R v Karger
[2001] SASC 64
•29 March 2001
R v KARGER
[2001] SASC 64
1................ MULLIGHAN J............ The accused was charged with the murder of Kerryn Jean Ostendorf at her home at Albion Terrace, Campbelltown on 17th January 1998. The case against him was that he gained entry to her house probably in the early hours of the morning of Saturday, 17th January 1998 and strangled her to death with an article of her clothing, a camisole, whilst she was lying face down on her bed.
It is beyond dispute that Ms Ostendorf was murdered by strangulation by ligature and the issue at the trial was whether the accused was the murderer. The time of death cannot be fixed with precision. Dr James, a forensic pathologist, was unable to fix the time of her death. She was last seen alive at about 3.45 am on 17th January 1998 when she drove a friend to his home at Newton, a nearby suburb. Ms Ostendorf was due to attend her grandson’s birthday party at about 2.00 pm. When she did not arrive, her daughter first telephoned her home at about 3.00 pm and did not receive an answer. She thereafter made a number of telephone calls without a response until about 7.45 pm. She and her husband went to the house at Albion Terrace at about 8.00 pm and found Ms Ostendorf’s body.
Obviously she was murdered after about 4.00 am and before about 3.00 pm. She was lying face down on the bed fully clothed. There were some indications of a sexual attack. Her skirt, underpants, blouse and brassiere had been cut and torn at the back so that her back, buttocks and legs were exposed. Her anus was dilated consistent with penetration by a blunt object which could have been a penis, finger or something else. There was no injury to the anus or any sign of semen.
Mr Pearman, a forensic scientist, examined the clothes. The back of her blouse had a vertical tear from the hem to the collar. The hem of the blouse gave the appearance of having been cut. The collar of the blouse had also been cut or torn. There was a second tear at the back of the blouse running parallel to the other tear. The back strap of the bra had been severed. The left waistband and right rear portion of the underpants had been severed by what appeared to be a cut rather than a tear. According to Pearman, this damage gave the appearance of having been caused by a pair of scissors. The rear of the skirt appeared to have been both cut and torn. Pearman expressed the opinion that a pair of scissors may have been used.
Various pairs of scissors were found at Ms Ostendorf’s house after the murder but examination of them did not reveal any adhering fibres, blood or hair. Inside her blouse and near the line of the cut and tear were two stains. One was about one millimetre in diameter and the other was about one millimetre by three millimetres in diameter. Mr Pearman examined them and concluded that they had the appearance of blood stains. A test known as a sangur test was applied to one or both of the stains, the evidence was not clear, and the result of the test was positive. That test cannot establish that stains are blood or even human blood but it confirmed that they could be blood. However, I was satisfied that DNA was present in, and extracted from, the stains and the results of analyses showed the same profiles as of the blood of the accused. Mr Pearman expressed the opinion that the stains were human blood.
The case against the accused was largely circumstantial. He was not seen at the scene of the crime and denied to the police any involvement in the murder.
It is unnecessary for present purposes to mention the various facts and circumstances from which, on the Crown case, an inference that the accused is the murderer could be drawn but I mention some of them. If the evidence was to be accepted by the jury, the Crown could prove that the accused knew Mrs Ostendorf, he had a sexual interest in her and that, using his mobile telephone, had telephoned her home telephone number on nine occasions on the night of her death commencing at 11.05 pm on the Friday night and ending at 3.52 am on the Saturday morning. None of those calls was answered as Mrs Ostendorf was not at home. It is likely that she arrived home soon after the last of these telephone calls. The accused lived nearby and the evidence was capable of proving that he had both the opportunity and a motive to kill her.
Two aspects of the circumstantial evidence were the subject of objection before the jury was empanelled. One piece of evidence is that which could have been given by a former girlfriend of the accused whom I shall refer to as “W”. The other was the results of DNA testing of the two stains at the crime scene and of the blood and nail clippings of Ms Ostendorf and of material gathered from part of her clothing and the blood of the accused and its significance.
I heard evidence from W and I concluded that part of the evidence sought to be led from her was inadmissible and could not, therefore, be led. The accused objected to the proposed evidence as to DNA profiling of the samples taken at the crime scene and of what has been described as frequency or probability evidence regarding those profiles and the profile of the DNA of the accused. I permitted a voir dire hearing regarding the challenge to that evidence. I heard a large body of evidence and after hearing argument ruled that the evidence was admissible and that there was no reason to exclude any of the evidence in the exercise of discretion. I indicated that I would give reasons for those decisions after the trial was completed on 7th February 2001 and I now do so.
The evidence of W
W was the girlfriend of the accused at the time of the murder. They did not live together. He was a soldier in the Australian Army and was stationed at Holdsworthy Army Base in Sydney, New South Wales. W lived in Liverpool which is a suburb of Sydney and her parents lived in Macclesfield which is a suburb of Melbourne. They met in about the middle of 1997 and commenced a close relationship about two months later which included a sexual relationship. The relationship ended about a month after the accused was arrested for the murder on 24th March 1998.
According to W, the sexual relationship was an open relationship which, she said, meant that she had told him that if he wanted to try different types of activities, then he could say so and they would discuss it. She had suggested some activities to him and had asked him if he would like to try them. She had suggested using handcuffs, a silk scarf and a vibrator. She had told him that she was open-minded and willing to try different activities. She made five statements to the police and there are some inconsistencies in what she told them about features of her sexual relationship with the accused. I have approached the question of admissibility of her evidence on the basis of the evidence given by her during a hearing before the jury was empanelled.
W said that she stayed with the accused at the single men’s quarters at the Holdsworthy Army Base overnight on a Friday or Saturday night at the end of October or the beginning of November 1997. She described the events of that night as follows. It was a very hot night. They obtained take away food which they ate in his room at about between 6.00 pm to 7.00 pm. After a time she removed her clothing except for a pair of “G-string” knickers. She lay on her back on the bed. The accused was playing a game on his computer. It was a small room without air conditioning. There was an electric fan near the bed. W went to sleep.
She awoke in the early hours of the morning. She felt that she was no longer wearing the knickers and saw that they had been cut through at each side. The front had been folded off her onto the bed exposing her vagina. W asked the accused what had happened to the knickers and he said that he had cut them off. She asked him why. At that time he was moving around on one side of the room. According to her, his answer was “cheeky” but she cannot now remember what he said. When asked what she meant by “cheeky”, she said that his reply had been more of a sexual connotation than anything else.
The accused walked a few paces to the bed. He was wearing jeans but not a top. W did not see him until he was sitting astride her thighs. She remained lying on her back. The knickers were removed. W could not remember who removed them. The accused then used them to tie her hands to the top of the bed. She did not resist and acknowledged that she may have removed and given them to the accused for that purpose. She expected that they would have sexual intercourse and she would willingly participate. W totally co-operated with the accused thus far.
The accused then lent forward as if to kiss W and produced a knife from behind his back. W described it as a Recontanto knife which she had seen on previous occasions in his room. The blade was about nine to ten inches in length.
W said that she could then feel the knife against her throat. The accused told her not to move as his knife was against her throat. She said that she was a little scared and felt vulnerable. He then moved the knife down the left side of her neck across the front of the middle of the neck and up the right side. Her skin was not broken. He also kissed her. She said that she could feel the blade of the knife but she could not say if it was the sharp edge or the blunt side of the blade which made contact with her. She asked him what he was doing and told him to get off her. She removed her hands from the tie and pushed the accused away. W told the accused that he was “a bloody psycho” in order to ease the tension.
