mmi 406: decision support systems and health care
TRANSCRIPT
![Page 1: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/1.jpg)
MMI 406: Decision Support Systems and Health Care
Osteroff, J. A., Pifer, E. A., Teich, J. M., Sittig, D. F., & Jenders, R. A. (2005). Improving outcomes with clinical decision support: an implementer's guide. Chicago, IL: HIMSS.
Contents Chapter 1: Identifying Stakeholders and Goals ............................................................................................ 2
Chapter 2: Cataloging Available Information Systems.................................................................................. 6
Chapter 3: Selecting and Specifying CDS interventions .............................................................................. 12
Chapter 4: Specifying and Validating the Details, and Building the Interventions ..................................... 18
Chapter 5: Putting Interventions into Actions ............................................................................................ 22
Chapter 6: Measuring results and refining the program ............................................................................ 27
![Page 2: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/2.jpg)
Page 1.1
Chapter 1: Identifying Stakeholders and Goals Key lessons
- A successful CDS program emerges from , and in turn supports, key organizational performance
improvement initiatives. Executive support for the program is vitally important
- Program goals are determined from a thorough environmental survey of pertinent initiatives
and stakeholders. A variety of external drivers and internal needs combine to drive priorities for
development of CDS interventions
- Governance and management structures and processes need to be established for the CDS
program elements which include intervention needs assessment, design, development, testing,
launch, and evaluation
- Champions – Key supporters who are trusted by your staff – are an important resource for
cultivating proper two way communication about, and support for, the CDS program
Page 1.2
People : The Keys to Success
- An essential first step in establishing a CDS program is identifying the key individuals,
committees, and positions upon which the program's success will depend, i.e. stakeholders
CDS Program Governance and Management
- Even in relatively modest CDS programs consisting of a few CDS interventions, there are
important decisions to be made and processed to be managed. CDS governance and
management structures are therefore required to fill these needs
- Developing a CDS program will often involve shifts in control and stakeholder interactions for
various care processes
- Ownership of process improvement by affected stakeholders should be fostered early in the
development of CDS governance and management structures, particularly for any stakeholders
that might be critical of the changes
CDS Program goals, and Clinical Goals and Objectives
- CDS program goals consist of strategic targets which the organization plans to address with CDS
interventions
Page 1.4
Generic outline of CDS program stakeholder groups and responsibilities
- Executive leadership: sets strategy , clinical standards, allocates resources
- Management / Oversight
![Page 3: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/3.jpg)
o Healthcare organization departments / functions: responsible for processes / outcomes
that will be affected by CDS program
Clinical departments (e.g. Laboratory, pharmacy, medicine, surgery
Organized medical staff
Cross cutting functions (e.g. quality, safety, disease management)
Clinical director of information systems / medical informatics
Other organizational committees and departments
o CDS oversight / Benefits realization: provides guidance / review of CDS initiatives
o CIO / IT steering: responsible for IT infrastructure for CDS
- Implementation / Project management: develops, deploys, monitors CDS interventions
o CDS specific: Overall responsibility to deploy / maintain CDS knowledge assets
o General IT: Overall responsibility to deploy / maintain clinical information systems
- End users and related positions
o End users: recipients of CDS interventions (e.g. Patients, nurses, pharmacists, physicians)
o Related staff: generate data for, or are affected by , interventions
o Subject matter experts: clinical authorities for content in CDS interventions
o Clinical thought leaders and champions (i.e. clinicians respected and listened to by
colleagues): promote dissemination
CDS program components: Steps and responsible parties
- Identify need for new / improved CDS interventions: Management / oversight with input from
executive leadership
- Design -> validate -> Develop -> test -> deploy CDS interventions: Implementation / Project
management with input from management, end-users, and related staff
- Evaluate intervention effect: Implementation / project management, using framework
developed by management / oversight and with attention to effects on end users and related
staff, and on the overall strategies, priorities, and clinical standards set by executive leadership
External Drivers
Quality Measures and Pay-for-Performance Programs
- Pay for performance and related clinical quality improvement initiatives rely on quality
measures. These measures address the extent to which healthcare services shown to improve
outcomes are actually being delivered to patients for whom they are indicated
Page 1.5
- An important step in determining objectives for a CDS program is a careful assessment of
external and internal quality initiatives that are affecting the organization
CDS Functional Requirements and HIT Product Certification
![Page 4: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/4.jpg)
- These efforts include healthcare information technology (HIT) product certification by the
Certification Commission for Health Information Technology (HIT), the Leapfrog CPOE testing
protocol, and minimum CDS requirements for electronic prescribing systems covered by the
Medicare Modernization Act
- The Office of the National Coordinator for Health Information Technology (ONCHIT) is charged
with facilitating the effective use of information technology to improve the quality, efficiency,
and safety of health care for all Americans.
