STATE OF FLORIDA

DIVISION OF ADMINISTRATIVE HEARINGS

BARBARA POR SRUR, M.D., )

)

Petitioner, )

)

vs. ) Case No. 96-4830

)

AGENCY FOR HEALTH CARE )

ADMINISTRATION, )

)

Respondent. )

)

RECOMMENDED ORDER

On July 31 and August 1, 1997, a formal administrative hearing was held in this case in Palmetto, Florida, before

1. Lawrence Johnston, Administrative Law Judge, Division of Administrative Hearings.

   APPEARANCES

   
   

   For Petitioner: John D. Buchanan, Jr., Esquire

   Henry, Buchanan, Hudson, Suber & Williams, P.A.

   117 South Gadsden Street Tallahassee, Florida 32302

   
   

   For: Respondent: Thomas Falkinburg, Esquire

   Moses E. Williams, Esquire

   Agency for Health Care Administration Fort Knox Executive Center No. 1

   2727 Mahan Drive

   Tallahassee, Florida 32308

   
   

   STATEMENT OF THE ISSUES

   
   

   The issue in this case is whether the Agency for Health Care Administration (AHCA) should recover from the Petitioner, Barbara

   Por Srur, M.D., alleged overpayment of Medicaid reimbursement for psychiatric services to Medicaid patients.

   PRELIMINARY STATEMENT

   
   

   On September 4, 1996, the AHCA issued a Final Agency Audit Report for January 1, 1991, through December 31, 1993, which reflected that the Petitioner owed AHCA $113,756.93 in overpayments of Medicaid reimbursement for psychiatric services, plus a $2,000 administrative sanction. On September 23, 1996, Dr. Por Srur filed a Petition for a Formal Hearing Pursuant to Section 120.57, F.S., and the matter was referred to the Division of Administrative Hearings (DOAH) on October 11, 1996. Based on the Joint Response to Initial Order, the case initially was scheduled for hearing on February 13-14, 1997.

   During the discovery deposition of the AHCA's auditor, the Petitioner learned that the auditor had relied on an independent review of the Petitioner's reimbursement claims conducted by a consulting psychiatrist, but AHCA would not disclose the identity of the consultant. The Petitioner filed a Motion to Compel Discovery on December 31, 1996. The AHCA did not respond the motion.

   On January 13, 1997, the Petitioner filed an uncontested Motion for Continuance based on a death in the family of Petitioner's counsel. On January 16, 1997, an Order Continuing Final Hearing and Compelling Discovery was entered. Final hearing was continued until May 15-16, 1997.

   On April 15, 1997, new counsel for the AHCA filed Respondent's Motion for Reconsideration of the Court's

   January 16, 1997, Order. The Petitioner responded in opposition, and reconsideration was denied.

   On April 30, 1997, AHCA filed Respondent's Motion to Amend Final Agency Audit Report. It requested leave to delete from the report claims for services rendered prior to September 4, 1991, to reduce the alleged overpayment sought to be recovered to

   $59,859.72, and to reduce the fine to $100. AHCA also filed a Motion for Continuance based on the Respondent's Motion to Amend Final Agency Audit Report and on AHCA's desire for a second independent review by another psychiatric consultant. The Petitioner did not oppose either motion, and they were granted. Final hearing was continued until July 31 through August 1, 1997.

   On May 1, 1997, AHCA filed Respondent's Motion to Bifurcate Final Hearing to postpone the receipt of evidence on the validity of the sampling technique and statistical formula used by AHCA to extrapolate the findings from specific claims it audited to the entirety of the claims submitted by the Petitioner during the relevant time period. The Petitioner filed a response in opposition to this motion. An Order Denying Bifurcation of Final Hearing was entered on May 13, 1997.

   The parties filed a Pretrial Hearing Stipulation on July 21, 1997.

   On July 22, 1997, the AHCA filed a Motion for Official Recognition of State of Florida Records. It sought official recognition of an Adjustment to Amended Final Agency Audit Report reducing the overpayment claim to $58,341.26. As reflected in the Findings of Fact, this "adjustment" actually was a desk audit that did not take into account any independent review by either consulting psychiatrist. The Petitioner objected and filed a response in opposition. AHCA filed a reply.

   At the outset of the final hearing on July 31, 1997, AHCA's Motion for Official Recognition of State of Florida Records was denied because the records were not the kind that can be officially recognized. It also was ruled that the question of their admission in evidence was premature.

   The Petitioner proceeded to present her case-in-chief, which consisted of her testimony and the testimony of three other witnesses. In addition, the transcripts of the deposition testimony of four other witnesses were admitted in evidence as Petitioner's Exhibits 3 through 7. (By agreement, the Petitioner was given permission to late-file her Exhibits 5 and 6.) Petitioner's Exhibits 1 and 2 also were admitted in evidence.

   (By agreement, the Petitioner was given permission to late-file her Exhibit 2.) Respondent's Exhibits 10 and 11 also were received as part of the Petitioner's case-in-chief.

   After the Petitioner's case-in-chief, AHCA moved ore tenus for a partial summary recommended order, and the motion was

   denied.

   
   

   AHCA called two witnesses and had Respondent's Exhibits 1, 6, 8, 9, 13-15, 17 admitted in evidence in its case-in-chief. By agreement, AHCA was permitted to late-file Respondent's Exhibit 17, the deposition testimony of an expert witness.

   Ruling was reserved on Respondent's Exhibits 12 and 16. It is now ruled that the objections to the admissibility of those exhibits are overruled, and they are admitted in evidence.

   At the end of the hearing, the parties asked for and were given 30 days from the filing of the transcript of the final hearing, if one was ordered, to file proposed recommended orders. AHCA then renewed its ore tenus motion for a partial summary recommended order, and the motion was again denied.

   The transcript of the final hearing was filed on August 19, 1997, but on September 10, 1997, the Petitioner filed an Unopposed Motion to Extend Time to File the Proposed Recommended Order until October 13, 1997, and the requested extension was granted.

   FINDINGS OF FACT

   
   

   The Petitioner's Practice

   
   

   1. The Petitioner, Barbara Por Srur, M.D., provided psychiatric services to Medicaid patients during the time period from January 1, 1991, through December 31, 1993. Some of these patients were outpatients seen in the Petitioner's private office, but most were either inpatients at Manatee Memorial

      Hospital or participants in the hospital's Partial Hospitalization Program (PHP). Under the PHP, patients essentially used the hospital for outpatient services; they were required to report to the hospital periodically for evaluation and treatment as an alternative to inpatient services.

   2. From a combination of these patient sources, the Petitioner's caseload was heavy. Many of her patients were adolescents, and many were Medicaid patients. Both adolescent and Medicaid patients tend to make relatively heavy demands on the time of a psychiatrist. Both tend to be relatively complicated psychiatric cases. In addition, Medicaid patients are more likely to require medical interventions that have to be coordinated with the delivery of psychiatric care, and their cases are more likely to require the psychiatrist to deal with various social workers also providing services to the patient, in addition to dealing with family members.

   3. The Petitioner had an agreement with another psychiatrist, a Dr. Goldman, to cover for each other so that both could take days off. When Dr. Goldman saw the Petitioner's patients, he would sign his name in the patient charts.

