Thursday, September 22, 2011

Amanda Knox and Raffaele Sollecito: There is no mixed blood

Part 32 in the Knox/Sollecito case

Introduction
The Conti-Vecchiotti report casts further doubt upon the reliability of the bra clasp and the kitchen knife, two of the strongest, yet seriously flawed, pieces of evidence against Ms. Knox and Mr. Sollecito in this case. With the appeals trial expect to wrap up shortly, it is time to reexamine the strength of the other biological evidence against them, especially the mixed DNA traces. An anonymous pro-guilt blogger did so several months ago. The related question of whether or not the luminol-positive areas (some of which had DNA profiles) were truly blood was the subject of a previous entry here and a discussion at Let’s Talk about True Crime.

Which samples contained mixed DNA?
There are two amorphous, luminol-positive spots in Filomena’s room (Reps. 176 and 177), although the alleles which may be Amanda’s are quite weak. There is a shoe print in the hallway (Rep. 183). In the shared bathroom there is a sample from the bidet drain, the cotton flock container on the floor, and from the inside of the washbasin (Rep. 137). There is no DNA from Amanda in the murder room, either by itself or mixed with Meredith’s DNA.

Did the jury in the first trial find the mixed DNA to be significant?
Barbie Nadeau wrote, "The defense's biggest mistake, according to interviews with jurors after the trial, was doing nothing to refute the mixed-blood evidence beyond noting that it is common to find mingled DNA when two people live in the same house. The jurors needed more than that. 'To have mixed blood, you have to both be bleeding,' one of them remarked to me after the verdict. It was obvious that Meredith was bleeding, but why was Amanda bleeding?'" ("Angel Face," pages 152-153)

Does mixed DNA indicate mixed blood?
The passage quoted above implies that at least one juror thought that the two were equivalent, but this is a fallacy. There are three possibilities with respect to mixed DNA and blood: neither source, one source, or both sources of DNA could be blood. A commonly repeated fallacy of this case is to equate a mixture of DNA with a mixture of blood (to assume possibility three). For example in a recent Dateline broadcast Barbie Nadeau said, “I have to say, I live in a house with 3 people, my two sons and my husband, I guarantee you I have no mixed blood with any of them, anywhere in my house. I don't bleed where they bleed, we don't bleed at the same time. There would never be my mixed blood, their blood and my blood anywhere ever.”

The false equivalence between mixed DNA and mixed blood dates back at least to the publication of the book Darkness Descending. In this book Colonel Luciano Garofano, a retired officer of the Carabinieri, said, “However, here is the electropherogram and you can see that the RFU value is very high, so the sample is undoubtedly blood, which is the body fluid that provides the greatest amount of DNA. In some cases you see higher peaks of Amanda's DNA than Meredith's. Amanda has been bleeding. Nor is it old blood, as the defence might say, because blood decays fast. We have the same result on the cotton-bud box. The light switch was over-scrubbed, but from the film the way the cotton-bud was good enough. There too we have mixed blood. So that's pretty significant for Amanda, Unfortunately for her, she bled at the same time Meredith was bleeding. That's a lot to explain." (Darkness Descending, page 371).

There are several reasons why Garofano’s interpretations are wrong. In response to a question of mine, Professor Dan Krane wrote, “Inferring tissue source from peak heights is just plain silly -- to the point of being absolutely outrageous. It hardly bears more comment than that, but if high peaks mean blood then what would you expect from semen which has a ten to one hundred fold higher concentration of DNA?” Professor Greg Hampikian concurred with the view that peak heights were not an indication of whether or not blood was the source of DNA. The peak heights for Raffaele’s profile on the cigarette butt were reported to me as being about the same height as those on the cotton box, and the former are presumably from saliva.

Colonel Garofano’s claim that the DNA from blood decays quickly is difficult to evaluate. A paper (Park et al., “Direct STR Amplification from Whole Blood and Blood- or Saliva-Spotted FTA without DNA Purification,” J. Forensic Sci., March 2008, Vol. 53, No. 2, 335-41) showed that 1-2 year-old blood samples gave strong signals in DNA profiling when stored in the form of FTA cards (which contain stabilizers); therefore, their study does not exactly refute what Colonel Garofano claimed, but it does not support his claim, either. However, this paper also showed that saliva gave tall peaks in DNA profiling, which is one more indication that peak height cannot be used to infer the biological origin of a sample. The rate of decay of a DNA sample depends upon so many factors that dating DNA by its degradation is not practical. Furthermore, even if one were to accept that DNA peaks from blood did degrade very quickly, one might have to conclude that the luminol-positive, mixed-DNA samples were not blood, inasmuch as the luminol was applied on 18 December, more than a month and a half after the crime.

