Opinion
NO. 96000
May, 1996
FINDINGS AND RULINGS ON DEFENDANT'S MOTION TO PRECLUDE EVIDENCE OF DNA TESTING
INTRODUCTION
Defendant was charged with the murder of State Trooper Mark Charbonnier on September 2, 1994 in Kingston. Prior to trial, defendant moved to exclude evidence of DNA testing by the Commonwealth. I conducted a two and one-half day evidentiary hearing before trial on the subject and, just prior to the start of trial, DENIED the motion with a short statement of reasons.
After trial, defendant was convicted. Because the admission at trial of DNA evidence linking the defendant to certain evidence found at the scene may be an appellate issue, I now write to set forth in detail my findings and rulings.
Most importantly, a .32 Walther handgun which fired the fatal bullet.
FACTS
Trooper Charbonnier was shot at approximately 3:15 a.m. shortly after stopping the defendant in his van on Route 3 in the southbound lane in Kingston.
Although seriously wounded in the abdomen, he was able to make a brief radio transmission which brought Trooper Sean Chicoine to the scene within minutes. In addition, a Boston Housing police officer, David Spiers, who was with his girlfriend at a nearby house, heard the shots and ran to the scene, arriving just prior to Trooper Chicoine.
Trooper Charbonnier was found wounded and lying at the left rear corner of the van, his nine millimeter handgun pointing to the embankment to the west side of the road. Clark was found "inchworming" up the embankment, also badly wounded and bleeding. A .32 Walther handgun was found about eight feet from Clark. Clark and Charbonnier were both taken from the scene for medical care.
Numerous other state and local police arrived at the scene very quickly and a thorough evidence gathering process began.
BLOOD SAMPLES
Lifecodes Corporation in Stamford, Connecticut performed the DNA testing in the case. On March 7, 1995, Lifecodes received thirteen samples sent from the state police crime lab in Boston, including samples of blood from the victim and the suspect, the suspect's clothing, samples collected from leaves and grass in the area, a swatch of blood from a watch found nearby, and a swatch of blood from the .32 handgun found near Clark. On September 14, 1995, Lifecodes received two additional samples from the handgun and Charbonnier's undershirt. Finally, on November 28, 1995, Lifecodes received from the crime lab a second sample of a blood stain taken from the watch found at the scene.
RFLP TESTING
DNA from several samples, namely, the defendant's shirt (Lifecodes FI 45888-63), blood found on the ground (FI 45891-66) and defendant's dungarees (FI 45896-71) was present in sufficient quantity to test by the so-called RFLP technique. Lifecodes concluded that these items matched the known blood sample from the defendant (FI 45886-61).
Restriction Fragment Length Polymorphism. This is the technique described and discussed in Com. v. Curnin, 409 Mass. 218 (1991); Com. v. Daggett, 416 Mass. 347 (1993); Com. v. Lanigan, 413 Mass. 154 (1992); and Com. v. Lanigan, 419 Mass. 19 (1994).
Also tested via the RFLP technique was the "swatch from the watch" received by Lifecodes on November 28, 1995 (FI 50383-75). The results also matched Clark's DNA pattern, or genotype. Again, there was no dispute about this during trial. The defendant stipulated that it was his watch which was found at the scene.
PCR TESTING
Because many of the samples contained insufficient DNA to test by the RFLP technique, most importantly, the blood on the gun, Lifecodes was requested to, and did, test by the DNA technique known as polymerase chain reaction (PCR). This technique is relatively new and has not yet been subject to appellate review in the Commonwealth.
Both PCR and RFLP are sample preparation techniques and are not tests to type genetic markers. However, for sake of simplicity, these terms shall be used liberally to describe genetic marker testing performed in this case.
As I will explain later, PCR testing has received overwhelming acceptance in numerous other jurisdictions.
PCR testing has been used in many areas of molecular and cellular biology, and its applications have ranged from medical diagnostics, allowing early detection of a variety of hereditary diseases, to basic genetic research. Reynolds, Sensabaugh and Blake, Analysis of Genetic Markers in Forensic DNA Samples Using the Polymerase Chain Reaction, 63 Analytical Chemistry 2 (1991). It has been used in numerous laboratories for forensic purposes since around 1990. Its use in a forensic setting has been validated by studies published in peer review journals. Comey and Budowle, Validation Studies of the Analysis of the HLA DQ Alpha Locus Using the Polymerase Chain Reaction, 36 Journal of Forensic Sciences 1633 (1991); Budowle, et al Validation and Population Studies of the Loci LDLR, GYPA, HBGG, D7S8 and GC (PM loci), and HLA-DQ Alpha Using a Multiplex Amplification and Typing Procedure, 40 Journal of Forensic Sciences 45 (1995) (Exhibit 2).
