“When you have eliminated the impossible, whatever remains, however improbable, must be the truth”
Ali S, et al., Efficacy of two hydrogen peroxide vapour aerial decontamination systems for enhanced disinfection of meticillin-resistant Staphylococcus aureus, Klebsiella pneumoniae and Clostridium difficile in single isolation rooms, Journal of
Hospital Infection (2016), http://dx.doi.org/10.1016/j.jhin.2016.01.016
I understand from contacts in Australia, New Zealand and elsewhere, that Rick Fentiman, Hygiene Solutions MD has been on a world tour, using the above paper to refute the mounting tide of concern as to the Deprox process. I will demonstrate that the tests at UCLH on which this paper is based were corrupted by a deliberate and calculated misrepresentation on the part of Mr Fentiman.
I have no qualifications in microbiology, genuine or otherwise. No doubt it appears very presumptuous for me to question a published, peer reviewed paper by a highly qualified and experienced team of microbiologists – however, my qualification to comment on the paper is not based on any particular expertise in any scientific discipline, it is that having worked for three months at Hygiene Solutions as a consultant on the Deprox process, I know in detail how the machine works, and what its limitations are.
In that period I had full access to the technical and in-house testing data for the machine, and conducted tests personally on the output and H2O2 concentrations generated by the process, which confirmed the previous test data on record. I also became well acquainted with the directors of the business, their technical capabilities, their business processes and their sales and marketing strategy.
I have no doubt whatever that the conclusion of this paper, that a 4.9% and a 30% VHP process will produce identical decimal reduction rates in a variety of pathogens is wrong. It is wrong because of the unverified premise that the Deprox unit was operating with 4.9% hydrogen peroxide solution.
The paper clearly states “ the aerial concentrations of HPV achieved by each system were not measured in this evaluation,” There is no reference to the concentration of the aqueous solutions being tested, and I think we can reasonably assume that this was not done.(1) I am sure that the experimental procedure, methodology, sample preparation and incubation, and the calculation of the tables of log reduction data are all correct and done to the highest standards – all this effort is rendered worthless however if the concentration of the working fluids are not 4.9% and 30% as claimed. The conclusions and recommendations of the paper are not presented as a comparison of Deprox versus Clarus, they are presented as a comparison of a 4.9% system versus a 30% system.
The authors of the paper then (and this is the only criticism I would make) trusted the system manufacturers to give an honest and accurate statement of the composition of their H2O2 based disinfectants. As this was an absolutely fundamental parameter, I respectfully submit that it should have been measured.
We have to accept that either;
- Some novel phenomenon is at work that either negates Henry’s law of partition (2), or the fundamental dose/response relationship (3) shown by living organisms to toxins, such that a six fold dilution of the active ingredient has no measurable effect on the efficacy of the disinfectant, or –
- One of the equipment manufacturers has deliberately misrepresented the composition of their disinfectant solutions.
Both these explanations are improbable. However, new information has come to light that was unknown to the authors of the paper at the time. This independently verifiable data shows that the Deprox manufacturer, and specifically the Hygiene Solutions director, Rick Fentiman, who was present at the tests, has a history of misrepresenting and falsifying the capabilities of the Deprox system, the composition of the Deproxin disinfectant, and his own educational qualifications. There is also very compelling evidence that he was involved in the submission of a “doctored” paper that was published in the Journal of Hospital Infection in 2012, in which results were claimed for Deprox, where in fact the system tested was the Glosair 400 from ASP (Johnson & Johnson). – see this email chain.
Much of the information referred to below has already been published on this site, but will bear brief repetition:
- Rick Fentiman (real name Warrick Fentiman), who is both Managing Director and Technical Manager of Hygiene Solutions Ltd, the Deprox manufacturer, publicly claims to be a “Cambridge graduate microbiologist” but in fact has no such qualifications, either from Cambridge University (4) or elsewhere. (Indeed, it may have been this apparent “qualification” that made S. Ali et al. decide that it was superfluous to test the disinfectant concentrations.)
- The log6 claim emphatically and repeatedly made for the Deprox system was (until the publication of the disputed UCLH paper) solely based on a test of an IC-4 machine, manufactured by Infection Control BV, which was running at vapour emission levels (ΔRH40) that the Deprox is not even capable of being set to. (5) Specifically, the Deprox units are set to give between 10% and 50% of the aerial H2O2 concentrations that the IC-4 machine produced. (ΔRH5 to ΔRH20)
- The composition of the Deproxin proprietary H2O2 solution that the Deprox normally uses was originally given correctly as 4.9% H2O2 and 0.2% Silver Nitrate. When it became evident that silver nitrate was unsafe for fogging applications, the composition data as given on the bottle label and MSDS information was changed to 4.9% H2O2 and 0.2% Silver. (The CAS number for silver nitrate, CAS: 7761-88-8 was still retained) (6). This avoided awkward questions about toxic silver nitrate, and gave a plausible explanation for the grey deposits that the process left in treated areas. It was suggested to customers that the antimicrobial properties of metallic silver particles were an additional benefit of the process! In fact, of course, the composition never changed, and it is easily verifiable from published data that the addition of finely divided silver to a H2O2 solution will catalyse immediate decomposition of the peroxide (7) – the “silver” claim is both untrue and impossible.
Further to this, there were some aspects of the way the UCLH tests were conducted that in the circumstances, are very suggestive:
At the time of the tests, Hygiene Solutions had a contract with UCLH, (8) and had 4 Deprox machines permanently on site, which were operated daily by Hygiene Solutions employees. UCLH had a definite rule that the equipment was not to be operated by their staff – hence no UCLH staff were trained in the use of the equipment.
