False positives and zombie lice!

A response to commenters on head lice treatments

A few months back, I did a series of posts in which I looked at what the available scientific literature had to say about head louse treatments. Two treatments which I discussed in detail, Quit Nits and Licefreee had not undergone FDA clinical trials and weren’t approved by the FDA. These products are apparently marketed under a loophole that exempts homeopathic medications from clinical trials required for FDA approval for other medications.
For many of the treatments sold in stores, there is nothing in terms of product names or active ingredients in scientific literature.  (You can find exactly how I conducted my search for evidence here.) Many manufacturers have internally run tests on their websites, and I discussed these in as much detail as I could. Unfortunately it seems I missed an internal technical report from Licefreee which was kindly pointed out to me in the comments on Do OTC Head Louse Treatments Work? Part 2. There was a lot of good discussion, and commenters seemed sincere to me. I’ve got no problem admitting there was information which I wasn’t aware of, and I see no reason why I shouldn’t look at new information when it arises to see if there’s any reason for me to correct my stance.
The commenters seemed interested in a discussion of what proving the efficacy of this treatment would involve, and I’m only happy to oblige. Additionally, I will discuss the possibility for false positives in this sort of research.

How do we know lice are dead?

With this sort of research, one of the potential false positives is demonstrating whether an insect is really dead. For humans, this isn’t hard. We have EEG machines to measure brain activity, and the fact that our blood is our main oxygen carrier requires our heart be constantly beating for us to stay alive. Thus, if there’s no measureable brain activity and if our heart’s not beating then it’s a pretty safe bet that we’re dead (most of the time). This, however, is not the case for insects. Insects, especially small insects, have a completely different way of breathing than humans. Insects transport oxygen directly to tissues via a series of hollow tubes or direct diffusion in smaller insects. CO2 is eliminated directly through the cuticle. This means that the heartbeat is much less important to their physiology than it is to ours. So one of the biggest problems in this sort of research is really determining whether the lice are dead or not.
Hukelbach et. al did something similar to what I did in my other post. They took a random assortment of products and set out to experimentally determine whether they cause mortality. However, during their literature review (basically what I did in my last post), they ran into a problem with how to judge whether lice were dead. We don’t have mini-EEG machines for lice, Hukelbach et. al used a really, really strict definition of ‘dead’ which basically throws the kitchen sink at the problem.:

A lot of head louse research uses similar definitions of ‘dead’, but much of this research simply isn’t very stringent when it comes to counting whether lice are deceased. For example, many researchers merely counted nonmoving lice as dead but few if any monitored the lice for recovery. In this paper, Hukelbach and his colleagues were interested in how many louse treatments resulted in false positives.
A lot of this research involves sitting over a dissecting microscope. Lice are small insects, but with a dissecting microscope things like gut or heart movement are pretty easy to see on living lice. The gut is this big, red stripe in the middle of the louse which is visible thanks to their translucent exoskeleton and bright red diet. Peristalsis in the gut occurs at a pretty constant rate in active lice. Lice also tend to be pretty active off the host, and most insects can be coaxed to crawl by prodding them with something or blowing on them. Gut peristalsis will occur in inactive but living individuals, while movement occurs in active individuals. In short, this is a good criterion which measures characteristics that are both voluntary and involuntary.
‘Some vital signs’ was defined if the lice were having trouble moving or righting themselves. In this instance, the pesticide clearly harmed the lice but the pesticide didn’t pack enough punch to kill the lice outright.
I would have liked to see the researchers look at the vital signs of the lice and statistically support what vital signs were indicative of lice which didn’t recover. Will lice which don’t have gut peristalsis recover? What about heart motion? This isn’t really a major objection, though… it’s something which is really more of a curiosity thing and isn’t a fatal flaw.

