This is an appeal of the decision of the Circuit Court of Cole County affirming a decision of the Administrative Hearing Commission denying Appellant EBG Health Care III, Inc. an increase in its Medicaid per diem reimbursement rate. The judgment is affirmed.
EBG Health Care III, Inc. (“EBG”) owns and operates a nursing home in Springfield. The home is known as Woodland Manor. The home is a 180-bed facility which is licensed and certified as a skilled nursing facility providing long-term residential care. EBG participates in the Medicaid program pursuant to § 198.045, RSMo 1986.
The nursing home entered the Medicaid program in 1973 under prior ownership. Except for a change in its reimbursement rate in 1984 due to a change in its level of care, the only changes in the reimbursement rate since 1981 have been in recognition of inflation. In 1984, the level of care-was changed from an intermediate care facility (ICF) to a skilled nursing facility (SNF), resulting in an increase in the reimbursement rate.
In 1988, while the home was owned by another party, the Missouri Division of Aging (DOA) determined that the home was not complying with state regulations. DOA removed 40 residents from the home, and forced the home to enter into a consent agreement. The consent agreement set standards for various care items, such as, for instance, requiring that only an RN or specially trained LPN would be permitted to install tube feedings. The consent agreement also required the hiring of an independent monitor and an independent consultant. It also set standards for nurse supervisors and established an enhanced ratio of nurse aides per patient for the different duty shifts.
EBG bought the facility in May, 1989, after the former owners had been operating under the consent agreement for more than one year. EBG entered into a new consent agreement with DOA. The new consent agreement did not mandate the enhanced care ratios set forth in the earlier consent agreements. In July, 1989, an inspection by DOA revealed no deficiencies, and the consent agreement was terminated.
The base rate applicable to the home was $33.85 for the overall reimbursement rate, of which $11.80 was the base rate for actual patient care costs, per patient per day. With the change in the level of care from ICF to SNF, and the inflation factors included, EBG was being reimbursed at the overall rate of $44.13, and the actual patient care portion of that was $15.54 (per patient) per day in 1989. EBG was spending for actual patient care, at the time it requested rate reconsideration, $26.80 per day. Thus, the actual cost difference was $11.26 per day. EBG requested an increase of $10.41 per day.
After denial by the Department of Social Services of EBG’s request for an increase in its Medicaid reimbursement rate, EBG sought review of the matter by the Administrative Hearing Commission under § 621.055 and § 208.156, RSMo 1986. The Administrative Hearing Commission ruled against EBG. The Circuit Court of Cole County affirmed the decision of the Administrative Hearing *145Commission, and EBG brings its appeal to this court.
On appeal, we review the decision of the Administrative Hearing Commission (“the Commission”) and not the decision of the circuit court. Welty v. State Bd. of Chiropractic Examiners, 759 S.W.2d 295, 297 (Mo.App.1988). Generally, our review is limited to determining whether the decision is supported by substantial and competent evidence upon the whole record, whether it is arbitrary, capricious, or unreasonable, or whether the Commission abused its discretion. Department of Social Services v. Our Lady of Mercy Home, 803 S.W.2d 72, 75 (Mo.App.1990). This court will not substitute its judgment on factual matters for that of the Commission. Barnes Hosp. v. Missouri Comm’n on Human Rights, 661 S.W.2d 534, 535 (Mo. banc 1983). The Commission judges the credibility of the witnesses. St. Louis County v. State Tax Comm’n, 406 S.W.2d 644, 649 (Mo. banc 1966). If the evidence would establish either of two opposing possible findings, we must uphold the factual determination of the Commission. Walker v. Supervisor of Liquor Control, 781 S.W.2d 113, 114 (Mo.App.1989).
