Monday, 30 November 2015

Health Informatics on FHIR

I'm taking this course offered by Georgia Institute of Technology, taught by Mark L. Braunstein, MD and I hope to share my class notes with all those interested. I'll also be taking other courses periodically and updating the blog as I go along.

Feel free to check out some of my notes and writings.

Health Data & Interoperability Standards

SNOMED-CT

Summary: 
Classifications vs. Ontologies
Basics of SNOWMED -CT

Classification: a list
Ontology: encompasses relationship among lists
- ICD-10, LOINC & SNOWMED-CT all include relationships

SNOWMED-CT relationships
Example: Heart Attack

311,000 concepts and 1.3 million relationships in this system. Very complex.


OID
Classifications of standard - ensure that operators know which database or system to be used/is using

MESSAGING STANDARDS

Outline:
Rationale for HL7 messages
Basics of HL7 messages
Recognize difference btween EDI/X12 (v2) and XML (v3)


Messaging standards - Hospital Order Entry/Results reporting

Sample message (EDI/X12) lab message from lab back to ward

Sample message (XML) same message
- easier viewing for people
- shown on browser (EDI/X12 before browsers existed)




DOCUMENT STANDARDS

Outline
Understand purpose of document standards
understand basics of HL7 CCDA


Key justifications: transition of care - errors made mostly between transfer from specialist to specialist. 80% of medical errors happened during the hand off between medical providers.

HL7 Consolidated Clinical Document Architecture


Objective data - physician done by inspecting patient.

Sections
 - entries

Reference Information Model of Healthcare (RIM)

Proposed stage 3 - providers must send electronic document for over 50% and receive 40%. Must use data in document to reconcile medications etc.


User interface. Can be used as portal for patients to see their own data too.

ADVANCED STANDARDS

Clinical Decision support
-  guide decisions for physicians and deliver most efficient care.

Example: ARDEN syntax


Blood thinner

Example: CIMI Clinical Information Modeling Initiative


Context management specification (CMS)
- Allow independent developed applications from many systems to synchronize and coordinate.


FHIR STANDARDS

Outline
Understand rationale for FHIR
Have basic understanding of web services and a simplified data model

To solve the problem of interoperability

Rise and Fall of HL7
- Eliot Muir: Complicated standards can be pushed for a while but ultimately market reject them. Even government will reject them." Cost too high. Too complex.

JASON AHRQ April 2014
- Should move to API based approach

FHIR - Fast Healthcare Interoperability Resources
- API (application programming interface) for exchanging Eletronic Heatlh records.
- Eaiser to implement because uses modern web-based suite of API technology
- represented by JASON or XML
- Can search patient, all of patient's problem
- Implication: can ensure real-time data gathering
- CRUD - create, read, update and delete


Example

Cool website - FHIR resource builder: http://clinfhir.com/

SMART on FHIR Apps


Arthritis App - get patient feedback, show trends and treatment effectiveness


Some also designed for patients: inform on risk etc.


Health Data & Interoperability Standards 1

Health Data and Interoperability Standards


Why is there a problem?















There is a problem because of the difference between syntax and semantics.


Example: National Drug Code system (FDA)




Medication is mainstay of all dieases so standardization across all of them is very important.

National Library of Medicine
      RxNorm (RXCUI unique number) to access the drug names
      UMLS – unified medical language system



Standards for Medical problems
Brief history: 
-                    Mid 1400 Northern Italy – Death certificates – name & age of deceased, cause of death certified by physician
-                    Black death 1629-1631: Death certificates used as source of data for analysis
-                    1661 London – Captain John Graunt interested in why children die – turned to death certificates to learn
-                    Massachusetts 1639 – death certificate made available for analysis
-                    London 1839 William Farr discovers interoperability
o   Each disease described in many ways, many terms (this is a problem!)
-                    London 1851 Great exposition, demonstrate top technology. Difficult to compare because no standards
-                    Chicago 1893, International List of Causes of Death (Bertillon Classification) He grouped disease by what we recognize today
-                    Denmark 1899, ICD (international classification of diseases) Decennial revisions, 10th in 1994, first disease “ontonogy” US want to use it in Oct 2015, only advanced industrial country to not use it.
-                    ICD-10: Radical departure. Introduce content model to facilitate automatic coding of diseases. More specific and detailed.




Another example:




Result: vast increase in complexity. 16X more codes. Lots of complaints from providers.

Example:

 














LOINC Codes – Logical Observation Identifier Names and Codes
            - developed by Regenstrief Institude in Indiana, leading health information center in US.



Friday, 20 November 2015

Personalized Health - Ivey Whitepaper

Personalized health

Ivey Whitepaper

Trends
  1. Consumers getting more empowered - through apps


Lessons
  1. Banking industry - personalized portfolio, multi-channel with online banking

Implications

  1. Focus on patient not the procedure
  2. Define value in terms of what people care about
  3. Put people behind decisions
  4. Tailor fit solutions to patients
  5. Connectivity
  6. More collaboration, less competition
  7. Provide personalized health information - enable better decision making
  8. Provide care differently to different segment

Heatlh Informatics (Georgia Tech) Week 1 cont'd

SYSTEMATIC ISSUES (explained by IOM)

Institute of Medicine
 - theres a quality gap
REASON  - too few ppl trained
REASON 2 - fragmented


Network view of primary care (georgia tech developed)
Average patient treated see average 14 different care provider
Interact with 229 providers
--> Paper records not good

Multiple conditions
Poor coordination: multiple conditions has 25-40% has greater odds of reporting errors
50% who seen 4 more physicians report error. --> laspe in communication during care transitions

Part of the Solution
Information Technology (IT) is not fully utilized.

