Opinion
RIVERA, J.This writ proceeding presents the question of whether a trial court must dismiss a petition to commit a person as a sexually violent predator (SVP) when the original SVP evaluations were prepared using an invalid protocol and replacement evaluations result in a split of opinion. We conclude the answer is no.
I. BACKGROUND1
While petitioner Roger Davenport was serving a term in a California state prison, prison officials referred him to the State Department of Mental Health (DMH) to determine if he met the criteria for commitment under the SVP Act (Welf. & Inst. Code, § 6600 et seq.).2 Two mental health professionals, Dr. Jeremy Coles and Dr. Thomas MacSpeiden, evaluated Davenport in accordance with a standardized assessment protocol developed by the DMH.3 Coles and MacSpeiden both concluded Davenport met the criteria for SVP commitment.
Based on the concurring evaluations, the San Francisco County District Attorney filed a petition to commit Davenport as an SVP. (§ 6601, subd. (d).) After a hearing, the trial court found probable cause to believe Davenport was an SVP.
While the matter awaited trial, the Fourth District Court of Appeal decided In re Ronje (2009) 179 Cal.App.4th 509 [101 Cal.Rptr.3d 689] (Ronje). Ronje *668held that portions of the standardized assessment protocol constituted an underground regulation—a regulation not adopted pursuant to the Administrative Procedure Act (Gov. Code, § 11340 et seq.; APA). (Ronje, supra, 179 Cal.App.4th at pp. 516-517.) Consequently, the standardized assessment protocol was invalid. (Ibid.) The Ronje court concluded the proper remedy to cure the use of the invalid protocol in pending cases was to (1) order new evaluations based on a valid assessment protocol and (2) conduct another probable cause hearing. (Id. at p. 519.)
In light of Ronje, the trial court ordered two new evaluations of Davenport. Coles and MacSpeiden reevaluated Davenport, presumably using a new assessment protocol adopted in compliance with the APA.4 Coles and MacSpeiden now disagreed as to whether Davenport met the SVP criteria. The DMH therefore appointed two new mental health professionals to evaluate Davenport. The result was another split of opinion.
Davenport moved to dismiss the proceedings on the ground the SVP petition was not supported by two valid concurring mental health evaluations. The trial court denied the motion and ordered a new probable cause hearing.
Davenport filed a petition in this court seeking a writ of mandate or prohibition compelling the trial court to grant his motion to dismiss the SVP proceedings. We summarily denied the petition and Davenport petitioned the Supreme Court for review. The Supreme Court granted review and transferred the case back to this court with directions to issue an order to show cause.
II. DISCUSSION
A. SVP Evaluations
The SVP commitment process begins when prison officials review the social, criminal, and institutional history of inmates convicted of certain sexual offenses. (§ 6601, subd. (b).) Inmates determined to be likely SVP’s are referred to the DMH for a “full evaluation.” (Ibid.) The evaluation is done in accordance with a standardized assessment protocol. (Id., subd. (c).) The protocol “shall require assessment of diagnosable mental disorders, as well as various factors known to be associated with the risk of reoffense among sex offenders.” (Ibid.)
Evaluations are conducted by two mental health professionals designated by the Director of the DMH. (§ 6601, subd. (d).) If the evaluators agree the *669person meets the SVP criteria, the director forwards a request for a commitment petition to the appropriate county. (Ibid.) In the event the mental health professionals disagree, the director must arrange for further examination by two “independent professionals.” (Id., subd. (e).) These professionals cannot be state employees. (Id., subd. (g).) A petition may then be filed only if the independent professionals concur the person meets the criteria for commitment. (Id., subd. (f).)
Petitions for SVP commitment are filed by the county’s designated attorney (usually the district attorney). (See § 6601, subds. (d), (h), (i).) The alleged SVP is entitled to a probable cause hearing to determine whether there is sufficient evidence to believe the person is likely to engage in sexually violent predatory behavior if released. (§ 6602, subd. (a).)
