**186 A Lancaster County jury found Kenneth Campbell met the statutory definition of a sexually violent predator (SVP) under South Carolina's SVP Act, S.C. Code Ann. §§ 44-48-10 to -170 (2018). Campbell appealed, and the court of appeals affirmed. In re Care & Treatment of Campbell , Op. No. 2016-UP-198, 2016 WL 2757342 (S.C. Ct. App. filed May 11, 2016). On certiorari, Campbell contends the court of appeals erred in affirming his civil commitment because the State inappropriately impeached the credibility of Campbell's expert witness by introducing evidence of a recent arrest warrant for an unrelated sex offender whom the expert had opined was unlikely to reoffend. We find the admission of testimony about a mere arrest warrant of an unrelated individual in a collateral matter unduly prejudiced Campbell and, therefore, reverse and remand for a new commitment proceeding.

I. Factual / Procedural History

The State referred Campbell to the SVP program due to his four alleged sexual assaults of three minor children with whom Campbell slept in the same house. For two of the assaults, the four-year-old victims recanted, and the State either dropped the charges or declined to press charges. For the remaining two assaults, one of which was committed while Campbell was **187 out on bond for the other, Campbell entered an Alford 1 plea to criminal sexual conduct with a minor in the first degree (CSCM-1st) and pled no contest to committing a lewd act on a child under the age of sixteen. He received an aggregate sentence of twenty years' imprisonment, suspended upon the service of twelve years' imprisonment and three years' probation.

Prior to Campbell's release, the State filed a petition pursuant to the SVP Act seeking Campbell's civil commitment for long-term control, care, and treatment. See S.C. Code Ann. § 44-48-30 (1) (defining an SVP as "a person who: (a) has been convicted of a sexually violent offense; and (b) suffers from a mental abnormality or personality disorder that makes the person likely to engage in acts of sexual violence if not confined in a secure facility for long-term control, care, and treatment"). The trial court made a determination of probable cause and appointed Dr. Marie Gehle to perform a psychiatric evaluation of Campbell. Dr. Gehle diagnosed Campbell with pedophilia but opined he was not at a high risk to reoffend. The State then obtained an independent evaluation from Dr. Ana Gomez.

At the jury trial, Dr. Gomez testified on behalf of the State and was qualified as an expert in psychiatry and forensic psychiatry. Dr. Gomez stated that after interviewing Campbell, conducting seven different psychiatric tests that accounted for various risk factors for reoffending, and examining the pertinent records in his file, she diagnosed him with pedophilic disorder, non-exclusive, indicating he was attracted to children of both sexes. She explained pedophilic disorder cannot be cured but can be managed through appropriate strategies and intervention.

*17 Dr. Gomez testified Campbell's proposed strategies to avoid reoffending-including finding religion and "walking away" from children any time he was around them-were wholly unrealistic. Additionally, Dr. Gomez expressed concern over Campbell's refusal to seek sex offender treatment while incarcerated, which is in itself a significant risk factor for reoffending. Dr. Gomez testified Campbell's pedophilic disorder caused **188 him serious difficulty in controlling his behavior, and his lewd act offense-committed while out on bond for the CSCM-1st offense-indicated his difficulty in controlling his behavior was ongoing.

As a result, Dr. Gomez opined Campbell was extremely likely to reoffend if he was not civilly committed. Further, Dr. Gomez testified the potential risk Campbell posed to future child victims was more imminent because, after his release, Campbell planned to live with his sister, and her grandchildren and great-grandchildren would frequently be sleeping in the same house as Campbell, as had his previous victims. In conclusion, Dr. Gomez testified it was her medical opinion Campbell met the criteria for designation as an SVP and he was in need of long-term control, care, and treatment at a secure facility.

Dr. Gehle then testified on behalf of Campbell and was qualified as an expert in forensic psychiatry. Dr. Gehle stated that after performing a similar interview and review of Campbell's file, she had also diagnosed him with pedophilic disorder, non-exclusive type. Dr. Gehle explained that in coming to her diagnosis, she had used only one of the seven psychiatric tests performed by Dr. Gomez. However, on that test, both doctors scored Campbell in the low- to moderate-risk group for reoffending, which equated to a rate of reoffending of 15.8% in the next five years and 24.3% in the next ten years, approximately the average rate for reoffending for all sex offenders.