According to W, she did not think any more about the incident until after her relationship with the accused ceased. Until then she trusted him and their sexual relationship continued.
The Crown proposed to lead this evidence as part of the circumstantial case against the accused in order to prove that he had previously been involved in an act of a sexual nature involving the cutting of clothes. It was claimed that the basis of admissibility of evidence of this previous incident was that it showed a particular sexual appetite of the accused of an unusual nature, namely the cutting off of clothes of a sexual partner during sexual activity. In The Queen v Von Einem (1985) 38 SASR 207, it was held that in a case of murder alleged to involve homosexual abuse by a male accused of a boy, evidence of homosexual appetites or inclinations of the accused of that nature were admissible. King CJ said at p211:
“The Crown allegation was that a fifteen year old boy had been abducted, drugged, held captive for five weeks and then murdered. He had suffered an anal injury. The motive suggested by the prosecution for the abduction and captivity was homosexual abuse and the motive suggested for the murder was the need to cover up what had been done. If the victim had been a girl, heterosexual appetite would have been an obvious motive. As the victim was a boy, it was relevant to prove that the accused had homosexual inclinations which could provide a motive for engaging in the conduct alleged against him. The gratification of a person’s appetites and inclinations frequently supplies the motive for his actions. If there is a question as to whether a person has committed a particular act, it is relevant to know that he is possessed of appetites or inclinations which would be gratified by that particular act. If, for instance, there has been a theft of a particular type of foodstuff or liquor, it is relevant to know that the person accused of the theft has a taste for that particular type of foodstuff or liquor. Where the appetite or inclination would or might be gratified by the commission of a particular type of crime, the existence of such a type of appetite or inclination is relevant as tending to establish a motive for the commission of that crime. In the passage cited below from his judgment in Reg v Kilbourne [1973] AC 729, at pp756-757, Lord Simon of Glaisdale referred to such a motive as ‘propensity to commit the particular offence charged’. If it is accepted that the deceased was held in captivity and in a sedated condition for at least part of the time and was subject to homosexual abuse, the possession of homosexual inclinations supplies a possible motive for the appellant being responsible for that activity. To my mind evidence of the possession by the appellant of homosexual inclinations was relevant in proof of his guilt of the charge.”
The passage referred to in the judgment of Lord Simon of Glaisdale in Reg v Kilbourne [1973] AC 729 is to be found on the following page of the judgment of King CJ in Von Einem. King CJ went on to say at pp212-213:
“The cases emphasise, of course, that what is prohibited by the rule is the proof of previous misconduct or bad character for the purpose of leading to the conclusion that the accused person possesses a disposition or propensity towards the commission of crime or crime of a particular type and that he is therefore likely to have committed the crime charged. What is relevantly precluded by the rule is not the proof of human tastes, preferences, appetites or inclinations, the gratification of which does not necessarily involve criminal conduct, as a possible motive for crime, but the proof of previous misconduct as the means of establishing the existence of those attributes.
Ms Abraham QC, who appeared with Ms Chapman for the prosecution, contended that the impugned evidence could be shown to be a relevant circumstance to be considered with the other circumstantial evidence in the case. It was relevant evidence and therefore admissible as there was no rule of exclusion which rendered the evidence inadmissible. The rules of exclusion designed to protect an accused person against disclosure of previous offending, bad conduct or character had, on Ms Abraham’s argument, no application because conduct of this nature, although unusual by the standards of ordinary people, did not amount to bad conduct or character. It did not amount to evidence of prior offending in view of the nature of the relationship between W and the accused.
The same issue was considered in R v Turney (1990) 52 SASR 438. In that case the accused was charged with unlawful sexual intercourse with a child. The victim gave evidence of a course of conduct characterised by oral and anal intercourse and evidence was led by the prosecution that the appellant had engaged in intercourse of that nature with the mother of the victim. The evidence was held to be admissible. King CJ, with whom Cox J agreed, said at p439:
“I acknowledge that demonstrated predilection on the part of an accused person for a bizarre form of sexual activity might in some circumstances, provide confirmation of an allegation that the accused person engaged in that form of sexual activity with an alleged victim. The variety of forms of sexual activity in which human beings engage is such, however, that the probative force of such evidence must be judged with the utmost caution. Experience in the courts and the literature on the subject alike disclose that anal and oral sexual intercourse are far more common than might be supposed. I think that it would be extremely dangerous to treat evidence of anal and oral sexual intercourse between the appellant and his wife as tending to confirm the alleged victim’s evidence that the appellant had had those types of sexual intercourse with her. I think that it would be quite unsafe for the jury to conclude that the occurrence of these forms of sexual intercourse between the appellant and his wife would render the allegations of the daughter, if they were not true, ‘an extraordinary coincidence’. I consider that the evidence of the mother as to what occurred between her and her husband have no tendency to prove the truth of the allegations against the appellant on the basis upon which they were admitted and left to the jury.”
And later at pp439-440:
“On the particular facts in R v King [1967] 2 QB 338 and R v Von Einem (1985) 38 SASR 207, evidence that the accused person was a homosexual, that is to say was sexually attracted to males, was held to be admissible. It by no means follows that evidence of an accused person’s sexual predilections will always, or even commonly, be admissible: R v Horwood [1970] 1 QB 133. The risks associated with such evidence are obvious. Its probative force must be determined in relation to the facts of the particular case. The forms of sexual activity in question here appear not to be uncommon and I do not think that evidence of a predilection for such forms of sexual activity provides a motive significantly different from or greater than that provided by the existence of general sexual appetite. Moreover, the existence of a predilection for a certain form of sexual activity by a man with his wife throws little, if any, light upon the existence of a predilection for that form of sexual activity, or any form of sexual activity, with his young stepdaughter. In my opinion the evidence could not be admitted on that basis.”
See also Duggan J at pp444-445.
In support of her contention that there was no rule of exclusion which rendered the impugned evidence inadmissible, Ms Abraham referred to the observations of King CJ in Turney that the exclusionary rule preventing the reception of evidence of prior criminal conduct (see Makin v Attorney-General for New South Wales [1894] AC 57, Sutton v The Queen (1984) 152 CLR 528 and Pfennig v R (1994-1995) 182 CLR 461) did not embrace conduct which was discreditable but did not constitute a criminal offence: see pp440-441. There are expressions of view by the learned authors of Cross on Evidence (6th Edition, 2000, para 21035, 21065 and 21070) and by Ligertwood, Australian Evidence (3rd Edition, 1998, para 3.19) to the contrary. It is argued that this exclusionary rule extends to non-criminal conduct. It seems that in R v Garrett (SASC, 29 May 1996, S5652, unreported), Doyle CJ, with whom Millhouse and Williams JJ agreed, accepted the wider operation of the exclusionary rule. However, it was held in that case that the possession of handcuffs, although relevant to particular issues at the trial, was not relevant propensity evidence. The possession by the accused of handcuffs did not establish a propensity to commit the offences with which he was charged.
