OPINION OF THE COURT
Read, J.
In this CPLR article 78 proceeding, petitioner Blossom View Nursing Home (Blossom) appeals from an Appellate Division order dismissing its petition seeking to bar respondents Commissioner of Health and Director of the Division of the Budget from auditing its patient review instruments (PRIs) for the years 1994 though 1996. The New York State Department of Health (DOH) did not announce its intention to commence audits of Blossom’s PRIs for any of these years until August 2002, and the parties dispute whether DOH may ever audit PRIs more than six years after they are filed. For the reasons that follow, we conclude that DOH is not required by its regulations to notice or commence an audit of PRIs within six years of filing; however, in light, of the facts of this case, DOH may not audit Blossom’s PRIs for 1995 and 1996, or use its audit of Blossom’s 1994 PRIs to determine Medicaid reimbursement rates.
I.
The Commissioner administers the State’s Medical Assistance Plan (Medicaid) and sets Medicaid reimbursement rates for nursing homes, more formally called “residential health care facilities,” for medical services provided to the indigent (Public Health Law § 2807 [3]; § 2808 [3]). Prior to 1986, the State’s Medicaid per diem reimbursement rate—the daily rate at which a facility can bill Medicaid for every Medicaid-eligible resident— was not in any way tied to the level of care required. As a result, nursing homes were thought to be discouraged from admitting individuals requiring more intensive care, who were as a consequence shunted into costly hospital beds. To reduce any rate-induced disincentive for nursing homes to provide services for individuals requiring institutional but not hospital care, DOH
developed and implemented the Resource Utilization Group-II (RUG-II) case mix reimbursement methodology, effective January 1, 1986.
RUG-II is a prospective system that establishes reimbursement rates by using a nursing home’s allowable costs in a base year or period,
adjusted for regional wage differentials, inflation and changes in the level of care required by its residents. RUG-II provides for reimbursement of capital and operating costs, the latter of which is composed of direct,
indirect
and noncomparable
components (10 NYCRR 86-2.10 [a] [7]; [b] [1] [ii]; [2]).
The direct component of operating costs is based upon a patient classification system that establishes 16 RUG-II categories grouped into five “hierarchies,” each of which is characterized by specifically diagnosed physical or mental conditions.
The five hierarchies are further divided to create the 16 RUG-II categories based on the resident’s functional ability to perform activities of daily living (ADL), measured by the resident’s score on an ADL index designed to reflect the need for supervision or assistance in eating, toileting and transferring
(see
10 NYCRR Appendix 13-A).
Each of the RUG-II categories represents a different combination of the two components—condition (the hierarchy) and functional ability (the ADL)—and reflects the costs associated with caring for nursing home residents classified within the particular category, expressed as a numeric value or case mix index (CMI). Generally, the higher the CMI, the more intensive and costly the care required.
The number of a nursing home’s residents classified in the various RUG-II categories determines the facility’s overall CMI and thus significantly influences its per diem Medicaid reimbursement rate. Consequently, it is essential for each resident’s condition and functional ability to be assessed accurately. This is accomplished by means of the PRI.
A qualified registered nurse completes the PRI, which summarizes the resident’s condition, including medical diagnosis and treatments, ADLs, behavior and specialized services required during the four weeks prior to the PRI’s completion. Each resident is assigned to one of the 16 RUG-II categories based on the information in the completed PRI, for which DOH has provided detailed instructions and clarification documents specifying certain “qualifiers” or criteria that must be met before any PRI question may be answered “Yes.” These qualifiers take the form of time period (the four weeks before the PRI’s completion), frequency (how often something needs to occur to meet the qualifier), the specific medical record documentation called for and exclusions (types of behavior or care to be disregarded when answering the question).
Nursing homes submit new PRIs electronically for all the residents in their care every six months, in accordance with a schedule set by DOH. Twice a year, at the halfway point between these “full-house” PRI filings, a nursing home must update its submission to DOH to account for new admissions and discharges (10 NYCRR 86-2.11 [b]). To insure the accuracy of a nursing home’s PRIs and thus the integrity of its reimbursement rates, DOH has developed a three-stage audit process. DOH has represented in this litigation that “PRI audits are performed at each facility [in the state] approximately every 18 months.”
