Wayne Lewis Charley was convicted, after a jury trial, of seven counts of sexual abuse of a child in Indian country, in violation of 18 U.S.C. §§ 1153, 2241, 2242, and 2246, and sentenced to life imprisonment. On appeal he contends, inter alia: (1) that Federal Rule of Evidence 414 is unconstitutional and that the admission, under that rule, of evidence of his prior sex offense conviction denied him due process; (2) that the district court erred by permitting five health care professionals, called by the government as lay witnesses, to (i) give expert opinion testimony without furnishing the defense with advance summaries of their testimony as required by Fed. R.Crim.P. 16(a)(1)(E); (ii) give expert testimony without adequate foundation, in violation of Fed.R.Evid. 702; (iii) improperly vouch for the credibility of the alleged victims, in violation of Rule 702; (iv) offer expert conclusions that had not been screened for reliability, in violation of Rule 702; and (v) offer testimony that assumed, *1269as a fact, that the alleged victims were in fact abused, in violation of Fed.R.Evid. 403; and (3) that the evidence was insufficient to convict him on Count I of the indictment. We conclude that the district court erroneously admitted portions of the testimony of the health care professionals, but that these errors were harmless. We therefore affirm Defendant’s convictions on Counts II through VII of the indictment. As to Count I of the indictment, we hold that the evidence was insufficient to support a conviction. Accordingly, we reverse the conviction on that count and remand for the entry of a judgment of acquittal on Count I.
BACKGROUND
On April 11, 1997, Dorothy Kalleco took her daughters, 13-year-old D.J. and 10-year-old J.J.,1 to the Crownpoint Healthcare Facility in Crownpoint, New Mexico. There she told D.J.’s pediatrician, Dr. Edward Junkins, that the girls had disclosed to her that an “uncle” had sexually molested them on numerous occasions over an extended period of time. II R. at 157. Defendant, 65-year-old Wayne Lewis Charley, the girls’ uncle by Navajo clan affiliation, was subsequently identified as the alleged abuser.
Dr. Junkins interviewed the girls out of the mother’s presence. They provided detailed accounts of anal and genital contact by their uncle both with his fingers and his penis. A physical examination showed no evidence of abuse. Both girls had intact hymens,2 and the anal and genital areas appeared normal with no visible bleeding, bruising, scarring, tears, tags or discharge. According to Dr. Junkins, this circumstance was not inconsistent with sexual abuse since children’s tissues heal quickly, although there may be residual scarring.
Dr. Junkins testified that the report of abuse had immediate significance to him as a possible unifying explanation for D.J.’s extensive history of atypical and medically puzzling physical complaints and conditions. These included persistent bed-wetting, chronic urinary tract infections, chronic stomach pain, headaches, breathing problems and, recently, numbness of her arm, a burning feeling in her face, and pain in the side of her face, arm and leg. One April entry in D.J.’s medical chart described her as anxious and crying. J.J.’s medical chart did not show a similarly extensive history of complaints, except for persistent bed-wetting.
While the girls were at the Crownpoint Healthcare Facility, Dr. Junkins called the Crownpoint Police Department and reported the incident to Sergeant Daniel Zuni. Officer Zuni responded and conducted a preliminary interview with the girls. In the course of the interview, the girls told Zuni that the sexual abuse had occurred primarily at their residence and secondarily at the Elite Laundromat, located in the Crownpoint Shopping Mall where, beginning in January 1997, Defendant worked several evenings each week from 7:00 to 10:00 P.M. Both locations are within the boundaries of the Navajo Indian Reservation. Officer Zuni also ascertained that Defendant was the alleged perpetrator. He then went to Defendant’s residence and determined that Defendant was a member of the Navajo Nation, as were the girls, thus making the alleged crime a federal offense.
Later that same evening, on April 11, criminal investigator Patricia Henry, to whom Officer Zuni had referred the matter, reported the circumstances to Federal Bureau of Investigation (FBI) Agent John D. Tanburg, who then commenced an investigation. On April 30,1997, Agent Tan-burg visited the girls’ residence and interviewed the girls in detail. During the visit, Tanburg was accompanied by anoth*1270er investigator who inspected the girls’ residence. At some point, Tanburg obtained a warrant, arrested Defendant, and placed him in a secure halfway house.
None of the law enforcement officers were able to discover any direct evidence of the reported incidents. There were no eyewitnesses; there was no physical evidence; and Defendant denied the accusations. However, from the outset of the girls’ disclosures, everyone involved, including those providing treatment, was aware that Defendant had been convicted in 1994 for sexually abusing his five-year-old granddaughter, and had been sentenced to one year in prison and three years of supervised release. He was released from federal custody on June 23, 1995.
In May 1997, about a month after the initial disclosures, the girls were referred to Western New Mexico Counseling Services in Gallup for supportive counseling. Joelle Baum, a registered nurse and registered health counselor, who also held a temporary license in art therapy, evaluated the girls and began a course of treatment for J.J. Kristine Lee Carlson, a clinical therapist and sexual assault coordinator for Gallup, New Mexico, began a course of treatment for D.J.
On June 4, 1997, the girls were seen and examined by Dr. Renee Ornelas, a pediatrician at the University of New Mexico. Like Dr. Junkins, Dr. Ornelas also found the girls’ genital and anal areas to be normal, evidencing no signs of abuse. But, after receiving information from Ms. Kalle-co about the girls’ medical history and reports of abuse, and after interviewing the girls, Dr. Ornelas concluded that the girls had been sexually molested.
On the same day as the Ornelas examination, June 4, 1997, a federal grand jury returned a seven-count indictment against Defendant, charging him with four counts of sexual abuse of a child in Indian country, in violation of 18 U.S.C. §§ 1153, 2242, and 2246, and three counts of aggravated sexual abuse of a child in Indian country, in violation of 18 U.S.C. §§ 1153, 2241, and 2246.3 Count I of the indictment charged Defendant with abusing D.J. “[o]n or about October 1995.” I R., Tab 9. Counts II through VII all charged Defendant with abusing either D.J. or J.J. “[o]n or about March 1997.” Id.
Well before trial in this case, the government notified defense counsel that it would call as witnesses certain health care professionals who had examined, counseled, or treated D.J. and J.J. In connection with that prospective testimony, the government also provided defense counsel with voluminous medical information and associated notes and records. However, it did not designate any of these witnesses as experts and, accordingly, did not provide defense counsel with any summaries of expert witness testimony pursuant to Fed. R.Crim.P. 16(a)(1)(E).
On October 24, 1997, the Friday preceding the start of trial, defense counsel filed a motion in limine seeking to prevent any of the designated health care professionals from offering expert opinion testimony regarding whether D.J. or J.J. were in fact sexually abused or were being treated for sexual abuse. The district court addressed the motion on October 27, 1997, the first day of trial, and denied it on the government’s representation that the witnesses in question were not being called as experts under Fed.R.Evid. 702, but as lay or fact witnesses under Fed.R.Evid. 701. After a colloquy .with defense counsel regarding how counsel could preserve his objections, the court indicated that it recognized defense counsel’s continuing objection to the testimony of the health care professionals.4
*1271The trial began that afternoon, and lasted three days. The government called twelve witnesses during the presentation of its case-in-chief. Among the first witnesses called were Dr. Junkins, Dr. Orne-las, Ms. Baum, and Ms. Carlson. As will be discussed more fully below, these witnesses testified regarding their visits with the girls, the symptoms exhibited by the girls, the treatment they provided the girls, and, with respect to Dr. Junkins and Dr. Ornelas, their diagnoses.
