Part 37 in the Knox/Sollecito
case
Part A: Errors in the principles of DNA forensics and
discovery
The Italian Court of Cassation, the supreme court of Italy,
released a report giving its reasons for overturning Amanda Knox and Raffaele
Sollecito's successful 2011 appeal and sending the case for another trial in
Florence beginning at the end of September.
The report by the Corte
Suprema di Cassazione (CSC) has been translated into English twice, and
a comprehensive critique has appeared. Page numbers below refer to the first translation. The CSC’s report made a number of highly questionable assertions or omissions
with respect to DNA contamination in general and with respect to the evidence
in this case:
“It was also ruled out by the same experts
that contamination occurred in the laboratory. Professor Novelli said that the
origin, the vehicle of contamination must be demonstrated: he specified to have
inspected 255 forensic sample extracts at the Polizia scientifica, had analyzed
all profiles and did not see any evidence of one single contamination; he
excluded absolutely persuasively that the contaminant could be present
intermittently and that DNA could remain suspended, and then fall on a
particular item.” (p. 67)
“but above all it [the discourse of justification] is based on the
erroneous belief that the burden of proof lies on demonstrating the absence of
contamination, whereas the demonstration data that emerged from the technical
advice was based on properly documented reporting activities carried out under
the eyes of the consultants that had nothing to detect, in a clean laboratory
environment, activities conducted according to methods tested, the results of
which could certainly be called into question…” (p. 69)
The moment of DNA contamination
is often not known
Should a mechanism of contamination have to be demonstrated? One problem with this argument is that the
mechanism of contamination may only be revealed years after the fact or not at
all. In the Jaidyn Leskie case (see
below) it was shown that the evidence items were in the same lab closely in
time, but the exact moment of contamination was elusive. Modern forensic DNA testing relies upon the
polymerase chain reaction (PCR) technique, which produces many copies from a
small number of DNA molecules. Dr.
Donald Riley wrote, “PCR is also very similar to what happens when a clinical
infection occurs. Clinicians have known for many years that a single germ
(bacterial cell or virus) contaminating a wound can produce a massive
infection. Similarly, a DNA molecule can contaminate (infect) a PCR and
become a significant problem. The ability of small amounts of DNA to
produce false and misleading results is well-known and well-documented within
the research community, where the technology originated. Anyone who has
caught a cold from an unknown source, or who has a pollen allergy should have
some sense of how easily PCRs are contaminated. Actually, it is probably
easier to contaminate a PCR than to catch a cold since unlike our bodies, PCRs
lack immune systems. The only protection PCRs have is the technique of
the analyst, use of control samples to monitor contaminants and careful
interpretation.” Dr. Riley went on to say, “When contamination
occurs there is rarely any way to confirm how it happened.” The argument that contamination must be proved
suggests an analogy: that doctors should demand that the patient prove the
exact moment at which he or she became infected before the doctor commences
treatment.
In his 2010 report on the Farah Jama case, Former Australian
Supreme Court Judge Frank Vincent wrote (p. 24), “Precisely how it
[contamination] may have happened cannot be determined as the deposition of the
minute quantity of material involved could have occurred in a number of
ways. It is possible to speculate about
the probability of transference through various mechanisms, but ultimately
pointless to do so.” Judge Vincent also
wrote (p. 45), “Whilst there is no absolute bar to conviction based solely on
DNA evidence, the better view is that a conviction should only be returned
where there is DNA evidence and at least one other item of evidence
present which is consistent with the guilt of the offender.” The tiny amounts of DNA involved and the ease
with which DNA transfers make it very difficult to pinpoint exactly when contamination
occurs. Attorney General Rob Hulls said,
“Mr Jama’s case, I think, highlights the
need for everyone in the criminal justice system to better understand the
nature and the limitations of DNA evidence. The perceived value of DNA
evidence means that extreme care must be taken at every stage of the process:
from collection of DNA to the handling of DNA, the testing of the materials and
indeed, the interpretation of the results, to the way in which the evidence is
presented to juries in criminal trials.”
Lack of good forensic technique
If the forensic police fail to follow generally accepted
guidelines in collecting evidence, should the evidence be accepted? The CSC wrote, “Prof. Novelli had agreed that
there are protocols and recommendations, but added that first of all the
operator had to contribute his common sense (ud. 6.9.2011, p. Transcription
59.), otherwise it put in question all the DNA analysis done from 1986 onwards.”
Does the CSC really believe that poor technique is insufficient ground
for considering the evidence to be unreliable?
