This document discusses leveraging technology to advance clinical documentation improvement (CDI) programs. It notes that current CDI programs focus primarily on revenue cycle management and have resulted in physician cynicism. The document advocates for more physician-engaged CDI programs that integrate with quality initiatives and utilize technology like computer-assisted physician documentation (CAPD) to provide real-time guidance to physicians. This could help address challenges under ICD-10 by capturing accurate clinical impressions and ensuring compliant documentation. The document concludes that successful CDI programs require advanced, clinically integrated technologies that fit with physician workflows.
2. 2
Advancing CDI Through
Leveraging Technology
Paul L. Weygandt, MD, JD, MPH,
MBA, CCS, FACPE
Vice President Physician Services
Nuance Communications, Inc.
Nick van Terheyden, MD
Chief Medical Information
Office – CLU
Nuance Communications, Inc.
3. 3
Learning Objectives
• At the completion of this educational activity, the
learner will be able to:
– Discuss challenges of EMR content awareness and
analysis, and current disconnected documentation
clarification processes;
– Explain methodologies to engage physicians in the
CDI process
– Describe how technology can assist with
documentation improvement and acceptance
– Identify status of current advanced CDI programs
and the opportunity for integration of evolving
technological innovations
5. 5
• Impact of documentation improvement
– Compliance
– Revenue cycle
The Documentation Gap
Hospital
inpatient
care
Physician
documen-
tation
Coding
process
Revenue
cycle
processes
CDI
programs
Gap
Error
recovery
Fraud
6. 6
Typical CDI Programs
• Early successes
– Typical hospital revenue cycle impact 2%–4%
– Compliance
• A revenue cycle initiative
– Managed by HIM under a strong coding influence
– Little communication with quality
• Focus: DRG “optimization”
– Specific focus only on those areas of documentation
impacting hospital reimbursement
7. 7
Typical CDI Programs
• Result:
– Cynicism from medical leadership/medical staff
– No fit with other physician/clinical initiatives
– 1–2 year success cycle
– Documentation specialists progressively
disappeared into cubicles
8. 8
• Focus of documentation improvement
– Compliance
– Revenue cycle
– Quality
A New Source of Physician
Engagement
Hospital
inpatient
care
Physician
documen-
tation
Coding
process
Revenue
cycle
processes
Quality/o
utcome
measure-
ment
Evolving
quality-
based
payment
CDI
program
9. 9
Physician Engagement:
The “Game Changer”
Typical CDI programs
• Success metrics
– Typical hospital revenue cycle impact 2%–4%
– Compliance
• A revenue cycle initiative
– Managed by HIM under a strong coding
influence
– Little communication with quality
• Focus: DRG “optimization”
– Specific focus only on those areas of
documentation impacting hospital
reimbursement
• Result
– Cynicism from medical leadership/staff
– No fit with other physician/clinical initiatives
– 1–2 year success cycle
– Documentation specialists progressively
disappeared into cubicles
“Physician-engaged” CDI
• Success metrics
– CMI improvement a metric of quality and revenue
– Improved compliance
– Typical CMI improvement 4%–8%
• A clinical initiative
– Integrated with clinical quality
– Clinical management, CMO accountability
• Focus: clinical accuracy
– Accurate severity capture for every admission
impacting reimbursement, clinical care, and
quality metrics
• Result
– “Ownership by the medical staff”
– Response rates approaching 100%
– Integrated with other physician/clinical initiatives
– Sustained results
– CDSs part of the clinical team
10. 10
“Physician-Engaged CDI”
• Current impact of advanced CDI
– Quality metrics ─ POA/HAC
– Core measures ─ Medical necessity
– Compliance ─ Patient safety
• Impact during ICD-10 implementation
– Fully functional computer-assisted coding
– Decreased fraud/abuse risk
– Physician engagement and satisfaction
• Evolving reimbursement methodologies
– Risk assumption, ACOs/derivatives, CMS-HCC system
13. 13
Basic Concepts
• Inadequate physician documentation has been a
challenge for accurate coding under ICD-9
• If uncorrected, that challenge will increase
dramatically under ICD-10
• Coding solutions, alone, cannot resolve the issue of
inadequate physician documentation
• Physician leaders must be able to engage their
colleagues in a proactive manner, establishing the
appropriate motivation and sharing necessary
knowledge to achieve success under ICD-10
15. 15
Current Clinical Documentation
and Coding Processes
Little operational integration of workflow
The physician world The HIM/revenue/compliance world
EHR Analytics
Quality
reporting
ComplianceCodingDocumentation
Patient
encounter
18. 18
Leveraging the EHR for Value
CLU
EHR
CAPD/
CA CDI
Analytics
Quality
reporting