The government also called FBI Agent Tanburg and Navajo investigators Zuni and Sampson Cowboy. Tanburg and Zuni told the jury about their respective investigations of the allegations, and Cowboy testified about Defendant’s 1994 abuse conviction.
Near the end of their presentation, the prosecutors called the two young girls to the stand. Both D.J. and J.J. testified, in graphic detail (mostly elicited by defense counsel on cross-examination), that Defendant had abused them in a variety of ways at various times and places. D.J. testified that Defendant molested her chiefly at her residence, and described sexual encounters with Defendant in her kitchen, hallway, and bedroom. D.J. testified that these incidents occurred primarily in the evening. She also testified to some incidents of molestation which occurred at locations outside her residence, in particular one instance in the parking lot of the Elite Laundromat in a truck. J.J., by contrast, testified that Defendant molested her chiefly at the laundromat, and described incidents in the employees’ room, the bathroom, and near the washing machines.
D.J. could not specifically recall whether Defendant had abused her in 1995, although she testified that the abuse had been occurring “[sjince [she] was small,” and had begun before Defendant went to prison in 1994 and continued after his release. Ill R. at 358, 363.
Following the testimony of the girls, the government called Dr. Timothy Kern, a urologist and surgeon who performed prostate surgery on Defendant on September 29, 1995. Kern testified that 95% of men who undergo such prostate surgery have no erectile problems after the surgery. On cross-examination, Kern stated that patients such as Defendant are told, upon release from the hospital, that they should not engage in sexual activity for four to six weeks afterwards, because ejaculation during this post-operative period “usually” although “[n]ot necessarily in every case” causes bleeding which requires urgent medical attention. IV R. at 442, 449. Kern, testifying from medical records, stated that Defendant had come in for a post-operative checkup in late October, and no evidence of bleeding was apparent.
The government’s final witness was Ms. Kalleco, who testified that, during the spring of 1997, she had often been out of the house in the evenings because she frequently traveled to Albuquerque to visit her sister who was in the hospital at that time. She stated that, on many of these occasions, the girls were home alone, because she took the girls’ siblings with her to Albuquerque. She also testified that Defendant frequently visited her house, ate some meals there, and on occasion asked questions designed to ascertain whether she expected to be away from the home in the evening.
The defensé’s primary theory of the case was that the girls’ mother was in a sort of *1272a feud with Defendant’s family over two separate incidents, one involving an automobile accident, and the other involving the sale of a pickup truck, and, in an attempt to get back at Defendant or his family, had pressured the girls into accusing Defendant of abuse. Secondarily, the defense argued that D.J. had invented the allegations of abuse to avoid having to go to school. Additionally, there was evidence that J.J. had told a social worker that, in accusing Defendant of sexual abuse, she and her sister were “just playing,” III R. at 323, and that D.J., when asked by doctors about the possibility of sexual abuse some years earlier, had denied that she was being abused.
The defense also called Bucky Belen, Defendant’s co-worker at the laundromat, who testified that he had seen the girls at the laundromat numerous times, where they would help Defendant and sometimes ask him for money, but that he had never seen Defendant act inappropriately toward them.
Defendant’s son, Dwayne Charley, testified for the defense, and stated that he provided Defendant transportation to work each night, either by driving Defendant there himself, or by driving his truck to his sister’s house and letting Defendant drive it to work from there. Dwayne Charley testified that, on the nights when he didn’t personally drive his father to work, he delivered the truck to his sister’s house at approximately 7:00 P.M., and that if Defendant had driven anywhere else on his way to work he would have been late.
The defense’s final witness was Charley himself, who, through a Navajo interpreter, denied ever abusing or in any way inappropriately touching either of the girls. Defendant also denied ever being at the girls’ house. Additionally, Defendant stated that while the girls were at his workplace, they would never help him and that they never asked him for money. Defendant also testified that he had told the FBI about his first offense, in 1994, and that he had never denied his involvement in the first offense to the FBI.
After the defense presented its case, the government presented one rebuttal witness, FBI Agent Greg Calles, who had been assigned to investigate Charley’s 1994 crime. Calles testified that Defendant had initially denied abusing his granddaughter, but that he later recanted and admitted molesting his granddaughter.
After deliberating for about three hours, the jury convicted Defendant on all seven counts listed in the indictment. At sentencing, the district court, applying the United States Sentencing Guidelines, sentenced Defendant to life imprisonment.
DISCUSSION
I. Evidence of Prior Conviction for Sexual Molestation
Evidence of Defendant’s prior conviction for sexual molestation of his granddaughter was admitted into evidence pursuant to Fed.R.Evid. 4035 and 414(a).6 Defendant contends that Rule 414 is unconstitutional, and that introduction of his prior conviction for a sex offense was so prejudicial that it denied him his due process right to a fair trial.
We recently held that Rule 414 is not unconstitutional on its face, “because Rule 403 applies to Rule 414 evidence,” and “Application of Rule 403 ... should always result in the exclusion of evidence that” is so prejudicial that it violates a defendant’s due process right to a fair *1273trial. United States v. Castillo, 140 F.3d 874, 883-84 (10th Cir.1998). Thus, when reviewing a district court’s decision to admit evidence under Rule 414 and Rule 403, we must conduct a “case-specific inquiry” into whether the district court “should have excluded the Rule 414 evidence under Rule 403.” Id. We will disturb a trial court’s decision to admit evidence under Rule 403 only for abuse of discretion. Id. at 884.
In this case, the district court made the following findings with respect to the admission of Defendant’s prior conviction for child sexual abuse:
For the record, I have made a balancing test under 403 and 41 [4], and I find the testimony of the previous sexual activity by the defendant to outweigh any harm to come to him under 403. Specifically, under relevancy, in the discussions under Rule 413 it says, [“]The proposed reform is critical to the protection of the public from rapists and child molesters, and is justified by the distinctive characteristics of the cases it will affect. In child molestation cases, for example, a history of similar acts tends to be exceptionally probative because it shows an unusual disposition of defendant—a sexual or sadosexual interest in children— that simply does not exist in ordinary people. Moreover, such cases require reliance on child victims whose credibility can readily be attacked in the absence of substantial corroboration. In such cases, there is a compelling public interest in admitting all significant evidence that will illumine the credibility of the charge and any denial by the defense. [”] So I have conducted that balancing test. As I previously stated to you I found it was more probative than not under 4[0]3.