DOH sends a list of nursing homes to be audited to an independent organization with which it has contracted to accomplish this work. For Stage I audits, DOH typically provides audit forms for 40 of a nursing home’s residents during the audit period, which the auditor completes independently to validate the ADL level assigned by the nursing home. The auditor usually completes a hierarchy verification sheet, or checklist, for 16 of the 40 residents, which addresses the qualifiers relevant to the resident’s specific classification. Auditors principally examine a resident’s medical records to complete the audit form, although they may also observe the residents whose PRIs are being audited or discuss the residents’ care with nursing home staff.
The Stage I auditor returns the completed Stage I audit forms to DOH with an explanation for any differences between PRIs after auditing and as originally submitted. DOH substitutes Stage I audited PRIs for those filed by the nursing home, and also determines whether there is a statistically significant variance (a change in RUG-II category for more than 25% of residents whose PRIs are audited or a change in CMI for the audited residents of .05 or greater), warranting Stage II review of an additional 80 residents.
A different auditor conducts the Stage II audit, which follows the same general format as the Stage I audit. The nursing home may dispute Stage I audit results with the Stage II auditor, who examines any documentation presented and either agrees with or overturns the Stage I auditor’s adverse findings, referred to as “controverted items.” The Stage II auditor may also seek to resolve any discrepancies between the audit findings and the nursing home’s PRIs through discussion with staff or observations of residents. The Stage II auditor returns the completed Stage II audit forms to DOH, which substitutes Stage II audited PRIs for those originally filed by the nursing home, and again determines whether there is a statistically significant variance, warranting a Stage III review of all remaining residents (except those in the lowest-rung RUG-II category).
Yet a third auditor conducts the Stage III audit, which follows the same general format as the Stage I and Stage II audits. The nursing home may seek to persuade the Stage III auditor to overturn controverted items from the Stage II audit. If after completion of the Stage III audit the nursing home is again found to have completed PRIs inaccurately to a statistically significant degree, DOH directs the nursing home to have its PRIs completed by an outside assessor at its own expense for at least one year (10 NYCRR 86-2.30 [f] [1] [ii]; [2]).
Finally, DOH’s procedures call for kicking off each stage of the audit with an entrance conference at which the auditor outlines the review’s purpose, scope and timetable. During the entrance conference, the auditor must secure the nursing home’s written determination as to whether an exit conference is requested. Nursing homes are entitled to an exit conference
at the conclusion of each stage of the audit so long as requested at the entrance conference. At the exit conference, the auditor conveys the provisional results of the review just conducted.
II.
On August 11, 1993, DOH’s contract auditor conducted a Stage I audit of Blossom’s July 1993 PRIs, and so examined records to determine the ADL levels of 40 residents and completed hierarchy checklists for 16 of them. The auditor placed 8 of these 40 residents in a different RUG-II classification, yielding a statistically significant decrease in the CMI for these 40 residents. Accordingly, a Stage II audit was scheduled for December 6, 1993.
The Stage II auditor examined the records pertaining to the remaining residents living at Blossom in July 1993 and, at Blossom’s request, reassessed the Stage I auditor’s findings. The Stage II auditor overturned three of the items controverted by Blossom in the Stage I audit, but agreed with the Stage I auditor as to the others. Further, the Stage II auditor concluded that PRIs for 32 additional residents were unsupported by documentation. There was an exit conference at which the auditor informed Blossom’s personnel of these provisional results.
Blossom’s administrator wrote a health care fiscal analyst at DOH on December 17, 1993 to complain about how the contract auditors had carried out the Stage I and II audits, and to dispute their findings. Although Blossom’s resident population was too small to qualify for a Stage III audit, DOH’s health care fiscal analyst nonetheless offered Blossom an additional review and thus another chance to contest Stage II audit results.
