On Appeal from the United States District Court for the Western District of New York Petitioner Donald J. Englert, II, who stands convicted in New York of engaging in a course of sexual conduct against a child in the first degree, see N.Y. Penal Law §130.75(1)(a), appeals from a judgment of the United States District Court for the Western District of New York (Siragusa, J.) denying him a writ of habeas corpus. See 28 U.S.C. §2254. In seeking such relief, Englert raises the same Sixth Amendment challenge to conviction that he unsuccessfully argued to state courts, i.e., that his trial counsel was constitutionally ineffective in failing to consult or call a medical expert to challenge the testimony of a government expert that the victim-child’s normal physical examination was consistent with past sexual abuse. Under the deferential standard of review applicable to such a Sixth Amendment claim, particularly when raised in a §2254 proceeding, we conclude that Englert fails to show that the state court unreasonably applied controlling Supreme Court precedent, i.e., Strickland v. Washington, 466 U.S. 668 (1984), in rejecting Englert’s ineffective-assistance claim. This court’s decision in Gersten v. Senkowski, 426 F.3d 588 (2d Cir. 2005) (holding state court to have unreasonably applied Strickland in rejecting ineffective-assistance challenge to counsel who, without investigation, conceded that physical evidence demonstrated sexual assault of child when qualified medical experts would have testified otherwise), compels no different conclusion because, as we there stated, and as controlling Supreme Court precedent now makes clear, “no per se rule” dictates that “expert consultation is always necessary in order to provide effective assistance of counsel in child sexual abuse cases,” id. at 609 (internal quotation marks omitted); see Harrington v. Richter, 562 U.S. 86, 111 (2011). In the particular circumstances of this case — where defense counsel, among other things, elicited a concession from the prosecution expert that the child’s normal physical examination was as consistent with a lack of abuse as with the alleged abuse — the state court did not unreasonably apply Strickland in finding that Englert was not prejudiced by his attorney’s failure to consult or call a medical expert and, thus, not denied effective assistance of counsel. See id. (recognizing that “[i]n many instances cross-examination will be sufficient to expose defects in an expert’s presentation”). AFFIRMED. REENE RAGGI, C.J. In 2013, petitioner Donald J. Englert, II, was convicted after a jury trial in Monroe County, New York, of engaging in a course of sexual conduct against a child in the first degree. See N.Y. Penal Law §130.75(1)(a). In unsuccessfully challenging that conviction in state court both on direct appeal and on collateral attack, Englert argued that he was denied effective assistance of counsel because his trial attorney failed to consult with or call an expert witness to rebut the prosecution’s medical expert. That expert testified that she had examined the child-victim and found no physical evidence of sexual abuse. Nevertheless, she opined that such normal findings were consistent with the child’s account of abuse ending more than six months earlier because any injuries would have had time to heal. When Englert made the same Sixth Amendment argument in petitioning for federal habeas relief from conviction pursuant to 28 U.S.C. §2254, the United States District Court for the Western District of New York (Charles J. Siragusa, Judge) denied his petition, ruling that the state courts had not unreasonably applied controlling Supreme Court precedent in holding Englert not to have been denied effective assistance of counsel. See Englert v. Colvin, No. 18-CV-6871 (CJS), 2022 WL 3214774 (W.D.N.Y. Aug. 9, 2022). Englert now appeals from that judgment. On de novo review, see Jordan v. Lamanna, 33 F.4th 144, 150 (2d Cir. 2022), we conclude that Englert’s habeas petition was properly denied because he fails to demonstrate, as required by the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”), see 28 U.S.C. §2254(d), that the state courts’ rejection of his ineffective-assistance claim was based on an unreasonable application of clearly established federal law, specifically Strickland v. Washington, 466 U.S. 668 (1984). In Strickland, the Supreme Court clearly established that a defendant raising an ineffective-assistance-of-counsel challenge to conviction must show both that “(1) counsel’s performance was objectively deficient,” and that “(2) petitioner was actually prejudiced as a result.” Harrington v. United States, 689 F.3d 124, 129 (2d Cir. 2012) (citing Strickland v. Washington, 466 U.S. at 687-88). New York courts did not unreasonably apply this law in concluding that Englert failed to make this showing. This court’s decision in Gersten v. Senkowski, 426 F.3d 588 (2d Cir. 2005) (holding state court to have unreasonably applied Strickland in rejecting ineffective-assistance challenge to counsel who, without investigation, conceded that physical evidence demonstrated sexual assault of child when qualified medical experts would have testified otherwise), compels no different conclusion because, as we there stated, and as controlling Supreme Court precedent now makes clear, “no per se rule” dictates that “expert consultation is always necessary in order to provide effective assistance of counsel in child sexual abuse cases,” id. at 609 (internal quotation marks omitted); see Harrington v. Richter, 562 U.S. 86, 111 (2011). In the particular circumstances of this case — where defense counsel, among other things, elicited a concession from the prosecution expert that the child’s normal physical examination was as consistent with a lack of abuse as with the alleged abuse — the state court did not unreasonably apply Strickland in finding that Englert was not prejudiced by his attorney’s failure to consult or call a medical expert and, thus, not denied effective assistance of counsel. See id. (recognizing that “[i]n many instances cross-examination will be sufficient to expose defects in an expert’s presentation”). Accordingly, we affirm the challenged judgment. BACKGROUND I. Trial A. The Prosecution Case In January 2013, Englert was tried in New York State court on an indictment alleging that, between 2005 and 2010, he engaged in a course of sexual conduct in the first degree with N.L., the pre-adolescent daughter of Englert’s then-girlfriend. See N.Y. Penal Law §130.75(1)(a).1 Trial evidence indicated that, in 2003, when N.L. was two-years old, the child and her mother began living with Englert. In 2005, N.L.’s mother gave birth to a son by Englert. The following year, the couple separated, apparently amicably because Englert continued to help care for his son and N.L., both at their mother’s home and at his own home, the latter visits frequently spanning weekends.2 Such visits continued until Englert moved to California in 2010. 