III. R. at 269-70 (quoting Fed.R.Evid. 413 advisory committee’s note (statement of Rep. Susan Molinari)) (emphasis added). These on-the-record findings are sufficient to explain the district court’s reasoning in admitting the statement, see Castillo, 140 F.3d at 884 (requiring that the district court “ ‘make a reasoned, recorded’ statement of its 403 decision when it admits evidence under Rule 413-415”) (quoting United States v. Guardia, 135 F.3d 1326, 1332 (10th Cir.1998)), especially in light of the fact that Defendant’s brief contains no argument that these findings are insufficient.7 By invoking the stated general reasons for the Rule’s enactment, the district court was implying that those reasons were particularly important in this case. Specifically, the district court alluded to the similarity between Defendant’s prior crime and the present charges, and to the fact that there was little direct corroborating evidence. See Guardia, 135 F.3d at 1331. After examining the record as a whole, we cannot conclude that the district court abused its discretion by admitting evidence of Defendant’s prior conviction.
II. Expert Witnesses
Defendant objects to the testimony of five health care professionals called as witnesses by the government: Dr. Junkins, Dr. Ornelas, Ms. Baum, Ms. Carlson, and school counselor Carolyn Joe.8 The government concedes that it elicited expert opinion testimony regarding sexual molestation of D.J. and J.J. from at least some of these witnesses, despite representing to the *1274court that they were merely being called as lay or fact witnesses.9
In his motion in limine, Defendant raised several objections to the testimony of these witnesses. First, Defendant argued that the government, by designating them as lay witnesses, was circumventing the requirements of Fed.R.Crim.P. 16(a)(1)(E), and asked the district court to sanction the government by limiting these witnesses to fact testimony or excluding their testimony altogether. Second, Defendant argued that the district court should exclude all opinions based on statements of the girls themselves because such opinions are unreliable and/or constitute impermissible vouching for the credibility of the girls. In a related argument, Defendant asked the district court to exclude, as prejudicial under Fed.R.Evid. 403, any testimony that the girls had been abused or referred for treatment of sexual abuse, contending that such testimony “improperly suggests to the jury the existence of child abuse.” I R., Tab 31 at 3. On appeal, Defendant presses each of these arguments, and argues additionally that the district court erred by allowing these witnesses to offer expert testimony without being properly qualified as experts.
As indicated above, our analysis of the testimony of these witnesses proceeds on the assumption that defense counsel made timely appropriate objections to relevant portions of their testimony, and that the district court overruled those objections. Accordingly, we review those rulings under an abuse of discretion standard. See Kumho Tire Co. v. Carmichael, — U.S. -, -, 119 S.Ct. 1167, 1176, 143 L.Ed.2d 238 (1999); United States v. Haslip, 160 F.3d 649, 653 (10th Cir.1998), cert. denied, — U.S. -, 119 S.Ct. 1346, 143 L.Ed.2d 509 (1999).
A. Rule 16 Violation
We agree with Defendant that the government violated Rule 16(a)(1)(E), which requires the government, at the defendant’s request, to “disclose to the defendant a written summary of testimony that the government intends to use under Rules 702, 703, or 705 of the Federal Rules of Evidence during its case-in-chief at trial.” A party failing to comply with the requirements of Rule 16 is subject to sanction by the district court, which “may order such party to permit the discovery or inspection, grant a continuance, or prohibit the party from introducing evidence not disclosed, or it may enter such other order as it deems just under the circumstances.” Fed.R.Crim.P. 16(d)(2). A district court’s decision to apply or not to apply sanctions “will not be disturbed absent abuse of discretion.” United States v. Wicker, 848 F.2d 1059, 1060 (10th Cir.1988) (citing United States v. Fernandez, 780 F.2d 1573, 1576 (11th Cir.1986)). “In selecting a proper sanction, a court should typically consider (1) the reasons the government delayed producing requested materials, including whether the government acted in bad faith; (2) the extent of prejudice to defendant as a result of the delay; and (3) the feasibility of curing the prejudice with a continuance.” United States v. Gonzales, 164 F.3d 1285, 1292 (10th Cir.1999); see Wicker, 848 F.2d at 1061. “Frequently it will be found that the party who requested disclosure has not been prejudiced and that no sanction is needed.” 2 Charles Alan Wright, Federal Practice and Procedure § 260, at 121-22 (1982) (citing cases). We note that the sanction requested by Defendant — exclusion of the witnesses’ expert testimony — is almost never imposed “in the absence of a constitutional violation or statutory authority for such exclusion.” Gonzales, 164 F.3d at 1292.
In this case, Defendant requested that the government turn over summaries of the testimony of all its expert witnesses. The government refused to comply with this request, because it maintained that each of its witnesses would be testifying as lay witnesses rather than as experts. *1275Nevertheless, between July and September 1997, the government turned over to the defense copies of all medical and counseling records from which the witnesses would be testifying. Therefore, the government argues that the defense suffered no prejudice because it knew well in advance of trial who was going to testify and the nature and purpose of the expected testimony. Defense counsel had the opportunity to interview the government’s witnesses prior to trial, and in fact questioned Dr. Junkins a few days before the trial began. Indeed, defense counsel himself has had a difficult time, even when directly questioned on the matter, identifying any particular prejudice. He does not identify, for example, any witnesses he would have called if the government had designated its health care professionals as experts prior to trial and furnished summaries of their testimony in addition to information already available to the defense. Nor does he specify any significant hindrance to his preparation for cross-examination of the witnesses in question. We note that in some ways, defense counsel may in fact have been better off with the actual records than with Rule 16 summaries.
In light of these facts, we cannot conclude that Defendant suffered any prejudice as a result of the government’s failure to provide Rule 16 summaries for its witnesses who admittedly gave expert testimony. When we couple this conclusion with the fact that there is no evidence that the government was motivated by bad faith, we are persuaded that the district court did not abuse its discretion in refusing to exclude the expert testimony of these witnesses.
B. Admissibility of the Testimony
Defendant’s remaining arguments—that the challenged witnesses impermissibly vouched for the girls’ credibility, gave unduly prejudicial statements, and testified without adequate foundation—are grounded in Fed.R.Evid. 403 and 702.10 Because the objections vary as to the testimony of each witness, we will consider the witnesses separately.
1. Dr. Junkins
Dr. Junkins qualified as an expert, although his credentials, contrary to proper procedure, were established piecemeal throughout his testimony. He is a board certified pediatrician. From 1994 through July 1997 he was employed by the Public Health Service serving as a physician at the Crownpoint Healthcare Facility in Crownpoint, New Mexico. At the Crown-point facility, Dr. Junkins primarily treated Native American patients, both children and adults, in a family practice setting, and was regarded by family practitioners there as a specialist in pediatrics to whom young patients with special problems were referred. After leaving Crownpoint, he was employed as a physician at the Primary Children’s Medical Center in Salt Lake City, Utah.
In addition to these credentials, Dr. Jun-kins also had specific experience with child sexual abuse cases. He testified about the form he fills out and the list of things he covers in suspected abuse cases, and how he followed that form in this case. He testified that normal physical examinations can still be consistent with abuse, stating that' “that’s been my experience with child abuse examinations.” II R. at 166. He testified that he looks for scars and bruising, telling the jury that “I look for all that. I often don’t find it, and so this did not surprise me in this case.” Id. at 166. He stated further that “a physical exam [in suspected abuse cases] is—you know, in my experience 90 percent of the time normal.” Id. at 220. And, again emphasizing *1276the basis of his opinion in experience, he testified:
Q. And that’s why, if you didn’t see a traumatized area within 72 hours, so what?
A. Oh, that’s what I — that’s been my overwhelming experience with this. I would have been surprised in fact.
Id. at 217-18 (emphasis added). Dr. Jun-kins was qualified to render expert opinions regarding child sexual abuse.