The laws governing forensic evidence must set the correct incentives in
any criminal justice system. If evidence
collection procedures are flawed yet the evidence is accepted anyway, there is
no impetus to collect it properly: The
same types of errors may convict innocent persons in other cases. In the Busco case, Professor Novelli sounded
a more cautious note than in the present one:
“According to Professor Giuseppe
Novelli, ordinarius of genetics at Tor Vergata, ‘There were
contaminations among items and for [=of?] the analyses on the corsage and on
the bra. The chain of custody of the items did not respect national and
international standards for the conservation of items.’” What common sense actually suggests is that the standards were put into place to
avoid generating a misleading result and that if one violates a particular
protocol there should be a solid, clearly defined reason. The CSC provided none; on the contrary, its words are a smokescreen rather than an argument.
The CSC effectively rejected the idea that the onus is on
the prosecution to show that the forensic police followed correct
protocols. Not everyone agrees, at the
very least. Based upon a number of cases,
a barrister in Australia argued that the prosecution must demonstrate that it
has done everything correctly in a preliminary hearing. Peter Faris wrote, “DNA should never be
admitted unless there is corroboration: that is, there must be some other
evidence of the identity of the offender.
As for the rest of DNA cases, the solution is to put the burden on the
prosecution.” Mr. Faris continued, “For
example, the prosecution would be required to produce evidence to disprove
contamination, whether at the scene or in the laboratory. Strict proof of continuity of exhibits would
be required. The prosecution would have to call evidence of the current
international statistical procedures. Proof by the defence of international
articles and learned writings could be relaxed.” These are much more sensible guidelines than anything provided by the CSC.
Negative controls and the
importance of electronic data files
Another serious objection to the court’s position is its belief
that a witness for the prosecution need only testify that controls were run, as
opposed to actually producing the negative control data in discovery. Negative controls are runs in which template
DNA is deliberately left out. A good way
to determine whether or not global contamination occurred is to examine the
negative controls in the form of electronic data files, because if any DNA
shows up, it must be from contamination.
Among the reasons why examination of negative controls should be done
using the electronic data files (EDFs), which are the raw data used to
construct electropherograms is that this allows the scientist to zoom in on
small peaks. Given the smallness of some
of the peaks in the bra clasp profile and of all of the peaks in the knife
profile, negative controls in the form of paper copies in which
the y-scale were set to two thousand RFUs for example, would be almost useless. Even if one accepts the dubious notion that
the defense bears the burden of proving contamination, then it is indefensible to
prevent the defense from having access to any and all documentation that bears
on this possibility, including but not limited to having the EDFs: It is also essential for the defense to
examine the laboratory protocols, instrument logs, contamination logs and
corrective action files. Professor William Thompson noted, “Under a guideline issued by the FBI’s DNA Advisory Board
in 1998, forensic DNA laboratories are required to “follow procedures for corrective
action whenever proficiency testing discrepancies and/or casework errors are
detected” and “shall maintain documentation for the corrective action.[19]” Obviously, the FBI’s rulings are not binding
on laboratories in another country.
However, good forensic science, like any science, doesn’t change as one
crosses a border from one country into another.
Good forensic principles, such as keeping a corrective action log,
deserve to be adopted universally.
Given their importance in detecting contamination, it is not
altogether surprising that negative controls themselves may be the subject of
forensic fraud. Professor Thompson continued,
“DNA analysts have recently been fired for scientific misconduct, and
specifically for falsification of test results, by a number of forensic
laboratories, including labs operated by the FBI,[14] Orchid-Cellmark (another
large private DNA laboratory),[15] the Office of the Chief Medical Examiner in
New York City,[16] and the United States Army.[17] In all of these cases, the analysts were
caught faking the results of control samples designed to detect instances in
which cross-contamination of DNA samples has occurred.” In the case of the fraud committed by
Jaqueline Blake, Professor Thompson argued that if her superiors had examined
the EDFs, it could have uncovered this misconduct.
The CSC made much of Professor Novelli’s testimony to the
effect that he found no evidence of contamination, but what criteria he used
were not provided. The CSC should state whether or not Professor Novelli used the electronic data files in
his examination. If Dr. Novelli did, then he had evidence that has been denied to the
defense, which raises the very troubling question of why the prosecution and its
witnesses should have access to something that was denied to the defense and
its expert witnesses. If he did not,
then his review was incomplete, and it makes his claim that there was no
contamination quite dubious at best.
Every DNA expert whom I have consulted or whose thoughts on the matter I
have read has been unequivocal in the importance of reviewing the EDFs; that
they should be released in routine discovery is a key recurring subject in this
blog. Dr. Mehul Anjaria wrote, "DNA
analysts import ‘raw’ data from capillary electrophoresis instruments into
software that assists in evaluating the DNA profiles. The DNA analyst can
review peaks by zooming in and looking at height, morphology, and location to
assist in determining if they should be reported as DNA alleles. The printed
data can be edited by the analyst to show only the peaks being reported. Thus,
it is imperative that a reviewer have access to the raw data and be able to
independently evaluate the raw data in the necessary software to determine if
the reported alleles are consistent with the actual data. The laboratory’s
interpretation guidelines are another necessity for the review."
