National Efforts for Clinical Decision Support (2nd)

Download Report

Transcript National Efforts for Clinical Decision Support (2nd)

National Efforts for Clinical Decision
Support (CDS)
Erik Pupo
Deloitte Consulting
Public Health as a CDS Variable
Three dominant myths in CDS usage
• Clinicians will use knowledge-based systems
if the programs can be shown to function at
the level of experts
• Clinicians will use stand-alone decisionsupport tools
• Diagnosis is the dominant decision-making
issue in medicine
– What does this patient have vs. what should I do
for this patient?
Applying Diagnosis to Public Health Alerts
National Focus – Standards and Knowledge
Artifacts
• Most existing CDS systems and their knowledge
bases have limited portability
• There is a need for a universal format(s) for CDS
knowledge that can be written once and
imported anywhere
• There is a need for a universal format for
encapsulating and accessing CDS capabilities
as a software service
National Efforts to Date for CDS
• Health eDecisions (development of HeD as a
standard for rules, alerts and guidelines)
• Clinical Quality Framework
– Clinical Decision Support (CDS) and electronic Clinical
Quality Measurement (eCQM) are closely related, share
many common requirements, and are both in support of
improving health care quality.
– Both need the ability to identify cohort of patients based on
logical combinations of patient data.
• CQM measures adherence to a standard plan of care.
• CDS guides a physician to follow a standard plan of care.
– Meaningful Use requires implementation of CDS rules to
improve the outcomes of certain eCQMs
Public health and CDS – what’s missing
• CDS alerts need to support multiple perspectives
• The standards used for the electronic representation of
CDS were not developed in consideration of public
health, and use different approaches to patient data and
computable expression logic.
– Technical and functional usability
• It is currently difficult to share logic between a public
health department and the CDS rules in a CDS system
or in an electronic health record (EHR)
– Lack of semantic interoperability across common CDS and EHR
data elements
Types of expected CDS inputs and outputs –
CDS vs Public Health
Example Inputs
Example Outputs from a CDS
Alert
What would a public health
alert look like?
Patient age,
gender, past
health
maintenance
procedures
List of health maintenance
procedures due or almost
due
Possible contamination of
medical equipment in a
acute care hospital
Medication
identifier, age,
gender,
weight, serum
creatinine level
Patient
summary
Recommended maximum
and minimum doses for
medication given patient's
estimated renal function
Possible bioterrorist attack
using existing medication
supply
Wide range of care
recommendations
Analysis of symptoms and
concerns
What’s missing – public health usability for
EHR and CDS
• Like with the thinking on CDS,
the use of public health alerting
mechanisms needs to manage
the usability of existing state
and local public health
workflows
– Technical reporting
– Functional usage of EHR
• National focus on public health
cognition and design of EHRs
to support public health
functions
What’s missing – Public Health and EHR
Terminology Harmonization
•
Encoding allows for alerts
– Granular encoding in
medication ordering allows
for alerts by drug name
– Granular encoding of
specific health concerns or
symptoms allows for alerts
by diagnosis name
•
Same thinking needs to
apply to public health
– More encoding of public
health events and vice versa
•
Harmonization of clinical
quality and clinical decision
support data models at the
national level
Summary – Multiple Efforts on Multiple
Fronts needed at the National Level
• Alerts
– Standardization of EHR formats for alerts need to
include public health input
• Rules
– Development of reusable rule sets to support public
health
– Common modeling of clinical decision support rules
using a public health “cognitive” model
• Guidelines
– Development of reusable guidelines from state and
local public health best practices