The matter was discussed by McHugh J in Pfennig v R at 512:
“The character and tendencies of a person are relevant in determining whether that person has committed the crime with which he or she is charged. But as a matter of policy the law generally excludes evidence of other incidents that reveals the criminal or discreditable propensities of the accused. Various reasons have been put forward to justify this exclusion. One reason is that it creates undue suspicion against the accused and undermines the presumption of innocence. Another is that tribunals of fact, particularly juries, tend to assume too readily that behavioural patterns are constant and that past behaviour is an accurate guide to contemporary conduct. Similarly, ‘[c]ommon assumptions about improbability of sequences are often wrong’ and when the accused is associated with a sequence of deaths, injuries or losses, a jury may too readily infer that the association ‘is unlikely to be innocent’. Another reason for excluding the evidence is that in many cases the facts of the other misconduct may cause a jury to be biased against the accused. ...”
I do not think it is necessary to consider the ambit of this exclusionary rule. It is not established by the evidence that in using the knife the accused committed a criminal offence. Although he did not give evidence, it is likely that in using the knife he lacked the necessary specific intent to commit an offence because of the nature of his relationship with W and also because the conduct itself could be regarded as consensual even though W would have preferred to have discussed, and approved of, it beforehand.
Furthermore, I do not think it was discreditable in the circumstances. To many it may be regarded as unusual, even disturbing, but it occurred in the context of a willing sexual relationship with features of experimentation and the unusual. Normalcy in sexual relations between consenting partners who are given to experimentation is to be determined by them and no-one else. As can be seen by the earlier brief description of features of their relationship, the use of aids for feigned subservience and stimulation were regarded as appropriate, and strange as it may seem to the more conventional mind, the cutting of knickers, tying of hands and the use of a knife as described was not unacceptable to W in her relationship with the accused.
I did not exclude the evidence because I considered that it offended the propensity rule. The basis for exclusion was that this one incident during the course of a consensual sexual relationship which, by that time, had existed for some months does not, to my mind, demonstrate a predilection on the part of the accused for the activity of cutting off clothes of a partner in the course of sexual activity and therefore could not amount to circumstantial evidence that the accused committed the murder. This one instance of conduct in the course of what might be described as an unconventional sexual relationship when judged by ordinary community standards, does not demonstrate a relevant sexual appetite which, to my mind, requires conduct on more than one occasion. The conduct did not establish a particular appetite or inclination or what may be described as a fetish, let alone one that is capable of consideration as a piece of circumstantial evidence that the accused was the killer. The conduct with the knife occurred on only one occasion. There had not been any earlier attempt to cut W’s clothing. There was no attempt to do so again following her reaction. The accused had been encouraged by W to experiment sexual matters, although, on her evidence, on the subject occasion he did not discuss the matter with her beforehand. Unlike proof of homosexuality, as in Von Einem, or an inclination for oral and anal intercourse evident over a lengthy period of time, as in Turney, evidence of one incident in the circumstances of a relatively extensive consensual sexual relationship does not, to my mind, establish an appetite, inclination or fetish. In the circumstances of the relationship between W and the accused, I did not regard the conduct as a bizarre form of sexual activity which, in itself, amounted to a piece of circumstantial evidence which could connect the accused with the murder.
For these reasons I ruled that the evidence was inadmissible. Consequently, it was not necessary to consider whether to exclude the evidence in the exercise of discretion.
The DNA Profiling
The DNA profiling was undertaken at the Forensic Science Centre in Adelaide by, and under the supervision of, Mr Pearman. I mention his qualifications later. Parts of the process were undertaken by other appropriately qualified forensic scientists.
Various samples of blood, hair and fingernail clippings were taken from the body of Mrs Ostendorf at the time of the post mortem on 18th January 1998. Samples of blood and hair were given voluntarily by the accused on 21st January 1998 after he had been questioned by investigating police officers. He was not arrested until 24th March 1998 but he was suspected of having committed the murder. Further samples were taken when he was arrested. The reference blood samples taken from Mrs Ostendorf and the accused are known respectively as KO1A and DJK1 and DJK2. The two stains on the back of the blouse of Mrs Ostendorf are known as KO22B and KO22C.
All of the samples, except KO22C, were subjected to DNA testing in February and March 1998 using a system known as Quadruplex. The tests showed that the DNA profile in KO22B is the same as that found in DJK1 and DJK2 and therefore he could not be excluded as being the source of that stain. That is a matter of considerable significance because the two stains were in the inside of the blouse and close to the tears. They were separated by a small crease which permits the conclusion that they are parts of the same stain. On the Crown case, their proximity to the tears suggests they were left by the murderer.
In early 1998 another system for DNA analysis was introduced into the relevant laboratory of Forensic Science Centre. It is known as the Profiler Plus system. All of the samples were tested for DNA using this system but no result was obtained with respect to KO22B, probably because all of the DNA had been extracted and used when the Quadruplex analysis was undertaken. The Quadruplex system inspects DNA at four loci and the Profiler Plus at ten loci, one of which is the same as a locus profiled by the Quadruplex system, known as vWA. The Profiler Plus inspects the amelogenin locus which indicates the gender of the donor of the sample, and eight other and different loci. As will be seen, a locus is a place or region on the DNA strand and loci is the plural of locus.
The stain KO22C and the reference samples KO1A and DJK1 were analysed using the Profiler Plus system in August 1999. A complete DNA profile at all ten loci was seen for the samples KO1A and DJK1. A DNA profile at a reportable level was seen at eight of the ten loci of the sample KO22C but that profile matched the DNA profile of the sample taken from the accused at the same loci. KO22C was also analysed on the Quadruplex system which revealed a match between the DNA profile of the accused and KO22C. The match was incomplete using the Profiler Plus system because the DNA profile of KO22C did not reveal a genotype at two loci in accordance with the reportable levels established by the Forensic Science Centre. A genotype is the combination of alleles, pieces of DNA at each locus, which are inherited from each parent. I return to that matter later.
The Forensic Science Centre also analysed the DNA of the accused, Mrs Ostendorf and the stain KO22B in March 1998 at the D1S80 locus. On the Crown case there was also a match between the DNA of the accused and KO22B at this locus although there was also a weak band inconsistent with the profile of the accused. On the Crown case this band could be a small fragment of DNA from Mrs Ostendorf as the wearer of the blouse.
The significance of the match between the DNA profiles of the accused and KO22B and KO22C is considerable as, if accepted, it would establish that the accused could not be excluded as the donor of the stains. If it was also accepted that the stains on the blouse occurred at the time of the murder, the accused could not be excluded as the murderer. The Crown proposed to lead evidence of population genetics and statistics to enable the jury to give weight to the match between the profiles of the accused and KO22C, but not KO22B or D1S80. The D1S80 locus was only used to see if the accused could be excluded by that analysis. There was no issue of admissibility about that evidence and the issues of interpretation and weight were clearly matters for the jury. The purpose of this evidence was to give due weight to the DNA evidence.
Later I mention the evidence relating to population genetics and statistics as the admissibility of that evidence was also the subject of challenge. At this stage I merely mention that calculations were made using the Australian Quadruplex Caucasian database and the South Australian Profiler Plus Caucasian database of DNA profiles and a formula known as a “match probability”, also known as a “conditional probability" using what is called the Equation 4.10 developed by appropriately qualified scientists and included in the American National Research Council Report in 1998 entitled “The Evaluation of Forensic DNA Evidence on the Sub-Population World” (“NRC Report II”).