Independent case reviews (including but not limited to an
examination of the EDFs) turn up problems in about two thirds of all cases,
according to Professor Dan Krane. The
Patrick Waring case in Australia is just one of many examples. Dr. Theodore Kessis commented on the Benjamin
LaGuer case: “It is highly improbable that any given forensic DNA laboratory will take
it upon itself to contact its accrediting bodies or the press and state for the
record how often they make mistakes… To best understand the weaknesses associated
with DNA testing we must rely upon the empirical, the occasions in which such
deficiencies are revealed either by the press or internal review of a lab’s
documentation of such problems by a defense expert. A close look at either
reveals that indeed many instances of DNA testing errors have lead to the false
conviction of individuals.”
Assertions from witnesses for
the prosecution cannot be taken at face value
What should we make of Professor Novelli's claim, which the CSC
accepted, of having examined hundreds of pieces of evidence and not observed
evidence of contamination? This is
problematic on several grounds. First,
it is unclear what Professor Novelli’s criteria or methods were for proving or
disproving contamination. Therefore,
there is a decided lack of transparency in the process. Second, Professor Novelli did not explain the
presence of multiple profiles on the clasp (see part B). Third, if the CSC accepts the judgment of a
prosecution witness over independent experts, it suggests that the court has a
strongly pro-prosecution bias. Nor can
it be persuasively argued that Professor Novelli’s credentials in the area of
DNA forensics are stronger than those of the independent experts, Conti and
Vecchiotti. Professor Novelli’s main
research interest is in the area of medical genetics with some additional
research into one specific area of forensic genetics (that of single nucleotide
polymorphisms), whereas Conti and Vecchiotti have focused more narrowly on
forensic genetics.
Fourth, forensic witnesses make false statements while
testifying. The CSC accepted the prosecution’s
claim that negative controls were performed, despite tacitly acknowledging that
they were not part of the case file. In
his report on the Leiterman case, Dr. Theodore Kessis highlighted the fact that
a negative control electropherogram showed contamination, and that Dr. Milligan
labeled it as such. Despite this, Dr.
Milligan later testified that no contamination occurred and that if it had, his
reports would have documented it. In
“Tarnish on the Gold Standard” Professor William C. Thompson wrote, “The DNA
analysts in the Houston Police Crime lab came up with an easy solution— they
simply failed to run extraction blanks (although they claimed in testimony that
they had run all necessary controls).” Thus personnel from crime labs sometimes falsely testify that they followed
protocols and saw no evidence of contamination.
Furthermore, even if the lab itself detected no problems, independent
case review often does (see above). The
implied standard of the CSC (proof by assertion of the forensic police and the
prosecution’s witnesses without documentation) flies in the face of the
principle that justice must be seen to be done.
It is negligent of the court to accept that when a prosecution
witness asserts something as true, the assertion should be sufficient as proof
that it is.
Ms. Stefanoni claimed that there
had been no contamination in her lab in seven years, but it is difficult to see
how a large lab could have an error rate of zero. Reporting on some crime labs in California, Maura
Dolan and Jason Felch wrote, “The
number [of errors] reported was small considering overall caseload -- 3,100
over five years – but [UC Irvine Professor William C.] Thompson said mistakes
caught by labs ‘undoubtedly’ make up a small fraction of errors. (In fact, he
said, labs that report the most are probably better run than those that claim
none.)”
The Six-day gap in testing
A central pillar of the CSC’s report (p. 68) is that the
six-day gap in testing the knife was sufficient to rule out laboratory
contamination: “Dr Stefanoni (technical consultant who wrote the advice art. 360
codaproc.pen.), heard also on appeal, had repeated that there was no evidence
of contamination: investigations on the knife had been conducted six days
before the last DNA trace of the victim, then the analysis had been blocked for
a further six days, a period deemed by the same expert Vecchiotti time to be
sufficient to prevent laboratory contamination, as declared in the SAL report,
wrongly reported as missing initially.”
Whether this assertion is true is difficult to verify without full
discovery taking place. It is not clear
whether all of the DNA electropherograms were released to the defense, and what
might be on the possibly unreleased electropherograms obviously cannot be
known. For the sake of argument,
however, we will assume that there was a six-day gap. The CSC cherry picked one statement to bolster
its preferred conclusion; both Conti and Vecchiotti still believe that the DNA
evidence is unreliable in spite of the gap. Moreover, the six-day argument does not
address the possibility of secondary/tertiary transfer before or during
collection of the knife at all (see part B below).
The assertion that a six-day gap is sufficient to exclude
laboratory contamination is flawed on at least four grounds. One, no primary or secondary source in the
forensic literature has made this claim, to the best of my knowledge. Two, any rule of thumb to that effect that
was generated with respect to ordinary amounts of DNA may or may not be valid
with respect to low template amounts of DNA.