Overestimation and underestimation of louse mortality

I copied and slightly modified figure 1 from the Hukelbach paper. Opening the figure in another window may be a good idea while reading this section, as I will be discussing some of the columns in detail.
The paper is far from perfect, but does raise some interesting questions about how stringent in-vitro tests must be to judge lice as dead. There are some problems with this paper that I’d like to get out of the way before the discussion. For instance, I would have liked to see the experiment repeated more so there would be some measure of variation for the times the lice stayed ‘knocked out’. Each of the bars on the graph counts as 25 lice, but there’s no measure of variation. Perhaps there were problems with keeping the lab supplied with lice, as populations can be variable. These problems aren’t minor, because with any sample there will be variation. If you’re doing something like this, it’s good to know where on the bell curve each of the results lies. Despite this I think this paper is good enough to raise some questions about stringency when scoring lice, although I wouldn’t draw any hard and fast conclusions about how long each of these products render lice immobile.
But even despite some fairly major problems, I don’t think this paper is completely junk. If I did, I wouldn’t be discussing it! I think they have some good, if imperfect results. Their experiments were run by submerging the lice in the supposedly insecticidal products for 20 minutes. Afterwards, the lice were observed at 30, 60 and 180 minutes to see what they had recovered. The negative control groups were untreated. The other paper I’m citing by the same group demonstrates that submersion in water has little effect on lice.
As a positive control, they used a product containing 1% permethrin that is marketed under the brand name Quellada which has undergone clinical trials and has been shown to be effective in eliminating head louse infestations in people. If you look at the 30 minute timepoint, you can see that the product has an apparrent efficacy of 40%, which raises up to about 80% by 180 minutes. If mortality was observed 30 minutes after treatment, this would have resulted in an underestimation of the product’s efficacy by claiming the product was half as effective as it was. On the converse, if one looks at the Neutralice column one will immediately see that the product’s apparent efficacy falls from 80% to under 20% as time goes on. If the results were interpreted at 30 minutes, this would have resulted in an efficacy estimate of four times what the product should have been estimated at. So the timepoint at which you observe mortality can be pretty important.
There is a third thing that I’d like to draw attention to, however. Lice Blaster, Neutralice and Tea Tree Oil all contain the same active ingredient, an ethanol distillation of the leaves of Melaleuca alternifolia which is known as Tea Tree Oil. This is used in malathion treatments as a putative synergist for the organophosphate pesticide. Lice Blaster contains 2% of this ingredient by volume (I assume), Neutralice contains 10% of this active ingredient and the Tea Tree gel contains 5% of this ingredient. The three products have no correlation in efficacy, which puzzles me. Is this a result of variation in the plants they obtained the extract from? Does the extract not kill lice? Could variation between samples be responsible for this? These questions remain to be answered to my satisfaction, and could have easily been answered with a better performed experiment.
All in all, the take-home message is that it’s possible for lice to seem dead soon after being immersed in water and that a recovery period is vital after scoring lice.

Criticism of Rossignol, 1999

Rossignol, 1999 can be viewed on the Teclabs website here. I am reproducing the methods section below with the intent of criticism under the Fair Use Clause of the US copyright law. Again, keeping this open in another browser may be a good idea. I am not copying the introduction or discussion, although they can both currently be viewed at the link. I am discussing the methods and data of this paper only, but I invite folks to discuss other portions of the paper in the comments.

According to the paper, Rossignol took two groups of lice and submerged them in either the product in question or in water. To determine if the lice were dead, he monitored gut peristalsis. He then monitored the lice until gut peristalsis stopped. When gut peristalsis stopped, he scored the lice as dead. There are some things in the paper which aren’t very clear.
After describing his criteria for death, he switches to ‘movement’ when discussing this in the results section. ‘Movement’ is a more vague term, and could mean the movement of the gut or heart, or movement of the lice themselves off host. Should I simply assume he’s talking about gut movement and not louse movement as we’d see in active lice taken off the host? He’s not very clear on this in the interpretation. Given the content of his study, I think we should give him the benefit of the doubt and assume that he’s being consistent in his criteria.
Submersion in water is used as a control. As I said earlier, I would prefer to see a vehicle control to determine if 10% salt was the real pediculicidal ingredient. There are other ingredients, like benzyl alcohol, which could be potentially pediculicidal. Since this is a test of the product and not salt in particular, I’ll let this slide.
Because the amounts of water control or the putative louse treatment aren’t listed, there’s no way to know how much of each was used so I can’t truly tell if the treatments were really consistent with one another. I suppose we can look past the possibility of inconsistent treatment during submersion because the lice just need to be covered in the treatment, but there are still some major issues with the time of submersion itself. He simply submerged the lice until he observed something which was a marker used for death at the time. The control group was submerged for more than half a day, while the Licefreee group was submerged for less than half an hour. He eventually got what he was looking for, but at no point did Rossignol indicate he observed the lice for recovery. This begs the obvious question: did the lice recover after this treatment? As we saw in the Hukelbach paper, this is something which happens.

Are there any other tests involving salt water in the literature which could indicate whether the lice are dead?