Dr. Gehle testified that although she agreed with much of Dr. Gomez's testimony and diagnosis, she disagreed Campbell was likely to reoffend. Dr. Gehle stated not every person convicted of a sex offense posed a high risk to the public upon his or her release, and Campbell's lack of prison referrals or disciplinary problems showed his ability to control his behavior on a day-to-day basis. Dr. Gehle also testified she was less concerned than Dr. Gomez about Campbell living with his sister and young children upon his release because it was unclear to her whether the children would actually spend the night or merely visit while other adult family members were present. Dr. Gehle then opined that, were Campbell to be released, there were safeguards in place to protect the public, such as placing him on the sex offender registry and monitoring **189 him while on probation, including through the use of a GPS anklet. Ultimately, Dr. Gehle concluded that while Campbell suffered from a mental abnormality, there was insufficient evidence to believe Campbell's abnormality made him likely to reoffend, and he therefore did not meet the criteria for designation as an SVP.

On cross-examination, the State questioned Dr. Gehle's exclusive reliance on the results of the single psychiatric test, particularly when Dr. Gomez had testified the test did not account for all of the risk factors associated with sexually reoffending, nor was the test intended by its creators as a stand-alone assessment. Furthermore, at the State's prompting, Dr. Gehle conceded Campbell had "meaningful risk factors" for reoffending, including a dysfunctional coping style, a resistance to rules and supervision, and a refusal to receive mental health treatment unless it was court-ordered. Finally, Dr. Gehle testified Campbell did not have an "ideal relapse prevention plan" due to his failure to receive sex offender treatment and his post-release "access to children" who would sleep in the same house as Campbell, similar to his prior victims. However, Dr. Gehle stated she gave Campbell credit for claiming he would "walk away" from children and not be around them in "that way."

On re-direct examination, Campbell attempted to rehabilitate Dr. Gehle's methodology for evaluating SVPs, emphasizing Dr. Gehle's vast experience in evaluating SVPs and having her reiterate her opinion that Campbell was unlikely to reoffend.

*18 On re-cross examination, the State's attorney asked if Dr. Gehle had ever wrongly opined an SVP candidate was unlikely to reoffend, to which Dr. Gehle responded she did not know. The State's questioning then focused on Dr. Gehle's pre-commitment evaluation of an unrelated sex offender, Michael Thomas. In doing so, the State's attorney handed Dr. Gehle her report on Thomas and requested Dr. Gehle read the portion of the report aloud where she had opined Thomas was unlikely to reoffend and, therefore, should not be civilly committed as an SVP. The State's attorney next handed Dr. Gehle an arrest warrant for Thomas dated approximately six months before Campbell's commitment proceeding and requested Dr. **190 Gehle read portions of the arrest warrant into the record. The arrest warrant stated Thomas was wanted for rape, and his DNA was a match for that of the alleged rapist. The State emphasized Thomas had reoffended within two years of his evaluation by Dr. Gehle and the resultant failure to commit Thomas as an SVP, stating that due to Dr. Gehle's error in opining Thomas should be released from custody, "another woman ha[d] been raped."

During closing arguments, the State's attorney concluded her remarks by reminding the jury that Dr. Gehle had been wrong before in opining other sex offenders were unlikely to reoffend, and that if Dr. Gomez was correct and Dr. Gehle was wrong again, Campbell was going to "get out and ... hurt another kid." Hammering that point home, she stated:

So I leave you with this: You have a person who on more than one occasion [ ] has sexually assaulted children. He takes no accountability [for] what he's done and he hasn't had sex offender treatment and he's refused it when it has been offered. He's going to go live in a house where people are going to allow him to be around children. You heard the testimony. What do you think is going to happen?

At the conclusion of the hearing, the jury found beyond a reasonable doubt that Campbell met the statutory definition of an SVP, and the trial court ordered Campbell's civil commitment. The court of appeals affirmed Campbell's commitment in an unpublished opinion.

II. Standard of Review

In general, the admission or exclusion of evidence is a matter left to the sound discretion of the trial court, whose ruling will not be reversed on appeal absent an abuse of that discretion. Carson v. CSX Transp., Inc. , 400 S.C. 221 , 229, 734 S.E.2d 148 , 152 (2012). Likewise, the scope of cross-examination is largely within the trial court's discretion. Bunch v. Charleston & W.C. Ry. Co. , 91 S.C. 139 , 142, 74 S.E. 363 , 364 (1912). "An appellate court will not disturb a trial court's ruling concerning the scope of cross-examination of a witness to test his or her credibility, or to show possible bias or self-interest in testifying, absent a manifest abuse of discretion." Yoho v. Thompson , 345 S.C. 361 , 365, 548 S.E.2d 584 , 585 (2001).

