TRICARE Operations Manual 6010.62-M, April 2021
Program Integrity
Chapter 13
Section 3
StatisticalSampling And Audits
Revision:
1.0AUDITASSISTANCE SOFTWARE
1.1The contractor shall utilizethe Regional Advanced Techniques Staff (RAT-STATS) software suiteto support its auditing process. There are a number of well-known,reputable software statistical packages (e.g., SAS, R) that maybe used in conjunction with RAT-STATS.
1.2RAT-STATSwhen performing audit sampling. RAT-STATS is a statistical samplingsoftware package created by the Office of Inspector General (OIG)to facilitate the auditing process.
1.3RAT-STATSoffers an assortment of capabilities which include, but are notlimited to, random number generation, various sampling techniques,sample size calculation, and extrapolation.
1.4The contractorshall download RAT-STATS from the Department of Health and HumanServices (DHHS) website. A number of important statistical considerationsrelated to the auditing process are summarized in paragraph 2.0.
2.0STATISTICALAUDIT METHODOLOGIES
2.1The contractor shall use atechnically sound statistical analysis method and sampling design.
2.2The contractorshall use current statistical methods to determine sample size,design of sample and analytical methods as determined by the scopeand the nature of the audit.
2.2.1The contractor shall performand design the audit individually on a case-by-case basis.
2.2.2The contractorshall eliminate zero paid claims from the universe before the sampleselection. This includes claims which were not denied, have allowableamounts, but zero dollars were paid.
2.2.3Simplerandom sampling involves using a random selection method to drawa fixed number of sampling units from the frame without replacement(i.e., not allowing the same sampling unit to be selected more thanonce).
2.3Stratified sampling involvesclassifying the sampling units in the frame into non overlappinggroups, or strata. The stratification scheme ensures that a samplingunit from a particular stratum is more likely to be similar in overpaymentamount to others in its stratum than to sampling units in otherstrata. Other sampling methods include, but are not limited to,multistage sampling, cluster sampling, and stratified cluster sampling.
2.4The contractorshall utilize a qualified statistician when developing statisticalsamples.
3.0PROBE SAMPLE AUDIT
3.1The probesample shall employ simple random sampling if it is to be used toevaluate the variability of overpayments in the claim universe.The results of the probe sample audit may trigger the need for thecontractor to perform a Statistically Valid Random Sample (SVRS)audit or a 100% audit.
3.2Statistically Valid RandomSample (SVRS)
3.2.1Statistical sampling is usedto investigate and project (e.g., extrapolate) the amount of overpayment(s) madeon claims when a 100% audit is not used. Statistical sampling iscommonly used in consideration of a large number of claims of potentiallyfraudulent claims and the dollar values associated with those claimsgiven available resources to conduct the audit.
3.2.2Thereare four key conditions that must be met in order for a SVRS toproduce a valid extrapolation. These conditions are:
•Properly defining the samplingunit, sampling frame, and the universe;
•Applying an appropriate samplingmethodology;
•Calculating a suitable samplesize; and
•Sample the units sampled inthe frame so that it is representative of the fraudulent claims.
3.2.3Focusingthe universe is performed by targeting specific claims which matchthe allegations of the case, and removing low dollar claims thatwill not be worth of recoupment.
3.2.3.1The contractor shall selectsampling units from the applicable sampling frames via a random samplingapproach once the claims universe has been focused and evaluatedto determine the sampling plan and methodology to be performed.
3.2.3.2The contractor shall electronicallygenerate sample sizes and random numbers customized for each auditperformed using RAT-STATS software referenced in paragraph 1.0.
3.2.3.3The focused claims universeshall be sorted in Internal Control Number (ICN) ascending orderwhen the software’s sampled numbers are applied.
3.2.3.4The contractor shall eliminatezero paid claims from the universe before the sample selection.This includes claims which were not denied, have allowable amounts,but zero dollars were paid.
3.2.4The contractorshall record and achieve seed number of the random sample(s) sothat the sample(s) can be recreated at a later date if necessary.The overall sampling plan and methodology may include a combinationof sampling approaches consisting of one or more SVRS or 100% claimsaudit(s).
3.3Stratified Sample
3.3.1When thefocused universe of claims includes diversity in a key characteristicsuch as paid dollar amounts or numbers of procedures billed withinthe claim, the efficiency of the audit is improved by employinga stratified sampling approach. If the potentially fraudulent claimsrepresent diverse subpopulations the audit will be stratified bythe contractor.
3.3.2The contractor shall employstratification when the universe of claims where the fraud is suspectedto have occurred has a standard deviation greater than the averageoverpayment. This characteristic of the claims universe will beidentified by the contractor using a random probe sample as describedin paragraph 3.0, or by evaluation of the averageand standard deviation of paid amounts using the software referencedin paragraph 1.0. The paid amounts representthe upper bound of possible overpayment for each claim when theaverage overpayment is unknown and can be used for estimation.