1. Victim Testimony To prove the charged abuse, the prosecution relied primarily on the testimony of N.L., by then 12-years old. N.L. recounted that, when she was four or five years old and still living with Englert, he began touching her breasts, buttocks, and vagina over her clothes, but then started touching her vagina with his finger under her clothes. After Englert and the child’s mother separated, he continued to engage in such abusive touching on occasions when N.L.’s mother left her and her half-brother in Englert’s care. N.L. testified that when she was eight or nine, the abuse escalated, with Englert trying to insert his penis into her vagina and, when that hurt, to insert it into her anus, or failing that, to rub it against her buttocks. By the time N.L. was ten, on weekend visits to Englert’s home, he would have her sleep in his bed with him where he would “usually” insert his penis into her vagina. Confidential App’x 1103-04. Also, he sometimes had her shower with him, and there pick her up and put his penis in her vagina. He also put cherry-flavored lubricant on his penis and inserted it into N.L.’s mouth. At other times, Englert placed his mouth on N.L.’s vagina. On more than one occasion, Englert showed N.L. pornographic videos on his computer before inserting his penis into her vagina or anus. N.L. testified that when she asked Englert not to touch her, he would question her love for him. Englert warned N.L. not to tell anyone about his conduct lest he get in trouble. N.L. stated that she did not then tell anyone because she was afraid. Englert continued sexually to abuse N.L. until he moved to California in November 2010. In May 2012, when N.L. was 11, she revealed Englert’s sexual abuse to a school friend who encouraged N.L. to tell her mother, which she then did. Thereafter, on May 24, 2012, N.L. was interviewed by a forensic examiner at the Justice for Children’s Advocacy Center in Batavia, New York, and then, on June 4, 2012, physically examined by a pediatric nurse practitioner. On cross-examination, N.L. testified that she was doing well in school, had many friends, and had a happy childhood. She also testified that, after Englert moved to California, she and her family were upset that he wanted his son to visit him there. Englert’s request was the subject of state court proceedings for some time before N.L. disclosed Englert’s sexual abuse to her mother. Although N.L. was not aware of those proceedings, she acknowledged knowing that, after making her abuse disclosure, her half-brother would not be visiting Englert in California. 2. Medical Testimony Cecilia Lyons, the nurse practitioner who physically examined N.L., testified to having conducted approximately 1,000 pediatric sexual abuse examinations over 14 years. In examining N.L., she used a colposcope, which can magnify the genital and rectal areas up to 10,000 times. Nurse Lyons reported that the results of her examination were “essentially normal,” with the edges of the child’s hymen appearing smooth, and with no lacerations or tears either to her hymen, genital, or anal area. Id. at 1474. Nurse Lyons opined that such normal findings were nevertheless consistent with N.L.’s report of sexual abuse more than six months earlier because the nurse “would have expected her injuries to have been healed” in the intervening time, id. at 1477-78, making it “not likely” that N.L. would have any visible injuries or scarring, id. at 1498. In so stating, Nurse Lyons referenced unspecified “research that shows how quickly hymens heal even when there is an injury, to the point where…they look normal like they did before they had the assault.” Id. at 1477. She testified that fewer than 3 percent of the children she had examined had disclosed sexual abuse within 72 hours of its occurrence, and she had found physical evidence of such abuse in fewer than 5 percent of cases. On cross-examination, counsel questioned Nurse Lyons about a litany of possible genital and rectal abnormalities and injuries that are indicia of sexual abuse. She acknowledged that none was evident in N.L.3 More significantly, she acknowledged that normal examination findings for N.L., which she opined were consistent with sexual abuse, were also consistent with an absence of sexual abuse. The nurse admitted that her consistent-with-abuse conclusion was informed in part by the child’s forensic interview and personal history. As for research indicating hymen regeneration or healing, Nurse Lyons conceded that she had not herself conducted such research and had never personally observed hymen regrowth.4 3. Relatives’ Testimony No prosecution witness testified to having seen Englert abuse N.L. Instead, her mother — who professed to having had no idea that her child was being abused during the charged five-year span — testified that in her own sexual encounters with Englert, he had favored anal and oral sex and had used a fruit-flavored lubricant for the latter. N.L.’s grandmother testified that on one occasion when she made an unannounced visit to Englert’s home, she saw then-five-year-old N.L. run up to Englert, who was then seated on a couch, pull a comforter off him, and grab at his boxer shorts. As the child did so, her grandmother saw that Englert appeared to have an erection. On cross-examination, the grandmother acknowledged making no mention of this incident before N.L. disclosed abuse by Englert because she was not sure what she had seen. Also on cross-examination, N.L.’s mother acknowledged that her daughter was a happy, well-adjusted child, who did well in school, had many friends, and had never received counseling for sexual abuse. N.L.’s mother testified that she thought her daughter would benefit from counseling and attributed her failure to receive it to the mother’s inability to find a female counselor who specialized in sexual abuse trauma. N.L.’s mother also acknowledged that in the year before her daughter disclosed any sexual abuse, the mother had been involved in “hostile” and “bitter” court proceedings with Englert over his desire to have his son visit him in California. Confidential App’x 1383, 1385. N.L.’s mother opposed such visitation, which, in fact, was not permitted after Englert was charged with abusing N.L. B. The Defense Case Testifying in his own defense, Englert denied engaging in any sexual conduct with N.L., insisting he viewed the child as his daughter. The defense’s theory of the case was that N.L. fabricated abuse allegations at her mother’s behest because of the couple’s dispute about their son visiting Englert in California. Englert testified that before he moved to California, he had enjoyed a good relationship with N.L.’s mother. But after he moved and the visitation issues arose, she told Englert that she was going to do whatever she could to prevent him from seeing his son. When Englert subsequently learned that New York authorities were investigating him for sexually abusing N.L., he returned to the state in August 2012 to attend a wedding and voluntarily went to the police in an effort to clear