IOM created a vision
Adoption - interoperability - analytics
>> creates learning health system


FEDERAL PROGRAMS:
OVERVIEW
Medicare and Medicaid.

HIT adoption historically low @ 1.5% hospitals, 4% of physicians (2008 and 2009)
2004 George Bush - raised the HIT issue as national priority "reduce cost, improve care". Goal to hit universal adoption by 2014.
President Obama - move incentives toward create better results (value based health care)
30B set aside for HITECH - to spur adoption of electronic records.
- 20B create medicare/medicaid incentives
- 2B office of the national coordinator

ONC - Office of National Coordinator
Key Programs: EHR Certification
- meaningful use
- health info exchange
- regional extension centers
- standards and interoperability
- research and demonstration projects

Programs to spur adoption
 - Vendors have to show qualification: EHR certification
 - Meaningful use
 - incentive payments (medicare and medicaid)

EHR CERTIFICATION

Quality improvement
- electronic prescribing, drug-drug,drug allergy interaction check, medication reconciliation
- provider order entry, patient reminder, patetient specifci education resources, automated measure calculation
- calculate and submit care quality

Need to show key qualities of care coordination
- electronic copy of health information
- timely access
- clinical summaries
- exhange information & patient summary record

Public health
 - submission to registries (cancer, reportable disease)
- electronic surveillance (epidemics, bioterrism)

EMR certification - how many certified?
 - professional's office
 - hospitals


All the systems designed by different standards - interoperability issue (how to make them talk to each other)

Meaningful Use
/eligible providers/
- physicians
- nurse practitioners
- certified nurse - mid-wife
- dentists
- physicians who practice in federally qualified health center, rural health center, led by physician assistant
- doctors of optometry


Meaningful Use - 3 stages
Stage 1: 14 core, 5 menu items - almost all have it
Stage 2: priority for now
Stage 3: still deciding

Healthcare Quality
(defined IOM)
The degree to which health service for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge"

One of most diagnosed disease is diabetes
1. process: did the doctor do the test?% of diabetics having the test
2. outcome: adequate control? % of diabetics above threshold

Chronic disease - incurable

Stage 1: at least half of patients must have online access to view record
Stage 2: 5% patients actually do that.
Stage 3: e-prescribing, clinical decision support (computer help physician avoid msitakes), computer-based physician order entry. Health information exchange, public health and clinical data registry reporting


INCENTIVE PAYMENTS

Two systems: medicare (no threshold) US care pay for 65+ citizens - carrot and sticks,
Medicaid (30% of patients, 20% for pediatricians) poor or disabled - just carrot

Medicaid no penalty - already among poorest providers, underserved. Mutually exclusive


carrot and stick. Better for early adopters, penalize late adopters. 


INCENTIVE REFORM

Value based reimbursement --replace-- number of procedures completed

What's wrong with financial incentives: The forces that have led to a global epidemic of over testing, over diagnosis, and over treatment are easy to grasp. Doctors get paid doing more, not less. We're more afriad of doing too little than of doing too much. And patients often feel the same way. They're likely to be grateful for the extra test done in the name of "being thorough" - and then for the procedure to address what's found. - Atual Gawande MD

Affordable care act (2010)
Medicare shared savings program will allow providers who voluntarily agree to work together to coordinate care for patients and who meet certain quality standards to share in any savings they achieve for the program.

Proof of concept
- 10 advanced sites
- 4 earned $29.4million reimbursement
- Marshfield Clinic earned half - health information technology

Accountable Care Organization (ACOs)
 - similar to Health Maintainence Organization
- potential savings

Pioneer ACOs - high performing sites
- 2 years of shared savings and shared losses
- successful program can move in year 3 to a population-based model

Value based care - not just the government
Anthem's Joe Swedish said it's value-based contracts include "enhanced payments for performance and shared risk or bundle payment arrangements."

Data matters!



Wednesday, 18 November 2015

Heatlh Informatics (Georgia Tech) Week 1


OUTLINE













INCENTIVE

1. Physicians get paid to do more.
 - incentivizes overuse of tests, technologies
 - make far more money doing procedures than interacting with patients

Structural difference between current and desired healthcare.



1. Specialist driven.
2. Fragmented and uncoordinated (due to specialist driven and extensive use of IT)

Patient-centered medical home (patient centric)
 - improve care coordination and communication


STRUCTURE

Problem 1: Single greatest cost of rising healthcare spending in the US is the growing prevalence of chronic disease.




Medicare provides to US citizen 65 or older.


Problem 2: Quality of our system appears to be lower than other OECD countries.

Good news: Chances of survival in US is greatest than anywhere else. (from health attack, acute, life threatening problems)

Bad news: Vulnerable elders receive 50% of recommended care with inconsistent quality, 210,000 preventable adverse events per year that contribute to death of hospital patients.

RESULT in other countries life expectancy higher than US.


Summary: US is putting a lot of money in healthcare, in order to reap greater profits. However, this does not always lead to the most efficient or best outcome for the patient, due to specialization and use of IT.

Health education through YT

Dr. Mike Evans uses youtube to deliver health topic white board sessions to the public audience.

https://www.youtube.com/watch?v=fqhYBTg73fw
http://www.evanshealthlab.com/