Internal Drivers
Page 1.6
- Support from clinicians who are concerned about specific clinical goals can help drive the
behavior change that will be required to deploy CDS interventions aimed at that goal
- Each CDS goal is broken down into more specific goals and objectives
CDS program successful factors
- Deep executive support for clinical quality improvement and belief in the value of information
technology to help achieve it
- A history of successful clinical information technology projects, fostering a strong belief among
clinicians and organizational leaders that CIS are valuable and desirable tools
- Excellent communication about the clinical and technological nature of the CDS program to all
involved stakeholders
- Involvement of key users and clinical champions well in advance of the implementation of any
new CDS program
- Strong support and rapid problem resolution by project staff before, and especially during, early
implementation
Page 1.8
Determining CDS Targets
- Sources for determining CDS targets
o Institutional analyses of quality, safety, patient satisfaction, cost, and regulatory
problems
o High level committees that prepare the overall response to environmental drivers
o Technology supported analyses and mining of local care and outcomes
o Interviews with clinicians, medical directors, and other stakeholders
o Survey assessing stakeholders' CDS related activities, needs, and priorities
o Direct observation of information needs in clinical settings
o Community based priorities and programs
Page 1.10
![Page 5: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/5.jpg)
Factors affecting the desirability of a CDS objective
- Clinical Objective Value Score = P+O+C+N+G-D-C
o P = Patient impact (individual population) (Quality, Safe, and cost effective care)
o O = Organizational Impact (Regulatory compliance)
o C = Clinical Impact (Enhanced workflow)
o N = Number of patients positively affected
o G = Gap between ideal and actual behavior pertinent to the intervention
o D = Difficulty associated with addressing the objective
o C = Cost of addressing the objective
Stakeholders Revisited: Identifying Champions
- Identifying and cultivating champions early in the process is an important component of success
- Champions have been defined as "the smallest group of person to whom everyone else will
listen"
Page 1.11
Working with Implementation Champions
- Engage these individuals to support the program
- A process should be put in place for developing and maintaining champions as a productive and
helpful resource for the CDS program
Not Everyone is a Champion
- It is essential to engage resistors and detractors in active dialogue
Page 1.12
Cost Justifying the CDS Program and Preparing to realize Benefits
Benefits Planning and Realization
- Key organizational steps to realizing benefits CIS
o Organizing for benefits with a benefits realization team
o Designing explicitly for benefits
o Planning for benefits by addressing change management issues during system
implementation
o Measuring the benefits
o Accountability for benefits
- Benefits realization team ideally includes individuals with technical expertise
- Creating a benefits realization function to underpin the team
Page 1.13
![Page 6: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/6.jpg)
Cost justification
Page 1.14
- The cost justification includes considering resources such as staff, dollars, time, and energy
currently applied in addressing a particular goal or objective.
- It also includes the results achieved from this investment, with the costs and results associated
with a more coordinated approach to applying knowledge resources to address the target
- Cost justifications and performance improvement targets should be based on available
literature when possible
Page 1.15
Step 1:
- Identifying and contact the key local committees, positions, and individuals currently in place
that will have a stake in the CDS program, either by proposing, validating, supporting,
communicating, or using the CDS interventions.
- Begin considering their current roles and activities and how they might relate to a more
coordinated CDS program.
- Consider also new positions and teams that might be needed to ensure the program's success
Step 2:
- Synthesize and validate a working list of CDS goals, and clinical goals, and objectives for your
CDS program. Define the baseline and target performance for the clinical objectives
Page 2.25
Chapter 2: Cataloging Available Information Systems Key Lessons
- The data types and features provided in the major CIS applications in your organization has
deployed help determine the types of CDS intervention that are available to you
- Powerful CDS is dependent on a reliable, fast, and usable CIS infrastructure
- Additional CIS features that impact CDS capabilities include availability of coded Data, use of
standard vocabularies, and ability to aggregate data from multiple sources
- There are six major types of interventions provided by CIS applications. Understanding these
different types, and their advantages and disadvantages, is critical in selecting and designing
optimal interventions to accomplish specific goals.
![Page 7: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/7.jpg)
Page 2.26
Discussion
- Strategic planning that involves IT and business and clinical leadership, and anticipates this
interdependent relationship between CIS and CDS, can help the organization navigate the
expensive and potential risky challenge of CIS procurement and deployment
- The CDS program will be most effective in addressing priority goals if it is approached as a key
strategic initiative. more clinical than IT, and is staffed and managed accordingly
Core CIS Elements
Basic Information System Infrastructure
- Hardware / Software
o Number of workstations / Terminals, Handheld/portable devices, printers, including per
bed, per clinician
o Quality: age, software / operating system version
o Reliability: e.g. % uptime
o Performance: speed of the core applications
- Network Connectivity
o Number and proportion of computers connected to internet and high speed internal
communication
- Wireless / Remote System
o Number and use of wireless devices
o Telemedicine infrastructure
o Remote access to information systems by clinicians
- Medical devices that generate patient data
o Number and type of medical instruments that gather and can deliver data about
patients
- Integration among clinical systems
o Number of different terminals / workstations required to access the full portfolio of
available applications