In this case, the typing of genetic markers using the PCR technique was effectuated through the use of three kits manufactured and distributed by the Perkin Elmer Corporation (Perkin Elmer). The kits are used by the laboratory to identify alleles (alternate forms of a gene) present at particular locations on a chromosome.
Perkin Elmer originally manufactured the Amplitype HLA DQ PCR Amplification and Typing Kit which was used to type specific alleles of the DQ alpha locus. Because of the relative frequency with which these alleles occur in a given population, DQ alpha typing was not particularly discriminating. The second kit put out by Perkin Elmer was the Amplitype PM PCR Amplification and Typing Kit or "polymarker kit," which served simultaneously to type alleles at loci LDLR, GYPA, HBGG, D7S8 and GC. A third kit, the Amplitype PM + DQA1 PCR Amplification and Typing Kit, types the DQ alpha and polymarker loci together. The latter two kits substantially enhance PCR's ability to discriminate.
To obtain results using RFLP testing, the scientist needs about one microgram of DNA. For PCR-based testing, reliable results may be obtained with two nanograms of DNA (Exhibit 5, tab 17, p. 16). In addition, PCR will work on "compromised" DNA. Thus, PCR is particularly useful in forensic settings where crime scene samples may be very minute and possibly degraded.
A microgram is one millionth of a gram.
A nanogram is one billionth of a gram.
A brief description of the PCR process is as follows. After the DNA from the sample has been extracted, it is added to a reagent solution and placed in a thermocycler which will heat and cool the solution. At 94° C, the "double helix" strands of DNA will "denature" or separate into two complimentary strands. The thermocycler temperature is then lowered to 60° C, at which temperature the primers join or "anneal" to a separated strand in a complimentary region of base pairs. The thermocycler temperature is then raised to 72° C, where the Taq polymerase enzyme pulls from the raw materials the appropriate chemical bases to extend the primers and create new DNA-product containing the sequence of interest. The process is repeated for 32 cycles resulting in production of more than one million copies of the original DNA. The basic scientific process is now well established and unchallenged.
Once the PCR process has been completed, the amplified solution is again denatured and pipetted onto a strip on which probes have been attached in the form of dots. The typing probes on the strips will bind or "hybridize" with the corresponding base patterns of known alleles at particular loci. If a particular allele is present at one of the loci used in PCR testing, a visible dot will appear on the strip.
Each strip has a control dot which will only light if there is a sufficient amount of DNA present in the solution. Dots that are as dark as or darker than the control dot are considered positive, indicating the presence of the corresponding allele. According to Perkin Elmer's protocols, dots lighter than the control dot "should be interpreted with care" (Exhibit 5, tab 16, p. 21, tab 17, p. 29 and tab 18, p. 5). What occurs to cause a dot to "light" if an allele is present is that the amplified DNA binds to a probe capturing a particular chemical in the dot. If the probe does not bind, the chemical is not captured and there is no change of color; hence, a particular allele is not present. There is an element of subjectivity in the scientist's interpretation of dot intensity.
In the hearing before me, defendant's expert, Dr. John Gerdes, a director in the DNA laboratory at the Immunological Associates of Denver (Exhibit 6), raised several criticisms of PCR based DNA testing's use in a forensic setting. No issue was raised as to the basic chemistry and biology; in fact, Dr. Gerdes testified that the Perkin Elmer kits, if used properly, will give proper results from single source samples containing adequate DNA. Defendant essentially contends that the possibility of contamination arising in forensic settings renders the technique unreliable for that purpose. Given the basic method of amplifying extremely small quantities of DNA, Dr. Gerdes argues that it is very easy to introduce extraneous DNA into the process, thus contaminating the samples and creating the potential for erroneous typing. Of most concern, according to Dr. Gerdes, is the accidental transfer of DNA from one item to another at the crime scene or at any time before the sample reaches the lab for analysis. Dr. Gerdes also contends that contamination can arise at the testing lab from reagents and other sources including mishandling or inadequate technician or equipment controls. These possibilities are generally acknowledged by those in the field; they disagree concerning whether the potential contamination problems are so serious that they render the entire procedure "unreliable".
