not enough, too big is too much ENCODE for clinical specificity and reliable classification Increased complexity of recording Data entry problem, Too much coding Duplication PEPPER (Portable Electronic Patient Profile and Episode Record & SALTED (Statistical Analysis of Long Term Episode Data)

78 (Portable Electronic Patient Profile and Episode Recorder)

79 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the concept of "episode of care" as a relevant model for data collection understand the potential limitations of electronic information systems in primary care understand the power and promise of electronic information systems in primary care

80 The Power and Promise of Standardised Data The Transition Project Dr. Henk Lamberts Department of General Practice - U of Amsterdam Netherlands data 43 general practitioners in 24 practices (1-year) 267,897 encounters (number of face-to-face visits) 389,709 reasons for encounter (max. 3/visit) 236,027 episodes - (186,559 new)

81 Transition Project Data Imagine symptom clusters, negations

82 The Power and Promise of Standardised Data Refinement and quantification of clinical judgement This can change the way that we practice medicine sensitivity and specificity of signs and symptoms Pre test likelihood PROMPTOR Effective diagnostic testing Promptor study 38%, 12% reductions

83 The Power and Promise of Standardised Data We understand what we do, and its value MOH understands & values what we do Real-time decision support Effect of lab testing and procedures Outcome measurements Practice better medicine

84 The large majority of personal health care needs the comprehensiveness the degree of integration of accessibility and of accountability can be assessed when episodes of care are classified with ICPC in a computer based patient record. From ICPC-2, Oxford University Press 1998

85 By the end of this session you will understand how the science of primary care is different from specialty and institutional care be able to define clinical decision support (CDS) know the evidence for effectiveness of CDS know the limitations of CDS understand the potential benefits of electronic information systems in primary care. understand the potential limitations of electronic information systems in primary care understand the concept of "episode of care" as a relevant model for data collection understand the power and promise of electronic information systems in primary care

86

87 Private sector Privacy/Confidentiality Proprietary Bottom line based Encounter epidemiology Supported by advertising Management information Institutional care Secondary data entry Public sector Anonymity Sharable/Standards Evidence based Episode epidemiology Supported by cost savings Clinical information Community care Point of service data entry

88

89 Opinion or RCT based clinical judgement Prescribing influenced by marketing Protocol/care map driven Lab testing based on opinion Discoordination of care Unable to foresee clinical or manpower needs Population based better clinical judgement Evidence and value based prescribing Information driven Laboratory testing with predictive value Integrated communication between professionals Better predictions Better resource allocation

90

91 If we make an adequate investment in the technologies and the human resources to deploy them, what shall we reap? If we keep on the way we have been, on the other hand, what will we lose? What can the system be tomorrow that it cannot be today (examples of new capabilities and deliverables of the health system)? What would the beneficial impacts be on the cost of health care, access to the system, the quality of the care provided, etc.? How would the lot of the health planners, the provincial ministries, the regional health organizations, the administrators, the providers, and the patients be improved? How will we know if we have succeeded?

92 infranet Project, University of Waterloo Smarter Health: The Value of IT in the Health System Feb Bob Bernstein Gary Hollingworth Gary Viner