Two new outside auditors undertook this third review on March 14, 1994. The auditors examined the records of 24 patients involving 27 controverted items from the Stage II audit and agreed with the Stage II auditor on all of them, although Blossom did not know this at the time. More significantly, the auditors prepared a report “detailing] what occurred” during their review. This report conveys the auditors’ distinct impression that Blossom’s staff was at least uncooperative if not downright intimidating and hostile. Upon completing their work, the auditors told Blossom’s administrator that they were finished, but needed time to compile their findings for the exit conference. When the administrator objected that the auditors “were not to leave the building until an exit interview was granted,” the auditors “explained again that [they] were not
prepared to do this yet and [Blossom] would hear from the [auditors’ home] office.”
On March 30, 1994, Blossom’s attorney wrote to the health care fiscal analyst at DOH to request the exit conference. His version of the “confusion as to what exactly happened on March 14,” demonstrates yet again the wisdom of
Rashomon:
perspective distorts reality and makes the absolute truth unknowable. According to the attorney, during the morning session of the review there was “[a] great deal of constructive dialogue” between the auditors and Blossom’s staff. In the afternoon, however, the auditors’ attitude changed and they “became difficult to work with.” This turn of events “dismayed” Blossom’s staff “since the review, until that point, had proceeded with the smooth exchange of assistance and information.” Worse yet, “the auditors left . . . without giving the nursing home staff any chance to discuss the patient records with the auditors or to see the results of the audit itself.”
Blossom’s attorney stated that he had since learned secondhand from DOH that the auditors “had no intention of ever returning” to conduct an exit conference, and that DOH was preparing Blossom’s CMI based on the March 14, 1994 review. He protested that “without question” Blossom was “not afforded its right” to an exit conference; he requested the exit conference in advance of DOH’s making any final decision regarding Blossom’s CMI.
No one from DOH ever replied to Blossom’s attorney’s letter. Indeed, Blossom did not hear from anyone at DOH on the subject of the Stage II audit until March 28, 2001—almost exactly seven years later—when the same health care fiscal analyst responsible for affording Blossom the March 14, 1994 review wrote to offer yet another reexamination. DOH offers no explanation for this seven-year gap other than “administrative oversight.”
On April 30, 2001, Blossom’s attorney (not the same one who had written to DOH before), accepted the offer of another review of the July 1993 PRIs and chided DOH for not promptly providing Blossom with the exit conference requested seven years earlier. Further, he disputed DOH’s right to use the audited PRIs to calculate the facility’s Medicaid reimbursement rate retroactively for periods from July 1993 forward.
On July 2, 2001, the Director of DOH’s Bureau of Financial Management and Information Support responded to Blossom’s
attorney by denying his “request” for DOH to accept Blossom’s PRIs without auditing their accuracy. Noting that although DOH “sincerely regret[s] that the final adjudication of the July[ ] 1993 PRI review has yet to occur, we cannot agree to disregard audit findings and circumvent longstanding protocols for determining when reviews are necessary.” He reiterated DOH’s offer to afford Blossom a “re-review” of the 27 controverted items from the Stage II audit, and pledged to “take all steps practical to move the audit process along as expeditiously as possible for subsequent periods” so as “to make the facility current at the earliest possible date.”
This second bonus review took place another eight months later, on February 22, 2002. Blossom did not question about half of the controverted items, apparently because supporting documentation for the July 1993 PRIs was in some cases no longer available. Of the remaining disputed findings, the auditors agreed with their predecessors on most, and with Blossom on only a few. As a result, the level of discrepancy between the July 1993 PRIs as submitted by Blossom and after audit was still statistically significant. Consequently, on March 28, 2002, DOH’s health care fiscal analyst directed Blossom to enter into a contract with an outside assessor approved by DOH to perform its next four rounds of PRIs (10 NYCRR 86-2.30 [f] [1] [ii]; [2]).
