¶ 1.
After a jury convicted him of second-degree sexual assault, Frederick Robertson discovered that the prosecutrix had been treated for depression with psychotic features around the time of the incident. Robertson now alleges that this newly discovered evidence could have been material to the jury determination of his guilt. The newly discovered evidence test is comprised of five factors. This appeal concerns how to implement the third and fifth newly discovered evidence factors — the materiality and different outcome at trial determinations — when a defendant seeks a postconviction in camera review of the prosecutrix's mental health records. We hold that with respect to the "materiality" factor the court should *353apply the Shiffra-Green1 materiality test. We also hold that if the court determines that the defendant has satisfied the first four newly discovered evidence factors and the defendant is thus entitled to an in camera review, the court should then apply the O'Brien2 "consequential evidence" test to determine whether to release the privileged records to the defendant. Finally, we conclude that Robertson is entitled to an in camera inspection of the alleged victim's psychiatric records. We therefore reverse the order denying the postconviction motion for the trial court to conduct an in camera review of the psychiatric records of the prosecutrix and remand for the in camera inspection.
¶ 2. In November 2000, Robertson and E.B. met each other for the first time shortly before a party that took place at the home of one of E.B.'s friends. At the party, the two talked and flirted until E.B. suggested that they proceed outside to a van that Robertson had borrowed. Once in the van, the two began kissing. After a short time, they noticed that other partygoers also were outside and Robertson suggested they drive to a different location around the corner for more privacy. Robertson instead drove to the parking lot of a different apartment complex and, once there, the two resumed kissing and engaged in oral sex. It is at this point that the two stories diverge and the alleged sexual assault occurred.
¶ 3. Robertson testified that at this point he then got on top of E.B. and that she did not resist his engaging in intercourse and she did not tell him "no" or *354otherwise indicate that she did not consent. E.B., however, testified to the contrary. She testified that she told him "no" several times before the intercourse began and that she screamed loudly and tried to stop Robertson. She testified that to stop her from screaming, Robertson put his hand over her mouth and told her that if she stopped screaming he would get off of her. She testified that he got off of her after she stopped screaming.
¶ 4. , Both Robertson and E.B. testified that after the sexual intercourse ended, E.B. ran from the van without putting her underwear and pants back on. After getting himself dressed, Robertson left the van and brought E.B. her clothes and told her that he was returning to the party. E.B. took her clothes, put them back on and ran to a nearby apartment where she told the residents that she had been assaulted and the police were then called. Robertson was later charged with second-degree sexual assault. -
¶ 5. At trial, the issue became whether E.B. had consented to the act of sexual intercourse and thus E.B.'s act of running from the van figured prominently in the closing arguments of both parties. In the first sentence of its closing argument, the State asserted that "this case is about credibility." The State then argued that the act of running from the van bolstered E.B.'s story that she was forced to engage in intercourse. The State rhetorically asked: "Is that somebody who is acting like they just engaged in consensual sex? I don't think so, and I think your experience in the everyday affairs of life will tell you that. That's consistent with somebody who was forced and that she did not want to do this."
¶ 6. Robertson's lawyer also addressed the act of running from the van, arguing that
*355for whatever reason — and I don't pretend to be able to explain this. For whatever reason, after vaginal intercourse was completed, including [Robertson] ejaculating, she freaked out. I don't know why she freaked out, but she certainly freaked. She jumped up; she ran from the van. Why did she freak? Why did she run? She just had sex with somebody she didn't know, somebody she just met that night. Maybe the reality that he had ejaculated and the possibility of pregnancy hit her, but certainly she freaked.
¶ 7. In its rebuttal, the State responded that Robertson had not.offered an explanation as to why E.B. ran from the van and that there was no evidence —but only speculation — to support the defense argument that she "freaked." The State then contended, "I can speculate too. I think she freaked. I think she freaked when she had sexual intercourse by force without consent. That's what freaked her out and that's why she acted the way she did."
¶ 8. The jury found Robertson guilty of second-degree sexual assault. In December 2001, Robertson filed a postconviction motion seeking an amendment of the judgment of conviction. Robertson requested that the court vacate the order leaving restitution to be determined. The court granted the motion and ordered the State to provide a proposed restitution amount to Robertson.
¶ 9. In the documentation supporting the proposed restitution amount, the State included a letter from E.B.'s treating psychiatrist. The letter stated that the psychiatrist had been seeing E.B. since December 1999 for "clinical depression with psychotic features." The letter further stated that "[s]he had an exacerbation of her clinical depression in the fall of 2000" and *356that the "rape happened in the midst of this exacerbation which intensified the clinical depression."
