Medicaid Performance Audit An Emerging Challenge for MCOs Harry Carstens Director, Compliance Molina Healthcare of Washington My Brief Resume Molina Healthcare of Washington: Compliance Director 2 years Blue Cross and Blue Shield: Compliance and Ethics 15 years Finance 7 years Operations Management 3 years 1
Audit Experience Center for Medicare and Medicaid Services Office of the Inspector General Defense Contract Audit Agency US Office of Personnel Management Washington State Department of Social and Health Services Washington State Health Care Authority Washington State Office of the Insurance Commissioner Public Accounting Firms NCQA URAC AGENDA What is a Performance Audit? Experience in Washington State Notable Issues Expected Findings Questions 2
What is a Performance Audit? Performance Audit Conducted by state government accountability office To evaluate the effectiveness of state agencies Guarantee that tax dollars are spent cost effectively 3
Performance Audit Authority Virginia: Joint Legislative Audit and Review Commission Make performance reviews of operations of state agencies to ascertain that sums appropriated have been, or are being expended for the purposes for which such appropriations were made and to evaluate the effectiveness of programs in accomplishing legislative intent. Washington: State Auditor the state auditor shall conduct independent, comprehensive performance audits of state government and each of its agencies, accounts, and programs to ensure accountability and guarantee that tax dollars are spent as cost-effectively as possible. Performance Audit Evolution Medicaid Managed Care CMS s Oversight of States Rate Setting Needs Improvement US GAO (2010) 1 Interim Report: Fraud and Error in Virginia s Medicaid Program VA JLARC (2010) 2 Compliance Audit Horizon New Jersey Health s Special Investigations Unit NJ State Comptroller (2011) 3 Medicaid Managed Care: Fraud and Abuse Concerns Remain Despite Safeguards OIG (2011) 4 Mitigating the Risk of Improper Payments in the Virginia Medicaid Program VA JLARC (2011) 5 Medicaid Managed Care Improper Payments, Fraud & Abuse Prevention and Detection WA State Auditor (2014) 4
Performance Audit Examples New Jersey State Comptroller Primary focus on fraud, waste and abuse Virginia Joint Legislative Audit and Review Commission Focus on improper payments in fee-for-service Washington State Auditor Focus on improper payments in managed care The Washington State Experience 5
Washington Focus Audit Focus Improper payments Errors vs. fraud and abuse Focus on clinical documentation Agnostic to managed care practices Washington Purpose Audit Purpose Generally: The state Medicaid agency s oversight of managed care organizations (MCO) Specifically: MCO Administrative costs Improper claims payments by the MCOs Tangentially: Program integrity processes False claims 6
Washington Timeline Began August 2012 Two phases Phase I Test for risk of improper payments Phase II Test for overpayments in identified risk areas Phase I completed / Phase II began January 2013 No final reports as of 2/4/2014 Washington Approach Phase I Test encounter data for outliers Identify risk areas Phase II Select samples for audit Audit claims paid against clinical documentation Extrapolate results to estimate overpayments 7
Washington Methodology Similar to CMS Payment Error Rate Measurement (PERM) 6 Sort claim by payment type (or risk category) Select stratified samples from each type or category Identify billing errors in the sample Extrapolate the error rate to estimate the total amount of error PERM Example 8
PERM Example Start with total dollar amount PERM Example Divide population into strata so each has equal dollar value. 9
PERM Example The sample size is the same for each stratum PERM Example Used Cumulative Square Root of the Frequency 10
Washington Phase I Test encounter data for risk of improper payments Encounter data submitted to the state during the audit year Current run of encounter data for the same period Claims data from the company s data warehouse, adding numerous data fields Identified areas with the highest potential risk for overpayments Washington Sidebar Note the change in language from improper payments to overpayments. Medicaid Program Integrity Manual Chapter 1 1035 7 An improper payment is any payment that should not have been made or that was made in an incorrect amount under statutory, contractual, administrative, or other legally applicable requirements. Incorrect amounts include overpayments and underpayments. 11
Washington Phase II Identified eight risk areas Eliminate net zero-paid claims from the population Select sample claims for each risk area Obtain clinical documentation Missing / incomplete documentation = error Test clinical documentation to support claim amount Re-price claims billed in error Calculate the dollar error rate Extrapolate an estimate of overpayment amount Washington Risk Areas Excessive Billing of Modifier 25 Duplicate Payment on Evaluation & Management Unbundling of CT Scans One Day Stay DRG Upcoding Upcoding of Evaluation and Management Atypical Antipsychotic Controlled Substance 12
Sidebar Removing $0 Paid What is the potential effect of removing zero paid claims from the data before sampling? Assume a total of 100 claims 5 claims paid $0 5 claims overpaid 90 claims paid correctly Five overpaid claims is 5% of 100 claims Five overpaid claims is 5.3% of 95 claims Removing $0 paid claims increases the probability of finding overpaid claims in the sample The sample is not representative of the population Sidebar Missing Documentation Treating missing or incomplete documentation as an assumed overpayment makes sense in a recovery audit. Providers should not be paid if they can t show evidence a service was performed. But should the same rule apply to managed care? MCOs are not the natural custodians of clinical documentation, and have limited ability to compel providers to produce documentation. Eventual results are based on assumptions. There is no certainty in the conclusion. 13