In addition, Dr. Gerdes asserts that the control dot fails as a control where the sample contains a mixture of DNA from more than one person. In substance, he disagrees with those scientists who ignore dot intensity less than the control dot; when a dot "lights up", however slightly, Dr. Gerdes would construe it as a "weak allele", indicating contamination.
LIFECODES CORPORATION
Lifecodes Corporation has been a leader in the forensic use of DNA identification procedures. Lifecodes conducts DNA testing for paternity and forensic uses. It has been doing RFLP testing since 1982, and has applied such testing to forensic uses since 1986. It has used PCR testing for forensic uses since 1991. Two witnesses from Lifecodes testified: Dr. Michael Baird and Lauren Galbreath.
Dr. Baird has a master's degree in biology and a Phd. in genetics. He has been employed at Lifecodes since 1982, and has been vice-president of laboratory operations since 1992 (Exhibit 1). He is very experienced in both RFLP and PCR based DNA testing. He has many publications to his credit. He oversaw and interpreted the results of the testing in this case. I was impressed with his credentials and found his testimony to be credible.
Lauren Galbreath was the forensic scientist who performed the actual testing in this case. She has a bachelor of science degree in biology, and has worked for Lifecodes since October of 1988 (Exhibit 3). Her duties involve processing evidentiary material. She has tested over 1500 samples using the RFLP technique, and over 150 samples pursuant to the PCR technique. She has testified in court at least 35 times concerning RFLP testing and 3 times concerning PCR-based testing. I also found her testimony to be credible.
Lifecodes has had in place at all relevant times detailed written clinical laboratory protocols which the scientists must follow (Exhibit 5, tab 13-15). Miss Galbreath followed the protocols applicable in this case. In addition, for the PCR testing, Lifecodes followed the protocols supplied by Perkin Elmer (Exhibit 5, tab 16-18).
PCR TESTING RESULTS
Because eight of the original 13 samples received by Lifecodes in March, 1995 contained insufficient DNA to obtain RFLP results, the samples were reanalyzed using PCR. Lifecodes' report of June 13, 1995 (Exhibit 5, tab 4) indicates that Clark's DNA pattern (FI 45886-61) matched the following samples: blood on suspect's shirt (FI 45888-63), blood on a leaf (FI 45890-65), blood on the hill(FI 45891-66), blood on a tree leaf (FI 45894-69), blood on the gun (FI 45895-70) and blood on the suspect's dungarees (FI 45896-71).
In addition, Lifecodes reanalyzed the additional samples received by the state crime lab on September 14, 1995, namely, blood samples from the gun (FI 49168-73) and the victim's undershirt (FI 49169-74). Both samples were analyzed using the Perkin Elmer polymarker and DQ alpha kit. The results from the gun matched the results from the blood of the defendant (Exhibit 5, tab 5).
On November 28, 1995, Lifecodes received a bloodstain sample taken from a watch found at the scene (FI 50383-75). The results of the PCR testing (polymarker/DQ alpha) on this sample were that the blood on the watch matched Clark's (Exhibit 5, tab 8). This sample was also tested by RFLP and a match found (Exhibit 5, tab 9).
FREQUENCY CALCULATIONS
Although each individual's DNA molecule is unique, testing only examines a small portion, and it cannot be said that the patterns of alleles found in the loci examined are unique to one person. Thus, to give meaning to a match, statistics must be generated to estimate the frequency with which a particular pattern of alleles is likely to be seen in a given population.
In this case, the Commonwealth offered the testimony of Dr. Frederick R. Bieber, a medical geneticist employed by Brigham Women's Hospital (Exhibit 4).
Dr. Bieber's doctorate was in population genetics, which is the study of gene frequencies in various populations. Dr. Bieber was familiar with the ceiling principle as explained in the National Research Council's 1992 report entitled DNA Technology in Forensic Science (Exhibit 10, chapter 3, pp. 82-85), and incorporated into Massachusetts law in Commonwealth v. Lanigan, 419 Mass. 15, 26-27 (1994). For Lifecode's RFLP matches of samples with Clark's blood, Dr. Bieber used the alleles discerned by two probes, D17S79 and D4S163, to calculate the frequency of the matched genotype. His conclusion was that one person out of 1009 could be expected to have the same genotype. The analysis complies with the ceiling principle.