What did Massei conclude with respect to the mixed DNA samples?
In contrast to the juror quoted above, the Massei report did not assume that mixed DNA was equivalent to mixed blood (pp. 278-279, English Translation). “It should then be highlighted that in that same bathroom various [300] trace specimens were found, of a mixed nature and testing positively for blood. It is true that, according to what was asserted and explained, it is not possible with a mixed trace specimen that tested positive for human blood to determine which of the trace’s contributors the blood belongs to. In this case, however, non-mixed traces were also found, which were shown to be of a haematological nature [i.e. blood] and turn out to have the biological profile of the victim.” The report continued (p. 279), “And it is probable - not necessary, but probable - that during the following act of scrubbing the hands to remove the blood, he/she left the mixed trace consisting of Meredith’s blood and of cells which had been removed by rubbing during the act of washing.”

The Massei motivations report acknowledged that Amanda had no wounds and therefore was not bleeding. It also noted that DNA by itself gives no indication of when it was deposited (see below). In summary Massei thought that the mixed DNA did not necessarily indicate mixed blood, but he believed that the traces were deposited simultaneously, at least partially on the basis of Amanda’s declaring that the bathroom was clean on the afternoon of 1 November. Such a position is problematic in that a clean bathroom does not necessarily imply a DNA-free bathroom. Moreover, there is no reason to rule out Amanda’s depositing the DNA on the morning after the crime in addition to the possibility that she deposited it before the crime.

Does mixed DNA have to be deposited at the same time?
In general the presence of DNA almost never gives an indication of how or when it was deposited. The abstract of an article (“DNA profiling of trace DNA recovered from bedding,” Forensic Science International, Volume 159, Issue 1, 25 May 2006, Pages 21-26) on DNA profiling states in part: “The results indicate that the DNA profile of an individual can be obtained from bedding after one night of sleeping in a bed. The DNA profile of the owner of the bed could also be detected in the foreign bed experiments. Since mixed DNA profiles can be obtained from trace DNA on bedding, caution should be exercised when drawing conclusions from DNA profiling results obtained from such samples.” This is a good example of mixed DNA that could not have been deposited simultaneously.

How common is mixed DNA?
Head of the US National Institute of Standards and Technology's genetics group, “[John] Butler has reviewed more than 5000 DNA samples from 14 US labs and found that mixing is a common occurrence: 34 per cent of the samples he studied included DNA from two people, while 11 per cent were three or four-person mixtures.” Although some fraction of the two-person samples are from the victim and the perpretrator, it is not reasonable to suppose that this is the case for all of them, let alone the three or four-person mixtures.

Are there other mixed DNA samples in this case?
In addition to the mixed DNA of Meredith and Amanda at the girls’ flat, there are also three mixed DNA samples containing Amanda’s and Raffaele’s DNA at his flat. One was found in Sollecito’s bathroom, one was found in his bedroom, and one was found on a pair of rubber gloves. The former two are also luminol-positive, but the identity of the luminol-reactive substance is not known. All three mixed samples are likely to be the result of cohabitation. Amanda’s and Raffaele’s DNA was also found on a cigarette butt at the girls’ flat (p. 193, Massei Report, English translation). The cigarette butt is also interesting in that some of the peaks comprising Amanda’s profile are moderate in intensity, despite possibly being the result of secondary DNA transfer (Amanda does not smoke cigarettes).

Do any of the samples contain DNA from a third party?
Sara Gino’s testimony indicated that in sample 177 in Filomena’s room there were alleles besides those of Meredith and Amanda. I have also examined a copy of the electropherogram. In the D19S433 locus, four alleles are marked: 12, 13, 16, and 16.2, but there are unlabeled alleles at 14 and 15 or 15.2. If one acknowledges that it might have been deposited at some other time than the murder, then one must also acknowledge the same possibility for Amanda's DNA.

