Introduction
DNA analysis was undoubtedly the most important development in forensic science in the twentieth century. Its ability to analyze small and environmentally challenged samples and to accurately establish their origins with a high degree of certainty was exploited to advantage by forensic scientists in the latter years of the century. However, forensic DNA analysis has another property that may eventually be even more important. Just as with fingerprints, the results of forensic DNA analysis can be stored in a databank. This can lead to many unique and important developments. The findings from most disciplines of forensic science come into play late in the investigative process where a suspect has been determined through other means and the police are looking for evidence that would either confirm or deny the person’s association with a crime and that could assist in any prosecution. DNA databanks are important because they assist at the beginning of the investigative process, enabling police to determine or eliminate suspects in the absence of other evidence. In addition, DNA databanks play another unique role. Even before the crime occurs, they can make a contribution which is of infinite value in economic and social terms, namely crime prevention. This can arise in many ways. First, it has been shown that violent offenders often commit several crimes (six or more according to one survey cited by the FBI) before they are finally apprehended. By making it possible to identify the perpetrator after their first or second offense, DNA databanks spare many potential victims. Secondly, by increasing the certainty with which dangerous violent offenders can be apprehended and successfully prosecuted, DNA databanks help to keep such people ‘off the streets’. Finally, the knowledge that they are very likely to be apprehended if they re-offend acts as a deterrent to some released convicts whose DNA information is in a databank.
The benefits of DNA databanks must be weighed against potential costs. Benefits are seldom achieved without causing some harm, and DNA databanks are no exception. In addition to the financial costs of their establishment and maintenance, DNA databanks also engender significant privacy concerns.
Accordingly, protection of privacy should be one of the pillars in the design of DNA databanks and the legislative framework surrounding them. An equally important design and legislative consideration is to maximize their efficiency and effectiveness as an investigative tool. Databank design and privacy concerns will be discussed, as will be the interrelationship of fingerprint and DNA databanks, benefits to police and society, and reaction of offenders. We will begin with a look at the history of DNA databanks.
History of DNA Databanks
On 1 August 1986, Dawn Ashworth was murdered in Leicestershire UK. The police involved in the search for the person who murdered her (and another victim from 1983) put a newly discovered tool, forensic DNA analysis, to work in their investigations. They created an informal DNA database by collecting blood samples from all the males in the town. As events unfolded, the database included all the males in town but one; Colin Pitchfork, the person eventually charged with the murder, had someone else pose for him in giving a blood sample. When that person later spoke about this in a pub, Colin Pitchfork was found out and the world’s first DNA database led, albeit indirectly, to his arrest.
From the time they first started conducting DNA testing in the mid 1980s, the London Metropolitan Police Laboratory started banking DNA information on an ad hoc basis. By the early 1990s this databank had become a key weapon in the detection of sexual crimes. However, in 1992 this databank was ordered to be destroyed when a legal action was brought before the European Commission on Human Rights under the privacy provisions of the European Convention.
The US Armed Forces Institute of Pathology established the first formal DNA databank in 1992. This databank, which was created to identify service men missing in action, proved successful in identifying casualties from the 1991 Desert Storm operation. By sharing their experience and through sponsorship of annual conferences on DNA databank and repositories beginning in 1992, the Armed Forces Institute of Pathology made a major contribution to development of forensic DNA databanks.
In the late 1980s, several states in the USA became the first jurisdictions in the world to pass DNA databank legislation. Other US states soon followed suit and the FBI developed the Combined DNA Index System (CODIS) to enable state databanks to communicate and share data with each other. Before long the first cold hit in a DNA databank occurred in Minnesota in 1991. A rape murder was solved by a match between DNA from blood and semen found at the crime scene and a databank sample from a person imprisoned for a burglary. The police went on record as saying that the case may never have been solved without the DNA databank. The wisdom of developing CODIS was shown in 1994 when the first cross-jurisdiction case was solved. By early 1999, all 50 American states had DNA databank legislation and a federal DNA databank was under development. Two factors initially limited the success of American DNA databanks. There was a lack of resources to analyze all samples collected from offenders and a long-established policy of not analyzing samples from cases without good suspects existed in many forensic laboratories. The DNA Identification Act of 1994 and other legislation provided funding for DNA databanks and forensic laboratories to overcome these obstacles as long as states comply with specified quality assurance standards, submit to external proficiency testing, and limit access to DNA information.
