I Found the Questions (Not the Answers) in a Bottle

A Today Show story on children getting their tongues stuck in steel water bottles (January 9, 2013) reminds us of how complex regulatory decision making can, and should, be. According to that story, a few children have gotten their tongues suck in steel water bottles and had to have the water bottles removed surgically putting the child at some risk in the process. With tongue planted firmly in cheek, I might suggest that the next regulatory step is to recall and then ban all steel water bottles, but I will resist and instead use the water bottle example to discuss the real life complexity of regulatory decisions.

In the past, I have discussed what safety means [link] and suggested that there are trade-offs and realities that render the concept of absolute safety to be meaningless for regulatory purposes. Furthermore, where you spend your safety dollar determines how effective you are at improving overall safety; eliminating minor hazards at great cost to resources may impair an agency’s ability to make things safer. Too often our regulatory dialogue is simplistic. We either “protect the public” or we are “baby killers.” It is important that such simplistic rhetoric not result in simplistic decision-making.

With some of these issues in mind, I looked for data on the numbers of steel water bottles in use and tried to determine the level of risk. (Put another way, the rate of incidents versus the number of products in use.) I could not find a lot of useful data. What I did find, is a very small number of reported incidents that led to the NBC report of tongue entrapment. Pinning down the number of steel bottles in use was equally difficult. Over the last several years, state and local politicians have banned BPA from water bottles in several jurisdictions and various retailers have effectively done so as well. These actions have taken place despite fairly wide-spread disagreement over the alleged “hazards” from BPA. (Not being a toxicologist myself, I spoke to a couple of former government toxicologists who I know have never been apologists for industry, and learned that the evidence of any real hazards from BPA is—to be polite—somewhat over-hyped.)

In addition to regulatory efforts to replace plastic bottles with other materials, I found environmental web sites that urge people to use reusable bottles in lieu of disposable plastic water bottles. Those sites actually touted the safety of steel water bottles.

In addition, various laws have also banned or regulated phthalates, particularly in children’s products. I found many firms offering stainless steel water bottles as the solution to either the BPA, phthalate, and/or environmental concerns. From all of this, one might hypothesize a substitution effect where steel bottles replaced plastic ones in the marketplace increasing use of steel bottles and replacing one set of risks with others. However, I did not find reliable data on trends in bottle production, so this hypothesis remains merely that.

As a potential regulator, I would want to find evidence that allows a careful analysis of the relative risk issues. Of course, as a regulator, I would want to look beyond just the substitution of tongue capturing steel bottles for those with BPA. I would want to evaluate concerns that the chemicals that are used in place of BPA or phthalates are less well understood and tested and may present more risk than BPA.

In theory, regulatory agencies look at the level of risk and have to satisfy certain statutory criteria before taking action to recall a product, but it is hard to look parents in the face and say “this risk is too small to regulate” after they have lost a child or sat anxiously beside their bedside for several days in the hospital suffering with them. This is true even when the parent or care giver failed to follow instructions or properly maintain a product. Our sympathy for their loss makes us want to blame the product for any such “foreseeable” behavior.

Our desire to avoid further such incidents, and to avoid criticism that may accompany perceived inaction, leads us to want to act fast to address such risks. Unfortunately, there can be trade-offs between speed and care that can impact the quality of the information available, and, therefore, the understanding of alternatives, and potential impacts of the desired “fix.” However, speed is often identified with doing the job well, whereas deliberation and care might be perceived as inaction or—on the part of industry—stalling.

Whether ultimately speedy action eliminates the risk as intended may not be as psychologically important as the attempt to do so. As a result, I turned from my musings over how we got here to how to deal with it. Do steel water bottles need to present the risk of injury they do? Could they be “fixed” in some way?

Would a warning to children not to stick their tongues in water bottles make any difference? Some have questioned recently whether Consumer Product Safety Commission (CPSC) has determined effectively that warnings do not work and products must always be changed or recalled if children are at risk. Further, at this time, we likely do not have sufficient data to determine whether keeping these water bottles away from certain sizes or ages of children would work. So, our ability to provide instructions to reduce risk would likely be limited to: “Warning: do not stick tongue into bottle. Tongue might become stuck and require surgery to remove.” At any rate, we can assume that at least some of the risk would not be eliminated by a warning, even one based on appropriate age of user data.

The Today Show story included a possible fix: putting a straw type top on the water bottles. Of course, that “fix” might not totally eliminate the risk either. The bottle user may not be satisfied with the flow rate and might remove the top to get more water more quickly. Is that residual risk “acceptable” or is any possibility of tongue entrapment leading to surgery unacceptable?