3.3.3The stratificationwill then be employed to split the universe of claims where thefraud was suspected to have occurred into smaller, non-overlappinggroups, or strata. Within a stratum, the claims are more similarto each other than to the other claims in the universe of fraudulentclaims.
3.3.3.1The contractor may use theirdiscretion in identifying the number and characteristics of strata neededfor an audit, such that each stratum has a standard deviation lessthan the average paid or overpayment amount. Each stratum’s subsetof claims will then be sampled using a random selection method ora 100% audit. A stratified audit when properly designed will requirethe sampling of fewer claims overall or have improved statisticalresults for the generally extrapolated amounts, because the variabilitywill be decreased in each stratum.
3.3.3.2The contractor may apply additionalsampling methods at their discretion including, but not limited to,multistage sampling, cluster sampling, and stratified cluster sampling.
3.4One-HundredPercent (100%) Claims Audit
The contractor may need toperform a 100% claims audit.
3.4.1Situationsmay include a small stratum of high dollar claims which should beaudited at 100% as part of stratified sampling approach. Alternatively,even lower dollar claims may need to be audited at 100% if the claims arenot similar (in terms of procedure, paid amount, or other characteristics)to a large group of other claims in the universe.
3.4.2In thevast majority of cases, the unit to be statistically sampled isthe entire claim (which includes all paid line-items).
3.4.3The contractorshall review the beneficiary’s entire Episode of Care (EOC) as partof the audit when the circ*mstances dictate.
3.4.3.1In this case, the contractorshall discuss the specifics of this approach with Defense HealthAgency (DHA) Program Integrity Office (PI) prior to selecting thesample(s) as there are ways of auditing the beneficiary’s EOC whilestill using the claim as the sampling unit.
3.4.3.2In other unusual circ*mstances,a probe sample audit may be required (e.g., an audit that is not statisticallyvalid). A statistically valid sample may or may not follow the probesample audit.
3.5ExternalAudit - Beneficiary Inquiry
3.5.1A secondary method of determiningprobable fraudulent practices is to conduct a verification of serviceswith beneficiaries. This may be used to supplement a claims auditmethod, and address 100% of the beneficiaries who received servicesfrom a provider within a recent period of no more than one year.If the provider is seeing more than 50 beneficiaries for which claimshave been submitted, a systematic sample may be used. (e.g., aninterval of every fifth, 10th, claim). Generally, no less than 50external verification letters shall be sent. A suspense period forresponses to the verifications letters shall be 30 business days,with a follow-up either written, or by phone by the 30th businessday. If the response is not received, the contractor may contactthe beneficiary by phone to receive responses.
3.5.2An externalaudit to the provider shall be conducted where a bill has been allegedlyaltered.
4.0AUDITS
4.1Auditsare performed to examine and verify the accuracy of claims.
4.2The contractorshall determine the type of audit appropriate for the particularcirc*mstances on a case-by-case basis.
4.3The contractor’sProgram Integrity staff who perform records reviews and audits shallinclude an appropriate number of Registered Nurses (RNs), or equallyqualified medically trained professional(s) and credentialed codingor credentialed medical billing professionals with the educationand anti-fraud experience.
4.3.1These personnel must have athorough knowledge of TRICARE regulatory provisions, policy, and standards.The reviewer shall document, in detail, the rationale for the auditfindings. The review must be dated and include the clinical specialtyand qualifications of the reviewer and the signature (not initials)and the legibly printed name of the reviewer.
4.3.2The contractorshall have a credentialed coding or credentialed medical billingprofessional or both on staff that has professional coding, reimbursem*nt,and anti-fraud education and experience that provides support inthe performance of medical records audits and program integrityfunctions.
4.3.3In cases where clinical issuesmay be included in an audit, the contractor’s audit team shall includea member with clinical expertise in the area under review.
4.4ReportingAudit Findings
4.4.1The contractor shall reportaudit findings in a clear and concise manner in an automated spreadsheet, accompaniedby a description of the audit with summary information in quantifiableterms.
4.4.1.1The contractor shall includethe findings on the DHA PI Audit Worksheet for each SVRS performed. Forreporting requirements, see DD Form 1423, Contract Data RequirementsList (CDRL), located in Section J of the applicable contract. Thesupporting audit spreadsheets shall provide the criteria used fordetermination of overpayments (e.g., no entry, not a benefit).
4.4.1.2An analysis of the frequencyof the occurrence of overpayments can lead to conclusions concerning furtherinvestigative actions. Other methods of analyses may be used concerningabusive practices.
4.4.2The contractorshall include individual audit sheets documenting individual findings(which will then be summarized in the audit worksheet(s) (e.g.,overpayment summary by claim line, audit summary report, extrapolation(sample) verification spreadsheet). Individual file folders, labeledwith identifying information, shall be generated as appropriateand must contain all applicable documentation or data required tosupport the audit finding, which will include but not be limitedto: claim copy, explanation of benefits, individual audit sheets, evaluationand management score sheet, and medical record documentation reviewedby the auditor.