Dr. Junkins properly testified to the facts of the girls’ medical history and treatment,11 his examination, and what the girls told him.12 He then offered two opinions to which Defendant objects. He testified, essentially, that while his physical examination revealed nothing remarkably abnormal, a normal examination is not necessarily inconsistent with a history of sexual abuse. More significantly, he linked the girls’' — especially D.J.’s — atypical history of physical complaints and manifestations to the alleged sexual abuse:
Q. But, now, in light of knowing her whole history, what do you think was related to the urinary tract infections, the kidney infections, the bed wetting, the abdominal pain, the paralysis, the nausea?
MR. FINZEL: Objection, Your Honor.
THE COURT: Overruled. Go ahead.
THE WITNESS: What was related to all those things, the unifying diagnosis that could — could possibly pull all of it together would be an abusive situation, an incredibly stressful situation, such as sexual or physical abuse could — could describe or explain all of those problems. Could.
II R. at 219 (emphasis added).
This was a more qualified version of an earlier statement by him that abuse “would” explain, and “was” a unifying diagnosis for, all of DJ.’s symptoms and history — immediately qualified by “helped” to bring the problems together. Id. at 162. Dr. Junkins also stated that the entire visit on April 11 caused it to become “clear to [him] as a pediatrician what had been going on and what’s the cause of all these problems.” Id. at 155. Taken as a whole, however, we regard the qualified opinion set out above (“could” explain), which was offered as a summation at the end of his testimony, to be the essence of Dr. Junkins’ opinion, and what the jury was left with by way of information.
The district court did not abuse its discretion by allowing these general and conditional opinions. A qualified medical expert in a sexual abuse case can, for example, “summarize the medical evidence and express an opinion that the evidence is consistent or inconsistent with the victim’s allegations of sexual abuse,” and “can inform the jury of characteristics in sexually abused children and describe the characteristics the alleged victim exhibits.” United States v. Whitted, 11 F.3d 782, 785 (8th Cir.1993); see also United States v. Wright, 119 F.3d 630, 635 (8th Cir.1997) (stating that a witness’s testimony that he “conducted a ‘Sexual molestation exam’ and that in his opinion the lack of medical evidence was not inconsistent with molestation” was admissible); United States v. Johns, 15 F.3d 740, 743 (8th Cir.1994) (allowing an expert to tell the jury about typical “emotional and psychological traits *1277for abuse victims,” and noting that such testimony was admissible because the witness “made no comment on [the alleged victim’s] credibility nor did [the witness] identify [the alleged victim] as a victim of child abuse”).
Defense counsel cross-examined Dr. Junkins in detail, was conversant with the subject matter, made no attempt whatsoever to challenge his experience, his training qualifications, or the basis for his opinions, and cites nothing to us on appeal which the defense was prevented from using to dispute any part of the testimony. There was no error in the admission of Dr. Junkins’ testimony.
2. Dr. Ornelas
Dr. Ornelas was also qualified to render expert opinions regarding child sexual abuse. Dr. Ornelas is a pediatrician at the University of New Mexico, and at the time of trial had been in practice for eleven years. For the last seven years, she had been the head of UNM’s Para Los Nios Program, which coordinates medical evaluations of children who are referred for concerns of sexual abuse.
Agent Tanburg reported his interview with D.J. and J.J. to Dr. Ornelas. On June 4, 1997, the day Defendant was indicted, Dr. Ornelas and her intake' social worker saw the girls, examined and evaluated them. The girls’ mother provided information regarding the girls’ medical problems and their reports of abuse. Although the record is not entirely clear on the point, it appears that Dr. Ornelas did not obtain copies of the girls’ medical records from the Crownpoint Healthcaré Facility or confer with Dr. Junkins or any other treating physician.13
At trial Dr. Ornelas did not testify to any further contact with or treatment of the girls beyond the examination and evaluation on June 4, 1997, except for a workup on the condition of the girls’ kidneys in order to see if there was an anatomical problem which caused their frequent problems with bed-wetting.
Dr. Ornelas testified that her examination of the girls showed no physical evidence of sexual abuse. She found no bruising, bleeding, scarring or tags, and she found that the hymens were intact. Defendant, for obvious reasons, does not challenge this portion of her testimony which, in. any event, was a factual recitation of the doctor’s observations, not an opinion. She opined generally that penetration does not necessarily occur in abuse cases. Rather, coitus can occur externally, between the labia. She also generalized that abuse can cause symptoms and behavior such as headaches, numbness, urinary tract infections, and bed-wetting.14
There was no error in allowing such generalized medical opinion evidence. Whitted, 11 F.3d at 785-86. Furthermore, the opinions depict matters so close to being self-evident that failure to conduct a gatekeeper inquiry prior to allowing the testimony was not error.
The problem with Dr. Ornelas’ testimony lies in the fact that she was permitted to give the jury her unconditional opinion that each of the girls was in fact sexually abused. She testified that her “final conclusion was that [D.J.] had been sexually abused,” III R. at 317-18, and that “[J.J.] also had a normal examination and also she — my conclusion is that she was also sexually abused,” id. at 323. Furthermore, Dr. Ornelas stated that “[t]he kind of contact that occurred is what’s called labial coitus.... ” Id. at 322 (emphasis added).
*1278She evidently based this opinion on what the girls said, to her and others, as indicated by the following exchange on cross-examination:
Q. Then your assessment of child •abuse is based on what [D.J.] told you about both [D.J.] and [J.JJ?
A. And what — the information from her mother and the information that we had that the children had given on prior interviews.
Q. Told the same — or—stories before?
A. Yes, well, had made disclosures before—
Q. Uh-huh.
A. —about sexual abuse.
Q. And, of course, the mother got the information she said from the children.
A. That’s correct.
Id. at 344.
Regardless of whether Dr. Ornelas based her conclusion on the girls’ allegations of sexual abuse, or on the girls’ history of atypical medical symptoms, her testimony is problematic. If the conclusion was based on the girls’ medical symptoms, questions surrounding the reliability of such a conclusion arise. On the other hand, if the conclusion was based on the girls’ allegations, Dr. Ornelas .was merely vouching for the credibility of the child complainants.
The opinion offered by Dr. Ornelas falls under Fed.R.Evid. 702. Among other things, that Rule imposes a special gatekeeping obligation on the trial judge to ensure that an opinion offered by an expert is reliable. See Kumho Tire, —■ U.S. at-, 119 S.Ct. at 1176. Indeed, “where [expert] testimony’s factual basis, data, principles, methods, or their application are called sufficiently into question ... the trial judge must determine whether the testimony has ‘a reliable basis in the knowledge and experience of [the relevant] discipline.’ ” Id. at 1175 (quoting Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 592, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993)). The trial judge has broad discretion, reviewable for its abuse, “to determine reliability in light of the particular facts and circumstances of the particular case.” Kumho Tire, — U.S. at -, 119 S.Ct. at 1179.15 The trial judge is granted great latitude in deciding which factors to use in evaluating the reliability of expert testimony, and in deciding whether to hold a formal hearing. Id. at 1176.