EBG had argued before the Department of Social Services that it was entitled to a rate increase due to a “change in ease mix” and also due to “extraordinary circumstances.” Before the Commission, EBG dropped the claim of extraordinary circumstances, and submitted its case only on the basis of a change in case mix. The term “change in case mix” means the change in the average acuity level of its residents. Regulation 13 C.S.R. 70-10.010 at § (4)(A)3.F. provides:
The prospectively determined reimbursement rate may be adjusted only under the following conditions:
# ⅝ ⅜ ⅜ # ⅜
F. When an adjustment to a facility’s rate is made in accordance with the provisions of § (7) of this rule[.]
(7)(D) provides as follows:
The Committee may review the following conditions for rate reconsideration:
a. Those costs directly related to a change in a facility’s case mix_
The DOA has a point system by which it measures the acuity level of the patients in a facility. Each patient is rated from 0-9 in nine different categories of care. Therefore, a completely self-sufficient resident would be a 0, and a patient requiring maximum care around the clock would be an 81. The DOA study in 1980 for this facility showed an average point per patient of 29.0. The study in 1989 showed the average point per patient had risen to 34.2. From August, 1986 to August, 1989, DOA surveys showed acuity levels from 30.99 to 35.06. Thus, there was an increase in the sickness of the patients from 1980 to 1989, but not such a large increase as to require that the facility in question be rated a skilled nursing facility. A patient whose acuity level is 54 or more requires a SNF. Since 1984, this nursing facility has been operated as a SNF.
The hearing before the Commission was conducted in September, 1990. At the hearing, Ewing Gourley, President and major shareholder of EBG, testified that, even after the consent agreements expired, EBG maintained the same level of staffing specified by the consent agreement because he believed it was necessary to take care of the residents’ needs. Mr. Gourley testified that, to operate the home on a level mandated by the consent agreement, it was necessary to have 57 staff members, whereas only 31 staff members would be the basic requirement- to provide the level of care necessary under the regulations.
Barbara Bullington, the administrator of the home, testified on behalf of EBG that she had reviewed records going back as far as 1980. She said there had been a “tremendous change” in the needs of the residents of the home. On cross-examination, she was asked whether the change from ICF to SNF would be the cause of the changes she had seen. She answered “not particularly,” since often there are ICF patients in a SNF, and often SNF patients in an ICF. She suggested that the change in the care needs of the residents was due to the fact that hospitals were releasing patients from hospital care earlier than was formerly the case.
*146Amy Frett, the facility’s director of nursing, testified that patients coming into the facility are more acutely ill than in 1980. She had prepared an exhibit which was a comparison of patient statistics for 1990 with patient statistics for the population in 1980. Instead of using the same categories used by the DOA in her exhibit, she had developed her own categories of patient conditions. Her testimony proceeded through each line of the exhibit. She compared the number of residents having different conditions in 1980 with the number found to have similar conditions on September 10, 1990, and described why such conditions required care or attention. She testified that 90 out of 149 patients were on bowel and bladder training. She said that the nursing home was using Foley catheters for incontinent patients to a much lower degree currently than in 1980. She suggested that having the patients on bowel and bladder training (meaning that they are taken to the restroom frequently on a regular schedule) was more in keeping with the greater emphasis of the United States Health Care Financing Administration on patient outcomes. The following are the categories and figures set forth in her exhibit:
WOODLAND MANOR — PATIENT STATISTICS
12/30/80 (Based on 73 Res.) 09/10/90 (Based on 149 Res.)