The objective of the evaluation process is to screen out individuals who plainly do not meet the SVP criteria. (See People v. Scott (2002) 100 Cal.App.4th 1060, 1063 [123 Cal.Rptr.2d 253] [Legislature has imposed procedural safeguards to prevent meritless petitions from reaching trial].) The actual legal determination that a particular person is an SVP, however, is made during the subsequent judicial proceedings, not during the screening process. (People v. Medina (2009) 171 Cal.App.4th 805, 814 [89 Cal.Rptr.3d 830] (Medina))
B. The Office of Administrative Law Determination
“The APA requires every administrative agency guideline that qualifies as a ‘regulation,’ as defined by the APA, to be adopted according to specific procedures. (Gov. Code, § 11340.5, subds. (a), (b).) The Office of Administrative Law (OAL) is charged with, among other functions, enforcing this requirement. (Gov. Code, §§ 11340.2, 11340.5, subd. (b).) If the OAL is notified or learns that an administrative agency is implementing a regulation that was not properly adopted under the APA, the OAL must investigate, make a determination, and publish its conclusions. (Gov. Code, § 11340.5, subd. (c).)” (Medina, supra, 171 Cal.App.4th at p. 813.)
Any regulation not properly adopted under the APA is labeled an “ ‘underground regulation.’ ” (Patterson Flying Service v. Department of Pesticide Regulation (2008) 161 Cal.App.4th 411, 429 [74 Cal.Rptr.3d 290].) The OAL’s (Office of Administrative Law) determination that a particular administrative guideline is an underground regulation is not binding on the courts, but it is entitled to deference. (Grier v. Kizer (1990) 219 Cal.App.3d 422, 435 [268 Cal.Rptr. 244], disapproved on other grounds in Tidewater Marine Western, Inc. v. Bradshaw (1996) 14 Cal.4th 557, 577 [59 Cal.Rptr.2d 186, 927 P.2d 296].)
*670In 2008, the OAL received a petition challenging a number of provisions in the “ ‘Clinical Evaluator Handbook and Standardized Assessment Protocol (2007)’ ” developed by the DMH pursuant to section 6601, subdivision (c). (Medina, supra, 171 Cal.App.4th at p. 814.) The OAL subsequently found the challenged provisions to be invalid underground regulations. “Although the OAL specifically restricted its inquiry to 10 provisions within the protocol [citation], its decision effectively invalidates the operative content of the protocol.” (Ibid.)
C. Ronje Determines the Protocol Is Invalid
Ronje, which appears to be the only published decision to reach the merits of the question, held the OAL correctly determined the assessment protocol was an invalid underground regulation. (Ronje, supra, 179 Cal.App.4th at pp. 516-517.) Although the People urge us to disregard Ronje on this point, the issue was not raised below, and indeed the parties and the superior court apparently proceeded on the assumption the 2007 assessment protocol was invalid. Given that the 2007 protocol has been superseded and that we would reach the same result regardless of its administrative validity or invalidity, we shall also proceed on the assumption the protocol was invalid.
The use of an administratively invalid assessment protocol, however, does not affect the superior court’s fundamental jurisdiction over an SVP proceeding. (Ronje, supra, 179 Cal.App.4th at p. 518; Medina, supra, 171 Cal.App.4th at pp. 815-817.) Dismissal is not the appropriate remedy in pending cases initiated after evaluations prepared using a non-APA-compliant protocol. (Ronje, supra, 179 Cal.App.4th at p. 518.) Instead, the Court of Appeal in Ronje directed the superior court to “(1) order new evaluations [under section 6601] using a valid assessment protocol, and (2) conduct another probable cause hearing under section 6602, subdivision (a) based on those new evaluations.” (Ronje, supra, 179 Cal.App.4th at p. 519.)
Once again the People question Ronje. The People suggest the remedy fashioned by the Ronje court “goes too far,” as the mere failure to comply with the APA did not compromise the reliability of the expert opinions that supported filing the petition in the first instance. Instead, the proper remedy, according to the People, was simply to direct the adoption of a protocol pursuant to the APA.
Given the “collateral” nature of the evaluation process (People v. Superior Court (Preciado) (2001) 87 Cal.App.4th 1122, 1130 [105 Cal.Rptr.2d 159] (Preciado)), the People’s suggestion cannot be lightly dismissed. We believe, however, it is unnecessary for us to reconsider the remedy our colleagues in the Fourth District settled upon. Applying the Ronje remedy to the facts of *671the instant case, we reach the same conclusion as the trial court: The split of opinions in the new evaluations does not require dismissal of the SVP petition.