**191 "An abuse of discretion occurs when the trial court's ruling is based on an error of law or, when grounded in factual conclusions, is without evidentiary support." Clark v. Cantrell , 339 S.C. 369 , 389, 529 S.E.2d 528 , 539 (2000). "To warrant reversal based on the admission or exclusion of evidence, the appellant must prove both the error of the ruling and the resulting prejudice, i.e., that there is a reasonable probability the jury's verdict was influenced by the challenged evidence or the lack thereof." State v. Commander , 396 S.C. 254 , 263, 721 S.E.2d 413 , 418 (2011) ; State v. Colf , 337 S.C. 622 , 625, 525 S.E.2d 246 , 247-48 (2000) ("The scope of cross-examination is within the discretion of the trial judge, whose decision will not be reversed on appeal absent a showing of prejudice.").

III. Discussion

Campbell argues the court of appeals erred in affirming the trial court's decision to allow the State to cross-examine Dr. Gehle about Thomas's arrest warrant because the arrest warrant was irrelevant to the ultimate issue in the case, i.e., whether Campbell was an SVP. Campbell additionally contends the admission of testimony about Thomas's arrest warrant was more prejudicial than probative, encouraging the jury to make its decision *19 based on fear of Campbell rather than whether he met the statutory criteria to be declared an SVP. While we find the arrest warrant fell within the broad scope of relevant evidence, we agree that, under these facts, the arrest warrant was more prejudicial than probative.

A. Relevance

Relevant evidence is that evidence having any tendency to make the existence of any fact of consequence to the ultimate determination of the action more or less probable than it would otherwise be without the evidence. Rule 401, SCRE. Considerable latitude and discretion must be allowed the trial court in determining the relevance and admissibility of impeachment evidence. State v. Williams , 263 S.C. 290 , 302, 210 S.E.2d 298 , 304 (1974). As a result, " 'anything having a legitimate tendency to throw light on the accuracy, truthfulness, and sincerity of a witness may be shown and considered in determining the credit to be accorded his testimony.' " State v. Jones , 343 S.C. 562 , 570, 541 S.E.2d 813 , 817 (2001) (citation omitted); see also Rule 611(b), SCRE ("A witness may be **192 cross-examined on any matter relevant to any issue in the case, including credibility.").

Here, Thomas's arrest warrant was a collateral matter because it could not have been presented during the State's case-in-chief to prove Campbell was an SVP. See State v. Bailey , 279 S.C. 437 , 439-40, 308 S.E.2d 795 , 797 (1983) (citations omitted) (holding evidence that the defendant's father and brother attempted to procure perjured testimony was improperly admitted because it could not have been presented as part of the State's case-in-chief and, as a result, was collateral to the defendant's guilt or innocence). However, given the "considerable latitude" with which we must review the trial court's relevance determination, we find Thomas's arrest warrant was in fact relevant to assist the jury in determining the weight to afford Dr. Gehle's testimony, and specifically her opinion as to whether Campbell was likely to reoffend. See, e.g. , Rule 611(b), SCRE (stating a witness may be cross-examined as to any matter related to any relevant issue, including credibility); Jones , 343 S.C. at 570 , 541 S.E.2d at 817 (holding any evidence that shows the accuracy, truthfulness, or sincerity of a witness may be admissible to help the factfinder determine the witness's credibility). As a result, the State was permitted to ask Dr. Gehle whether she was aware if she had ever erred in her SVP evaluations, and whether she was aware of a specific error. 2

B. Unfair Prejudice

Even when relevant, evidence may be inadmissible if its probative value is substantially outweighed by the danger **193 of unfair prejudice. Rule 403, SCRE ; State v. Wiles , 383 S.C. 151 , 158, 679 S.E.2d 172 , 176 (2009). "Evidence is unfairly prejudicial if it has an undue tendency to suggest a decision on an improper basis, such as an emotional one." State v. Wilson , 345 S.C. 1 , 7, 545 S.E.2d 827 , 830 (2001) (citing State v. Alexander , 303 S.C. 377 , 382, 401 S.E.2d 146 , 149 (1991) ). The determination of prejudice must be based on the entire record and will generally turn on the facts of each case. State v. Stokes , 381 S.C. 390 , 404, 673 S.E.2d 434 , 441 (2009).