Could the forensic team have run controls?
Besides the issue of how samples were collected in general, the forensic police could have done substrate controls, where they examined areas for DNA that were a few inches away from putative blood stains, as materials scientist Dr. Mark Waterbury suggested. If they had found Amanda’s DNA in some of those locations, it would have been suggestive of innocent DNA deposition.

They could also have performed or cited studies of DNA deposition in bathrooms, if such studies existed already (I am not aware of any). There are a number of ways that such studies could be performed. For instance, one could take blood from person A and place it in person B’s bathroom, then collect DNA samples. If one found mixed DNA from A and B, it would strengthen the hypothesis that Amanda deposited DNA in the normal course of everyday living.

Is there precedent for mixed DNA arising through contamination?
Many cases of contamination show a mixture of DNA from the analyst and a potential suspect, as discussed in the previous blog entry. One case, the murder of Jane Mixer, showed contamination from two potential suspects, Gary Leiterman and John Ruelas. However, Ruelas was four years old at the time of the murder and lived in a different city. Therefore, this is probably a case in which both profiles arose from contamination.

Could the way the DNA was sampled have resulted in mixed DNA?
The Massei Motivations report (p. 278, English translation) indicates that the defense thought that the mixed traces were meaningless: “All the more so since the samples had been taken using the same blotting paper which had been used for various parts of the bidet and the sink.” Even Colonel Garofano (a strongly pro-prosecution commentator on the case) was dismayed at the way the washbasin trace was collected, noting, “The fact that the sample was collected by wiping both the edge and the plughole is dangerous. You’re likely to find all sorts of stuff in the plughole.” (p. 370, “Darkness Descending”)

Did the police take every precaution to avoid contamination?
No, there are several ways in which the work could have been improved. Ms. Stefanoni’s view was that liquid samples are liable to cross-contamination, but dry traces are not. In the English translation of the Massei report (p. 203) it says that Stefanoni “specified” that gloves were changed “every time an object was touched that was particularly soaked with blood, and when it was obvious that the gloves would be soiled;” On pages 204-205 she indicated that the presence of a liquid is necessary to bring about contamination by touch.

Ms. Stefanoni’s view is out of the mainstream. On page 38 of John Butler's textbook “Forensic DNA Typing,” he wrote, “Use clean latex gloves for collecting each item of evidence. Gloves should be changed between handling of different items of evidence.” At Forensic Magazine in the article “Evidence Handling and Collection” Dick Warrington wrote, “Go about collecting evidence. I can’t say enough about avoiding cross contamination. Put on gloves, use gloves, change gloves. Do that every time you touch a piece of evidence. Likewise, use disposable tweezers, scalpels, etc. Change these each time they are used, as well.” Warrington also wrote an article for Forensic Magazine called “DNA Collection and Packaging,” that discussed the use of gloves and tweezers to avoid contamination. Orchid Cellmark’s guidelines state, “Use clean latex gloves for collecting each item of evidence. It is recommended the gloves be
changed between the collection of each item of evidence.” If the police handled an item of evidence with Amanda’s DNA then handled an item with Meredith’s DNA, the glove could carry Amanda’s DNA into the other sample.

Are there innocent explanations for the mixed DNA in Filomena’s room?
The luminol work that first identified some of the areas that later were shown to contain mixed DNA traces was performed on December 18, 2007. By this time many police personnel had been in the girls’ flat, and many of Meredith’s items had been tossed about. This raises the possibility that the forensic police tracked the genetic material of either Knox or Kercher into Filomena’s room from the hall. The forensic police who were recorded on 18 December wore one-piece tyvek garments but did not appear to have any outer shoe covering. Former FBI agent Steve Moore noted that they did not change shoe covers going from one room to another that that this creates the potential for cross-contamination. This is especially worrisome in that several members of the team are quite close to the dried bloodstains in Meredith’s room, as can be seen in parts 10 and 11 of a series of videos taken on 18 December 2007.

In addition the luminol-positive spots are only presumptive blood; these tested negative by tetramethylbenzidine, a second type of presumptive test, and there is no record of confirmatory blood testing. Therefore, it is open to debate whether or not the luminol-positive substance is even blood. One photograph of the luminol-positive footprints in the hallway also show blue specks on the ruler and on the boot of one of the forensic police officers. It is unclear what the luminol-positive substance was in this case, or whether it could have contaminated other items of evidence.