¶ 10. Based on this information, Robertson filed a motion for postconviction discovery, asking the trial court to conduct an in camera review of the records of E.B.'s psychiatric treatment, inform Robertson of E.B.'s doctors' conclusions and release to him information consequential to the case. The trial court denied Robertson's motion for postconviction discovery.3 The court agreed that the jury could have viewed E.B.'s leaving the van as an indication that she was trying to escape from whatever had occurred in the van. However, the court concluded that it was not sure that the information from E.B's psychiatric records would shed any further light on her flight from the van. Robertson appeals from the judgment of conviction and the post-conviction order.
¶ 11. Our first task is to examine whether Robertson met his burden to compel the trial court to conduct an in camera review of E.B.'s psychiatric records. This question necessarily involves a determi*357nation of the appropriate test to be applied when a defendant seeks an in camera inspection of psychiatric records in a postconviction setting. The parties agree that the principles articulated in State v. Shiffra, 175 Wis. 2d 600, 499 N.W.2d 719 (Ct. App. 1993), State v. Green, 2002 WI 68, 253 Wis. 2d 356, 646 N.W.2d 298, and State v. O'Brien, 223 Wis. 2d 303, 588 N.W.2d 8 (1999), govern the disposition of this case. The thrust of their dispute lies in how to reconcile those principles with the five-factor newly discovered evidence test this court applied in State v. Behnke, 203 Wis. 2d 43, 53-54, 553 N.W.2d 265 (Ct. App. 1996). We: therefore begin our analysis with a discussion of thosáí four cases.
¶ 12. In Shiffra, this court addressed the issue of whether a defendant can gain access prior to trial to a complaining witness's psychiatric history and counseling records. Shiffra, 175 Wis. 2d at 605. We began our analysis of the issue by acknowledging the competing rights and interests involved when a defendant seeks discovery of confidential records. Id. On the one hand, a criminal defendant's right to due process, in particular, the right to a meaningful opportunity to present a complete defense, is implicated. Id. On the other hand, the State has an interest in protecting a patient's privileged records from being disclosed. Id. We concluded that an in camera review of the privileged records achieved the proper balance between the competing rights and interests of the State and the defendant. Id.
¶ 13. We next established that to be entitled to an in camera inspection of privileged records before and during trial, the defendant must make a preliminary showing that the sought-after evidence is material to his or her defense. Id. We explained that the posttrial materiality test set forth in Pennsylvania v. Ritchie, 480 *358U.S. 39, 57 (1987), i.e., asking whether the evidence would have had an effect on the outcome of the trial, was too difficult to apply before trial. Green, 2002 WI 68 at ¶ 32; see also Shiffra, 175 Wis. 2d at 607-08. Rather, we concluded that to compel an in camera review, the defendant must demonstrate that the evidence is relevant and may be helpful to the defense or is necessary to a fair determination of guilt or innocence. Shiffra, 175 Wis. 2d at 608.
¶ 14. Recently, in Green, another case involving a request for an in camera review of counseling records, our supreme court affirmed our reasoning in Shiffra and clarified the threshold the defendant must satisfy to be entitled to an in camera review. Green, 2002 WI 68 at ¶ 34. Green instructs that a defendant must set forth a specific factual basis demonstrating a reasonable likelihood that the records contain relevant information that is necessary to a determination of guilt or innocence and that is not merely cumulative to other evidence available to the defendant. Id. Information is necessary to a determination of guilt or innocence if it tends to create a reasonable doubt that might not otherwise exist. Id. This test, pertaining to access to privileged mental health records, has been applied only in the context of a pretrial request for an in camera inspection.
¶ 15. In O'Brien, a case predating Green, our supreme court concluded that the rationale underpinning Shiffra applied with equal force in the postconviction setting where a defendant sought to test physical evidence. See O'Brien, 223 Wis. 2d at 320. There, the defendant, after being convicted of sexual assault, sought to test blood and semen samples and anal swabs and smears, arguing that this physical evidence could support his defense that the victim had consented to *359fellatio intercourse and could support his denial of anal intercourse. Id. at 313. The court explained that it was well established that under the Due Process Clause, criminal defendants must be given a meaningful opportunity to present a complete defense and that the court had previously recognized the right of a defendant to use postconviction discovery when the evaluation is of evidence that is "critical, relevant, and material." Id. at 320 (citation omitted). The court then held that a criminal defendant has a right to postconviction discovery of physical evidence when the sought-after evidence would be relevant to an issue of consequence,4 but that the remedy only should be extended to a case where the evidence would create a reasonable probability of a different outcome at trial. Id. at 323. Although O'Brien involved a postconviction discovery motion, the defendant sought discovery of physical evidence and the supreme court did not address whether the test set forth was to be applied by the trial court when it was evaluating a request for an in camera review of privileged mental health records.
