*339Petitioner Raleigh Radiology LLC ("RRAD") and Respondents N.C. Department of Health and Human Services, Division of Health Care Regulation, Healthcare Planning and Certificate of Need (the "Agency"), and Duke University Health System ("Duke") all appeal an amended final decision of the Office of Administrative Hearings regarding a Certificate of Need ("CON") for an MRI machine.
I. Background
In early 2016, the State Medical Facilities Plan determined a need for one fixed MRI machine in Wake County and began fielding competitive requests from various applicants. Duke and RRAD each filed an application for a CON with the Agency in April 2016.
In September 2016, the Agency conditionally approved Duke for the CON and denied RRAD's application.
In October 2016, RRAD filed a Petition for Contested Case Hearing. Duke was permitted to intervene in the contested case.
In November 2014, a contested case hearing was held before an administrative law judge (the "ALJ").
On 16 March 2018, the ALJ issued its Final Decision, reversing the decision of the Agency and ordering that "[t]he Certificate of Need shall be awarded to [RRAD]."
Duke and the Agency timely appealed. RRAD also timely cross-appealed.
II. Standard of Review
We review a final decision from an ALJ for whether "substantial rights of the petitioners may have been prejudiced[.]" N.C. Gen. Stat. § 150B-51(b) (2018). We use a de novo standard if the petitioner appeals the final decision on grounds that it violates the constitution, exceeds statutory authority, was made upon unlawful procedure, or was affected by another error of law. N.C. Gen. Stat. § 150B-51(b)(1)-(4), (c) (2018). And we use the whole record test if the petitioner alleges that the final decision is unsupported by the evidence or is "[a]rbitrary, capricious, or an abuse of discretion." N.C. Gen. Stat. § 150B-51(b)(5)(6), (c) (2018).
III. Analysis
Duke and the Agency argue that the ALJ erred in conducting its own "comparative analysis review" of the two CON applications. We review this question of law de novo . Cumberland Cty. Hosp. Sys. v. N.C. Dep't of Health & Human Servs. , 242 N.C. App. 524, 527, 776 S.E.2d 329, 332 (2015).
Section 131E-183 of our General Statutes sets forth the procedure the Agency should use when reviewing applications for a CON. N.C. Gen. Stat. § 131E-183 (2016). Specifically, the Agency uses a two stage process.
First, the "the Agency must review each application independently against the criteria [set by its regulations] (without considering the competing applications) and determine whether it 'is either consistent with or not in conflict with these criteria.' " Britthaven, Inc. v. N.C. Dep't of Human Res. , 118 N.C. App. 379, 385, 455 S.E.2d 455, 460 (1995) (citing N.C. Gen. Stat. § 131E-183(a) ). Each "applicant for the issuance of a CON has the burden of demonstrating compliance with the review criteria[.]" E. Carolina Internal Med., P.A. v. N.C. Dep't of Health & Human Servs. , 211 N.C. App. 397, 404, 710 S.E.2d 245, 251 (2011).
Second, where there are competing applications which have passed the first step, "the Agency must decide which of the competing [conforming] applications should be approved" based on various "comparative" factors. Britthaven , 118 N.C. App. at 385, 455 S.E.2d at 461. "There is no statute or rule which requires the Agency to utilize certain comparative factors." Craven Reg'l Med. Auth. v. N.C. Dep't of Health & Human Servs. , 176 N.C. App. 46, 58, 625 S.E.2d 837, 845 (2006). But, rather, the Agency has discretion to choose which factors by which it will compare competing applications. Id .
Where an unsuccessful applicant appeals the Agency decision in a CON case, the ALJ in a contested case does not engage in a de novo review, but simply reviews for correctness *340of the Agency decision, pursuant to N.C. Gen. Stat. § 150B-23(a). E. Carolina Internal Med. , 211 N.C. App. at 405, 710 S.E.2d at 252. In fact, "there is a presumption that 'an administrative agency has properly performed its official duties.' " Id. at 411, 710 S.E.2d at 255 (quoting In re Cmty. Ass'n , 300 N.C. 267, 280, 266 S.E.2d 645, 654 (1980) ).
In the present case, the Agency reviewed each application for the CON independently. Britthaven , 118 N.C. App. at 385, 455 S.E.2d at 460 (citing N.C. Gen. Stat. § 131E-183(a) ). Such review revealed that Duke's application conformed with all criteria and that RRAD failed to conform with respect to certain criteria. At that point, assuming that RRAD's application indeed failed to conform to certain criteria, it would have been appropriate for the Agency to proceed with issuing the CON to Duke. Nevertheless, the Agency, as stated in its seventy-four (74) pages of findings, additionally "conducted a comparative analysis of [Duke's and RRAD's applications] to decide which [one] should be approved," assuming that RRAD's application did satisfy all of the criteria. See id. at 385, 455 S.E.2d at 461.
The Agency, in its discretion, used seven comparative factors in reviewing the CON applications: (1) geographic distribution, (2) demonstration of need, (3) access by underserved groups, (4) ownership of fixed MRI scanners in Wake County, (5) projected average gross revenue per procedure, (6) projected average net revenue per procedure, and (7) projected average operating expense per procedure. This comparative analysis led the Agency to approve and award the CON to Duke.
However, in the contested case hearing, the ALJ deviated from the above factors and used two additional factors: (1) the types of scanners proposed by each applicant, and (2) the timeline of each proposed project. Of note, there was evidence that RRAD's proposed MRI machine was superior to the machine which Duke would use. It is this deviation and the reliance on additional comparative factors by the ALJ which we must conclude was error.
Indeed, adding two additional comparative factors is not affording deference to the Agency, but rather constitutes an impermissible de novo review of this part of the Agency's decision. Such a substitute of judgment by the ALJ is not allowed. E. Carolina Internal Med. , 211 N.C. App. at 405, 710 S.E.2d at 252.
Evidence was provided that the factors utilized by the Agency have been used in two previous MRI CON decisions and that the additional factors used by the ALJ have not been a part of the Agency's policies and procedures for many years. We note that information pertaining to RRAD's allegedly superior MRI machine was not included in RRAD's application, though it was otherwise presented at the Agency public hearing, but without an expert testifying as to the machine's medical efficacy. Even so, the Agency has the discretion to pick which factors it evaluates in conducting its own comparative analysis. Craven Reg'l Med. Auth. , 176 N.C. App. at 58, 625 S.E.2d at 845. Further, regarding the timeline factor used by the ALJ, there was testimony that the Agency puts little, if any, weight to this factor as the factor disadvantages new providers. The ALJ did not determine that the Agency acted arbitrarily and capriciously, but rather simply substituted his own judgment in weighing the factors. However, we cannot say that it was an abuse of discretion for the Agency to rely on the factors that it did.
Separately, RRAD argues that the Agency erred by concluding that its application was not conforming. But even assuming that the Agency incorrectly determined that RRAD's application did not conform to certain criteria, such error was harmless as the Agency proceeded with a comparative analysis of both applications as if RRAD's application did comply.
Therefore, we reverse the Final Decision and reinstate the decision of the Agency.1
*341IV. Conclusion
The ALJ erred in disregarding the comparative analysis of the Agency and conducting its own comparative analysis. Thus, we reverse the Final Decision and reinstate and affirm the decision of the Agency awarding the CON to Duke.2
REVERSED.
Judge BRYANT concurs.
Judge ARROWOOD concurs in result only.