- Should also consider overall usability: User's baseline interactions with the underlying hardware
and software that will convey the CDS intervention
Page 2.27
CIS Applications
- CIS interventions are most effective when tightly integrated into workflow
- This involves incorporating them into the clinical information systems that underpin patient care
processes
CIS Applications pertinent to CDS Intervention
![Page 8: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/8.jpg)
- Departmental data management
o Pharmacy info system
o Lab info system / results reporting system
o Radiology information . results reporting system
- Clinical records and patient management
o EMR: ambulatory, inpatient, for patients
o Department-oriented (e.g. anesthesia, cardiology
o Care tracking systems: Emergency department, operating room
o Medication administration and documentation
- Ordering
o Computerized provider order entry
o E-prescribing systems
- Data aggregation
o Data warehouse
o Clinical data repository
o External and internal registries
- Clinical content
o Reference / knowledge sources for clinicians
o Health information for patients
o Health risk assessment tools
- Financial / administrative
o Charge capture system
o Billing system
o Scheduling / registration system
o Directories
Standard Vocabularies
- A robust CDS program will likely involve a variety of specific interventions delivered via several
different information systems
- Integrating the various components of DS interventions, such as pertinent patient data and
clinical knowledge, often require common underlying vocabulary and coding scheme
Page 2.28
- A structured vocabulary provides relationships among the concepts and details about those
concepts
- This permits automated aggregation of related data by the CDS system and intelligent
interpretation of those data
Page 2.29
- The Common Terminology Services (CTS) standard of HL7 provides a standard for exchanging
messages with a standard vocabulary in electronic format
![Page 9: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/9.jpg)
- Information systems can use the CS standard, in combination with appropriate vocabularies, to
translate from one code set to another and to look up a code given its name (and vice versa)
- Using consensus vocabularies facilitate reporting of data to regulatory and accreditation
agencies
- It also allows exchange of data among organizations without the need to translate from one
non-standard coding scheme to another
Some standard vocabularies pertinent to CDS
- SNOMED-CT: Systemized Nomenclature of Medicine Clinical Terms (Multiple Areas)
- ICD-9-CM: International Classification of Diseases Clinical Modification (Diagnoses)
- CPT-4: Current Procedural Terminology (Procedures)
- LOINC: Logical Observation Identifiers Names and Codes (Laboratory tests)
- HCPCS: Healthcare Common Procedure Coding System (Supplies)
- CDT: Code on Dental Procedures and Nomenclature (Procedures)
- NDC: National Drug Codes (Drug: Names)
- RxNorm (Drugs: Names + Dose forms)
- NDF-RT: National Drug File Reference Terminology (Drugs: Mechanism of action, physiologic
effects)
- HIPAA code sets (Billing administrative functions)
- (HL7) : Health Level Seven
Page 2.30
- Some HCO create a local vocabulary that maps local or CIS vendor codes to standard schemes
- Central clinical data repository (CDR) properties:
o What types of data are available
o How quickly is the repository updated after data appears in the individual source
system?
o Does the repository make use of standard vocabularies
o Are the data stored with their original vendor terminology
o Is the CDR data organized so that needed data of a given type from all sources can be
retrieved together
Page 2.31
- Some organizations maintain a distinct data warehouse into which data are stored periodically,
but not necessarily in real time
- Many create registries of information about patients with chronic diseases to help manage
these patients
Data Types and Key Elements:
- Patient centric: Medications, allergies, lab results
![Page 10: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/10.jpg)
- Practice centric / Site specific: Local antibiotic resistance patterns and sensitivities
Page 2.32
CDS Intervention Types
- New Taxonomy:
o Makes it much easier to identify the best type of intervention for meeting specific
classes of clinical objectives
o Easier to classify and apply research into CDS impact
Page 2.33
Clinical Decision Support intervention types
- Documentation forms / Templates:
o provide complete documentation for quality / continuity of care, reimbursement, legal
o Reduce errors of omission by displaying items for selection
o Provide coded data for other data driven CDS
o Subtypes
Patient self assessment forms
Nursing patient assessment forms
Clinical encounter documentation forms
Departmental / multidisciplinary clinical documentation forms
data flow sheets
- Relevant data presentation
o Optimize decision making by ensuring all pertinent data are considered
o Organize complex data collections to promote understanding of overall picture and to
highlight needed actions
o Subtypes:
Relevant data for ordering or documentation
Choice lists
Practice status display
Page 2.34
Retrospective / aggregate reporting or filtering
Environmental parameter reporting
- Order creation facilitators
o Promote adherence to standards of care by making the right thing the easiest to do
o Subtypes
Single order completers including consequent orders
Order sets
![Page 11: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/11.jpg)
Tools for complex ordering
- Time based checking and protocol/pathway support
o Provide support for multi-step care plans, pathways, and protocols that extend over
time
o Subtypes:
Stepwise processing of multi-step protocol or guideline
Support for managing clinical problems over long periods and many encounters
Page 2.35
- Reference information and guidance
o Address recognized information needs of patients and clinicians
o Subtypes:
Context insensitive
General link from EMR or clinical portal to a reference program
Context sensitive
Direct links to specific, pertinent reference needed
- Reactive alerts and reminders
o Provide immediate notification of errors and hazards related to new data or orders
entered by CIS user or the CIS itself, or passage of a time interval during which a critical
event should occur
o Help enforce standards of care
o Subtypes:
Alerts to prevent potential errors or hazards
Alerts to foster best care
Step 1:
- Conduct and document an inventory of your organization's CIS that could play a role in
delivering CDS interventions.