   4. The Petitioner signed a Medicaid Provider Agreement on September 1, 1988. By doing so, she agreed "to keep for 5 years complete and accurate medical and fiscal records that fully justify and disclosed the extent of the services rendered and billings made under the Medicaid program" and agreed to furnish

      them to AHCA upon request. She also agreed to "abide by the Florida Administrative Code, Florida Statutes, policies, procedures, manuals of the Florida Medicaid Program and Federal laws and regulations."

   5. Manatee Memorial maintained the patient charts for inpatient and PHP psychiatric patients. Usually, the Petitioner would have Manatee Memorial furnish her office with copies of the admission and discharge summaries on her inpatients and PHP patients, but the Petitioner herself would not maintain any other medical records on those patients. Sometimes Manatee Memorial would neglect to send the Petitioner her copies of the admission and discharge summaries on her inpatients and PHP patients.

   6. The Petitioner stipulated that she never documented time intervals in the medical records she maintained on her patients. The Petitioner probably was not alone in this practice. However, the evidence was clear that some psychiatrists, including the Petitioner's own expert witness, document time spent in individual psychotherapy in order to ensure Medicaid reimbursement.

   7. When the large number and amount of Medicaid claims being submitted by the Petitioner for the period from January 1, 1991, through December 31, 1993, came to the attention of the AHCA's Medicaid Fraud Detection Section, the matter was referred to an auditor for investigation and possible audit. The auditor had AHCA's fiscal agent, Consultec, generate an FLQPLPSY report

      estimating the hours per day (Medicaid and non-Medicaid) billed by the Petitioner during this time period. Based on the FLQPLPSY report, AHCA decided to proceed with an audit of the Petitioner's Medicaid reimbursement claims for the period from January 1, 1991, through December 31, 1993.

      AHCA's Audit Strategy

      
      

   8. Rather than audit all paid claims submitted by the Petitioner during the audit period, AHCA chose to audit a sample of the total claims and use a statistical formula to extrapolate the results of the sample audit to the total number of claims. It clearly was acceptable and valid to proceed in this fashion.

   9. AHCA next decided to sample the Petitioner's claims using cluster sampling rather than random sampling or some other method. In cluster sampling, clusters are randomly sampled. In the context of the Petitioner's Medicaid reimbursement claims, the Petitioner's patients were treated as clusters of claims attributable to those patients. Cluster sampling was the most efficient sampling method for use in this context. Although the Petitioner's patients differed in several respects--including whether they were inpatients, outpatients, or PHP patients-- cluster sampling was acceptable and valid.

   10. To accomplish the cluster sample, AHCA asked its fiscal agent, Consultec, to generate an ad hoc report on the total number of Medicaid claims billed by the Petitioner (11,673), the total number of patients to whom the bills were attributed (524),

       and the total number of Medicaid patients dollars paid to the Petitioner ($383,486.50) on those claims during the audit period. Next, the auditor had an office automation specialist utilize a computer program formula developed by Robert Peirce, the Administrator of AHCA's Office of Medicaid Program Integrity and a qualified expert in statistics, to generate a choice of sample size alternatives as a function of error proportion and proportion of point estimate prospectively recovered. From the choices given, the auditor selected 34 clusters (i.e., patients) as the sample size. The evidence proved that the sample size chosen by the auditor was acceptable and valid.

   11. The auditor then asked Consultec to randomly select 34 of the 524 Medicaid patients to whom the Petitioner's paid Medicaid bills were attributable. Consultec used a sequential random sample method to select the 34 clusters (patients). It would appear from the evidence that Consultec essentially divided

       524 by 34, got 22.28, and selected every 23rd patient. The resulting cluster sample yielded 741 claims to audit, approximately the same number of claims per patient as for the total number of claims and patients (524 and 11,673). This supports a finding that the sequential random sample method used was acceptable and valid.

   12. The statistical formula used by AHCA to extrapolate its audit findings from the 741 sample claims to the 11,673 total claims also was acceptable and valid. The formula used a "t"

       value from the Distribution of "t" Table of 1.6923602, which was appropriate for use when selecting 34 out of 524 clusters. The formula yielded not only a point estimate of total overpayment but also a lower total overpayment amount with a 95% confidence level. In using the formula, the AHCA seeks to recover only the lower total overpayment amount with a 95% confidence level.

       First KePro Consultant and the Final Audit Report

       
       

   13. AHCA's auditor next proceeded to audit the claims included in the sample. To do this, she obtained medical records from the Petitioner and from Manatee Memorial Hospital. The auditor then decided that she needed the assistance of a consulting psychiatrist to analyze the medical records to determine whether the Petitioner had billed for services using the most appropriate American Medical Association Physician's Current Procedural Terminology (CPT) codes. The billings and medical records were referred to an organization called KePro, which had a contract with the AHCA, for assignment to one of its consulting psychiatrists.

   14. After comparing the billings with the medical records, the KePro consultant reported to the AHCA auditor that many services were billed incorrectly. As will be seen, the report of this KePro consultant was not put in evidence, and the record is not clear as to all of the procedure code changes recommended by this consultant. However, the KePro consultant often reported to

       the AHCA auditor that the correct CPT codes provided for less reimbursement to the Petitioner.

   15. The AHCA auditor used the consultant's report to determine the overpayments associated with the 741 audited claims. She also rejected claims not supported by any medical records and claims for services rendered by a health care professional other than the Petitioner. The AHCA auditor then used the statistical formula to determine the point estimate of total overpayment and the total overpayment with a 95% confidence level.

   16. The AHCA auditor then prepared a Provisional Agency Audit Report, dated January 31, 1996, which provisionally determined that the point estimate total overpayment was

       $162,374.42 and that the total overpayment, with a 95% confidence level, was $142,680.44. In addition, AHCA assessed a $2,000 fine.

   17. Upon review of the Provisional Agency Audit Report, the Petitioner and her current billing clerk discovered that the AHCA had rejected many of the Petitioner's audited claims as not having the necessary supporting documentation because the AHCA had not obtained all of the medical records from Manatee Memorial. The Petitioner had her billing clerk go to Manatee Memorial to obtain the medical records for her inpatients and PHP patients and send them to the AHCA auditor. The Petitioner was able to supply the auditor with many of the missing records. But

       to the dismay of the Petitioner and her billing clerk, some records could not be found.

   18. Based on the receipt of the additional documentation from the Petitioner, AHCA issued a Final Agency Audit Report on September 4, 1996, which reduced the point estimate total overpayment to $125,669.85 and reduced the total overpayment with a 95% confidence level to $113,756.93. The fine remained the same.

       Second KePro Consultant

       
       

   19. When the Petitioner attempted to depose the KePro consultant during discovery in this proceeding, AHCA refused to identify the consultant. The Petitioner eventually obtained an Order Compelling Discovery, but instead of making the first KePro consultant available for deposition, AHCA chose to refer the case to a second KePro consultant.

   20. Subsequently, on April 30, 1997, AHCA filed Respondent's Motion to Amend Final Agency Audit Report. It requested leave to delete from the report claims for services rendered prior to September 4, 1991, which was five years prior to the Final Audit Report. This was done because the Petitioner was only required to keep supporting medical records for five years.