his name. Soon after, he was arrested. Englert’s brother and current girlfriend testified, the former reporting a positive relationship between Englert and N.L.; the latter stating that Englert’s relationship with N.L.’s mother grew hostile after Englert moved to California. The girlfriend also denied that Englert enjoyed anal sex, testifying that when he tried it once at her urging, he was repulsed. The defense called no medical expert. Instead, in addressing the jury, counsel emphasized the lack of physical evidence of abuse and the normal findings of N.L.’s genital and rectal examination. Counsel urged the jury to use common sense in considering this medical evidence, arguing that physical evidence of abuse would be expected if N.L. had, in fact, been sexually victimized over a number of years.5 Counsel also highlighted evidence showing that N.L. was happy, well-adjusted, and excelling in school, with no behavioral or mental health issues. He argued that this too was inconsistent with years of sexual abuse and reminded the jury that prosecution witness Perkowski had testified that almost all of the nearly 3,200 child sex abuse victims he had encountered had required counseling. The jury found Englert guilty, and on April 9, 2013, the state court sentenced him to 22 years’ imprisonment.6 II. State Direct Appeal With the assistance of new counsel, Englert appealed his conviction, arguing, inter alia, that he had been denied the effective assistance of counsel guaranteed by the New York and United States Constitutions because trial counsel failed to present expert medical testimony to explain the significance of the normal findings of N.L.’s physical examination and to counter Nurse Lyons’s testimony that these findings were consistent with sexual abuse. In affirming Englert’s conviction, the Appellate Division rejected this claim, ruling that “[t]he failure of defense counsel to obtain the testimony of an expert does not constitute ineffective assistance of counsel because defendant has not shown that such testimony was available, that it would have assisted the jury in its determination or that [defendant] was prejudiced by its absence.” People v. Englert, 130 A.D.3d 1532, 1533 (N.Y. App. Div. 4th Dep’t 2015) (internal quotation marks omitted). The New York Court of Appeals denied further review. See People v. Englert, 26 N.Y.3d 967 (2015); People v. Englert, 26 N.Y.3d 1144 (2016) (denying reconsideration). III. State Collateral Challenge to Conviction A. Englert’s Submission Proceeding pro se, Englert moved under N.Y. Crim. Proc. Law §440.10 to revoke his conviction, again arguing, inter alia, that trial counsel had been constitutionally ineffective in failing to consult with or call a medical expert.7 In support of this motion and his request for an evidentiary hearing, Englert submitted his own affidavit and seven almost identically worded affidavits from various relatives, each of whom recounted pre-trial conversations with defense counsel about consulting with or hiring experts to rebut prosecution experts and N.L.’s allegations of abuse. Englert and his relatives claimed that trial counsel told them that it was “not his responsibility” to consult or hire an expert because “[t]he burden of proof rested with the prosecution.” Confidential App’x 303-07, 309-10. One of Englert’s cousins stated that trial counsel told her that “he didn’t have sufficient time to locate and hire an expert before the trial began.” Id. at 308. Englert also submitted an affidavit from Dr. Jeffrey Bomze, a physician with experience in forensic pediatrics and child sexual abuse cases. Insofar as Nurse Lyons testified that N.L.’s normal examination findings were consistent with sexual abuse, Dr. Bomze did not reject the possibility. Indeed, he acknowledged that “the absence of any identifiable trauma does not rule out abuse,” id. at 319, and that Nurse Lyons “testified correctly that the large majority of sexually abused children have no positive findings on anogenital examination if the exam is delayed beyond the first several days or longer,” id. at 321. Instead, Dr. Bomze faulted Nurse Lyons for “concluding with certainty that sexual abuse occurred in this case” “without adequate evaluation,” as indicated by the “lack of detailed descriptions of the anogenital examinations” conducted and the failure to consider “[t]he full context of the child’s medical and psychosocial histories” or other possible explanations for examination results. Id.8, 9 As for Nurse Lyons’s testimony that hymens can heal even after injury from sexual abuse, Dr. Bomze agreed that the “majority” of hymenal lacerations heal, but characterized the nurse’s opinion as incomplete or confusing because “complete lacerations of the hymen” and “deep lacerations of the posterior hymen” may not fully heal. Id. at 321-22. Dr. Bomze stated that if he had been retained by the defense, he would have advised trial counsel to request N.L.’s complete medical history to determine if there were any documented medical or psychosocial issues after the alleged abuse given the delayed disclosure. He would have advised counsel to highlight for the jury the types of trauma associated with sexual assault that were not found in N.L.’s case, and he would have pointed counsel to (unspecified) studies indicating that residual findings of abuse are more frequently found in children subjected to anal abuse than in those subjected to vaginal abuse. As to Nurse Lyons’s testimony that a lack of physical evidence of abuse is consistent with abuse, Dr. Bomze stated he would have advised trial counsel, on cross-examination, to ask the nurse whether it “is also consistent with a child who has not been abused.” Id. at 325.10 He also would have advised counsel to call a medical expert to testify that Nurse Lyons’s examination of N.L. revealed no physical evidence of abuse, that the nurse erred in attributing “no significance” to that absence, and that her consistent-with-abuse conclusion was “very misleading for the jury” because it could be interpreted to mean that N.L. was “definitely abused.” Id. at 325-26. B. State Court Decision On June 4, 2018, the New York Supreme Court denied Englert’s §440.10 motion without holding a hearing. See N.Y. Crim. Proc. Law §440.30(1)(a), (4) (directing court to consider whether motion is determinable without hearing to resolve factual questions and listing grounds for denying motion without hearing). The court observed that Englert had unsuccessfully raised a similar ineffective-assistance challenge on direct appeal. Insofar as he now offered affidavits to support his collateral challenge, the court was not convinced. Noting that “nothing” in Dr. Bomze’s affidavit “contradict[s] Lyon[s]‘s key conclusions,” the court concluded that Englert was not prejudiced by trial counsel’s failure to consult or call him. Confidential App’x 380-81. The court explained that Dr. Bomze did not “set forth an opinion that it would not be possible for there to be a physical exam of the victim six months