In response to a question of mine Dan Krane replied, “Was Meredith's DNA processed in the lab before
the knife? If so, then I don't accept that six days is sufficient to
exclude contamination in the lab. That would just be bad practice, plain
and simple.” Meredith’s reference
profile was indeed generated before the six-day gap and was obviously much
higher in amount that the knife profile, which falls into the low template
range. This is doubly problematic. Generally, one wants to run the reference
profile last, because it is in high quantity, but low template quantities of
DNA are especially worrisome. In
response to a previous question of mine Dan Krane wrote, “There is absolutely
no question but that contamination is a much greater problem in LCN cases than
conventional DNA testing. The reasons that it is a greater problem are both
because it is easier to detect contaminants ([Sara] Gino's point) and because
it is easier to transfer (and to transfer without knowing) smaller amounts of
DNA than larger amounts of DNA.”
Three, imagine that a forensic worker uses a tool such as a
ruler in his or her examination of one piece of evidence, transferring DNA to
it. Then the ruler is not used again for
a week. If it were not properly cleaned,
the ruler would be quite capable of transferring DNA to a new item of
evidence. A recent study of the surfaces
and instruments used in autopsies indicates that concerns about contaminated
surfaces are well-founded. The authors
wrote, “Using DNA-free swabs, we took samples from instruments used during
autopsy and autopsy tables. Surfaces and instruments were routinely cleaned
before sampling. Swabs were subjected to
different PCRs to quantify the total amount of DNA and to amplify individual
specific STR-markers. In most samples,
alleles that could be linked to bodies that had been autopsied before were
found. Furthermore, we could show that a
DNA transfer from the autopsy table to a body was detectable in four out of six
cases investigated.” A review article on
trace DNA cautioned, “Fingerprint brushes
are able to transfer amounts of DNA between exhibits that could generate
profiles and may retain biological evidence for a considerable period of time
[205,206].” Dick Warrington also
warned of the dangers of cross-contamination from instruments. In his
report on the Gary Leiterman conviction, Dr. Theodore Kessis wrote (p. 9), “It
must be noted however that contamination
errors have been documented where no direct processing link between sample and
contaminant have been established, raising the specter that a source of contamination can linger in a
laboratory for some time.”
Four, in the Jaidyn Leskie case there was a two-day gap in
between examination of the condom used in the alleged rape of Ms. P and of
Jaidyn Leskie’s clothing at the laboratory.
The police ruled out the possibility that Ms. P was responsible for
Jaidyn’s death, and Professor Krane’s examination of the electronic data files
suggested that the possibility of a coincidental match was quite unlikely. Therefore, the only plausible explanation is
contamination at the stage of handling evidence. There was also a one-day gap in the
examination of two women in the Farah Jama case, and these two examinations are
likely to be the cause of the observed DNA contamination, as discussed in the
Vincent Report. Given that one-day and
two-day gaps are demonstrably insufficient to ensure a lack of contamination,
then why would a six-day gap be sufficient?
Observation of the collection or
testing of the evidence
In the passage from p. 69 (“documented reporting activities carried out under the eyes of the
consultants that had nothing to detect…”) quoted above the court seems to
put emphasis on the observation of the collection and seemingly also the
laboratory work by the consultants employed by the defense. This raises some serious questions. With reference to the collection techniques
of the forensic police, the court’s stance is nonsensical. On the one hand, the court notes that the
consultants did not object. Yet the
court also wrote, “the
vehicle of contamination must be identified
in order to defuse the data offered by the technical consultants, it not being enough to assume
insufficient professionalism of the operators in sampling…” (pp. 68-69) The court implies that pointing out technical
flaws using recordings of the evidence collection long after the fact would not
be good enough, despite the fact that this is what consultants or independent
experts who are brought into the case later on would have to do. Equally serious is the fact that no one can
see DNA. Therefore, when one observes
the collection of the evidence, all one can comment on is whether or not the
police used good technique (a great deal turns on the meaning of the word
assumed). It seems as if poor technique
is not sufficient in the court’s eyes unless it somehow demonstrates the
vehicle of contamination, which it is quite unlikely to be able to do. On the one hand, the best technique
does not guarantee that contamination will not happen; on the other hand,
errors in technique increase the chances that it will.
If the court also means say that one must raise objections
during the observation of testing the items of evidence, then their stance is
equally problematic. Dan Krane observed,
“Having the opportunity to witness the testing of samples is of marginal
utility at best. Reviews of the underlying data for DNA tests often reveals
alternative interpretations of the evidence samples, especially in
circumstances were small amounts of DNA are involved and it is difficult to
distinguish between signal, noise, and technical artifacts. Observing testing
rarely provides any more insights than what should be possible from a review of
contemporaneous notes that should be part of a lab's case file. Witnessing
testing is far from a cure-all. Problems such as contamination of samples can
easily arise before a sample arrives in a laboratory yet could not be detected
by an expert observing the testing process itself.”
