The DNA of Technology for Chronic Disease Management

1
The DNA of Technology forChronic Disease
Management

• David J. Morin CEO and Co-Founder
• Cielo MedSolutions LLC

2
What I Will Cover

• Elements to consider when evaluating technology
  for chronic disease management

3
The Goal of Technology

• A lot has been said about technology and chronic
  disease management (CDM) /quality improvement
  (QI)
• But the goal isnt the implementation of
  technologies per se
• It is high-quality, patient-centered care
• The technology is the catalyst it is what you
  can organize around and through which you can
  enable change

4
On the Road to QI

• If improved quality of care is your journey,
  think of

• Your technology as your vehicle (the enabler)
• Your care data as the vehicles dashboard (what
  you are managing)
• CDM/QI programs as a road map (how you get there)

5
Technical Components

• Disease management can be enabled through tools
  like
• Electronic health records
• Disease registries
• Clinical decision support
• Through functionality like
• Care reminders at the point of care
• Population management functionality to reach due
  patients
• Performance feedback reports to monitor care
  delivery
• Patient education reports
• But, you need the right underlying elements
  within the technology for success

6
The Technology DNA

• Correct and complete information.
• For all your patients and patient problems
• Collected, presented and delivered effectively.
• Giving everyone the right information and tools
  to drive care improvement (including the
  patient)
• Easily implemented, adoptable in bites.
• And adaptable to future needs.

7
Correct and Complete Information

• Phillips and Klinkman1 say your data must answer
• Who has ________ ? disease registries
• the basis for point-of-care decision support and
  quality assessment
• Who gets ________ ? the probability of specific
  diagnoses from common presenting symptoms
• basic clinical epidemiology in primary care
• requires episodes of care
• What is the context in which the care is
  provided?
• competing demands, social problems, patient goals
  and priorities
• multimorbidity
• What happened out there?
• track care across settings primary to specialty
  care, office to hospital

8
A Data Model

• Phillips and Klinkman refer to a primary care
  information model simple building blocks to
  capture complex reality1

9
Correct and Complete Information

• Make sure your data is telling you whats really
  going on
• Administrative diagnosis data has issues when
  used for clinical documentation and decision
  support
• 50 inaccuracy in administrative data (Jollis, et
  al, 1993)2
• 43 inaccuracy in administrative data (Peabody,
  Medical Care, 2004)3
• Billing and reimbursement coding mindset
  restricts improvement activities (Langley J.,
  Beasley C. 2007)4
• ICD-9 limited in fit for primary care
• 45 of presenting problems dont fit (White,
  1969)5
• ICD-9-CM captures considerably less than half of
  the information considered important (Chute, C.
  1995)6
• Lack of documentation regarding severity

10
Correct and CompleteInformation

• ICD-10
• 155,000 terms still not 100 coverage
• SNOMED-CT
• Over 344,000 concepts too much?
• ICPC - International Classification of Primary
  Care
• 95 fit to primary care with specificity
• Symptom and social problem diagnoses
• ENCODE
• 10,000 primary care clinical terms
• Chronic, acute, family history, social problem,
  symptom
• Mapped to ICD-9, ICD-10, ICPC

11
Correct and Complete Information

• Capture all patient problems
• Manage to the patient, not to the disease
• Be organized by patient not disease, but
  responsive to disease populations (Austin, 2007)
  7
• Registry of the Day not a really good idea
• Expensive, time-consuming, slows benefit
• Results in silos of data (this is not your goal)
• Capture both billable and non-billable diagnoses
• Know the source of the data
• attribution, administrative or self-reported

12
Correct and Complete Information

• Know where the patient is relative to the care
  they need
• must have context
• a response to a reminder for an evidence-based
  guideline is, in many cases, not a binary
  response (Y/N)
• you aint done till youre done

13
Correct and Complete Information

• Examples
• Colorectal cancer screening many times the
  first occurrence of this reminder leads to a
  discussion on the options on this screening.
  Patient usually goes home to decide and screen is
  ordered after a 2nd discussion.
• Reminder is flagged discussed on first visit,
  ordered on second visit. Only when the
  screening is completed is guideline considered
  done.
• A1C evaluation usually the patient is given a
  lab requisition to have blood drawn and tested at
  a later date.
• Reminder is flagged ordered on first visit.
  Only when a result from the test is returned is
  guideline considered done.

14
Collected, Presented and Delivered Effectively

• All patient encounters must utilize the
  technology
• If not all-patient, will not become routine in
  care delivery, adoption will suffer
• Presenting information has to be simple and fit
  into the existing workflow
• Shellhase (2003)8 found that 75 of physicians
  using an EHR ignored or did not observe flashing
  reminders for preventive services
• How many clicks and/or screens to get to the info
  you need?