       
       

   21. In fact, rather than deleting claims, AHCA left the older claims in the sample claims being audited but did not look

       for overpayments among those claims. As a result, none of the claims of 7 of the 34 patients in the cluster sample were reviewed for overpayments; nor were some of the claims of the remaining 27 patients. The result was a finding of no overpayments among these older claims.

   22. Since the total number of clusters and claims audited were assumed to remain the same (34 and 741, respectively), AHCA used the same statistical formula to determine the point estimate of total overpayment and total overpayment with a 95% confidence level. As a result of the Respondent's Motion to Amend Final Agency Audit Report, the alleged point estimate of total overpayment was reduced to $85,160.91, and the total overpayment with a 95% confidence level was reduced to $59,859.72. The fine was reduced to $100.

   23. The Petitioner did not oppose AHCA's Motion to Amend Final Agency Audit Report, and it was granted. Nonetheless, AHCA proceeded with a second KePro review of the claims that for services rendered after September 4, 1991. The second KePro consultant disagreed with the first KePro consultant in several respects, and the review by the second KePro consultant would have resulted in an increase in the overpayment claim to

       $71,026.63. But the evidence was that, based on agency policy not to increase a claim for overpayment as a result of a second consultant review, AHCA continued to claim just $59,859.72, plus a fine of $100.

   24. Although AHCA never based a claim on the review by the second KePro consultant, the Petitioner deposed the second KePro consultant and retained her own consulting psychiatrist to review the work of the second KePro consultant. AHCA's deposition of the Petitioner's expert revealed his harsh criticism of the work of the second KePro reviewer for numerous alleged errors and inconsistencies.

   25. Careful comparison of the testimony of the Petitioner's expert to the other evidence in the case raises a question as to whom the Petitioner's expert was in fact criticizing. Pages 74 through 87 of Petitioner's Exhibit 5 (which is the testimony of the Petitioner's expert) reveal the difficulty in answering that question. As a specific example, lines 15-20 of page 76 assert that the second KePro consultant chose 99231 as the most appropriate CPT code for the services rendered to Patient 8 on May 14, 1993, whereas neither Respondent's Exhibits 13 nor 14

       --both of which reflect the worksheet, or "tool," of the second KePro consultant--indicate that code 99231 was selected by the consultant.

       The Desk Audit

       
       

   26. After the depositions of both the second KePro reviewer and the Petitioner's expert, the AHCA auditor prepared a desk audit on the advice of AHCA's legal department.

   27. The desk audit did not take into account the independent review by either KePro consultant. Even so, AHCA

       assumed that the desk audit would be adequate, because the grounds for disallowance of charges were limited to one or more of the following three reasons: (1) no time interval was documented in the medical records; (2) no medical records were found; or (3) the Petitioner did not personally render the service. As will be explained, AHCA's assumption was only partially correct.

   28. When completed on July 21, 1997, AHCA called the desk audit an Adjustment to Amended Final Agency Audit Report. From among the claims for services rendered before September 4, 1991, a total of $5380 of claims was rejected. AHCA again used the same statistical formula to determine the point estimate of total overpayment and total overpayment with a 95% confidence level, since the total number of clusters and claims audited were assumed to remain the same (34 and 741, respectively). As a result of the Adjustment to Amended Final Agency Audit Report, AHCA reduced the alleged point estimate of total overpayment to

       $84,751.34 and reduced the alleged total overpayment with a 95% confidence level to $58,341.26.

   29. AHCA chose to proceed on the basis of the desk audit alone and, over objection, was permitted to do so. At final hearing, AHCA attempted to exclude from evidence any reference to either of the KePro reviews on the ground that they allegedly were irrelevant to the desk audit. But those objections were overruled; as will be seen, the first two audits were not

       irrelevant to the desk audit. But apparently because of the positions their legal strategies required them to take, the parties' evidentiary presentations and proposed recommended orders did not clearly and comprehensively address their positions as to each claim in the sample. As a result, a time- consuming and painstakingly careful review of all of the evidence was required in order to determine whether there were overpayments on the 741 individual claims included in the desk audit.

       No Records Found

       
       

   30. It was possible for the desk auditor to determine from her own review of the audited claims, without the assistance of a consultant, when there was a complete absence of medical records to support the Petitioner's billings. Claims for reimbursement could simply be disallowed in those instances. The desk audit rejected a total of $1,858.50 of claims due to missing records.

   31. Although the Petitioner's witnesses criticized the KePro reviewers for overlooking medical records, the Petitioner did not attack the desk audit on the same ground. Instead, the Petitioner's defense to this ground was that the missing records were inpatient and PHP patient records which she was relying on Manatee Memorial Hospital to maintain.

   32. The evidence suggested that the Petitioner is not alone in relying on hospitals to maintain medical charts for inpatients. However, the Petitioner did not prove that AHCA has

       a policy of excusing doctors from the requirement of documenting claims for Medicaid reimbursement when a hospital fails to maintain those records.

   33. In some cases, there were gaps of missing medical records on Manatee Memorial inpatient and PHP patients. The Petitioner's expert consultant opined that, in those cases, the Petitioner should be given "the benefit of the doubt" on the theory that the Petitioner must have seen her patients a minimum number of times during the time periods represented by the gaps in the records.

   34. Although the Petitioner's defense conceded the missing records alleged in the desk audit, careful review of the evidence revealed a few instances in which the desk auditor disallowed claims on the ground of missing records, when in fact records appeared to have been presented to AHCA for review.

   35. The desk auditor disallowed the following bills on this ground:

       Patient 5 (E.S.), 4/1/93, CPT Code 90853 for $28 Patient 6 (J.G.), 8/31/92, CPT Code 90843 for $35 Patient 10 (D.H.), 2/12/93, CPT Code 90853 for $28

       
       

       Although not apparent from a review of the medical records included in Respondent's Exhibit 15, the second KePro consultant indicated that he reviewed medical records on all of those claims.

   36. As to the claim on Patient 6, the medical record for the 8/31/92 bill was dated 8/28/92. In many other instances,

       AHCA gave the Petitioner credit when there was such a minor discrepancy on the billing date, as the second KePro consultant proposed to do in this instance. The consultant stated that the record did not justify reimbursement under code 90843 but did justify $17.50 of reimbursement under code 90862.

   37. Deducting the $73.50 represented by the foregoing discrepancies from the total of $1,865.50 of claims rejected in

       
       

       the desk audit, the actual overpayment among the audited claims due to missing records was $1792.

       Services Not Rendered By Petitioner

       
       

   38. It also was possible for the desk auditor to determine from her own review, without the assistance of a consultant, when services were not performed by the Petitioner. In the desk audit, claims for reimbursement were disallowed when the medical chart was signed by Dr. Goldman, who was covering for the Petitioner. The desk audit rejected a total of $642 of claims for services not rendered by the Petitioner.

   39. The Medicaid Physician Provider Handbook consistently provided through the audit period that: "All aspects of [psychiatric] services must be rendered personally by the psychiatrist." (emphasis in bold in the original) The Petitioner's expert agreed that the Petitioner should not have been billing for services rendered by Dr. Goldman.

   40. The Petitioner's witnesses criticized the KePro

       reviewers for not giving the Petitioner credit when chart notations were in her handwriting, but she omitted to sign the chart. But it did not appear from a careful review of the evidence that the desk audit failed to give the Petitioner credit in any such case.

