IN RE: the Care and Treatment of Douglas GIRARD. In the Matter of the Care and Treatment of Eugene Mallard.
Nos. 103,505, 103,506.
-- January 11, 2013
Sam S. Kepfield, of Hutchinson, argued the cause and was on the briefs for appellants.Kristafer R. Ailslieger, deputy solicitor general, argued the cause, and was on the brief for appellee.
If expert opinion testimony based on scientific methods or procedures is offered as evidence in Kansas state courts, the offering party must satisfy the Frye test. See Frye v. United States, 293 F. 1013 (D.C.Cir.1923). That test requires a “showing that the basis of a scientific opinion is generally accepted as reliable within the expert's particular scientific field .” State v. Shadden, 290 Kan. 803, 819, 235 P.3d 436 (2010).
In these proceedings to declare Douglas Girard and Eugene Mallard sexually violent predators, they ask us to instead apply the Daubert test to the actuarial risk assessments used by the State's expert witnesses in helping to predict the odds of their reoffending. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). The Daubert test essentially makes Frye's test of “general acceptance” simply one factor to be considered in the admissibility calculus. 509 U.S. at 594.
In both their cases, the district court essentially ruled that Frye applied to the actuarial risk assessments. It admitted the scientific opinion testimony which was partially based on these statistical calculations of risk. The Court of Appeals affirmed, with a majority holding that neither Frye nor Daubert applied because the actuarial assessments were not scientific. In re Girard, 45 Kan.App.2d 1109, 1111, 257 P.3d 1256 (2011). We granted the defendants' consolidated petitions for review and have jurisdiction under K.S.A. 20–3018(b).
We hold Frye applies and the actuarial risk assessments survive Frye's scrutiny. The district court therefore is affirmed.
Girard and Mallard were both convicted of aggravated indecent liberties with a child. The State filed petitions for their continued confinement as sexually violent predators under K.S.A. 59–29a01 et seq. To obtain this designation the State was required to prove, among other things, that the men were likely to commit repeat acts of sexual violence because of a mental abnormality or personality disorder. See K.S.A. 59–29a02(a). Both men filed motions in limine to exclude the State's expert witnesses' testimony predicting the odds of Girard and Mallard reoffending. The motions were heard at their respective commitment hearings.
The State's psychologists evaluated both men and opined they met the criteria to be considered sexually violent predators. Their evaluations relied in part on two actuarial risk assessment instruments, the STATIC–99 and the MnSOST–R. The evaluators used these instruments to determine the rate, expressed as a percentage, at which offenders with characteristics similar to Girard and Mallard had reoffended. The assessments themselves do not expressly provide a recidivism estimate for the particular offender being evaluated. According to the State's expert, John Reid, the instrument-based assessments are simply among the factors considered by the evaluators in their sexual predator determinations for the defendants. Other factors include treatment reports, mental health and criminal records, and personal interviews.
In both cases, Reid testified that the actuarial risk assessment instruments, and their use, are generally accepted as reliable within the psychological community. In Girard's hearing, Reid testified that according to a clinical research study, 95.1% of evaluators reported using such instruments “most of the time” or “always.” According to Reid, 73.2% of evaluators rated actuarial assessments as “essential” to an evaluation.
In Mallard's hearing, Reid testified that the MnSOST is generally accepted in the psychological community and that “[t]here is lots of literature” that says the MnSOST–R should be used. Mallard's expert psychologist, Stanley Irving Mintz, countered that actuarial risk assessments are “controversial” because there is “a wide range of differences of opinion” about them. While Mintz testified that he does not use such assessment instruments, he nevertheless conceded that they “are widely used by other psychologists and psychiatrists in many institutions obviously in Kansas and others” and “are used at Larned and elsewhere, Canada, so forth.” For Mallard, the district court held that the Frye test applied to the instrument-based assessments. But it also held that they were admissible under either Frye or Daubert.
For Girard, the district court held that the instrument-based assessments were admissible independent of Frye or Daubert because they were not scientific tests but statistical analyses of various factors. In the alternative, the court held the assessments were admissible because Frye applied and they met its requirements.
After the court rejected the defendants' arguments urging application of the Daubert test to exclude the psychologists' opinion testimony, it found that Girard and Mallard both met the statutory criteria of a sexually violent predator. Both men were committed to Larned State Hospital's Sexually Violent Predator Treatment Program.
After consolidation of the appeals, the Court of Appeals affirmed. The majority held that Frye governs the admissibility of expert scientific opinion in Kansas, and it therefore would “not apply the tests set forth in Daubert until instructed to do so” by the Supreme Court. In re Girard, 45 Kan.App.2d at 1111. But after examining decisions from some other jurisdictions, it also essentially ruled the use of actuarial risk assessment instruments was not scientific. See, e.g., State ex rel. Romley v. Fields, 201 Ariz. 321, 35 P.3d 82 (Ct.App.2001). Rather, they “merely help the professional draw inferences from historical data or the collective experience of other professionals who have assessed sex offenders for potential recidivism.” 45 Kan.App.2d at 1112 (citing, Fields, 201 Ariz. at 328). As a result, no test applied.