Conclusions
Mixed DNA is commonly observed and is not equivalent to mixed blood. In general DNA samples cannot be dated, and any two profiles within a sample may have been deposited at different times. The mixed DNA in the bathroom may have been created by Meredith's blood falling on Amanda's biological matter that was already there. The chances of this happening might have been lessened if the forensic police had taken a smaller trace with respect to the washbasin, for example. Dirty gloves or dropped swabs (which happened elsewhere) made have mixed DNA during collection. The police or the inhabitants of the flat may have tracked Meredith's blood into Filomena's room. The evidentiary value of these mixed DNA samples is very low.

Sunday, September 4, 2011

The likelihood of DNA contamination

Part 31 in the Knox/Sollecito case

The subject of DNA contamination has been a frequent topic of this blog. DNA contamination is again near the forefront of the Kercher murder and the second trial of Amanda Knox and Sollecito. The court-appointed, independent experts, Drs. Conti and Vecchiotti, have issued a report that raises the strong possibility of contamination. The odds that a forensic DNA sample are contaminated are hard to determine yet very important. Let us first examine a specific statement made with respect to environmental contamination and this case, then look at the question more generally. A subsequent entry will examine the collection practices of the Rome lab and Conti’s and Vecchiotti’s evaluation of their work in this case.

Professor Giuseppe Novelli, a researcher into medical genetics and forensic DNA profiling, said, «Il contaminante va dimostrato, dove nasce e dove è. Il gancetto contaminato dalla polvere? Più probabile che cada un meteorite e butti giù questo tribunale» "The contaminant needs to be demonstrated, where it comes from and where it is. The clasp contaminated by dust? It's more likely that a meteorite comes down and knocks down this courthouse." (translation by komponisto)

Let us assume that Dr. Novelli is specifically referring to household dust. The potential for dust being an issue in forensic DNA testing first came to light with the publication of a paper from Bonnie Brown and coworkers: Toothman et al., “Characterization of human DNA in environmental samples,” Forensic Science International 178 (2008) 7–15. These workers sampled dust from offices, research laboratories, and classrooms.

Figure 3 in this paper is an electropherogram of dust from a classroom, and it shows peaks of 1000 to 3500 RFU corresponding to 3-6 alleles in some loci associated with shorter pieces of DNA. As the length of the DNA fragments increases (moving from left to right on the elecropherogram), the height of the peaks decreased. The authors point out that this is consistent with DNA template that is partially degraded. The combination of a sample’s being a mixture and its yielding only partial profiles makes it very difficult to identify individual profiles.

The authors wrote, “Results of this study have implications regarding the processing of forensic samples. First, the presence of genotypeable human DNA in dust illustrates a significant
potential contamination source in forensic investigations. Twenty-five of 36 samples contained sufficient input human DNA for STR analysis using the AmpFlSTR® Profiler PlusTM assay (~1.0 ng), and 36% (including low-input samples) produced alleles at multiple loci. These results demonstrate that even though anti-contamination measures may be in place at a crime scene and the laboratory, trace DNA derived from dust in the vicinity of other evidence is capable of producing signals higher than background noise in STR analyses.”

Trace DNA and the environment
Secondary transfer is the movement of DNA from a donor to an intermediate object then to another object from which it is collected. Some forensic scientists only classify it as contamination when it occurs subsequently to the object’s being taken into custody; however, other workers prefer to treat innocent secondary transfer as equivalent to contamination on the basis that neither is relevant to the investigation. Both secondary transfer prior to an object’s being collected and contamination need to be understood more thoroughly for trace DNA to be used routinely in DNA forensics. Van Oorschot and colleagues wrote, “Greater effort needs to be made by police/crime investigators to investigate how a DNA sample arrived at the location where it was found, as well as by scientists to better understand the impact of activities on the relative amounts of DNA from particular sources at a crime scene… Some preliminary contributions to our knowledge of transfer in relation to residential burglary and street robbery have recently been made [67].