Here, no reliability determination was made at all with respect to Dr. Ornelas’ unconditional opinion that D.J. and J.J. were sexually abused. As a practical matter, that issue might have been disposed of simply by sustaining the objection, on foundation grounds, to the question seeking to elicit Dr. Ornelas’s opinion. At that point, government counsel might have approached the subject in a more acceptable way, and if not, a bench conference could have ended the line of questioning. Or, the subject could have been explored and passed on prior to trial, in the sound discretion of the trial judge. See Kumho Tire, — U.S. at -, 119 S.Ct. at 1176. But, as it happened, nothing was done here. The record does not disclose, for example, what data would support ruling out all causes except sexual abuse for the girls’ physical complaints, or to what degree Dr. Ornelas relied on her purely subjective views. Cf. id. at 1177. Indeed, Dr. Ornelas herself, as indicated above, ordered a work-up to determine if an anatomical problem was causing the girls to wet the bed. Ill R. at 324-25. Thus, if Dr. Ornelas’s opinion was based on the girls’ medical history, there is simply no *1279showing on this record, as required by Rule 702 and Kumho Tire, that such testimony is reliable. See Gier v. Educational Serv. Unit No. 16, 845 F.Supp. 1342 (D.Neb.1994) (conducting a reliability inquiry and determining that expert opinion testimony that sexual abuse in fact occurred was, in that particular case, not reliable), aff'd, 66 F.3d 940 (8th Cir.1995).
On the other hand, if Dr. Ornelas’ opinion was largely based on crediting the girls’ account, whether disclosed to her or others, she was essentially vouching for their truthfulness. In general, expert testimony which does nothing but vouch for the credibility of another witness encroaches upon the jury’s vital and exclusive function to make credibility determinations, and therefore does not “assist the trier of fact” as required by Rule 702.16 See United States v. Azure, 801 F.2d 336, 339-40 (8th Cir.1986); see also United States v. Shay, 57 F.3d 126, 131 (1st Cir.1995); Whitted, 11 F.3d at 785-86;17 4 Weinstein’s Federal Evidence § 702.03[5] (1998). Indeed, the government concedes that, in this case, testimony to the effect that D.J. and J.J. were “telling the truth ... would be impermissible.” Appellee’s Br. at 26.
Most courts that have considered the issue have concluded that expert testimony, based on the statements of the alleged victim, that sexual abuse in fact occurred is inadmissible under Fed.R.Evid. 702 (or similar military or state evidentiary rules) because, in such cases, the expert offering the opinion is merely vouching for the credibility of the alleged victim.18 Thus, if Dr. Ornelas largely based her opinion on the statements of the girls, then under the *1280foundation (or lack thereof) presented in this case, we consider it inadmissible. Therefore, regardless of whether Dr. Ornelas’s conclusion was based on the girls’ medical history or on their allegations of abuse, its admission was erroneous.
3. Joelle Baum and Kristine Lee Carlson
Joelle Baum and Kristine Lee Carlson testified as mental health experts. As an initial matter, we conclude that, as licensed mental health counselors, Ms. Baum and Ms. Carlson were properly qualified to testify as they did.19 Aside from alleging that these counselors were not properly qualified, Defendant objects to their testimony on three grounds. First, he argues that it was error to let Ms. Baum and Ms. Carlson testify at all because evidence of the referral of D.J. and J.J. to Western New Mexico Counseling Services for treatment of sexual abuse prejudicially presumes the fact in issue — 'that sexual abuse occurred. Second, he contends that they impermissibly vouched for the girls’ credibility. Finally, he asserts that their testimony contained numerous baseless statements (other than the fact of referral) that assumed the fact of sexual abuse.
With respect to Defendant’s contention that evidence of the referral itself was prejudicial, we are, under the circumstances of this case, compelled to disagree. Ms. Baum testified that the referral came about because “the girls had disclosed sexual abuse and that they were having a real difficult time and needed ... supportive counseling to help them cope.” Ill R. at 239. This does not go to the truth of the matter of sexual abuse, but, rather, to the reason for the referral. Likewise, statements by both Ms. Baum and Ms. Carlson that they were treating trauma associated with sexual abuse relate to the treatment. While we agree with Defendant that some prejudice flows from the premise underlying this testimony, the additional testimony given by Ms. Baum and Ms. Carlson with respect to what they observed and did caused this testimony regarding referral to be more probative than prejudicial under Fed.R.Evid. 403. Cross-examination is the tool for clarifying any ambiguities on the point. Thus, on cross-examination, defense counsel emphasized that the referral was based only on an allegation of sexual abuse.20
As to Defendant’s contention that the counselors vouched for the credibility of the children, we note that the counselors permissibly testified regarding the facts of their observations of the girls, the nature of the treatment given, and what the girls said and did in connection with their treatment. For instance, Ms. Baum described J.J. as distressed and fearful when telling about really bad dreams she was having. She stated that J.J. drew a house and said she was having bad memories about experiences that had happened there. She also stated that J.J.’s bed-wetting had decreased after counseling.
As to D.J., Ms. Baum testified that she was experiencing paralysis of her legs, *1281that her brother had to carry her into the session, and that she seemed very and anxious and had difficulty talking about the situation. D.J. also said she was having bad memories from past experiences in her house. Both girls expressed fear of being at home without their mother there.
Ms. Carlson testified that D.J. complained of nightmares, was very fearful, did not like to wear dresses, and related distressing flashbacks and images. Also, Carlson testified that D.J. feared being away from her mother, disliked being touched, was afraid of strangers, and expressed fear of Defendant’s wife. Finally, D.J. talked about ways to be safe at home and identified things to protect her such as ancestors, her dog, dragons and lions. She then created artwork in the form of a large shield. All of these factual observations and statements given in treatment were properly admitted.
Both Ms. Baum and Ms. Carlson did express opinions: that the girls needed supportive counseling; that bed-wetting can be a symptom of anxiety; that once traumatized a child is slow to trust an adult; and that D.J.’s symptoms were consistent with the symptoms of sexual abuse victims. These opinions, some of which were general and others qualified, were within the expertise of the witnesses. They were both relevant and sufficiently reliable, and would assist the jury, as required by Fed.R.Evid. 702. This testimony did not constitute impermissible vouching. See Whitted, 11 F.3d at 785-86.
However, Ms. Carlson’s opinion that D.J.’s symptoms were more consistent with the symptoms of children who have been sexually abused than with the symptoms of children who witness physical abuse of their mother was erroneously admitted. No sufficient foundation was laid for this kind of expert analysis, and no reliability inquiry was undertaken. See Kumho - U.S. at -, 119 S.Ct. at 1175-76.
Finally, Defendant’s third argument has merit. The counselors were permitted to offer testimony that assumed the fact of abuse&emdash;a determination to be made by the jury. Ms. Baum stated that “it was becoming increasingly hard for the children to be in the[ir] house because some of the abuse had happened in the house.” Ill R. at 243. Ms. Carlson talked about the environment “where the abuse occurred,” id. at 251, and engaged in the following exchange with government counsel:
A. The home environment seems to have changed during that time and become more stabilized, so that the abuse that I have been dealing with her at this point in time seems to be recent.