The evidence which the Crown purported to lead from Mr Pearman and Dr Buckleton, also a forensic scientist, was that there was a match probability of one in a little over 90 billion. According to Mr Pearman, this means that having seen the match between the DNA profiles of the accused and the crime scene stain KO22C on one occasion, the probability or likelihood of seeing it again with a second and unrelated person with the same DNA profile as the accused is more than one in 90 billion. The proposed evidence was that in the present context, the stain KO22C is greater than 90 billion times more likely to match the profile of the accused if he left it, than if an unknown and unrelated person left it. It was proposed that Dr Buckleton would express the result of the calculation in this way: it is about 90 billion times more likely if the accused donated the stain KO22C, than if a random person donated it.
Grounds of the Challenge to Admissibility
The accused challenged the admissibility of the DNA and statistical evidence on various grounds and sought a voir dire hearing for that purpose. After hearing argument, I ruled that the questions which could be investigated were whether:
The Quadruplex system and the Profiler Plus system for DNA analysis are recognised and accepted by the relevant scientific community as reliable.
Mr Pearman has sufficient expertise to use these systems and to assess the results produced by them and if the procedures adopted by him in using these forms of analysis are recognised and accepted by the relevant scientific community as reliable.
Mr Pearman has sufficient expertise to express opinions proposed to be led from him in evidence as to the degree of probability of a match of the DNA profiles of the blood of the accused and KO22C occurring through chance or coincidence.
The two databases were valid for the purpose of expressing opinions as to that degree of probability
Mr Pearman and any other witness to be called by the prosecution for that purpose had sufficient expertise to express opinions about the validity of the database or databases and to correctly interpret them for the purpose of expressing such opinions.
Evidence was required to prove each of these matters and I permitted such evidence to be given on a voir dire hearing.
That ruling was given well in advance of the commencement of the trial, including the voir dire hearing. When that commenced, the accused particularised his case on the voir dire as follows:
“Further Particulars of Rule 9 Notice
1...... Profiler Plus is not recognised or accepted by the scientific community as reliable.
1.1... The scientific community cannot recognise or accept a genetic profiling process without being able to test it.
1.2 The Perkin Elmer Profiler Plus system is a commercial kit which has not been able to be evaluated by the scientific community because validation or other studies performed by, or relied on, by Perkin Elmer have not been released, and evaluation of its proprietary systems and software have not been permitted.
1A.... Further and alternatively, the Perkin Elmer Quadruplex system and the Perkin Elmer Profiler Plus system
(a).... have not been shown to comply with TWGDAM and NATA requirements, including but not limited to:
h Minimum samples.
h..... Peak heights on the Gene-scan and/or Genotyper.
h As to amplification without valid positive control samples.
h..... As to extraction and processing without a reagent blank.
h Contamination.
h..... Internal validation studies.
(b) have not been shown to have valid application outside the analysis parameters and protocols vouched for by Perkin Elmer; in particular, but not limited to:
h..... Minimum samples.
hPeak heights on the Gene-scan and/or Genotyper.
h..... As to amplification without valid positive control samples.
hAs to extraction and processing without a reagent blank.
h..... Contamination.
hInternal validation studies.
h..... Substrate controls.
hOff ladder alleles.
h..... Establishing and applying the sizing standard.
hSaturation.
h..... Matrix.
hEditing of the Genotyper function.
2...... Mr Pearman does not have sufficient expertise to use such systems or to interpret the results, because his conduct and supervision of the systems:
h..... Reveal a lack of understanding of the process and of essential steps in the process, through disregard of the manufacturer’s protocol.
h Ignore the absence of validated laboratory protocols.
h..... Reveal a disregard of scientific method.
h Reveal a lack of scientific objectivity.
4...... The validity of the database or databases.
4.1... The SA Caucasian database (PP99) cannot be used to express opinions based on “probabilities”, because the total of useable samples is too small for statistical purposes.
4.2 The samples used for the Quad 98 and Quad 99 databases must be shown to be useable.”
There was no paragraph 3.
During the course of the hearing, I was informed that it was not suggested that the Quadruplex system is not recognised and accepted by the relevant scientific community as reliable and accurate, but that the Defence contended that Mr Pearman did not have sufficient expertise to use either the Quadruplex system or the Profiler Plus system in accordance with the particulars set out in paragraph 2 of the particulars and at a later stage the accused was permitted to add the following particular to paragraph 2:
“h.... Reveal a failure to prevent contamination, the presence of contamination and the failure to use appropriate reagent blanks and controls.”
Ms Abraham did not claim to be prejudiced by that amendment.
During the voir dire the Defence mounted a detailed challenge to the reliability and accuracy of not only the Profiler Plus system of DNA analysis, but of fluorescence technology using DNA with short tandem repeats (“STR”), which I later describe, the competence of the Forensic Science Centre and the competence and expertise of Mr Pearman and the scientists at the Centre. The method of calculation of the probability of finding a known DNA profile elsewhere in the population was challenged on the basis that such evidence is inadmissible, or, alternatively should not be admitted in the exercise of discretion.
The practices and procedures of the Forensic Science Centre were examined in exhaustive detail. The foundation of the challenge to the DNA analysis was the opinions of Dr Atchison and Dr Davis to whom reference is made later. As will be seen, I regarded that foundation as without substance. Many serious allegations were made against Mr Pearman and the Forensic Science Centre based upon those opinions which I found to be groundless. It is necessary to express my reasons for the rulings which I made in more detail than would normally be the case. I have not attempted to discuss the evidence or even many of the submissions of the Defence in anything like the detail advanced during the hearing, although I have considered all of the evidence and the submissions on behalf of the prosecution and the accused.
Background of DNA Technology
In order to appreciate the issues raised by the challenge to admissibility, it is necessary to have an understanding of the recent developments in DNA technology. In R v Jarrett (1994) 62 SASR 443 at pp445-446, I described DNA and its significance in the forensic context. I also described the polymerase chain reaction (PCR) technique of amplifying or copying a small region of DNA millions of times and the technique for looking at a small number of the type of base pairs which are very variable from person to person. There is no need to repeat those observations. That fundamental science has not altered.
However, DNA analysis technology has made substantial advances since the development of the technology used in Jarrett. The DNA analysis was undertaken in that case in 1992 and the systems used in that case inspected four loci, or places on the DNA strands, separately and used essentially manual processes for inspection of the DNA fragments which are known as alleles. As can be seen from the description of the process in Jarrett with each of the loci then used, the inspection was undertaken with the naked eye by a suitably qualified person using a standard allelic ladder containing known allele sizes with the exception of one locus, HLA DQA1, which involved a different process. That locus was not used in the present case.
DNA analysis using this type of technology has been used extensively in the forensic context and commonly accepted in the courts as can be seen in various reported cases and, as yet, unreported cases: see Jarrett and R v Humphrey (1999) 72 SASR 558 and R v Smith (1998) 71 SASR 543 in this State and the cases referred to in R v Pantoja (1996) 88 ACrimR 554 by Abadee J at pp572-574 which show that evidence of this nature is accepted in courts in New South Wales, the United Kingdom, the United States of America and Canada. Other more recent cases show the acceptance of evidence of DNA profiling: in New South Wales, R v Milat (1996) 87 ACrimR 446 and R v Lisoff (NSW CCA, 22nd November 1999, unreported), in Victoria, R v Noll (VSCA, 7th October 1999, unreported) and The Queen v Vivona (VSC CCA, 12th September 1994, Jgt No 82 of 1994, unreported) and in Queensland R v Fitzherbert (QCA, 30th June 2000, unreported).