Dr. Bieber also calculated the frequency statistics regarding the polymarker and DQ alpha matches of samples with Clark's blood. He used four population data bases set out in the Budowle, et al validation study (Exhibit 2). Using the allele frequencies as shown in these peer reviewed data bases, and multiplying the frequency data for each allele (the so-called product rule), he arrived at the following results: The frequency statistics for the polymarker matches obtained from Clark's shirt (FI 45888-63), the blood-stained leaf (FI 45890-65), blood swatch from the top of the hill (FI 45891-66), swab from blood sample on gun (FI 45895-70) and blood sample from Clark's dungarees (FI 45896-71) were 1 in 1,210 for the African-American database, 1 in 135 for the Caucasian database, 1 in 169 for the Southeast Hispanic, and 1 in 310 for the Southwest Hispanic. The frequency statistics for the combined polymarker and DQ alpha matches obtained from the swab from the blood sample on the gun (FI 49168-73) and the bloodstain taken from Clark's watch (FI 50383-75) are 1 in 12,273 for the African American database, 1 in 4,362 for the Caucasian database, 1 in 7,147 for the Southeast Hispanic database and 1 in 17,033 for the Southwest Hispanic database.
Specific Criticisms in this Case
Through the testimony of Dr. John C. Gerdes and Dr. William Shields in the pretrial hearing, the defendant offered the following criticisms of the testing and the results in this case.
Defendant chose not call those witnesses at trial.
Dr. Gerdes
1. Ms. Galbreath failed to clean her scissors adequately between opening the respective samples received at the laboratory.
2. Ms. Galbreath extracted reference items (i.e. known blood samples) at the same time as samples. Dr. Gerdes argued that reference items should be analyzed separately from evidentiary samples to avoid contaminating the samples.
3. Ms. Galbreath did not quantify the amount of DNA with which she had to work for the PCR tests.
4. Ms. Galbreath did not run so-called substrate controls to determine if the substrate (i.e. unbloody portions of clothing) contained extraneous DNA. This, according to Dr. Gerdes, could be a red flag of contamination that may have occurred outside of the laboratory.
5. Ms. Galbreath did not run a reagent control to determine whether the reagent used in the testing contained DNA.
6. Lifecodes subjects itself several times per year to proficiency testing run by and pursuant to a protocol established by the College of American Pathologists. Lifecodes claims that it has not yet erred in PCR proficiency testing. Nonetheless, Dr. Gerdes find fault with the testing because the laboratory knows that the samples are submitted for proficiency test purposes (i.e. the tests are not "blind"), and the samples are not "difficult", i.e. like the ones often encountered in forensic analysis.
7. Dr. Gerdes contends that his observations of the Perkin Elmer test strips show "weak alleles", indicating a mixture of human DNA, hence contamination.
Dr. Shields
Dr. William M. Shields, an expert in population genetics (Exhibit 7), did not quarrel with Dr. Bieber's use of the ceiling principle to calculate the frequency statistics on the RFLP matches. He contended, however, that the ceiling principle should also have been used to calculate the frequency statistics regarding the PCR matches because population subgroups have differing frequencies of alleles at the loci used for PCR. Using seven population data bases which he has computer access to, and applying the ceiling principle (i.e. using only the highest allele frequencies in any of the population subgroups), he calculated the likelihood of random match with the suspect's blood genotype as 1 in 505.
Because Dr. Shields' own numbers make it very unlikely that the blood on the gun was anyone other than defendant's, the question becomes quite academic; and presumably for that reason, defendant did not recall Dr. Shields as a witness at trial.
The ceiling principle is not used because it is a scientifically valid method of estimation; rather, it is an ultra conservative method that avoids bias against a criminal defendant and thereby avoids an alleged, and rather technical, controversy among population geneticists about population substructure. See Com. v. Lanigan, 419 Mass. 15, 20-27 (1994). Dr. Shields concedes that the NRC report's discussion of the ceiling principle did not occur in the context of PCR testing.
