Friday, January 6, 2012

What are the ethical roles and responsibilities for chemical laboratory safety?

Janet Stemwedel is a current professor of philosophy; (updated*) before she moved into philosophy of science, she obtained a Ph.D. in physical chemistry. She asks a series of questions on ethics in laboratory safety, regarding the Sheri Sangji case: 

"So, who is ethically responsible for keeping an academic chemistry lab safe? And what exactly is the shape this responsibility takes — that is, what should he or she be doing to fulfill that obligation?
  • What’s the responsibility of the principal investigator, the scientist leading the research project and, in most cases, heading the lab?
  • What’s the responsibility of the staff research assistant or technician, doing necessary labor in the lab for a paycheck?
  • What’s the responsibility of the graduate student in the research group, trying to learn how to do original research and to master the various skills he or she will need to become a PI someday? (It’s worth noting here that there’s a pretty big power differential between grad students and PIs, which may matter as far as how we apportion responsibility. Still, this doesn’t mean that those with less power have no ethical obligations pulling on them.)
  • What’s the responsibility of the institution under whose auspices the lab is operating? When a safety inspection turns up problems and issues a list of issues that must be corrected, has that responsibility been discharged? When faculty members hire new staff research assistants, or technicians, or graduate students, does the institution have any specific obligations to them (as far as providing safety training, or a place to bring their safety concerns, or protective gear), or does this all fall to the PI?
  • And, what kind of obligations do these parties have in the case that one of the other players falls down on some of his or her obligations?
If I were still working in a chemistry lab, thinking through ethical dimensions like these before anything bad happened would not strike me as a purely academic exercise. Rather, it would be essential to ensuring that everyone stays as safe as possible."

Here's how I would structure things (and how I have seen things structured in the past):

The principal investigator: The PI is ultimately responsible for running a safe lab. The best way to ensure that is to directly train the most senior members of the lab, to "train the trainer." From that point forward, the PI needs to ensure that senior graduate students and postdocs have a training and mentorship role in chemical safety and are properly executing that role. The PI should set up a safety/maintenance program, lay out the program and the responsbilities in clear, simple terms. The PI has the most prominent enforcement role; walking into their lab, they should spot safety problems and ask that they be corrected immediately. The PI has the most prominent foresight role; they should be able to see incoming safety problems from new chemistry, procedures or policy and be able to respond appropriately.

Postdocs and senior graduate students: Postdocs and senior graduate students (having been previously directly mentored by the PI) should be responsible for setting the day-to-day safety tone of the laboratory. They should be good examples and resources for newer lab members. While not directly responsible for the safety of younger graduate students, they should have a prominent advisory role and be willing to intervene (to the point of contacting the PI, if necessary) in order to keep lab members from doing egregiously unsafe things. They also have a foresight role, in that they should also be able to spot incoming safety issues and note them with the other group members.

Younger graduate students: They should be responsible for finding out and understanding the chemical safety risks they are taking on in their work. They are responsible for asking if they don't understand an issue and making sure that they don't proceed blindly. They also have the uncomfortable responsibility to question and/or dissent if they're being asked to do something unsafe.

The institution: The institution is ultimately (legally) responsible for the safety of its students; they're also the only administrative check on the power of the PI. If the PI cannot run a safe laboratory, they are the only people who can intervene. The institution has not discharged their responsibilities if they merely point out safety deficiencies without following up. The institution (via the EH&S office) has the responsibility to supply both basic chemical safety training (e.g. helping the PI develop a chemical hygiene plan) and for providing a 3rd party for graduate students and postdocs to turn to.

What if things don't work well? That's a problem with the way I've structured things: this system works well when you have a good PI. If you don't, well, that's an issue. While the institution may have the legal and/or ethical responsibility to intervene, I have a difficult time imagining a situation in which they can or will.

I'm not sure this answers Professor Stemwedel's questions in a satisfactory manner, but it's definitely how I see things. Readers, am I crazy?

*Thanks for DM for the correction of grammar. 


