Page 3 of 6
Mudge's protestations might have sufficed, were it not for the confluence of two factors: the Madden scandal and the lab's need for reaccreditation. In August 2010, the lab was due to renew its accreditation for an additional five years. In light of the media feeding frenzy around Madden, the board figured it had better take another look at the crime lab's operations, and ASCLD inspectors paid a visit to San Francisco.
ASCLD, based in North Carolina, is not known for being an exacting watchdog. It is actually an umbrella group that encompasses a trade organization and a for-profit consulting group that helps member labs pass inspections, and its accreditation process is entirely voluntary.
The accreditation division is headed by Ralph Keaton, a former agent at the now-disgraced North Carolina State Bureau of Investigation, where a state attorney general's investigation recently uncovered 200 cases where lab analysts tainted or misrepresented forensic evidence. Amy Driver, a former Los Angeles Police Department forensic scientist who runs an industry blog, BulletPath, describes ASCLD as a "good ol' boy system."
Nevertheless, investigators reported troubling findings in their September report on San Francisco's crime lab, which was first released by the SFPD in response to inquiries from SF Weekly. Foremost among them were dual confirmations of the whistleblower's allegations that records of a sample switch had been destroyed and that lab security had been inadequate.
Despite Mudge's previous assurances that lab security was airtight, accreditation board investigators found that quality manager Marty Blake "readily admitted" that "the lab had problems with the biometric fingerprint access system working correctly" prior to November 2009, when a new card-access system was installed. "At times before the proximity card system was installed she observed that doors to the building and laboratory secure areas were propped open," the report noted.
And the DNA mixup? "Interviews with laboratory staff verified that the sample mixup did occur," the report stated. Investigators found that the last tube in a batch of testing samples, which should have been clear — what analysts call an "extraction negative" — was found to be colored. With Gabriel's approval, Nelson switched the labels on that tube and an adjacent clear tube, believing they were the only two that had been out of place. According to the report, "The DNA typing results were reviewed by the supervisor and the case showed no discrepancies in the DNA typing results, no further corrective actions were taken by the laboratory."
This in itself was an extremely sloppy way to fix the problem, according to Jim Norris, a Redwood City–based forensics consultant who headed the SFPD Forensic Services Division for nine years before retiring in 2004. A misplaced tube in a sample batch could potentially indicate that all of the DNA evidence during that round of testing is out of place and mislabeled, he said, and requires that testing be performed from scratch to ensure accurate results.
"You can say, 'I know how the mistake was made, and I know which two tubes are which,'" he says. "But are you really sure? The correct thing to do there is go back and start over."
An equally if not more serious breach of lab protocols stemmed from the lack of records. While some record of the switch may have remained in the lab's computer system, the inspectors noted that "no documentation of the incident" could be found in the case files, where they could be viewed by prosecutors or defense attorneys. Oddly enough, the report stated that this destruction of records was carried out "according to the DNA Unit standard practice," leading to a finding of "noncompliance with [ASCLD's] requirement to maintain examination documentation."
Keaton, in a telephone interview from ASCLD's office, says the agency was not pleased to learn that the sample switch took place after Mudge's indications to the contrary. "Obviously, that doesn't make us happy," he says. "If we felt that we had a consistent pattern of misrepresentation, we would find that extremely serious."
To date, SFPD officials say, they have not sought to determine which cases might have been affected by Nelson's error. According to Mudge's August 2009 letter to ASCLD, Nelson processed 33 cases in the last quarter of 2008.
Police say the switch was a minor accident that was quickly corrected. "It's like we're at a social party," said Capt. Donna Meixner, who was promoted in June and put in charge of the SFPD's Forensic Services Division. "You're Peter, I'm Donna. They mix up our name tags. But you're still Peter, and I'm still Donna."
Despite the fact that records of the mistake by Nelson were destroyed, an internal-affairs investigation into the incident showed that there was "absolutely no coverup in this case," according to SFPD spokeswoman Lt. Lyn Tomioka. She noted that ASCLD ultimately chose to reaccredit the crime lab in October, its findings notwithstanding.
Tomioka acknowledged that the initial denials of the lab mishap "looked bad" in light of investigators' later findings, but said the lab has made great strides since then in increasing the transparency of its operations. "We have moved so far forward with the lab that this would be like churning up old history," she said, noting that the SFPD had promptly provided ASCLD's most recent findings to SF Weekly when asked.
Nelson, approached by a reporter after testifying at a recent court hearing, declined to comment on the sample switch. "I'm not at liberty to discuss it with you," she said. Mudge, who has left his position at the crime lab but still works at the SFPD, could not be reached for comment.