R06.2 F43.0 I10 06BY3ZC J03.0 A78 03HK0MZ R52 0SG1430 COMPUTER-ASSISTED CODING AGENDA Evaluating and Understanding the Technology Review of Lessons Learned from Early Adopters Workflow and Analytics with CDI Enhanced CDI with CAC WHAT IS COMPUTER-ASSISTED CODING, REALLY? Computer-assisted coding (CAC) is the use of computer software that automatically generates a set of medical codes for review, validation, and use based upon provider clinical documentation. There are CAC solutions in the market today that do not meet this basic definition and do not suggest codes. Remember, just because a solution is marketed as CAC, does not mean it meets the criteria that AHIMA has defined. **Delving into Computer Assisted Coding American Health Information Management Association (AHIMA)
WHAT IS NATURAL LANGUAGE PROCESSING (NLP)? NLP is the technology behind CAC, and is the key determining factor as to whether a CAC solution may succeed or fail. There are multiple technologies used for NLP in the CAC industry, ranging from basic terminology matching to advanced artificial intelligence. NLP is the brains of the system that assigns the codes to be presented in the user interface. You cannot have CAC solutions without some form of NLP. NLP ENGINES ON THE MARKET Many different NLP approaches: Each NLP technology performs differently, using different approaches to text evaluation. Educate yourself on the various types of NLP used in CAC and understand the deployment methodology for each engine type. Medical terminology matching Pattern matching Symbolic/statistical hybrid** NLP DEPLOYMENT STRATEGIES The deployment strategy of the NLP engine is important to understand because it directly affects the process of development, enhancements, support, code updates and customization based on a site by site basis. Different deployment strategies require various levels of resource commitment from both the vendor and the customer. Some NLP engines are completely managed by the vendor on a cloud based model.
CAC INTEGRATION ENCODER Existing Coder Tools Integrated. EMR/EHR Patient chart used for coding is sent to CAC to process. Coder Is presented with Electronic Charts with Suggested Codes. ADT When the Patient is Registered it sends ADT info to CAC. CAC Abstraction System All Codes and Info Sent from CAC BASIC EXPECTATIONS OF CAC Increase Coder Productivity Smart Workflow for Coding and CDI Consolidation Systems used for Coding Physician Query from Coding and CDI Tracking Responses Integrated Existing Encoder Or Encoder Replacement Enterprise-Wide Management and Reporting Both ICD-9 & ICD-10 Dual Code Suggestions ICD-10 Based Physician Guidance RECOMMENDATIONS BEFORE REVIEWING CAC Coding Workflow CDI Workflow Challenges Time Consuming Processes Where Documents Are Located Systems Used By Coders
WHAT ARE YOUR CAC NEEDS? In what areas will CAC be utilized? Inpatient, Outpatient and/or Professional What processes would you like to enhance or add to coding or CDI? Workflow Scrubbing Automatic Abstraction Review of Charging Electronic Physician Query IMPLEMENTATION CONSIDERATIONS Is there a waiting list to start your project or can you start right away? How does the vendor role out CAC? Phased Approach Big Bang Does the version you are installing support ICD-10 or will there be an upgrade to support it? When will ICD-10 be available to begin dual coding? What is the vendor s definition of dual coding? WORKFLOW CONSIDERATIONS Do you have to switch your encoder or can you use your existing encoder? Can you integrated to your current coding queues? Are there different platforms for inpatient, outpatient and/or professional coding? Is CDI a separate module?
WHAT IS LIVE AT CUSTOMER SITES? Many solutions demonstrated may not be live. Is the product being demonstrated live? If so, where and how many sites are live? Is it live beta, pilot, or production? Is it live across all of inpatient and outpatient, or limited to specific disciplines? CAC is changing so rapidly and several vendors have shifted their NLP and CAC strategies one or more times, so it is important to validate what is actually live and where. CAC FOR ICD-10 Gap Analysis Results Procedure Codes Coded Not Coded Diagnosis Codes Coded Not Coded Procedures Diagnoses
Gap Analysis Information Coding results for FY2011 at 23 hospitals were considered Analyzed the Top 200 procedure codes and diagnosis codes These Represented 80% of the codes billed ICD-10 CM Documentation Type Of Encounter (Initial Or Subsequent) Applied Specificity (Did The Patient Lose Consciousness?) Acute Versus Chronic Relief Or Non-relief (Intractable Versus Non-intractable) ICD-10 CM Documentation External Cause (What Caused The Accident?) Activity (What Was The Patient Doing When Injured?) Location (Where Was The Patient When Injured?)