- For each system, note its pertinent CDS capabilities, coding systems and current usages
o Include in your survey all the different clinical content and knowledge resources that are
available in your organization to support clinical care and decision making
Page 2.36
o Your organizations clinical knowledge resources provide an initial content base for
achieving the clinical goals within the CDS program
o The knowledge asset survey provides a foundation for making this assessment
Page 3.41
![Page 12: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/12.jpg)
Chapter 3: Selecting and Specifying CDS interventions - Clinical objectives can be classified into objective classes. Different objective classes are
optimally addressed by a subset of the full spectrum of CDS intervention types. Categorizing
objectives into classes can help determine the best intervention to meet a particular need
- Different interventions are most effective for achieving specific objectives when delivered at
specific points in clinical workflow
- The final selection of the best intervention to meet a particular need is based on specific
selection factors reflecting availability of the intervention in your CIS, ease of implementation,
acceptability, and impact of each type of intervention
- Careful attention is required to benefits realization during intervention selection and
implementation. It is essential to remain sharply focused on the quantified care improvement
that is expected and ensure that the emerging interventions will deliver this benefit and enable
its measurement
Page 3.42
Clinical Objective Classes
- Most clinical objectives can be grouped into a few specific classes, where each class lends itself
to specific types of CDS interventions
- Mapping specific clinical objectives to objective classes, mapping those classes to CDS
opportunity points in workflow, and then selecting specific CDS interventions pertinent to the
CDS opportunity points can help you find the best interventions to achieve those objectives
Workflow Opportunities: Optimal Timing of CDS
- In the clinician and patient workflow, it is important to consider when are the most effective
times to deliver CDS interventions
- Different objective classes are best addressed with CDS interventions at different points in this
workflow
Page 3.43
Classes of objectives and specific examples
- Prevent Errors
o Prevent errors of commission
Decrease likelihood of prescribing an overdose of warafin for patients with liver
disease
Prevent allergic reactions, drug interaction
Reduce inappropriate repetition of tests
o Prevent errors of omission
Reduce inattention to critical lab values and other test results
Avoid missed steps in a necessary sequence
![Page 13: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/13.jpg)
Treat patients more rapidly once a condition or problem becomes apparent
- Optimize Decision Making
o Optimize choice of individual tests and therapies based on additional considerations
Improve compliance with on-formulary medications
Improve likelihood that less expensive drugs are chosen when appropriate
o Improve compliance with simple care guidelines
o Improve appropriateness of overall workup and treatment plan for a given situation
Page 3.44
o Improve compliance with complex short term multi-step protocols
o Optimize treatment of chronic conditions over time; long term management
- Improve Care Processes
o Improve documentation of care, including eliciting key data from history and physical
examination
o Improve patient education, empowerment, and satisfaction with care
o Improve communication among caregivers
- CDS can often be very effectively delivered to patients themselves, adding a powerful tool for
fully engaging patients in their own care between health system encounters
Mapping Objective Classes to Possible intervention types
Page 3.45
Deciding Among Several Available Intervention Types
- Figure 3-3 is a detailed framework for mapping clinical objectives to CDS intervention types that
can be applied at particular stages in the workflow
- Stages in workflow:
o Pre Admission
o Admission / Nurse assess
o Clinician's H&P and plan
o Documentation
o Orders
o Order handling
o Drug admin / Therapies / Procedures
o Discharge
o Post Visit / Home Care
o Results arrive
o Consult request
![Page 14: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/14.jpg)
Selection Factors
- Each objective class in Figure 3-3 maps to several different intervention types and points in
workflow
- Even after ruling out interventions that are not available in your environment, there may still be
a number of options available
- The analysis of care process issues should point to places in workflow where breakdowns occur
and where there are opportunities for process improvement
- If you can identify the steps in the care process where improvements is most needed, then
corrective knowledge interventions at those particular steps might be particularly helpful for
improving overall performance for the clinical objective of interest
Page 3.46
Objective classes and workflow opportunities
Prevent Errors:
- Prevent errors of commission
o Occur when an action taken is incorrect for some reason
o CDS Intervention Type:
Alerts to prevent errors
Relevant data for ordering
Choice lists
Single order completers
Support for complex ordering and calculations
Flowsheets
Page 3.47
- Prevent errors of omission
o When a needed test of therapy is not prescribed for a patient
o CDS Intervention Type:
Order sets
Single order completers and consequent orders
Clinical encounter documentation forms
Nursing assessment forms
Practice status display
Alerts to foster best care
Multistep protocol support / Clinical pathways
Page 3.48
Optimize Decision Making
![Page 15: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/15.jpg)
- Optimize choice of individual tests and therapies based on additional consideration (such as
cost-effectiveness)
o CDS Intervention Type
Order sets
Choice lists
Alerts to foster best care
Consequent orders
Reference material
o Here, the user is initiating a specific action, and CDS is used to optimize the specific
choices and parameters
o The intervention is most convenient when presented proactively
Page 3.49
- Improve compliance with simple care guidelines
o CDS Intervention Type
Clinical data flowsheets
Alerts and reminders to foster best care
Order sets
Multidisciplinary
Clinician encounter forms
Patient self-assessment forms
Reference information (context-sensitive preferred)
Page 3.50
- Optimize treatment of chronic conditions over time (long term management)
o CDS Intervention Type
Flowsheets
Alerts to foster optimal care
Aggregate reporting
Order sets
Patient self-assessment forms
Reference material
Multi-step protocol support/Clinical pathways