  1. I pretty much agree, and I think that is the way things are (more or less) supposed to work. But, like you said, you need good people (especially a committed PI) to make the system function. Of course, the system generally lacks meaningful methods of enforcement, so it is not uncommon to see breakdowns...everywhere.

  2. I don't think you're crazy (and I'm glad you made this very useful contribution to the conversation).

    As you note, the institution have some responsibilities to the students and technicians in the lab that overlap with the PI's responsibilities. Part of this is your standard safe workplace stuff -- making sure people in their employ have the resources they need to survive the terms of their employment -- and part of it is more under the aegis of "our program exists to create the next generation of excellent chemists," which brings with it a responsibility for the program (not just the individual advisors) to transmit best practices.

    And, the institution has a responsibility to conduct its scientific business in such a way that the public is not harmed by fires or hazardous chemical releases or whatever. (Maybe this is also connected to a broader responsibility not to make science look bad -- not to give the public reason to believe that chemists and the universities who employ them are reckless jerks who are more likely to pay fines when caught than to bother with prevention.)

    Something I'd love to see more discussion of, from you and your commentariat, is what kind of support for students and techs the institution should be offering. What if the grad student knows s/he doesn't have the training, equipment, etc. to do task X safely, has communicated this to the PI, and the PI still insists that the grad student do task X? (I'm worried about echos of the lack of support students asking reasonable questions in the Sames lab got from the Columbia Univ. chemistry department...)

  3. As far as I know Professor Stemwedel's PhD in Chemistry is neither expired nor revoked....

  4. These suggestions are all well and good. But if some substance is generated in the lab and someone is subsequently killed while working under this program will the SOPs provide adequate cover for all concerned? If not who will go to jail for not foreseeing this accident?

  5. From my few years in safety-related work I believe that every one of us is ultimately responsible for our safety. The advisors/management/institution is responsible for creating a "safety culture" where people are encouraged to ask, required to apply, and rewarded for achieving safety. The safety measures are reasonable an applied with reason, and the organization acknowledges that the unexpected happens and is proactively looking for potential issues.
    Students and new employees need to be introduced to this culture, but the culture itself needs to exist. Hard to do, but there is no other way. You can't inspect people into culture, only into compliance.

  6. A9:33a:

    That's an interesting hypothetical. Let me see if I can restate it: "If an previously unknown-to-be-toxic substance is generated in a laboratory, and laboratory workers are harmed, should there or will there be legal consequences for the PI?" Let's further up the ante and say that the hypothetical is set in LA County.

    Assuming that the foreseeable EH&S requirements for the laboratory are documented to be in place (work practices, administrative controls, PPE), I would think that the PI should be free from legal consequences. I wager that the DA and the police officers may sniff around, but there would be no path to prosecution. Do you disagree?

  7. I've always been surprised at how little EH&S actually did in the academic labs I worked for. They would go through an inspection once a year, which we prepared for, but after people got lazy and there were all sorts of violations. The inspections were CYA for the university.

    If the inspectors walked around at a random time, they would have seen all sorts of unsafe activity. This taught students a bad lesson: Safety was an administrative game but nothing to actually be concerned for.

    Unfortunately, a good safety officer is draconian safety officer (in academia). It's up to the university at all levels to empower them to enforce good practices. However, the university should also create a system where the safety department helps constructively by encouraging low level employees (and grad-students) to ask for help in rectifying safety issues and provide real aid without judgement.

  8. Hi Chemjobber,

    Well when it comes to capricious actions, government officials know no bounds, IMO.

    There are many scenarios: a student added waste to a waste can and a large quantity of chlorine dioxide was generated which subsequently explodes killing everyone in the lab; a student decided on his own to make chlorine dioxide without consulting anyone, because he was curious about the properties of chlorine dioxide, and it subsequently blows up killing the student; the student runs a reaction where chlorine dioxide is generated but was never noted before in the literature and the student is killed when he decides to distill the reaction rather than just wash it; a reaction that is known to generates chlorine dioxide is processed to remove the chlorine dioxide, but the flask of worked-up product explodes on the rotovap killing the student; no one in the biochem lab has any idea about chlorine dioxide, and inadvertently pulls a bottle of sodium chlorite from the shelf and adds it to an acidic solution, the subsequent explosion kill several students in the lab; a student opens a bottle of long stored material that has slowly generated chlorine dioxide over the years and then dies in an explosion when the stopper is removed; and on and on I could go about a single molecule.