to one year after the last incident of abuse, and for it to show ‘normal’ findings despite the years and frequency of abuse.” Id. at 381 (emphasis in original). And, while recognizing that Dr. Bomze faulted the process by which Nurse Lyons conducted her physical examination of N.L., the state court observed that he did not seriously challenge “the substance” of her examination findings. Id. As for Englert and his relatives’ claims that trial counsel declined to consult an expert, the court deemed the veracity of those claims “questionable.” Id. at 385 n.15. Even crediting those affidavits, however, the court determined that counsel had not provided Englert with ineffective representation. Rather, the record showed trial counsel was “fully engaged, knowledgeable, and aggressive in his cross examination” of the prosecution’s witnesses and, specifically as to Nurse Lyons, counsel “posit[ed] to the jury the possibility that if the victim had been abused to the extent she testified, then the physical examination would have shown the evidence of injury.” Id. at 383-84. On November 15, 2018, the Appellate Division denied Englert leave to appeal the denial of his §440.10 motion. IV. Federal Habeas Proceedings On December 3, 2018, Englert, proceeding pro se, filed the instant habeas petition again arguing, inter alia, that his trial counsel was constitutionally ineffective. On August 9, 2020, the district court denied the petition. As to the ineffective-assistance claim, the district court held that the state courts did not unreasonably apply Strickland v. Washington, 466 U.S. 668, in rejecting the claim because, at the first step of Strickland analysis, the record showed trial counsel to have been “versed in the indicia of abuse,” to have questioned Nurse Lyons extensively about her observations of N.L.’s genital and rectal area, and to have “employed the reasonable strategy of demonstrating the absence of such indicia [of abuse] in N.L.’s case to imply that no such abuse occurred.” Englert v. Colvin, 2022 WL 3214774, at *9. The district court thus deemed it unnecessary to address the prejudice step of Strickland analysis. After Englert timely filed this appeal, this court appointed counsel and issued a Certificate of Appealability on the following questions: “[1] whether the district court erred in denying Englert’s claim that counsel was ineffective for not consulting or calling an expert medical witness and [2] whether the district court erred in denying Englert’s claim that counsel’s cross examination of the state’s medical expert was ineffective.” Certificate of Appealability, Englert v. Lowerre, No. 22-2016 (2d Cir. Mar. 15, 2023), Dkt. 31. In addressing these questions, appointed counsel treats the second as a subpart of the first, i.e., arguing that trial counsel’s cross-examination of Nurse Lyons was ineffective because he had failed to consult a medical expert. Thus, we treat the questions together in explaining why Englert’s §2254 petition was properly denied on the merits. DISCUSSION I. Standard of Review We review de novo the district court’s denial of Englert’s §2254 petition for relief from his state court conviction. See Jordan v. Lamanna, 33 F.4th at 150. Where, as here, a state court has rejected on the merits a claim thereafter raised in a §2254 petition, AEDPA permits a federal court to grant habeas relief only where the state court’s decision was “contrary to, or involved an unreasonable application of, clearly established Federal law,” 28 U.S.C. §2254(d)(1), or was “based on an unreasonable determination of the facts in light of the evidence presented,” id. §2254(d)(2). A state court decision is “contrary to” clearly established federal law “if the state court arrive[d] at a conclusion opposite to that reached by [the Supreme] Court on a question of law” or the state court “decides a case differently than [the Supreme] Court has on a set of materially indistinguishable facts.” Williams v. Taylor, 529 U.S. 362, 412-13 (2000). A state court decision involves an “unreasonable application” of clearly established federal law “if the state court identifies the correct governing legal principle from [the Supreme] Court’s decisions but unreasonably applies that principle to the facts of the prisoner’s case.” Id. at 413. This “highly deferential” standard of review is “difficult to meet” in that it “demands that state-court decisions be given the benefit of the doubt.” Cullen v. Pinholster, 563 U.S. 170, 181 (2011) (internal quotation marks omitted); accord Fischer v. Smith, 780 F.3d 556, 560 (2d Cir. 2015); see also Burt v. Titlow, 571 U.S. 12, 20 (2013) (stating that on §2254 review, federal courts “will not lightly conclude that a State’s criminal justice system has experienced the ‘extreme malfunctio[n]‘ for which federal habeas relief is the remedy” (quoting Harrington v. Richter, 562 U.S. at 102)). Thus, under AEDPA, the determinative question “is not whether a federal court believes the state court’s determination was incorrect but whether that determination was unreasonable — a substantially higher threshold.” Schriro v. Landrigan, 550 U.S. 465, 473 (2007); see McCray v. Capra, 45 F.4th 634, 640 (2d Cir. 2022) (“[A] federal court may not issue a writ of habeas corpus simply because it thinks the state court ‘applied clearly established federal law erroneously or incorrectly.’” (quoting Williams v. Taylor, 529 U.S. at 411)). In other words, to obtain §2254 relief, a petitioner must show “that the state court’s ruling on the claim being presented in federal court was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.” Harrington v. Richter, 562 U.S. at 103. Thus, “the existence of ‘reasonable arguments on both sides’ is ‘all [the state] needs to prevail in [an] AEDPA case.’” Jordan v. Lamanna, 33 F.4th at 151 (quoting White v. Woodall, 572 U.S. 415, 427 (2014)). Further, under AEDPA, “clearly established Federal law,” 28 U.S.C. §2254(d)(1), “refers only to the holdings of the Supreme Court extant at the time of the relevant state court decision,” Jackson v. Conway, 763 F.3d 115, 134 (2d Cir. 2014) (internal quotation marks omitted); see Williams v. Taylor, 529 U.S. at 412. While this court may rely on our own prior decisions for the limited purpose of “ascertain[ing] whether [we] ha[ve] already held that the particular point in issue is clearly established by Supreme Court precedent,” we may not rely on our decisions or those of other circuits “to refine or sharpen a general principle of Supreme Court jurisprudence into a specific legal rule that th[e] Court has not announced.” Marshall v. Rodgers, 569 U.S. 58, 64 (2013); accord Rodriguez v. Miller, 537 F.3d 102, 106-07 (2d Cir. 2008) (noting “[n]o principle of constitutional law grounded solely in the holdings of the various courts of appeals or even in the dicta of the Supreme Court can provide the basis for habeas relief” under AEDPA). Here, the “clearly established” federal law applicable to Englert’s ineffective-assistance claim is the two-part test articulated by