In 1994, legislation enabling establishment of a DNA databank was passed in the UK and in 1995 the Forensic Science Service opened the world’s largest and most successful DNA databank. The tremendous success of this databank spurred many other countries to pass DNA databank legislation. Numerous other countries around the world now have operational DNA databanks.
Design of DNA Databanks
In order to withstand court challenges and be viable in the long term, DNA databanks require enabling legislation. Such legislation will reflect the political, social, economic and constitutional environment of the jurisdiction in which it is enacted. Since no two jurisdictions are exactly alike for all factors, differences will naturally occur in the resulting legislation. This, in turn, will lead to differences in the design of the databanks. The following discussion highlights the various options available.
Indices
Two indices form the core of virtually all DNA databanks. The first of these, usually termed the crime scene index, contains DNA information obtained from bodily substances connected with crime scenes. The second, often known as the offenders index, contains DNA information from known samples of bodily substances taken from criminal offenders. DNA databanks in some jurisdictions contain additional indices such as missing persons DNA, or anonymous samples used to generate population statistics and establish evidential value.
Who is required to provide samples?
Depending on the jurisdiction, only a small subset of offenders may be included in the convicted offenders index (e.g. only those persons convicted of major sexual offences), there may be a long list of offenses for which persons convicted are required to give samples, or (as is the case in the UK), all persons charged with any criminal offence could potentially have their DNA information entered into a databank. In most jurisdictions, the process of conviction (or charge) is sufficient of itself to require an offender to provide a sample. Some jurisdictions, on the other hand, require that a special application be made to the presiding judge. Canada has two classes of offenses; primary offenses (most major crimes of violence) for which those convicted are compelled to provide samples, and secondary offenses (less serious offenses) where judicial discretion applies. Legislation in many jurisdictions includes provision for retroactive collection from persons convicted of offenses prior to enactment of the legislation and/or for retrospective collection from those persons who committed crimes before the legislation came into force but were convicted subsequently.
What samples are taken
Blood, being the most reliable bodily substance for forensic DNA analysis, is the most common sample collected. In many jurisdictions, liquid blood is used since tubes of blood are often collected routinely as part of the prison intake process. Other jurisdictions collect bloodstains with a finger prick, which is less intrusive and eliminates the requirement to have medical professionals involved in the collection process. This can save a considerable amount of money and is more convenient with respect to collection from convicted persons who receive a sentence that does not involve prison. Finger prick samples can now be collected on FTA®-treated paper. This greatly simplifies subsequent DNA extraction, and improves health and safety by immobilizing blood-borne pathogens. In some jurisdictions collection of blood is felt to be too intrusive. Two other sample types are then available – plucked hairs and buccal swabs. Some jurisdictions permit collection of two or three sample types and encourage collection of a second sample type in case the first does not provide satisfactory results due to improper collection or storage.
What is retained in the databank?
The large databank created by the US Armed Forces Institute of Pathology is unique in that only bodily substances (liquid blood samples) are stored. The bank does not store DNA information since analysis is conducted only when unidentified remains are found in relation to a military incident. All other forensic DNA databanks store results of DNA analysis of collected samples of bodily substances. In nearly all jurisdictions, the bodily substances themselves are also stored. This enables the databank to remain current in the event that a major change in analytical technology renders existing DNA information obsolete. However, the bodily substances themselves engender far greater privacy concerns than the limited DNA profiles used for forensic identification purposes. Accordingly, a few jurisdictions do not permit retention of samples after analysis; DNA databanks in such jurisdictions consist of DNA information only.