On May 3, 2002, Blossom’s administrator requested a “waiver for Part 86-2.30,” essentially citing DOH’s tardiness in completing the audit of the July 1993 PRIs. On June 12, 2002, the Director responded, stating that it was “unclear” whether Blossom was requesting that DOH “forego processing of the audit results in total” or “rescind its instruction that the facility utilize an independent assessor for PRI submissions between July 2002 and April 2003.” In either event, he denied the request.
In August 2002, DOH notified Blossom of its intention to audit Blossom’s January 1994 PRIs in November 2002. Blossom’s attorney wrote the Director and the fiscal health care analyst to object that DOH could “no longer conduct a timely, lawful audit of the 1994 rate period,” asked DOH to withdraw its audit request and threatened a lawsuit if it did not. On October 17, 2002, the Director denied the request and stated that the audit would take place on November 18, 2002. Further, he warned, “If medical records are not made available to the auditor, we cannot presume the facility’s PRIs are correct. Consequently, all issues relating to hierarchies would need to be denied and all issues relating to [ADLs] would need to be reduced to their lowest level.”
When further efforts to persuade DOH to withdraw or postpone the impending audit proved futile, Blossom commenced this proceeding in Supreme Court on November 15, 2002. Blossom sought judgment declaring respondents’ “announced intention” to audit 1994 PRIs to be arbitrary, capricious, irrational and contrary to state and federal law because “untimely as a matter of law”; declaring any “attempt” by respondents to audit PRIs for 1995 and 1996 to be arbitrary, capricious, irrational and contrary to state and federal law; ordering respondents to recalculate its Medicaid reimbursement rates for 1994, 1995 and 1996 using the CMI derived from the PRIs originally filed for these years; and enjoining respondents from auditing Blossom for “all periods” in 1994, 1995 and 1996.
By order to show cause dated November 18, 2002, Supreme Court granted Blossom a temporary restraining order barring respondents from auditing PRIs filed for the time period from January 1, 1994 through December 31, 1996. On January 3, 2003, Supreme Court issued a decision in Blossom’s favor; on February 4, 2003, Supreme Court entered an order barring respondents from auditing Blossom’s PRIs for 1994 through 1996 and directing them to recalculate Blossom’s Medicaid reimbursement rates for those years in accordance with the PRIs as originally filed. The Appellate Division reversed on the law and dismissed the petition. We subsequently granted Blossom leave to appeal, and now reverse.
III.
Prior to April 1, 1983, Medicaid audit responsibilities were “inefficiently distributed among various State agencies including the Departments of Social Services and Health and the Offices of Mental Hygiene and Mental Retardation and Developmental Disabilities” (Mem of State Exec Dept, reprinted in McKinney’s 1983 Session Laws, at 2415, 2417). Accordingly, as of April 1, 1983 all Medicaid-related audit activities were vested in the Department of Social Services (DSS), including auditing the fiscal and statistical records and reports filed by nursing homes to support their Medicaid reimbursement rates
(see
L 1983, ch 83, §§ 4, 9; Social Services Law §§ 364, 368-c; 18 NYCRR 517.3 [a] [1]). DOH continued to set and adjust Medicaid rates
(see
Public Health Law § 2803 [2]; § 2807 [3]; § 2808) and so also, after the advent of RUG-II, undertook to review PRIs (10 NYCRR 86-2.30 [e] [3], [5]).
As of October 1, 1996, responsibility for administering the Medicaid program was shifted from DSS to DOH
(see
L 1996, ch
474, §§ 233-248; Social Services Law § 363-a [1]; Public Health Law § 201 [1] [v]). DSS’s rules and regulations pertaining to transferred Medicaid functions continued in full force and effect as rules and regulations of DOH (see L 1996, ch 474, § 242). DSS retained the Medicaid audit function until April 1, 1997, when it was renamed the Department of Family Assistance, and the “medicaid audit function pursuant to sections 364 and 368-c of the social services law” was placed with DOH (see L 1997, ch 436, part B, § 122 [a], [e]).
