Medical Call Data Validation

Transcription

1 Medical Call Data Validation January 30 February 2, 2018 Palm Beach County Convention Center West Palm Beach, FL The Path to Data Excellence

2

3 Medical Call Data Validation Presented by: Bob Vaughan and John Foust Objectives Usage Discuss NCCI s need for high quality Medical Call data to perform legislative pricing and research studies. Data Quality Understand common data quality concerns and the ongoing efforts to improve quality. Data Validation Understand the medical data validation test approach used by NCCI and how results are communicated to data providers. 2 1

4 Agenda 3 Purpose of the Medical Data Call 2

5 Legislative Analysis Legislative Pricing Analyses and Regulator Requests Annually Changes to physician fee schedules Inpatient hospital comparing different Medicare based scenarios Outpatient hospital and Ambulatory Surgical Center (ASC) changes to reimbursement rates Drug formularies Opioid utilization Air ambulance reimbursement 5 Legislative Pricings Common Analysis Proposed Fee Schedule Changes by a State Impact on future cost Maximum Allowable Reimbursement (MAR) = Conversion Factor x Medicare Relative Value Units (RVU) Weighted average percentage change by share of procedures subject to fee schedule 6 3

6 Research Studies Medical Data Call Research Committee Medical Cost Trends: Then and Now Investigating the Drivers of the 2015 Workers Compensation Medical Severity Decline How Is Medical Inflation Measured? And Why Should I Care? Do Drug Fee Schedules Based on AWP Have an Effect on Prices Paid for Drugs in Workers Compensation? 7 Medical Cost Trends: Then and Now In recent years, the growth in US healthcare costs has outpaced the growth in WC medical costs Recently, Medicare has put greater upward pressure on WC hospital prices The in network share of payments for physician services has been steadily increasing Prescription Drug Fee schedules and Pharmacy Benefit Managers (PBMS) help control drug costs 8 4

7 Investigating the Drivers of the 2015 Workers Compensation Medical Severity Decline Paid medical costs per claim declined 1% in Accident Year 2015 A 3% decline in paid costs per claim for physicians accounts for most of the 2015 medical cost decline Hospital and Emergency Room costs increased in Accident Year 2015 Approximately half of the states had medical severity decreases in Accident Year State Medical Data Reports NCCI s State Medical Data Reports provide a data source for regulators and other industry stakeholders. 10 5

8 Common Data Quality Topics Data Quality Key Data Elements Data Quality Paradigm Data Quality Processes Edit Strategy Rejection of Files Aggregate Quality Review Key data elements must be consistent on a per claimant basis from all submitters. Key Data Elements Carrier Code Policy Number Policy Effective Date Claim Number Creation of Claim ID 12 6

9 What NCCI is Doing Comparing Medical Data Call to Unit Statistical Data: Linking (Claim Number) Large difference in Medical Paid Amount 13 Incorrect Reporting of Key Data Elements Transactional data = No key field change process 1 Suspend future data submissions 2 Correct historical data 3 Validate next submission Consult with NCCI 4 Allow future data submissions 14 7

10 Medical Call Transactions Transaction Code 01 Original 02 Cancellation 03 Replacement Use The first reporting of a medical transaction Only one may be submitted for a medical transaction Used to delete or cancel record Used to revise previously reported non key field values Report current cumulative values, not the change in value 15 Data Quality Duplicate Records Additional Reimbursements Amount Charged = $75 Reporting Options: Amount Paid = $50 Additional Reimbursement = $10 1. If possible, submit one record with what value in the Paid Amount? $60 2. Submit a replacement record Same key fields including Bill ID and Bill Line ID Current cumulative value ($60, not $10) Submit a cancellation record and a new original Same key fields on cancellation as original record Current key fields and cumulative value on new original Bill ID and Bill Line ID may be unique 16 8

11 Claim Number Additional Reimbursements Replacement Example Transaction Code Bill ID Line ID Paid Procedure Code Amount Charged by Provider Paid Amount Original must be in the same submission or on NCCI database (Quality Validation Edit ) Transaction Date of the replacement record is the date that the additional reimbursement was made in the payer s system Must be Transaction Date on the record being replaced (Return Record Edit ) 17 Additional Reimbursements Cancellation Example Claim Number Transaction Code Transaction Date Bill ID Line ID Paid Procedure Code Amount Charged by Provider Paid Amount Original must be in the same submission or on NCCI database (Return Record Edit ) Transaction Date of the cancellation and new original records is the date that the additional reimbursement was made in the payer s system New key fields reported on the new original record 18 9

12 Data Quality Duplicates Mirror Duplicates Bill ID 101 Payer s system assigns Bill ID 201 Two reporting options: Only submit one of the two bills/records; if both bills are created in the same quarter and the first has not been reported, submit the second bill only Submit cancellation record and new original record 19 Duplicate Billing Claim Number Transaction Code Transaction Date Bill ID Line ID Paid Procedure Code Amount Charged by Provider Paid Amount Here s a common example of how this might be reported. How should this be corrected? 20 10

13 Duplicate Billing Cancellation Example Claim Number Transaction Code Transaction Date Bill ID Line ID Paid Procedure Code Amount Charged by Provider Paid Amount Original must be in the same submission or on NCCI database (Return Record Edit ) Transaction Date of the cancellation and new original records is the date that the second bill was created in the payer s system New key fields reported on the new original record 21 Data Quality Trending Industry Averages QUARTERLY Unusual percentage of: Duplicate transactions Pharmacy, inpatient and outpatient hospital, and physician cost by state of jurisdiction Outliers 22 11