# of Res. % # of Res. %
Elopement 1 1.4% 7 4.7%
Blind 0 0% 1 .7%
Decubitus 4 5.5% 24 16.1%
Tube Feeding 1 1.4% 4 2.7%
Feeding — -Total 12 16.4% 56 37.6%
Feeding — Assist 7 9.6% 37 24.8%
Feeding — Self 43 58.9% 51 34.2%
N on-Ambulatory 24 32.9% 83 55.7%
Bladder Incontinency 16 21.9% 91 61.1%
Bowel Incontinency 16 21.9% 85 57.0%
Confused/Disoriented 27 37.0% 114 76.5%
Restorative Program 14 19.2% 65 43.6%
Routine Lab 7 9.6% 101 67.8%
Requires turning every 2 hrs. 10 13.7% 90 60.4%
Transfer with Assistance 15 20.5% 102 68.5%
Seizures 1 1.4% 11 7.4%
Behavior Disruptive/Combative
Violent 4 5.5% 25 6.8%
Restraint 14 19.2% 77 51.7%
Catheter 11 15.1% 18 12.1%
Bowel & Bladder Training 8 11.0% 90 60.4%
Assistance with Daily Living
—Assistance Needed 32 43.8% 51 34.2%
—Total Care 20 27.4% 94 63.1%
—Self 21 28.8% 5 3.4%
Diabetic 8 11.0% 21 14.1%
Ambulatory 35 47.9% 32 21.5%
Ambulatory with Assistance 14 19.2% 38 25.5%
Limb Contracture 3 4.1% 32 21.5%
Special Skin Care 15 20.5% 53 35.6%
Speech Difficulty 0 0% 41 27.5%
Direct Physical Therapy 15 20.5% 4 2.7%
Amputations 1 1.4% 4 2.7%
Oxygen 0 0% 9 6.0%
*147On cross-examination it was shown that many of the statistics concerning the 1980 patients were based on her memory and the memory of some others rather than on actual records. In 1980, Ms. Frett was in nursing school and worked as an aide in the nursing home. There was substantial overlap in the 32 categories she used in developing her exhibit. For instance, “eonfused/disori-ented,” “transfer with assistance,” “behavior disruptive/combative/violent,” “restraints” and “assistance with daily living” are categories which would tend to substantially overlap. There were also categories listed as to which the facility had no care program, such as “speech difficulty” and “amputations” (amputees, of course, would be included in the nonambulatory category and other categories as well).
When Ms. Frett’s statistics for 1990 were totaled up for certain categories, the total of residents shown on the exhibit was different from the number of actual residents at the time of the survey. For instance, with regard to feeding, the exhibit showed that the four categories of feeding total up to 148 instead of 149, which is the number of residents. Also, the categories designated “ambulatory,” “ambulatory with assistance,” and “non-ambulatory” add up to 153. As to all such items where there appeared to be a discrepancy on the exhibit, Ms. Frett acknowledged there must be an error of some kind, but she maintained that all the figures were checked for accuracy.
The Department presented the testimony of Donna Fischer, a medical review nurse with the Division of Aging. She testified that she, along with two others, conducted an on-site inspection of the facility in August, 1990 (one month before the AHC hearing, and several weeks before Amy Frett’s September 19, 1990 survey). Ms. Fischer testified that she and the other inspectors each reviewed charts and then interviewed residents. Altogether, they reviewed the circumstances and records of 121 out of the 149 patients.
Ms. Fischer testified that of the patients she reviewed, she found none on any bladder and bowel training program. She said that any who were documented to be incontinent all had Foley catheters. She said she saw no documentation of patients with recent seizures. She said that having a patient with a documented seizure in the past was of little or no significance for the care demands because patients susceptible to seizures are on medications which generally control the seizures.
In its ruling, the Commission noted that a rate adjustment may be allowed for a change in case mix where the increased costs are “directly related” to a change in case mix. The Commission found that EBG had shown that it experienced increased costs for direct care. However, it held that EBG had failed to relate those increases directly to a change in case mix.
Direct relation contemplates more than coincidence ... if the provider shows only that it has incurred increased costs and that it has experienced a change in case mix, it has hot established a direct relation between the two events. Rather, the provider must establish a causal connection. It must show that the change in case mix caused the increased costs.... Although EBG has shown that it did experience a change in case mix as well as increased costs, it has failed to link the two.
On appeal, EBG contends that the Commission erred in its interpretation of the applicable regulation “when it concluded as a matter law that appellant did not link its increased costs to its case mix.” EBG contends the Commission misinterpreted or misapplied the law and regulations in its determination and asserts that it satisfactorily proved its entitlement to a rate adjustment.
