@ PCG |Health Public Focus. Proven Results.” New Mexico Human Services Department Behavioral Health Provider Audits Final Report CONFIDENTIAL June 21, 2013 atc costing Gi he an At Act Cepeh ny EgSine of New Mexia Homan Seviees Deparment EU Dehra ve Ae EXECUTIVE SUMMARY {In February 2013, the New Mexico Human Services Department (HSD) contracted with Public Consulting Group Ine, (PCG) to aut fieen (15) mental health and substance abuse providers statewide. In 2012, these providers constituted approximately 87% ofall Core Service Agency (CSA) spending for Medicaid and non-Medicaid behavioral health servies!. PCG's audit ‘consisted of three main components 1) Clinical Case File Ault ~ a review of case fle documentation, including safing ‘qualifications and eredenils; 2) Billing Systems Audit ~ a review ofthe billing system itself, as wel as the protocols ‘and processes employed by the provide; and, 3) Enerprse Audit ~ a review of the organization and its key stakeholders, third party contracts, and other stakeholder relationships. ing. an approsch developed and refined through auditing behavioral health providers e nationally and ttlored to New Mexico's payment rules and regulations, PCG's multi-faceted suit arived atthe following Findings: 1) Clinical Findings: Wentfed more than $36.0 million in overpayments to these 15 providers over 8 thee-year period from 2009-2012. This amounts to nearly 15% ofall payments made to these providers. A 2003 Congressional General Accounting Office (GAO) report stated that Medicaid fraud, waste, and abuse is expected to be 3% 1 9% oF all payments. PCG recommends the collection of these overpayments 2) ITiiling System Findings: No material findings, though PCG did idetfy weaknesses in provider billing processes, including lack of audit tails when it eomes to changes made in systems. Generally, PCG recommends that providers tighten billing process ‘controls 3) Enterprise Findings: \dentfed potential confiets of interests of some individuals and some of the audited provides. PCG recommends thatthe State of New Mexico further ‘review instances of potential conflicts of interest «Contos gen ot pve pont a een tt ye ew Mes kee ere rg ee a fet, malate rataae co ameedsceeage a Ce acter etn te ey eo =x PagelI mall ect pum cys Behav Hei Prove Aude Sa ‘Execute Suny Summary of Clinical Audit CG's clinical ease fle review utilized two different methodologies foreach provider: 1) Random sampling of provider claims ~ Audit of 150 randomly sampled claims that were submitted by the providers. The sampling methodology allows for a statistically valid extrapolation of the findings. 2) Consumer ease file review ~ A review ofa full year's worth of case file document for selected consumers, These findings are not extrapolated, but ean be used to identify deficiencies that cannot be identified when viewing a single claim, PGs clinical case fle review revealed moderate to significant levels of non-compliance with state payment rules and regulations. Generally, the provides reviewed inthis audit lack many of the appropriate safeguards against overiling and would benefit from targeted technical assistance. Additionally, PCG's findings reveal deficiencies in accuracy of clinical documentation, which signifies potential quality of care concerns that shouldbe further reviewed bythe State of New Mexico, PCG utilized an audit too! developed and refined through auditing behaviors! health providers rationally and tailored to New Mexico's payment rules and regulations. For the randomly sampled clsims PCO uilized a statistically significant extrapolation methodology to identify ‘more than $33.8 milion in overpayments to these |5 providers over a three-year period from 2009-2012. With te consumer case file, oF “longitudinal,” reviews PCG identified an ‘addtional S2.1 millon in overpayments to these 15 providers over the same thre year period, for total extimated overpayments of $36.0 milion (nesrly 15% of claims paid during this petiod). Below are non-compliance rates and extrapolated overpayments by provider ToeitRemiais fini ne , vnleria aoeee [ae eee woue | ae ee | ee ane ae wet ae eee ee fe te feat ue ee ae a fea | ae pee fae tae oer ee ee fees [ae | eee ee if ee feel eee eee foe ae eee te ae ee ee ee ee ae reese eee ee ee ee ee tee ee fae pa tee got fae eee [em ae ae eee aS eee Its important to note that only the more egregious errors were used to extrapolate the amounts owed across the universe of claims for these providers. & more sct review ofthe randomly sampled provide claims originally indicated a non-compliance rate of 74%, PCG classified a numberof thes findings as “poor documentation practices” that should be remedied through a combination of tainings, technical assistance, and clinical and management assistance. “These errors included missing signatures, inadequate case note competion, and below standard preparation of plans of care. Had PCG used these erors in the extrapolation, the resulting ‘overpayment amounts would have been much greater. PCG considers the extent ofits findings to bea significant concer forthe Sate of New Mexico. In 2 2003 repor” the Congressional General Accounting Office (GAO) estimated that fraud, waste, and abuse amounted to between 3% and 9% of total Mediesidsrending. Using this GAO ‘study as base, this audit reveals overpayments that are double what can be expected. coca Acovtng Ofc, “ajo ManagenentChaleges nd Popa Risks Deparment of Heath an Human Services" 2003 bap wor gn gov esetsa40237027 pot agemill fil meine tits chav eats Provide Audis FURL Cons Executive Say Summary of TPBillng