Sporadic and global
contamination
The CSC wrote, “he
[Novelli] excluded absolutely persuasively that the contaminant could be
present intermittently and that DNA could remain suspended, and then fall on a
particular item.” Despite what the CSC believes, contamination may be
sporadic (“intermittent”), as well as global.
Dr. Donald Riley wrote, “Negative controls also can't rule out
contamination of individual samples. The individual samples lack
individual signs of contamination if it occurs.” Other authors whose works are quoted
elsewhere in this entry also made the point that contamination can be
sporadic.
Conclusions, Part A
The CSC’s report suggests an almost willful misunderstanding
of DNA profiling. Its failure to put any
significant burden on the prosecution and forensic police is out of step with
other nations. The CSC’s stance that
contamination must be proved fails to recognize that the exact moment of
contamination may never be established even after thorough studies, as were
conducted in both the Jaidyn Leskie and Farah Jama cases. A claim that a
six-day gap in testing is itself assurance of a lack of contamination is
unsupported in the literature and the time gaps in those two cases (during
which contamination must have taken place) undercut their contention. Even if one accepted the dubious argument
that contamination must be proved, then it is incumbent upon the prosecution to
turn over the electronic data files (including negative controls), machine
logs, standard operating procedures, and other relevant forensic data. Perhaps the most serious error the court made
is to accept the assertion that there was no contamination without the methods
or standards of proof even being defined.
Under this court’s implicit rules, the prosecution would need only to
offer DNA evidence and perhaps testimony from a friendly witness about the lack
of contamination to ensure that the DNA evidence would be accepted in any
criminal trial. Even without such
testimony, the lack of insistence upon unfettered disclosure would severely
hinder the defense from ever successfully challenging forensic DNA evidence on
the grounds of contamination or misinterpretation of the data. Under such rules false convictions based upon
faulty DNA evidence are inevitable because a defense based upon a theory of
contamination would be virtually impossible to prove.
Part B: Specifics of this case
Kitchen knife
The CSC ignored innocent DNA transfer to the knife as an explanation
for the knife DNA profile. It is
possible that Meredith’s DNA was deposited onto the knife before or during
collection by secondary or tertiary transfer, as was suggested by Dr. Alexander
Kekule, among others. Professor Gregory
Hampikian’s study of DNA transfer via gloves is further evidence of the plausibility
of such mechanisms. Dried blood is a
good source of DNA because small particles can flake off easily, and clothing
is a known carrier of DNA. One can
envision tertiary transfer from airborne DNA to officer Gubbiotti’s clothing,
to his gloves to the knife. However, it
is not up to the defense to prove a particular route of contamination; instead,
it is up to the prosecution to show that they did everything correctly, as
discussed in part A. There was no reason
whatsoever for anyone to unpack and repackage the knife. Compounding the error, it was unwise to
assign officer Gubbiotti to anything having to do with evidence collected at
Sollecito’s flat, given the fact that he had been to the women’s flat on the
same day. Reporting on the Patrick Waring case Estelle Blackburn wrote, “In
court, police conceded they had not followed best practice in the case. Various officers said that the Central Park
scene was left unguarded from 1.25am on the night, it was a week before it was
searched, and the same officers had visited the homes of the girl and the
accused which allowed for contamination of evidence.” What is
not best practice in Perth is not best practice in Perugia.
The CSC's report does not seriously consider the lack of
blood on the kitchen knife the prosecution has argued was one of the murder
weapons. Yet the lack of blood plus the
presence of starch call into grave question whether the DNA seen in the profile
was actually physically on the knife. It
is doubtful that one can clean a knife of blood and not of DNA, whether the
scratch on the knife observed by Stefononi were real or not. Detergent water should lyse human cells,
releasing DNA and other cellular contents.
Bleach is used routinely to destroy unwanted DNA in molecular biology
laboratories. If the knife were cleaned
with bleach, the remaining DNA might show evidence of degradation (peaks associated
with long DNA fragments would be smaller than those associated with short
fragments). Yet there is no obvious
trend of smaller peaks moving from left to right in the knife
electropherogram. Moreover, there is no
evidence that the knife was carefully cleaned; on the contrary, the presence of
starch suggests that cleaning was not particularly assiduous. Forensic scientists Elizabeth Johnson and
Gregory Hampikian wrote, “it is
unlikely that all chemically detectable traces of blood could be removed while
retaining sufficient cells to produce a DNA profile consistent with the
victim.”