   41. On the other hand, careful review of the evidence did reveal two instances when claims disallowed by the desk auditor should have been allowed because service in fact was rendered by the Petitioner.

       
       

   42. The desk auditor rejected a claim for $32 for a code 90260 on Patient 30 (M.H.) on 12/14/91 because the service was rendered by Dr. Goldman. But the Petitioner's current billing clerk pointed out that, in rejecting the claim, the desk auditor overlooked a medical record signed by the Petitioner on 12/15/91 which supported the claim. The Petitioner did not bill separately for the service rendered on 12/15/91. As already indicated in Finding 36, in many other instances, AHCA gave the Petitioner credit when there was such a minor discrepancy on the billing date. It is found that this claim should not have been disallowed on this ground.

   43. The desk auditor also rejected a claim for $35 for a code 90843 on Patient 3 (J.J.) on 1/11/93 because it was not apparent from a review of the medical records included in Respondent's Exhibit 15 that any service was rendered by Dr.

       Goldman. However, the Petitioner's current billing clerk testified that there was a medications record omitted from Respondent's Exhibit 15 which supported a claim for $17.50 for a code 90862. It is found that AHCA did not prove that a claim for

       $17.50 for a code 90862 should not be allowed.

       
       

   44. Deducting the $49.50 represented by the foregoing discrepancies from the total of $642 of claims rejected in the desk audit, the actual overpayment among the audited claims due to billings for services not rendered by the Petitioner was

       $592.50.

       
       

       No Time Interval Documented

       
       

       1. Requirement to Document Time

          
          

   45. Section 10.9 of the June 1991 update to the Medicaid Physician Provider Handbook on record-keeping and medical records provided in pertinent part: "If time is a part of the procedure code description being billed, then duration of visit shown by begin and end time must be included in the record." Section

       11.10 on Psychiatric Services also provided in pertinent part: "All procedures with time descriptions apply to the minimum time in rendering the service. Required documentation in the record to show time interval is necessary."

   46. According to the evidence, the August 1992 update to the Medicaid Physician Provider Handbook included an additional section entitled Time, which provided:

       The inclusion of time in the old visit codes has been implicit in prior editions of CPT. Beginning in 1992, the inclusion of time as an explicit factor is done to assist physicians in selecting the most appropriate codes to report their services. However, the times indicated in each specific E/M code are average amounts of time a physician may spend with a patient. Thus, the actual content of the service should be used in determining the most appropriate code except in cases where the counseling and/or coordination of care dominates the patient encounter (more than 50%). The extent of counseling and/or coordination of care should be documented in the patient's records.

       Similar to the 1991 Handbook, Section 11.10 of the 1992 version on Psychiatric Services provided in pertinent part: "All procedures with time descriptions apply to the minimum time in rendering the service and required documentation in the record to show time interval is necessary." But, according to the

       
       

       evidence, Section 10.9 from the 1991 Handbook was omitted from the 1992 version.

   47. According to the evidence, the section entitled Time from the 1992 Handbook was deleted from the 1993 version (updated in July, 1993), and Section 10.9 from the 1991 Handbook was added back. Section 11.10 from the 1992 Handbook was renumbered 11.12, but the language remained unchanged.

   48. Notwithstanding the unclear language in the 1992 version of the Handbook, it appears that the intent of the

       applicable Handbook during the entire audit period was to require documentation of the time interval to support a claim for reimbursement under CPT codes having time as part of the procedure's description.

       b. Codes 90844 and 90843.

       
       

   49. Under the applicable Medicaid fee schedules, code 90844 was reimbursed $35 from 1/1/91 to 9/30/91, $75 from 10/1/91 to 12/31/91, and $52.50 from 1/1/92 to 12/31/93; code 90843 was reimbursed $20 from 1/1/91 to 9/30/91, $50 from 10/1/91 to 12/31/91, and $35 from 1/1/92 to 12/31/93.

   50. According to the CPT, code 90843 specified approximately 20 to 30 minutes of individual psychotherapy, while code 90844 specified approximately 45 to 50 minutes of individual psychotherapy.

   51. The evidence proved that time was part of the description of the 90843 and 90844 procedures and that it was necessary to document a time interval in order to be reimbursed

       
       

       under those codes. There was no evidence that time was part of the description any other pertinent CPT code.

   52. As previously found, the Petitioner stipulated that none of her medical records documented a time interval. Notwithstanding this stipulation, it was noticed in careful review of the evidence that the Petitioner actually was given credit for having documented a code 90843 time interval on at

       least one occasion--Patient 6 (J.G.) on 2/12/92. The same careful review of the evidence confirmed that the Petitioner never documented time intervals on a claim for reimbursement under 90844 or 90843 which AHCA disallowed because no time interval was documented.

       1. The KePro Consultants Allowed Codes 90844 and 90843

          
          

   53. As previously mentioned, AHCA did not disclose the identity of the first KePro consultant and opposed any evidence as to either KePro review on grounds of relevance. But the relevance objections were overruled, and the evidence was that both KePro consultants often allowed reimbursement under code 90843 notwithstanding the requirement that time intervals be documented.

   54. The evidence was clear that, despite the absence of documentation of time intervals, the second KePro consultant frequently approved 90843 codes billed by the Petitioner, sometimes downcoded from 90844 to 90843, and sometimes changed (usually lower-paying, but sometimes higher-paying) codes not requiring documentation of time intervals to 90843 codes.

       
       

   55. The Petitioner did not elicit testimony from the first KePro consultant and did not introduce the first KePro review into evidence, but the Petitioner's billing clerk testified on the results of the first KePro review. According to her testimony, the first KePro consultant sometimes upcoded from a

       lower-paying code 90260, which did not require documentation of a time interval, to a code 90843 notwithstanding the absence of the required documentation of time intervals.

       1. AHCA Audit Policies on Use of Consultants

          
          

   56. Prior to the desk audit, AHCA intended to defer to its consultants. The evidence was that it was AHCA's policy to defer to the judgments of consultants on whom it intended to rely in rejecting bills included in an audit.

   57. The evidence was that, in part, AHCA's deference to its consultants followed an unwritten AHCA policy of being lenient to providers by not further downcoding after a consultant already had downcoded once from a higher-paying code (usually code 90844) to a code 90843. There also was evidence that AHCA was adhering to a litigation strategy of supporting all of the judgments of consultants upon whom it intended to rely in rejecting bills, even if those judgments were not entirely correct.

   58. In this case, AHCA initially intended to rely on the first KePro consultant in rejecting bills included in an audit; AHCA did not initially plan on two KePro consultant reviews. When it initiated the second KePro review, AHCA intended to rely on it. But with the second KePro review came unanticipated problems.

   59. The first problem arose when the second KePro review resulted in a higher overpayment. Because of AHCA's decision not to increase the assessment as a result of the second review, it

       found itself in the position of trying to use the second KePro review to substantiate the results reached in the first review.

   60. The second problem arose when not only did both two consultants sometimes recommend payment of 90843 codes despite the absence of the required documentation of time intervals, the two KePro consultants did not always agree on the proper codes to use.