the Supreme Court in Strickland v. Washington, 466 U.S. 668. See Cullen v. Pinholster, 563 U.S. at 189. This test requires a defendant challenging conviction based on ineffective assistance of counsel to show both that “(1) counsel’s performance was objectively deficient,” and that “(2) petitioner was actually prejudiced as a result.” Harrington v. United States, 689 F.3d at 129 (citing Strickland v. Washington, 466 U.S. at 687-88). At the first step of Strickland analysis, courts “‘strongly presume[]‘” that counsel “‘rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.’” Cullen v. Pinholster, 563 U.S. at 189 (quoting Strickland v. Washington, 466 U.S. at 690). To overcome this presumption, a petitioner “bears a heavy burden,” United States v. Barrett, 102 F.4th 60, 72 (2d Cir. 2024) (internal quotation marks omitted), because “[t]he question is whether an attorney’s representation amounted to incompetence under ‘prevailing professional norms,’ not whether it deviated from best practices or most common custom,” Harrington v. Richter, 562 U.S. at 105 (quoting Strickland v. Washington, 466 U.S. at 690). To establish prejudice at Strickland’s second step, a petitioner must demonstrate “that there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland v. Washington, 466 U.S. at 694. “A reasonable probability is a probability sufficient to undermine confidence in the outcome,” id., which requires “a ‘substantial,’ not just ‘conceivable,’ likelihood of a different result,” Cullen v. Pinholster, 563 U.S. at 189 (quoting Harrington v. Richter, 562 U.S. at 112). As the Supreme Court has observed, when Strickland applies “in tandem” with AEDPA, our review of the state court’s denial of a petitioner’s ineffective-assistance challenge to conviction is “doubly” deferential, certainly at the first step of analysis. Harrington v. Richter, 562 U.S. at 105 (internal quotation marks omitted); see Burt v. Titlow, 571 U.S. at 15 (stating that “doubly deferential” standard requires giving “both the state court and the defense attorney the benefit of the doubt”).11 II. Strickland Analysis A. The Performance Step: Counsel’s Obligation To Consult with or Call a Medical Expert in Child Sexual Abuse Cases At the first step of Strickland, Englert bears a heavy burden in arguing that trial counsel was constitutionally ineffective in failing to consult with or call a medical expert to counter Nurse Lyons’s testimony because, as the Supreme Court has stated, “strategic decisions — including whether to hire an expert — are entitled to a ‘strong presumption’ of reasonableness.” Dunn v. Reeves, 594 U.S. 731, 739 (2021) (emphasis added) (quoting Harrington v. Richter, 562 U.S. at 104); see United States v. DiTomasso, 932 F.3d 58, 69-70 (2d Cir. 2019) (“Trial counsel’s [a]ctions or omissions…that might be considered sound trial strategy, including decisions not to call specific witnesses — even ones that might offer exculpatory evidence — [are] ordinarily not viewed as a lapse in professional representation.” (internal quotation marks omitted)). To be sure, in making such decisions, defense counsel has a “duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary.” Strickland v. Washington, 466 U.S. at 691. Nevertheless, we evaluate the reasonableness of an attorney’s investigative choices with a “heavy measure of deference to counsel’s judgments,” id., mindful that the “same investigation” is not “required in every case” and “[i]t is rare that constitutionally competent representation will require any one technique or approach,” Cullen v. Pinholster, 563 U.S. at 195 (internal quotation marks omitted). In here arguing that his trial counsel must be deemed constitutionally ineffective for not consulting with or calling a medical expert, Englert relies not on Supreme Court precedent but on this court’s decision in Gersten v. Senkowski, 426 F.3d 588. In that AEDPA habeas challenge to a petitioner’s New York child sexual abuse conviction, this court held that the state court unreasonably applied Strickland in rejecting an ineffective-assistance claim based on trial counsel’s failure to consult or call a medical expert. See id. at 607-11. But that does not mean the same conclusion applies here. In Gersten itself, this court made plain that “no per se rule” makes “expert consultation…always necessary in order to provide effective assistance of counsel.” Id. at 609 (internal quotation marks omitted). More recently, the Supreme Court has made this point even more forcefully. See Harrington v. Richter, 562 U.S. at 111. In there rejecting a §2254 challenge to conviction based on defense counsel’s failure to consult with or call a forensic blood expert, the Court stated that Strickland does not require “for every prosecution expert an equal and opposite expert from the defense.” Id. While the Court recognized that, in some cases, counsel could be deemed ineffective for failing to consult with or call an expert, it emphasized that there were “countless” ways to provide effective representation, id. at 106, and that strategic considerations — such as the possibility of shifting the jury’s attention to “esoteric matters of forensic science” or “transform[ing] the case into a battle of the experts” — could justify not calling a defense expert, id. at 108-09; see Strickland v. Washington, 466 U.S. at 688-89 (“No particular set of detailed rules for counsel’s conduct can satisfactorily take account of the variety of circumstances faced by defense counsel or the range of legitimate decisions regarding how best to represent a criminal defendant.”). Indeed, in Harrington v. Richter, the Supreme Court observed that “[i]n many instances cross-examination will be sufficient to expose defects in an expert’s presentation.” 562 U.S. at 111. As earlier noted, AEDPA permits federal habeas relief from a state conviction based on legal error only if a state court’s legal rulings were contrary to or unreasonably applied federal law “as determined by the Supreme Court[.]” 