RUFFIN, Presiding Judge.

James Freels suffers from a severe form of cerebral palsy. On September 3, 1999, the then five-year-old Freels, through his parents and a treating physician, petitioned the Georgia Department of Community Health (the "Department") seeking reimbursement under Medicaid for his hyperbaric oxygen therapy ("HBOT"). After the Department denied his petition, Freels requested a hearing before an administrative law judge ("ALJ"). The ALJ upheld the Department's refusal to reimburse Freels, and the Department then issued a final decision denying Medicaid coverage for Freels' HBOT. Freels appealed to the superior court, which reversed the Department's final decision. We granted the Department's application for a discretionary appeal, and for the reasons set forth below, we affirm in part, reverse in part, and remand the case to the superior court.

"[O]n appeal our duty is not to review whether the record supports the superior court's decision but whether the record supports the initial decision of the local governing body or administrative agency. . . ." Our function "is to determine whether the superior court has in its own final ruling committed an error of law." The "any evidence" standard applies to the superior court's review of the findings of fact of the ALJ.

The record shows that Freels' parents aggressively sought treatment for his cerebral palsy. Freels underwent two years of occupational therapy, speech therapy, and physical therapy, and then began a series of assisted movement exercises referred to as "patterning." In February of 1999, shortly after Freels turned five, his parents learned of HBOT from a volunteer who had assisted with Freels' patterning exercises. Freels' family arranged for a series of forty-two treatments using HBOT.

Norman Harbough, Freels' physician, submitted a request to the Department for reimbursement for the costs of Freels' HBOT from Medicaid. In a November 30, 1999 letter to Harbough, the Department denied the request, explaining that, "[o]ur physician panel has determined that HBOT in Cerebral Palsy is experimental/investigational and is not a generally accepted practice at this time." Harbough asked for reconsideration, but the Department upheld the denial because there was "no scientific data to support the medical necessity for HBOT in this case."

After the Department's medical review decision, Freels asked for an administrative hearing. At the hearing, Freels presented the testimony of Dr. Paul Harch and his father David Freels. Dr. Harch is a board certified physician in the field of hyperbaric medicine. Dr. Harch also has extensive experience in using a brain blood flow imaging technique known as a SPECT scan. SPECT is an acronym for single photon emission computed tomography, which is a functional image of the blood flow in the brain. Dr. Harch admitted, however, that the use of SPECT scans as a reliable source of objective data is in dispute.

According to Dr. Harch, HBOT involves enclosing the patient's body in a pressurized vessel containing pure oxygen, which causes an increase in the amount of oxygen dissolved in the patient's blood. Dr. Harch explained his theory behind the effective use of HBOT in treating cerebral palsy. After the brain is damaged in patients with cerebral palsy, Dr. Harch testified, some brain cells may become idle. The delivery of more oxygen to the damaged area may encourage growth of blood vessels and thus restore function to the cells. This theory, Dr. Harch concluded, is consistent with his review of SPECT scans generally showing more normal or increased brain blood flow in patients who have undergone a series of HBOT treatments.

Dr. Harch performed a neurological exam on Freels. He looked at Freels' SPECT scan brain images taken before HBOT and those taken after a number of treatments. The images showed a "generalized improvement in the flow, but specifically . . . in the general regions of his speech motor area." Dr. Harch noted that the difference in the two scans shows a change from an asymmetric to an even distribution in the brain blood flow. Dr. Harch was impressed with the difference in images given the improvement in Freels' speech. Dr. Harch testified that HBOT was safe and effective.

James Carroll, a board certified neurologist specializing in child neurology, was called as a witness by the Department. Dr. Carroll testified that there was no biological reason why HBOT would cause cells to develop in the brain where they were not surviving. He also stated his opinion that SPECT scans only showed blood flow in the brain and could not be related to function. Dr. Carroll testified that even though there may have been evidence of improvement in Freels' condition during the HBOT therapy, the improvement could not necessarily be attributed to the HBOT. He further testified that he had no reason to believe the HBOT therapy would correct or ameliorate any of Freels' medical problems and that HBOT therapy was not an acceptable standard treatment for children with cerebral palsy.