   61. Faced with these problems, AHCA decided not to rely on either KePro consultant in rejecting bills included in the audit. Instead, upon the advice of its legal department, AHCA chose to conduct a third audit--a limited desk audit that did not rely on any consultant.

   62. It is found that AHCA's normal policy of deferring to its consultants did not control the desk audit to the extent of requiring AHCA to continue to use impermissible 90844 and 90843 codes. In correcting those errors, AHCA was not changing its policies; rather, AHCA was adapting them to fit the peculiar position in which it found itself in this case. It is found that, under these circumstances, it was appropriate for AHCA to correct the erroneous use of the impermissible 90844 and 90843 codes by the KePro consultants.

   63. It also is found that AHCA still should have followed its policies of choosing the most appropriate CPT codes in downcoding from codes 90844 and 90843 and of utilizing consultants to determine the most appropriate codes. While the

       consultants had the expertise to make those determinations, the desk auditor admittedly did not.

       1. Desk Auditor's Inappropriate Downcoding

          
          

   64. In the desk audit, whenever the Petitioner billed a code 90843 or 90844, the desk auditor automatically changed the codes to either a code 90841 (if the service was rendered before December 1, 1991) or a code 90862 (if the service was rendered on or after December 1, 1991) and assessed the resulting overpayment.

   65. The desk auditor's use of code 90841 is not disputed. Code 90841 was another individual medical psychotherapy code, but it did not specify time. The desk auditor only used it in place of code 90843 during the short time code 90841 was in effect-- between October 1 and November 30, 1991. There was no evidence that code 90841 was inappropriate in of those instances.

       Besides, code 90841 was reimbursed $50 during the time it was in effect, the same reimbursement allowed for code 90843 during that time.

   66. To the contrary, the desk auditor's automatic use of code 90862 in all other cases was not appropriate because it ignored the judgments of the KePro consultants and was contrary to the evidence.

       1. Direct Evidence of Appropriate Downcoding

          
          

   67. It is found from the testimony of the Petitioner's current billing clerk that the first KePro consultant often chose

       
       

       code 99231 or code 99232 as the most appropriate alternative to a 90844 or 90843.

   68. Hospital code 99231 required two of the following three key components: a problem-focused interval history; a problem- focused examination; and a medical decision that was straightforward or of low complexity. According to the applicable CPT, code 99231 included counseling and/or coordination with other providers (such as nurses), usually where the patient was stable, recovering or improving. It paid $29 starting 10/1/91 through 12/31/93.

   69. Hospital code 99232 required two of the following three key components: an expanded problem-focused interval history; an expanded problem-focused examination; and a medical decision that was of moderate complexity. According to the applicable CPT, Code 99232 was reserved for cases where the patient was not responding to therapy adequately or had developed a minor complication. It paid $31 starting 10/1/91 through 12/31/93.

   70. The CPT advised that code 99231 typically required the psychiatrist to spend approximately 15 minutes on the floor and that code 99232 typically required the psychiatrist to spend approximately 25 minutes at bedside and on the floor. However, time is not considered to be part of the description of either code 99231 or 99232.

   71. According to the first KePro consultant, the following are the procedures that AHCA should have downcoded from 90844 to 99231 (instead of to 90862):

       
       

       Patient 33 (R.G.) 03/03/93

       03/08/93

       03/12/93

       03/17/93

       
       

       Each of those downcodes generate overpayment in the amount of

       
       

       $23.50 (the difference between the $52.50 claimed for code 90844 and the $29 allowed for code 99231), for a total of $94.

   72. According to the first KePro consultant, the following are the procedures that AHCA should have downcoded from 90843 to

       99231 (instead	of 90862):
       Patient 3	(J.J)	12/30/92
       		01/18/93
       		01/20/93
       		01/27/93
       Patient 8	(E.B.)	05/17/93
       		05/19/93
       		05/31/93
       		06/02/93
       		06/04/93
       		07/05/93
       		08/13/93
       		08/20/93
       Patient 15	(B.D.)	06/26/93
       		06/27/93
       		06/28/93
       		07/05/93
       Patient 24	(J.D.)	03/10/92
       Patient 28	(T.H.)	12/21/92

       Patient 33 (R.G.) 03/23/93

       03/25/93

       03/27/93

       
       

       Each of those downcodes generated overpayment in the amount of $6 (the difference between the $35 claimed for code 90843 and the

       $29 allowed for code 99231), for a total of $126.

       
       

   73. According to the first KePro consultant, the following are the procedures that AHCA should have downcoded from 90843 to 99232 (instead of to 90862):

       Patient	8 (E.B.)		08/09/93
       			06/16/93
       			06/23/93
       			06/30/93
       			12/20/93
       Patient	13	(M.B.)	09/30/92
       Patient	15	(B.D.)	06/29/93
       			07/06/93
       Patient	29	(B.E.)	11/10/92
       			11/18/92
       Patient	33	(R.G.)	03/11/93
       			03/18/93

       
       

       Each of those downcodes generated overpayment in the amount of $4 (the difference between the $35 claimed for code 90843 and the

       $31 allowed for code 99232), for a total of $48.

       
       

   74. In most cases, it was impossible to tell from the testimony of the Petitioner's expert what code he would have assigned to particular services instead of 90844 or 90843. In

       most cases, he testified in terms of numbers of services for each code, not in terms of particular dates of service. For the same reasons, it cannot be found that he disagreed with the first KePro consultant as to the use of codes 99231 and 99232 instead of codes 90844 and 90843.

   75. The Petitioner's current billing clerk clearly agreed with the downcoding recommended by the first KePro consultant.

   76. As previously found, the second KePro consultant frequently approved 90843 codes and changed 90844 codes to 90843, and it cannot be ascertained from the evidence what alternative codes the second KePro consultant might have been inclined to choose. But in some cases, in particular in regard to Patient 3 (J.J.), the second KePro consultant agreed with 99231 codes assigned by the first KePro consultant. In addition, the second KePro consultant recommended downcoding from 90843 to 99231 (instead of 90862) on five occasions not mentioned in the testimony of the Petitioner's billing clerk on the recommendations of the first KePro consultant in regard to Patient 3--January 4, 13, 15, and 22 and February 1, 1993. Each of those downcodes would add another $6 of overpayment to the findings of the first KePro consultant, for a total of $30.

   77. It is noted that in Respondent's Exhibit 14, the second KePro consultant's report (or "tool"), conflicted with his deposition testimony in regard to Patient 3 (J.J.) on January 22, 1993. He testified that it should have been a code 90853, which

       is a group psychotherapy code that was reimbursed $28 in 1992 and 1993. But the deposition testimony, which would substantiate a

       $7 overpayment in that instance, is rejected in favor of $6 overpayment reflected in Respondent's Exhibit 14.

   78. In several cases, it appeared that the second KePro consultant disagreed with the first KePro consultant's recommendation to downcode from 90843 to 99232 (instead of to 90862). The second KePro consultant thought that no code was appropriate for the bills on the following patients because he did not see an indication of individual therapy with the patient in the medical records: Patient 8 (E.B.) on June 16, 23, and 30, 1993; Patient 15 (B.D.) on June 29, 1993; Patient 29 (B.E.) on November 10, 1992; and Patient 33 (R.G.) on March 11, 1993. In those cases, the coding suggested by the first KePro consultant is accepted as supported by the greater weight of the evidence. (It does not appear that absence of individual therapy precluded the use of code 99232).

