Wednesday, September 12, 2012

More on random drug testing...

Last week I was on CBC discussing random drug testing, noting that there was no evidence random drug testing reduced injuries. You can see a summary of the research that I posted the other day after several folks requested it. 

I provided a link to this summary to the organization behind the random drug testing because they wanted a copy of the studies I'd reviewed. Tonight I got a letter from them. I responded and I thought posting both bits of correspondence (since the letter to me was copied to CBC!) might be insightful for students interest in how evidence contributes (or not) to policy making.
-- Bob Barnetson (not Bartensen!)

September 11, 2012

Pat Atkins, Administrator
Drug & Alcohol Risk Reduction Pilot Project

Dear Dr. Bartensen:

Thank you for taking the time to speak with me on the phone this morning to discuss your guest appearance on the CBC Edmonton radio show on September 5, 2012.

During the CBC segment you indicated that you were not able to find any studies that link workplace safety and alcohol and drug testing, and could not find any evidence of the deterrent effects of random testing. When I asked if you could provide me with copies and links to the research you mentioned on the radio, you said you could not remember any specific studies and that you based your conclusion on what you learned from a journal browse in the library.

As the Administrator of Alberta’s Drug and Alcohol Risk Reduction Pilot Project (DARRPP), I have accumulated some of the current research on the subject of alcohol and drug testing in the workplace, and I thought it would be helpful to share some of this information with you so you can refer to it in the future.

I also have some other relevant information to share relative to alcohol and drugs generally.

First, I am attaching an article from Cornell University which indicates a strong link between the implementation of alcohol & drug testing and increased safety in the construction industry. It is entitled “Construction companies find benefits in testing for drugs, Cornell study finds.” This study found that the implementation of drug testing reduced injury rates by 51% within two years of implementation.

I would also like to refer you to the US Department of Transportation (DOT) study, also attached. Random alcohol and drug testing was implemented by the DOT for motor coach carriers in 1995, and, in 2008, they had achieved a 47% decrease in positive tests for drugs and a 60% decrease in positive tests for alcohol for workers who were in a random testing program for both drugs and alcohol. This same study showed that in 2008, for workers in random testing, the post-incident positive test rate for drugs was down to 1.24%.

In addition, a separate study of the U.S. road transportation industry, which looked only at fatal crashes where alcohol was involved during this same time period showed a 23% decrease for workers who were random-tested. This shows that random testing can be successful in deterring workers from being impaired on the job and endangering themselves and others.

In our conversation I pointed to recent events at Toronto Transit. In 2007 a serious accident occurred on the job where the operator was killed and several workers seriously injured. In a post-incident test, the operator was found to have measureable levels of THC, likely consumed during his work shift. Following this incident, a comprehensive A&D testing program was approved by the TTC, which was implemented in 2010. At that time, random testing was recommended but not approved. In 2011, another accident occurred resulting in the death of a passenger. The driver was charged with criminal negligence causing death and possession of cannabis. Following this event, the TTC approved random testing. I think these studies and findings are quite relevant.

In Northern Alberta, many organizations have had extensive safety programs in place for decades, making every effort to take actions to ensure the health and safety of workers and the public. Most have also had A&D testing programs in place for 10 years or more, as well as education programs, training for workers and supervisors, availability of EAP and treatment programs (often paid for by the organization), speakers and symposiums for workers and supervisors. In spite of these efforts, drugs and drug paraphernalia are still being found in camps, on heavy equipment and on work sites, and workers continue to test positive in post-incident tests. Pre-site access testing still results in substantial numbers of positive tests, in spite of the fact that workers are aware that the testing will occur.

One organization also found that close to six times more alcohol and drug abuse problems are revealed in substance abuse expert assessments after incidents occur than pre-emptive tests prompted by supervisor observations – and this is in spite of years of extensive education and supervisor training. Continuing to maintain the status quo and ignore the pressing need to improve workplace safety would be irresponsible on the part of site owners, who have a legal obligation to maintain a safe workplace. Given the data we have about the positive effects of random testing as a deterrent and its clear, measurable impact on improving workplace safety, the introduction of random testing for safety sensitive positions in Alberta’s oil sands and construction industries is a reasonable and responsible action.