In their review of trace DNA in forensics Van Oorschot and colleagues suggested six remedies to address the problem of contamination:
“1. perform more studies similar to those of Raymond et al. [67], Cook and Dixon [202], Dowlman et al. [203] and Toothman et al. [201] in order to learn more about the occurrence and persistence of DNA on particular surfaces in different environmental conditions” Understanding how prevalent and persistent background DNA is in the environment is far from being completely understood. Dr. Novelli declined to comment for this report, but his comparison with a meteorite is premature, at best.

The relationship between DNA sample size and the ease of transfer
In their 2010 review article Van Oorschot and colleagues wrote, “Contaminant DNA may appear as either the major or minor sample within a mixture or, alternatively, may overwhelm the target DNA completely.” One explanation for this somewhat counterintuitive statement is that contamination in the laboratory may introduce DNA from previous PCR amplifications by a number of possible routes. However, the two main DNA profiles of interest (Meredith’s on the knife blade and Raffaele’s putative profile on the bra clasp) both involve relatively small amounts of DNA; therefore, we will focus on small samples.

The discussion of low copy number (LCN) testing from the Crown Prosecution Service noted, “This increased sensitivity means ultra-clean laboratories are needed for the testing to minimise contamination of the sample by DNA from any other source.” The New Zealand Institute of Environmental Science and Research has spent $1 million building anticontamination areas for low copy number (LCN) DNA forensics. The New Zealand Herald wrote, “The bogey is contamination. The very sensitivity of the technique which enables it to extract a DNA profile from the tiniest sample also makes it extremely vulnerable to contamination. Stringent measures are needed to minimise that risk… We live in a ‘soup’ of DNA, explains ESR forensic programme manager Keith Bedford. ‘If I were to shed dandruff, massive amounts of dna could fall ... hair could carry DNA. The way I am speaking at the moment, we could probably detect DNA on this pad in front of me.’”

Sara Gino testified for the defense in the trial of the first instance, and some of what she had to say is pertinent to this issue. From the Massei report (p. 258, English translation): “She reaffirmed that [the risk of] contamination exists, and emphasised that in minimal quantities of DNA there is not necessarily a greater risk of contamination but it was easier to notice the effects of the contamination and be misled (‘...It's not that the risk of contamination is greater; but it is easier to see the contamination...’ page 92).” In response to a question on this subject, Professor Dan Krane responded, “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.”

Some examples of DNA contamination
Farah Jama was a young man accused of rape on the basis of his DNA seemingly being found on the alleged victim. Mr. Jama is black and at 21 was too young to have entered the club at which the incident occurred, which catered to people over 28. Moreover, the alleged victim did not recall seeing a black man at the club that night. Yet, as Milanda Rout wrote, “But the judge and the jury did not buy his alibi, despite supporting evidence from his father, brother and friend. Instead, they believed the forensic scientist who testified there was a one in 800 billion chance that the DNA belonged to someone other than the accused man.” After Mr. Jama spent more that eighteen months in prison, he was released because prosecutors said that they could not rule out contamination. The contamination event may have occurred during two forensic medical examinations, one of the victim and the other of Mr. Jama on an unrelated matter that occurred one day earlier.

Perhaps the most thoroughly studied case of contamination is that which occurred in the Jaidyn Leskie case. This blog has covered the Leskie case on two previous occasions. The DNA of a woman who probably never left her village was found on the clothing of the submerged body of a toddler. She was a mentally challenged woman who may have been raped, which is why her DNA was being examined. Because the woman was such an exceedingly unlikely suspect, the only reasonable explanation was contamination. Contamination has been documented on several occasions when evidence items from unrelated cases are examined within a few days in the same lab.

Russell John Gesah was charged with the rape and murder of a mother and child. Kathleen Skeen wrote, “A Victorian Police Forensic Services Centre review found clothing with Mr Gesah's DNA from an unrelated offence had been examined on the same day and same surface as clothing from the Tapp case." The Gesah case, and the murders of Jane Mixer and Jane Durrua (see below) are all examples of DNA cold hits. The Gesah case prompted the State of Victoria to reexamine thousands of cases (see below).

Gregory Turner might have been convicted of murder on the basis of DNA evidence. However, a forensic worker contaminated a key piece of evidence with his and her DNA. She also acknowledged contamination in two other cases on which she had worked. An interesting aspect of the Turner case is that the DNA from the victim came from her fingernails, and Mr. Turner’s DNA came from his wedding ring. These facts suggest that the presence of liquids is not necessary to bring about cross-contamination, in contrast to the implications of Patrizia Stefanoni’s testimony in the present case.

