R v FULLER - [2013] SADC 150

District Court of South Australia
(Criminal)
R v FULLER
[2013] SADC 150
Ruling of His Honour Judge Stretton (ex tempore)
30 October 2013
EVIDENCE - ADMISSIBILITY AND RELEVANCY - OPINION EVIDENCE - EXPERT OPINION - QUALIFICATIONS OF WITNESS
The accused is charged with rape of a 14 year old girl. A part of the case against him is DNA evidence linking him to samples recovered from the vagina and underpants of the complainant. At the close of evidence the defence sought the exclusion of DNA evidence, that had been given without objection, on the basis that a new software component "STRmix" of the DNA analysis process had not been properly established as accurate, that the Forensic Science Centre witnesses were not qualified to give evidence about it, and for other reasons.
HELD:
1. The STRmix component of the DNA analysis process has been accepted as reliable amongst other impartial and disinterested experts notwithstanding its relative novelty.
2. The Forensic Science Centre witnesses are qualified to give the evidence they gave about the processes used to extract and analyse DNA in this case.
3. Evidence of the DNA analysis in this case, including the results, are more probative than prejudicial.
4. The application for exclusion of DNA evidence obtained utilising the STRmix program is rejected.

R v FULLER
[2013] SADC 150

This is a very late application for the exclusion of a significant proportion of expert evidence that has already been given without objection in this case. I’m giving this ruling immediately because we are at the very late stages of the case. The jury has concluded hearing the evidence. Tomorrow is scheduled for the final addresses of counsel and simply put, I cannot delay this decision. The interests of justice require that it be made immediately.
The accused is charged with three counts of rape of a 14 year old girl. Part of the prosecution case is forensic evidence said to connect the accused with the vagina and underpants of the complainant. A part of that forensic evidence is DNA evidence.
The prosecution led, without objection from Mr Mancini, two experts in various aspects of the DNA process. They were qualified in a straightforward way by the prosecutor. They then proceeded, without objection, to give evidence about aspects of the DNA process that appeared to relate to their expertise.
DNA reports relating to assessment of the now challenged evidence were tendered, without objection, as were explanatory appendices that relate to the technology used for the various testing processes. Much of the DNA extraction and analysis process is common to all the DNA evidence in this case, however, from the middle of 2012 a new software program component was implemented at the Forensic Science Centre. That program is one component of the analytical stage of the DNA process. It is essentially that new software component that is the subject of this 11th hour challenge.
It is regrettable that it’s come at this very late stage, particularly since it is plain that the defence had known about this new software component since they received the report of Sandra Lee Goldup dated 21 May 2013 which Mr Mancini tells me he received shortly after that time. At p.6 that of report, it is clear that the calculations now being made utilise a new software program called “STRmix”. The report states that STRmix has been developed and validated within Australia and New Zealand. It goes on to explain, in relatively short form, what the STRmix program does. Other parts of the process and the general DNA analysis concept remain the same, as you can see from the remaining common parts of that report. The recent change can conveniently be seen by comparing P12, which represents a description of the earlier process of DNA processing and analysis prior to the new program, and P13, which represents a description of the new process involving the the STRmix program which is essentially new and updated statistical calculation software.
The trial therefore proceeded, the expert witnesses recited their qualifications without objection and then gave their substantive evidence pursuant to their stated qualification without objection. The various results of their analysis in tabular form were tendered as P7, P9, and P10. An older version P11 was also tendered at the request of the defence.
Because there was no indication whatsoever of a challenge to the qualification of those experts at the time, nor to their ability to give the substantive evidence that they gave, no-one was on notice that there might be an application to exclude the evidence on the basis of a want of qualification to comment, a want of proper qualification as an expert or that the system that they implemented had not been proved sufficiently, nor that the output of the system that they used might be in any other way so wanting as to be inadmissible.
The Crown case closed and the defence case opened. Part of the defence case included a forensic scientist who gave evidence criticising the Crown forensic evidence in various respects. In essence, he said that the results were not as reliable or certain as the prosecution experts asserted. He also gave evidence criticising aspects of the new STRmix program. The issue of the reliability and certainty of the STRmix program and any results generated by it was plainly joined. It seemed that it would be a classic jury question.
The challenge now to the expertise of the Crown’s expert witnesses and to the very admissibility of the Crown’s expert evidence has to be decided by me at a very late stage of events. The Crown was not aware of this challenge at any stage when they called their evidence relating to the qualifications of their experts or the evidence that those experts gave concerning the now challenged component of the DNA analysis.