28 U.S.C. §2254(d)(1). Thus, Gersten and similar earlier precedents of our court here relied on by Englert12 must be understood in light of Harrington v. Richter’s teachings as to both the Strickland deference properly accorded counsel’s choices in deciding whether to consult or call experts and the AEDPA deference properly accorded state courts on habeas review of such a claim. See 562 U.S. at 105. Consistent with that double deference at the first step of Strickland analysis, a §2254 petitioner claiming that counsel was constitutionally ineffective in failing to consult with or call an expert witness must show that “every fairminded juris[t] would agree that every reasonable lawyer would have” consulted or called a medical witness in the particular case. Dunn v. Reeves, 594 U.S. at 740 (emphasis in original) (internal quotation marks omitted).13 Englert fails to make that showing here. That is evident from significant differences between this case and Gersten. In Gersten, trial counsel “conceded” that the physical evidence of examination offered by the prosecution “was indicative of sexual penetration without conducting any investigation to determine whether this was the case.” 426 F.3d at 608. As this court noted, had counsel reasonably investigated the medical issues, he would likely have discovered “exceptionally qualified medical experts” who would testify that the physical evidence, in fact, “was not indicative of sexual penetration and provided no corroboration whatsoever of the alleged victim’s story.” Id. By contrast, here, there is no dispute that N.L.’s physical examination revealed no signs of sexual abuse. On that record, the prosecution’s witness, Nurse Lyons, could opine only that such normal examination findings were nevertheless consistent with the child’s allegations of sexual abuse more than six months earlier because any injuries would have had time to heal. Englert’s trial counsel — far from conceding that point — on cross-examination elicited from Nurse Lyons admissions that (1) normal examination findings were also consistent with “no abuse,” Confidential App’x 1489;14 and (2) the nurse’s consistent-with-abuse opinion depended, at least in part, on “the history…given” by N.L., id. at 1490. Thus, far from failing to challenge the prosecution expert’s medical conclusions or even conducting an ineffective cross-examination, defense counsel neutralized the physical examination evidence, thereby reducing — if not eliminating — any corroborative value the prosecution might try to assign it. Counsel thus shifted the focus of the trial to the credibility of essentially uncorroborated witnesses, particularly N.L. and her mother, and to their possible motives to testify falsely. In short, this is one of the “many instances” referenced by the Supreme Court in Harrington v. Richter where cross-examination was, by itself, “sufficient to expose defects in an expert’s presentation.” 562 U.S. at 111. On this record, one can hardly conclude that counsel’s use of cross-examination, rather than expert testimony, to neutralize Nurse Lyons’s consistent-with-abuse opinion, was an approach that “no competent lawyer would have chosen.” Dunn v. Reeves, 594 U.S. at 739 (holding that “[e]ven if there is reason to think that counsel’s conduct ‘was far from exemplary,’ a court still may not grant relief if ‘[t]he record does not reveal’ that counsel took an approach that no competent lawyer would have chosen” (quoting Burt v. Titlow, 571 U.S. at 23-24)). Much less does it appear that “every fairminded juris[t] would agree” that proceeding as counsel did without consulting or calling an expert amounted to professional incompetence. Id. at 740 (emphasis in original) (internal quotation marks omitted); see Harrington v. Richter, 562 U.S. at 105. Further, in contrast to Gersten, in this case there is no reason to think that if trial counsel had consulted a medical expert, that expert would have offered opinions materially at odds with Nurse Lyons’s. The submitted affidavit from Dr. Bomze questioned some of Nurse Lyons’s examination methods — e.g., in his view, she should have documented the child’s examination positions and obtained the child’s full medical and psychosocial history before expressing any opinion — but, in the end, he agreed with the nurse’s critical conclusions: i.e., that (1) the child’s physical examination findings were all normal; (2) such normal findings do not rule out sexual abuse; (3) it is possible for physical signs of abuse, including hymen injuries, to heal over time; and (4) a majority of sexually abused children show no signs of abuse on a physical examination when they delay disclosing abuse beyond the first several days. Counsel was not incompetent for failing to call a medical expert who, thus, would have reinforced the prosecution’s argument that normal examination findings did not foreclose abuse. Indeed, had a defense expert testified consistent with Dr. Bomze’s opinions, that would have made it harder for counsel to pursue the defense strategy of appealing to jury common sense in arguing the unlikelihood of a pre-adolescent child being sexually abused over almost five years and showing no signs whatsoever of physical injury — an argument critical to counsel’s urging reasonable doubt as to N.L.’s testimony.15 While Dr. Bomze’s assertion that certain severe hymen injuries do not readily heal might have supported counsel’s argument, any such support would have been easily undermined by the doctor’s acknowledgment that “the large majority of sexually abused children have no positive findings on anogenital examination if the exam is delayed beyond the first several days or longer” after abuse. Confidential App’x 321.16 Trial counsel cannot be deemed incompetent for not introducing such equivocal expert testimony, particularly given its potential for triggering a distracting “battle of the experts” on “esoteric matters of forensic science,” when counsel reasonably aimed to neutralize the physical examination evidence — which he did — and to focus jury attention on witness credibility and motive to falsify. Harrington v. Richter, 562 U.S. at 108-09. Further, insofar as Dr. Bomze faulted Nurse Lyons for testifying with “certainty” about N.L. being abused, Confidential App’x 321, the characterization is mistaken. Nurse Lyons testified that the child’s normal examination results were consistent with her account of abuse and that she would not have expected to see injuries in light of N.L.’s delayed disclosure, not that the nurse found abuse demonstrated to some reasonable degree of medical certainty. Finally, Dr. Bomze stated that, had he been consulted, he would have urged Englert’s trial counsel to question Nurse Lyons about normal examination results also being consistent with the absence of child sexual abuse. But trial counsel did just that, eliciting the concession that neutralized the physical examination’s normal findings. See supra at 23. Thus, it appears that Englert’s ineffective-assistance challenge fails at step one of Strickland analysis. In nevertheless urging otherwise, Englert points to his own and his family members’ affidavits, which indicate that counsel’s failure to consult with or call a medical expert was based not on any considered trial strategy, but on a lack of time or inclination to pursue the matter. In considering that argument on AEDPA review, we note the state court’s initial skepticism about the veracity of these substantially identical affidavits. We also note the absence of any affidavit from Englert’s trial counsel and any explanation from Englert as to why none was obtained. See Dunn v. Reeves, 594 U.S. at 740-42 (deeming absence of evidence from trial counsel “particularly significant” to evaluating ineffectiveness claim given “range of possible reasons…counsel may have had for proceeding as they did” and concluding that state court not obliged to accept petitioner’s “blanket assertion on an incomplete evidentiary record” that no reasonable strategy supported counsel’s actions (internal quotation