Nov P12-23 Nov Fea 10.17

Tarnish on the ‘Gold Standard:’ Understanding Recent Problems in Forensic DNA Testing • In Virginia, post-conviction DNA testing in the high-pro- file case of Earl Washington, Jr. (who was falsely convicted of capital murder and came within hours of execution) contradicted DNA tests on the same samples performed earlier by the State Division of Forensic Sciences. An outside investigation concluded that the state lab had botched the analysis of the case, failing to follow proper procedures and misinterpreting its own test results. The outside inves- tigators called for, and the governor ordered, a broader investigation of the lab to determine whether these problems are endemic. Problematic test procedures and mis- leading testimony have also come to light in two addition- al capital cases handled by the state lab. 2 • In 2004, an investigation by the Seattle Post-Intelligencer documented 23 DNA testing errors in serious criminal cases evidence has long been called “the gold standard” handled by the Washington State Patrol laboratory.3 DNAof forensic science. Most people believe it is virtually infallible — that it either produces the right result or no • In North Carolina, the Winston-Salem Journal recently result. But this belief is difficult to square with recent news published a series of articles documenting numerous DNA stories about errors in DNA testing. An testing errors by the North Carolina extraordinary number of problems related State Bureau of Investigation.4 to forensic DNA evidence have recently come to light. Consider, for example, the • The Illinois State Police recently can- following: celled a contract with Bode Technology Group, one of the largest • The Houston Police Department (HPD) independent DNA labs in the country, shut down the DNA and serology section expressing “outrage” over poor quality of its crime laboratory in early 2003 after a work.5 television expose revealed serious deficiencies in the lab’s procedures, deficiencies • LabCorp, another large independent that were confirmed by subsequent investi- lab has recently been accused of botch- gations. Two men who were falsely incrim- ing DNA paternity tests.6 inated by botched lab work have been released after subsequent DNA testing While these scandals are bad proved their innocence. In dozens of cases, enough, the problems with DNA evi- DNA retests by independent laboratories dence do not end there. A close look at have failed to confirm the conclusions of the field shows that DNA testing errors the HPD lab. The DNA lab remains closed have been popping up all over the coun- while an outside investigation continues.1 try. Many of the mistakes arise from By William C. Thompson 10 WWW.NACDL.ORG THE CHAMPION cross-contamination or mislabeling of While there have always been bad error in DNA testing is so low as to be DNA samples. Problems of this type have labs, their shoddy work has been diffi- negligible, but growing evidence sug- been documented in Minnesota,7 North cult to detect because the worst labs tend gests otherwise. Carolina,8 Pennsylvania,9 Nevada,10 and to be found in jurisdictions that have An important source of evidence on California.11 Tellingly, one of the private historically shielded crime labs from the nature and frequency of these prob- labs hired to retest DNA evidence in cases external scrutiny. For example, it is now lems is “contamination logs” and “cor- that were botched by the Houston Police recognized that the Houston Police rective action files” that are maintained Department Crime Lab has itself pro- Department (HPD) Crime Laboratory by some DNA laboratories. Under a duced false matches due to sample mix- did grossly inadequate, incompetent and guideline issued by the FBI’s DNA U 12 ups. A particularly ominous sign of biased DNA and serology work for well Advisory Board in 1998, forensic DNA N underlying problems is that accidental over a decade before a team of television laboratories are required to “follow pro- D transfers of DNA among samples from journalists exposed the problems in late cedures for corrective action whenever 18 E different cases being processed by the 2002. Defense lawyers did not (and proficiency testing discrepancies and/or R same laboratory have produced several probably could not) expose the lab’s casework errors are detected” and “shall S 13 false “cold hits.” problems because Harris County maintain documentation for the correc- T While most of the problems are due (Houston) judges routinely denied tive action.”19 While many laboratories A to inadvertent mistakes, a number of requests for discovery of underlying lab- ignored this guideline, some laboratories N cases involving dishonesty have also come oratory notes and for expert assistance (probably the better ones) have begun to D to light. DNA analysts have recently been in evaluating DNA evidence. Indeed, keep records of instances in which, for I fired for scientific misconduct, and