D. Application of Ronje to the Facts of This Case
Davenport believes dismissal is the appropriate remedy in his case, notwithstanding Ronje’s explicit rejection of that remedy, because of the split of opinions in the more recent evaluations. Davenport argues the Ronje court “clearly and unequivocally” held the government was required “to go back to the beginning” and assess him in accordance with section 6601 with a legally valid protocol. In other words, according to Davenport, the Ronje court’s order to perform new evaluations indicated the “SVP evaluation process was to begin anew.”
We start by observing that if by beginning anew Davenport means the SVP evaluation process starts again, as if the commitment petition had never been filed, the Ronje court could have easily reached that result by directing the superior court to dismiss the commitment petition. Ronje, however, did not reach that result because the flaw in the protocol had no effect on the fundamental jurisdiction of the court: “The trial court has the power to hear the petition notwithstanding the error in using the invalid assessment protocol. Dismissal therefore is not the appropriate remedy.” (Ronje, supra, 179 Cal.App.4th at p. 518.)
The initial evaluations in Ronje and in the instant case had served their purpose by the time the Director of the DMH forwarded the request to file a commitment petition. In fact, although the SVP Act requires the director to send the two evaluations to the county’s designated counsel (§ 6601, subd. (d)), the act does not require that the evaluations be alleged or appended to the commitment petition. (Preciado, supra, 87 Cal.App.4th at p. 1128.) After the petition has been filed, the People’s burden is not to prove two evaluations exist, but to prove the alleged SVP is a person likely to engage in sexually violent predatory criminal behavior. (Ibid.) That burden may be carried by presenting the testimony of one or more experts at trial on behalf of the People. (See People v. Scott, supra, 100 Cal.App.4th at p. 1063 [two concurring experts are a procedural prerequisite to commence the petition process but the SVP Act does not expressly require two experts to testify at trial on behalf of the People].)
The new evaluations prepared in this case pursuant to Ronje are comparable to updated or replacement evaluations authorized by section 6603, subdivision (c). That subdivision permits the county’s attorney to request an updated evaluation if necessary to prepare the case for commitment, or a *672replacement evaluation if an original evaluator is no longer available to testify. Although section 6603, subdivision (c), anticipates the possibility of a split of opinion in updated or replacement evaluations, it does not explicitly state the consequences of such a split. The answer, however, is found in Gray v. Superior Court (2002) 95 Cal.App.4th 322 [115 Cal.Rptr.2d 477] (Gray).
In Gray, the petition for SVP commitment was supported by two concurring evaluations, but later evaluations performed pursuant to section 6603, subdivision (c), resulted in splits of opinion among a number of different evaluators. The alleged SVP therefore argued the subsequent splits of opinion had undermined the “foundational underpinnings” of the petition. (Gray, supra, 95 Cal.App.4th at p. 324.) When the trial court disagreed and allowed the case to proceed, the alleged SVP filed a writ petition in the Court of Appeal. (Id. at p. 325.)
The Court of Appeal concluded it was unlikely the Legislature intended for dismissal when postpetition SVP evaluations resulted in a split of opinion. (Gray, supra, 95 Cal.App.4th at p. 328.) “[W]e think it more likely that the required new evaluations are intended for informational and evidentiary purposes.” (Ibid.) The appellate court agreed the case should proceed: “Once a petition under the Act has been filed, and the trial court (as here) has found probable cause to exist, the matter should proceed to trial. In other words, once a petition has been properly filed and the court has obtained jurisdiction, the question of whether a person is a sexually violent predator should be left to the trier of fact unless the prosecuting attorney is satisfied that proceedings should be abandoned.” (Id. at p. 329.)
A similar result followed in People v. Superior Court (Salter) (2011) 192 Cal.App.4th 1352 [121 Cal.Rptr.3d 873] (Salter), a recent case involving a mentally disordered offender (MDO). (See Pen. Code, § 2960 et seq.; MDO Act.) The purpose of the MDO Act is the same as the SVP Act: to protect the public from dangerous felony offenders with mental disorders and to provide mental health treatment for their disorders. (People v. McKee (2010) 47 Cal.4th 1172, 1203 [104 Cal.Rptr.3d 427, 223 P.3d 566].)