Prior to conviction, a defendant is presumed innocent in the eyes of the law. See State v. Posey , 269 S.C. 500 , 503, 238 S.E.2d 176 , 177 (1977) (declaring such a statement "elementary"). Thus, a mere arrest warrant in no way proved Thomas in fact committed the offense for which he was arrested and, standing alone, could only have *20 a minimal impact on Dr. Gehle's credibility. Accordingly, while it was relevant that Dr. Gehle may have opined in error that Thomas was unlikely to reoffend, Thomas's arrest warrant had very low probative value as to whether Campbell was an SVP.

In contrast, the manner in which the State used the arrest warrant was highly prejudicial to Campbell. The State grossly mischaracterized the results of Dr. Gehle's evaluations of Thomas and Campbell. As Drs. Gomez and Gehle both testified, the results of the common psychological test they performed on Campbell indicated he was in the low- to moderate-risk group. Individuals in that group have a rate of reoffending of 15.8% in the next five years and 24.3% in the next ten years, a rate which is consistent for that of all sex offenders. Therefore, by concluding neither Thomas nor Campbell met the statutory criteria to be classified as an SVP, Dr. Gehle did not guarantee either man would never reoffend. Rather, Dr. Gehle concluded the men were unlikely to reoffend. However, the State asserted otherwise in its cross-examination and closing argument, stating that, based on an arrest warrant alone , Dr. Gehle had been "wrong" in her evaluation of Thomas; that as a result, "another woman ha[d] been raped;" and that Campbell was likewise bound to "hurt another kid."

Moreover, the State exacerbated the prejudicial effect of the arrest warrant during its closing argument. Specifically, the State emphasized that on multiple occasions, Campbell had **194 assaulted children sleeping in the same house as him, refused to accept responsibility for his actions, declined to receive sex offender treatment while imprisoned, and was planning to live in a house in which minor children would regularly sleep. The State's attorney then asserted, "You heard the testimony. What do you think is going to happen?" 3 This rhetorical question-the last statement the jury heard prior to its deliberations-was a naked attempt by the State to appeal to the jurors' emotions and cloud their ability to impartially weigh the evidence. See Tappeiner v. State , 416 S.C. 239 , 254 n.8, 785 S.E.2d 471 , 478 n.8 (2016) (finding an emotional plea to jurors that an accused rapist was a bad actor and could not be trusted to watch the jurors' own family members had a strong prejudicial impact on jurors' impartiality in part because "the emotional plea was the very last thing the jury heard before beginning its deliberations ... [and therefore] was likely at the forefront of the jurors' minds when beginning their discussions"); Von Dohlen v. State , 360 S.C. 598 , 609, 602 S.E.2d 738 , 744 (2004) (holding the State must tailor its closing arguments so as not to appeal to the personal biases of the jury or arouse the jurors' passions or prejudices).

Weighing the minimal probative value of Thomas's arrest warrant against the prejudice resulting from the State's mischaracterization of the import of the warrant, we hold the admission of the warrant unfairly prejudiced Campbell because it had an undue tendency to suggest a decision on an improper basis, namely fear he would reoffend and harm another child. See Wilson , 345 S.C. at 7, 545 S.E.2d at 830 (citing Alexander , 303 S.C. at 382 , 401 S.E.2d at 149 ) (stating evidence is unduly prejudicial if it suggests a decision on an emotional basis, rather than a factual one). Additionally, we find the error was not harmless because the case against Campbell amounted to a "battle of the experts," in which Dr. Gomez opined Campbell presented an imminent danger to the community, and Dr. Gehle opined Campbell was unlikely to reoffend. As such, we find the improper denigration of Dr. Gehle's credibility was reasonably likely to have affected the **195 outcome of the trial. Cf. Tappeiner , 416 S.C. at 253-54 , 785 S.E.2d at 478-79 (finding the State's improper vouching for the victim's credibility and its emotional appeal that the defendant was a bad actor who could not be trusted to watch the jurors' own family members was not harmless error because the case was "entirely dependent on a credibility determination between the prosecution's witnesses and the defense's witness," and therefore, it was "likely the emotional plea, particularly in conjunction with the solicitor's *21 improper vouching for Victim's credibility, swayed the jurors' view of the facts and resolution of the contradictions in the witnesses' testimonies").

IV. Conclusion

The decision of the court of appeals upholding Campbell's SVP status and his involuntary commitment is reversed and remanded for a new commitment proceeding.

REVERSED AND REMANDED.

KITTREDGE, HEARN, FEW and JAMES, JJ., concur.