Reliability of genetic testing

10.29 Some genetic tests are not entirely reliable for a number of reasons, both technical and non-technical. The stigmatisation and discrimination that may flow from genetic testing are central concerns for this Inquiry, and these concerns become acute when a test is unreliable. Issues related to the reliability of genetic information, including the interpretation of test results, are considered in Chapters 3, 11 and 23. Scientific reliability of genetic testing—the scientific or technical reliability of the test—is discussed below.

Scientific reliability

10.30 The scientific reliability of a genetic test may be affected by a number of factors including sample contamination, incorrect laboratory testing procedures, mislabelling, and transcription errors. Although there has been considerable attention paid in recent years to developing policies in relation to the ethical and lawful use of genetic information, there has been less discussion about the impact of erroneous information.

10.31 Every laboratory testing procedure, no matter how well-established, involves the possibility of error. This is equally true of genetic testing. For example, the PCR method of DNA amplification allows minute quantities of DNA to be replicated in a way that facilitates testing. Yet there is a danger that the sample may be contaminated with extraneous genetic material, such as from previously amplified products or from the operator, thus generating copies of irrelevant DNA. There are also occasional errors with the sequence fidelity of amplified products, resulting in reading errors.[20]

10.32 The scientific reliability of a genetic test is measured by the ‘sensitivity’ and ‘specificity’ of the test. These are technical terms but, in essence, refer respectively to the statistical likelihood that a ‘true-positive’ will return a positive test result and that a ‘true-negative’ will test negative. Clinicians and patients desire 100% accuracy. However, few laboratory tests are currently more than 98% sensitive and specific.[21]

10.33 Moreover, every test result requires individual interpretation, with a further opportunity for error to be introduced. Because genetic tests are considered to be ‘scientific’, many non-experts may invest excessive confidence in their significance and predictive value.[22] As a result, a small number of people who take genetic tests will receive inaccurate information about their genetic status. Whether this involves the trauma of a false positive or the spurious re-assurance of a false negative, either type of error is likely to have crucial implications for the individual, who might plan his or her life on the basis of the test result. In addition, because members of a family share genes and DNA, any error in a genetic test may have long term implications both for the person tested and his or her family. Predictive tests are particularly vulnerable to this difficulty because of the likelihood of a long delay before the error is recognised.


10.34 Another problem with the reliability of genetic testing procedures is the possibility of fraud. The Inquiry has received a number of submissions suggesting that laboratory protocols do not sufficiently protect against intentional interference with laboratory samples tested.[23] For example, a lawyer described a client’s allegation that an opposing party in family law proceedings had bribed a laboratory or otherwise falsified parentage test results.[24]

10.35 As discussed in Chapter 11, the scientific reliability of genetic testing is regulated by accreditation standards, which are administered by NATA and other bodies. However, the accreditation of a laboratory in accordance with the best technical and scientific standards is no guarantee against intentional deception by its employees. Although specific examples of possible laboratory fraud have been brought to the Inquiry’s attention, the Inquiry has no evidence of the incidence of fraudulent testing in Australia. Nevertheless, the possibility of fraud identified in the above submission indicates that this matter is one of continuing concern.

[20] R Trent, Molecular Medicine: An Introductory Text (2nd ed, 1997) Churchill Livingstone, 20.

[21] R Linsk, Consultation, Sydney, 20 February 2001.

[22] In the forensic context, see Ch 44.

[23] N Turner, Submission G083, 14 January 2002; Confidential Submission G074ACON, 10 January 2002.

[24] N Turner, Submission G099, 22 February 2002. See further Ch 35.