Page 3.52
- Improve appropriateness of overall workup and treatment plan for a given situation
o CDS Intervention Type
Reference information (context sensitive)
Order sets
Flowsheets
Clinician documentation forms
![Page 16: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/16.jpg)
Alerts to foster care plan
Aggregate reports
Clinical pathways
Page 3.53
- Improve compliance with clinical pathways, complex multistep protocols, and other very specific
multistep
o CDS Intervention Type
Clinical pathways
Alerts
Multidisciplinary documentation forms
Order Sets
Page 3.54
Improve Care Processes
- Improve documentation
o CDS Intervention Type
Nursing assessment forms
Clinician documentation forms
Patient Self Assessment forms
Multidisciplinary Forms
Page 3.55
- Improve patient education and empowerment
o CDS Intervention Type
Patient self assessment forms
Relevant data display
Multidisciplinary documentation forms
Alerts to foster better care
Order sets
Reference information (context sensitive)
Page 3.56
- Improve Communication among caregivers
o CDS Intervention Type
Multidisciplinary documentation forms
Alerts to foster best care
Multi-step protocol support
![Page 17: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/17.jpg)
- The final choice of which intervention(s) to use in addressing a specific need depends on several
factors:
o Ease of development: How hard or easy it is to develop the intervention
o Acceptability: Likelihood of the intervention being accepted by the intended decision
makers
o Impact: Likelihood of the intervention leading to the desired effect
o Special considerations about each intervention type may make some types more or less
desirable
- Figure 3-4 explores these selection factors for each intervention type outlined
Page 3.65
- These figures also intended as a framework for driving and analyzing research on the pros, cons,
and effectiveness of different types of CDS interventions in a variety of circumstances
Benefits Planning Revisited
- It is critical to keep in mind the specific benefits you are trying to accomplish as outlined
- This has been referred to as "designing for benefits" and involves making achievement of the
desired objectives an explicit criterion in intervention design
- Use case scenarios are useful for linking design with outcomes in this way
Use case scenarios
- Use case scenarios are step by step written descriptions of how users interact with a system and
the results of these interactions
- Use case scenarios for an intervention should cover several different examples of likely user
system interactions and can be divided into a Low, medium, and High complexity progression
Page 3.66
Measuring Benefits
- Early attention to these measurement issues can help ensure the most practical and efficient
path to the needed information , or prompt development of alternative measurement strategies
Step 1
- Assign an objective class to each clinical objective for which you are developing a CDS
intervention
Step 2
- Use the objective classes to help identify optimal CDS interventions and workflow
opportunities addressing each corresponding objective, given the capabilities of your available
information systems.
![Page 18: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/18.jpg)
Page 4.69
Chapter 4: Specifying and Validating the Details, and Building the
Interventions - There are many logistical details about an intervention that need to be carefully established to
ensure that it leads to the desired benefits. The nuances of who will receive the intervention
and exactly how it fits into their clinical workflow are among the keys to success
- Feedback on these details from stakeholders, including recipients and other s affected by the
intervention, should be sought and adjustments made as needed. Those accountable for the
interventions and their effects should be comfortable with the plan and its likelihood of success
- As the interventions are validated and built, make sure that required organizational processes or
structural changes are put into place
- Putting in place a robust plan for gathering and responding to feedback from those receiving or
affected by the interventions will help ensure that the rich dialog with stakeholders that was
cultivated during the development stages continues after launch.
Page 4.70
- A central goal of intervention development is to ensure successful integration into clinical
workflow
- Questions
o When and how is the intervention triggered
System initiated interventions include those triggered by new data or by the
passage of time
User initiated interventions include information delivered in response to a user
request
o What criteria are evaluated for system initiated interventions to determine whether
specific information is presented to someone after the intervention is triggered
o Where will the patient data come from that are required by the intervention logic
o What exact information will be delivered to cause the desired outcome
o How is the information delivered
Page 4.71
o Who will receive the information generated by the intervention
o What are the feedback channel whereby intervention recipients can communicate
concerns and input about intervention usability and value
Optimizing Intervention Specifications
Optimizing Effectiveness
- Provide clear and practical recommendations in any intervention that offers advice to the user
![Page 19: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/19.jpg)
- Link advice to action opportunities as much as possible to optimize translation of the
information into the desired action
- Link interventions to each other when appropriate
- Consider the cascade of events that the intervention is designed to initiate and ensure that the
organization is prepared for these events
- Consider the communication factors for interventions directed to patients
Optimizing Safety
- Consider unintended negative consequences that the intervention might have on the process
and outcomes of care
Page 4.72
- Develop a fail safe plan in the case the CIS underlying the CDS interventions become temporarily
unavailable
- Minimize intervention overload, especially with alerts
Optimizing Acceptability and Use
- Ensure that the intended recipients can access and act on the information with the greatest
ease and least disruption possible
- Provide evidence and rationales supporting any clinical recommendations
- Provide appropriate customization and tailoring of the intervention
- Provide incentives wherever possible for effective use of CDS interventions
Special Considerations for Alerts
Page 4.73
- When is the most opportune time to deliver the alert
- Alert escalation: Who will get the message if an unsolicited alert is delivered during a time when
the intended target is not available to receive it?
