That has prompted D.C. Police Chief Cathy L. Lanier to order officers to take urine samples from suspected drunk drivers. That is a more cumbersome and expensive way to handle the 120 tests a month the District runs. Urine tests cost police $75 apiece, compared with less than $10 for each breath test.
Drunken-driving arrests for the first quarter of 2011 were down by about 40 percent from the same period last year, according to the police union. Lanier and the attorney general’s office have not attributed that decline to the lack of breath tests. But some frontline officers, who have been responsible for hundreds of arrests in previous years, said the confusion in the system leaves officers hesitant to make arrests.
The District’s breath testing program unraveled in February 2010, when city officials acknowledged that errors in the old analyzers had overstated drivers’ scores in about 400 convictions.
D.C. officials have known since October that the battered breath test operation needed an overhaul, not incremental fixes. That conclusion came from a team of national experts who volunteered to help the city and contacted the attorney general’s office last fall, interviews show.
But questions of how to pay for the rebuilding and who will supervise it have remained unresolved.
A lack of accuracy was not the only weakness exposed by the testing debacle. In regard to manuals, training systems, scientific rigor and quality control, the program came up short of those of other states, the experts said.
In Maryland, Virginia and other states, breath alcohol testing is the responsibility of a centralized forensic science unit. Lanier and the attorney general’s office have asked the D.C. medical examiner’s office — the city’s forensic arm — to run the bulk of the program.
But without more money, more staff and more expertise, the medical examiner has said she is reluctant to take over a broken system.
Neither the attorney general’s office nor the police department could say with precision how the shift away from breath tests has affected enforcement or court outcomes.
Drunken-driving caseloads had been steadily increasing, according to records in D.C. Superior Court. There were 1,391 drunken-driving cases in 2008, compared with 1,707 last year.
But caseloads slowed this year, with 174 drunken-driving cases in January and 135 in February.
Arrests also appear to have slowed in the District, with 155 as of early March, compared with 271 for the same period last year, according to the police union. The police department could not provide comparable statistics.
Without breath scores, prosecutors have used other evidence to obtain convictions, including field sobriety tests and drivers’ admissions they had been drinking, the attorney general’s office said in written responses to questions from The Washington Post.
Prosecutors also said the testing errors have caused few cases to be overturned because other evidence preserved them.
But several defense lawyers said that looking at case dispositions isn’t a way to fully measure the turmoil in the system.
After the high test scores were revealed, it was left to defendants, including about 200 who had served jail terms, to go back to court if they wanted to make challenges. Many people don’t have the money or time to try to get convictions overturned, even after they learned last year that their cases involved incorrect scores, several defense lawyers said.
Under D.C. law, a driver can be convicted of driving while intoxicated based solely on breath alcohol test results of at least .08, which makes the scores critical.
Of the roughly 400 convictions flagged last year, only about 40 DWI sentences were challenged, according to a case-by-case review by The Post. Nearly all of those resulted in a subsequent plea to driving under the influence — a charge that does not rely on having a test score in evidence.
About 30 civil cases challenging the breath testing program also have been filed.
Creating a national model
The team of experts focused on the breath test program has been volunteering time to the District and had expected that the process would have moved further along, Stephen Talpins said.
Talpins is a lawyer and former Miami prosecutor who reviewed the District’s procedures with the heads of the breath alcohol testing programs for Virginia, Florida and Wisconsin.
“We’re trying to get several things done,” including finding the money and setting a schedule for experts to build a D.C. program that is not only better but also could be a national model, Talpins said. “We don’t want to convict people of a crime we aren’t sure they committed.”
Unlike Maryland and Virginia, the District does not have a law that builds in layers of supervision and safeguards for a breath test program.
D.C. law says that breath test machines have to be checked every three months for accuracy without saying by whom or under what standards.
The job fell to the police department. And within the department, control fell to one officer who managed the program for 14 years — and did not detect testing errors that occurred after he made adjustments to the machines’ motors starting in 2008.
A contractor hired Feb. 1, 2010, to help with the breath testing program ran accuracy checks and discovered the inflated tests scores on his second day on the job.
“It was a mess,” said Ilmar Paegle, the consultant.
Some scores were inflated by 40 percent, Paegle told prosecutors last April in a memo — which is above the 20 to 30 percent error rate regularly cited later in public statements by police and prosecutors.
Paegle is a retired U.S. Park Police officer who had worked in that department’s breath testing program. The District hired him for a year under a federal grant to be the temporary manager of the breath test program after the officer who had run it shifted in 2009 to an assignment he had requested.
When he took over the job, Paegle tested the equipment to see whether the machines correctly recognized the alcohol concentration in solutions in which the strength was known.
The machines didn’t.
“That was my ‘oh crap’ moment,” Paegle said.
The machines were pulled out of police districts Feb. 4, 2010, although officers on the street were not told what the problem was.
Prosecutors began offering pleas in court without explicitly telling defense lawyers that there was an accuracy problem with the machines.
By June, the attorney general’s office had assembled a list of cases that included results from machines that were wrong.
“Anyone can punch in the wrong numbers at times, but you check your work for accuracy and that shows you your mistake,” Paegle said. “That didn’t happen, and [breathalyzer] instruments that never should have left the shop went out to officers to use in arrests.”
The attorney general’s office said it had been told that no disciplinary actions over the testing failures had been taken by the police department as of late February. But citing D.C. personnel rules, the chief’s office declined to say anything more about the investigation.
The machines gave accurate readings once the mistaken calibration was detected and corrected, Paegle said, but by then, they bore such a taint that the department replaced them in March 2010, and Paegle began retraining officers on new equipment.
But the program still lacked rigorous outside supervision.
In a joint appearance in July after the testing error blowup, Lanier and former attorney general Peter Nickles announced that the medical examiner’s office would “take a more active role in overseeing maintenance and operation” of the program.
That language was more expansive than the terms that appear in the written agreement signed a month before, a review of the formal memorandum shows.
The medical examiner’s office agreed in June to provide “scientific and limited administrative oversight of MPD’s breath alcohol testing program” until 2015. It would review, but not design, procedures, according to the agreement, which also said that the medical examiner’s office would not oversee training on testing procedures or directly work with the machines.
Beverly Fields, a spokeswoman for the D.C. medical examiner, said the medical examiner would not comment about breath test issues past or present “due to pending litigation.”
Lanier announced the shift to urine testing Feb. 4, a year to the day after the widespread mistakes were confirmed and four days after Paegle’s contract with the city ended.
Lanier said she and the attorney general’s office have turned to the expert panel to design procedures, supervise a breath test program and perhaps have an interim plan in place that could shift to the District’s centralized forensic laboratory, which is set to open in fall 2012.
“It was our hope that we would have protocols established in a shorter period of time,” Lanier said in an e-mail reply to questions about the continuing lag.
Acting Attorney General Irvin B. Nathan said, “We intend to create a program of which the District can be proud and that can be used as a model.”
What role the medical examiner will play is not clear.
In recent testimony, the medical examiner, Marie-Lydie Pierre-Louis, said her office had not been set up to run breath alcohol testing and did not have the money or expertise to take over that job.
She appeared before one of the several hearings held by D.C. Council member Phil Mendelson (D-At Large), who has continued to press police and prosecutors for an “unambiguous explanation of the state of drunk-driving prosecutions in the District.”
“This is just bad for the District’s reputation,” Mendelson said. “We don’t need this tarnish.”