Colorado's Blood Fiasco Twice as Bad as Originally Thought

It’s rather dis­con­cert­ing to find out that the error rates in blood alco­hol test­ing not only are flawed, but twice as bad as orig­i­nally thought. In DUI tri­als, there is a pre­sump­tion of guilt, so any­time you have a sit­u­a­tion where sup­pos­edly “objec­tive”, “impar­tial” and “irrefutable” evi­dence shows a sys­temic error that could lead to false con­vic­tions, things need to be dis­closed quickly.

The num­ber of flawed blood alco­hol tests has dou­bled since a Colorado Springs Police Crime Lab inter­nal audit last year dis­cov­ered that some results were inflated, author­i­ties said last week.

Results in 167 DUI cases have been called into ques­tion, up from the ini­tial 82 that were dis­closed in December, accord­ing to 4^th Judicial District Senior Deputy District Attorney Frederick Stein.

Stein could not say yet what the impact of the erro­neous tests would be on the crim­i­nal cases involved, but added that the DA’s office hopes have a report com­pil­ing those out­comes within the next two weeks.

On Dec. 11, the police depart­ment reported that a rou­tine audit had dis­cov­ered the errors and began both an inter­nal and exter­nal inves­ti­ga­tion into how reported test results came in higher than the actual level.

At the time, the depart­ment started re-testing about 1,000 blood alco­hol test results taken since January 2009.

Meanwhile, pros­e­cu­tors con­tacted indi­vid­u­als in the cases that were based upon the incor­rect test results. The District Attorney’s office also fielded about 140 phone calls in December on a hot line set up for peo­ple who wanted to know if theirs was among the affected cases. Authorities have not released the names of the defen­dants in those cases.

The Colorado Bureau of Investigations is con­duct­ing the exter­nal inves­ti­ga­tion while the police depart­ment con­ducts its own inter­nal review. Both of those inves­ti­ga­tions are ongo­ing, police spokesman Lt. David Whitlock said Friday.

“I can’t say any­thing defin­i­tive about the under­ly­ing cause,” Whitlock said. “What has occurred is that we’ve nar­rowed it down to the point where we believe it is likely a human error. We believe very strongly that it’s not an equip­ment fail­ure or some­thing inher­ent in the lab itself.”

He would not say whether inves­ti­ga­tors believe the errors were made by an indi­vid­ual or sev­eral people.

In January, the police depart­ment also dis­closed that a refrig­er­a­tor that held DNA evi­dence had mal­func­tioned over a four-day period in December. Police, how­ever, said experts did not believe the 235 sam­ples of blood, semen and urine had been ren­dered unus­able by the accident.

One Colorado Springs lawyer who spe­cial­izes in DUI cases said the depart­ment needs to let the pub­lic know what caused the blood alco­hol errors to hap­pen in the first place.

“If you iden­tify the prob­lem then you should be able to fix it,” said Timothy Bussey, a for­mer pros­e­cu­tor. “But if you can’t, then that’s a problem.”

Bussey filed a Colorado Open Records Act request­ing detailed infor­ma­tion on the lab errors. In response, the depart­ment pro­vided him with some records, but not most of what he requested. He has filed an appeal.

He did obtain one doc­u­ment, how­ever, that gives a clue as to what inves­ti­ga­tors believe might be the source of the prob­lem: a dis­crep­ancy in the con­cen­tra­tion of the chem­i­cal n-Propanol, aclear, polar water-soluble sol­vent
with a very mild odor.

According to a Dec. 9 memo from Crime Lab Supervisor Ian Fitch, evi­dence sug­gested that in the flawed batches, less n-Propanol was added to the unknown blood sam­ples than to those being used as a stan­dard and control.

Whitlock could not esti­mate how long the inter­nal inves­ti­ga­tion will take, but said the depart­ment is eager to share what it finds out.

“We know this is some­thing that is a con­cern to the pub­lic and involves our cred­i­bil­ity,” he said. “We’re eager to tell them how we’ll fix it.”

Article from the Colorado Gazette.

I was con­tacted by a jour­nal­ist out of Colorado Springs, a John Ensslin, who had obtained copies of inter­nal memos from Colorado Springs Police crime lab. Mr. Ensslin wanted some help trans­lat­ing the con­tent of the memo into plain English. I spoke with him tele­phon­i­cally, and then proof read his fol­lowup arti­cle here.

First off, a tip of the hat to local lawyer Tim Bussey for think­ing like a reporter and fil­ing a Colorado Open Records Act request.

While the police depart­ment did not give Bussey every­thing he wanted, they did deliver a Dec. 9 memo from Crime Lab Supervisor Ian Fitch, who wrote the following:

“A root cause analy­sis inves­ti­ga­tion, which included input from Colorado Bureau of Investigation (CBI) Agent Tim McKibben, strongly sug­gested that the high blood alco­hol con­tent val­ues reported in the pre­vi­ously affected cases resulted from a dis­crep­ancy between the con­cen­tra­tion of inter­nal stan­dard, n-Propanol, added to the stan­dard curve com­po­nents (stan­dards) and pur­chased con­trols (con­trols), and that added to the unknown blood sam­ples being tested. Evidence sug­gests that in batches affected by the non­con­for­mance in ques­tion, less n-Propanol was added to the unknown blood sam­ples than to the standards/controls. This would explain the inac­cu­rate high ethanol val­ues reported, since the amount is cal­cu­lated rel­a­tive to the n-Propanol.

Regardless of exactly how the lab man­aged to cause an infla­tion of reported blood alco­hol results, it demon­strates an all to com­mon prob­lem with foren­sic sci­en­tists now days. Many of them have a seri­ous lack of under­stand­ing of basic chem­istry con­cepts. They can be taught how to per­form a series of steps, like fol­low­ing recipe. When they have to under­stand and trou­bleshoot the steps of the process, many just sim­ply don’t under­stand. They can’t trou­bleshoot a faulty instru­ment. They can’t mix reagents prop­erly. Blame it on the explo­sive pop­u­lar­ity of “foren­sic sci­ence” degree pro­grams, many arrive to work lack­ing the min­i­mum under­stand­ing of chem­istry needed to be a good foren­sic scientist.

It’s amaz­ing that defense attor­neys, more and more, have to resort to pub­lic infor­ma­tion requests under the Freedom of Information Act, to get poten­tially excul­pa­tory infor­ma­tion out of crime labs. So much to “inde­pen­dent” and “unbiased”.