The specific DOH and DSS regulations cited by the parties as relevant to this appeal are 10 NYCRR 86-2.7 (relating to audits of fiscal and statistical records and reports), promulgated by DOH in 1976; 18 NYCRR 517.3 (relating to audits of fiscal and statistical records and reports), promulgated by DSS in 1985; 10 NYCRR 86-2.30 (relating to PRIs), promulgated by DOH in 1985; and 10 NYCRR 415.22 (relating to patients’ clinical records), promulgated by DOH in 1990.
There is considerable overlap between 10 NYCRR 86-2.7 and 18 NYCRR 517.3. This is not surprising since these provisions both address auditing fiscal and statistical records and reports, the task transferred from DOH to DSS in 1983, and then returned from DSS to DOH in 1997. Thus, 10 NYCRR 86-2.7 provides that
“[a] 11 fiscal and statistical records and reports shall be subject to audit. All underlying books, records and documentation which formed the basis for the fiscal and statistical reports, filed by [the nursing home] with the [State],
shall be kept and maintained by the facility for a period of time not less than six years from the date of filing, or the date upon which the fiscal and statistical records were to be filed, whichever is the later date.
In this respect, any rate of payment . . . will be construed to represent a provisional rate until such audit is performed and completed, at which time such rate or adjusted rate will be construed to represent the audited rate” (emphasis added).
Similarly, 18 NYCRR 517.3 (a) (1) specifies that for cost-based providers such as Blossom
££[a]ll fiscal and statistical records and reports . . . used for the purpose of establishing rates . . . and all underlying books, records, documentation and reports which formed the basis for such fiscal and statistical records and reports are subject to audit. All underlying books, records and documentation which formed the basis for the fiscal and statistical reports filed by a provider with any State agency responsible for the establishment of rates of payment or fees
must be kept and maintained by the provider for a period of not less than six years from the date of filing of such reports, or the date upon which the fiscal and statistical records were required to be filed, or two years from the end of the last calendar year during any part of which a provider’s rate or fee was based on the fiscal and statistical reports, whichever is later. . . .
[A]ny rate of payment certified or established . . . will be construed to represent a provisional rate until an audit is performed and completed, or the period within which to conduct an audit has expired without such audit having been begun or notice of such having been issued, at which time such rate or adjusted rate will be construed to represent the final rate as to those items audited” (emphasis added).
18 NYCRR 517.3 (a) (2) concomitantly specifies that ££[a]ll required fiscal and statistical reports are subject to audit
for a period of six years from the date of their filing or from the date when such reports were required to be filed, whichever is later”
(emphasis added); and 18 NYCRR 517.3 (c) directs that
££[n]otification by the department to the provider of the department’s intent to audit shall toll
the six-year period for record retention and audit.
The department shall not notify a provider of its intent to audit more than six years from the date of filing of the fiscal and statistical reports to be audited” (emphasis added).
With respect to PRIs, 10 NYCRR 86-2.30 (e) (3) states that “[t]he patient review form (PRI) and any underlying books, records, and/or documentation which formed the basis for the completion of such form shall be subject to review by [DOH].” Further, “[DOH], in order to ensure accuracy of the [PRI], may also
conduct timely on-site observations and/or interviews of patients ¡residents and review of their medical
records” (10 NYCRR 86-2.30 [e] [5] [emphasis added]). Finally, 10 NYCRR 415.22 (b) requires a nursing home resident’s “[c]linieal records” to “be retained
for six years from the date of discharge or
death” (emphasis added).
As Blossom points out, PRIs and “any underlying books, records, and/or documentation which form[ ] the basis for [their] completion” (10 NYCRR 86-2.30 [e] [3]) in common with “fiscal and statistical records and reports” and “[a]ll underlying books, records and documentation which form[ ] the basis for . . . fiscal and statistical reports” (10 NYCRR 86-2.7; 18 NYCRR 517.3 [a] [1]), significantly influence a facility’s Medicaid reimbursement rate; and six years is indisputably the period for record retention and audit of fiscal and statistical reports and their supporting documentation as well as for the retention of a resident’s clinical records after discharge or death. It does not follow, however, that DOH must audit PRIs within six years of filing.