¶ 16. In Behnke, the only published case we are aware of that involves a postconviction motion for an in camera review of privileged records, we analyzed the issue of whether a defendant is entitled to an in camera inspection using the five factors courts consider when deciding whether to grant a new trial in the interests of justice based on newly discovered evidence. Behnke, 203 Wis. 2d at 53-54. The five factors courts consider in a newly discovered evidence determination are: (1) the *360evidence must have come to the moving party's knowledge after trial, (2) the party must not have been negligent in seeking to discover it, (3) the evidence must be material, (4) the evidence must not be cumulative, and (5) it must be reasonably probable that a different result would be reached on a new trial. Id. We concluded that while the defendant had met the first, second and fourth factors, the third factor, the materiality determination, required him to satisfy the Shiffra materiality test and he had failed to do so. Behnke, 203 Wis. 2d at 54. Consequently, we did not even reach the question of how to implement the fifth factor in a postconviction in camera request.
¶ 17. Here, Robertson contends that the Shiffra-Green preliminary materiality test, asking whether there is a reasonable likelihood that the evidence is relevant and necessary to a determination of guilt or innocence, applies when a defendant seeks a postcon-viction in camera review and that if the in camera review is granted, then the trial court applies O'Brien's "consequential evidence" test to determine whether to release the records to the defendant. The State, on the other hand, seems to assert that Behnke dictates that both the Shiffra-Green materiality test and the O'Brien "consequential evidence" test, which resembles the fifth newly discovered evidence factor, govern when a defendant seeks a postconviction in camera review, and thus to even be entitled to an in camera review of confidential records, a defendant must demonstrate a reasonable likelihood that the records contain relevant information that would have changed the outcome of the trial. The State's argument, however, is inconsistent with the United States Supreme Court's teachings in Ritchie and ignores the dichotomy that exists in Wisconsin law between the threshold a defendant must *361satisfy to be entitled to an in camera review and the test the trial court applies when it reviews the materials at issue in its chambers.
¶ 18. In Ritchie, the defendant was charged with sexual assault of his minor daughter. Ritchie, 480 U.S. at 43. Before trial, the defendant subpoenaed records from the protective service agency investigating the daughter's allegations. Id. The agency refused to release the records on the grounds that they were privileged and the trial court denied the defendant's motion to enforce the subpoena. Id. at 43-44. On appeal, the defendant claimed that the agency's refusal to turn over the records and the trial court's denial of his motion for disclosure thwarted his right to confront his accuser. Id. at 45.
¶ 19. The Supreme Court held that the defendant was entitled to an in camera review of the agency files. Id. at 58. The Court reasoned that criminal defendants have a right to put evidence before the jury that might influence the determination of guilt; and the government has an obligation to turn over evidence that is both favorable to the accused and material to guilt or punishment. Id. at 57. The Court concluded that evidence in the files would be material "if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different." Id. (citation omitted). The Court also rejected the State's argument that to be entitled to an in camera review the defendant had to make a particularized showing of what information he or she was seeking or how it would be material. See id. at 58 n.15. The Court seems to instruct that the test for the preliminary showing to get the in camera review is somehow less burdensome than the test used for determining whether evidence is material during the in *362camera review. See id. According to the Court, the defendant is only required to make a plausible showing that the information exists and that it would be material and favorable to his or her defense. Id. The Court then remanded the case to the trial court for an in camera review to determine whether the daughter's file contained information that probably would have changed the outcome of the trial. Id. at 61.
¶ 20. We emphasize that the defendant in this case, like the defendants in Green and Shiffra, is trying to make a preliminary showing to compel an in camera review by the trial court. As such, the defendant's preliminary burden for seeking in camera review must be less stringent than the test applied by the court during its in camera inspection. See Ritchie, 480 U.S. at 58 n.15.
