Principles of Health Data Analytics

&
EHR
2
3
4
5
Healthcare Data Sources and Basic
Analytics
• Electronic Health Records
• Biomedical Image Analysis
• Sensor and Biomedical Signal Analysis
• Genomic Data Analysis
• Clinical Text Mining
• Mining Biomedical Literature
• Social Media Analysis

6
Clinical Text and Data Mining
o Data is what we collect and store, and knowledge is
what helps us to make informed decisions.
o The extraction of knowledge from data is called data
mining.
o Data mining can also be defined as the exploration and
analysis of large quantities of data in order to discover
meaningful patterns and rules.
o The ultimate goal of data mining is to discover
knowledge. 7
Why Not Traditional Data Analysis?
1- Tremendous amount of data, Algorithms must be highly scalable to handle
such as terabytes of data
2- High-dimensionality of data
– Micro-array may have tens of thousands of dimensions
3- High complexity of data
– Data streams and sensor data
– Time-series data, temporal data, sequence data
– Structure data, graphs, social networks and multi-linked data
– Heterogeneous databases and legacy databases
– Spatial, spatiotemporal, multimedia, text and Web data
– Software programs, scientific simulations
8
Data Mining: Confluence of Multiple Disciplines

9
Typical Data Mining System

10
Artificial Intelligence (AI)

o AI in healthcare is the use of complex algorithms and

software to emulate human cognition in the analysis of
complicated health related data.

o Specifically, AI is the ability for computer algorithms to

approximate conclusions without direct human input.

11
Machine Learning (ML)

o ML is a sub-field of AI, ML is a data driven approach focus

on creating algorithms that has the ability to learn from the
data without being programmed.

o Example: the Child and the hot stuff.

12
 Machine Learning (ML) Tasks

The most popular ML tasks are:

o Classification: to predict a class of an object (for example
spam emails based on the content, or female/male based
on the web activity).
o Regression: to predict a continuous value for an object (for
example Colorectal Cancer incidence for next year).
o Clustering: to group similar objects together.
o Data Mining.
13
14
AI Application in Healthcare
Clinical decision support system is an application that

uses pre-established rules and guidelines that can be

created and edited by the healthcare organization, and
integrates clinical data from several sources to
generate alerts and treatment suggestions”.

15
Electronic Health
Record (EHR)
Purpose of a Patient (Medical) Record
Recall observations
Inform others
Instruct students
Gain knowledge
Monitor performance
Justify interventions
Away for communicating between staff

17
What is EHR
o EHR is the systematized collection of patient and population
electronically-stored health information in a digital format. These
records can be shared across different health care settings.

o EHR is a real-time, patient-centered documentation software

intended to make health info available instantly and securely to
authorized users whenever and wherever they need it.

18
Evolution of EHR

o Computerized Medical Record.

o Computer Based Patient Record.
o Electronic Medical Record (EMR).
o Electronic Health Record (EHR).

19
Electronic Medical Record (EMR)
Electronic Medical Record: An electronic record of
 health-related information on an individual that can
be created, gathered, managed and consulted by
authorized clinicians and staff within one healthcare 

organization.

20
Personal Health Record (PHR)
 Personal Health Record: “An electronic record of health

related information on an individual that conforms to nationally

recognized interoperability standards and that can be drawn from
multiple sources while being managed, shared and controlled by
the individual.” 
 The PHR is a collection of health information by and for the patient.

There is overlap between the EMR and the PHR.

21
The Relationship Between EHR, PHR, EMR

 The EHR is the larger system

that includes the EMR and PHR

and interfaces with multiple
other electronic systems locally,
regionally and nationally.

22
EMR vs EHR
EMR EHR
A record of medical care created, A repository of individual health
managed, and maintained by one records that reside in numerous
health care organization (intra- information systems and locations
organizational) (inter-organizational)

Integration of health care data from a Aggregation of health-related

participating collection of systems information into one record focused
from one health care organization around a person’s health history, i.e.,
a comprehensive, longitudinal record

Consulted by authorized clinicians Consulted by authorized clinicians

and staff within one health care and staff across more than one health
organization. care organization

Data continuity throughout one Data interoperability across different

health care organization organizations

23
Main Components of (EHR)
• A collection of related fields forms a record.
• A collection of related records is a file.
• The numbers 120/80 are a form of data, but when we define the
numbers as a patient’s blood pressure reading then they become
information.
• Health data is health facts that are collected about a patient or group
of patients that describe a health issue.

24
Categories of (EHR)

 Administrative: such as patient scheduling,

admission and registration.
Clinical: include the collection, storage, and display

of clinical information.

25
Electronic Health Record Key Components
The following components are desirable in any EHR system. The

reality is that many EHRs do not currently have all of these

functions.
 Clinical Decision Support Systems (CDSS) to include alerts,
reminders and clinical practice guidelines. CDSS is associated with
computerized physician order entry (CPOE).
 Secure messaging (e-mail) for communication between patients and
office staff and among office staff.
 Referral management feature.

 Retrieval of lab and x-ray reports electronically.

 Retrieval of prior encounters and medication history.

 Electronic patient encounter.

 The ability to input or access information electronically.

 Remote access from the office or home.

 Electronic prescribing.

 Problem summary list that is customizable and includes the major

aspects of care: diagnoses, allergies, surgeries and medications.
 Backup systems in place.
What is CPOE
 CPOE is a computer application that enables a provider to place

patient orders via the computer for further processing.

 CPOE is also an electronic prescribing system.

 Coupled with a clinical decision support system, CPOE has the

capability of applying rules-based logic to assist the provider with

making optimal ordering decisions.
 CPOE will manage patient care more effectively, will reduce medical

errors and improve healthcare delivery efficiency.

28
Why EHR ?
o Paper records are severely limited
o Expensive to copy, transport and store.
o Easy to destroy.
o Difficult to analyze and determine who has seen it.

o The need for improved efficiency and productivity

o The goal is to have patient information available to anyone

who needs it, when they need it and where they need it.

29
o Enhance Quality of care by enhancing patient safety, this

enhancement comes through:

o Improved legibility of clinical notes.

o Improved access anytime and anywhere.

o Reduced duplication of tests.

o Reminders that tests or preventive services are overdue.

o Clinical decision support that reminds us of patient allergies, the

correct dosage of drugs, etc.

30
o Enhance Quality of care and improve patient safety, this

enhancement comes through:

o Improved legibility of clinical notes.

o Improved access anytime and anywhere.

o Reduced duplication of tests.

o Reminders that tests or preventive services are overdue.

o Clinical decision support that reminds us of patient allergies, the

correct dosage of drugs, etc.

o Time Saving, Effort and money saving.

31
Barriers to Electronic Health Record Adoption
The most important barriers against EHR adoption are:
o Financial Barriers.

o Physician resistance.

o Work flow changes.

o Integration with other systems.

o Lack of standards.

o Inadequate proof of benefit.

o Privacy and security concerns.

32
EHR Basic Functions
o Health Information and Data (Problem Summary List) 

o Demographics.

o Insurance information .

o Lab results and graphing capability.

o Vital signs and graphing capability.

o Clinical reminders and alerts.

o Future and past appointments.

o Medical problems.

o Allergies.

o Medications.
33
Thank You