Until last year, no one had any reason to doubt Dr. Elizabeth Johnson's work. A molecular biologist who had done postdoctoral work in hematology at M.D. Anderson Cancer Center, Johnson was hired in 1991 to set up a DNA testing laboratory for the Harris County Medical Examiner's Office. One testing technique that Johnson helped perfect has attracted notice in other parts of the country and was instrumental in a 1995 murder conviction in northern California.
But when her findings in a brutal double killing in Pasadena didn't conform to investigators' assumption about who committed the crime, Johnson suddenly found her expertise being called into question. In fact, at least two law agencies -- the Sheriff's Department and the Pasadena Police Department -- stopped utilizing Johnson's services early this year after the District Attorney's Office launched a review of her work procedures. While awaiting the results of that review, both departments have been contracting with more expensive private labs for DNA testing.
Johnson's troubles began in April of last year with the killings of sisters Martha Durrett and Linda Harrison, who were bludgeoned to death inside the Pasadena home they shared. Joe Durrett, Martha Durrett's estranged husband, quickly became the prime suspect for Pasadena homicide detectives, who realized early on that they would need extensive forensic evidence to pin the murders on him. One week after the killings, the investigators received the evidence they had been hoping for, and shortly thereafter, Joe Durrett was charged with two counts of capital murder. He was arrested at Ben Taub Hospital, where he was recuperating from wounds he had sustained two days after the murders in a mysterious shooting outside his Houston home.
The charges against Durrett were based on the results of tests by the ME's Office on strands of hair found in a bloody clump in one of Martha Durrett's hands. The medical examiner's forensic hair analyst determined that some of that hair matched samples of Durrett's hair. By combining those test results with reports that Durrett had been harassing the women before the murders, detectives were confident they had built a strong case.
But after Elizabeth Johnson took a closer look at that forensic evidence, the case against Joe Durrett began to unravel.
When evidence from the murders of Martha Durrett and Linda Harrison was brought to the ME's crime lab, Johnson was in California testifying in an evidentiary hearing on behalf of the Santa Cruz County district attorney. The California judiciary traditionally has been reluctant to accept new scientific techniques in criminal investigations, and Johnson was in court to explain a DNA protocol she and a colleague had recently developed.
A polymerase chain reaction -- or PCR -- is one of two types of DNA analyses. PCR, which is used when there is not much DNA available for testing, is not new, but Johnson claims to have developed a technique for taking the process a step further.
"It's a pretty simple idea," she says. "It takes a pre-existing system. You modify that to add in a set of DNA primers that will target an additional DNA region. It improves the sensitivity. So instead of getting just one bit of information, you get two."
That second bit of information, Johnson explains, will reveal whether the donor of the DNA sample is a man or a woman. Additionally, she says, with her protocol, test results can be obtained from much smaller samples of DNA.
Both of the added benefits from Johnson's protocol played a significant role in the Santa Cruz County Sheriff's Department's investigation of a 1993 murder of a woman who had been dismembered and whose body parts were stuffed into garbage bags and tossed into Monterey Bay. A co-worker of a man arrested in the killing told investigators that he had witnessed the suspect throwing a hacksaw from a truck window on the day of the murder. The hacksaw was found, and a California state crime lab found a small amount of blood on the blade -- but not enough to determine if the blood was that of the victim.
Assistant district attorney Mike Bartrum had heard of Johnson's gender testing technique and forwarded blood samples from his investigation to Harris County's DNA lab. "She tested our sample and it tested positive for female blood," says Bartrum. "That was introduced in our trial and was very significant in its outcome."
The suspect was convicted of the woman's murder. Bartrum says it's hard for him to imagine anyone -- especially prosecutors -- questioning Johnson's methods.
"Her credentials are flawless," he says.
After returning from California, Johnson began conducting DNA tests on the blood samples from the double murder in Pasadena. What she discovered over the next few months surprised her.
"I started finding that the preconceived notions of whose blood was where were not factual," Johnson says.
For example, Johnson says her testing revealed that the blood found in Joe Durrett's home and on his clothes shortly after the double murder was his, not the victims'.
Johnson then began examining the hair samples. Contrary to the comparison done by Amy Haralson, the lab's hair analyst, Johnson's testing indicated that the hair found in Martha Durrett's hand was not her husband's. Johnson believes the discrepancy was attributable to a desire by Haralson to provide the prosecution with the evidence it wanted.
"About 20 hairs were recovered from the victim's hand," says Johnson. "Only a few were ever examined. I don't think she ever compared them to the victims or revealed the existence of the other hairs to the detectives or the District Attorney's Office."
In August, with her tests concluded, Johnson began the job of informing the investigators that their case against Joe Durrett was flimsy, at least from a forensic standpoint. The two Pasadena homicide detectives assigned to the case "looked like they had just had a dagger stuck through their hearts," Johnson recalls. A few days later, she spoke with Craig Goodhart, the assistant district attorney who was handling the case. He did not take the news well.
"He said, 'You mean I arrested this man, took him before a grand jury which indicted him and put him in jail because somebody over there told me the hairs matched, and now you're telling me they don't match?' And I said, 'Yes, that's basically what I'm telling you.' He was not a happy camper."
(Both Goodhart and Haralson declined to comment, citing the ongoing investigation of the Pasadena murders.)
