More convictions are affected than Houston prosecutors might want to admit.

Preliminary findings last month showed 40 percent of DNA cases examined and 22.5 percent of blood-test cases scrutinized between 1987 and 2002 had major errors.

And that's only part of the story.

Michael Bromwich, an independent investigator hired by the city in 2005, is extending his inquiry seven years further back, to 1980, casting doubt on hundreds more cases. ... According to Mr. Bromwich's reports, poorly trained lab workers faked or misinterpreted tests, withheld exculpatory findings and gave false testimony in court.

Why haven't more convicted defendants been exonerated? Brownwich faults defense lawyers for not paying enough attention to closed cases that may have been affected by faulty crime lab evidence.

One problem, critics say, is the difficulty that inmates face getting and keeping attorneys and navigating the system, which provides limited funding for free legal work and imposes hurdles to control frivolous claims.

It's also a fact that neither defense lawyers, often representing indigent defendants by taxpayer-funded court appointment, nor district attorneys, whose staffs tend to be lean, have the resources to review a large number of cases quickly.

It's a question of providing resources where they're needed.

"There has to be focused, systematic attention brought to bear on these cases," said Barry Scheck, co-founder of the Innocence Project, which has done its own work in Houston cases.

"This is the single worst forensic scandal in the history of American justice, and it's taking place in a jurisdiction [Harris County] that has executed more people than any state except Texas and Virginia."