ICD-10 PCS Documentation What Was The Procedure? What Was The Body Part? The Approach? The Device or Devices Used? Qualifiers? Documentation Is Lacking Need For More Specific Documentation From Physicians Today ICD-9 Doesn t Require Physicians To Specify Laterality. For Example: When coding for a foreign body in the nose ICD-9 currently does not require documentation other than, Foreign Body of the Nose. Example Of ICD-9 CM Specificity A provider sees a patient for a [open] [fracture] of the [radius]. ICD-9 Code Description 813.52 Other Open Fracture of Distal End of Radius (Alone)
Example Of ICD-10 CM Specificity A provider sees a patient in a [subsequent encounter] for a [nonunion] of an [open] [fracture] of the [right] [distal] [radius] with [intraarticularextension] and a [minimal opening] with [minimal tissue damage] ICD-10 Code S52.571M Description Other intra-articularfracture of lower end of right radius, subsequent encounter for open fracture type I or II with nonunion CAC FOR CDI Goals Of CDI Identify And Clarify: Missing, Conflicting, Or Nonspecific Physician Documentation Support Accurate DX/PCS Coding, DRG Assignment, SOI, And Expected ROM Promote Health Record Completion Improve Communication Of The Healthcare Team. Provide Education
Challenges With CDI Patient Chart Review, Where To Start? When To Re-Review Clinical Vs. Coding Knowledge Working DRG Querying Ongoing Requirements What IF Technology Could Distribute a Work List Automatically for CDI Specific DRGs/MS-DRGs Unspecified Diagnosis Charts with Protocols Payer CAC Helps With CDI Challenges The "search and compare" functions assist CDI in reviewing vast amounts of data in a more efficient manner. Improve MCC/CC capture rates Auto-suggested codes further improve productivity and accuracy Word search functions allow CDI staff to search for terms that should be documented based on clinical data. Physician queries can be tracked more effectively
CAC - NOT Just For Coders The traditional concurrent review role of the CDI specialist is expanding. Due to increasing demands to demonstrate quality initiatives in real time, the role will become even more important in the future. CAC Powered by NLP Natural Language Processing (NLP) software for healthcare identifies medical language found within the text of the documented patient encounter. NLP converts patterns of language and applies the correct ICD or CPT code. The entire patient record is considered and not just single documents within the chart. Free form text is processed and does not require structure or changes in physician behavior. Not all CAC is powered by NLP. CAC Helps With CDI Challenges Computer-Assisted Coding can be used to identify possible opportunities for greater specificity in documentation. Add a concurrent clarification query into an active chart Ability to track and monitor CDI activities. **Clinical Documentation Improvement Toolkit Whitepaper, AHIMA
How Does CAC Help CDI? Provides a working DRG, expected reimbursement and the total charges of a patient upon admission. Supports a complete overview of the entire record and pointing out relevant clinical information, with suggested codes. Provides dashboards to retrieve and respond to queries for an efficient and timely process during the care process. We Already Have A CDI Software? Computer-assisted coding integrated with CDI technologies can enable advanced data analytics for evaluating quality outcomes data and aligning it with financial performance. Many hospitals have been able to identify key areas for improvement and fast-track solutions to address problem areas, often resulting in a two to three percent increase in case mix index! CAC With CDI Scenario Patient Chart Scenario: A patient is admitted to the acute care setting in the hospital. The admitting diagnosis is change in mental status. The physician also documented that the patient has a fever of >100.4⁰F. Clinical indicators include tachycardia >90 BPM, Respiratory rate of >20 breaths per minute and an elevated white blood count of >12,000.
CAC Outcome NLP reviews the patient chart, sees that the patient is febrile, elevated white blood count, increased respirations. Since an elevated white blood count of >10,000 is a clinical indicator of sepsis this chart would get sent to CDI for further review. A CDI specialist queries the physician for further specificity. Insufficient Documentation Insufficient documentation means that the provider did not include pertinent patient facts (e.g., the patient s overall condition, diagnosis, and extent of services performed) in the medical record documentation submitted. Lack of complete and accurate diagnoses reporting on every daily progress note contributes to instances of Insufficient documentation Customer Case Study Adventist Health System: Largest Not-for-Profit, Protestant Healthcare Organization In The US Healthcare Facilities In 10 States 7,700 Licensed Beds 55,000 Employees 8,700 Physicians 2 Hospital Divisions (7 Regions, 44 Campuses) Total Operating Revenue Of $5.7 Billion
Customer Case Study Adventist Health System is currently using CAC to automate the grouping of diagnosis and procedures to create a working DRG every 24 hours. The working DRG information is posted to [EMRs] and is available to case managers and clinical documentation specialists in an effort to expedite their monitoring of transfer DRGs and proper documentation capture. -Migdalia Hernandez, RHIT, Adventist s Director of Corporate HIM **From No Code Left Behind, in For the Record Magazine RANGE OF IMPROVEMENTS The range improvement should be 25% or more. Accounts receivable and coding backlogs and reductions are more noticeable as coders become more comfortable with the technology. R06.2 F43.0 I10 06BY3ZC J03.0 A78 03HK0MZ R52 0SG1430 Thank You!!