delivered the Opinion of the Court.
¶1 This ease, a companion to People v. Jacobson, 2017 CO 28, 393 P.3d 548, requires us to determine whether a trial court abused its discretion by refusing to poll the jury about whether jurors had seen a news report about the case that had been posted online and ran in a local newspaper. Here, the trial court gave repeated admonitions not to seek out news about the case—including just before the newspaper released the story. Thus, the trial court did not abuse its discretion by refusing to poll the jury after a newspaper published a prejudicial news report with limited distribution. We therefore reverse the court of appeals and affirm the defendant’s conviction.
I. Facts and Procedural History
¶2 Defendant Emmett Andrew Larsen is the father of S.L. and the grandfather of S.L.’s children, A.H. and K.H. S.L., along with A.H. and K.H., moved to Colorado to live with Larsen. While in Colorado, K.H. told a therapist that her uncle, T.J., had sexually abused her. A Department of Human Services (“DHS”) caseworker interviewed both A.H. and K.H.
¶3 During the interview, K.H. said T.J. had abused her but did not accuse Larsen of abusing her. A.H., on the other hand, told the caseworker that Larsen had touched her breasts and vaginal area. In a follow-up interview, A.H. again accused Larsen of abusing her. K.H. again stated that Larsen had never touched her inappropriately. Larsen was arrested, and the People charged Larsen with one count of sexual assault on a child by one in a position of trust as part of a pattern of abuse and two counts of sexual assault on a child by one in a position of trust.
¶4 During jury selection, the trial court admonished the prospective jurors to avoid media coverage of the case:
You must not read, view, or listen to any reports about the case in the press or on radio, television, or any other type of electronic media. And this includes the Internet. You are not to Google anything about this case. You are not to do any individual investigation about anything.
The trial court, after the jury was sworn in, again admonished the jury members to avoid media coverage of the ease:
[D]o not read or listen to any accounts or discussions of the ease that may be reported by newspaper or other publications or by television or radio. And that also includes any Internet information, any court blogs. You are not to go Google anything about this case.
¶5 Then, at trial, AH. testified—in contrast to her interviews—that Larsen had touched her breasts but not her vaginal area. K.H. testified—also in contrast to her interviews—that Larsen had, in fact, inappropriately touched her breasts. In a transcribed phone call between Larsen and S.L. admitted at trial, Larsen admitted that he had touched K.H.’s breast. The trial court also admitted a recorded phone call in which Larsen admitted that he had inappropriately touched his daughter when she was young.
¶6 After the close of evidence, but before the trial court dismissed the jurors for the weekend, the trial court noted that a news reporter was in attendance. The trial court took this as an opportunity to again admonish the jury to stay away from news reports about the ease:
Also want to let you know that there was a newspaper reporter in the courtroom. And I don’t know if there is anything likely to be in the newspaper, but I will remind you that you are not to read anything in the newspaper about this case. I would also instruct you that if you have loved ones at home that start—who may read the article and start talking to you that you are to avoid any discussion about anything that may be in the newspaper or in any other media.
¶7 After the weekend, the trial court was notified by defense counsel that an article about the case had been posted to the Internet and printed in a local newspaper. Defense counsel then asked the trial court to poll the jury about whether any members had read the article. The trial court declined to do so, stating that (1) it trusted the jury; (2) if any of the jurors had read it they would say so; and (3) any questions about a specific *546article may generate curiosity and make it more likely that a juror would seek it out.
¶8 The jury convicted Larsen of the sexual assault of A.H. (both the pattern of abuse and position of trust charges) and acquitted him of the charge involving K.H. The trial court sentenced Larsen to eight years to life in prison. The court of appeals reversed Larsen’s conviction, holding that the trial court abused its discretion under Harper, 817 P.2d 77, by refusing to poll the jury. People v. Larsen, 2015 COA 157, ¶ 2, — P.3d -. The People sought certiorari, which we granted.1 We now (1) hold that the trial court did not abuse its discretion; (2) reverse the court of appeals; and (3) affirm Larsen’s convictions.
II. Analysis
¶9 Trial courts have “broad discretion in deciding the ultimate issue of whether the media reports prejudiced the defendant’s right to a fair trial.” Harper, 817 P.2d at 83-84. Therefore, we review the trial court’s decision not to poll the jury for an abuse of discretion.