    If you peruse the ten volumes of the Encyclopedia of Explosives (~8000 pages) you will encounter compounds you never imagined, let alone thought to be explosive. How many professors have done that so they can properly monitor the infinite possibilities for accident like the ones I described above?

    Now for the many ways to poison ones self…….

  9. A11:10a:

    There are many ways to look at the case. Here are two:

    1) Because the particulars (see below) of the case, Professor Harran is being charged with a crime. Academic chemistry will be affected somewhat by this case, but not nearly as much as the DA/public might hope.

    2) Because of our overlawyered culture, Professor Harran is being charged with a crime. Because of this, academic chemistry will lose a significant amount of freedom and research productivity will suffer.

    Not being a PI (but having been through the academic chemistry process), I believe #1 is closer to reality.

    You may disagree. It probably speaks more to our political/legal outlooks than anything else.*

    Particulars: young woman, an slightly large reaction setup for a total synthesis laboratory, poor technique, severe burns, death from severe burns, recent safety inspections that went poorly, an angry family, a public that is ignorant of general culture of academic chemistry, a district attorney position that is more politically sensitive than most.

    *I'll note, by the by, that I think American life is somewhat unhealthily over-regulated and we could all take a step back from the courthouse.

  10. Well, only the “recent safety inspections that went poorly” aspect of the situation you described above gets addressed by your safety suggestions. Personally I have never seen a safety inspection ever prevent an accident. Accidents are called accidents because that is what they are. They always have a cause or they would not happen. However, I have never met anyone clairvoyant to predict an accident before it occurs.

    The lessons we should all be learning: 1) never say anything to anyone after an accident; 2) make sure before you speak with anyone you hire a criminal attorney to advise you before speaking about any safety incident; 3) no doubt your attorney will advise you to say nothing to anyone ever; 4) put in place an many safety SOPs as you can dream up to provide everyone with fail safe CYA; 5) assume someone will be blamed no matter who is at fault because someone must always be the fall guy when an accident occurs.

    If I were the defense attorney, I would call in PIs and working chemists from every major synthesis lab in the country and put each on the stand. I would ask them about their laboratory experiences with dangerous chemicals and about how the learned to handle these materials, plus what experiences they had which were near misses and whether they could foresee those near misses. I would call in Dupont chemists who function in a nearly perfect safety infrastructure and ask about laboratory incidences they have had despite their suffocating safety culture.

    I believe, like you, that safety concerns will get some heightened visibility and SOPs will be put in place in academic labs. However, hyper vigilant safety is impossible to maintain unless you are working with known, highly dangerous stuff like fluorine or cyanide and even then bad stuff happens. People will slip back into a normal state of work until the next accident happens, then sure as the sun rises in east someone will get blamed and a now newly appearent trail of non-vigilant safety violations will point the finger at that someone.

    You know sometimes bad things happen to you in life whether you deserve it or not. This case is just about spreading around the bad life stuff to Professor Harran because a pound of flesh is required for this incident.

  11. A 8.50a:

    What if the grad student knows s/he doesn't have the training, equipment, etc. to do task X safely, has communicated this to the PI, and the PI still insists that the grad student do task X?

    I think this is an important point. I'm sure many of us have worked in labs where the pressure for results is intense. While most PIs would never insist that a student does a task that the student has expressed concern over, the need to get something done, or produce results quickly can mean that students can make risky choices on whether to proceed with a reaction or spend a few hours researching the safety implications. My PI would never insist that I work alone in the lab, late at night (and has said that this is unsafe) but does expect results at a rate that means not doing so is not an option.


looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20