The findings about the agency charged with overseeing the accreditation of all municipal DNA labs could throw thousands of criminal cases into doubt.

A state lawmaker complained that the Texas DPS "misled" the Legislature, and a criminal defense lawyer speaking for a statewide organization demanded a multimillion-dollar review of convictions.

The audits, which the Chronicle obtained through open records requests, were conducted last year at the DPS labs in Houston, Austin, El Paso, Garland, Lubbock, Corpus Christi, McAllen and Waco. They revealed procedural flaws, security lapses and shoddy documentation at several of the labs, which could result in faulty DNA profiles and possibly send innocent people to prison or allow the guilty to escape justice.

The findings include:

• DNA analysts who do not understand how to interpret test results.
• The failure of analysts to run blank samples to make sure instruments are not contaminated with DNA residue from previous tests.
• DNA reports that do not include important statistical probabilities.
• Possible cross-contamination of blood samples.
• A lack of lab security.
• Failure to document critical analytical procedures at the state lab that inputs DNA profiles into the FBI national DNA database.

"The only thing that can be done is to do the same thing that's being done in Houston," warned Stanley Schneider, chairman of the Texas Criminal Defense Lawyers Association's crime lab strike force. "We've got to get someone to go back and see if (DPS) made mistakes. And it's going to cost millions and millions of dollars."

In response to the Chronicle's findings, the state lawmaker who pushed through a 2003 bill mandating tougher lab oversight last week subpoenaed five years worth of DPS records and said he expects to begin a new round of hearings as early as mid-April.

"I have lost confidence in DPS and their ability to oversee these labs," said state Rep. Kevin Bailey, D-Houston and chairman of the House Committee on General Investigating. "Clearly we're going to have to hold hearings and ask them to come forward and give us their analysis of what's going on."

As part of last year's reforms, the Legislature entrusted DPS with overseeing the accreditation of all public DNA labs in the state by September 2005. That includes the municipal labs, like the one operated by the Houston Police Department, as well as the state-run labs, which process evidence for state agencies and local cities and counties that cannot afford their own labs.

But the revelations by the Chronicle, combined with previously reported problems at municipal crime labs in Houston, Bexar County and Fort Worth -- as well as the the DPS's decision to quietly suspend testing at its lab in McAllen for three months last summer -- have raised questions about the quality of information that his committee received from DPS officials and suggest that new solutions are needed.

"I personally feel that we were misled," he said.

DPS officials in an e-mail response late Friday stood by the quality of the work of the DNA labs, but emphasized that none of the problems resulted in the false identification of a suspect. They attributed many of the problems in the audit to documentation rather than faulty lab work.

The reforms were prompted by the DNA debacle in Houston. Last year, the Houston Police Department and the Harris County district attorney's office began independent retesting of almost 400 cases involving evidence processed by the police department's DNA division. That process continues. So far, one man, Josiah Sutton, has been released from prison after retests excluded him as a suspect in a rape for which he served 4 1/2 years in prison.

The City Council to date has authorized $4.6 million for the retesting.

Perhaps the most troubling item in any of the DPS's 2003 internal audits is found in the report on its Houston laboratory. The team of three DPS auditors found problems in interpreting DNA results when the evidence included multiple contributors.

"The interpretation of mixtures is not fully understood by some of the DNA analysts," the team reported in March 2003. On Friday, a DPS spokeswoman e-mailed that the interpretation was understood by all but one analyst.

Schneider noted that the misinterpretation of a DNA mixture was the key problem in the Sutton case.

"This is one of the basics of advanced DNA," Schneider said. "It's simple if done properly. If not, it's a problem. How can this lab function if that's the problem?"

A forensic consultant who works in the private sector concurred.

"Not understanding the interpretation of mixtures could give you a wrong result," said California-based DNA analyst Nora Rudin, who reviewed the DPS audits at the Chronicle's request. She, too, pointed out that it was faulty interpretation of mixtures that contributed to the closure of HPD's DNA lab.

The Houston audit also found problems with "macro hair examinations," the side-by-side comparisons of strands of hair, and in some cases the local analysts mistakenly concluded that they did not have enough hair remaining to test. Rudin noted that there have been recent cases in which DNA tests on hair have contradicted the other comparisons.

"If a hair has been lost," she said, a review "is impossible."

At the DPS lab in McAllen, testing was suspended from June 16 to Sept. 26, 2003, after auditors cited problems. Key among them was the allegation that the sensitivity of an instrument used to determine DNA profiles was not established before working on evidence. The shutdown led to a review of evidence processed by the lab in nearly 300 cases. Although DPS officials did not inform most law enforcement officials or any of the defendants or their attorneys about the probe, the agency says the review turned up no mistakes.

Alejandro Madrigal, the former supervisor of the McAllen DNA lab who was suspended -- for administrative reasons, the DPS says -- and then demoted, maintains that none of the faults found with his lab was as severe as the mixture interpretation problem in the Houston DPS lab.

"Those mixtures can get kind of hairy, and if you don't understand them, you've got a problem," Madrigal told the Chronicle before his reinstatement. "To me, this means they need more training (in Houston)."

Additionally, inspectors said the external vault at the Houston DPS lab did not have a protective fence around it and that "it can be easily approached at the rear or front without detection."

The problems extend beyond Houston and McAllen, however. Forensic consultant Rudin pointed to the El Paso audit, which indicates that statistical probabilities for possible DNA matches are provided "upon request" -- a practice she describes as highly unusual and one that could potentially lead to inexact evidence being introduced into criminal proceedings.

"Without a statistic, a match is meaningless," Rudin said. "So if they're issuing a lot of reports without statistics ... then that's misleading."

The audits also turned up significant problems in Abilene, Waco, Lubbock, Garland and at two labs in Austin. The problems range from bad methodology, storage and record-keeping to severe backlogs that could mean some trials have to go forward before evidence testing is complete.

Rudin also questioned the quality of the auditing. She pointed to an August 2003 recommendation by audit team captain C. Glen Johnson for lab personnel to use data from another DPS facility to validate work in the El Paso lab. Johnson did not respond to a request for comment.

DPS officials e-mailed that they stood by that system of inspection.

"That upset me more than anything I read," Rudin said. "And it was from an inspector, which tells me that the inspectors have no clue. The whole point (of the audits) is that you're supposed to do them in your own lab. That's what internal validation is. And that doesn't mean within a laboratory system. It means within a physical plant."

Regardless of the severity or quantity of problems found by the inspectors, Rudin said, almost every audit report concludes that the lab in question "is producing quality work."

"It's the same language every time," Rudin said. " `It's fine. They're doing a service. They're doing quality work.' And on every single one of them it was so gratuitous it stopped having any meaning."

Schneider's overall assessment was more pointed.

"If crime labs were in the private sector, they'd all be shut down," the Houston attorney said. "Business would not tolerate this kind of functioning."