THE SCOPE OF REVIEW: THE JUDICIAL GATEKEEPING ROLE
The trial judge's role in evaluating the admissibility of expert testimony has changed considerably in the wake of Daubert v. Merrell Dow Pharmaceuticals, Inc., 113 S.C. 2786 (1993). In the process of rejecting the rule of Frye v. United States, 54 App. D.C. 46, 293 F. 1013 (1923), the Supreme Court indicated that the trial judge's role would be to make "a preliminary assessment of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue." Daubert v. Merrell Dow Pharmaceuticals, Inc., 113 S.C. at 2796. Some factors which the court said would bear on the inquiry were whether the theory or technique can be and has been tested; whether it has been subjected to peer review and publication; the known or potential rate of error; the existence and maintenance of standards controlling the technique's operation; and, finally, "general acceptance" in the relevant scientific community. "The inquiry envisioned by Rule 702 is, we emphasize, a flexible one. Its overarching subject is the scientific validity — and thus the evidentiary relevance and reliability — of the principles that underlie a proposed submission. The focus, of course, must be solely on principles and methodology, not on the conclusions they generate." Id. at 2797.
In closing, the Supreme Court noted two concerns. First, that the abandonment of Frye would lead to the admission of "pseudo-scientific" assertions which would befuddle the jury. The court left this to what it called the conventional devices of the adversary system, namely cross-examination, presentation of contrary evidence, and careful instructions on the burden of proof.
The other concern, the polar opposite, was that the recognition of a "screening role" for the judge would exclude all but "scientific orthodoxy". The court recognized that a "gatekeeping role" for judges, no matter how flexible, would inevitably on occasion prevent the jury from learning of authentic insights and innovations, but that, nevertheless, was the balance struck by the rules of evidence.
Judge Rehnquist, in a brief concurrence/dissent, stated that "I do not doubt that Rule 702 confides to the judge some gatekeeping responsibility in deciding questions of the admissibility of proffered expert testimony. But I do not think it imposes on them either the obligation or the authority to become amateur scientists in order to perform that role." Id. at 2800.
In Commonwealth v. Lanigan, 419 Mass. 15, 24-27 (1994), the SJC rejected the Frye test and adopted the reasoning of the Supreme Court in Daubert. The SJC noted that the general proposition set forth in the Daubert opinion seemed sound, but the opinion gave little guidance for the application of that proposition to the facts of the given case. Lanigan II, while recognizing the trial judge's gatekeeping role, did little to explain what that role entails.
The basic question which Daubert and Lanigan II leave open is how high is the threshold for "reliability". Presumably the proponent of the scientific evidence must do more than merely make a prima facie showing that a particular technique is reliable. In other words, merely because a sufficiently qualified expert supports the reliability of the evidence does not, ipso facto, make it admissible. On the other hand, the proponent of the evidence does not have to convince the judge that his/her experts' opinions are correct; that is the function of the jury. See In re Paoli R.R. Yard PCB Litigation, 35 F.3d 717, 744-45 (3rd Cir. 1994). What the judge must look for is whether there are good grounds for the expert's opinion based on the methods and procedures of science. The evidentiary standard of reliability is lower than the merits standard of correctness. Id., at 744; see also Bowers v. Northern Telecom, Inc. 905 F. Supp. 1004, 1007 (N.D. Fla. 1995). As a former colleague on the Superior Court and now U.S. District Court judge has put it, the trial judge's gatekeeper responsibility is to ensure "that the fact-finding process does not become distorted by what is popularly called 'junk science'." Whiting v. Boston Edison Co., 1995 W.L. 115885 at 9 (D. Mass. 1995) (unreported).
DNA CASELAW: OVERWHELMING ACCEPTANCE
Numerous appellate courts in other jurisdictions have considered the admissibility of PCR based DNA evidence. All have involved the DQ alpha locus. Virtually all have ruled that the evidence was admissible. The cases of which I am aware are as follows:
1. Spencer v. Commonwealth of Virginia, 240 Va. 78, 393 S.E.2d 609 (1990) (finding of trial judge that PCR evidence reliable will not be disturbed if based on credible evidence.)
2. State of Washington v. Russell, 125 Wn.2d 24, 882 P.2d 747 (1994) (trial judge correct in finding that the underlying principle and techniques of PCR had been generally accepted by the scientific community. Supreme Court rejected defense argument that problems with use of PCR in forensic setting, including differential amplification, misincorporation and contamination, are so serious that testing fails to meet Frye standard).