Unlike his colleagues, he believed that opinion testimony based on the instrument-based assessments should be subject to the Frye test. But he agreed with the majority's result because he concluded that Girard and Mallard did not challenge on appeal the assessments' admissibility under Frye. 45 Kan.App.2d at 1112 (Malone, J. concurring).
We granted Girard's and Mallard's petition for review to address their issue of first impression.
Issue: Actuarial risk assessments are subject to Frye.
To establish that an individual is a sexually violent predator under the Kansas Sexually Violent Predator Act, K.S.A. 59–29a01 et seq., the State must prove, among other things, that the individual was “convicted of or charged with a sexually violent offense and who suffers from a mental abnormality or personality disorder which makes the person likely to engage in repeated acts of sexual violence.” K.S.A. 59–29a02(a). The narrow issue on appeal concerns the admissibility of the State's expert witness testimony opining about the defendants' odds of reoffending through use of actuarial risk assessment instruments.
Standard of review
The admission of expert testimony is generally subject to an abuse of discretion standard. Shadden, 290 Kan. at 819. But de novo review is appropriate here because Girard and Mallard ask us to reconsider the application of the Frye standard raising an “abstract question of law.” Kuhn v. Sandoz Pharmaceuticals Corp ., 270 Kan. 443, 455–56, 14 P.3d 1170 (2000) (whether district court failed to correctly apply Frye standard presents an “ ‘abstract question of law’ “ and de novo review appropriate). See also Shadden, 290 Kan. at 819; Martinez v. Milburn Enterprises, Inc., 290 Kan. 572, 579, 233 P.3d 205 (2010) (this court exercises de novo review when reviewing a district court's legal conclusions).
Kansas uses the Frye test
Kansas has “long recognized that the test for the admission of scientific evidence is that applied in Frye v. United States, 293 F. 1013 (D.C.Cir.1923).” State v. Heath, 264 Kan. 557, 577, 957 P.2d 449 (1998); see Shadden, 290 Kan. at 819. The Frye test requires that before expert scientific opinion may be admitted into evidence, the basis of the opinion must be generally accepted as reliable within the expert's particular field. See Shadden, 290 Kan. at 819; State v. Graham, 275 Kan. 176, 184, 61 P.3d 662 (2003).
In both cases the district court ultimately concluded that the two actuarial risk assessment instruments passed the “general acceptance” standard of Frye. Under the facts of their cases, we agree. The State's expert testified that the instruments are generally accepted and widely used. Even Mallard's expert conceded that the instruments are “widely used by other psychologists and psychiatrists in many institutions.” Their testimony is consistent with findings of other state courts. See Roeling v. State, 880 So.2d 1234, 1238–39 (Fla.App.2004) (all experts including the appellant's expert testified at trial that actuarial assessments are widely used in evaluating possible sexually violent predators); In re Commitment of Simons, 213 Ill.2d 523, 535, 821 N.E.2d 1184 (2004) (noting that experts in at least 20 states “rely upon actuarial risk assessment in forming their opinions of sex offenders risks of recidivism”).
We also recognize that a number of other state appellate courts have generally concluded that actuarial risk assessments pass the Frye test scrutiny. See Simons, 213 Ill.2d. at 536 (collecting cases holding that actuarial risk assessments meet the Frye general acceptance test); In re Commitment of R.S., 173 N.J. 134, 136, 801 A.2d 219 (2002); Ortega–Mantilla v. State, 898 So.2d 1164, 1168 (Fla.App.2005). Additionally, the general acceptance of actuarial risk assessment has been thoroughly litigated in several states. Simons, 213 Ill.2d. at 537. And “the academic literature contains many articles confirming the general acceptance of actuarial risk assessment by professionals who assess sexually violent offenders for risk of recidivism.” Simons, 213 Ill.2d. at 541.
Girard and Mallard correctly concede that the general trend seems to be to uphold admissibility of such evidence under Frye. See also Janus and Prentky, Forensic Use of Actuarial Risk Assessment with Sex Offenders: Accuracy, Admissibility and Accountability, 40 Am.Crim. L.Rev. 1443, 1471, 1497 (2003) (actuarial risk assessments are admissible under either Frye or Daubert ). But see Fields, 201 Ariz. at 328 (Frye test does not apply to actuarial assessments because they are not novel scientific evidence.).