Developments
The first major development in DNA technology since the acceptance and use of PCR technology was the Short Tandem Repeat (STR) method of DNA analysis which was developed over the last ten years or so. It is now widely used around the world.
Before mentioning STR technology, it is appropriate to briefly describe the processes of extraction, quantification and amplification of DNA which are common to all current systems, including those employing silver staining which is a method of inspection used before fluorescence technology and which I shall later briefly describe.
DNA must be extracted from the relevant sample which may be a reference sample or a crime scene sample. There are differences in the process depending upon the nature of the sample, for example blood, urine, saliva or semen. In the present case the Forensic Science Centre was concerned with what were accepted as probably blood stains, some tissue and blood reference samples. The method of extraction used was the Chelex method which is widely used and accepted by forensic scientists. An extract from the sample is placed in a known volume of chelex with the result that DNA is extracted from that part of the sample. This procedure is undertaken in near sterile conditions using a biological safety cabinet and gloves and gowns which are, or should be, free of contamination. Dedicated sterile instruments are used. At the Forensic Science Centre the process occurs in one of three rooms designated for that purpose and the extracted DNA is kept in a sealed container in a refrigerator in a secure environment until required. Protocols and procedures are in place to avoid contamination. At the extraction stage, a control sample known as a reagent blank is also prepared. It is comprised of the reagents used in the extraction process but contains no DNA. It is run through the various stages of the analysis procedure so that it may be seen if the reagents are contaminated.
The next stage in the process is to attempt to quantify the amount of DNA which has been extracted, although not all laboratories undertake this step. It is desirable to know the quantity of DNA in the extract because there is an optimal level of DNA required for the amplification process, which is the next stage, so that the sample of amplified DNA submitted for analysis is not overloaded or saturated. If too much DNA is added, there is what is known as over-amplification or distortion which prevents or complicates interpretation at the analysis stage. If very low levels of DNA are amplified, a phenomenon known as “stochastic effect” occurs which may result in not all alleles being seen. Stochastic means random and is discussed later. The aim at the quantification stage is to provide an estimate of how much DNA is in the extract so that the amount of DNA added to the reaction mix at the next stage can be regulated. The systems used by the Forensic Science Centre in the present case are the QuantiBlot and ACES systems which are accepted and widely used by forensic scientists.
At the Forensic Science Centre quantification is undertaken in the main laboratory. A small amount of the extraction solution containing the DNA, about five microlitres, is taken and placed in a separate tube.
The DNA sample is placed on a membrane to which is added a probe labelled with, in the QuantiBlot system, a chemiluminescence detection system. The amount of chemiluminescence is proportional to the amount of DNA in the sample and is detected by an x-ray film. Control samples are run at the same time with known amounts of DNA as points of reference. The ACES system operates in much the same manner but a radioactive probe is used instead of chemiluminescence. The ACES system is regarded by some scientists as not as precise in measuring the quantity of DNA as the QuantiBlot system in that it tends to underestimate the amount of DNA in the sample.
Neither system is accurate. They are not particularly sensitive. Both tests can show no DNA to be present when it is present and a full profile may be obtained on analysis as occurred in the present case when using the Quadruplex to analyse the sample KO22B in 1998. I accept the evidence of Mr Pearman that the main purpose of quantification is to ensure that too much DNA is not amplified.
The next stage in the process is the amplification stage where extracted DNA is loaded into the reagents for the PCR. The sample is then loaded into a thermal cycler and heated and cooled at various stages using specified temperatures. This stage of the process occupies about two to three hours. This process mimics the copying process in the human body. A part of the extracted DNA duplicates, which is repeated time and time again until millions of copies have been made. Consequently there is usually ample DNA for analysis and profiling even when there was initially a low quantity of DNA in the sample. The copies are artificial but nevertheless accurate copies. The amplification process is also widely accepted and used by the forensic science community.
I now mention STR and fluorescence technology.
STR regions on the DNA in a chromosome are described as polymorphic DNA loci. They contain repeated nucleotide sequences. Nucleotides are base pairs and arrange themselves along the DNA molecule. The bases, A G T and C, only behave in a certain way in that A always pairs with T and G with C and at these types of loci they have identical repeat patterns along each strand of DNA. These repeat patterns occur differently on the same chromosome among individuals. They are particularly discriminating and are inspected by forensic scientists for the purpose of human identification. There are other regions of DNA which, by chemical reaction, dictate the characteristics of the human body and are consequently of interest to medical scientists. They tend not to differ much from person to person and are of little interest to forensic scientists.
STR loci consist of simple tandemly repeated sequences of one to six base pairs in length which may exhibit a high degree of length polymorphism due to variation in the number of repeat units displayed. They occur frequently throughout the DNA in the chromosome and have smaller repeat units than other regions of DNA. Because of their abundance, polymorphic nature and ease of amplification, they provide a source of highly imperative loci for use in human identification. The smaller size enables the PCR to amplify very small amounts of DNA for analysis. Because they are relatively short in size, up to about 350 to 400 base pairs, they are less susceptible to degradation than larger pieces of DNA. I accept that the “abundance, hypervariability and amenability” of STR loci to amplification by PCR “make them ideal markers for use in the identification of individuals”[1].
[1] Kimpton et al, “Evaluation of an automated DNA profiling system employing multiplex amplification of four tetrameric STR loci”: International Journal of Legal Medicine (1994) 106 302-311
The repeat sequence at one locus may be, for example, AATT and may be repeated four times, in which case it would be known as a four repeat unit. A sequence may be repeated five times and be known as a five repeat unit or there may be some other sequence. It is the repeat pattern which determines the length of the DNA fragment at a particular locus. One individual may have three tandem repeats at a particular locus inherited from one parent and eight tandem repeats inherited from the other parent or some other combination. These lengths of DNA are known as alleles and in this example are measured according to length, determined by the number of repeats. The allelic designations, or profile, at this locus would be 3 8. If the numbers of tandem repeats at each locus were eight and seven, the profile would be 8 7. In these two examples the alleles are of different length or size and the person is known as a heterozygote at that locus. Where the alleles are the same size, the person is known as a homozygote at that locus, and the profile would be a single digit, say 10.
Because the alleles are of different lengths, there is a technique to separate each of them, known as electrophoresis, which is mentioned shortly. Each locus is given a different name which identifies the particular chromosome and the precise location on that chromosome which is being described. An example is the D3S1358 locus. “D3” refers to the chromosome. “S1358” refers to the location on that chromosome. This designation and its significance are known to forensic scientists. A purpose of identifying the loci is so that anyone inspecting a particular locus is aware of the particular location on the chromosome which is being inspected.
Initially the inspection of STR loci was undertaken by what is described as the silver staining process. DNA is negatively charged. The amplified DNA is loaded onto a gel and the alleles are excited by an electric current through the electrophoresis process and move through the gel according to their size. The smaller pieces move faster than the larger pieces and are dispersed along the lane in the gel accordingly. An allelic ladder is loaded into the gel which is a sample containing the known alleles at the particular loci being inspected. Through the silver staining process, which need not be discussed in any detail because it is not relevant to any issue in the present case, the alleles appear as bands adjacent to the corresponding allelic size in the ladder and can be seen with the naked eye. That technology has been commonly used throughout the world. This system permits only one locus to be inspected at a time but various samples may be analysed simultaneously in the same gel.