Accordingly any potential challenge to admissibility concerning the specific breadth of each expert’s qualification, the exact degree of wider acceptance of this updated DNA process and, in particular, the STRmix program, was not apparent to anyone. If there had been a voir dire on these topics, the witnesses’ minds would have been directed to that issue, and there would have been examination and cross-examination tailored to that issue. In other words the qualification of the experts and the degree of specific acceptance or otherwise of the system of DNA analysis utilised would have been consciously addressed. Because it was not, I am in a less informed position than I otherwise might be.
No-one has sought to call any evidence on the application itself. I have, however, in the course of argument, been taken to passages in the trial evidence and, in particular, I have been reminded of the evidence Dr Taylor did give about the validation process for STRmix.
I have already referred to Exhibit P13 which is in evidence before me. It is to the effect that the STRmix software component of the DNA analysis process has been developed and validated within Australia and New Zealand. Dr Taylor was asked a single question in examination-in-chief as to the validation process for STRmix. At p.634 he said:
‘Once we had the software programmed with the models in it that we desired, we then obviously had to quite rigorously test all aspects of the software so we checked the calculations produced by the software by hand, to hand calculations, to make sure they were the same, we had the code checked by a professional IT company, we compared the results of STRmix to Cuban interpretations of the same profiles. We also had it beta tested by a number of independent forensic labs around Australia and New Zealand and we have tested it on a number of known mixtures, that is, artificially produced mixtures with a known source.’
He went on in his evidence to say that the validation process was undertaken over six months. He also gave evidence that it was initially utilised in both South Australia and the Institute of Environmental Science and Research in New Zealand. All other forensic laboratories in Australia are now using the software except three who are currently at the stages of validation and implementation. Those three are Northern Territory, Tasmania and ACT.
The inference I draw from that evidence, unchallenged as it was, is that STRmix has been validated, accepted and implemented in each of those other jurisdictions. I conclude that it wouldn’t be being used unless it was regarded by those laboratories as accurate and reliable. I am satisfied that, as Dr Taylor said, the STRmix code has been tested by a number of independent forensic laboratories in Australia and New Zealand and has also been implemented after validation by a number of other forensic science centres around Australia.
Accordingly, in my view, the prosecution have satisfied an onus - which I interpolate to say they didn’t know that they were facing - to show that STRmix has been accepted as reliable amongst other impartial and disinterested experts notwithstanding its relative novelty.
A part of the STRmix process is calibration or integration - I’m not sure what the correct terminology is - in the local jurisdiction or environment, essentially so that the software fits or suits the hardware and system of the local jurisdiction concerned. There is no basis for me to suspect that that component of STRmix was not validated along with every other component by both the independent beta testing conducted independently at the request of the Forensic Science Centre, or by the validation processes that I conclude have been carried out in New Zealand and in several other Australian jurisdictions.
I conclude that the STRmix system has been sufficiently accepted as reliable amongst other impartial and disinterested experts to be capable of producing admissible evidence, and that the resulting evidence is admissible in this case.
I have reviewed the qualifications and experience of the experts. I conclude that they were qualified to give the evidence that they did give.
I turn now to the submission by Mr Mancini that the evidence of the prosecution forensic witnesses concerning the use and operation of STRmix doesn’t have any or any sufficient weight and ought accordingly to be excluded, and the submission that that evidence, at least that which is comprised by Exhibit P10, is misleading or potentially misleading or is more prejudicial than probative.
In this case the further issues that have now been particularised by Mr Mancini as possible bases for excluding the DNA evidence have been well-known to the defence throughout and well before the trial. Those issues include the possibility of secondary transfer, the degree to which a DNA sample located somewhere is probative of actual guilt, the use of the STRmix software together with it’s accuracy and reliability, and so forth. The defence have squarely addressed all of those at trial. They did not object to the evidence that was called about it, rather they met it head-on with their own expert.
In my view there is no reasonable possibility that the jury will be, tacitly or in any other way, misled into thinking that the prosecution evidence is unchallengeable or necessarily needs to be accepted by them. Those issues are, in my view, classic jury issues about which both sides have called evidence and are able to be reasonably and appropriately considered by the jury in their deliberations.
The evidence is plainly probative, and not prejudicial to the defence in any way beyond its legitimate forensic weight in support of the prosecution case.
Accordingly, the defence application to exclude the STRmix evidence comprised in P7, P9, P10 in its entirety, the evidence of Sandra Goldup concerning STRmix and the evidence of Duncan Taylor in its entirety is dismissed.