This is an issue of first impression in Kansas. But we find similarities between actuarial risk assessments and some other types of evidence to which Kansas appellate courts have applied Frye. See, e.g., Kuhn v. Sandoz Pharmaceuticals Corporation, 270 Kan. at 460 (collecting “Kansas Frye test cases [which] have addressed a variety of scientific techniques”). Particularly persuasive are appellate decisions concerning testing or assessing through the use of probabilities such as State v. Isley, 262 Kan. 281, 936 P.2d 275 (1997).
In Isley, this court rejected defendant's challenge to the admission of DNA evidence which argued that the derivation of statistical probabilities based upon an expert witness' chemical analysis did not meet the Frye test. The court first noted that “DNA testing meets the Frye test in Kansas.” 262 Kan. at 285. It ultimately concluded “the statistical analysis portion of both RFLP [restriction fragment length polymorphism] and PCR [polymerase chain reaction] testing meets the Frye test.” (Emphasis added.) 262 Kan. at 290.
Similarly, in Smith v. Deppish, 248 Kan. 217, 813 P.2d 929 (1991), this court held that DNA profiling results through RFLP analysis based upon probability calculations was subject to Frye. See State v. Hill, 257 Kan. 774, 895 P.2d 1238 (1995) (PCR analysis met Frye test). Indeed, several commentators have compared DNA typing and its probabilities to actuarial assessments of sexually violent predators. See 40 Am.Crim. L.Rev. at 1490–91.
Kansas courts have also applied Frye to blood analysis and accompanying statistical probability testimony. See State v. Washington, 229 Kan. 47, 55, 622 P.2d 986 (1981) (Frye applied to Multi-system method of blood analysis of polymorphic enzymes from which expert opined on the probability that blood found in victim's apartment was defendant's); Tice v. Richardson, 7 Kan.App.2d 509, 644 P.2d 490 (1982) (Frye applied to expert testimony opining that, after using formula for calculating probability that defendant was plaintiff's actual father as compared to a random man selected from population as a whole, plausibility of paternity in defendant was 99.96%; based on results using human leukocyte antigen tests, expert opined defendant was father).
Finally, while using actuarial risk assessment instruments in the area of sexual offender recidivism is not the exact equivalent of testing of DNA and blood, we believe the assessments are nevertheless a science. See Goddard v. State, 144 S.W.3d 848, 852 (Mo.App.2004) (“[A]ctuarial instruments ․ are generally classified as scientific evidence.”). As the Goddard court explained:
“ ‘[T]he STATIC–99 is a risk assessment instrument that combines ten factors. An individual's scores on these factors are summed, and the total score is compared to a table that shows the reoffense frequencies associated with each score. The table indicates, for example, that a score of 5 is associated with a frequency of sexual recidivism (over a five year follow-up period) of 33%. The highest risk category shown on the table-scores of 6 or above-is associated with a measured frequency of sexual recidivism (over a 5 year period) of 39%.’ Eric S. Janus, Examining Our Approaches to Sex Offenders & the Law: Minnesota's Sex Offender Commitment Program: Would an Empirically–Based Prevention Policy be More Effective, 29 Wm. Mitchell L.Rev. 1083, 1095–96 (2003).” 144 S.W.3d at 850 n. 2.
See 40 Am Crim. L.Rev. at 1465, 1471 (referring to actuarial risk assessments as science and “a serious enterprise, backed by sophisticated empirical methodology”).
For all these reasons, we affirm the district court's ruling that the actuarial risk assessments pass Frye scrutiny. So the rationale of the Court of Appeals majority is moot.
Implicit in our holding is a rejection of Daubert's application to our facts. In Daubert the United States Supreme Court held that the Frye test of 1923 was superseded by the adoption of the Federal Rules of Evidence in 1972. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. at 587. The Court created the Daubert test to make the admissibility standard for expert testimony consistent with the Federal Rules of Evidence, particularly Rule 702 governing expert testimony. Daubert demoted Frye's test of “general acceptance” from “an absolute prerequisite to admissibility” to simply one factor to be considered in the admissibility calculus. 509 U.S. at 588.
But Kansas has not adopted the Federal Rules of Evidence. And we disagree with Girard and Mallard that the language in our statute regarding expert testimony, K.S.A. 60–456(b), is substantially similar to Rule 702. We also disagree that this statute and the federal rule contain the same intent—which Girard and Mallard do not identify for us. Finally, we observe that Kansas has not adopted Daubert. See, e.g., State v. Heath, 264 Kan. 557 (Frye is the proper standard for admission of scientific evidence); Shadden, 290 Kan. 803 (applying Frye ). The defendants do not persuade us to depart from that path.
We conclude that in both cases the district court correctly admitted the expert testimony which relied in part upon actuarial risk assessments to opine Girard and Mallard were sexually violent predators.
The judgment of the Court of Appeals affirming the district court is affirmed. The judgment of the district court is affirmed.
The opinion of the court was delivered by NUSS, C.J.:
Judge MALONE concurred.