Another fundamental failure of the report stems from the
fact that the knife profile is a low template sample. The CSC wrote,
“The
discourse of justification, as maintained by the plaintiffs, did not take account
of the authoritative voices of dissent concerning the presence of contaminating
agents; adequate explanation was not offered as to how this assumption had to
cover only some (the most demanding in terms of defense) examined tracks and
not others; but above all it is based on the
erroneous belief that the burden of proof lies
on demonstrating the absence of contamination, whereas the demonstration data
that emerged from the technical advice was based on properly documented
reporting activities carried out under the eyes of the consultants that had
nothing to detect, in a clean laboratory environment, activities conducted
according to methods
tested, the results of which could certainly be
called into question, but for their probative value, not for the operations
carried out by preceding contradictory technique, from which did not emerge
critical profiles at the time, but only in retrospect (about the decision at
First Instance had dwelt from p. 289 to p. 298 on an abundance of topics only
partially refuted in an appropriate manner, so that equally significant were
the observations of Dr Stefanoni, brought to the attention of the court of
Second Instance, at the hearing on 6.9.2011).”
The implicit assertion
that the Rome lab was clean enough for low template DNA forensic typing is made
dubious by the fact that dedicated facilities have been constructed to carry
out this sort of work. To combat
the increased risk of contamination in low template DNA profiling, special
facilities and handling techniques are needed, as noted in the New Zealand
Herald: “The ESR has spent $1 million building special anti-contamination areas
at its premises in Auckland, Wellington and Christchurch. Protocols are being
developed for crime scenes where the LCN technique is used and for the handling
of samples from collection through to courtroom. LCN crime scenes will be divided into cold,
warm and hot zones hot being the crime zone. Clothes are put on and discarded
at each zone to minimise the risk of contamination.” The crown prosecution service wrote, “The FSS LCN test requires an ultra-clean
laboratory and so is more expensive and less widely offered than the standard
test.... The site of this bespoke laboratory is remote from other DNA Units,
operates stringent entry requirements, is fitted with positive air pressure and
specialist lighting and chemical treatments to minimize DNA contamination.”
In the article “Setting Up a PCR Laboratory” Theodore E.
Mifflin wrote, “Air handling. For extremely high-performance PCR laboratories
that will be involved with detecting very-low-prevalence DNA or RNA molecules
(e.g., infectious disease agents in clinical samples), additional measures may
be necessary to prevent contamination from the air being recirculated between
the pre- and post-PCR laboratories.” Dr.
Mifflin’s main focus is pathology, but his points about low levels of DNA are
germane to low template DNA forensics in that both situations use polymerase
chain reaction (PCR) to amplify very small quantities of DNA. With respect to the knife profile it is
especially worrisome that Ms. Stefanoni presumably used a Speed Vac to
concentrate the DNA. The sample would
have been vulnerable to airborne contamination upon release of the vacuum or to
any DNA left in the Speed Vac from previous samples. At the very least, a control should have been
run that was also subjected to the same concentration step.
Indeed, the need for specialized facilities and handling
techniques when working in the low template region is recognized even by the prosecution’s own expert witness. A review article coauthored by Giardini,
Spinella, and Novelli stated, “Thus in these conditions [less than 100
picograms of DNA] there is a greater probability of artefacts, partial profiles
with fewer alleles, contamination, preferential allele amplification, the complete
absence of one allele (allele drop-out) in heterozygous loci and the
nonspecific generation of extra alleles (allele drop-in) [78,79]…. Very few
laboratories perform low template DNA typing properly, because it requires
dedicated facilities and great experience, although there are several published
methods for the interpretation of such profiles [80-82].” At the time of the murder, the Rome lab was
not even certified for standard DNA testing.
If a sample is handled in a regular DNA facility, without the special
precautions available in dedicated low template DNA facilities, the risk of
contamination is unacceptably high. Even
if all such precautions are followed, some prominent DNA scientists find LCN
DNA unreliable for forensic profiling.
Nothing in the CSC’s report can be construed even as recognition (let
alone an honest assessment) of the special problems inherent in low template
DNA forensics or the lack of special precautions taken in the Rome lab.
bra clasp
The CSC’s views on the bra clasp profile are at least
equally problematic. This blog has
previously discussed DNA contamination on several occasions. Van Oorshot and colleagues wrote, “From a
theoretical perspective, any DNA deposit that is not immediately relevant to
the crime being investigated can be viewed as contamination. In this light,
gross or sporadic contamination may appear at any point: (1) before the crime
has been committed; (2) in the interval between the crime and securing the
crime scene; (3) during the investigation of the scene; and/or (4) within the
laboratory.” There are alleles in the
autosomal DNA profile that are unidentified (not belonging to Meredith Kercher,
Raffaele Sollecito, or Amanda Knox).
There are also extra alleles in the YSTR profile that do not belong to
Raffaele Sollecito. Even using a
conservative threshold of 50 RFU, one locus has DNA from two additional males. If one uses a threshold closer to that
employed in the knife electropherogram, the number of loci with two additional
alleles (therefore two additional donors) goes up. It is unrealistic to posit that all of these
alleles were deposited during the crime, leading to the conclusion that it was
deposited in a way that is unrelated to the crime. Therefore,
the bra clasp has already been shown to be contaminated by the Van Oorshot
definition.