marks omitted)); see also Burt v. Titlow, 571 U.S. at 22-23 (“[T]he burden to ‘show that counsel’s performance was deficient’ rests squarely on the defendant.” (quoting Strickland v. Washington, 466 U.S. at 687)). We need not pursue those points further because, even if we were to credit Englert’s affidavits and, at step one of Strickland, decide that further inquiry into counsel’s reasons for not consulting or calling a medical expert was warranted, that would not entitle Englert to habeas relief because, in any event, he fails to demonstrate the prejudice required at Strickland’s second step. Thus, we proceed to that step, mindful of the Supreme Court’s instruction that “[t]he object of an ineffectiveness claim is not to grade counsel’s performance,” and ‘[i]f it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice,…that course should be followed.” Strickland v. Washington, 466 U.S. at 697; accord Waiters v. Lee, 857 F.3d 466, 478-79 & n.21 (2d Cir. 2017) (rejecting Strickland claim for failure to demonstrate prejudice and declining to resolve performance prong, even though court had “no reason to disagree” with state court’s plausible explanations for counsel’s failure to call expert).17 B. The Prejudice Step: the State Court Did Not Unreasonably Apply Strickland’s Prejudice Prong Englert argues that, with the benefit of expert consultation or testimony, trial counsel would have been able to rebut Nurse Lyons’s testimony, undermine N.L.’s credibility, and support his fabrication defense. He maintains that Nurse Lyons’s testimony was “critical” to the prosecution’s case because it explained to the jury why they should not be troubled by the lack of physical evidence of abuse, and that Dr. Bomze’s affidavit “established the opposite of Lyons’[s] conclusions,” i.e., “that it would be expected to find evidence of sexual abuse” in a child who had been continually abused. Appellant’s Br. at 44, 54 (emphasis in original). We are not persuaded that every fairminded jurist was obliged so to conclude. To explain, we note at the outset that Englert overstates Dr. Bomze’s affidavit. Therein, Dr. Bomze did not opine that he would expect to see physical evidence of injury in a child who claimed sexual abuse akin to that reported by N.L. Nor did he dispute Nurse Lyons’s conclusion that the absence of such physical evidence was, nevertheless, consistent with N.L.’s account of abuse because injuries would have had time to heal. To the contrary, Dr. Bomze corroborated the critical portion of Nurse Lyons’s testimony when he stated that “the large majority of sexually abused children have no positive findings on anogenital examination if the exam is delayed beyond the first several days or longer.” Confidential App’x 321. While Dr. Bomze may have faulted certain statistics cited by Nurse Lyons in support of her opinions, in the end, he agreed that “the absence of any identifiable trauma does not rule out abuse.” Id. at 319. Indeed, Dr. Bomze does not even opine that identifiable trauma is likely in this case. Thus, as noted supra at 22-24, this case is not akin to Gersten v. Senkowski, where the petitioner proffered expert evidence that completely refuted the prosecution’s physical evidence purportedly showing forced sexual penetration of the child-victim. 426 F.3d at 611-12; cf. Pavel v. Hollins, 261 F.3d 210, 227-28 (2d Cir. 2001) (faulting defense counsel, pre-AEDPA, for not calling defense expert who would have stated “unequivocally” that victims’ medical conditions were “not consistent” with their claims of abuse (internal quotation marks omitted)). Because Englert can point to “no evidence” from Dr. Bomze “directly challenging…conclusions reached by the prosecution’s expert[],” the state court reasonably determined that his Strickland claim fell short of establishing the substantial likelihood of a different result necessary to demonstrate prejudice. Harrington v. Richter, 562 U.S. at 112. In urging otherwise, Englert submits that consulting an expert such as Dr. Bomze would have allowed trial counsel to rebut Nurse Lyons’s testimony that hymens can regenerate. But this fails to demonstrate prejudice because Dr. Bomze agreed that “the majority” of hymenal lacerations do heal over time. Confidential App’x 321-22. And while Dr. Bomze faulted Nurse Lyons for not distinguishing “complete” or “deep” hymenal lacerations, which “may not fully heal,” id. at 322 (emphasis added), he did not opine that such lacerations would be expected if a child had been abused or that the absence of finding unhealed lacerations undermined N.L.’s claim of abuse. Englert argues that by not consulting with or calling Dr. Bomze, trial counsel was unable to adduce evidence that anal abuse of a child should result in some bleeding,18 and that physical indications of anal abuse can persist longer than comparable indications of vaginal abuse. Even assuming such testimony might have provided additional fodder for challenging N.L.’s credibility, it would not require every fairminded jurist to question conviction. As Dr. Bomze himself acknowledged, a large majority of sexually abused children have no positive findings of injury on anogenital examination when, as in N.L.’s case, they delay disclosing the abuse. Moreover, N.L. testified that Englert’s penis penetrated her buttocks “only a little bit,” undercutting the value of Dr. Bomze’s opinion regarding injuries in this case. Confidential App’x 1065. To be sure, if called as a witness, Dr. Bomze would have reiterated for the jury that N.L.’s physical examination findings were normal and could have emphasized that such findings did not necessarily establish abuse. But trial counsel effectively made those points by securing Nurse Lyons’s concession that N.L.’s normal examination findings were also consistent with no abuse and by appealing to jurors’ common sense about the expected physical effects on a preadolescent child of years of sexual abuse. As noted supra at 24-25, calling Dr. Bomze would have made it more difficult for counsel to appeal to juror common sense as that would mean two medical practitioners would have agreed that most children who delayed reporting sexual abuse do not show physical injuries. On this record, then, the jury did not need to hear from Dr. Bomze, and trial counsel did not need to consult with him, to understand that the absence of physical evidence of abuse is not proof that a child has been abused, and that Nurse Lyons’s testimony on the point was, at best, neutral and, therefore, inconclusive.19 See Harrington v. Richter, 562 U.S. at 112 (holding state court reasonably concluded evidence of prejudice insufficient where uncalled expert would not have testified differently from concession trial counsel extracted from prosecution expert); Wong v. Belmontes, 558 U.S. 15, 23-24 (2009) (declining to consider uncalled expert testimony persuasive of Strickland prejudice where jury did not need to hear from expert to understand evidence and could rely, instead, on “common sense”). In these circumstances, we