14 Data Quality Trending Industry Averages ANNUALLY Differences in sum of Paid Amount from Medical Data Call and Financial Call data and NAIC data on a calendar year basis Outliers 23 Data Quality Outreach Example DRG Code 999 Outlier Detection Example: Overuse of Diagnostic Related Group (DRG) Code 999 Ungroupable instead of specific DRG codes or the per-diem code 24 12

15 Data Quality Outreach Inpatient Hospital Bills A Hospital Bill may be reported to NCCI in a number of ways depending on how it is reimbursed DRG payment record reported with total charges Revenue Code transactions reported with no charges or Individual charges reported on Revenue Code transactions DRG payment record reported with no charges 25 Reporting hierarchy: Data Quality Outreach Provider Zip Code Zip Code that impacts reimbursement Zip Code where service was performed Billing Zip Code, unless billing house or Pharmacy Benefit Manager MS TN AL GA SC NC LA FL 26 13

16 Medical Marijuana Paid Procedure Codes Other Reporting Topics MM001 Reimbursement to Claimant MM002 Reimbursement to Dispensary Place of Service Code DS Dispensary Place of Service Code 02 assigned to Telehealth Network Service Code B Pharmacy Benefit Manager Compound Drugs Paid Procedure Code J7999 Compounded drug, NOC 27 Repackaged Drugs State fee schedule requires repackaged drugs to be reimbursed based on NDC code of original manufacturer Paid Procedure Code Secondary Procedure Code Original NDC Code Repackaged NDC Code This is standard reporting for the Medical Data Call: The Paid Procedure Code is the one associated with the Paid Amount 28 14

17 Data Exclusions Groups with substantial issues may be excluded from pricings or studies. Examples of issues leading to exclusion: High percentage of duplicated transactions Incorrect formatting of Paid Procedure Codes or Taxonomy Codes that impacts the ratios 29 Recap 1. What is medical data used for? 2. What are two ways to address duplicate records? 3. Key fields consistently reported establish the ID

18 Questions 31 Validation Tests 16

19 Validation Processes Editing Quality Tracking Quality Validation Usage and Trend Analyses Validation Tests 33 Validation Tests Carrier group data is evaluated each quarter Different criteria/scenarios per test Percentages of data by record count or total paid Comparison to industry Identify outliers (unusual percentages) 34 17

20 Formalizing Trend Analyses Quarterly runs can detect arising issues early Automation allows more time for reviewing output and refining criteria System identifies repeat outliers not needing to be contacted again 35 Validation Process Identify outlier reporting behavior Analyze carrier data to find illustrative examples Update validation test system based on explanation NCCI Medical Data Validator Issue Description & Detail Data Explanation Carrier/ Medical Data Provider Contact(s) Investigate medical data systems: Determine data was correctly reported, or Update system to correct issues Detailed discussion helps NCCI determine how to refine our tests

21 Types of Validation Tests Record Level Tests National Drug Code (NDC) Unit Reporting Paid Duplicates Jurisdiction to Provider State Comparison Claim Level Tests Multiple Accident Date Claims 37 NDC Units What quantities/unit counts should be reported for NDCs? Pill counts Standard package count Quantity as per NDC standard measurement Most prescriptions are for 15, 30, 45, 60, 90, or 120 pills Few prescription bills should report 0, 1, or invalid units 38 19

22 NDC Units Outlier Common Units (15, 30, 45, 60, 90, 120) Number of Carrier Groups Percentage of Records Reporting Common Units 39 NDC Units Outlier 0, 1, or Invalid Units Number of Carrier Groups Percentage of Records Reporting 0, 1, or Invalid Units 40 20

23 See handout. NDC Units Outreach Example 41 See handout. NDC Units Outreach Example 42 21

24 See handout. NDC Units Outreach Example 43 See handout. NDC Units Outreach Example 44 22

25 See handout. NDC Units Outreach Example 45 Duplicate Transactions Not properly creating cancellations/replacements Additional reimbursement Duplicate billings Bill ID 101 Bill ID

26 Duplicate Transactions Outreach Example See handout. 47 Duplicate Transactions Outreach Example See handout

27 Jurisdiction to Provider State Comparison Jurisdiction State Code: state under whose Workers Compensation Act the claimant s benefits are paid Provider Zip Code: reimbursement, location, billing service Not billing house, pharmacy benefits manager, case management firm Identify Provider State based on Zip Code and compare to Jurisdiction State 49 Jurisdiction to Provider State Jurisdiction State Provider State Comparison FL FL FL ME IA SC GA FL 50 25

28 State Comparison Outreach Example See handout. 51 Multiple Accident Date Claims Accident Date: the date the claimant was injured For occupational disease, use the earlier of last day of work without the disability or the last day of coverage Review all transactions related to a claim Identify claims with more than one Accident Date Blank is not considered a different Accident Date 52 26

29 Multiple Accident Date Claims Outreach Example See handout. 53 Questions 54 27

30

31 Supplemental Information

32

33 Presenter Biographies Bob Vaughan, AIDM, has worked at NCCI for 20 years. He is a consultant in the Data Services Department in NCCI s Data Resources Division. Bob participates in many projects related to NCCI s data tools and serves as coordinator for the Data Collection Procedures (DCP) Subcommittee. Bob is also responsible for data reporting communications and manuals. He is actively involved in industry committees focused on data reporting. John Foust is the medical data steward for NCCI s Medical Data Validation Department. Medical Data Validation encompasses collecting, editing, storing, and validating data that is associated with NCCI s Medical Data Call. He provides Medical Data Call systems support and project leadership as well as data quality review. John has worked in various departments in NCCI s Data Resources Division over the past 12 years, including on the Unit Data Collection and Validation Teams and in the Voluntary Data Services Department.