Q. Although you’re aware of abuse over an extended time of [D.J.], sexual abuse?
A. Oh, yes. Yes.
Id. at 258.
Q. And so if she had expressed a desire to not going to school or had problems that would prevent her from going to school, what would have been the reason for those based on your counseling sessions with [D.J.]?
A. I would have connected that&emdash;I do connect those to the trauma of the sexual abuse.
Id. at 258-59.
While the latter quote may also have been an opinion, it was also more particularly a statement of fact. It was error to admit these statements which were manifestly outside the counselors’ direct knowledge.21 Such statements had minimal pro*1282bative value, and, in the context of this referral for treatment of problems allegedly connected with sexual abuse, they were unquestionably prejudicial. They also im-permissibly invaded the province of the jurors. Accordingly, they were inadmissible under Fed.R.Evid. 403, and it was error to allow them.
C. Harmless Error
Finally, we must determine whether the district court’s erroneous decisions to admit the conclusions of Dr. Orne-las, the prejudicial statements of Ms. Baum and Ms. Carlson, and the physical abuse-related opinion of Ms. Carlson were harmless. A non-constitutional error, such as a decision whether to admit or exclude evidence, is considered harmless “unless a substantial right of [a] party is affected.” Fed.R.Evid. 103(a). We have stated that an error affecting a substantial right of a party is an error which had a “ ‘substantial influence’ on the outcome or [which] leaves one in ‘grave doubt’ as to whether it had such effect.” United States v. Rivera, 900 F.2d 1462, 1469 (10th Cir.1990) (en banc) (quoting Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)). When conducting our harmless error analysis, we review the record as a whole. See United States v. McVeigh, 153 F.3d 1166, 1204 (10th Cir.1998), cert. denied, — U.S. -, 119 S.Ct. 1148, 143 L.Ed.2d 215 (1999). Harmless error analysis is not the same as insufficiency of the evidence analysis.22 See United States v. Tome, 61 F.3d 1446, 1455 (10th Cir.1995) (stating that “[t]he question is not whether, omitting the inadmissible statements, the record contains sufficient evidence for a jury to convict the defendant”). The burden of proving that an error is harmless falls on the government. See Rivera, 900 F.2d at 1469 n. 4.
Harmless error is not an easy issue in this case. Defendant took the stand in his own defense and categorically denied ever inappropriately touching either of the girls. There was no physical evidence of abuse, and no eyewitnesses to the alleged misdeeds.
On the other hand, both girls testified, each corroborating the other. Both were old enough to understand the true nature of their accounts. D.J. especially had an undisputed and documented history of atypical medical conditions and complaints that was consistent with sexual abuse. In addition, both girls displayed symptoms of emotional distress that were consistent with sexual abuse. J.J. was distressed and fearful, and drew a picture of a house where she had had bad experiences. D.J. also had bad memories associated with a house, and complained of nightmares, expressed fear of strangers and Defendant’s wife, and did not like to wear dresses.
Defendant had a prior conviction for sexual abuse of a child, for which he served one year in prison and three years on supervised release. Indeed, because Defendant was released from prison for this offense in 1995, and was on supervised release at the time of the events in question here, the prior conviction and sentence were proximate in time to the allegations underlying this case.
It is also undisputed that the girls were often left home alone in the evening. Thus, the evidence established that the Defendant had both the inclination and the opportunity to commit the crimes.
*1283Furthermore, there were three aspects of Defendant’s testimony that were flatly contradicted by other witnesses. He testified that he was never at the girls’ house, a fact contradicted by the girls and by Ms. Kalleco; he testified that the girls, when at his workplace, would never help him and never ask him for money, a fact contradicted by defense witness Bucky Belen; and he testified that he had never denied his involvement in the 1994 sexual abuse offense to the FBI, a fact contradicted by Agent Calles. As to these facts, the jury necessarily had to weigh Defendant’s credibility.
The problem facing us is the impact that the testimony of Dr. Ornelas, Ms. Baum, and Ms. Carlson might have had on the jury’s deliberations under the circumstances outlined above. The admission of the expert conclusion of a highly credible physician that the girls were in fact sexually abused, coupled with the admission of the testimony of two counselors who assumed that the girls had been abused, and stated it as a fact, undoubtedly had some effect on the jury’s deliberations. However, nearly all of Dr. Ornelas’s, Ms. Baum’s, and Ms. Carlson’s testimony was properly admitted at trial, and, as discussed above, there is no evidentiary problem with the testimony of any of the other witnesses. Thus, only a small, albeit important, portion of the testimony admitted at trial was erroneously admitted.
Defendant calls to our attention several cases in which courts, when faced with similar situations, have found that the admission of expert testimony, in sexual abuse cases, which vouches for the credibility of other witnesses constitutes reversible or even plain error, and our own research has unearthed several other such cases. See, e.g., Whitted, 11 F.3d at 786-87 (plain error); United States v. Birdsall, 47 M.J. 404, 410 (C.A.A.F.1998); see also, e.g., Commonwealth v. Colin C., 419 Mass. 54, 648 N.E.2d 19, 23 (Mass.1994); State v. Gokey, 154 Vt. 129, 574 A.2d 766, 772 (Vt.1990). But see, e.g., Brown v. State, 523 So.2d 729, 730 (Fla.Ct.App.1988) (holding the error harmless “in light of the overwhelming evidence of guilt”); Commonwealth v. Rather, 37 Mass.App.Ct. 140, 638 N.E.2d 915, 920 (Mass.App.Ct.1994) (holding the error harmless due to eyewitness corroboration of the victims’ testimony). However, after carefully studying these cases, we conclude that each of them, even the cases finding the error harmless, is in important respects distinguishable from the case at hand. These cases do not appear to compel a result, on the facts presented to us here, in one direction or the other. Our problem is compounded by the fact that the government, which has the burden of showing that the error is harmless, failed to address the issue in its brief.
However, after reading and re-reading the record, we are persuaded that the errors in admitting portions of the testimony of Dr. Ornelas, Ms. Baum, and Ms. Carlson did not affect a substantial right of a party. See Fed.R.Evid. 103(a). In light of the strength of the properly admitted testimony, outlined above, and the relatively modest amount of erroneously admitted testimony, we cannot say that the erroneously admitted portions of the testimony substantially affected the trial’s outcome, and we are not in grave doubt that the testimony had such effect. Any errors in the admission of evidence committed by the district court were harmless.
III. Insufficiency of the Evidence on Count I
Finally, we address Defendant’s final argument — that the government presented insufficient evidence to convict him on Count I'of the indictment. In conducting this inquiry, we must “determine whether the evidence, considered most favorably to the government, is sufficient to allow a reasonable jury to find [Defendant] guilty beyond a reasonable doubt” on Count I of the indictment. United States v. Perez, 959 F.2d 164, 168 (10th Cir.1992); see also United States v. Wacker, 72 F.3d 1453, 1462-63 (10th Cir.1995) (stating that all evidence is to be *1284viewed “in the light most favorable to the government” and that “[w]e reverse only if no rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt”).