ComplianceCodingDocumentation
Patient
encounter
ICD-10
knowledge
CA
compliance
CA quality
reporting
CA data
analytics
CAC
Voice/
direct text
entry
25. 25
CAPD: A Revolutionary New Solution
Voice input:
“the patient has acute
respiratory failure”
CAPD response: Multiple
correlates of acute on
chronic respiratory failure
identified within narrative
documents
Physician determines
and documents
additional specificity
“acute on chronic
respiratory failure”
Acute Respiratory Failure
Concurrent medical record
corrected: “acute on chronic
respiratory failure”
26. 26
Fits Physician Workflow
• Interactive clarifications while physician is documenting
“Patient has altered
mental status, abnormal
liver function, and
treatment with
lactulose”
CAPD identifies clinical
correlates of hepatic
encephalopathy and
presents documentation
alternatives in a compliant
manner
Physician reviews
clinical finding and
documents clinical
opinion “hepatic
encephalopathy”
Hepatic encephalopathy
Accurate
diagnosis for:
• Severity
• Quality
• Outcomes
• Payment
Altered Mental Status
Concurrent medical
record indicates
specific diagnosis of
“hepatic
encephalopathy”
28. 28
On the Horizon
• Progress notes
– Meaningful expression of a physician’s “clinical
impression”
– Should systems be designed for E/M coding?
– How can we avoid error/fraud?
29. 29
On the Horizon
• Problem lists
– When should they be created (the ED)?
– When should they be available?
– Only definitive diagnoses?
– Where do they reside?
– Who owns them?
– Resolved conditions …
30. 30
On the Horizon
• Discharge summary
– Inclusion of diagnoses managed during the
admission (resolved conditions)
– Inclusion of diagnoses provided by consultants
(which may conflict with attending)
– Availability on day of discharge
– Physician workflow/support
32. 32
Summary
• ICD-10 will impact every medical record interaction
– All physician documentation should be viewed through the
ICD-10 lens
– The transition to ICD-10 will require advanced, clinically
integrated CDI programs
• Watch out for technologies that could impact fraud, abuse, and
error
– Copy and paste, point and click, etc.
– We need to capture the physician’s “clinical impression”
• Physician engagement and satisfaction is critical
– We must positively impact physician workflow
– We must avoid rework
– We must leverage technology at the point of care
33. 33
Thank you. Questions?
In order to receive your continuing education certificate(s) for this
program, you must complete the online evaluation. The link can be
found in the continuing education section at the front of the
workbook.
Notas do Editor
The ability for clinician to capture documentation in any workflow and any device is essential to capturing the whole patient story
And capture is only the beginning
First we must make clinical data digital…
…then we must make digital data meaningful
And that’s no small feat!
[this slide automatically transitions to next slide build]
Natural Language Understanding (aka Clinical Language Understanding – CLU) is evolving speech from capture and recognition, to understanding
To do that, we have NLP technology that enables deep understanding
It extracts the clinical facts from the narrative (Voice or Text) and transforms those facts into codified standard terminology while still retaining the original captured story
Note from Editorial Services: There’s a typo in some of the text that’s currently outside of the slide (“penicillan”, should be “penicillin”) but I can’t seem to edit the text. Not sure if it’s important as slide show doesn’t really focus on this text (zooms by too fast to see typo), but figured I’d point it out anyway.
The Clinical Language Understanding Engine normalizes and translates your data into standard medical terminology – how does it do this…
Our CLU engine contains the largest medical knowledge base in the world
What differentiates our CLU is the optimal combination of our four best of breed technologies; ontology, syntactic parsing, using both NLP rules-based and statistical engines
The knowledge elements represent the standard controlled medical vocabularies, SNOMED-CT, ICD-9, ICD-10, RxNorm for medications
The medical concepts and relationship types represent information gathered from hundreds of thousands of clinical documents analyzed and codified in a way that reflects how these different types of data are related to each other
Our CLU technology engine has been tested to within 94% accuracy and precision well beyond any other system in the healthcare IT market. The engine is not only accurate, but flexible, as what your needs are tomorrow are not necessarily what your needs are today either from a quality, audit, or federal regulatory requirements perspective. The engine constantly analyzes and learns with each new document it processes, always improving, always growing.
