Monday, October 1, 2012

Chemist faces charges for "dry-labbing"

Dookhan faces more than 20 years in prison on charges of obstruction of justice and falsely pretending to hold a degree from a college or university. She testified under oath that she holds a master's degree in chemistry from the University of Massachusetts, but school officials say they have no record of her receiving an advanced degree or taking graduate courses there.

Last last week and over the weekend, more details of her case have arrived. From the AP:
Chemist Annie Dookhan was "Superwoman," a colleague at a Massachusetts state crime lab used to joke. She seemed unstoppable in her quest to please prosecutors, police and her bosses, testing two to three times more drug samples than anyone else, working through lunch and not bothering to put in for overtime. "The kind of person, if you owned your own business, you would want to hire her," a supervisor would later tell police. 
Beginning about four years ago, suspicions arose about the way she seemed to plow through so many cases so fast. After that, a supervisor complained he never actually saw her in front of a microscope. But her superiors let her work on. Now, the startling explanation has come spilling out: Dookhan told investigators she faked test results on drug samples and cut other corners.
Dookhan's alleged confession and the missed warning signs were detailed in state police reports obtained this week by The Associated Press. As early as 2008, a supervisor noticed Dookhan's testing numbers were high. He spoke to her superior, but nothing happened. In 2009, the supervisor took his concerns to another superior, saying he never saw Dookhan in front of a microscope.  
Dookhan was handling a staggering number of samples. An average chemist could test 50 to 150 samples a month, but Dookhan was doing more than 500, according to monthly reports, a lab employee told police later. At one point in 2011, a top official even gave Dookhan a special project to try to slow her down. Other lab employees worried about their own jobs and their prospects for promotion because their productivity wasn't as high as Dookhan's.
She was eventually suspended from lab duties after getting caught forging a colleague's initials on paperwork in mid-2011. Dookhan resigned last March as the Public Health Department investigated. Over the summer, police interviewed Dookhan at her home, and she admitted faking test results for two to three years, forging signatures and skipping proper procedures, according to the reports.  
Dookhan said she cheated by "dry labbing," or identifying a drug sample as a narcotic by looking at it instead of testing it. She said she routinely tested only five out of every 25 samples, and deliberately turned a negative sample into a positive for narcotics a few times. At least one defense attorney has also accused her of overstating the weight of some of the seized drugs to expose defendants to heavier penalties.
From the Boston Globe, a fairly detailed look at some of the tests she was accused of faking:
But the State Police report suggests Dookhan herself may have, at times, served in another quality assurance role, as the lab’s quality control chemist, who typically runs daily tests to ensure scales are calibrated and machines are running properly.“These machines could have been used by other chemists, who did not even know that the machines were not properly verified,” said Workman.
That possibility, Workman said, would call into question a much larger universe of drug tests beyond the 60,000 Dookhan is believed to have run during her tenure.... 
A one-page overview of the lab’s routine procedures, provided to the Globe from someone familiar with its operations before it was closed in August, indicates that a primary chemist is assigned a sample from the lab’s evidence room, weighs it, and performs testing to preliminarily identify it. That person then prepares vials of the substance to send to a secondary chemist, known as the lab’s GC/MS person, who analyzes the vials in a machine to confirm the identification. 
That process, known as gas chromatography–mass spectrometry, first separates drug samples into their component parts. For instance, it separates cocaine from the baking powder, laxatives, or other substances drug dealers typically cut the drug with to inflate its weight. Next, the machine analyzes the chemicals and prints out a characteristic pattern, similar to a bar code, that is used to identify a sample. For instance, it compares the pattern for the sample of presumed cocaine being tested to the pattern of a known sample of cocaine. 
The GC/MS chemist analyzes the printout and records the results on a control sheet, which is initialed by that chemist and the primary chemist on the case, and then the primary chemist returns the original samples and all documentation to an evidence officer. “The GC/MS [machine] produces a printed, recorded analysis of the sample, but unless you are highly trained you may not be able to know if it’s traceable to a specific sample of evidence,” said Justin McShane, a Pennsylvania criminal defense attorney and senior instructor in gas chromatography-mass spectrometry at the American Chemical Society, a trade group of more than 164,000 chemists. 
Because of the complexity, and lack of cameras recording the preliminary tests, McShane said, it is possible for an unscrupulous chemist to dupe colleagues and prosecutors and defense lawyers, who depend on the work but are rarely given more than a simple card that indicates whether the evidence was positive or negative for ­illicit drugs.
McShane said he routinely hears “horror stories” from chemists he trains about unrelenting pressure to test more samples. Dookhan allegedly confessed to State Police that she forged colleagues’ initials and contaminated samples to “get more work done,” according to their report.
“You are judged by numbers in the lab,” McShane said. “There is a culture of pressure to get it done with no new ­resources. But there is no ­excuse for [cheating] at the end of the day.”
I think my concerns about overproductivity are bearing out. (Of course, as readers would note, that is not immediate evidence of pencil-whipping. A pink flag, not a red one, I say.) More soon.