- How long will an intervention be available to the recipient if it is stored in a patient's electronic
record
- How urgent is the intervention
- Which medium will be used to convey the message
- What is the expected average and maximum number of alerts per patient and clinician?
- Will unwanted interactions among alerts occur because of their number and nature?
Page 4.74
- What is the proper notification threshold to maintain adequate sensitivity yet minimize the
negative workflow effects
![Page 20: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/20.jpg)
- Is it appropriate and technically feasible to allow recipients to opt out of specific alerts in
general or for a specific patient
- When and how can previously issued alerts be retracted
- How are alerts dismissed or overridden
Use cases and organizational issues
- The use cases can help you visualize how these issues will affect user interaction with
intervention, which can in turn be used to help refine the use cases if needed
Involving Stakeholders
- Those who will affected in some way by the CDS interventions are in a great position to help
filter all the considerations into the intervention specifications
Page 4.75
- Pay careful attention to resistors and detractors as their concerns can be a vital source of
information about deployment issues that might interfere with the intervention success
- Transparency is key
- Maintaining broad representation and a transparent process for specification development can
help ensure that any remaining detractors are seen as not reflecting the organization's overall
commitment to the intervention
Validating the Specifications
- After addressing the above considerations in your intervention specifications, you are ready to
validate the plan with stakeholders
Accountability for CDS Results and their Measurement
- A picture of the responsible parties for your individual and collective CDS interventions begins to
emerge from your organization's version of the CDS roles and responsibility
Page 4.76
- An organization wide benefits realization team made up of individuals with a broad array of
pertinent expertise can be a much more powerful approach
- The team ideally should possess expertise in clinical , technical, administrative, financial, and
medical informatics related health services research areas and should report to a group of key
executives
Building the Interventions and Feedback mechanisms
Content, Mechanics, and Organizational Issues
Page 4.77
![Page 21: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/21.jpg)
- Intervention development will consist primarily of utilizing CDS features provided by your CIS
and CDS vendors and adding content as needed
Feedback Mechanisms
- It is important to build feedback mechanisms for assessing the use and response to the
interventions, particularly unsolicited ones such as alerts
- Other desirable feedback mechanisms
o Prompt implementation team responses
o Clearly informing users as to when they can expect to see their suggestions incorporated
into the application, if a change is needed
- Proactively seeking and encouraging feedback can lead to much less aggravation
- Channels for gathering feedbacks
o Direct observations of uses interacting with the system
o Subjective user feedback
o Input from clinical champions
o Objective measurements of intervention usage
Page 4.78
Project Management
- Project management is essential for interventions with any complexity and / or significant
workflow impact
Step 1
- Define and document the parameters and logistical details for each chosen intervention to
optimize stakeholder acceptance, ease and cost of implementation, effect on workflow, and
magnitude of benefit. Document these specifications
o The completed form is used to obtain validation and approval by key stakeholders in the
proposed plan and then to guide those who will be responsible for building the
interventions
Page 4.79
Specification form for validation
- Intervention Name
- Clinical Objective
- Desired Action
- Baseline performance
- Desired outcome
- Associated interventions focused on objective
- Workflow step
![Page 22: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/22.jpg)
- Specific CDS intervention and pertinent CIS applications
- Approach
Page 4.80
- Clinical background
- Selection criteria
- Exclusion criteria
- Target population for intervention
- User Interface
- Monitoring
- Evaluation
- Primary Stakeholders
- Clinical champion for this project
- Urgency / required delivery time
- Whose jobs do you expect to be affected by this project
- What are possible adverse consequences of implementing this project
Step 2
- Validate the proposed CDS interventions with appropriate stakeholders and obtain necessary
approvals
o Ensuring that the proposed CDS interventions, both individually and collectively, are
acceptable to key stakeholders and recipients, fit into workflow, are likely to achieve
their desired outcome, and will not adversely affect other healthcare processes
Step 3
- Develop the planned CDS interventions. Build in technological and people based mechanisms
for gathering and processing feedback from intervention recipients
Page 5.85
Chapter 5: Putting Interventions into Actions - Careful and complete testing of all new CDS functionality , along with other CDS and CIS systems
that might be affected, is essential to ensure that the completed interventions performed as
expected. The details of this testing depend on the type of CDS intervention, the nature of the
clinical content, and the underlying technology used to create the CDS intervention
- Make sure that everyone who will use, or might be affected by, the new CDS functionality is well
informed about project status and upcoming events to minimize surprises and optimize their
constructive participation
![Page 23: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/23.jpg)
- Anticipate users' training needs and provide training opportunities that fulfill them as
conveniently as possible for the users
- Consider the nature for both the CDS intervention and the clinical and technical environment
when determining the speed, scope, and order for rolling out new CDS interventions
Page 5.86
CDS Intervention testing
- The details of your testing approach will vary somewhat depending on the type of CDS
intervention, the nature of the content, and the underlying technology of the CDS intervention
Quality Assurance approach
Use Cases and testing Scenarios
- Typically, one or two basic use cases will be developed that relate to the intervention's primary
objective.