The prisoner in Salter was identified as an MDO and sent to a state hospital for treatment. (Salter, supra, 192 Cal.App.4th at p. 1355.) Pursuant to Penal Code section 2970, the hospital’s medical director later asked the district attorney to file a petition to extend the prisoner’s commitment. After the petition had been filed, the prisoner was reevaluated by two experts, who reached opposite conclusions as to whether the prisoner qualified for involuntary treatment under the MDO Act. (192 Cal.App.4th at p. 1356.) An acting medical director at the hospital reported the results to the superior court and *673recommended against civil commitment. The superior court granted the prisoner’s motion to dismiss. (Ibid.)
The Court of Appeal reversed, finding the petition was viable when filed and that when there are conflicting medical opinions, the People are entitled to a jury trial to resolve the conflict in the evidence. (Salter, supra, 192 Cal.App.4th at pp. 1357-1358.) The court could see “no reason in logic or law to allow either party to deprive the other of the right to a jury trial that is allowed in the MDO legislation.” (Id. at p. 1359.)
The reasoning of Gray and Salter is persuasive. The original petition was properly filed and the superior court obtained jurisdiction. Pursuant to Ronje new evaluations were ordered. The effect of that order was not to begin the proceedings anew. Instead the matter may properly proceed to a new probable cause hearing as ordered by the trial court and, if probable cause is found, to trial.
It is important to keep in mind that neither Ronje nor the OAL evaluated the substance or the reliability of the 2007 assessment protocol. (See Ronje, supra, 179 Cal.App.4th at pp. 515, 520.)5 Davenport likewise does not point to any substantive defect in the 2007 protocol, or any prior protocol, and there is no evidence here that the use of a procedurally invalid protocol had a material effect on the conclusions in the original evaluations.6
Given the substantial risk of serious harm that could result from releasing a potential SVP to the public, dismissal is a drastic step. SVP evaluations may change over time for reasons other than that an individual no longer qualifies as an SVP. For example in Gray, the alleged SVP refused to be interviewed after the original set of evaluations, almost certainly rendering the later evaluations less precise. (Gray, supra, 95 Cal.App.4th at p. 330.) Given the procedural safeguards in place—a probable cause hearing, a jury trial, a unanimous verdict—there is no need to dismiss the commitment petition and start the SVP evaluation process from the beginning in this case.
*674III. DISPOSITION
The order to show cause is discharged. The petition for writ of mandate or prohibition is denied.
Ruvolo, R J., and Reardon, J., concurred.
Petitioner’s petition for review by the Supreme Court was denied March 28, 2012, S200079. Kennard, J., and Werdegar, J., were of the opinion that the petition should be granted.
The background, essentially the procedural history of the case, is taken from the original pleading (writ petition). The People have not disputed the facts as alleged in the petition.
All further statutory references are to the Welfare and Institutions Code unless otherwise noted.
“The State Department of Mental Health shall evaluate the person in accordance with a standardized assessment protocol, developed and updated by the State Department of Mental Health, to determine whether the person is a sexually violent predator as defined in this article.” (§ 6601, subd. (c).)
At the People’s request, we have taken judicial notice of both an earlier protocol (2007) and an undated “Standardized Assessment Protocol for Sexually Violent Predator Evaluations,” which according to the Attorney General is currently in use after compliance with the APA. (See Cal. Code Regs., tit. 9, §§ 4000, 4005.)
“The 2008 OAL Determination No. 19 concerned only whether the assessment protocol constituted a regulation under Government Code section 11342.600 and stated, ‘[n]othing in this analysis evaluates the advisability or the wisdom of the underlying action or enactment.’ [Citation.] The 2008 OAL Determination No. 19 advised that the OAL ‘has neither the legal authority nor the technical expertise to evaluate the underlying policy issues involved in the subject of this determination.’ ” (Ronje, supra, 179 Cal.App.4th at p. 515.)
The 2007 protocol postdates the initial evaluations of Davenport in 2006. (See Ronje, supra, 179 Cal.App.4th at p. 516.) No one, however, has indicated that any earlier protocol either was adopted in compliance with the APA or was defective in any substantive way.