       1. Indirect Evidence of Appropriate Downcoding

       
       

   79. Due to the manner in which this case was presented, the foregoing findings based on direct evidence do not account for many instances in which the AHCA desk auditor automatically downcoded from 90844 or 90843 to 90862. However, in many of those instances, the patient appears to have been hospitalized or in the PHP program, and it is possible that a higher-paying hospital code could have been proper. Without expert evidence to

       the contrary, AHCA did not prove that the Petitioner was not entitled to the highest-paying code available in those instances.

   80. In the following instances, $31 for a code 99232 was the most reimbursement to which the Petitioner could have been entitled instead of the $35 for the code 90843 she billed, for an overpayment of $4 each:

       Patient 3 (J.J.) -- 01/21/93 $4

       
       

       Patient 8 (E.B.) -- 05/07/92

       05/12/92

       05/14/92

       05/21/92

       07/02/92

       08/11/92

       08/16/92

       08/18/92

       12/17/92

       12/22/92 $40

       
       

       Patient 13 (M.B.) -- 09/08/92

       09/09/92

       09/11/92

       09/15/92

       09/16/92

       09/22/92

       09/23/92

       09/25/92

       09/29/92 -$76

       10/02/92

       10/07/92

       10/13/92

       10/14/92

       10/16/92

       10/20/92

       10/21/92

       10/23/92

       10/28/92

       10/30/92

       Patient	15	(B.D.)	--	06/30/93
       				07/01/93
       				07/02/93
       				07/03/93
       				07/04/93	-$20
       Patient	28	(T.H.)	--	12/23/92
       				12/28/92
       				01/04/93	-$12
       Patient	29	(B.E.)	--	11/06/92
       				11/11/92
       				11/25/92
       				12/09/92
       				12/16/92	-$20
       Patient	31	(B.L.)	--	02/25/92	-$4
       Patient	33	(R.G.)	--	03/01/93
       				03/02/93
       				03/04/93
       				03/05/93
       				03/07/93
       				03/09/93
       				03/13/93	-$28

       
       

       The total of these overpayments is $204.

       
       

   81. It was noticed that the second KePro consultant did not think any code was appropriate for Patient 3 (J.J.) for

       January 21, 1993, because the medical records did not indicate individual therapy with the patient. But it did not appear from the evidence that the absence of individual therapy precluded the use of code 99232.

   82. In the following instances, $31 for a code 99232 was the most reimbursement to which the Petitioner could have been entitled instead of the $50 for the code 90844 she billed, for an overpayment of $21.50 each:

       
       

       Patient 33 (R.G.) --	03/10/93
       	03/15/93
       	03/19/93
       	03/22/93
       	03/24/93
       	03/26/93
       	03/29/93

       -$150.50

       
       

       The total of these overpayments is $150.50.

       
       

   83. Based on the evidence, the highest-paying hospital code available to the Petitioner before hospital codes 99231 and 99232 went into effect was code 90260. Code 90260 was for intermediate hospital services. It was in effect through February 29, 1992. Until December 31, 1991, it was reimbursed $32; for the first two months of 1992, it was reimbursed $22.

   84. In the following instances, $32 for a code 90260 was the most reimbursement to which the Petitioner could have been entitled instead of the $50 for the code 90843 she billed, for an overpayment of $18 each:

       Patient 6 (J.G.) -- 12/02/91

       12/04/91

       12/06/91

       12/09/91

       12/11/91

       12/13/91

       12/16/91

       12/18/91

       12/20/91

       12/23/91

       12/27/91

       12/30/91 -$216

       
       

       Patient 30 (M.H.) -- 12/13/91

       12/16/91

       12/18/91

       12/20/91

       12/23/91

       12/27/91

       12/30/91 -$126

       The total of these overpayments is $342.

       
       

       Patient	6 (J.G.) --	01/03/92
       		01/06/92
       		01/08/92
       		01/10/92
       		01/13/92
       		01/15/92
       		01/17/92
       Patient	30 (M.H.) --	01/03/92
       		01/06/92
       		01/08/92
       		01/10/92
       		01/13/92
       		01/15/92
       		01/17/92
       		01/20/92

   85. In the following instances, $22 for a code 90260 was the most reimbursement to which the Petitioner could have been entitled instead of the $35 for the code 90843 she billed, for an overpayment of $13 each:

       
       

       -$91

       
       

       -$104

       
       

       The total of these overpayments is $195.

       
       

   86. It can be inferred from the evidence that, except as set out above, the 90862 codes used by the AHCA auditor were proper. They occurred at times when the patient was not hospitalized and when, based on the evidence, none of the hospital codes were appropriate. The following overpayments result from downcoding from 90843 to 90862 (a difference in reimbursement of $17.50 each):

       
       

       Patient 3 (J.J.) -- 02/10/93

       02/17/93

       02/24/93 -$52.50

       Patient	6 (J.G.) --	05/21/92
       		07/30/92
       		09/17/92
       		10/22/92
       		10/30/92	-$87.50
       Patient	13 (M.B.) --	08/11/92
       		08/14/92	-$35.00

       
       

       The total of these overpayments is $175.

       
       

   87. It was noticed that the second KePro consultant did not think any codes were appropriate for Patient 3 (J.J.) for February 10, 17, and 24, 1993, because the medical records did not indicate individual therapy with the patient. But it was clear from the evidence that the absence of individual therapy did not preclude the use of code 90862.

       Summary of Findings as to Audited Claims

       
       

   88. Altogether, there were $3749 in overpayments among the 741 claims in the sample, or $5.059379217274 per claim. See Findings 37, 44, 71-73, 76, 80, 82, and 84-86. These overpayments are distributed among the patients making up the sample as follows:

       Patient	1	$ 56.00
       Patient	3	219.00
       Patient	4	283.00
       Patient	5	168.00
       Patient	6	530.00
       Patient	7	31.00
       Patient	8	291.50
       Patient	9	112.00
       Patient	10	28.00
       Patient	11	0 00
       Patient	13	290.00
       Patient	14	25.50
       Patient	15	276.00
       Patient	18	84.00
       Patient	21	87.00

       Patient	22	0.00
       Patient	23	0 00
       Patient	24	122.50
       Patient	25	84.00
       Patient	26	0.00
       Patient	27	156.00
       Patient	28	88.00
       Patient	29	98.00
       Patient	30	294.00
       Patient	31	67.00
       Patient	32	0.00
       Patient	33	358.50

       
       

   89. Multiplying the mean overpayments per claim by the 11,673 claims in the sample, the point estimate overpayment is

       $58,992.92 (as compared to the $84,751.34 reflected in the Adjustment to Amended Final Agency Audit Report that resulted from the desk audit.) Using the same statistical formula as before, and the list of overpayments per patient set out in Finding 88, AHCA can convert this lower point estimate overpayment to a lower total overpayment amount with a 95% confidence level.

       Continued Validity of Statistical Sample and Formula

       
       

   90. The Petitioner took the position that deletion of claims for services rendered from January 1 through September 4, 1991, invalidated the statistical sample and formula used by AHCA. As a result of the deletion of these claims, seven of the original 34 sample clusters were not examined; nor were some of the claims in the remaining 27 cluster samples.

