The Near Future of Medicine: Justintime Information at the PointofCare Academic Concepts vs Products

1
The Near Future of Medicine Just-in-time
Informationat the Point-of-CareAcademic
Concepts vs Products
2
How do we learn?

• Adults learn by solving problems
• Our problems clinical questions
• CME can highlight advances and make us aware of
  our deficits
• Answering clinical questions in the context of a
  real case at the point of care is how we learn

3
Clinical questions

• Theyre common Average 1 question for every
  other patient
• Theyre important
• Only 1 in 3 questions pursued
• Found answer 80 of time
• Guess at 70-80 of information needs!
• Journals only used to answer 2 of 1101 questions
  in busy practice
• Books, colleagues used most often
• ½ of the time lt60 seconds was devoted to
  answering the question
• Ely JE. Analysis of questions asked by family
  doctors regarding patient care. BMJ
  1999319358-61

4
Clinical Questions

• Internal Medicine Residents
• 2 for every 3 patients
• 29 pursued
• textbook (31) journals (21) attendings (17)
• Patient expectation, fear of malpractice
  associated with seeking answer
• Lack of time (60), forgot (29).
• Am J Med 2000109218-33.

5
Information Sources for the point of care

• Keep in mind the usefulness equation
• Usefulness Relevance x Validity
• Work

6
POEM

• Patient-Oriented
• Evidence
• that Matters
• matters to the clinician, because if valid, will
  require a change in practice

7
Minimizing WorkInformation Sources

• Just-in-Case information
• Libraries, Medline, MDConsult, WebMD, StatRef,
  Up-to-Date, other databases
• A superstore of information
• Focus a complete inventory of information
• Benefit Much information is always in stock to
  meet many needs
• Detriments Even the simplest needs require time
  to access the information

8
Minimizing Work Information Sources

• Just-in-Time information
• Highly filtered information sources with rapid
  access
• A Seven-Eleven -- not everything, but quick and
  what you need most of the time
• Focus the best, most commonly needed information
• Benefit Rapid access (less than one minute)
  ease of use
• Detriments Reliance on the filtering mechanism

9
Key resources at the point-of-care for
hunting

• Desktop computer
• Cochrane Library
• Essential Evidence Plus
• Dynamed
• Best Evidence/Clinical Evidence/Clinical
  Inquiries
• Stat!-Ref
• Up-to-Date
• Handheld computer
• Essential Evidence Plus
• PDR, Lexi-Comp Drug Database, 5-min. Clinical
  Consult
• Low tech
• Wall charts, pocket cards, and bulletin boards
• Pearl books

10
Two Tools Needed to Master Information

• A method of being alerted to new information
• A tool for finding the information again when you
  need it
• Without both
• You dont know that new info. is available
• You cant find it when you do
• Clinical example- Riboflavin for migraines

11
Quality First-Alert Systems

• 1. How is the information filtered?
• Patient- vs disease- oriented?
• Specialty-specific?
• Comprehensive? Which journals?
• Does it matter (change my practice?) or is it
  simply news?
• 2. Is the information valid?
• must have levels of evidence (LOE) labels
• Beware Trojan Horse!

12
Quality First-Alert Systems

• 3. How well is information summarized?
• 2000 - 3000 words accurately in 200 words
• 4. Is the information placed into context?
• Much more than abstracts
• Translational Validity

13
First-Alert SystemRisks

• Spyware Epocrates, PDR for the PPC
• Trojan Horse whos paying when its free?
• Abstracts only Journal Watch, BMJ Updates,
  Journals to Go, Tips from other Journals,
  Clinical Updates
• No relevance/ validity filter
• You can have information free and you can have
  it uncensored, but you cant have it both ways.
  No Free Lunch!

14
Randomized
Analyzed on an intention to treat basis
Individuals assessing outcomes were blind to
treatment
15
Essential Evidence Plus The Clinical Awareness
System

• Daily POEMs/Essential Evidence Plus
• Specialty-specific
• Comprehensive
• Coordinated (hunting and foraging together)
• Specific and reproducible criteria for relevance
  and validity
• Available at the point-of-care
• All backed up by LOEs
• POEMs for Primary Care, Pediatrics, Internal
  Medicine

16
? Quality of Information in Reviews

• Study of 36 review articles using 10 criteria for
  determining rigor
• Overall rating of rigor
• Experts correlation 0.23
• Non-experts correlation 0.78
• ? expertise of writer stronger prior opinion ?
  less time spent on review ? lower quality
• Study of quality of 35 review articles on type 2
  DM
• Ave. score 1/15 best score 5/15
• Experts should do research
• Non-experts should write due to less bias
• Oxman AD, Guyatt GH. The science or reviewing
  research. Ann N Y Acad Sci 1993703125-33.
  Shaughnessy AF, Slawson DC. What happened to the
  valid POEMs? A survey of review articles on the
  treatment of type 2 diabetes. British Medical
  Journal 2003327266-9.

17
(No Transcript)
18
Translational ValidityCan We Trust Review
Articles?

• Reporting of the UKPDS by 35 review articles
• 85 of reviews Didnt tell readers that good
  glucose control doesnt decrease mortality
• All reported that good control decreased
  complications
• None reported that almost all (84) benefit due
  to decreased rates of retinal photocoagulation
  (no change in blindness rate- important NonValid
  POEM)

19
Translational ValidityCan We Trust Review
Articles?

• Only 18 (NNR 6) metformin decreased
  mortality, independent of BS control (Valid POEM)
  
• None reported lack of any benefit (micro- or
  macrovascular) of insulin/ sulfonylureas in obese
  diabetics- Important NonValid POEM
• Only 13 (NNR 8) reported that blood pressure
  control is more important than BG control (Valid
  POEM)

20
Lending a Hand to Patients with Type 2 Diabetes
Metformin/?ASA
Blood pressure
Cholesterol
Glucose control?
Smoking
Slawson DC, Shaughnessy AF. Lending a Hand to
Patients with Diabetes. Data from Vijan S.
Treatment of hypertension in type 2 diabetes
mellitus blood pressure goals, choice of agents,
and setting priorities in diabetes care. Ann
Intern Med 2003 138593-602.
21
GS Valid POEM vs PROSE?

• All individuals with type 2 diabetes have
  insulin resistance. Insulin resistance is the
  problem. Metformin (and insulin and
  sulfonylureas) treats only the symptom of
  hyperglycemia. Therefore, the thiazolidinediones
  should be used as the very first-line agent . .
  . J Fam Pract 200251984
• This apparent ability of TZDs to rejuvenate
  beta-cells is a compelling reason to use these
  agents as initial therapy for type 2
  diabetes. Am J Med. 2003 Dec 8115 S8A20S-23S.
• PROSE Prescribed Recommendation(s) based On
  Substandard Evidence

22
Drilling for the Best Information
23
The Clinician of the Future

• I know a lot, therefore I am
• Replaceable by a computer
• I think, therefore I am
• Never replaceable by computer
• Hand held computer stethoscope of the future

24
(No Transcript)