The next development in DNA technology relevant for present purposes is fluorescence technology. At the PCR or amplification stage specially designed primers with fluorescent tags attach to the alleles or DNA fragments. They remain with the DNA in the gel and during the process of electrophoresis. Instruments have been developed to undertake the electrophoresis and detection processes. They are known as sequencers and enable the detection of alleles by the use of a laser which agitates the fluorescence in the primers which is captured by a camera and transmitted and recorded by a computer. The first of these sequencers is known as an AB1 373 DNA sequencer (“the 373”). Subsequent models are the AB1 Prism 373A sequencer (“the 373A”), AB1 Prism 310 sequencer (“the 310”) and the AB1 Prism 377 sequencer (“the 377”). There are differences in the technology and process between the 310 and the others which need not be discussed as the 310 played no part in the DNA analyses in the present case. These sequencers were developed and manufactured by Perkin Elmer, a corporation in the United States of America, which has been prominent in the development of modern DNA technology. Other sequencers were developed by other manufacturers but they have no relevance to any issues in the present case.
The principle of electrophoresis in this technology is the same as with the silver staining process. As the electricity is applied, the fragments of DNA or alleles have moved according to their size. As has been mentioned, the fluorescent dye attached to the primer has become part of the amplified DNA so that there is a fluorescent tag attached to every piece of DNA. The primers are specific as to location. They bind to a particular locus of the DNA in accordance with their design and the dye attached to the primers is detected during the analysis process. The position of the allele is compared to an allelic ladder which is comprised of known alleles. The size of alleles is determined. The consequence is the designation of a profile at a particular locus.
Computer software was developed to assist in the interpretation of the data collected through the sequencers. Initially this software was known as genescan and later seems to have developed that name as a brand name, “Genescan”, and I shall refer to it accordingly.
The Genescan analysis software converts the data obtained through the fluorescence process. It tracks each sample so as to ensure that the computer has analysed the correct portion of the sample in the particular lane. It produces a table of results and also shows them in a graphical format, called an electropherogram, showing peaks rising from the base line in the colours of the system which I mention later when discussing multiplex technology. The table of results includes the colour of the peak, the time of the run, the height and area of the peak, and also the Relative Fluorescent Unit (“RFU”) value. Usually the greater the degree of fluorescence, the higher the peak on the electropherogram shown as an RFU. With Genescan it is at this stage of the process that the analysts interpret the results. Later, mention is made of artefacts and other issues which can arise in the interpretation of the results and the need for expertise of the scientist at the analysis stage. At this stage, it is sufficient to say that the analyst interprets the Genescan results and may manipulate the software so as to assist in that function.
When the laser excites the dye in the primer attached to the amplified DNA, the emitted light passes through a type of prism which separates out the light according to its wave length so that each specific wave length falls upon a designated portion of the detection camera. Usually a light wave, although basically of one of the particular colours, say blue, encompasses other light waves which spill over into another dye, say green. Whilst the majority of the light falls on its designated portion of the detection camera, and is detected as an allele, the portion of light which falls into another colour is picked up by other portions of the detection camera. The computer software will recognise that other portion of light and it may be represented in the data as a small peak in another colour. There is software in the system known as matrix which is established and adjusted when the system is installed and which is designed to take into account that portion of light which falls on to other parts of the detector. It effectively subtracts them from the collection software. However, there will be times when the matrix does not subtract all of that type of light and it will be shown as a small peak on another line on the electropherogram of the Genescan known as “pull up” which is mentioned later.
The next significant development in DNA technology for forensic scientists was the use of multiplex systems. Before this time, the DNA at each selected locus was amplified and inspected individually. Multiplex technology initially permitted amplification and analysis of more than one locus at a time using the techniques of electrophoresis and fluorescence and the collection software. Multiplex systems were developed to inspect different STR loci. Early such technology was known as a triplex which inspected the loci D21S11, FGA and the amelogenin gender locus. Another triplex system inspected the loci D3S1358, D21S11 and FGA.
Kits were produced by Perkin Elmer for these triplexes. They had different colours of dye being blue or green and were respectively known as the Blue Kit or the Green Kit each inspecting different loci.
These multiplex systems were further developed over the years by Perkin Elmer and, one of its competitors, Promega Corporation in the United States of America and the Forensic Science Service in the United Kingdom. The Forensic Science Service developed a quadruplex system in the early 1990’s and Perkin Elmer developed the Quadruplex system which was used in tests in the present case. These systems were widely used by forensic scientists in the United States of America and the Forensic Science Service. They permitted the inspection of four loci in each DNA sample simultaneously known as THO1 and FES using blue dye in the primers and vWA and F13 using green dye.
Advantages of Fluorescence Technology
I accept the evidence of Mr Pearman that there are advantages in using fluorescence technology. It is described as a semi-quantitative technique which means that it is powerful in the interpretation of mixtures which occur when different people contribute DNA to the same stain or sample. The process removes some of the subjectivity of the scientist when looking at weak bands and deciding whether to designate such bands to be alleles. An internal sizing standard is used which offers a considerable level of precision. The process is able to correctly designate alleles at a locus which only vary by one base pair. The system removes the need for manual transcription of results and possible human error in cases involving contemporaneous testing of many samples.
Use of Scientific Articles and Publications
During the hearing many scientific articles and publications were admitted into evidence which show the development of STR fluorescence technology and the acceptance by the forensic science community of its reliability and accuracy for the purpose of DNA analysis for human identification. I have already briefly referred to one of these articles and later in these reasons I refer to other such articles for other purposes, including the acceptance of the Profiler Plus system by that community for that purpose. These articles and publications were proved through Mr Pearman as part of the body of scientific knowledge relevant to matters in issue on the voir dire. Mr Wells QC, who appeared with Mr Mayne for the accused at the pre-trial hearings, contended that these articles and publications may only be used for a limited purpose and referred to decision of McGarvie J in PQ v Australian Red Cross Society & Ors [1992] 1 VR 19 in support of that contention. In PV v Australian Red Cross, McGarvie J had to consider the use which could be made of the data set out in authoritative scientific publications. He said, at p34:
“By ‘information in authoritative scientific publications’ I mean information of the type which scientific experts of the relevant categories ordinarily treat as data on which they may rely in forming opinions and making decisions within the area of their expertise. Included in such data are facts and opinions stated in articles or reports in scientific publications or in statements by organisations, public authorities or persons regarded by such experts as having knowledge and expertise in the relevant area. Such data includes facts in tables or statistical material on which such experts ordinarily rely.
It is made clear in Borowski v Quayle [1966] VR 382 that expert witnesses may not only base opinions they give in evidence on such data, but may give evidence of fact which is based on such data. Expert witnesses may do this although the data on which they base their opinion or evidence of fact will usually be hearsay information in the sense that they rely for such data not on their own knowledge but on the knowledge of someone else: see R v Abadom [1983] 1 WLR 126, at pp129-32. The considerations which justify that principle are stated in the passages from Wigmore on Evidence on which Gowans J relied in Borowski, at pp386-8. See also R.W. Baker, The Hearsay Rule, p165.