conclude that Dr. Bomze’s testimony would not have “so clearly ‘alter[ed] the entire evidentiary picture’ that the [state] court’s decision” that Englert failed to demonstrate prejudice “is indefensible.” Waiters v. Lee, 857 F.3d at 484 (quoting Strickland v. Washington, 466 U.S. at 696). The same conclusion obtains with respect to Englert’s argument that Dr. Bomze’s testimony would have assisted the jury in understanding the significance of N.L.’s lack of social, emotional, and psychological issues. As an initial matter, Englert is mistaken in arguing that trial counsel was “prevented” from arguing that N.L.’s happiness, success in school, and lack of mental health issues were “inconsistent with long-term child sexual abuse.” Appellant’s Br. at 48. Counsel not only intimated as much in his cross-examination of N.L., her mother, and Nurse Lyons, but also, in summation, argued that, as a matter of common sense, such an absence of social, emotional, and psychological issues was inconsistent with years of sexual abuse. Further, counsel reinforced that point by reminding the jury that prosecution witness Perkowski had testified that almost all of the approximately 3,200 sexually abused children he had encountered as a clinical social worker needed professional counseling.20 Had Dr. Bomze been called to testify, he might have identified certain social, emotional, or psychological problems associated with children who have been sexually abused (e.g., “irritation, anxiety, depression, sexually acting out, school problems,” Confidential App’x 321), but such testimony was unlikely to have undermined the outcome of the case. Notably, nowhere in his affidavit does Dr. Bomze directly assert that the absence of such indicia undermines a claim of abuse — the conclusion trial counsel was urging the jury to reach by reference to their own common sense. Nor can Englert establish that the state court unreasonably applied Strickland’s prejudice test in failing to recognize that consulting an expert such as Dr. Bomze could have helped trial counsel cross-examine Nurse Lyons more effectively. Without the assistance of a medical expert, trial counsel engaged in an extensive cross-examination of Nurse Lyons that, as noted, effectively neutralized her consistent-with-abuse conclusion. While Dr. Bomze might have advised trial counsel to establish that Nurse Lyons looked for and failed to find certain indicia of sexual abuse in N.L.’s genital and anal areas, trial counsel effectively did as much by taking the nurse through a litany of possible injuries and abnormalities and having her confirm that a physical examination — under 10,000 times magnification — revealed none. As for other areas of cross-examination suggested by Dr. Bomze, particularly relating to Nurse Lyons’s examination methods and procedures, the state court reasonably determined that Englert was not prejudiced by trial counsel’s failure to pursue such inquiry. As respondent observes, challenging Nurse Lyons’s physical examination as incomplete or otherwise unsound might have suggested that her normal findings were inaccurate and that she had missed indicia of abnormalities or injuries supportive of N.L.’s abuse claim. This would hardly have served Englert’s interests. In sum, because Englert failed to show how consulting or calling a medical expert would have materially undermined Nurse Lyons’s testimony beyond that achieved by trial counsel on cross-examination or otherwise would have been substantially likely to yield a different trial outcome, see Cullen v. Pinholster, 563 U.S. at 189; Harrington v. Richter, 562 U.S. at 111-12, the state courts did not unreasonably apply Strickland in concluding that Englert was not denied “a trial whose result is reliable,” Strickland v. Washington, 466 U.S. at 687.21 Accordingly, on AEDPA review, Englert’s ineffective-assistance challenge to conviction under §2254 was correctly denied for lack of prejudice. C. The State Court’s Decision Was Not Based on an Unreasonable Determination of the Facts We briefly address Englert’s contention that the state court’s rejection of his ineffective-assistance claim on §440.10 review was based on “an unreasonable determination of the facts in light of the evidence” and, thus, does not preclude federal habeas relief. 28 U.S.C. §2254(d)(2). The argument fails on the merits. First, Englert asserts that the state court “misidentified” Nurse Lyons’s key conclusions and “disregarded” portions of Dr. Bomze’s affidavit in determining that Englert did not establish prejudice. Appellant’s Br. at 52, 55. We locate no such errors in the state court’s characterization of the evidence and, in any event, Englert fails to show that all fairminded jurists would have otherwise found or stated the facts, as required to obtain §2254 relief. See Cardoza v. Rock, 731 F.3d 169, 178 (2d Cir. 2013) (holding that §2254(d)(2) requires showing that “reasonable minds could not disagree that the trial court misapprehended or misstated material aspects of the record in making its finding”). Second, Englert argues that it was unreasonable for the state court to reject his ineffective-assistance claim without holding an evidentiary hearing because he satisfied the requirements for a hearing under New York law. See N.Y. Crim. Proc. Law §440.30(1), (4). Englert does not explain how the lack of an evidentiary hearing rendered the state court’s factual determinations unreasonable, and any procedural error by the state court in denying a hearing under state law is not cognizable on federal habeas review. Cf. Estelle v. McGuire, 502 U.S. 62, 67-68 (1991) (emphasizing “it is not the province of a federal habeas court to reexamine state-court determinations on state-law questions”). Finally, because Englert’s claims of state court errors of law and fact fail on the state court record, the district court correctly denied his request for an evidentiary hearing in that forum before denying him habeas relief under §2254. See Cullen v. Pinholster, 563 U.S. at 182-83; Shoop v. Twyford, 596 U.S. 811, 819 (2022); Garner v. Lee, 908 F.3d 845, 860 (2d Cir. 2018). CONCLUSION For the foregoing reasons, we conclude that the district court correctly denied Englert’s 28 U.S.C. §2254 request for habeas corpus relief from a New York State conviction for engaging in a course of sexual conduct against a child in the first degree, see N.Y. Penal Law §130.75(1)(a), on the ground of ineffective assistance of counsel. The state courts’ rejection of Englert’s ineffective-assistance claim was not based on either a misstatement of the facts or an unreasonable application of controlling Supreme Court precedent, i.e., Strickland v. Washington, 466 U.S. 668, particularly insofar as Englert fails to demonstrate prejudice from trial counsel’s failure to consult with or call a medical expert. In these circumstances, no federal habeas relief is warranted. See 28 U.S.C. §2254(d). Accordingly, we AFFIRM the district court’s judgment denying Englert’s petition for a writ of habeas corpus.