Defendant asserts that there is insufficient evidence in the record which points toward abuse of D.J. “on or about” October 1995. The record evidence on this point is as follows: D.J. stated at trial that Defendant had been abusing her “[s]ince [she] was small,” III R. at 358; D.J. testified that Defendant had abused her both before he went to prison and after he got out of prison; D.J. testified that Defendant abused her “more than five times,” and that not all of these times were in 1997, id. at 394; and J.J. testified that Defendant had abused D.J. when the girls were little.
However, there is no direct evidence of even any contact between Defendant and D.J. during the entire year 1995, let alone evidence of sexual abuse during October of that year. Defendant was in prison for the first half of 1995, and D.J. testified that she could not remember any abuse in 1995, or even being with Defendant at any time in 1995. In addition, Dr. Kern testified that Defendant, due to his prostate surgery, would likely have demonstrated signs of bleeding had he .tried to ejaculate during the period between September 29, 1995, and late October 1995, and that no such signs of bleeding were apparent.
We have stated that, where, as here, time is not an element of the offense, and where “the phrase ‘on or about’ is used in an indictment in connection with a specific date ..., if the prosecution proves that the offense was committed within a few weeks of the date, the proof will be deemed sufficient to hold [the] defendant responsible for the charge.” Kokotan v. United States, 408 F.2d 1134, 1138 (10th Cir.1969). Indeed, “[a] variance between the date alleged in the indictment and the date of the commission of the offense as shown by the evidence is generally not fatal.” United States v. Harmon, 486 F.2d 363, 366 (10th Cir.1973). The important consideration is that there must be some evidence which tends to show that the defendant committed the charged offense on “a date reasonably near to the specified date” alleged in the indictment. Castillo, 140 F.3d at 885; see also United States v. Tsinhnahijinnie, 112 F.3d 988, 991 (9th Cir.1997).23 Thus, even if we assume, ar-guendo, that because of his prostate surgery Defendant could not have committed sexual abuse during the entire month of October 1995, that does not necessarily compel a finding of insufficient evidence on Count I. Evidence tending to show that Defendant committed the crime within a few weeks of — or some other interval which, under the circumstances of the case, could be considered reasonably near to — October 1995 may be enough under our precedents. See Kokotan, 408 F.2d at 1138.
However, no such evidence exists in this case. Had D.J. been able to more specifically recall abuse which could be indexed to late 1995, such as, for instance, abuse soon after Defendant’s release from prison, or abuse while D.J. was in a certain grade in school or while D.J. had a certain schoolteacher, or abuse coincident with some family event, our conclusion would likely be different. On this record, however, we are compelled to conclude that the government has not met its burden of proving Defendant guilty on Count I. See Tsinhnahijinnie, 112 F.3d at 990-92 (finding the evidence insufficient in a similar case, where the alleged victim could not recall abuse in 1992 as charged in the indictment, and there was no other evidence of abuse in 1992). Accordingly, we *1285reverse Defendant’s conviction on Count I and remand to the district court for the entry of a judgment of acquittal.
CONCLUSION
For the foregoing reasons, we conclude that the district court erred by admitting the conclusions of Dr. Ornelas that the children had in fact been sexually abused, the prejudicial statements of the counselors assuming that abuse had occurred, and the physical abuse-related conclusion of Ms. Carlson, but that these errors were harmless. Furthermore, we conclude that the government presented insufficient evidence to,support the jury’s verdict as to Count I of the indictment. Accordingly,we REVERSE Defendant’s conviction on Count I, and REMAND the case to the district court with instructions to enter a judgment of acquittal on Count I.24 We AFFIRM Defendant’s convictions on all other counts.
. The names of the alleged victims are fully disclosed in the record and in the briefs. However, to the extent it may assist in preserving their privacy, we will use only the girls' initials in this opinion.
. Dr. Junkins recorded a slight irregularity at the base of D.J.'s hymen, but drew no direct conclusion from that fact. A later examination by Dr. Renee Ornelas, using magnification, showed no abnormality.
. The first four counts charged Defendant with abusing "Jane Doe A," intended to refer to D.J. I R., Tab 9. The last three counts charged Defendant with abusing "Jane Doe B," intended to refer to J.J. Id. The last three counts are "aggravated" sexual abuse because the victim, J.J., was under the age of 12 at the time of the abuse. See 18 U.S.C. § 2241(c).
. Neither the specific grounds for objection in the motion in limine nor the precise grounds *1271or terms upon which the grant of a continuing objection was claimed are beyond dispute. To the contrary, it appears that defense counsel did not uniformly present the district court, at appropriate times, with the precise objection upon which it sought a ruling. However, the government raises no issue with respect to these subjects on appeal. See Ap-pellee’s Br. at 19 n. 7. Our analysis, therefore, proceeds on the assumption that the defense made specific, timely, and proper objections to the disputed testimony and that those objections were overruled. The one exception is the testimony of the school counselor, Carolyn Joe, and Defendant concedes, as discussed below, that he failed to object to her testimony at trial.
. Rule 403 states as follows:
Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.
. Rule 414(a) states as follows:
In a criminal case in which the defendant is accused of an offense of child molestation, evidence of the defendant's commission of another offense or offenses of child molestation is admissible, and may be considered for its bearing on any matter to which it is relevant.
. Defendant's brief on this point is almost entirely devoted to arguing that Rule 414 is unconstitutional on its face, an argument foreclosed by our holding in Castillo.
. It is doubtful that Carolyn Joe, a school counselor, can even be categorized as a health care professional, and even if she can, there is some question as to whether she in fact offered any expert testimony or merely testified as a lay witness under Fed.R.Evid. 701. Her testimony chiefly dealt with D.J.'s excessive absences from school. In any event, Defendant admits that he did not object at trial to Joe's testimony. See Appellant's Reply Br. at 2; Appellee’s Br. at 19 n. 7. Accordingly, we review the district court’s decision to admit Joe’s testimony for plain error, see United States v. Enjady, 134 F.3d 1427, 1435 (10th Cir.), cert. denied, - U.S. -, 119 S.Ct. 202, 142 L.Ed.2d 165 (1998), and find none. Our discussion below is therefore confined to the testimony of the other four challenged witnesses.
. Proposed amendments to Fed.R.Evid. 701 expressly exclude testimony based on specialized knowledge. By its terms, Fed.R.Evid. 702 covers such testimony.
. Rule 702, which governs the admission of expert testimony, states as follows:
If scientific, technical, or other specialized. knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.
. The medical records, marked for identification as Government’s Exhibits 1 and 2, were extensively incorporated into the trial record through Dr. Junkins' testimony, but they were never offered or admitted into evidence as exhibits.
. Treating professionals may relate statements made to them by the alleged victims, as long as the statements were made for the purposes of diagnosis or treatment. Such statements are admissible under Fed.R.Evid. 803(4), as exceptions to the hearsay rule, provided a proper foundation is laid. See United States v. Tome, 61 F.3d 1446, 1450-51 (10th Cir.1995); United States v. Joe, 8 F.3d 1488, 1494-95 (10th Cir.1993).
. The government examined Dr. Ornelas with respect to exhibits marked for identification as Government’s Exhibits 7 and 8, which consisted of D.J.’s and J.J.’s respective records taken in connection with the Ornelas evaluation. However, those records were not offered or admitted as evidence at trial.
. The following exchange is an example.