An expert witness, in relying on data in authoritative publications, is not confined to confirming or correcting a recollection of what is stated in the data. The witness may rely on the data without a previous knowledge of it. An example is the reliance that may be placed on tables and the like: see Borowski, at pp387-8. The data relied on may be a statement of fact or opinion.
When an expert witness bases evidence on data in an authoritative scientific publication it is the evidence of the witness which is thus put before the court. The publication itself is not evidence of the truth of statements it makes as to data. If the witness refers to or quotes from an authoritative publication as correctly stating a fact, what is referred to or quoted is part of the testimony of the witness: Sussex Peerage Case (1844) 11 Cl & Fin 85, at pp 114-17; 8 ER 1034, at pp1046-7; Collier v Simpson (1831) 5 C & P 73; 172 ER 883; Cocks v Purday (1846) 2 Car & Kir 269; 175 ER 111; Concha v Murrieta (1889) 40 Ch D 543, at p 554; Federal Commissioner of Taxation v Hamersley Iron Pty Ltd (1980) 33 ALR 251, at pp273-4; Baker, The Hearsay Rule, p164 and Gillies, The Law of Evidence in Australia, pp354-5.
As a statement in an article in a learned scientific journal is only before the jury as part of the testimony of an expert witness, it is for the jury to decide whether an expert witness in evidence adopted or acknowledged the correctness of the statement. If it is before them in that way, they assess it in the same way as any other part of the evidence of the expert witness: Concha v Murrieta (1889) 40 ChD 543, at p554.”
Obviously the admissibility of any piece of evidence depends upon the purpose for which it is to be used. In the present case the purpose for which the articles were admitted was to establish the nature, development and general acceptance of STR fluorescence technology by the forensic science community, including the use of particular instruments and software. That is a question of fact. The purpose of admitting the articles was not to prove some particular data not accepted by a witness at the hearing, but to assist in the matters which I have mentioned. Once such articles and publications have been proved and verified and accepted as forming part of the body of knowledge of the relevant scientific community as they were in the present case by Mr Pearman, they may be used for that purpose. Furthermore, they may be used as evidence supporting the evidence of witnesses about such matters and as a basis for the opinions of such witnesses. They may be used as evidence of the data contained in them upon which the witness has relied. In The Queen v Vivona (12th September 1994, Butterworths Unreported Judgments), the Court of Criminal Appeal of Victoria cited, with approval, the following passage from Wigmore on Evidence, 3rd Ed, vol 2, 784-5, para 665(b):
“The data of every science are enormous in scope and variety. No one professional man can know from personal observation more than a minute fraction of the data which he must every day treat as working truths. Hence a reliance on the reported data of fellow-scientists, learned by perusing their reports in books and journals. The law must and does accept this kind of knowledge from scientific men. On the one hand, a mere layman, who comes to court and alleges a fact which he has learned only by reading a medical or a mathematical book, cannot be heard. But, on the other hand, to reject a professional physician or mathematician because the fact or some facts to which he testifies are known to him only upon the authority of others would be to ignore the accepted methods of professional work and to insist on finical and impossible standards. Yet it is not easy to express in usable form that element of professional competency which distinguishes the latter case from the former. In general, the considerations which define the latter are (a) a professional experience, giving the witness a knowledge of the trustworthy authorities and the proper source of information, (b) an extent of personal observation in the general subject, enabling him to estimate the general plausibility, or probability of soundness, of the views expressed, and (c) the impossibility of obtaining information on the particular technical detail except through reported data in part or entirely. The true solution must be to trust the discretion of the trial judge, exercised in the light of the nature of the subject and the witness’ equipments. The decisions show in general a liberal attitude in receiving technical testimony based on professional reading.”
I accepted the evidence of Dr Buckleton that forensic scientists use either the product rule or the match probability, or Equation 4.10, approach and the latter is used extensively in the British Isles, and in laboratories in the United States of America, New Zealand and some laboratories in Australia. The Equation 4.10 method is accepted by forensic scientists. It is the most conservative of the other methods of calculation available. Both Mr Pearman and Dr Mitchell accepted that the Equation 4.10 was an accepted approach.
The Forensic Science Centre used Equation 4.10 in the present case with an Fst value of three per cent and a confidence limit of 95 per cent. Using software, this match probability was one in 90.558 billion.
Fst is the measure of co-ancestry or relatedness of alleles and reflects the probability that alleles within a population have a common ancestor. It is necessary to know the level of that ancestral relationship between people. Population geneticists undertake that task by using an Fst value. Scientists have estimated this level of relatedness among Caucasians as less than one per cent. The use of three per cent is conservative. Dr Buckleton explained that an Fst value of three per cent is based upon the premise that all Caucasians are related to the level of second cousins, which is obviously factually incorrect and therefore conservative. According to Dr Buckleton, this percentage value easily accommodates for any degree of relatedness within the Caucasian community.
The confidence level of 95 per cent was explained by Mr Pearman and Dr Buckleton in this way. If 200 persons are profiled and allele frequencies are determined on that basis and the process is again undertaken with 200 different persons, slightly different values would be obtained for the allele frequencies simply because different people within the same population are being sampled. That sampling variation is taken into account by use of a confidence level which defines, with some certainty, that the true value lies between a common value and a rare value. This means that it is certain that 95 per cent of the time the true allele frequency will fall between the upper limit and the lower limit. In the present case the upper confidence limit was one in 90 billion and the lower confidence limit was one in 272 billion and so the figure adopted is the conservative and lower figure.
Mr Pearman presented this figure as a likelihood ratio which, he said, means that the probability or likelihood of seeing a second unrelated person with the same DNA profile is more rare than 1 in 90 billion which, in the present context, means that the crime scene stain KO22C is greater than 90 billion times more likely to match the profile of the accused if he left it, than if it was left by an unknown and unrelated person. Dr Buckleton put it this way. He said that it is about 90 billion times more likely if the accused donated the stain KO22C, than if it was donated by a random person.
Both Mr Pearman and Dr Buckleton said that the two databases were acceptable and valid for this purpose and that it was acceptable to combine the THO1, FES and F13 loci of the Quadruplex loci with the seven Profiler Plus loci at which the accused showed a profile. Both of these witnesses also said that the selection of a Caucasian database was appropriate as the accused is Caucasian and that the racial group where it is most likely to see a second profile, the same as the accused, is in the same racial group. Therefore, to use the race of a suspect is the more conservative database to use.
Also, both Mr Pearman and Dr Buckleton expressed the opinion that the frequency of alleles in both databases could be used in the calculation using Equation 4.10. Their evidence was that the creation of racial databases is accepted and is the usual practice by forensic scientists and is preferable to the creation of general databases.
I now turn to the grounds of the challenge by the defence.
Whilst the particulars of the challenge do not raise an issue about the accuracy of the South Australian Profiler Plus database, it had to be considered whether that database was accurate for the purpose for which it was to be used. I concluded that it was and the remaining few errors were of no consequence. That is the effect of the evidence of Mr Pearman and Dr Buckleton. I mention that at the trial evidence was given of fresh calculations having been made after the few remaining errors had been corrected and the match probability was greater than one in 90 billion.
I need not repeat the particulars of the first ground. I accepted the evidence of Dr Buckleton and Mr Pearman that the DNA profile of the accused is likely to be more common within his own race, the Caucasian race, than in any other and therefore the use of the Caucasian database was not only appropriate but was most favourable to him and accorded with the practice accepted by the forensic science community.