The CSC noted that a cigarette butt was the only other piece
of evidence where Sollecito’s DNA was found (p. 68). They go on to make the fallacious and foolish
equivalence that found DNA equals all DNA.
In other words the court believes that the cigarette is the only source
of Sollecito’s DNA in the women’s flat.
Yet Mr. Sollecito left fingerprints and presumably used a towel to wash
his hands before preparing food. Unless
the forensic police sampled everywhere in the flat, there is no reason to
believe that they found all of Mr. Sollecito’s DNA.
They also discount the problem of the six-week gap between
the crime and the collection of the clasp, saying that the house was “boarded
up.” This is highly misleading: the police themselves moved Meredith’s
mattress and other items from one place to another. The bra
clasp itself moved more than a meter from the point at which it was first seen
to the point at which it was collected.
The reason that the forensic police should do their work first is to
collect samples before other police officers potentially contaminate the
scene. The American Bar Association’s
standards for collecting evidence reads in part, “Standard 2.1 Collecting DNA evidence from a crime scene or other
location
(a) Whenever a serious
crime appears to have been committed and there is reason to believe that DNA
evidence relevant to the crime may be present at the crime scene or other
location, that evidence should be collected promptly.”
The reason for being prompt is to minimize the chances of
contamination: “The most important
aspect of evidence collection and preservation is protecting the crime scene from the time the first officer or responder arrives
until the last piece of evidence has been noted and collected without being contaminated.”
Furthermore, the bra clasp was allowed to decompose during
the time that the forensic police stored it.
If it could be retested, it is possible that additional evidence of
contamination would surface. The CSC
appears oblivious to this additional handicap that the defense faces, due to
the impossibility of retesting. Some
commenters have argued that when a piece of evidence goes missing or decomposes
while in police care, the court should be obliged to accept the defense’s
interpretation of the evidence. That
principle would produce a more just state of affairs than exists presently.
Discovery
The CSC wrote (p. 69), “…in a context in which negative controls were made by Dr. Stefanoni,
checks that had been stated too superficially to be missing by the experts,
simply because they were not attached to the report.” This is a tacit admission that the negative
controls were never turned over to the defense.
An anonymous expert in Italian law (one who is unaffiliated with this
case) told me that an Italian lawyer can get anything he wants in
discovery. Yet the electronic data files
were never released to the defense during the trial of first instance, nor is
there any evidence that they have been released since. Perhaps negative controls were simply not run
(this would not be the first time in the history of DNA profiling that controls
were simply left out). Even if one
claims that the negative controls were made available in some format, it is
certain that they were not made available in the form of EDFs.
Moreover, the lack of forensic discovery was more pervasive
than solely the failure to release the EDFs. In an article from 25 April 2010 Fiona
Ness quoted Greg Hampikian “’It’s fairly routine in the US that I send a
request and get what I want. But in the Knox case I haven’t been able to get a
copy of the standard operating procedures of the lab and without that, it’s
hard to see if they even followed their own guidelines.’” Not only the EDFs, but also the standard
operating procedures and machine logs should have been released but were
not. In addition it is possible that a
number of electropherograms were also withheld, some from the same time as the
knife electropherogram.
The CSC was silent about the multiple failures of the
prosecution to turn over forensic data and about Ms. Stefanoni’s false testimony
about the quantity of DNA on the knife.
Coupled with its fecklessness over Stefanoni’s stonewalling with respect
to the results with tetramethylbenzidine, the court’s position contradicts any
claim to the effect that an Italian lawyer can get what he wants. The CSC should have insisted upon complete
release of all forensic information as a matter of principle, but even more so
given the problematic, questionable nature of the DNA and presumptive blood evidence
in this case. Moreover, to demand that
the defense prove contamination without simultaneously demanding complete
release of all forensic data, especially
the negative controls, is absurd:
Lack of complete discovery has repeatedly undermined the defense in this
series of trials; therefore, the trials have been grossly unfair. The Duke lacrosse case might have played out
very differently if North Carolina did not have an open discovery law, even one
that is not perfect. Instead of
correcting the failures to comply with reasonable discovery requests in the
present case, the CSC has put its stamp of approval on them. If I were an Italian citizen, I would be
asking myself whether the CSC would uphold the right to discovery more
generally or would it be similarly obsequious to the prosecution.
Forensic technique
The CSC wrote, “So the objective data collected indicating
the absence of evidence (already highlighted in the judgment of first instance from
p. 281 onwards, which made reference to the video recording of transactions
that took place with the precautions of reporting protocols of the forensic
team, accustomed to interventions of this nature) giving credit to the
hypothesis of contamination…” They also
wrote, “but above all it is based
on the erroneous belief that the burden of proof lies on demonstrating the
absence of contamination, whereas the demonstration data that emerged from the
technical advice was based on properly documented reporting activities carried
out under the eyes of the consultants that had nothing to detect, in a clean
laboratory environment, activities conducted according to methods tested…” (p.
