Limiting the Intrusiveness of Forensic DNA Sampling

By Mark A. Rothstein 

As of December 1st of this year, New York State will expand its computer database of DNA samples from convicted felons. By increasing the number of crimes for which a DNA sample is required, about 25,000 new samples will be added each year. Similar plans are being promoted in many other states as well as at the federal level, with pressure increasing to include in federal and state databases DNA information about a wide range of convicted criminals and even people who merely have been arrested for crimes.

It is undisputed that DNA identification is a powerful tool that can both help to convict the guilty and to exonerate the innocent. The difficult question, from both a legal and policy standpoint, is to what extent the legitimate ends of law enforcement justify the intrusive means of obtaining and storing individual genetic information. Thus far, the courts have upheld the requirement that incarcerated felons submit a DNA sample for inclusion in law enforcement databanks. The legal challenges, however, have involved laws requiring sample collection from individuals convicted of serious felonies, such as murder and violent sex crimes. Even if the courts would be willing to uphold wider DNA sampling, there are many policy questions to be resolved before there would be widespread public support for expanding DNA collection.

One of the first questions is whether people are comfortable with the routine submission of DNA samples for law enforcement. Is there a perceived or real difference between fingerprinting, commonly performed in various contexts (including the military and employment) and widely accepted, and DNA testing? If DNA testing is considered different, is it because the sample is obtained from blood, that "genetic" material is obtained, or that the sample could reveal a wide range of predictive health information? Another question is at what point short of obtaining DNA samples from everyone (which would be relatively easy to do at birth when newborn blood samples are routinely obtained for medical testing) is it appropriate to draw the line for DNA testing? An argument could be made that it is actually fairer to obtain DNA samples from everyone than to require samples from, for example, all individuals arrested for crimes (but not convicted) because a higher percentage of racial and ethnic minorities are arrested.

Even as we begin the important public policy debate on the limits of forensic DNA testing, it would be extremely valuable for legislators or forensic laboratories implementing DNA testing legislation to take the following three steps to limit the intrusiveness of forensic DNA sampling.

First, all DNA samples should be obtained by using cheek swabs. It is not necessary to use blood samples, and using cheek swabs is fast, easy, painless, and less invasive.

Second, all DNA identification should use exclusively noncoding regions of DNA. Only genes "code" or signal for the development of physical or mental traits. Located in between the genes on human chromosomes, there are vast spaces of genetic material, sometimes called "junk DNA." The sequences of this DNA have no known relationship to an individual's current or future health or physical or mental development. The minor variations among people in these sequences have no other scientific value than for identification. It is now common, although not necessarily standard practice, to use only noncoding regions of DNA in the identification process. The advantage of this approach is in ensuring genetic privacy. Even if an individual's identification sequence were inadvertently or unlawfully released to a third party, such as an employer or insurer, the information would have no value in determining the current or predicting the future health status of the individual. It would be comparable to a genetic identification number.

Third, all sample material should be destroyed upon analysis. The only information that should be entered into computerized DNA identification databases is a sequence of identifying numbers, along with the name of the individual from whom the sequence was obtained. Retaining original specimens leaves open the possibility that they could be reanalyzed for wider purposes, including testing for information contained in genes. Destroying the samples immediately after analysis eliminates this problem. If the DNA database is used to identify a potential criminal suspect, and there is any question about the accuracy of the original DNA identification, it would be relatively easy to obtain another sample from the individual and to perform another analysis.

Implementing these recommendations will not avoid the important legal and policy debate about balancing effective law enforcement with the civil liberties of citizens. Nevertheless, these three modest, easily adoptable measures, will ensure that whatever the parameters of the DNA identification system, it will be as nonintrusive as possible.

08/31/99