under a policy of the Harris County example, samples are mixed up or DNA N specifically for falsification of test results, District Attorney’s Office, that defense from one sample is accidentally trans- G by a number of forensic laboratories, lawyers rarely challenged, the defendant ferred to another samples. R including labs operated by the FBI,14 could not even get a copy of laboratory The surprise for defense lawyers E Orchid-Cellmark (another large private reports in his case until the trial began. who have managed to gain access to these C 15 DNA laboratory), the Office of the Crime labs in Virginia and North files is how voluminous they are. Errors E Chief Medical Examiner in New York Carolina have also received little scrutiny occur regularly. Files from Orchid- N 16 17 City, and the United States Army. In all in the justice system due to severe limi- Cellmark’s Germantown, Maryland T of these cases, the analysts were caught tation in those states on the availability facility, for example, show dozens of faking the results of control samples of discovery, funding for independent instances in which samples were contam- P designed to detect instances in which experts, and funding for indigent inated with foreign DNA or DNA was R cross-contamination of DNA samples has defense in general. somehow transferred from one sample O occurred. But these problems could not to another during testing. I recently B L So what is going on with DNA test- remain hidden forever. Journalists have reviewed the corrective action file for an E ing? How can we explain this sudden rash played a big role. In Houston, the prob- accredited California laboratory operat- M of problems with “the gold standard” of lems were first exposed in a series of ed by the District Attorney’s Office of S forensic science? How can a test that has exposes by television reporters who were Kern County (Bakersfield). Although long been advertised as virtually infallible assisted by academic experts. Excellent this is a relatively small laboratory that I produce so many errors? And what is investigative journalism also helped processes a low volume of samples N behind the recent spate of dishonesty expose problems in Washington State, (probably fewer than 1,000 per year), F among DNA analysts? The answers to Virginia and North Carolina. Another during an 18-month period, it docu- O these questions are, in my view, intercon- important factor has been post-convic- mented multiple instances in which R nected. Some serious underlying prob- tion DNA testing. A number of errors (blank) control samples were positive for E lems with DNA testing that have existed have come to light because post-convic- DNA, an instance in which a mother’s N for a long time are beginning to come to tion DNA tests contradicted tests per- reference sample was contaminated with S light. What we are seeing is not a sudden formed by government crime labs. The DNA from her child, several instances in I deterioration in the quality of DNA test- work of Peter Neufeld, Barry Scheck, which samples were accidentally C ing. It is the inevitable emergence and and their colleagues at the Cardozo Law switched or mislabeled, an instance in D recognition of problems that existed all School Innocence Project has been which an analyst’s DNA contaminated N along but heretofore were successfully instrumental in exposing problems, par- samples, an instance in which DNA A hidden. In this article, I will describe these ticularly in Texas and Virginia. extracted from two different samples was underlying problems, comment on why accidentally combined into the same T they are occurring, and discuss what Surprising Frequency tube, falsely creating a mixed sample, and E defense lawyers can do about them. Of Cross-Contamination And an instance in which a suspect tested S Sample Mix-Ups twice did not match himself (probably T I Bad Labs Another problem now emerging due to another sample-labeling error). N One chronic problem that is now into the light is an unexpectedly high The errors documented in these files G being recognized is the uneven quality of rate of laboratory errors involving mix- are disturbing, in part, because they forensic DNA laboratories. Laboratories up and cross-contamination of DNA probably represent just the tip of an omi- vary greatly in the care with which they samples. Errors of this type appear to be nous iceberg. The documented errors validate their methods and the rigor chronic and occur even at the best DNA are, of course, those that the laboratory with which they carry them out. Quality labs. This is a problem that forensic sci- itself caught and corrected. In most control and quality assurance proce- entists have largely managed to keep instances, these errors produced unex- dures that are followed religiously in under wraps (perhaps because it is pected results that flagged the problem, some labs are ignored or followed inter- always embarrassing).

Load more