   91. Notwithstanding the deletion of these claims, the number of clusters and claims remained the same for purposes of the statistical sample and formula used by AHCA. The actual

       effect of deleting these claims was to presume that they did not include any overpayments, a presumption to the benefit of the Petitioner. This did not invalidate the formula or the "t" value used in the formula.

   92. The decision not to examine claims for services rendered from January 1 through September 4, 1991, would only have invalidated the statistical sample and formula used by AHCA if examination of them would have resulted in underpayments, instead of "zero overpayment."

   93. Given the limited scope of the desk audit, examination of the deleted claims clearly would not have resulted in underpayments. Clearly, no underpayment could result from the disallowance of claims because no supporting documentation could be found or because Dr. Goldman, not the Petitioner, performed the service. Likewise, there was no evidence that there were appropriate alternatives to codes 90844 and 90843 that reimbursed more than those codes.

   94. The Petitioner also claimed that there were other bills, besides those for services rendered between January 1 and September 4, 1991, that were omitted from AHCA's audit. If so, the failure to examine those bills could call into question the validity of the statistical sample and formula used by AHCA in the audit.

   95. The Petitioner's current billing clerk testified, based on a review of the Petitioner's billing records, that some

       services that appeared from those records to have been billed were not included in the audit. Apparently, some of those bills were sent to AHCA's fiscal agent, but it was not proven that any of them were paid. (To the contrary, the Petitioner and her current billing clerk testified that the Petitioner was not paid for approximately $22,000 of bills that allegedly were sent to AHCA's fiscal agent during the audit period.) Apparently, other services were not billed at all. (The Petitioner's expert valued unbilled services at approximately $1633.)

   96. The Petitioner's evidence did not overcome AHCA's more convincing evidence that the 741 bills that made up the sample were all of the paid bills on all 34 patients in the cluster sample.

   97. AHCA only intended to audit paid bills; it never intended to audit unpaid bills, much less services never billed. Since the audit covered all paid bills on all 34 patients in the cluster sample, the statistical sample and formula used by AHCA in the audit was valid notwithstanding the existence of unpaid bills or services never billed that were not included in the audit.

       Petitioner's Claim of Offsets

       
       

   98. The Petitioner also took the position that, in conducting the audit and assessing overpayments, AHCA did not give her proper credit for offsets outside the scope of the desk audit itself. If true, failure to give her proper credit would

       invalidate the audit of the claims in the sample. In addition, examination of the deleted claims could have resulted in underpayments that could invalidate the statistical sample and formula used by AHCA in the audit.

       
       

   99. One offset claimed by the Petitioner was for the value of services she either never billed or billed but was never paid. See Finding 95.

   100. Another offset claimed by the Petitioner was for her use of codes (other than 90844 and 90843) when higher-paying codes allegedly were more appropriate. (The Petitioner also alleged that some codes she used should have been upcoded to 90843, as the two KePro consultants sometimes did; but as previously found, it was inappropriate to use either 90844 or 90843 since the Petitioner never documented any time intervals.)

   101. The evidence was clear that, until the final hearing in this case, the Petitioner never sought an adjustment from AHCA for these alleged offset claims.

   102. The 1991 Medicaid Physician Provider Handbook had a Policy for Adjustment Requests stating: "You must submit an adjustment request (when additional money is due to the provider due to provider filing error) within one year of the original claim's payment date." It is not clear from the evidence whether the 1992 and 1993 versions of the Handbook included this policy statement.

        CONCLUSIONS OF LAW

        
        

   103. Section 409.913, Florida Statutes (Supp. 1996), describes the program AHCA is supposed to operate in order to oversee the activities of Florida Medicaid recipients and providers and, among other things, recover overpayments. In pertinent part, it provides:

        
        

        (1) For the purposes of this section, the term:

        
        

        * * *

        
        

        (d) "Overpayment" includes any amount that is not authorized to be paid by the Medicaid program whether paid as a result of inaccurate or improper cost reporting, improper claiming, unacceptable practices, fraud, abuse, or mistake.

        (10) The agency may require repayment for inappropriate, medically unnecessary, or excessive goods or services from the person furnishing them, the person under whose supervision they were furnished, or the person causing them to be furnished.

        * * *

        
        

        1. In making a determination of overpayment to a provider, the agency must use accepted and valid auditing, accounting, analytical, statistical, or peer- review methods, or combinations thereof. Appropriate statistical methods may include, but are not limited to, sampling and extension to the population, parametric and nonparametric statistics, tests of hypotheses, and other generally accepted statistical methods. Appropriate analytical methods may include, but are not limited to, reviews to determine variances between the quantities of products that a provider had on hand and available to be purveyed to Medicaid recipients during the review period and the quantities of the same products paid for by the Medicaid program for the same period, taking into appropriate consideration sales of the same products to non- Medicaid customers during the same period. In meeting its burden of proof in any administrative or court proceeding, the agency may introduce the results of

           such statistical methods as evidence of overpayment.

        2. When making a determination that an overpayment has occurred, the agency shall prepare and issue an audit report to the provider showing the calculation of overpayments.

        3. The audit report, supported by agency work papers, showing an overpayment to a provider constitutes evidence of the overpayment.

        
        

   104. It has been held, and has been stipulated by the parties, that AHCA has the burden of proof in this case. See Southpoint Pharmacy vs. Dept. of Health, etc., 596 So. 2d 106,

        108 (Fla. 1st DCA 1992). It is clear from the statute that AHCA can meet its burden of proof through the appropriate use of statistics.

        Appropriateness of Statistical Sample and Formula

        
        

   105. As found, it was appropriate for AHCA to use cluster sampling and extrapolation of findings from auditing the sample claims through use of the statistical formula employed. As also found, it was proven that the cluster sample was valid.

   106. The Petitioner seeks to apply the decision in the case of Ramirez vs. State, 651 So. 2d 1164 (Fla. 1995), to the statistical sampling and statistical formula used by AHCA in this case. But Ramirez involved the admissibility of expert opinions concerning new or novel scientific principles. To the contrary, the evidence in this case was that the statistical sampling and statistical formula used by AHCA was standard.

        Claim of Offset Untimely

        
        

   107. The Petitioner's criticism of the statistical sampling and statistical formula used by AHCA in this case was based in

        part on the Petitioner's claim that she is entitled to credit for underpayments that should be offset against overpayments.

        However, Florida Administrative Code Rule 10C-7.030(6)(a) required that all Medicaid claims, including resubmittal of clean claims, be submitted within 12 months of the date of service in order to be considered for payment. Florida Administrative Code Rule 10C-7.030(6)(b) required that all adjusted claims be submitted within 12 months of the original payment date in order to be considered for payment. In addition, as found, the Medicaid Physician Provider Handbook had a policy requiring adjustment requests (when additional money is due to the provider due to provider filing error) to be made within one year of the original claim's payment date. The Petitioner agreed to abide by this requirement; yet, she did not ask for adjustments at any time before the final hearing. It is concluded that the Petitioner is barred from asserting offset claims at this time.