- The basic use cases should derive from business needs
- The others result from taking a hard look at the basic case in light of rich clinical knowledge and
practice experience
- Use case variants are an important way to catch needed design changes before launch
Page 5.87
- You can now develop the narrative use cases into formal testing scenarios
- These scenarios describe exactly how the user will interact with the intervention and the results
of that interaction
Unit testing
- Software developers perform this testing to ensure that the application will deliver the CDS
intervention as it has been built and that it functions as designed, independent of other parts of
the CIS
- It involves applying the testing scenarios to ensure that valid inputs produce valid and
appropriate outputs
- Unit testing for more complex applications may involve the use of special software for
debugging and testing.
- This testing is typically the responsibility of the application developers, not the software QA staff
Integration testing
- Systems developers and administrative staff ensure that all of the individually developed system
components work as designed when integrated with the existing CIS infrastructure
- Regression (verification) testing might also be necessary when the new intervention involves
significant modification to the underlying CIS.
![Page 24: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/24.jpg)
- This selective testing ensures that all new software is working as expected and that no new bugs
have been introduced that could result in the failure of previously working functions
User Acceptance testing
- This final pre launch testing is performed by clinical end users to ensure that the system works
for users as intended and meets all of the planned business and clinical requirements
- Representatives of all the pertinent user groups should be included
Beta testing / pilot launch
- For very complex or intrusive interventions, it can be useful to release the intervention initially
to a few selected clinical users
Page 5.88
Full Live Evaluation
- The feedback channels will serve as the means whereby the implementation team can assess
use response to the system
Launch Communications
- Communications about the launch set the tone for the acceptance
- Historically, lack of sufficient communication during these early stages has been one of the most
common causes of problematic and failed CIS and CDS implementations
- Launch related communications should be two-way. It is important to not only to convey the
CDS goals, strategies and tactics, but also to listen for concerns and opportunities to improve
the approach
Page 5.89
- Champions can play a crucial role in communicating with the user community about new CDS
interventions
- Clinical proponents of an intervention who are thought leaders should be an initial source of
communication to all other affected users before launch.
- Maintaining and using effective communication channels is an ongoing process
- It is essential to be alert for any generalized negative response to intervention components and
for individual detractors
Training
Page 5.90
- Investigate training resources provided by your CIS and CDS content vendors
- Consider a train the trainer approach
- Make training documentation available prior to training sessions
![Page 25: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/25.jpg)
- Schedule training conveniently
- Tailor training duration and intensity to user needs and constraints
- Pay careful attention to expressed concerns and non verbal communication during this training
- Remember that most "training" will occur after launch
Intervention Rollout
- In some cases, CDS intervention will be launched at the same time as these systems
- Waiting until the underlying CIS infrastructure is solidly and successfully in place before
overlaying potentially problematic CDS interventions is the prudent course in many cases
Page 5.91
- More complex or multi-pronged initiatives might require a phased launch
- For interventions such as alerts, it can be helpful to launch them initially with a small number of
users or in selected locations
- Such limited live testing can uncover problems not previously identified from development
testing
- Factors to consider when choosing pilot launch locations include
o How representative they are to the environments
o Size
o Availability of support staff
- The overall goal for the rollout plan is to simultaneously optimize the speed at which benefits
are realized and the magnitude of those benefits
- Specific interventions that offer the greatest prospect for returns with the least disruption are
logical starting points
- In instances where the policy might be somewhat unpopular, consider decoupling its
enforcement with the corresponding CDS intervention launch
Page 5.92
- It is wise to have some type of celebration to mark a successful launch
o It serves as a milestone event that makes all users aware that this implementation
round is essentially complete
o It can be a time to review expected intervention benefits with the user community
o It can help reinforce the partnership between the implementation teams and the user
community on using the CDS program as a helpful tool in achieving key objectives
Step 1:
- Before launch, test the CDS intervention content, mechanics, and logistics
o Make sure that use cases still accurately reflect the important business and clinical
needs implicit in the clinical objectives
o the testing scenarios will effectively challenge the interventions with these needs
![Page 26: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/26.jpg)
o Verify that any processing done to incorporate clinical knowledge into the CDS
intervention has not changed its clinical meaning or effect
Step 2:
- Develop and implement an intervention rollout plan that addresses user communications,
training, and feedback, as well as responsibility for intervention monitoring
o Make sure that by the time any interventions are launched, ongoing routines to monitor
and support the intervention's content and delivery mechanics are already in place
o The plan should include the party responsible for the intervention and an initial
monitoring frequency
Page 5.93
- Pre-launch Testing Worksheet
o Intervention type
o Intervention Name
o Test Scenario
o Date / Tester
o Results
o Notes
Page 5.94
- Intervention Launch plan
o Clinical Information System
o Intervention Name
o Communication / Training Plan (Pre Implementation)
o Feedback Plan (Post implementation)
o Intervention Maintenance Plan
o Notes
Page 5.95
o A key component of the rollout plan is whether and how interventions will be launched
in phases
Step 3:
- Gather, document, and address user feedback before, during, and after intervention rollout
o Help prioritize intervention enhancements into predictable schedule
o Help communicate to users that their input is being heard and taken seriously
![Page 27: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/27.jpg)
o Can be used in discussions with management to help justify resource needs for ongoing
system enhancements
- Implementation Status
o Information System
o Intervention Name
o testing Started
o testing Complete
o Planned launch Date
o Actual Launch Date
o Locations /users
Page 5.96
- Feedback Issues and resolution
o Information System
o Intervention Name
o Feedback date and User
o Feedback
o Plan / Resolution
o Target Date / Actual Date
o Priority
Page 6.99
Chapter 6: Measuring results and refining the program - Key stakeholders and organizational leadership can help establish reasonable measurement
intervals and expectations for improvement. Evaluating intervention effectiveness requires
both quantitative and qualitative approaches
- Log files and other approaches to tracking intervention use can be helpful for monitoring
intervention effects. Details of interest include when and how the intervention was invoked,
where the user and patient were at the time, when in the workflow it happened, and how the
information delivered was handled and applied
- Plan an iteratively refining interventions to improve their use and benefits. Opportunities will
unfold based on lessons learned and data amassed during each implementation round, evolving
capabilities in your CIS infrastructure and available CDS tools, new clinical knowledge, and the
changing environment in healthcare
- It is important to have a systematic approach to managing organizational knowledge assets.