15
Collected, Presented and Delivered Effectively

• Inaccurate or untimely information will lead to
  frustration and adoption will suffer
• Examples with regards to reminders
• Prompt for A1C, but patient had already been
  given lab req.
• Prompt for pap smear, but it is not due for six
  months
• Lack of comorbidity data, wrong evidence-based
  guideline presented (like diabetes and renal
  disease vs. diabetes)
• Prompt for mammogram, but mammogram already
  delivered

16
Collected, Presented and Delivered Effectively

• Care reminders must take into account the correct
  variables
• Examples include

17
Delivering Actionable Data to Improve Care

• Just reporting a score, good
• Reporting such that you can increase your score
  (and increase quality), priceless
• You must have at your fingertips timely,
  accurate, actionable and forward-looking data to
  continually drive improvement across the
  population
• A care report should help you to DO something

18
Want This Type of Report?
19
How About This Type of Report?
20
Or This Type of Report?
21
Reporting

• A good reporting module should
• be the front windshield, not the rearview
  mirror
• give you access to your data
• allow you to monitor the population but action
  the individual
• provide an easy way to modify and configure
• enable analysis differently than how data is
  collected
• data collection to the guideline reporting to
  the quality program

22
Giving the Entire Care Team the Right Tools to
Drive Improvement

• Physician-directed primary care team managing
  patient care
• Your technology should support a Team Sport
  concept
• Everyone in a practice has a role in improving
  care quality
• Everyone in a practice should have tools to
  improve care quality
• Reporting and actionable data provides that
• Some examples
• Pre-visit planning
• Patient outreach
• Data sharing

23
Giving the Patient Simple and Effective Tools to
Participate

• Patient Health Summary/Care Plan
• Individualized document showing patients status
  on key indicators and needs
• List of future needed services with dates
• In simple language, no guess work
• Personal Health Record (PHR) patients
  electronic file of health data
• This sounds easy, but in reality is tough to fit
  into workflow
• Which PHR will you support? - Over 100 efforts
  underway to build a PHR
• How are you going to access it?
• Do you want outside devices plugged into your
  network?
• What if it isnt simple to get the information?
• Can you trust the information it? - Whats the
  source?

24
Adaptable to Future Needs

• Dont implement technology to support todays
  needs, implement technology to support both
  todays and tomorrows needs
• All-problem, all-patient registry as new
  quality programs emerge, an all-problem,
  all-patient registry supports them on-the-fly
• Use of a simple data model a database that is
  easy to understand and query is one that is easy
  to write reports against. A database with 100s
  of tables is very difficult to use for report
  writing (and very expensive)

25
Adaptable to Future Needs

• Support of data sharing standards your
  technology must be able to both send and receive
  data to other systems and entities
• Table-based versus programming-based decision
  support engine adoption of new guidelines can
  be done in a matter of hours

26
Adoptable in Bites

• Keep it simple! (aka dumb it down)
• Go after this in a phased approach
• Start with something simple that will provide a
  win and is easily implemented
• Rollout new pieces in phases
• Make sure all stakeholders have buy-in and a
  voice in design and rollout
• Keep each phase manageable, well-defined and
  focused
• Over-communicate and get at fears right away

27
In Summary

• Technology is a tool to help with your chronic
  disease management/quality improvement program
• When evaluating technology, focus on its DNA as
  much as usability and features
• Keep things simple, easy and effective
• Ensure you are buying for both today and the
  future
• It will work! It will improve care delivery! It
  will have a positive return on investment!

28
References

• 1 Phillips R, Klinkman M. Health IT to Support
  the Patient-Centered Medical Home
  www.ncvhs.hhs.gov/071127p1.pdf
• 2 Jollis JG, Ancukiewicz M, DeLong ER, Pryor DB,
  Muhlbaier LH, Mark DB. Discordance of Databases
  Designed for Claims Payout versus Clinical
  Information Systems Implications for Outcomes
  Research Ann Intern Med. 1993 Oct
  15119(8)844-50.
• 3 Peabody JW, Luck J, Jain S, Bertenthal D,
  Glassman P. Assessing the Accuracy of
  Administrative Data in Health Information
  Systems Med Care. 2004 Nov42(11)1066-72.
• 4 Langley J, Beasley C. Health Information
  Technology for Improving Quality of Care in
  Primary Care Settings. Preparted by the
  Institute for Healthcare Improvement for the
  National Opinion Research Center under contract
  No. 290-04-0016. AHRQ Publication 07-0079-EF.
  Rockville, MD Agency for Healthcare Research and
  Quality. July 2007
• http//healthit.ahrq.gov/portal/server.pt/gateway/
  PTARGS_0_1248_661809_0_0_18/AHRQ_HIT_Primary_Care_
  July07.pdf

29
References

• 5 White K. Improved Medical Care Statistics and
  the Health Services System Public Health Reports
  Vol. 82, No. 10, October 1967
• 6 Chute C. Moving Toward International
  Standards in Primary Care Informatics.
  www.ahrq.gov/research/pcinform/dept3.htm
  November 1995
• 7 Austin B. A Tour of the Model Clinical
  Information Systems and Decision Support. Dec
  10 2007. www.improvingchroniccare.org/downloads/re
  designing_chronic_illness_care__the_ccm.ppt
• 8 Schellhase KG, Koepsell TD, Norris TE.
  Providers' reactions to an automated health
  maintenance reminder system incorporated into the
  patient's electronic medical record J Am Board
  Fam Pract. 2003 Jul-Aug16(4)350-1.