The Department also called child neurologist Frank Berenson, who testified that it has not been established that HBOT therapy is medically beneficial for the treatment of children with cerebral palsy. He corroborated Dr. Carroll's testimony that SPECT scans do not measure the functional activity of the brain, and he stated his opinion that it was medically implausible that HBOT could change the injury to the brain caused by cerebral palsy. Based upon this, and other evidence, the Department decided that it would not reimburse Freels for the treatment.

1. The superior court reversed the Department's final decision because it was affected by error of law. Specifically, the superior court found that the Department applied the wrong legal standard by focusing on whether HBOT was an accepted treatment that was medically necessary. According to the superior court, the proper inquiry was whether HBOT was necessary "to correct or ameliorate a physical or mental defect or condition" regardless of whether it is an accepted medical practice. The Department contends the superior court erred in finding that the legal standard used by the Department in reaching its final decision failed to comport with Medicaid requirements. We disagree.

The General Assembly has designated the Department as the agency authorized to adopt and administer the plan for medical assistance under the federal Medicaid program. Medicaid is

a cooperative venture of the state and federal governments. A state which chooses to participate in Medicaid submits a state plan for the funding of medical services for the needy which is approved by the federal government. The federal government then subsidizes a certain portion of the financial obligations which the state has agreed to bear. A state participating in Medicaid must comply with the applicable statute, Title XIX of the Social Security Act of 1965, as amended, 42 U.S.C. § 1396, et seq., and the applicable regulations.

"The interpretation of a statute by an administrative agency which has the duty of enforcing or administering it is to be given great weight and deference." Nevertheless, the Department must comply with the applicable federal law, and, having chosen to participate in the Medicaid program, the State must provide the services required under the program.

Federal law governing the Medicaid program provides that eligible recipients under the age of 21 are entitled to early and periodic screening, diagnostic, and treatment ("EPSDT") services. Specifically, 42 U.S.C. § 1396d (r)(5) provides that EPSDT services include: "Such other necessary health care, diagnostic services, treatment, and other measures . . . to correct or ameliorate defects and physical and mental illnesses and conditions discovered by the screening services, whether or not such services are covered under the State plan." The parties agree that Freels is eligible for EPSDT services.

In its final decision, the Department noted that it "reimburses only for services which are medically necessary and within accepted professional standards." The Department also recited the definition of medically necessary services found in its policies and procedures manual:

Medically necessary services are those services which are reasonable and necessary in establishing a diagnosis and providing palliative, curative or restorative treatment for physical and/or mental health conditions. Services meeting professional standards are those which, in the opinion of [a] recognized peer, under usual circumstances contributed to a satisfactory outcome in the health status of the recipient. The services provided, as well as the type of provider and setting, must be appropriate to the specific medical needs of the recipient; and there must be no other equally effective, more conservative or substantially less costly course of treatment available. The determination of medical necessity shall be made in accordance with currently accepted standards of medical practice.

The Department denied Medicaid coverage to Freels because it found that, "[b]ased upon the evidence presented, Petitioner failed to satisfy the requisite burden of proof that HBOT treatments are an acceptable standard of medical practice and has not proven that HBOT is medically necessary for Petitioner."

The Department contends that its core standard for reimbursement, "palliative, curative or restorative treatment," is not necessarily at odds with the language of 42 U.S.C. § 1396d (r)(5), and that it is free to promulgate rules for the application and administration of the Medicaid program so long as the rules are consistent with the Medicaid program. The findings of the Department, however, refer to Freels' failure to establish HBOT as "medically necessary," a term defined at length in the quoted excerpt from the policy and procedures manual, indicating that the Department based its decision on its manual and not the applicable federal statute. And, even if we assume that medical services which provide "palliative, curative or restorative treatment," are functionally the same as medical services which "correct or ameliorate defects and physical and mental illnesses and conditions," the federal statute does not require that a treatment also be "an acceptable standard of medical practice" to be eligible for reimbursement. As the superior court ruled, "[i]nstead of requiring proof that HBOT is the accepted standard medical practice, or that it meets the definition of medical necessity reserved for adult Medicaid recipients, the [Department] should have focused its inquiry on whether HBOT was necessary to correct or ameliorate [Freels'] physical condition." The Department's findings show that the proper legal standard was not used in making its reimbursement determination, and we affirm the superior court's reversal of the Department's decision on this basis.