  1. Sorry not tied directly to this post but interested is your take of Abbott CEO CV "errors" (as to over inflation of credentials

  2. Oh what a tangled web we weave...

    Amazing story. Dookhan is rewarded for her bad behavior, so of course she keeps doing it. It's not surprising that people were questioning her work all the way back to 2008. Tests take as long as they take to complete. A 3x productivity rate has to be a major red flag. Dookhan was able to get away with it, and now every case (not just hers) back to when she was hired is going to be under scrutiny.

  3. Two things that suck because of this:

    1) People may have been wrongfully convicted of crimes they didn't commit.

    2) Criminals probably will go free when their cases are overturned due to her data.

    This will cost the state millions of dollars to have to either re-try cases or settle the lawsuits due to her 'productivity'.

    What a nightmare.

    1. I think any case handled by the lab in which she worked will have to be re-examined. She didn't just sabotage her own cases, she hurt everyone else's since she was in charge of instrument calibration which she didn't do. This will be a big, expensive mess.

      I doubt there will be much they can do. Many samples will have been destroyed by now, but defense attorneys certainly have enough reasonable doubt to get a lot of convictions overturned. I imagine MA will have to appoint special counsel to handle this.

  4. Should be an interesting stay for her in the pen, considering she helped send some of those people there.

  5. To avoid problems like this, the state of Pennsylvania requires that accredited environmental laboratories have an ethics policy and give ethics training to employees within 2 months of hiring and at least every 14 months thereafter for all employees. Management is to develop a proactive program to prevent fraud and unethical behavior.

  6. I'm interested in what changes would have to be made to make sure this doesn't happen.

    One already mentioned was video of the testing. That seems invasive to me but I guess it would be reassuring.

    Another would be for HR to follow up on potential hires degree and previous work claims. Probably get a transcript sent direct from the university and get employment confirmation from past employers. Actually calling people's references, too.

    I think a third way would be to introduce "blind" testing. That is, the testing employee doesn't know what the suspected substance is. I assume sometimes you can tell by looking at it, especially if you're experienced but there should be no getting around the initial testing (which was skipped in this case).

    More cross-checking, having others sign off that they witnessed the test and it was done accurately. That's probably be more employees but is probably cheaper than all the money this rogue employee will cost the state.

    1. You don't need background checks and video surveillance. That is perhaps you do, but getting rid of managers excessively fond of brown-nosing employees would do much more good.

  7. Now to the more important news. CJ, did you see that ACS and me personally were featured in today's Dilbert?

  8. This is why I'm terrified of political talking points requiring things like drug testing for people to receive government benefits. There are so many questions people who are in favor of it can't or won't answer (most I've talked to have little or no scientific background): who pays for the tests? Who oversees the testing or performs QC? How do they confirm positives vs false positives? What is the time frame for this stuff?

    For most of us our bench work only impacts our own productivity. But in drug testing labs, people's lives are on the line.

  9. Another argument against the death penalty. Even DNA-based evidence is not to be trusted when handled by flawed human beings.

  10. This should give laid-off pharma researchers a glimmer of hope: there will be a bonanza for expert witnesses in MA as defense attorneys re-examine evidence against their clients.

  11. This is qvxb in the spam-filter evasion mode. To avoid problems like this, the state of Pennsylvania requires that accredited environmental laboratories have an ethics policy and give ethics training to employees within 2 months of hiring and at least every 14 months thereafter for all employees. Management is to develop a proactive program to prevent fraud and unethical behavior.