¶ 21. We also observe that in O'Brien, the defendant sought to test physical evidence. The potential outcomes of scientific testing of physical evidence, like the testing sought in O'Brien, are relatively limited and easier for the defendant to predict and thus it is easier to demonstrate materiality. Given the confidential nature of mental health records, it makes it very difficult in most cases for the defendant to predict what evidence will be found during a review of the records and thus how it may be material to the case. See Green, 2002 WI 68 at ¶ 35.
¶ 22. We therefore conclude that the O'Brien "consequential evidence" test should not be used to decide whether to conduct an in camera review. A defendant requesting confidential records during postconviction discovery should be required to meet the preliminary Shiffra-Green burden. As with pretrial requests for *363confidential records, applying the Shiffra-Green test to the posttrial setting strikes the appropriate balance between the interests of the defendant and the State. It affords the defendant an opportunity to have a judge, an independent overseer, review the records while still protecting the privacy of the alleged victim. Then, if the defendant has shown an entitlement to an in camera review based upon the first four factors of the newly discovered evidence test, the trial court should apply the O'Brien "consequential evidence" test to determine whether the material it reviews during its in camera inspection should be disclosed to the defendant. We have every confidence in the trial courts to make a proper determination as to whether the disclosure of the information is necessary based on the competing interests involved in such cases. See Shiffra, 175 Wis. 2d at 611.
¶ 23. The State argues that this conclusion is inconsistent with our decision in Behnke. This is simply not the case. Like Behnke, here we have adopted the materiality test from Shiffra, as it has been clarified in Green, and applied it to the third newly discovered evidence factor. See Behnke, 203 Wis. 2d at 54. Additionally, as we have indicated, in Behnke, the defendant was engaging in what can be characterized as a "fishing expedition" and we did not reach the question of whether the fifth newly discovered evidence factor— whether it is reasonably probable that a different result would be reached on a new trial — should have been applied. See id. From our discussion of Green and Ritchie it is clear that the fifth newly discovered evidence factor, which is similar to the O'Brien "consequential evidence" test, normally does not come into play until the trial court is actually conducting the in *364camera inspection in its chambers.5 As we discussed, in Ritchie, the Court concluded that when deciding whether to release the records to the defendant, the trial court should determine whether the evidence was material, meaning the court should ask whether the result of the proceeding would have been different had the evidence been disclosed to the defendant, a test that mirrors the O'Brien test. Thus, contrary to the State's assertions, our approach not only is consistent with our decision in Behnke, but also will bring Wisconsin into alignment with the teachings of the Supreme Court in Ritchie.
¶ 24. The State next submits that if we adopt the two-pronged test and use the newly discovered evidence factors and Shiffra-Green in the first instance and O'Brien in the second instance, we must apply a deferential standard of review because this case involves a newly discovered evidence determination by the trial court. However, in a typical newly discovered evidence case, an appellate court will independently determine whether a denial of a new trial based on the new evidence violates the defendant's due process rights. State v. Coogan, 154 Wis. 2d 387, 394-95, 453 N.W.2d 186 (Ct. App. 1990). Whether the defendant submitted a preliminary evidentiary showing sufficient for an in camera review implicates a defendant's constitutional right to a fair trial and thus raises a question of law. Green, 2002 WI 68 at ¶ 20. We review questions of law de novo. Id.
*365¶ 25. Having set forth the appropriate test, we now apply it to the facts of this case. The parties here do not dispute that Robertson discovered the existence of the records after trial, he was not negligent in discovering the records and that the evidence in the records would not be cumulative to evidence presented at trial. Rather, the focus is on whether Robertson has satisfied his burden pursuant to the Shiffra-Green test.
¶ 26. As we have stated, to be entitled to an in camera review of confidential records, a defendant must set forth a specific factual basis demonstrating a reasonable likelihood that the records contain relevant information that is necessary to a determination of guilt or innocence and not merely cumulative to evidence already available to the defendant. Green, 2002 WI 68 at ¶ 34. Mere speculation or conjecture as to what information is in the records is not sufficient. Id. at ¶ 33. The Shiffra-Green test essentially requires the court to look at the existing evidence in light of the request for an in camera review and to determine "whether the records will likely contain evidence that is independently probative to the defense." Green, 2002 WI 68 at ¶ 34.