Johnson didn't give much more thought to the Durrett case until last September. Shortly after Labor Day, she received a tip that the Pasadena detectives planned to visit her lab to retrieve the blood and hair samples so they could send the evidence to CellMark, a private forensic laboratory in Maryland, for re-examination.
To protect the integrity of the samples and prevent their contamination, Johnson and an assistant set about photographing and sealing the evidence.
"Not only is it unwise to let evidence go out unsealed," says Johnson, "I did not want there to be any possibility whatsoever of anyone tampering with that evidence before it got to CellMark."
The detectives did seize the evidence and sent it off to CellMark. A few months later, the police and prosecution got their second opinion: DNA experts at CellMark agreed with Johnson that the blood and hair samples from the crime scene did not link Joe Durrett to the murders. Prosecutors had no choice but to drop the charges against Durrett.
"Goodhart and the detectives were so sure that Joe Durrett was guilty," says Johnson. "And maybe he is. And maybe he isn't. But the evidence has to be there. And the evidence is not there."
After the results from CellMark seemed to vindicate her findings, a relieved Johnson assumed that was the last she would hear of the Durrett case. She was wrong.
In February, the District Attorney's Office subpoenaed Johnson before a county grand jury to turn over her files on five cases in which her lab findings resulted in the freeing of suspects. Among those five cases -- three rapes and two murders -- was the Pasadena double murder. Johnson accuses Goodhart -- whose name was on the subpoena for her records -- of being behind an attempt to publicly discredit her for not being a "team player."
"Craig Goodhart has spread rumors calling my reputation and my protocol into question," charges Johnson, who says that her modified protocol was used in only three of the five cases in question.
Johnson does not believe that District Attorney Johnny Holmes initiated the review of her work, and, according to the district attorney, she is correct on that score.
"I've got no dog in this fight," says Holmes, explaining that the investigation into Johnson's work was launched after complaints about Johnson from several of his prosecutors -- not just Goodhart.
"I want to find out if there is a problem. If there is, I want something done about it. If there's not, I want everybody to shut up."
Holmes says Johnson's work on the five cases is being reviewed by Dr. Ranajit Chakraborty, a professor in the biological sciences division of the University of Texas Medical School. Chakraborty refused to speak with the Press. He and Johnson, however, have crossed paths before, having offered expert testimony for opposing sides in the civil trial of a lawsuit filed over a sexual assault. One of the main points of contention in the suit -- which was settled out of court -- was Johnson's modified DNA protocol.
Johnson describes Chakraborty as "a brilliant mathematician," but says having him review her work is inappropriate, given the past differences between them.
Somewhat surprisingly, Holmes tends to agree that Chakraborty could possibly be biased against Johnson's protocol. If Chakraborty's report -- which is overdue -- is critical of Johnson, Holmes says he will have the cases reviewed a second time by another expert.
Holmes also frankly admits his life would be simpler if the reviewer concludes that Johnson's techniques are sound. If there is a problem, he says, "it will really open up a can of worms," possibly resulting in challenges to hundreds of convictions in criminal cases based at least partly on Johnson's lab work.
But Joe Durrett's lawyer believes there's a much simpler explanation for the review: an obsession by the District Attorney's Office with hanging the Pasadena murders on her client.
"I think they are doing this to go after Joe Durrett," says Katherine Scardino. "They had to throw in the other names just so it wouldn't look like that's what they're doing. They'd like to ruin [Johnson's] career. But they are desperately trying to find some reason to re-file [murder charges] on Joe Durrett."
Indeed, Pasadena police report that they have sent the blood and hair samples in the Durrett investigation back to the CellMark labs in Maryland for yet a third round of testing. One detective says those tests will also review the microscopic hair analysis first performed by Amy Haralson. Asked if his department had been mislead about how comprehensive Haralson's initial tests had been, Sergeant Kelly Payne replied, "Apparently."
In a possibly related matter, the FBI is conducting an investigation into allegations of sabotage and potential evidence tampering at the ME's Office. According to a source close to the ME's Office, it was discovered that reagents in the DNA lab had been mixed together on two occasions late last year. A reagent is a substance that is added to a solution of another substance to precipitate a chemical reaction. If a mixing of reagents went unnoticed, it is possible that DNA test results could have been compromised.
"This is not something that could have happened accidentally," says the source, who did not want to be identified. "It's clear somebody came in and messed with the reagents," which were kept under lock and key.
Johnson refused to comment on the investigation.
The allegation of possible tampering was forwarded to the district attorney, but Holmes passed it on to the FBI to avoid the appearance of a conflict of interest posed by his ongoing review of Johnson's work. He says a case could be made that suspects' civil rights were violated if criminal evidence were tampered with, thus justifying the federal jurisdiction.
Special agent Doug Shipley confirmed that the FBI is conducting the investigation, but would not elaborate.
Meanwhile, Elizabeth Johnson is under the impression that her days at the ME's Office are numbered, regardless of the outcome of the review of her work.
"The situation has gotten to the point where I don't want to be there too much longer," admits Johnson. "I just like to leave on my own terms and expose the twits for what they are.
Get the This Week's Top Stories Newsletter
Every week we collect the latest news, music and arts stories — along with film and food reviews and the best things to do this week — so that you’ll never miss Houston Press' biggest stories.