Are there other design alternatives? A little research showed that companies working with the CPSC have done many recalls of water bottles with pop-up tops that can be chewed on or for other reasons come off and potentially cause a choking hazard. (Pause to wonder: Are there more tongue capturing steel bottles because manufacturers of bottles looked at previous recalls and decided that they did not want to deal with such a problem?)

Obviously, we can come up with several design alternatives that might eliminate this particular risk in a few minutes. The question for the regulator is whether in eliminating the risk before our face (no pun intended) we are ignoring or creating other risks that—given more time and information—we could have addressed. Unfortunately, we are all smarter about perceiving low probability events relating to unusual usage after they have occurred. (Of course, with 20-20 hindsight such events are always “foreseeable.”) Even in products like drugs that are tested extensively for safety and effectiveness before marketing, safety effects may not show up until after there are millions of users and lower probability events occur in use.

Should steel water bottles be totally eliminated because no instructions or design changes totally eliminate the risk or we are uncertain about such changes? While people might see that question as absurd, they do so because they can perceive that there are trade-offs of utility versus risk. They may also see the possibility of improvements in existing bottles that might eliminate the risk. (Some might even believe that the risk in question is very low and not worth addressing. We might call this the “stuff happens” view.) However, there are likely those who would take a more simple view of safety, particularly in the context of considering a possible recall: steel water bottles present a serious risk, innocent children are injures, and steel water bottles should, therefore, be eliminated. Period.

Finally, I wondered whether rulemaking or a recall is the tool a regulator would use to address the steel water bottle risk. Which is more important, speed, or care? Does it matter whether you are addressing just one water bottle manufacturer who has had an incident, or if you are addressing water bottles across the board? (Of course, if there are other manufacturers with identical designs, should you be addressing just the one with a known injury?) Does it matter whether you are solely addressing the risk that is currently before you, or trying to make water bottles and the public safer overall?

CPSC often responds to even a small number of incidents by recalling a product, sometimes to ban or regulate it later. While the intention of such efforts is a good one the results could be perverse. As noted above, addressing one risk might create others. Or in the alternative, over time it may prove that the quick fix did not get the job done, and more time could go into figuring that out and ultimately taking action then a more careful initial approach.

In addition, product utility might be affected even in addressing a very low level risk. Should there be any consideration of the utility of a product feature that presents a risk? Does the limited examination of a product in the context of a recall lead to the best consideration of such issues? Does it matter whether the alternatives consumers are left with present risks of their own? Or from a policy point of view, does spending effort on minimal risks detract from an agency’s ability to be effective against more serious problems.

CPSC rulemaking is designed to obtain input from industry and to ensure some consideration of solutions, costs, and benefits before action. However, this process can take a long time and use a lot of agency resources. So the agency may ask itself whether the risk is high enough to merit the speed inherent in other solutions such as recalls? Would a recall provide protection with the expenditure of fewer resources? Likely a regulator should also wonder whether the technical issues are sufficiently understood to yield a good result from a recall or whether more consideration, input, and study is appropriate. Finally, they could ask whether the rulemaking or a “quick fix” allow proper considerations of the trade-offs inherent in “fixing” the safety problem.

Ultimately, as regulators of water bottles or any other product, we should want to know whether our good intentions lead to good results. Does the approach we take really make people safer, or does it merely allow us to say “we tried” while we have little effect, or a negative effect on risk? Regulators can usually see when their reluctance to regulate has led to continued risk and seek to regulate further. (CPSC has taken this approach several times in recent years.) However, they should be able to know and admit when they acted too fast with too little information, went too far, or created new risks in their desire to make things safer.

A “good regulatory decision” is one that enhances safety. It requires consideration of many complex factors including the trade-offs inherent in regulating something. However, sometimes good regulation gets lost in the din of voices from those who oversimplify safety decisions into morality plays. While it may seem that the world is full of those who advocate for “safety” allied against “heartless baby killers”, this kind of simplistic presentation of issues often leads to poor regulatory choices and in some cases likely decreases, rather than enhances, safety. At least this is the “message in the bottle” that I read.

Eric Stone is the principal in the Law Firm of Eric Stone, LLP. His practice includes health and safety regulatory law before CPSC, NHTSA and other state and federal agencies. He can be reached at ericstone@ericlstonelaw.com, phone: 301-424-0270, website: www.ericlstonelaw.com.