EBG attacks in particular the following language in the Commission’s decision:
[The provider] must show that the change in case mix caused the increase costs. This can only be done by identifying the increased costs and tracing them item by item back to the situation requiring the increased expenditure. For example, a provider might show that a rise in the number of bedridden resident has resulted in a need for additional personnel to regu*148larly reposition the residents and that such need increased the provider’s personnel costs. Such a showing would necessarily require quantitative documentation.
While a reading of this portion of the opinion in isolation might suggest that the Commission is requiring a very detailed, very stringent test of causation, a reading of the entire opinion of the Commission in context shows that the Commission was simply trying to articulate a common sense understanding of what is required to show that the increased costs were directly related to a change in case mix. The opinion of the Commission shows that the Commission was bothered by difficulties in EBG’s evidence. The Commission noted that the records for 1980 were incomplete, yet these records formed the basis of the comparison statistics. The Commission also noted that the exhibit presented by Ms. Frett, and upon which EBG principally relied, was shown to contain “gross errors” and methodological weaknesses. Further, the exhibit and the accompanying testimony concerning the bowel and bladder training programs were impeached by the testimony of Ms. Fischer, the DOA medical review nurse, who participated in a review just two weeks before the date of compilation of Ms. Frett’s exhibit. Thus, it appears that the Commission was not quarreling with the kind of evidence presented by EBG — that is, testimony showing that increased acuity level of the residents had required more staffing costs — but rather that the Commission simply found EBG’s evidence to be less than credible.
The Division of Aging’s independent survey of the acuity of the patients showed an increase in acuity from 29.0 to 34.2. While this represents an increase in acuity, it is not an overwhelming increase. It is not necessary for the facility to be a skilled nursing facility to care for patients with an average acuity of 34.2. Although the statistical exhibit prepared by Ms. Frett was admissible as a statistical compilation under § 536.070, RSMo 1986, the Department of Social Services was entitled to challenge the exhibit as to its mode of preparation, the weight to be given the exhibit, and the conclusions which can reasonably be drawn from it. The exhibit was designed in such a way as to cause the appearance that there was a major change in the acuity level of the patients. Ms. Frett had created 32 categories of care instead of the nine commonly used by DOA. Some of the categories had little to do with care needs, and many of the others involved overlapping. Some of the figures were necessarily erroneous. Furthermore, the assertion by the exhibit and by Ms. Frett’s testimony that 90 residents were on bowel and bladder training was undercut by the testimony of the DOA’s medical review nurse.1 The absence in evidence of the underlying records themselves, as opposed to the purported statistical summaries of the records, handicapped the Commission’s ability to determine the accuracy of the evidence which suggested a significant increase in the cost of care due to a change in case mix. This handicapping of the Commission worked to the detriment of EBG, since EBG bore the burden of proof.
It was shown that EBG provides more staffing than is required by the regulations. While this enhanced level of care is of benefit to the residents, as the Commission noted, the law does not require the Department of Social Services to pay for the enhanced level of care unless it is shown that the increased staff is necessary due to the change in the level of acuity. The applicant bears the burden of proof. Section 621.055, RSMo 1986. Here, the burden was not carried to the satisfaction of the Commission.
The decision of the Commission is supported by the evidence. It was not arbitrary, capricious, unreasonable or an abuse of *149discretion. We will not substitute our judgment for that of the Commission. Barnes Hospital, 661 S.W.2d at 535. The Commission judges the credibility of the witnesses. Smarr v. Sports Enterprises, Inc., 849 S.W.2d 46, 47 (Mo.App.1993).
Judgment is affirmed.
BRECKENRIDGE, J., concurs.
BERREY, J., dubitante.
. Ms. Fischer’s testimony also suffered from some "methodological weaknesses." She had been one of three persons participating in the review. She did not interview all 121 of the residents surveyed. We do not know how many she interviewed; presumably, it was roughly one-third. It is not known whether she reviewed the records of the survey work of the other two participating in the survey; she may have been testifying only as to the patients she reviewed. Nevertheless, the impact of her testimony, that she found no residents on bowel and bladder training (in contrast to Ms. Frett's testimony that 90 out of 149 residents were on such training) was for the Commission to determine.