69) Yet
one of the defense consultants, Professor Potenza, did raise objections to the
testing. Therefore, if the court is
asserting that no one objected at this stage of the process, it is simply
mistaken.
The CSC believes that the evidence collection was handled
properly; this is one of the most remarkable and troubling positions it took in
the section of the motivations document on the genetic investigations. For example, how could the CSC know whether
or not the lab was clean? The independent
experts, Conti and Vecchiotti, listed some of the recommendations concerning
forensic evidence. “handling of the
objects must be reduced to the minimum possible, and the objects must not be
reopened [once packaged], not even for interrogation purposes..” and “Once
sealed, the containers must not be reopened outside of the laboratory
environment.” This guideline was ignored
for the knife taken from Sollecito’s flat, as discussed above. Conti and Vecchiotti also quote guidelines
concerning gloves that will come as no surprise to readers of an entry here, whereas Stefanoni’s beliefs on how frequently gloves are well outside the
mainstream. Alex Heigl wrote, “There was
laughter in the courtroom at one point when the jury was shown a video of the
detectives collecting DNA evidence, doing the opposite of what the experts
[Conti and Vecchiotti] had just described as the correct way.” Douglas Bremner wrote, “Professor Conti
showed a film of the evidence collection that they said they had analyzed frame
by frame in Amanda Knox case. ‘There are a number of circumstances that don’t
follow protocol or proper procedure’ and the work of the scientific police was
not supportable.” Does the CSC really
believe that these errors never happened, or do they just expect a casual
observer of this case to take their word for it?
"Everything is possible"
The CSC wrote, “The Court of Second Instance, supported the
probable contamination advanced by experts, based on the "anything is
possible", which is not an expendable [usable] argument, because of
its generality…” The CSC plucked this
phrase spoken under needlessly hostile cross examination and quoted it without
context; it is a dismaying and troubling misrepresentation of Conti and
Vecchiotti’s position. The
Conti-Vecchiotti report used a variety of sources to document the manifold
failures of the forensic team (see above).
What they did not do, and what no one should have expected them to do,
was to find a particular moment where contamination must have occurred.
The CSI effect and the cases of
Farah Jama and Lukis Anderson
DNA evidence is based upon solid science, but there are some
limitations inherent in DNA profiling that deserve due consideration from the
justice system. Despite these
limitations DNA evidence can exert an effect on the minds of judges and jurors
that is out of proportion to the weight it deserves. This blog has previously covered the case of
Farah Jama on several occasions, but the recent case of Lukis Anderson also
underscores the problem of unintended DNA transfer leading to false
imprisonment. A spokesman for the
ambulance associated with the paramedics that picked up Mr. Anderson said, “We don’t have any definitive answers as to what
may have caused the transmission of DNA going from one person to another.” The Anderson case is also additional evidence
that one cannot easily determine a route of DNA transmission, even when it is
obvious that it happened.
Conclusions, part B
The CSC accepted a number of extremely dubious arguments
with respect to the DNA evidence in the Knox/Sollecito case while at the same
time failing to address good ones. The CSC
did not answer the question of how there could be DNA on a knife without
blood. This paradox alone is strong
evidence for contamination. Remarkably
the CSC did not acknowledge the obvious and independently verified failures of
the forensic police to follow good practice, both with respect to the knife and
the bra clasp. Its position with respect
to the bra clasp, that there was no way for one particular sample of Raffaele’s
DNA to have made its way to the clasp, is at best terribly misguided. The CSC also did not acknowledge that the
multiple profiles on the bra clasp force one either to claim many unknown
assailants or to acknowledge that it is contaminated. On top of these other questionable positions,
by failing to compel the release of the forensic data, the CSC eviscerated Ms.
Knox and Mr. Sollecito’s right to challenge the evidence against them. The court’s stance not only puts Ms. Knox and
Mr. Sollecito at risk of a false conviction, but doing so also greatly
increases the odds that future defendants will be convicted on the basis of DNA
evidence that is equally erroneous. One
might hope that a nation’s supreme court would be a bastion against the state’s
encroachment on an individual’s liberty, but the CSC’s report gave no evidence
that such hope is justified; instead, the court has acted as if it has fallen
prey to the CSI effect.
Former Australian Supreme Court Justice Vincent summarized
the problem perfectly in his report on the Farah Jama case (p. 11). “In
other words, the DNA evidence was, like Ozymandias' broken statue in the poem
by Shelley, found isolated in a vast desert. And like the inscription on the
statue's pedestal, everything around it belied the truth of its assertion. The
statue, of course, would be seen by any reasonably perceptive observer, and
viewed in its surroundings, as a shattered monument to an arrogance that now
mocked itself. By contrast, The DNA evidence appears to have been viewed as
possessing an almost mystical infallibility that enabled its surroundings to be
disregarded. The outcome was, in the circumstances, patently absurd.”