This is the fatal flaw in this technical report. They didn’t look to see if the lice would recover. As we saw in Hukelbach, 2008, lice can become inactive and apparently dead for a surprising amount of time. The fact that lice can survive submersion in conditions like this shouldn’t be a surprise. I shower once or twice a day, depending if I’ve hit the gym. While this is fairly routine, one should try to imagine what it’s like for a louse. This would be roughly the equivalent of a daily, caustic tsunami washing through the habitat. Yet, they continue to persist.
Even in the peer-reviewed literature this has caused some problems in interpretation of dead lice because there’s no great standard. Because there was no measurement of louse recovery, there’s no way to know if these lice were truly dead and thus no way to know whether they’d recover enough to plague some poor kid. In other words, this paper only shows that the lice go dormant soon after immersion in the Licefree product. By not monitoring for recovery, they have not proven this product kills lice. By not running a clinical trial, they have not shown they can eliminate lice in their natural environment.
The science behind the marketing is simple: lice go dormant soon after immersion in water as a coping mechanism to survive flooding of their environment. Most people don’t have an intimate knowledge of insect physiology and are completely unaware of this trick. Hukelbach’s group has done quite a bit of work documenting this effect, showing that lice can recover from a dormant state after immersion in water containing high amounts of salt or chlorine with no ill effects.

You can see this at work in the video above. According to their website, the spray is 1/100 salt by volume. The hair gel, as mentioned previously is 10% salt by volume.The same group which performed the work showing that lice can recover has also investigated whether salt water can kill lice while looking at whether seawater would kill lice. They looked at the louse killing potential of 12% or 25% weight/volume salt water solutions. According to the paper, all lice recovered shortly after being taken out of the solution. The sample sizes for these tests were much smaller and done differently than in the Licefreee label because a comparison wasn’t the intention of the paper. While the test was only 10 lice for each of the relevant treatments, they exhibited complete recovery after 20 minutes of submersion. If salt water was fatal to the lice at a 1% concentration within three minutes, we would expect complete mortality after 20 minutes of submersion in a 25% solution with no recovery. Instead, contrary to the claims of the Licefreee product, we see a short period of dormancy with complete recovery.
Between this and the anecdotal accounts of tests described in the conference proceedings mentioned in the comments, I think there is more than enough reason to say there is no convincing evidence this product works.
This is an in-vitro trial, similar to Hukelbach 2008. At best, it would only prove the product in question would kill lice under ideal laboratory conditions. To show that a product can cure lice, clinical trials must be conducted which measure the product’s effectiveness under real-world conditions. In-vitro tests are needed to show that the product has a rational basis for a clinical trail, but do not prove that the product kills lice on the patient’s head well enough to eliminate an infestation. Clinical trials are also required to show that there are no adverse effects on the patient, such as skin irritation. This happens occasionally with Tea Tree Oil.
To be fair, this methodology may have been the standard practice at the time for in vitro trials. Even if this is the case, that doesn’t change the fact the test was imperfect and prone to false positives because they didn’t demonstrate the lice would recover from the treatment. Without a clinical trial you’ve still only got in vitro evidence which isn’t sufficient to say that any product would work in a real system. There are many good reasons for any pharmaceutical to go through clinical trials and the loopholes which allow products to go to market with evidence as scant as we’ve seen in these two posts should be closed.
Edit: 6/12/2012 10:07 PM
I just became aware of this but it appears that as of January 17th 2012 TecLabs has completed a clinical trial of Licefreee! spray, the product discussed in this article. At the time of this posting, however, there have been no results reported. The data generated from these studies takes time to compile, crunch and generally go through in a rigorous manner. With a bit of luck, I would ideally like to see the data generated from this study laid out for all the world to see in a peer-reviewed journal article.
I remain skeptical of 1% NaCl as a louse treatment for the reasons I discussed in these two posts, and I think 13 years (half the time I’ve been alive) is a bit late in the game for this to be happening. Nevertheless, I laud and applaud the company for taking the steps to apply rigorous science to the products they sell. These things are a bit of a risk, as the results may require a bit more investment in R&D to develop or refine the recipe if needed. However, this is ultimately a small price to pay to add another proven weapon to our antilouse arsenal.
Canyon D, & Speare R (2007). Do head lice spread in swimming pools? International journal of dermatology, 46 (11), 1211-3 PMID: 17988347
Heukelbach J, Canyon DV, Oliveira FA, Muller R, & Speare R (2008). In vitro efficacy of over-the-counter botanical pediculicides against the head louse Pediculus humanus var capitis based on a stringent standard for mortality assessment. Medical and veterinary entomology, 22 (3), 264-72 PMID: 18816275