A Horizontal Slice Through Medical Informatics

1
A Horizontal Slice Through Medical Informatics

• Mgt 459d Digital Health Telemedicine
• Class 2
• David Dilts

2
Questions Focus

• Questions of the Day
• What technologies are available to follow a
  patient through the system? What are their
  benefits and limitations?
• Focus 
• Clinical Pathways / Care paths

3
Textbook Chapters

• Ch. 7 Protocol-based decisions support
  evidence-based medicine
• Ch. 8 Designing and applying protocols
• Ch. 9 Computer-based protocol systems in
  healthcare

4
Systems

• Are models
• Have behavior
• Are embedded in an environment
• Have structure
• Have function
• May have a purpose
• May be arbitrary

5
Generic Model of How to Model
Real World
Artifact
Model of The World
Artifact Model
6
Evaluation Framework
Problem Domain
IV Setting Validity
I Conceptual Validity
Replication
Literature
Conceptual Model
Problem Setting
Field Tests Data Analysis
Development
III Operational Validity
II Verification
Created System
7
Levels of Knowledge

• Data facts, raw values, little context
• 103 degrees
• Information data with context
• The patients temperature is 103 degrees
• Knowledge information with relationships
• If a patient is 3 years old, has a temperature
  of 103 degrees, and the sniffles, she most likely
  has the flu
• Wisdom knowing what to do with knowledge
• Having the flu does not mean a patient needs
  antibiotics.
• Note a major problem with most systems they
  deal with data, not information

8
Electronic Medical Record

• The basic of all medical informatics
• Generally only data with a data dictionary
• Generally contains no knowledge, or wisdom
• Generally designed functionally instead of
  customer-centric
• Owned by the care provider, not the customer

9
Ideas and trends in medical records (Dwyer, 1999)

• The future scenario Clinicians could
  automatically
• Choose a practice guideline
• Review all patient records for major quality of
  care issues
• Screen all patients for drug interactions
• Scan articles relevant to the patient condition
• Make patient record available for clinical
  research
• Note all clinician based!

10
Protocol-based Critical Pathways

• a.k.a., clinical practice guidelines, care maps,
  practice parameters, care pathways, care paths,
  etc.
• a formal, documented process for treating a
  condition
• a set of standards of practice
• are management plans that display goals for
  patients and provide the corresponding ideal
  sequence and timing of staff actions to achieve
  those goals for optimal efficiency (Pearson,
  1995, p. 941.)
• systematically developed statements to assist
  practitioner and patient decisions about
  appropriate health care for specific clinical
  circumstances. (Institute of Medicine)

11
History of Critical Pathways

• Developed initially from CPM PERT
• Then flowcharting methods
• Then Critical Pathways, but for nursing care
• Transitioned into Hospitals
• Drivers DRGs and HMOs
• Today, 80 of US hospitals use critical pathways
  of some type

12
Goals of Critical Pathways

• Best practice versus wide-practice styles
• Defining standards of hospital duration (DRG
  issue)
• Examine interrelationships among steps (i.e.,
  BPR)
• Provide a game plan for all involved parties
• Collect data
• Decrease documentation burdens
• Improve patient satisfaction

13
Why use Critical Pathways?

• Researchfind out what actually works
• Start of modern medicine and epideminology
  (medical accounting)
• Responsibility
• Identification of who is responsible for what
• What to do in what case
• Reduce inappropriate variation in clinical
  practice patterns
• Increase consistency, and hence quality
• Reduce cost of care
• Legal Liability Reduction (?)

14
Elements of Clinical Pathway Structure

• Entry criterion
• Functional flows
• Flowcharts The flow of work
• Decision trees Yes/no decisions
• Petri Nets State transition diagrams
• Discuss specific decisions and tests required at
  different steps
• Results oriented

15
Design Principles

• Use a multidisciplinary approach
• i.e., use all major stakeholders (including
  patients)
• Make all assumptions explicit
• Do not be too specific (Figure 8.2, p. 111)
• Reflect the skill level knowledge of the user
• Protocols must be constantly reviewed

16
When not to use Pathways

• When clear goals cannot be isolated
• When probabilities or utilities cannot be
  evaluated
• When there is ill-defined results
• When decisions are not separable
• When there is high variance in unmeasurable
  aspects of the input

17
Passive versus Active Protocols

• Passive
• Only a source of information, not intrinsically
  part of the care process
• A set of helpful guidelines
• Active
• Shape the delivery of care
• Explicit rules and guidelines of care

18
Promises Pitfalls of e-medical records (Ann
Intern Med 1998)

• Define appropriate use
• Ensure security confidentiality
• Guide patients in effectiveness of use
• Proactively assess medico-legal liability
• Enable for use by multilingual and multicultural
  populations

19
Measurement Problems

• It is easier to measure costs than to measure
  quality
• How to measure input?
• Patient satisfaction / Quality care

20
Legal Issues

• Physician Licensure
• Physicians are licensed, not HMOs
• Economic credentialing
• Cannot use practice guidelines for physician
  credentialing (e.g., you cannot fire a physician
  for not following a critical pathway)
• Malpractice
• Follow the pathway and get sued for poor quality
  (standards of care) OR
• Dont follow the pathway and get sued for poor
  quality!

21
Ideas Trends in Medical Informatics (Dwyer 1999)

• Barriers to progress
• Quality of information on the net
• EBM research has shown that practitioners have
  only 30 minutes per week to review evidence
• Security confidentially
• Future possibilities
• Reduction in medical errors
• Focus on the clinical encounter

22
Questions about Critical Pathways (Pearson, 1995)

• How to deal with variances from the path
• Autonomy versus standardization
• Malpractice Risk is not followed
• Research and education versus individualized care
• Effectiveness Measurement

23
Questions about critical pathways (Pearson, 1999)

• Are they any better than other practices?
• Not necessarily (Holmboe 1999 study that compared
  to other CI initiatives)
• Medicines response, Holmboe et. al., 1999 was
• Only a single medical condition
• Unique to the hospitals involved
• Not a randomized study

24
Ideas Trends in Medical Informatics (Dwyer 1999)

• Standardization
• integrate these data seamlessly
• HL7
• Unified Medical Language System (a code of codes)
• Patient-Centered Health Care
• Internet-based care
• Promises Pitfalls
• Barriers to progress
• Future possibilities

25
The Major Questions

• Are critical pathways an effective means to
  reduce the costs of health care, improve quality,
  and reduce malpractice liability?
• OR
• Are they just being used to appease powerful
  interests-with limited promise perhaps
  potential negative cost, quality, and liability
  implications?

26
Two Views

• most observers characterize available evidence
  as inconclusive at best (Allen et al., 1997)
• Versus
• Proper skepticism does not mean that we should
  allow cynicism to block the search for ways to
  bring evidence into medicine, to reduce
  unnecessary variation in care, and to improve the
  use of effective treatments. Whether critical
  pathways should have a prominent role in these
  efforts is unknown, but while that is being
  determined, such programs will remain in place
  ( Pearson, 1999)
• guidelines are not working in the physicians
  favorthey are more often used against physicians
  than in their defense. (Brennan, 1994, quoted in
  Pearson, 1999)