As an initial matter, PRIs are not “fiscal and statistical records and reports.” The latter are the “cost reports” that nursing homes must file annually with DOH pursuant to 10 NYCRR 86-2.2. Nor are PRIs the “underlying books, records and documentation which form[ ] the basis for . . . fiscal and statistical reports,” which include such things as employee payroll records, mortgage and loan documents, and purchase records. For one thing, fiscal and statistical reports must be completed in accordance with generally accepted accounting principles and certified by an accountant (10 NYCRR 86-2.4, 86-2.5). These requirements make no sense with respect to PRIs, which are instead subject to a three-stage audit process (10 NYCRR 86-2.30 [e], [f], [g]) and certification by “the [facility] operator and the nurse assessor responsible for [PRI] completion” (10 NYCRR 86-2.30 [c] [3]). Further, DOH promulgated 10 NYCRR 86-2.30 in 1985, a time when DSS—not DOH—was responsible for auditing a nursing home’s fiscal and statistical reports.
Courts normally “defer to an administrative agency’s interpretation of its regulations if not irrational”
(Matter of Sylcox
Nursing Home & Health Related Facility v Axelrod,
184 AD2d 986, 988 [1992],
Iv denied
80 NY2d 761 [1992]). Deference is especially fitting here where DOH has promulgated a separate regulation governing PRIs, which does not impose any specific deadline for PRI audits and instead simply instructs that they must be timely conducted
(see
10 NYCRR 86-2.30 [e] [5]).
Nonetheless, Blossom protests that the “only rational interpretation that can be gleaned” from the relevant regulations is that 10 NYCRR 86-2.30 (e) (5) incorporates the six-year limitations period specified in 10 NYCRR 86-2.7 and 18 NYCRR 517.3 (a) and (c). Otherwise, nursing homes would be subject to financial uncertainty and would have to cope with PRI audits many years after residents had died or been discharged (which usually occurs within 2.5 years of arrival), staff had left and the rules for completing or auditing PRIs might have changed. Consequently, Blossom argues, any audit of PRIs conducted more than six years after filing places the nursing home in an untenable position. The auditor must review residents’ clinical records to validate PRIs, but the nursing home may have lawfully destroyed these records for at least some of the residents in the audit population.
These are valid points. It is also true that significantly delayed or protracted PRI audits harm the public fisc by thwarting prompt recoupment of any Medicaid overpayments. As we pointed out in
Matter of Elcor Health Servs. v Novello
(100 NY2d 273, 280 [2003]), however, “[t]hat [DOH’s] interpretation might not be the most natural reading of the regulation, or that the regulation could be interpreted in another way, does not make the interpretation irrational.”
Although PRI audits are not subject to the six-year limitations period specified in 10 NYCRR 86-2.7 and 18 NYCRR 517.3 (a) and (c), “timely” is not synonymous with “timeless.” In this case, DOH sought to audit Blossom’s January 1994 PRIs more than six years after they were filed solely because DOH neglected to wrap up its Stage II audit of Blossom’s July 1993 PRIs until 2002, and PRI audits take place sequentially. Because DOH offers no better explanation than “administrative oversight” (meaning inadvertence, not supervision) for the seven-year hiatus in its nearly nine-year long audit of Blossom’s
July 1993 PRIs, we hold that any audit of Blossom’s PRIs filed in 1995 and 1996 is untimely as a matter of law.
Accordingly, the order of the Appellate Division should be reversed, with costs, the proceeding converted to a declaratory judgment action and judgment granted declaring that respondents may not audit or adjust Blossom’s CMI for the period January 1, 1995 through December 31, 1996, and that Blossom’s Medicaid reimbursement rates for the period from January 1, 1994 through December 31, 1996 be determined by implementing the CMI as calculated using the PRIs originally filed by Blossom for 1994, 1995 and 1996.
Chief Judge Kaye and Judges G.B. Smith, Ciparick, Rosenblatt, Graffeo and R.S. Smith concur.
Order reversed, etc.