2. The Department also contends the superior court erred in reversing the Department's final decision on the grounds that the Department's findings are "clearly erroneous in view of the reliable, probative, and substantial evidence of record," and requiring the Department to reimburse Freels for his HBOT therapy. We agree.

Under the Administrative Procedure Act, the administrative agency's findings are judicially reviewable if they are "clearly erroneous in view of the reliable, probative, and substantial evidence on the whole record." OCGA §§ 50-13-19 (h)(5). Our courts have held that the "clearly erroneous" and "any evidence" standards of review on factual issues are synonymous and that the superior court cannot substitute its judgment for that of the ALJ as to the weight of the evidence on questions of fact. On appeal our duty is not to review whether the record supports the superior court's decision but whether the record supports the initial decision of the local governing body or administrative agency.

Here, the superior court essentially rejected the testimony of the expert witnesses called by the Department. The superior court found that neither Dr. Carroll nor Dr. Berenson had experience in the approved neurological uses of HBOT therapy "or in the methods upon which treatments can be shown to effect increased brain function," and concluded that these witnesses provide no evidence to refute the testimony showing that HBOT was correcting Freels' cerebral palsy. The superior court further found that the Department erred in giving weight to the testimony of witnesses who were unfamiliar with the medical literature on the use of HBOT on neurological indications and the use of SPECT scan analysis as a diagnostic tool to measure increased brain function. The record shows otherwise.

Dr. Berenson and Dr. Carroll reviewed material concerning the use of HBOT for patients suffering from cerebral palsy. Both doctors testified that SPECT scan analysis does not measure brain function. In Berenson's opinion, it was biologically implausible that HBOT could change the injury to Freels' brain that occurred through his cerebral palsy. The testimony also indicated that any improvement shown by Freels was not necessarily caused by the HBOT therapy.

The superior court's determination that Dr. Carroll and Dr. Berenson were not qualified to testify about the efficacy of HBOT therapy, and thus could not be relied upon by the Department, is unfounded. It stands to reason that the less utilized and more experimental the medical treatment, the smaller the pool of medical practitioners with any direct experience with the treatment. But that should not make it impossible for the Department to show that an experimental treatment such as HBOT is medically ineffective. Here, the Department chose to present the testimony of medical doctors who had extensive experience in treating cerebral palsy and were familiar with the mechanics of HBOT and the theory behind its application in treatment of brain injury. The trier of fact was entitled to use its discretion in determining whether these witnesses qualified to testify as experts. Doctors Carroll and Berenson demonstrated they were qualified to form an opinion as to the effectiveness of HBOT therapy, and the Department was entitled to rely on their testimony even though it was at odds with the testimony of Dr. Harch. Accordingly, the superior court erred in ruling that the reliable, probative, and substantial evidence showed that HBOT was correcting and ameliorating Freels' condition and in ordering the Department to reimburse Freels for his HBOT.

In sum, because the Department's final decision was affected by error of law, we affirm the order of the superior court reversing the Department's final decision. We nonetheless reverse the ruling of the superior court to the extent that it usurped the role of the factfinder in discounting the testimony of the Department's expert witnesses. We remand the case to the superior court with instructions that the superior court in turn remand the case to the Department for determination of whether, in light of this opinion, Freels is entitled to reimbursement for his HBOT.

Judgment affirmed in part and reversed in part and remanded with directions. Barnes, J., and Pope, Senior Appellate Judge, concur.

RUFFIN, Presiding Judge.

DECIDED NOVEMBER 19, 2002.