Systems Aual CG did not identify any specific instances of tampering with the providers’ billing systems. This finding mst be qualified for several reasons. First, PCG was unable to complete 8 comprehensive review of all billing systems as one of the billing systems vendors, Anasazi, prohibited providers from sharing system manuals, as they were considered proprietary (noted in that PCG viewed from Anasazi to one ofthe audited providers). Aditionaly, PCG lading identified areas of weaknesses in provider practices, ‘© Lack of audit tral for the creation of and changes made w claim records in provider billing systems; + Lack of audit irl for any changes made to the 837 reports ling system outputs) prior to finalizing inthe Automated Clearing House portal Summary of Enterprise Ault Lastly, PCG's enterprise audit sought to a) provide the state witha clearer view of how its provider system is organized and b) identity any potential appearances of conflicts of interest for the organization and its key board members and employees. The enterprise audit revealed that some providers may have potential conflicts of interest that should be further reviewed by ‘the State of New Mexico. Examples of the types of potential confit of interest and areas that, CG recommends further research include ‘+ Unusual compensation andor benefits to some key stakeholders; ‘© Key stakeholders’ relationships with releted parties with financial interests in veanset ‘+ Some arrangements with thied parties are unclear as © the level of effor and compensation for some executives and, + Non-dislosure of ll thied party contracts. Scorecard and Risk Tier Results Based on the elnial case file compliance outcomes and findings related to IT controls, PCG developed, in conjunction with HSD, a “seorecard™ for each provider. Below, PCG has eePRL oNSLING 7 ‘organized the providers’ scorecard resus in relation to each other. The scorecard ranges from “Significant Non-Compliance” to “Compliant.” Camplance PCG then used these provider scorecard ratings to categorize providers into “Risk Tier” replete With recommended state ations, a8 flows: Findings that include missing Provide trainings and elnical, documents te. assistance as needed. 7 Significant volume of findings that] Provide trainings and clinical include missing documents assistance as needed. ‘Potentially embedcliical management {improve process. | Significant findings, including Provide —talnings — and lineal signifean quay of care findings. assistance as needed. ‘+ Potentially embedclinical management to improve processes. Potential change in management. "> Mandatory changein management. PaeeVmii Sua often Mesen iM Il wena uae got ovis Based on PCG's scorecard methodology, each ofthe 1S providers was categorized into a Risk ‘Tle, the results of which are shown below. 1 ‘© Provide trainings and clinical assistance as needed. 2 > Provide Wanings and clinical assistance as | M,C,1,D,J,L,HyandN needed. + Potentially embed clinical management to improve processes. 3+ Provide trainings and elinieal assistance as] E,G,A,F,K,O and B needed, ‘+ Potentially embed clinical management to lmprove processes. Potential change in managerent q ‘= Mandatory change in management ‘See NOTE, below NOTE». Please note that Ter 4: Credible Allegation f Fraud isa determination tht can only ‘be made bythe Sate of New Mexico. PCG ualized results from its clinical case file audit and Thing. system audit to develop the scorecard, which translated into providers being ‘categorized in Tiers 1, 2, and 3. The State of New Mexico may determine that information ‘provided inthe case fle, IT/biling system, and enterprise audits constiutes a re-categorization ‘of one or more providers into a higher riskier, including Ter 4. Rackgeon {In February 2013, the New Mexico Human Services Department (HSD) determined the need for 18 comprehensive clinical and billing auit of select providers within its behavioral heath system and engaged Public Consulting Group (PCG) to conduet these audis. Claims data mining bythe state's behavioral health vendor revealed a significant numberof potential biling abnormalities. These potential billing abnormalities included, but were not limited to, the following data and case file “findings” Tae vi‘Soe ofNew Meso Homan Serves Department Belair Heath Provider Aa Exectve Samay i a ~ Cross billing at diferent locations for the same member pote uncertain as o who rendered the service (iF rendered at all); + Insufficient documentation; = Cross billing mulkiple codes and double billing (eg individual and group therapy); Upcoding individual therapy (compared to the average time billed per code inthe peer group) Excessive billing for psychosocial rehab; including requesting authorization for a consumer on medical leave; = Suspicious high volume days per one code; ovebilling for inappropriate codes; psychosocial rehabilitation billed for large units on a given date to one clinician; excessive hours per day billed by practitioner; excessive hours of service billed per patent per code; billing for services duplicative in nature; = eentitying Provider a the rendering clinician; [No medal necessity reviews to determine basis for long-term psychotherapy Forging clinician recorés to incorporate more ime than truly performed; (Out of home placement services outside norm of service; doubtful medical needs ~ Billing outpatient services the same day as bundled services, ally overlapping time: [Not all of the aforementioned potential billing issues can be addressed with a single audit, panicularly when an objective ofthe aut isto identify recoupable overpayments, In order to recoup across a universe of pad claims, a more comprehensive review is required, Nerowly focusing on one particular suspicious trend in «provider's claims history inhibits the ability of the auditor and the state to extrapolate those results aeross the entire claims history, Rather than tempting to address each provider's uniquely identified issues, PCG worked with HSD to develop a comprehensive approach that would serutnie individual providers holistically (as ‘opposed to looking ata few aberrant trends that may or may not run afoul of policy even if substantiated) and the system at large. This approach was characterized by three main goals: 1) Identify potential credible allegations of fraudulent activity. 