Q. Okay. So you’re saying that painful urination would be a symptom consistent with sex abuse?
A. Yes. It certainly could be.
Ill R. at 351 (emphasis added).
. Proposed amendments to Rule 702 are currently under consideration. Should these amendments take effect, trial judges conducting the gatekeeping inquiry will be required to undertake a three-part reliability inquiry in connection with all expert testimony. Gatek-eeping judges must make certain that all expert testimony admitted at trial "is sufficiently based upon reliable facts or data” and "is the product of reliable principles and methods,” and that "the witness has applied the principles and methods reliably to the facts of the case.” Proposed Fed.R.Evid. 702, quoted in 67 U.S.L.W. 2588 (Apr. 6, 1999).
. Courts have also held that testimony which vouches for the credibility of a witness violates other evidentiary rules, such as Rule 608(a)(1) or Rule 403. Some courts have held that, although Rule 608(a)(1) “permits testimony concerning a witness’s general character or reputation for truthfulness,” it "prohibits any testimony as to a witness’s truthfulness on a particular occasion.” State v. Rimmasch, 775 P.2d 388, 391 (Utah 1989) (construing Utah R. Evid. 608); see also United States v. Azure, 801 F.2d 336, 341 (8th Cir.1986); State v. Wood, 194 W.Va. 525, 460 S.E.2d 771, 778 (W.Va. 1995) (construing W. Va. R. Evid. 608); People v. Koon, 713 P.2d 410, 412 (Colo.Ct.App.1985) (construing Colo. R. Evid. 608). And at least one court has held, in a sexual abuse case, that testimony that bolsters the credibility of the complaining witness violates Rule 403's balancing test. United States v. Funds Held in the Name or for the Benefit of John Hugh Wetterer, 991 F.Supp. 112, 120-21 (E.D.N.Y.1998).
. The government attempts to distinguish Whitted by arguing that the testifying doctor in that case was a hired expert, while the testifying doctors in this case were the girls’ actual treating physicians. This distinction misses the point. The crucial factor that separates admissible opinion testimony from impermissible vouching is the basis upon which the doctor forms his or her opinion, not the status of the doctor.
. See Whitted, 11 F.3d at 785-86; United States v. Birdsall, 47 M.J. 404, 409-10 (C.A.A.F.1998); Johnson v. State, 292 Ark. 632, 732 S.W.2d 817, 821 (Ark.1987); State v. Batangan, 71 Haw. 552, 799 P.2d 48, 52 (Haw.1990); State v. Bressman, 236 Kan. 296, 689 P.2d 901, 907-08 (Kan.1984); State v. Black, 537 A.2d 1154, 1157 & n. 1 (Me.1988); Commonwealth v. Colin C., 419 Mass. 54, 643 N.E.2d 19, 22-23 (Mass.1994); People v. Beckley, 434 Mich. 691, 456 N.W.2d 391, 407-08 (Mich.1990); Goodson v. State, 566 So.2d 1142, 1146-47 (Miss.1990); State v. Lucero, 116 N.M. 450, 863 P.2d 1071, 1075 (N.M.1993); State v. Kallin, 877 P.2d 138, 140-41 (Utah 1994); State v. Gokey, 154 Vt. 129, 574 A.2d 766, 768; 771-72 (Vt.1990); State v. Haseltine, 120 Wis.2d 92, 352 N.W.2d 673, 676 (Wis.Ct.App.1984). Other state courts, however, disagree, and have allowed such testimony to be admitted. See Broderick v. King’s Way Assembly of God, 808 P.2d 1211, 1216-17 (Alaska 1991); Glendening v. State, 536 So.2d 212, 220 (Fla.1988); State v. Hester, 114 Idaho 688, 760 P.2d 27, 31-32 (Idaho 1988); State v. Geyman, 224 Mont. 194, 729 P.2d 475, 479 (Mont.1986); Townsend v. State, 103 Nev. 113, 734 P.2d 705, 708 (Nev.1987); State v. Figured, 116 N.C.App. 1, 446 S.E.2d 838, 842-43 (N.C.Ct.App.1994); State v. Stowers, 81 Ohio St.3d 260, 690 N.E.2d 881, 883 (Ohio 1998); State v. Wilson, 121 Or.App. 460, 855 P.2d 657, 660 (Or.Ct. App.1993); Johnson v. State, 970 S.W.2d 716, 721 (Tex.App.1998). Several of these cases, however, did not address whether the vouching problem was implicated by such testimony.
. These witnesses also held themselves out as "art therapists.” Ill R. at 237, 251. The district court made no findings as to whether these witnesses qualified as expert art therapists. However, neither Ms. Baum nor Ms. Carlson offered any expert opinion based upon art therapy. Their references to pictures drawn by the girls, such as, for instance, pictures of a house where the girls said they felt afraid, were factual observations, and not interpretations or diagnoses based on specialized knowledge of art therapy.
. We note that some courts have held that a counselor's testimony that "[the alleged victim] was referred to me for sexual abuse recovery counseling” constitutes impermissible vouching and is therefore inadmissible. State v. Haslam, 663 A.2d 902, 905-06 (R.I.1995); cf. Commonwealth v. McCaffrey, 36 Mass.App.Ct. 583, 633 N.E.2d 1062, 1067-68 (Mass.App.Ct.1994). We disagree with the analysis of these courts. A statement by a counselor that the alleged victim was referred for sexual abuse counseling is not necessarily the same as a statement that the victim is telling the truth, or even that the victim was in fact sexually abused, and therefore does not necessarily rise to the level of impermissible vouching.
. At least, no foundation for these statements appears in the record. It could be that the counselors were making these statements based on their belief that the girls were telling the truth. In that case, their testimony would constitute impermissible vouching. See Whit-ted, 11 F.3d at 785. It could be that the counselors were inartfully trying to relate to the jury out-of-court statements made to them by the girls in the course of diagnosis or treatment, admissible under Fed.R.Evid. *1282803(4). See Tome, 61 F.3d at 1449-50. But they do not come across as such in the record.
. The dissent agrees with us that the proper harmless error standard is different from the standard used in insufficiency of the evidence cases, and reemphasizes the point. As indicated above, we expressly apply the Kotteakos harmless error standard here. While we, like the dissent, are not totally free from doubt about whether the erroneously admitted testimony may have had some influence on the outcome of the case, we part company from the dissent because, as discussed below, we do not have grave doubt that the errors had a substantial effect on the outcome. Grave doubt, by definition, does not include doubt at every level, and substantial influence, by definition, does not mean any or some influence.
. Often, in determining whether the date of the evidence in the case is "reasonably near” to the date alleged in the indictment, courts will consider whether the defendant's "substantial rights” were "prejudiced” by the variance. See Tsinhnahijinnie, 112 F.3d at 991; United States v. Brewer, 1 F.3d 1430, 1437 (4th Cir.1993). We think it evident that Defendant was prejudiced by the temporal variance between the evidence and the indictment in this case.
. Acquittal on Count I will have no effect on the sentence imposed in this case. Defendant was sentenced to concurrent sentences of 240 months on each of Counts I-IV, and to concurrent sentences of life imprisonment on each of Counts V-VII. Even with the acquittal on Count I, Defendant still must serve, concurrently, three life sentences and three sentences of 240 months' imprisonment.