¶ 27. Robertson proffered the letter from E.B.'s doctor in support of his motion for postconviction discovery. In the letter, the doctor stated that E.B. had been diagnosed with depression approximately one year before the alleged sexual assault and that psychotic features had accompanied her depression. The doctor also noted in the letter that E.B. had an exacerbation before the alleged sexual assault. While we acknowledge that if a person has depression it does not automatically translate into an inability to accurately or truthfully perceive or relate events or otherwise make a *366person less credible, this letter indicates that the doctor had diagnosed E.B. with more than just depression. Though the letter does not describe the psychotic features, the standard text used in psychiatric diagnosis specifies that a person who has depression with psychotic features:
[i]ndicates the presence of either delusions or hallucinations (typically auditory).... Most commonly, the content of the delusions or hallucinations is consistent with the depressive themes .... Less commonly, the content of the hallucinations or delusions has no apparent relationship to depressive themes.
American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders 412 (4th ed. Text Revision 2000).6
¶ 28. As the State submitted in its closing arguments, "this case is about credibility." The State used E.B.'s act of running from the van to bolster her credibility and thus her testimony that she did not consent to sexual intercourse. The information in the records concerning E.B's psychiatric treatment and the nature of the psychotic features presented by her depression could explain her behavior in a way that was not possible to do during trial. Robertson could not offer any reason for why E.B. "freaked out" and ran from the van, thereby leaving only one plausible explanation for E.B.'s behavior — that she had just been forced into sexual intercourse. Providing an explanation could in turn rebut or weaken the commonsense *367explanation offered by the State and thus could affect E.B.'s credibility and lend credence to Robertson's defense of consent.
¶ 29. The State argues that the overall tenor of the doctor's letter shows that he was under the assumption that E.B. was "raped" and that he would not have used the word if he thought that E.B.'s behavior in fleeing the van half-clothed was the result of a delusion or hallucination rather than her response to having been sexually assaulted. This argument ignores the context in which the letter was written. The letter was written in response to a request for restitution amounts after Robertson had been convicted, making the word "rape" a possible reference to the legal determination that Robertson had committed the crime. Further, because Robertson did not learn of the existence of the diagnosis until the postconviction restitution proceedings, he did not have the opportunity to examine the doctor and ask what he meant or what was the basis for his conclusions.
¶ 30. The State also attempts to compare this case to Behnke. In Behnke, the defendant, who was convicted of sexual assault and battery for causing injury to the victim's eye, face and chest, learned after trial that prior to the attack the alleged victim had received treatment because of psychiatric trauma and that she cut or bruised her arm, using pain, to stop her fear. Behnke, 203 Wis. 2d at 48, 52-54. We refused to conduct a posttrial in camera review of the records because we were troubled by what we labeled the "spread effect theory" — that if a person is acting out in a particular fashion by abusing oneself in a certain way, it is enough of a probability that he or she is abusing himself or herself in other ways too — thus justifying a look at his or her mental health records to be sure. Id. at 54-55. *368Here, the State argues that like Behnke, the possibility that E.B. was suffering from some type of delusion or hallucination when she fled from the van was simply too attenuated from the information contained in the letter to justify an in camera review. However, unlike Behnke, where the defendant had no basis for asserting that the victim would injure herself anywhere but her arm, Robertson's assertion that E.B. could have suffered from delusions or hallucinations that contributed to her behavior is supported by her doctor's diagnosis and medical literature.
¶ 31. We therefore conclude that Robertson has met his burden of making a preliminary showing that there is a reasonable likelihood that the records contain relevant information necessary to a determination of guilt or innocence. As the State itself observed, the case came down to credibility and the trial court concluded that the jury could have viewed E.B.'s leaving the van as an indication that she was trying to escape from whatever had occurred in the van. Robertson has presented evidence demonstrating that E.B.'s psychiatric difficulties might affect both her ability to accurately perceive events and her ability to relate the truth. These problems are relevant and necessary to a determination of guilt or innocence because they bear directly on both E.B.'s credibility and Robertson's defense of consent. Thus, Robertson is entitled to an in camera inspection of the mental health records.
¶ 32. The State next argues that if we conclude Robertson is entitled to a posttrial in camera review of E.B.'s records, then we should certify the case to the supreme court to determine the remedy if E.B. on remand refuses to waive her privilege and permit an in *369camera review. The State's argument is premature. E.B. has not yet refused to waive her privilege in the records and thus the issue is not yet properly before this court. Courts act only to determine actual controversies — not to announce principles of law or to render purely advisory opinions. See State ex rel. Ellenburg v. Gagnon, 76 Wis. 2d 532, 535, 251 N.W.2d 773 (1977). We therefore decline to certify the issue to the supreme court.
By the Court. — Judgment affirmed; order reversed and cause remanded with directions.