3. State of Washington v. Gentry, 125 Wn.2d 570, 888 P.2d 1105 (1995) (It is now settled law in Washington that the PCR technique of DNA analysis passes Frye test. Criticisms of the test in the particular case went to weight, not admissibility).
4. State of Montana v. Moore, 885 P.2d 457 (Mont. 1994) (PCR test met Daubert standard of reliability; alleged errors in particular test went to weight; defendant failed to demonstrate whether tissue was contaminated in the instant case).
5. State of Nebraska v. Carter, 246 Neb. 953, 524 N.W.2d 763 (1994) (PCR match results should have been excluded because expert's method of calculating frequency lacked general acceptance. Expert did not have degree in population genetics. He used data from 1990 paper to calculate frequency of DQ alpha 4,4 allele).
6. Harrison v. State of Indiana, 644 N.E.2d 1243 (Ind. 1995) (Although trial judge should have held Frye hearing on PCR test results, the defendant not entitled to new trial because he consented to admissibility of results at a pretrial hearing before results were obtained, and results were not of primary importance to the murders of which he was convicted).
7. State of Kansas v. Hill, 257 Kan. 774, 895 P.2d 1238 (1995) (Citing to several appellate decisions in other jurisdictions, court held PCR testing admissible).
8. People v. McSherry, 14 Cal.Rptr.2d 630, 11 Cal.App.4th 1157 (Cal.App. 2 Dist. 1992) (Appellate Court upheld denial of motion for new trial based on new PCR tests that showed semen on panties of sexual assault victim did not match the defendant's; court relied on overall strength of prosecution case.).
9. People v. Amundson, 34 Cal.App.4th 1151, 41 Cal.Rptr.2d 127 (Cal.App. 4 Dist. 1995) (Held that PCR analysis was sufficiently established to have gained general acceptance in field, the testimony with the respect to the technique in its application had been offered by qualified experts and correct scientific procedures were used in the particular case).
10. People v. Morganti, 43 Cal.App.4th 643, 50 Cal.Rptr.2d 837 (Cal.App. 1st Dist. 1996) (PCR analysis held to be generally accepted in the relevant scientific community; absence of substrate control went to the weight and not the admissibility of the evidence).
11. McLaughlin v. State of Texas, Court of Appeals of Texas, Fourteenth District, Houston, 1993 Tex. App. LEXIS 379, 1993 WL 22050 (Defendant attacked the application of PCR testing in his case, arguing DNA extraction improperly performed; Held, defense criticisms went to weight, not admissibility).
12. Campbell v. State of Texas, 910 S.W.2d 475 (Tex.Crim.App. 1995), 1995 Tex.Crim. App. LEXIS 68. (Held, appellate court could not say trial judge erred in admitting DNA evidence).
13. State of Ohio v. Penton, 1993 Ohio App. Lexis 1937, 1993 WL 102507 (Ohio App. 3 Dist. 1993) (PCR testing results held properly admitted; conflicting testimony of Dr. Gerdes that PCR not a reliable tool in forensic science went to weight).
14. State of Oregon v. Lyons, 124 Or. App. 598, 863 P.2d 1303 (1993) (Defendant argued PCR method was generally accepted, but not in field of forensics. Held, potential contamination presents an open field for cross-examination, but did not indicate PCR was inappropriate for forensic use. Possibility of human error present, but safeguards against error existed and were followed. That the majority of PCR articles were produced by scientists with professional or financial interest in the success of PCR went to articles' credibility, not their existence. Held, PCR admissible).
15. Seritt v. State of Alabama, 644 So.2d 1 (Ala.Cr.App. 1994) (Upheld admissibility of PCR testing in murder trial).
16. State of Missouri v. Hoff, 904 S.W.2d 56 (Mo.App.S.D. 1995) (PCR testing held generally acceptable in scientific community and admissible. No error).
17. People v. Lee, 537 N.W.2d 233 (Mich.App. 1995) (Defendant argued PCR was inadmissible due to errors caused by enzyme Taq polymerase, and potential for contamination. Held, evidence admissible. Possibility of contamination at scene an area for cross-examination and went to the weight of the evidence. Concerning laboratory contamination, as long as adequate safeguards were taken, the results were reliable. Whether proper procedures were followed in particular case is question for the jury. If serious errors appear, court may rule test to be inadmissible).