Analytical methodology
The first DNA databanks were based on restriction fragment length polymorphism (RFLP) analysis. Newer databanks use polymerase chain reaction (PCR) analysis of short tandem repeats (STRs). Many older databanks are in the process of conversion to the newer methodology as it is faster and more amenable to automation. Several databanks are increasing the use of robotics to automate parts of the analytical process. It is important that sufficient DNA loci be analyzed to ensure that the databank returns only one match and not several. A North American standard of 13 STR loci has been developed and is presently in use in many countries around the world. Britain and some other European countries use a somewhat different set of loci but there is sufficient overlap between the two standards to enable some sharing of data. In the future, increased standardization is anticipated in response to the increasingly international nature of modern crime.
Statistical evaluation
It is very important to assess the value of a match between the DNA profiles associated with a crime scene and those of a suspect developed through a databank search. This topic has been quite controversial, and there has been a gradual evolution of views as the meaning of a match and the appropriate questions to be answered became better understood. Initially, many people felt that since it is much more likely that a crime scene profile will match with someone in a large databank than it would with any one suspect developed through conventional means, a databank search would drastically reduce the evidential strength against an individual found to match. This was the view taken by the US National Research Council. Their 1992 report recommended that when a suspect is identified through a databank search, additional loci should be analyzed (in both the crime scene and suspect profiles) and that ‘only the statistical frequency associated with the additional loci should be presented at trial’. Their 1996 report suggests that evidence obtained from searching a data bank of size n would be n times weaker than the same evidence obtained by other means. There is a logical inconsistency in this view; a match to only one person from a crime scene profile searched against a databank of every person on earth would be extremely probative, whereas the above argument would suggest that it would have no probative value. It has been shown that this problem arises because the approach advocated by the National Research Council addresses the question: ‘how likely is it that someone in the databank would match if all were innocent?’ rather than the relevant question: ‘given that a databank search has been conducted, how strong is the evidence against the unique individual found to match?’. When this more appropriate question is asked, the answer shows that the DNA evidence is actually slightly stronger after a positive databank search than it is when a suspect is developed through other means, and accordingly, there is no need to use additional loci for court purposes. (It should be noted that whereas the DNA evidence is stronger after a databank search, the overall case against the suspect may very well not be; hence an understanding of the problems of incorporating the DNA evidence with the non-DNA evidence in such cases can be particularly important. In addition, police investigating cases in which a suspect is developed only through a DNA databank match need to guard against the biasing effect this information might have on other aspects of their investigation.)
Other design considerations
Some databanks use bar codes to identify samples, thereby providing an additional measure of privacy protection and minimizing the potential for sample mix-up. Some databanks require the convicted offender to provide a fingerprint along with the sample of bodily substance. This then enables fingerprint and criminal history files to be checked to verify the identity of the offender and ensure that an alias has not been used. Many databanks are required to purge samples and DNA information from certain offenders who have been granted a pardon. Use of bar codes or other means of rapidly locating and identifying samples is particularly important in such jurisdictions.
Critical success factors
Regardless of the design of a DNA databank, the following factors are critical to ongoing successful operation. The enabling legislation must be substantial and robust, and it should be continually amended and updated. For example, the trend in many jurisdictions is to start with a limited number of designated offenses and later, once the databank has established a good reputation and has the ability to expand, to amend the legislation by adding additional offenses. It is extremely important to ensure that all samples authorized by law are indeed obtained and that they are promptly and properly collected. Proper administration and sample tracking must be established with appropriate documentation. Quality control and quality assurance procedures must be put in place and should be reviewed by a recognized accrediting organization. Loci used in ongoing casework must correspond to loci used in the databank. Finally and most importantly, forensic laboratories must have the capacity to examine cases with unknown suspects.