This includes policies (e.g. covering periodic content review) and infrastructure (e.g. committees
and knowledge management tools) to ensure that the quality, currency, and appropriateness of
all the content in CDS interventions
![Page 28: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/28.jpg)
Page 6.100
- The effects of CDS interventions must be analyzed carefully to ensure that the considerable
resources required for their implementation yield the intended results
- This analysis will demonstrate whether interventions are being used as expected and as
appropriate
- It can help quantify return on investment
Evaluating Alerts
Tracking Alert Firing and User responses
- Track the details of when and how the alert is invoked or fired
o Details of what exactly is triggered
o Where the user and patient are at the time
o when it occurs in the workflow
o how the information delivered is handled
- There will likely be a significant number of rejected alerts in the log
- The implementation team can examine alert override reasons for patterns among various
clinicians or even entire departments
Page 6.101
Managing Organizational CDS Knowledge Assets
- Ensuring the quality, currency, and appropriateness of the collection of content assets is
essential to keep the knowledge intervention safe and consistent with best clinical practices
Currency and Appropriateness
- Changes to the content and technical components in one part of the information system may
require repeated testing to ensure that the other inter-related parts of the system continue to
behave as expected after the change
- It is essential to ensure that the content underlying the CDS interventions remains synchronized
with this best practices
Page 6.102
- Assigning an "expiration date" to th knowledge components of all CDS intervention can be a
useful strategy to help keep the knowledge base current
- These time limits should be set to correspond with the anticipated "shelf –life" of the
intervention's knowledge and trigger content review by an appropriate domain expert after a
pre-determined time since the last review has elapsed
- Preserving superseded knowledge, using techniques such as versioning, helps to link knowledge
to interventions even in retrospect years after the intervention
![Page 29: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/29.jpg)
Page 6.103
- Versioning may help people understand what happened during the course of a CDS intervention
if this context is ever called into question
The Knowledge Asset Portfolio
- It is helpful to consider your organization's CDS content assets as a single portfolio of knowledge
resources.
- These assets might exist in several different forms integrated into clinical information systems
- Organizations have both paper based and electronic medical reference information
- This universe of local content assets may come from a variety of sources
- It is useful to have a systematic approach to procuring and managing this information, taking
into account the inter-relationships among its components
- These changes necessitate corresponding updates to pertinent components of the entire suite
of CDS interventions, which your knowledge management mechanisms should enable in a timely
and efficient manner
Page 6.104
Step 1:
- Assess intervention use and usability on an ongoing basis
Page 6.105
o Quantifying how often each intervention is utilized provides a backdrop for examining
its effectiveness in achieving target objectives
o Approaches that combine automated statistics with data from surveys and surrogate
markets may be needed for usage data
o Any intervention that yields more costs and problems than benefits despite reasonable
remediation efforts probably should be discarded in favor of an alternative CDS
approach to the objective, or a reassessment of the objective
Step 2
- Evaluate intervention impact on target objectives
o The substantial efforts in the entire CDS program are directed towards realizing
measurable and important improvements in targeted healthcare processes and
outcomes
Page 6.106
- Use and Usability Issues Log
![Page 30: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/30.jpg)
o Intervention name
o Usage and Usability Issues (Source / Channel)
o Date Noted
o remediation Plan (Responsible Party, Date Resolved)
o Priority
Page 6.107
o The assessments should be quantitative and compared to baseline measurements
o Key definitions might vary across systems, making it difficult to create a unified picture
of intervention effects
o It is important to keep straight whether the targets and performance measures pertain
to the objective or to a component intervention
o Improved compliance with the intervention is expected to improve performance toward
the clinical objective
Page 6.108
o Enhancements assessed qualitatively are also important and should be documented as
well
Step 3:
- Continually enhance the value to users and impacts on goals and objectives of individual CDS
interventions and the overall CDS program
Page 6.109
- Performance against objectives
o Clinical Objective
o Desired Action
o Intervention name
o Baseline Performance
o Desired Improvement
o Actual Performance Improvement
o Other Effects
Page 6.110
- CDS program enhancement plans
o High level clinical goal
o Clinical Objective
o Intervention name
o Effectiveness Summary
o Issues and Usability Summary
![Page 31: MMI 406: Decision Support Systems and Health Care](https://reader036.vdocument.in/reader036/viewer/2022062301/62a5578e0ddb9a30f26c5247/html5/thumbnails/31.jpg)
o Enhancement Plans