18. State of Lousiana v. Spencer, 663 So.2d 271 (La.Ct.App. 1995) (PCR testing was properly admitted).
19. Redding v. State of Georgia, 464 S.E.2d 824 (Ga. 1995) (PCR test properly admitted; while possible problems could arise, no evidence that they did in this particular test).
20. Harmon v. State of Alaska, 908 P.2d 434 (Alaska Ct.App. 1995) (PCR test properly admitted).
RULINGS
PCR based DNA analysis passes muster when analyzed under virtually all of the factors which Daubert, Lanigan II and other cases have identified as bearing upon the reliability of scientific evidence.
Testability. PCR based DNA evaluation is testable. It has been tested repeatedly, and such testing has shown the technique to be reliable . In a sense, it was tested in this very case because defendant unquestionably bled on his shirt and dungarees, and the PCR tests both resulted in genotypes matching his known sample. Peer Review. Numerous validation studies of PCR testing of forensic samples have been published in peer review journals.
Error Rate. Lifecodes subjects itself to proficiency testing in PCR analysis according to the protocol of and under the auspices of the College of American Pathologists. It has not yet erred. Because virtually everything that human beings do is subject to occasional error, presumably mistakes will occur in the future. However, the alleged nonexistence of error rate, however that concept may be defined, is not a persuasive argument to exclude this evidence. The Appeals Court has recently held that "[w]eakness in the laboratory's proficiency testing went to the weight to be ascribed to the evidence of match, not to its admissibility." Com. v. Teixeira, 40 Mass. App. Ct. 236, 240 (1996).
Standards. Written protocols govern the operation of Lifecodes' laboratory. Testing in this case was done in accordance with these protocols, and the Perkin Elmer protocols for PCR testing.
General Acceptance. PCR testing is generally accepted in the relevant scientific community.
Non Judicial Uses. PCR is employed regularly for purposes other than litigation.
"Fit". Although forensic use of PCR presents different and somewhat difficult problems, mainly the risk of sample contamination, the technique is applicable to forensic use and has been regularly so applied for several years.
Qualification of Experts. Doctor Baird and Ms. Galbreath are very well qualified.
Acceptance by Other Courts. As indicated, appellate courts in other jurisdictions are virtually unanimous in the acceptance of PCR based testing, at least as to the DQ alpha locus. Although there is as yet no appellate case on the polymarker loci, I am not persuaded to exclude this evidence on that basis. In fact, no specific challenge to problems unique to the polymarker method was raised in the hearing before me.
Subjective Interpretation of Dots. As indicated previously, the interpretation technique based on dot intensity has some subjectivity to it. Dots less intense than the control dot are to be interpreted with caution, which does leave a certain amount of judgment and discretion to the scientist. It is true, as Dr. Gerdes testified, that less intense dots could indicate weak alleles which could indicate a DNA mixture due to contamination. Such less intense dots could also be an artifact of the testing process. Nevertheless, so far as I am aware, no appellate court has yet rejected PCR testing for that reason.
As far as Dr. Gerdes' criticisms are concerned, I find and rule as follows. Dr. Gerdes' criticisms concerning Ms. Galbreath's opening samples with allegedly inadequately cleaned scissors, extracting reference items at the same time as samples to be tested, failure to quantify the amount of DNA present, and failure to run substrate controls go the weight of the evidence, not its admissibility. The jury is fully capable of dealing with this type of challenge to laboratory evidence.
With reference to the allegation that Ms. Galbreath did not run a reagent blank control, I do not find this to be so. Upon the credible testimony of Ms. Galbreath and Dr. Baird that the Perkin Elmer protocols were faithfully followed, I find that the reagent blank controls were run through the amplification process.
Even if no reagent blank control was run, this would be a question of weight and not admissibility.
Concerning proficiency testing and the failure to establish "error rate", I am satisfied that Lifecodes' proficiency testing through the College of American Pathologists is adequate to pass the Daubert "reliability" standard. The concept of "error rate" does not yet have such a firm, fixed meaning in the forensic testing community that the absence of a procedure to establish it is a fatal flaw to the admission of the evidence. See Com. v. Teixeira, 40 Mass. App. Ct. 236, 240 (1996).