Washington Methodology 95% confidence level and 10% precision rate If tested 100 times, 95 tests will produce results within ± 5% Adequate, but not stringent Shift focus from improper payments to overpayments Remove denied and zero paid claims from the population Select stratified samples from population for each risk area Review clinical documentation for each sample claim Identify billing errors Extrapolate error rate to the population Issues with the Audit Approach 14
Issues Caveat This discussion is about the methodology and not about any person or agency involved with the audit. The Washington State Auditor is highly regarded for impartiality, fairness and competence. The contracted auditors have demonstrated exemplary professionalism and competence. Issues Document Source Who should obtain supporting clinical documentation? The auditor is empowered by state and federal law The MCO is restricted by state law and by contract There is an advantage to the MCO to obtain records Opportunity to review records that will be audited There are disadvantages to the MCO A potential strain on relationship with provider Significant cost Bear the burden of missing/incomplete documentation 15
Issues MC is not FFS The audit is agnostic to managed care practices The managed care claims lifecycle is longer than FFS No consideration of prior authorization Claims denied through prepayment edits are not audited Claims recovered through post-payment review are excluded No consideration for fraud and abuse recoveries Payment errors are over-represented in the audit samples Extrapolated results are inflated Issues Statistics 95% confidence level and 10% precision rate If tested 100 times, 95 tests will produce results within ± 5% Adequate, but not stringent Confidence and precision rely on certain rules Sample must be randomly selected from the population The sample must be observed (i.e. not assumed) Eliminating some instances (i.e. $0) violates randomness Assuming values (i.e. missing documentation) violates observation principle Impossible to achieve stated confidence and precision 16
Issues Statistics Matter Issues Professional Judgment Competent medical coders sometimes disagree Providers coding specialists convert clinical notes to claims Managed care edits filter out likely errors Post-payment review finds more coding errors Program integrity finds more errors Multi-level appeal processes 17
Issues Clinical Judgment States set criteria for clinical competence Competence is earned through demonstration Education Examination Practice Continuing education Competence is rewarded through licensure Some states restrict claims review to their own licensed clinicians Some states require active practice to qualify to review clinical documentation Some states require certification in specialty under review Issues State Regulations MCOs pay claims not chart notes Some states restrict access to clinical documentation MCO must accept information from any reliable source MCO may not routinely require clinical documentation on all cases Administrative simplification 18
Issues Audit Accuracy The American Hospital Association RACTrac Survey 8 (May 2012) 2,220 hospitals participated 36% of RAC denials appealed 75% of appealed RAC denials were overturned 27% of RAC denials were overturned Issues Auditor Knowledge Reject ER claims based on Modifier 25 Significant, separately identifiable evaluation and management (E/M) service by the same physician on the day of a procedure Medicare requires that modifier 25 always be appended to the emergency department (ED)E/M code (99281-99285) when provided on the same date as a diagnostic medical/surgical and/or therapeutic medical/surgical procedure(s). (Emphasis in the original) CMS Transmittal A0040 9 19
Issues Auditor Knowledge Auditor applied MS-DRG to Medicaid Claims MS-DRG is Medicare-specific May not apply to Medicaid patients Particularly for newborns, pediatric, and maternity cases As stated previously, we do not have the expertise or data to maintain the CMS DRGs for newborns, pediatric, and maternity patients. Medicare Program; Proposed Changes to the Hospital Inpatient Prospective Payment Systems and Fiscal Year 2008 Rates Federal Register, May 3, 2007 10 Issues Demographics The Medicaid population is distinct from other populations High percentage of women and children Complexity and comorbidities Irregular access to care Mental health issues Chemical dependency 20
Issues Demographics... the design and delivery of health services to Medicaid enrollees can be particularly challenging enrollees are often irregular users of the system; their diagnoses are frequently complicated by risk factors and social issues that render treatment plans ineffective... Medicaid Medical Management: A Complex Challenge for States 11 Deloitte Center for Health Solutions Expected Findings 21
Findings From the Virginia findings: More oversight of managed care is needed to ensure rates exclude improper payments Improper payments in capitated rates present a growing risk requiring effective oversight Oversight of MCO program integrity activities is insufficient to prevent risk of improper payments Oversight of MCO expenditure data is insufficient to ensure improper payments do not inflate capitated rates Findings VA Department of Medical Assistance response to JLARC audit report 5 22
Findings VA Department of Medical Assistance response to JLARC audit report 5 Questions 23
End Notes 1. http://www.gao.gov/assets/310/308487.pdf 2. http://jlarc.virginia.gov/reports/rpt404.pdf 3. http://www.nj.gov/comptroller/news/docs/horizon_report.pdf 4. http://oig.hhs.gov/oei/reports/oei-01-09-00550.pdf 5. http://jlarc.virginia.gov/meetings/october11/medicaid.pdf 6. http://www.cms.gov/research-statistics-data-and-systems/monitoring-programs/medicare-ffs-compliance- Programs/PERM/Downloads/CMSPERMmanual.pdf 7. https://www.cms.gov/regulations-and-guidance/guidance/manuals/downloads/mpi115c01.pdf 8. http://www.aha.org/content/12/12q1ractracresults.pdf 9. http://www.cms.gov/regulations-and-guidance/guidance/transmittals/downloads/a0040.pdf 10. http://www.cms.gov/medicare/medicare-fee-for-service-payment/acuteinpatientpps/downloads/cms-1533-p.pdf 11. http://www.deloitte.com/assets/dcomunitedstates/local%20assets/documents/us_chs_medicaidmedicalmanagement_0908w.pdf 24