As has been mentioned, the evidence of Dr Buckleton as to the appropriate Fst value was that it is about one per cent. The use of the value of three per cent is conservative by a substantial margin and therefore it is unnecessary to calculate a precise figure. Also the evidence of Dr Buckleton, which I accepted, established that independence between loci in the sub-population of interest, which includes the perpetrator, need not be established when using Equation 4.10 which is a view widely accepted by forensic scientists. Dr Mitchell expressed a different view. He said that the Fst value does not take account of any possible independence between alleles at loci or between loci. Also, he said that the two databases could not be combined as the same protocol for establishing racial characteristics had not been used and because the same individuals were not profiled in the databases. Dr Atchison expressed the same view.
I thought that these differing opinions were matters of fact to be resolved by the jury at the trial assisted by appropriate explanations and directions: R v Jarrett at pp451ff; see also R v Lisoff [1999] NSW CCA 364 at p14. Of course, if I accepted the evidence of Dr Atchison and Dr Mitchell and rejected the contrary evidence of Dr Buckleton and Mr Pearman and concluded that their views did not form part of the body of knowledge of the forensic science community, a question of admissibility could arise. I did not form that view. Consequently, I regarded the evidence as admissible despite the differing opinions.
Dr Atchison and Dr Mitchell both expressed the view that even with the Equation 4.10 approach, linkage equilibrium must be established. It may be seen that there is a difference of opinion between them and Dr Buckleton and Mr Pearman. There was a reference to scientific authority by Dr Buckleton in support of his view which I need not repeat. It was submitted by Ms Abraham that Dr Mitchell lacked expertise in the use of Equation 4.10 for forensic purposes and his opinion should not be preferred, and due to a lack of understanding on the part of Dr Atchison that his opinion should not be accepted.
I regarded this issue as a matter for the jury. I did not reject the evidence of Dr Buckleton or Mr Pearman which could have raised an issue of admissibility.
There was also a clear difference of opinion between Dr Buckleton and Mr Pearman and Dr Mitchell and Dr Atchison as to the use of both databases. I also regarded that matter as an issue for the jury. I did not reject the evidence of Dr Buckleton or Mr Pearman.
I now turn to the second ground of the challenge which is that evidence of the use of the likelihood ratio approach is inadmissible because it is meaningless without using the calculation known as Bayes’ Theorem. This theorem is complex and need not be set out for present purposes. It may be seen in Aitken and Storey, The Use of Statistics in Forensic Science at p120. A likelihood ratio is a component of the theorem and involves calculations of the probability that a suspect is the offender and the probability that the suspect is not the offender. The ratio of those two probabilities is the likelihood ratio for the purpose of Bayes’ Theorem. The theorem is a mathematic model for the reaching of a conclusion by taking into account all relevant evidence in the case. It was considered by the Court of Criminal Appeal in R v Adams [1966] 2 CrAppR 467. The defence had been permitted at trial to lead evidence from an expert as to the use of Bayes’ Theorem in a rape case involving DNA evidence, the Crown having accepted that it was a valid method for looking at non-statistical matters in statistical terms. Rose LJ, in giving the judgment of the Court, observed at p481:
“... the theorem can only operate by giving to each separate piece of evidence a numerical percentage representing the ratio between probability of circumstance A and the probability of circumstance B granted the existence of that evidence. The percentages chosen are matters of judgment: that is inevitable. But the apparently objective numerical figures used in the theorem may conceal the element of judgment on which it entirely depends. More importantly for present purposes, however, whatever the merits or demerits of the Bayes Theorem in mathematical or statistical assessments of probability, it seems to us that it is not appropriate for use in jury trials, or as a means to assist the jury in their task. In the first place, the theorem’s methodology requires, as we have described, that items of evidence be assessed separately according to their bearing on the accused’s guilt, before being combined in the overall formula. That in our view is far too rigid an approach to evidence of the type that a jury characteristically has to assess, where the cogency of (for instance) identification evidence may have to be assessed, at least in part, in the light of the strength of the chain of evidence in which it forms part. More fundamentally, however, the attempt to determine guilt or innocence on the basis of a mathematical formula, applied to each separate piece of evidence, is simply inappropriate to the jury’s task. Jurors evaluate evidence and reach a conclusion not by means of a formula, mathematical or otherwise, but by the joint application of their individual common sense and knowledge of the world to the evidence before them.”
The use of the theorem was rejected by the Court, which re-affirmed the usual approach to scientific evidence as being appropriate, that is to assess the evidence and reach a conclusion about it. That decision was affirmed in Doheny and Adams at p375.
It was argued by the Defence that because Bayes’ Theorem may not be used in cases of this nature, nor can a likelihood ratio because it is part of the theorem. Dr Atchison said that the likelihood ratio is part of Bayes’ Theorem. Dr Buckleton said that the likelihood ratio is provided to courts without implying the use of Bayes’ Theorem. I accepted that evidence. He referred to passages of the second report of the National Research Council in 1999 on the Evaluation of Forensic DNA Evidence (NRC Report II) which support that view and he mentioned other scientific publications and tests which, he said, also support that view.
In my view, this ground of the challenge also fails. I accepted the evidence of Dr Buckleton that reports are made using a likelihood ratio in the United Kingdom, parts of the United States of America and some laboratories in Australia. In the use of a likelihood ratio by forensic scientists in the manner described by Dr Buckleton and used by the Forensic Science Centre, other evidence, for example opportunity, alibi or confession, is not taken into account as would occur with the use of Bayes’ Theorem.
I mention one other matter which was advanced by Dr Atchison. He said that as the denominator of the likelihood ratio points to the hypothesis that the profile seen does not come from the suspect but from some unrelated person, it is necessary when calculating the denominator to take account of laboratory error rates in so far as they may produce a false match between the DNA of the suspect and that found in the crime scene sample. It is the errors which are uncorrected which must be taken into account. Dr Atchison could not advance any way in which such an error rate could be calculated. He said that it was logically impossible to do so because the errors to be considered are unknown errors. The consequence, he said, is the likelihood ratio cannot be calculated.
I rejected this evidence. There is no support for this opinion to be found in the literature as disclosed in the evidence. Dr Buckleton said that an error should not be taken into account and that view is expressed in NRC II and other publications. His view is that the possibility of error is a separate matter. Mr Pearman said that there is much evidence to show that an error rate cannot be calculated. He could not see any difference between a frequency or a likelihood ratio with regard to an error rate.
I accepted their evidence. If the tribunal of fact concludes that there is a reasonable possibility of an error causing an incorrect match, be it in the handling of samples, or in the laboratory or in any other way, there could not be a finding of a match between the DNA of the accused and that in the crime scene sample to the required degree of proof, with the consequence that the DNA part of the case would collapse and there could be no need to consider the statistical evidence.
As to the third ground I saw no reason to exclude the evidence of statistical calculation in the exercise of discretion. I concluded that the evidence was admissible. The discretion does not fall to be exercised unless there is prejudice “additional to, or distinct from, the detriment to the accused’s interests involved the probative force of the evidence”: Duke per King CJ at p48. There was no such additional prejudice. I repeat the observations I made in Jarrett at pp457-458.
For these reasons I admitted the evidence of Mr Pearman and Dr Buckleton which the Crown proposed to lead.
20
8
0