The murder of Jane Mixer was initially attributed to a serial killer. When Gary Leiterman’s DNA was found on the decades-old evidence, he was convicted. However, the presence of the DNA matching then four-year old John Ruelas on the same item of evidence (despite Ruelas’s living in another city) strongly points to this being another example of contamination. This illustrates another important principle. One does not always know the precise moment that contamination occurred, but one can infer contamination when the direct deposit of DNA is shown to be highly unlikely.

A seemingly solved cold case that turned out to be contamination involved the 1968 murder of Jane Durrua. Jerry Lee Bellamy’s DNA was found when the evidence was tested in 1999. Evidence against Mr. Bellamy in an unrelated case was tested on the same day as items from the Durrua case. The actual evidence of contamination was not conclusive, but despite this, charges against Mr. Bellamy were dropped. Alleged serial killer Robert Zarinski was later arrested, but he died before he could be tried.

The difficulties in quantifying the frequency of DNA contamination
Not all labs document contamination events. Some labs argue that contamination that is identified with the use of negative control experiments does not count as contamination. Negative controls will spot wholesale contamination events but will not necessarily catch sporadic contamination. These facts make it difficult to quantify how frequently contamination occurs. However, it does not seem to be an especially rare event. Professor Thompson is a lawyer who specializes in probability theory as it relates to DNA profiling. Maura Dolan reported that he is among the leading authorities on laboratory errors in the United States. In response to a request from the Los Angeles Times to review the records from some California forensics labs, Thompson said, “’on a regular basis, laboratory personnel make mistakes that could lead to false identifications’ of suspects.” He also indicated that what has emerged in recent years is just “the tip of the iceberg.”

In 2008 Professor Thompson wrote an article, “The Potential for Error in Forensic DNA Testing (and How That Complicates the Use of DNA Databases for Criminal Identification"). “Doubt was also cast on a number of convictions in Queensland when a forensic scientist who had previously worked for a state forensic laboratory publicly expressed concerns about the reliability of the lab’s work. He told The Australian newspaper that it was not uncommon for the lab to mix up DNA samples from different cases.[62] For example, he said that analysts’ own DNA, from blood samples used as analytical controls, often was mixed up with (or found its way into) casework samples, creating false matches: “[Q]uite often my (colleague) would walk down the aisle and say, ‘I’ve just committed another rape on the Gold Coast.’”[62] The analyst said that while many such errors were caught, sample limitations made it impossible to resample or retest in some questionable cases.” These remarks underscore the notion that DNA contaminations are not a rare event.
[62. A. McDonald, “DNA evidence claim clouds Australian convictions,” The Australian, July 8, 2006.]

In response to the Russell John Gesah contamination incident (see above), the Victorian police reexamined their cases involving DNA forensics. During the period from 1988 to 2008 the Victorian police service handled 7000 cases involving DNA. According to Peter Gregory and coauthors, “In 2003, Mr Scheffer told an inquest on Moe toddler Jaidyn Leskie that since late 1999, 39 cases had been identified as requiring "diagnostic and corrective action", with most involving contamination.

Finally, testimony reported by Annabelle McDonald (in The Australian) implied that mixing up samples is a not uncommon event. Although mislabeling of samples (if that is what mixing up means) is not itself contamination, it has the potential to lead to the same erroneous judicial result. A mislabeling in Nevada was uncovered during an independent review of the Lazaro Sotolusson case. In addition Dwayne Jackson was also the victim of a similar mistake at the Las Vegas forensics lab.

Conclusions
There is not yet enough information on environmental contamination to make conclusive statements about how common environmental contamination is; however, DNA is deposited routinely in all sorts of ways that are unrelated to criminal activity. The authors of a recent study believe that environmental dust can give rise to extra alleles in evidence samples. The frequency of contamination is difficult to quantify, but it is not an especially rare occurrence. The chances of contamination are greater for DNA in the low template range than they are for larger samples. Historical examples of contamination suggest that it is more likely to occur when items of evidence are processed closely in time. Contrary to the implication of Dr. Novelli's remarks, it is rarely the case that the exact mechanism of contamination is proven.