¶10 In Harper, this court laid out a three-step process for trial courts to follow when a prejudicial news report comes out during the course of a trial that contains otherwise inadmissible evidence. Id. at 83. First, the trial court should determine whether the coverage is actually prejudicial and contains inadmissible evidence. Id. Second, if the report could prejudice the jury, the court should canvass the jury to learn if any of them learned of any impermissible information. Id. Finally, if any jurors admit to learning of outside information, the court should question them outside the presence of other jurors to learn what effect the prejudicial information had on the juror. Id. The issue in this case focuses on the first prong.
¶11 On the first prong, when determining whether a report may have prejudiced the jury, courts should consider five relevant factors: (1) whether the report contained inadmissible, prejudicial information; (2) how closely related the report is to the matters in the trial; (3) the timing of the report; (4) the likelihood the jury was exposed to it; and (5) the likely effectiveness of any instruction not to read, watch, or listen to reports “in light of the nature and manner of dissemination of the news reports.” Id. at 84.
¶12 The court in Harper took issue with the common presumption that jurors follow a trial court’s instruction, reasoning that the presumption was not applicable in a jury exposure situation because prejudicial information “could have come to the attention of a juror inadvertently.” Id. at 82. However, this is true only where it may “reasonably be believed” that the report came to the jury’s attention. Id. at 84
¶13 The report at issue in this case was entitled, “Custody fight reveals a highly dysfunctional family.” The report contained the following information: (1) one of the victims testified that she and her sister were currently living with Larsen; (2) this testimony prompted an attempt by child welfare workers to place the victims in protective custody; (3) the trial judge stopped the child welfare workers from removing the victims from Larsen’s home despite allegations of sexual abuse because a no-contact order was never issued; (4) a summary of the allegations against Larsen; (5) the victims’ mother’s testimony, including that she did not believe that the victims were abused, despite her *547claims that Larsen had molested her as a child; and (6) the victims’ mother’s testimony that Larsen had abused her beginning at age two and that the abuse included “full intercourse.” The court of appeals held that this last piece of information was prejudicial and thus required the trial court to poll the jury, Larsen, ¶¶ 19-22.
¶14 Although it was already before the jury that Larsen had admitted to molesting his daughter, he only admitted to inappropriate touching around the time his daughter was fifteen years old. In contrast, the news report stated that Larsen’s daughter accused him of abusing her beginning at age two and that the abuse involved full intercourse—far more prejudicial accusations than what Larsen had admitted.
¶15 Therefore, the news report contained prejudicial information relevant to the trial, and it was broadcast during the trial. Those factors favor a finding of error. However, outweighing those factors is the low likelihood that jurors read or were exposed to the reports. The reports had limited distribution—they were printed only in one local newspaper and on one website. Further, the trial court gave several specific admonitions to the jury to avoid news coverage of the case, including on the Friday before the weekend during which the newspaper published the report in print and on the internet. “With repetitive, specific admonitions and a limited distribution, there was no reasonable likelihood that the jurors had been exposed to prejudicial news reports.” Jacobson, ¶ 13. Also present here, similar to Jacobson, is the trial court’s concern that additional questions to jurors about a specific news report may have been counterproductive—i.e., that by questioning the jurors about the report it may have encouraged jurors to seek out the report. Id. Therefore, the trial court did not abuse its discretion when it refused to poll the jury about the prejudicial news report.
¶16 Further, in the alternative, had a juror been exposed to the prejudicial information and therefore decided Larsen had a propensity for sexually assaulting young female relatives, it would have been reflected by across-the-board guilty convictions. But the jury convicted Larsen only of sexually assaulting one of the victims, not both. See Martin v. People, 738 P.2d 789, 795-96 (Colo. 1987) (holding that a split verdict is an indication that prejudice did not affect the jury’s verdict).
III. Conclusion
¶17 The trial court did not abuse its discretion when it declined to poll the jury following a prejudicial news report. Therefore, we reverse the court of appeals and affirm Larsen’s convictions.
JUSTICE HOOD concurs in part and dissents in part, and JUSTICE MÁRQUEZ and JUSTICE GABRIEL join in the concurrence in part and dissent in part..Specifically, we granted certiorari on three questions:
1. Whether the court of appeals erred in holding that under Harper v. People, 817 P.2d 77 (Colo. 1991) the trial court had abused its discretion by failing to poll the jury.
2. Whether the court of appeals erred in holding that under Harper admonitions alone were not sufficient to neutralize the potential for unfair prejudice and requiring trial courts to poll jurors anytime a media report is published mid-trial.
3.Whether the court of appeals erred in holding that the proper remedy for a finding that the jury should have been polled concerning a prejudicial mid-trial news article is a remand for a new trial and not a remand to determine whether the jurors were actually exposed to the article.
Because we find the first question dispositive, we decline to reach the second and third questions.