4.5Medical Necessity Audits forMedical or Mental Health Claims
4.5.1The contractor shall performmedical necessity audits that are performed by a RN, or equallyqualified medically trained professional, who can make medical judgmentsbased on professional education and experience. This means RNs orqualified Physician’s Assistants (PAs) for medical claims. A qualifiedLicensed Practical Nurse (LPN) or Licensed Vocational Nurse (LVN),working directly under the close supervision of an RN or PA, maybe used, if the contractor submits the LPN’s or LVN’s full resumeand a detailed scope of authority and responsibility to the ContractingOfficer’s Representative (COR) for approval before the LPN or LVNassumes a medical review role. For mental health claims, a clinicalpsychologist, psychiatric nurse practitioner, a psychiatrist oran equally qualified professional shall perform the audit.
4.5.2Thesepersonnel shall have a thorough knowledge of TRICARE regulatoryprovisions, policy, and standards.
4.5.3The reviewershall document, in detail, the rationale for the audit findings.
4.5.4The reviewshall be dated and include the clinical specialty and qualificationsof the reviewer and the signature (not initials) and the legiblyprinted name of the reviewer.
4.5.5Claimsthat the reviewer cannot make a determination on shall be referredto the contractor’s medical staff, or an external consultant atno cost to the Government. Use of medical staff or consultants isexpected and required not only for initial reviews but post-paymentanalyses and audit requests from DHA PI.
4.5.6The contractorshall involve physicians, consultants with a specialty appropriateto the case and allegations whenever the case is complex, physiciansor consultants at no cost to the Government.
4.5.7Othertypes of audits shall be performed to suit the allegations or aberrantbilling practices such as probe, non-invasive, EOC, or calendarand are left up to the determination of the contractor. This shallinclude also utilizing other investigative techniques such as licenseverification and Internet research.
4.6PrescriptionRecords Audit for Pharmacy Claims Or Pharmacy Claims Audit
4.6.1Auditsshall be performed by a qualified trained professional, who canmake judgments based on professional education and experience suchas a certified pharmacy technician, a pharmacist, Doctor of Pharmacy oran equally qualified trained professional.
4.6.2Thesepersonnel shall have a thorough knowledge of TRICARE regulatoryprovisions, policy and standards.
4.6.3The reviewershall document, in detail, the rationale for the audit findings.
4.6.4The reviewshall be dated and include the clinical specialty and qualificationsof the reviewer and the signature (not initials) and the legiblyprinted name of the reviewer.
4.6.5Claimsthat the reviewer cannot make a determination on shall be referredto the contractor’s pharmacy staff (or if available, medical staff)or an external consultant at no cost to the Government. Use of pharmacystaff or an external consultant is expected and required not onlyfor initial reviews but post-payment analyses and audit requestsfrom DHA PI.
4.6.6The contractor shall performother types of audits to suit the allegations or aberrant billingpractices such as probe, non-invasive, etc.
4.6.7The contractorshall utilize other investigative techniques such as performingpurchase verification, license verification, and Internet research.
4.7DentalNecessity Audits for Dental Claims
4.7.1Dental necessity audits mustbe performed by a qualified trained professional, who can make judgmentsbased on professional education and experience such as a certifieddental technician, a dentist, or an equally qualified trained professional.
4.7.2Thesepersonnel shall have a thorough knowledge of TRICARE regulatoryprovisions, applicable contract, policy and standards.
4.7.3The reviewershall document, in detail, the rationale for the audit findings.The review shall be dated and include the clinical specialty andqualifications of the reviewer and the signature (not initials)and the legibly printed name of the reviewer.
4.7.4Claimsthat the reviewer cannot make a determination on shall be referredto the contractor’s dental staff or an external consultant at nocost to the Government. Use of dental staff or consultants is expectedand required not only for initial reviews but post-payment analysesand audit requests from DHA PI.
4.7.5Othertypes of audits shall be performed to suit the allegations or aberrantbilling practices (e.g., probe, non-invasive, EOC). This shall alsoinclude utilizing other investigative techniques such as licenseverification, and Internet research.
4.8The contractorshall ensure compliance with the Alcohol, Drug Abuse and MentalHealth Administration (ADAMHA) Reorganization Act, Public Law 102-321(July 10, 1992) and implementing regulations including 42 Code ofFederal Regulations (CFR) Part 2, when data requested includes servicesrelated to substance abuse or mental health and medical recordsrequests.
4.9Request for Mental Health Records- Psychotherapy Notes
The contractorshall not request that a provider submit psychotherapy notes asdefined by “notes recorded by a mental health professional whichdocument or analyze the contents of a counseling session and thatare separated from the rest of the medical record”. The refusalto submit such information shall not result in the denial of theclaim or allegations of fraud or abuse. The provider is responsiblefor extracting information to support that the claim is for reasonableand necessary services. If the provider does not submit informationto substantiate that the services were medically necessary, theclaim shall be denied.
- END -