2) ent egulatry compliance lvls of behavior! heat providers 3) entity areas of weakness that must be strengthened prior to the implementation of Centennial Care. PCG was tasked with conducting onsite audits of selected providers to examine case files supporting specific claims, IT systems and processes, and edherence with compliance protocols, {nd to examine existing relationship, financial or other, among providers and other entities, The ‘onsite audits were conducted in February and March and included interviews with relevantanil vmaticatee — @ rau Cons Behavioral Healy Provider Ate Excetve Summary provider staff, collection of hard copy and electronic file documents related to the sbove mentioned areas, and examination and manual testing of IT systems. The onsite visits were supplemented by desk reviews of collected documentation at a locaton separate from the provider site, PCG's approach is deserted in more dealin the body ofthe fll report. Key Findings While each provider is unique wih respect to clinical findings, PCG identified certain cormmon themes across many ofthe 5 provides reviewed, which ae described below. For each provider, 2 section is included in the appendix that shows the detailed clinical findings specific to that provider, PCG's findings include: ‘+ More than $37.3 million in overpayments for these 15 providers over a three end a half year period July 2009-January 2013). This extrapolated overpayment amounts 10 15% of {ota payments from state sources to these providers during this time period ‘+ Non-compliance with many New Mexico state rules and regulations. Pervasive issues that PG identified across providers include: e Randomly Sampled Clains 1© Community Support Workers lacked evidence of completion of the requted ‘raining per the service definition, ‘© Assessments (psychosocalpsychatic evaluations) were not up to date (within last 12 months) to determine if the consumer continued to meet the need of the rendered service. Incomplete critical information suchas Five Axis diagnosis ‘+ Substance abuse history was absent for most consumers with a dual iagnosis of mental health and substance abuse ‘©. Treatment plans were not up-to-date and individualized per consumer. Updated treatment plans are necessary to determine any changes to goalsfobjectves in addition o progress or lack of progress by the consumer. Without continuously ‘updated treatment plans, it is impossible to determine if the treatment imervetion till meet the behavioral health needs ofthe consumer. * GoalsObjectives were not measurable and did not document achievable target dates based onthe consumer's needs. Service specific clinical interventions used to reach goalslobjectves were absent @e fim i Sin of Now Meso Hanan Soe Depernent unui gongunve avira eh roi: Ae Execute Sma * Discharge plans and estimated length of teament were not documented ‘© Consumer Documentation ' Consents for medications rendered were abser. Documentation frequently did not describe the clinical interventions, rogres o lack of progress toward zoals,andnex steps in treatment ‘Interventions inthe progress notes didnot alvays lnk to the consumer tweatment plan or support the program definition of the billed service. ‘Progress notes did not contain a start and sop time of @ duration that would enable determination as to whether the bile time was accurate, * Billed unite did not match the units documented onthe progress notes. Intensive Outpatient Program progress notes di not contin the treatment ‘modalities used a required in the service defriton * Documented evidence of the required treatment team was absent for most team services. Longitudinal File Review Findings SefetyRisk Assessments were not completed or updaed for consumers who were assessed t0 have curent or past suicidal ideations (S.), homicidal ideatons (Hl), ‘selTharm or domestic violence issues ‘Treatment plans were not up-to-date and individualized per consumer. * Plans conained the same gols/objectves for more than 12 months, Potential overutilization of services without documented justification of the service relate to extensive length of stay. Consumer Documentation Documentation frequently did not describe the clinical interventions, progress or lack of progress toward goals and 2ext steps in treatment. 1 Progress notes did not contain a start and stop time or @ duration that ‘would enable a determination as to whether the billed time was accurate, ‘Billed unis did not match the units documented on the progress notes. ‘+ Weaknesses identified in providers" billing processes. PCG identified weaknesses in internal claims processes. PCG was unable to complete a comprehensive review of al billing systems as one particular billing software vendor was unwilling to allow providers to share with PCG important documentation and information abou the system.