Privacy Considerations
The DNA molecule holds the key to all that makes an individual unique. DNA information can reveal secrets of a person’s past (such as who his or her biological parents were), and can predict predisposition to genetic-based diseases. Were such information to fall into the hands of the wrong people, or to be used in the wrong way by otherwise well-meaning people, much harm could ensue. Privacy advocates fear that samples from DNA databanks will be used in research aimed at identifying a ‘criminal gene’. They are also concerned that the ‘law enforcement purposes’ for which forensic DNA databanks were intended will be gradually broadened to include use by immigration authorities, child support enforcement officials and other government agencies, and that this ‘surveillance creep’ will eventually lead to everyone being required to have their DNA information on file. For these reasons, those concerned about privacy have carefully scrutinized DNA databanks and their enabling legislation. Fortunately, legislators and people responsible for the operation of forensic DNA databanks have generally been cognizant of privacy concerns and have built in numerous safeguards. First and foremost, loci chosen for forensic DNA analysis and databanks are deliberately selected to be from noncoding regions of the DNA molecule with no known direct link to any genetic disease or trait. Secondly, many databanks use bar codes and other means to ensure that there is no direct link between a person’s DNA information and their personal identifiers, thereby making it difficult for any one person to illegitimately obtain the information. Third, authorizing legislation often prescribes fines or prison terms to those people whose action or inaction causes improper release of DNA information or to those who tamper with DNA samples. Fourth, some jurisdictions involve privacy advocates in the drafting of legislation and in the oversight of DNA databank operations.
Interrelationship of Fingerprint and DNA Databanks
DNA databanks serve as a complement to, rather than a replacement for, long-established fingerprint databanks. DNA evidence is most commonly associated with crimes of violence although DNA can also often be found in property crimes. Conversely, fingerprints are most commonly found in property crimes but can also often occur in crimes of violence. Instead of having just one identification tool, police investigators now have a pair. If fingerprints are not found at a particular crime scene, the chances are that DNA will be and vice versa. Both fingerprints and DNA databanks can be used to link crimes committed by serial offenders. If a particular crime scene has both DNA and fingerprints, longer linkages can be established, e.g. three crimes linked to a scene through fingerprints can be connected with four others linked through DNA.
Reaction of Offenders
When police obtain a tool as powerful as DNA databanks, criminals can be expected to offer strong reactions. Some of these are positive. Knowledge that their DNA information is in a databank is sufficient to cause some offenders to refrain from criminal activity. There have even been reported incidents of people who were convicted before the date when a DNA databank was established volunteering to give samples as a demonstration of their determination to change their ways. Having their DNA information in the databank also benefits released offenders who can be quickly eliminated as suspects whenever an unsolved crime occurs in their neighborhood, thereby reducing the amount of hassle they receive from police.
However, as might be expected, the reaction of the majority of offenders is negative. Some try to mount court challenges to the authorizing legislation. Since most legislation has been carefully drafted in a consultative manner, such challenges are almost invariably unsuccessful. Offenders must thus adapt to the new reality and find ways of circumventing it. The first attempt at such adaptation occurred when Colin Pitchfork sent someone else to give a blood sample for him. Anecdotal evidence of other adaptations has included an increased use of condoms in sexual assaults (which brings other benefits to society), and attempts to destroy evidence through burning bodies etc. Some criminals have even learned to use DNA databanks to their advantage by planting someone else’s DNA at crime scenes.
Benefits to Police and Society
DNA databanks can help police quickly solve crimes that would previously have taken a long time to solve or may have remained unsolved. They can link crimes within and across jurisdictions, thereby establishing that a serial offender is at work and increasing the chances of eventual solution. They can also save police much time and expense. When investigating unsolved crimes, police no longer need to conduct time-consuming interviews with previous offenders, who can be quickly and effortlessly eliminated by a quick search of the databank. On the other hand, if a match is obtained, this generally provides enough evidence for arrest, thereby eliminating the need to mount expensive surveillance or undercover operations.
Much as DNA databanks benefit police, it is society as a whole which reaps the greatest benefit -reduction in victimization.