What this means to you is that information gathered through the process of dictation and transcription via our Nuance documentation creation solutions passes through the CLU engine to be codified in a way that makes narrative unstructured data meaningful and actionable.
Offers you a more sophisticated approach to tackling the most complex documentation challenges. With revolutionary Clinical Language Understanding (CLU) technology and over 20 years of innovation driving Clinical Documentation Improvement (CDI) success, Nuance Healthcare solutions powered by JA Thomas Advanced Practice Clinical Documentation Improvement strategies go beyond just using documentation to support coding. Automated, clinically appropriate guidance and clarifications are presented to physicians at natural points during their workflow.
It is imperative that physicians document the care provided as well as the rationale behind their actions.
Unlike retrospective coding based programs, our solution offers a concurrent review process, with appropriate clinical guidance allowing physicians to accurately capture compliant documentation, complexity levels and severity levels in their documentation from the moment the patient enters the healthcare system.
.
It is designed to monitor physician documentation and only when necessary interactively prompt for additional information to help
improve document quality in the context of ICD-9 today and ICD-10 ‘tomorrow’
subtly train physicians on the details required for ICD-10
help organizations achieve appropriate reimbursement that better reflects the level of care and treatments provided
How does it work? Very simply, CAPD:
Uses Nuance Clinical Language Understanding engine to analyze clinical documents
Applies JATA Clinical Documentation Improvement guidelines
Automatically generates clarifications based on JATA Clinical Documentation Improvement content
Presents clarifications during physician’s document creation workflow
In our example, while dictating the physician indicates respiratory failure. For those CDI specialists in the room you know that more detail may be required to support the appropriate code generated for this diagnosis. Physicians, you know you will likely receive a communication from your CDI team asking for more details. CAPD identifies that this unspecified respiratory failure may require more details, and intuitively asks the physician to define the acuity – acute or chronic, acute on chronic, and by responding to the queries the physician will update the documentation and a much more detailed complete document is sent through the system to be coded.
Physician starts dictating in to EHR using Dragon Medical 360 | Network Edition
Dragon Medical 360 | Network Edition alerts CAPD that physician started dictation for a specific patient stay
CAPD shows outstanding queries for patient stay
Physician sees queries and continues dictating
Physician can reprocess updated report anytime to see updated queries
Physician completes report and exits
CAPD processes final report and generates an audit trail
The Front End/Direct Dictation workflow features use of Dragon Medical 360 | Network Edition and its interaction with the CAPD editor to provide physicians with instant feedback while documenting. The CLU engine works in the background to evaluate the content of the dictation, evaluating the text for missing information, detect missing diagnoses or procedures strongly suggested by data presented, unclear associations between relevant findings or unspecified diagnoses. It then highlights them for physician consideration and correction. then generates a clarification that proposes a specific diagnosis or procedure strongly suggested based on evidence from one or more of the patient’s documents collected from this admission. The physician can make the corrections using Dragon, review and electronically sign the document. A more detailed clinical document is available for CDI specialists and coders to review and group codes for reimbursement.
This workflow is the most compelling solution as it brings the clarifications to the physician while they are documenting, allowing questions to be addressed without disrupting the clinical thought process and in the context of what’s happening with the patient at the time.
Case Study:
Consider for example a new patient to your healthcare system seen in your Emergency Department for altered mental status and severe abdominal pain. Through the course of their initial assessment, documentation includes a past medical history of cirrhosis, lab values including low hemoglobin, high SGOT and melenic stool, an EEG order and prescription for rifaximim and lactulose. While completing the note, the physician is presented with automated clarifications proposing diagnoses of hepatic encephalopathy and anemia of blood loss, as suggested by the evidence shown in the patient’s documentation. The physician is able to confirm the relevance of the hepatic encephalopathy and the blood loss anemia and make the appropriate adjustments in the patient’s documentation.
Capturing all diagnoses makes it possible to set the principal diagnosis correctly at the beginning of the patient’s stay, and helps to ensure that appropriate treatment and care shorten the overall length of stay, guide post discharge care and reduce potential readmission. The severity of illness and risk of mortality will be accurately documented and proper details will be available for coding either manually or using Computer Assisted Coding for the correct DRG assignment.
Additionally, benefits of documented secondary diagnoses include:
Accurate mortality index, observed and expected mortality stats (O/E)
Appropriate severity of illness (SOI) for public report cards
Accurate APR-DRG Severity of Illness and Risk of Mortality
Favorable preparation for ICD-10