        Cf. Bank of South Palm Beaches vs. Stockton, 473 So. 2d 1358, 1361 (Fla. 4th DCA 1985)(even an equity court cannot "disregard established law in order to reach a fair result").

        Requirement to Keep Records

        
        

   108. As found, the Petitioner signed a Medicaid Provider Agreement "to keep for 5 years complete and accurate medical and fiscal records that fully justify and disclosed the extent of the services rendered and billings made under the Medicaid program" and agreed to furnish them to AHCA upon request. She also agreed

        to "abide by the Florida Administrative Code, Florida Statutes, policies, procedures, manuals of the Florida Medicaid Program and Federal laws and regulations."

   109. Section 409.913(7)(f), Florida Statutes (Supp. 1996), requires Medicaid services to be "fully and properly documented." Florida Administrative Code Rule 59G-9.030(5)(a) provides: "If such documentation is not furnished by the provider, it will be concluded that the provider was not and is not entitled to payments for any products or services comprehended by the unfurnished documentation." It was clear from the evidence that AHCA properly disallowed claims when no medical records could be found to document the services claimed to have been rendered.

   110. The Petitioner claimed, through the testimony of her expert, that the Petitioner should have been given "the benefit of the doubt" when there were large gaps in the records. There is no basis in law for doing this in the face of the clear statute and rule provisions. Cf. Bank of South Palm Beaches vs. Stockton, supra.

   111. The Petitioner also claimed that she should be excused from failing to support claims with proper documentation on the ground that Manatee Memorial was responsible for maintaining records that would document the claims. It may be true, as argued by the Petitioner, that Manatee Memorial had its own, independent duty to maintain patient records for inpatients and PHP patients who were seen by the Petitioner. But the

        Petitioner's obligation to furnish documentation sufficient to support her Medicaid claims was hers alone and grounded not only in statutes and rules but also in her contractual agreement with AHCA. The Petitioner cannot avoid her obligations by blaming Manatee Memorial. See Atchley vs. First Union Bank of Florida,

        576 So. 2d 340, 344 (Fla. 5th DCA 1991)("liability for misfeasance cannot be avoided by the person who obligated himself originally to perform the contract").

        Services Not Rendered By Petitioner

        
        

   112. It also is clear that the Petitioner cannot claim Medicaid reimbursement for services rendered by Dr. Goldman. As found, the Medicaid Physician Provider Handbook, which the Petitioner agreed to follow, consistently provided throughout the audit period that: "All aspects of [psychiatric] services must be rendered personally by the psychiatrist." (emphasis in bold in the original) In addition, Florida Administrative Code Rule 59G-4.230(4)(a)9. provided: "A physician shall bill and accept payment for psychiatric services only when those services were personally rendered by the psychiatrist, as defined in 59G- 1.010(238), Florida Administrative Code, or by a resident under the personal supervision of a psychiatrist who is a member of the medical faculty at a teaching hospital as defined in 408.07(49), Florida Statutes." Even the Petitioner's expert agreed that the Petitioner should not have been billing for services rendered by Dr. Goldman.

        Policies Affecting Requirement to Document Time Intervals

        
        

   113. The Petitioner attempted to characterize the desk audit as a change of policy to require documentation of time intervals for reimbursement under codes 90844 and 90843. To the contrary, as found, the evidence as a whole proved that AHCA always had a policy to require documentation of time intervals for reimbursement under those codes.

   114. AHCA also had policies: (1) not to further downcode after a consultant already had downcoded once from a higher- paying code; and (2) to adhere to a litigation strategy of supporting all of the judgments of consultants upon whom it intended to rely in rejecting bills, even if those judgments were not entirely correct. The Petitioner claimed that the desk audit also violated those policies.

   115. As found, under the peculiar circumstances of this case, AHCA's normal policy of deferring to its consultants did not control the desk audit to the extent of requiring AHCA to continue to use impermissible 90844 and 90843 codes. Rather, it was appropriate for AHCA to correct the erroneous use of the impermissible 90844 and 90843 codes by the KePro consultants.

        Choosing Appropriate Alternative Codes

        
        

   116. It also was found that AHCA still should have followed its policies of choosing the most appropriate CPT codes in

        downcoding from codes 90844 and 90843 and of utilizing consultants to determine the most appropriate codes. The desk audit would have violated those policies.

   117. AHCA argued that the two KePro consultant reviews were irrelevant to the desk audit. AHCA cited Daugherty vs. Latham, 139 Fla. 477, 190 So. 2d 742, 748 (1939), in support of its argument. That case held only that a "party cannot be required to sue those against who he does not wish to proceed, nor state another case than that upon which he elects to rely." It is not authority for excluding necessary evidence as to the two KePro reviews.

        Adjustments to Desk Audit

        
        

   118. As found, after correcting errors in the desk audit, there were $3749 in overpayments among the 741 claims in the sample, or $5.059379217274 per claim. Multiplying the mean overpayments per claim by the 11,673 claims in the sample, the point estimate overpayment is $58,992.92 (as compared to the

        $84,751.34 reflected in the Adjustment to Amended Final Agency Audit Report that resulted from the desk audit.)

        
        

   119. As found, the errors in the desk audit did not invalidate the statistical sample or formula used in the audit. Using the same statistical formula as before, and the list of overpayments per patient set out in Finding 88, AHCA can convert

this lower point estimate overpayment to a lower total overpayment amount with a 95% confidence level.

RECOMMENDATION

Based upon the foregoing Findings of Fact and Conclusions of Law, it is

RECOMMENDED that AHCA use the same statistical formula to convert the reduced total point estimate overpayment of

$58,992.92 to a total overpayment amount with a 95% confidence level and enter a final order requiring the Petitioner to repay the new reduced overpayment.

RECOMMENDED this 9th day of December, 1997, in Tallahassee, Leon County, Florida.

J. LAWRENCE JOHNSTON Administrative Law Judge

Division of Administrative Hearings The DeSoto Building

1230 Apalachee Parkway

Tallahassee, Florida 32399-3060

(904) 488-9675 SUNCOM 278-9675

Fax Filing (904) 921-6847

Filed with the Clerk of the Division of Administrative Hearings this 9th day of December, 1997.

COPIES FURNISHED:

John D. Buchanan, Jr., Esquire Henry, Buchanan, Hudson,

Suber & Williams, P.A.

117 South Gadsden Street Tallahassee, Florida 32302

Thomas Falkinburg, Esquire Moses E. Williams, Esquire

Agency for Health Care Administration Fort Knox Executive Center No. 1

2727 Mahan Drive

Tallahassee, Florida 32308

Sam Power, Agency Clerk

Agency for Health Care Administration 2727 Mahan Drive

Tallahassee, Florida 32308

Douglas M. Cook, Director

Agency for Health Care Administration 2727 Mahan Drive

Tallahassee, Florida 32308

Jerome W. Hoffman, General Counsel Agency for Health Care Administration 2727 Mahan Drive

Tallahassee, Florida 32308

NOTICE OF RIGHT TO SUBMIT EXCEPTIONS

All parties have the right to submit written exceptions within 15 days from the date of this Recommended Order. Any exceptions to this Recommended Order should be filed with the agency that will issue the final order in this case.

Docket for Case No: 96-004830

Orders for Case No: 96-004830