Finally, the fact that the PCR method requires that the scientist interpret dot intensity against a control dot does not render the technique unreliable. Defense experts have not pointed me to any literature which criticizes the method. In this case, when an unusual third allele appeared in the known sample taken from the victim after he had been transfused in emergency life saving attempts at the hospital, the laboratory obtained another known sample of the victim's blood before the hospital procedure, ran another test, and the mystery allele disappeared. With respect to other tests on samples where dots less intense than the control dot appeared, the laboratory was correct to ignore them. They were most likely an artifact of the method of developing the strip rather than contamination. Again, the jury is capable of coping with this issue.
Humans can have at most two gene forms, one from the father and one from the mother. The presence of a third allele suggests a mixture of DNA from more than one person.
FREQUENCY CALCULATIONS
Dr. Bieber did not apply the so-called ceiling principle in calculating frequency of occurrence statistics for the PCR matches. He used the four data bases described in the Budowle, et. al. article (Exhibit 2). By multiplying the frequencies of the alleles found at each locus (the product rule), he calculated how frequently the particular genotype was likely to occur in Caucasian, African American, Southeast Hispanics and Southwest Hispanics populations.
On those samples in which Lifecodes tested using the polymarker/DQ alpha kit, including most importantly the handgun, the frequency figures were as follows:
African American: 1 in 12,273
Caucasians: 1 in 4,362
Southeast Hispanics: 1 in 7,147
Southwest Hispanics: 1 in 17,033
The NRC discussion of the use of the ceiling principle applies only to RFLP testing, and the NRC never recommended its use for PCR results. The frequency of occurrence of RFLP alleles in a given population data base is very small. The NRC recommended using the largest frequency found in 15-20 population databases, or 5%, whichever was larger. Until such time as 15-20 population databases were compiled, the NRC recommended the use of a modified ceiling principle; whereby, the use of the largest frequency found in the available databases would be used, or 10%, whichever was larger. If the allele frequencies are over 10%, a 95% confidence itnerval would be applied to this statistic making it even more conservative. Dr. Beiber used the modified ceiling principle to calculate the RFLP frequency statistics in this case. The ceiling principle in theory yields the same frequency for a genotype, regardless of the suspect's ethnic background, because the reported frequency represents a maximum for any possible ethnic heritage (Exhibit 10, p. 85).
With PCR testing, the scientist is dealing with a limited number of alleles which occur relatively frequently in a given population (Exhibit 2). Virtually all alleles occur much more frequently than 5%, the upper bound used in the ceiling principle, and most occur more frequently than 10%, the upper bound used in the modified ceiling principle (Exhibit 2, tables 4 and 7, p. 50 and 52).
The use of the ceiling principle in calculating frequency occurrence statistics in connection with RFLP testing is mandatory in Massachusetts. Commonwealth v. Lanigan, 419 Mass. 15, 26-27 (1994). But no case law of which I am aware requires its use in connection with PCR testing. At least one Superior Court judge has ruled that it is not applicable to PCR-based DNA testing. Commonwealth v. Rosier, Berkshire Superior Court No. 94-0700, (Memorandum of Decision and Order by Ford, J. dated January 29, 1996). In fact, the use of the ceiling principle, even for RFLP purposes, is beginning to be questioned by courts in other jurisdictions. Armstead v. State of Maryland, 342 Md. 38, 673 A.2d 221 (Ct.App.Md. 1996) (For RFLP frequency statistics, there was no error in presenting product rule statistics along with ceiling principle statistics to the jury).
Furthermore, the ceiling principle was not recommended by the NRC or adopted by the SJC because it was generally accepted as reliable in the scientific community. Rather, it was adopted as a practical solution to avoid an arcane academic dispute rendering inadmissible very significant evidence. I have not been made aware of any academic articles which have advocated the use of the ceiling principle in connection with PCR results. The ceiling principle is social policy, not science.
Finally, as I understand Daubert and Lanigan II, it is beyond the charge of the trial judge to adopt rules of social policy when dealing with the admissibility of scientific evidence. My responsibility is to determine whether the proffered evidence is reliable. Here, I concur with the testimony of Dr. Beiber that the application of the product rule with peer review population data bases to PCR results produces accurate and reliable estimates of frequency of occurrence of the relevant genotype.
Patrick F. Brady Justice of the Superior Court
DATED: May, 1996