DARRPP will also provide accurate Alberta wide statistical information relative to alcohol & drugs. The scope and efficacy of such information ought to allow intelligent and insightful conclusions to be drawn. This is part of the reason this is a “pilot project”

Employers are also responsible for respecting the human rights and privacy of employees. DARRPP participants all accept their obligations under the Alberta Occupational Health and Safety Act, the Alberta Human Rights Act, and the Personal Information Protection Act. For your reference, I am also attaching DARRPP’s participant guidelines, which clearly document these obligations.

Random workplace alcohol and drug testing is a powerful tool that has been proven to deter irresponsible behavior, prompt people with dependencies to get help, reduce injuries and save lives. As DARRPP’s Administrator, I can assure you that the pilot project has been designed to respect human rights and privacy, and I am confident that the testing will be implemented responsibly by participating employers.

As a former oil sands worker and resident of Fort McMurray for over 30 years, I am concerned at the alcohol and drug-related risks that are present in Alberta’s oil sands and construction industries. I trust you will find this information useful. Please let me know if you need additional information.

Pat Atkins, Administrator
Drug & Alcohol Risk Reduction Pilot Project

CC: CBC Edmonton Radio


Thanks for your letter.

I’m not entirely sure what to make of your letter, but I did want to clarify a few things,

First, the assertion I made on the radio and on the phone was that I could find no studies that linked random workplace drug testing (as opposed to alcohol testing) to injury reductions and that I could find no studies that linked drug use to increases in injury rates. When you asked for the studies, I indicated that I didn’t happen to have a list, rather I had simply spent some time in the library reading studies.

But I did undertake to provide you with a list. I emailed you a link to my blog (where I posted the studies I reviewed) alone with a summary of each. You’ll find that they do indeed support my assertion. Most of interest to you will be the systematic review done for the government of Alberta in 2006—it is the most recent and most comprehensive.

Second, the Cornell study which you referred me to (and which I provided you a copy of since you hadn’t bothered to get a copy or read it!) did not refer to random drug testing (which is what is at issue). Rather, it refers to drug testing generally (which may or may not include random testing). Additionally, it isn’t a very good study (low response rate, small sample). So it doesn’t support random drug testing as a way to reduce injuries, which would have been evident if anyone at DARRPP had bothered to read it.

Third, I can’t comment on the DOT study (it was not attached) other than to observe that a study that finds random testing led to a reduction in positive tests does not tell us anything about whether random testing resulted in a reduction in injuries. All it tells us is that positive test results dropped.

Fourth, the study about random alcohol testing is not relevant to random drug testing. Alcohol and drugs appear to be quite separate issues and I have little concern with alcohol testing if there is some good reason to do so.

Fifth, the TTC examples are interesting, but again they do not provide any evidence that random drug testing reduces injury rates or the much more basic issue of whether drug use (outside of work or while at work) results in an increased rate of injury. An equally valid interpretation is that TTC got spooked and decided to cover their ass via testing after the second incident.

Sixth, your assertion that this is a “pilot study” is problematic in a couple of way. This study is not occurring in a laboratory with willing participants and disinterested researchers. You members are compelling their workers to participate in this study (which is unethical). These “researchers” have multiple and often conflicting interests at play which compromise their judgment. There is no reason to believe that your intervention will be effective based on the available data. Having such a belief is generally a precondition to experimenting on humans, where the experiment could result in significant harm (in this case, reputational and career consequences). And the rationale you are using justify invading the privacy of workers (i.e., that drug testing will make the workplace safer) assumes the hypothesis you are researching is true. Yet all of the available evidence indicates the opposite! 

Seventh, you note that:
“Random workplace alcohol and drug testing is a powerful tool that has been proven to deter irresponsible behavior, prompt people with dependencies to get help, reduce injuries and save lives.”
On the issue of random drug testing, this simply is not true. There is no evidence of this! What you are making is an ideological statement. You are entitled to have this opinion, but when you take action that affects the livelihood of others, you need to be able to substantiate such a statement and you have not done so.


1 comment:

  1. I’ve got to say it’s rather a novelty to arrive at a relatively unique blog like this, great effort. I expect I’ll be coming back again soon.