The paper states: “The HPS1 unit was operated by a trained engineer (Bioquell), while the HPS2 module was operated by hospital staff following training by a dedicated member of the issuing manufacturer (i.e. Hygiene Solutions).”
I have two independent witnesses that the two Deprox machines were used in the trial were not the machines already on site, but were specially prepared at the Kings Lynn depot, with all new piezo discs and calatytic deactivation media.(9) The machines were driven up to UCLH personally by Rick Fentiman, who stayed for the duration of the tests then drove the machines back to Lynn. Apparently no other Hygiene Solutions staff were involved. The “dedicated member” therefore was Rick Fentiman, and he instructed and supervised some unidentified member of the “hospital staff” in the operation of the equipment “on the spot”. (10)
The paper says: “However, during this study, both parties provided storage of equipment and hydrogen peroxide stock solutions off-site.” In the case of the Deprox units, this was the large van in which they were transported.(14) Clearly then there would have been opportunity for Mr Fentiman to have filled the internal storage tank of the test machines with a 30% solution,(11) and disposed of or diluted any residual fluid after the test, without either his own employees or the UCLH staff being aware of the substitution.(12).
It is pertinent that (unknown to the UCLH) the Deprox has a substantial internal storage tank, of about 8 litres capacity.(13) The evaporation unit draws from the bottom of this tank, and the 2 litre Deproxin refill bottles trickle feed in to the top of the tank. I assume that for the sake of authenticity, a genuine Deproxin refill was inserted in to the top of the machine for the tests, hence even if S. Ali et al. had tested the concentration of the fluid, they would have found it to be as stated. As concentrated H2O2 solution is substantially denser than water, a trickle of dilute solution in to the top of the tank would have no significant effect over the course of a few cycles of the machine.
UCLH themselves are now very embarrassed by the paper, and are trying to control the damage by a communication ban. Dr Shanom Ali, the paper’s corresponding author, has been gagged by the UCLH for the past six months – Dr Ali politely tells enquirers that he is not permitted to discuss the matter – and refers the caller to the UCLH Press Office – who are also not permitted to discuss the matter…
Annette Jeanes, Director of Infection Prevention & Control at UCLH and one of the co-authors of the paper, also said she was not permitted to discuss the matter, but did confirm that “persons were investigating whether the tests had been rigged.”
In U.S. criminal law, means, motive, and opportunity is a common summation of the three aspects of a crime that must be established before guilt can be determined in a criminal proceeding. Respectively, they refer to: the ability of the defendant to commit the crime (means), the reason the defendant committed the crime (motive), and whether the defendant had the chance to commit the crime (opportunity).
The means and opportunity are easily established. The particular transport and storage arrangements, and the surprising fact that the busy MD of a company with 60+ employees undertook the whole project personally and without the assistance of the employees or equipment already on site at UCLH, gave the requisite time and privacy.
Motive is less obvious, as while the outcome of the test was obviously commercially very valuable to Hygiene Solutions, it is also true that had the ruse been discovered, it would (and may still) do such reputational damage to the company that it would be forced to discontinue the product.(The VW emissions fraud provides an interesting parallel – cleverly rigged tests to gain an advantage over the competition, followed by scandal and public humiliation.)
What then would have induced Mr Fentiman to take such a risk? Why undertake the tests in the first place? There were a number of factors that forced Hygiene Solutions in to commissioning the tests, and that made it imperative that the results vindicated their log 6 claim.
By late 2015 the lucrative £18,000/month (8) UCLH contract was in deep trouble. Since the Deprox contract started, the C. difficile rates in the hospital had soared, with around 50 extra cases per year, as compared to a national trend that was declining or flat. Exacerbating this embarrassing fact, according to UCLH staff, spot checks on Deprox treated rooms revealed persistent C. difficile contamination. The risk was not only that of losing the contract – had the Deprox failed the test, Hygiene Solutions could have been held legally liable for the additional 100 or so C. difficile cases that had occurred since the contract started.
By the end of 2015 Prof. Peter Wilson and his colleagues had lost confidence in the Deprox HPV process and were trialing three different UV light based systems as a replacement. In the December 2015 UCLH newsletter, Prof Wilson and team gave an enthusiastic endorsement of UV as a much better alternative than HPV.
“It looks like something out of Star Trek,” said consultant Professor Peter Wilson, “but is also straight forward, safe and easy to use. Our trial research shows it can kill the majority of bacteria in just half an hour, compared to three hours using Hydrogen Peroxide Vapour (HPV).” UV light can also be used in bay and ward areas, unlike HPV which is only suitable when used in a sealed room. “Ultra violet has the added benefit of being far quicker and more flexible. Clostridium difficile forms spores, which are extremely resistant to ordinary disinfectants and normal cleaning methods so this is an exciting development.” Three models have undergone trials and a decision on whether to introduce UV light to UCLH is likely to be made in the next few months.
As events turned out, Hygiene Solutions did lose the contract to an alternative and much cheaper UV based system – but they successfully avoided any implication of liability for the unexplained epidemic of C. difficile that accompanied their tenure at the hospital.
Means, motive and opportunity are this all clearly established. But is it “beyond reasonable doubt” that the crime took place?
Ultimately, everything hangs on this question:
Was the H2O2 concentration in the trialled Deprox machines 4.9% or 30%?
What is the evidence that it was 4.9%?
The only evidence for this is the word of Rick Fentiman, self proclaimed Cambridge graduate microbiologist who has an extended history of fraudulent misrepresentation in respect of the Deprox product.
What is the evidence that it was 30%?
The log reductions achieved by the two machines were extremely closely matched across the whole gamut of organisms and simulated soiling levels. The only plausible explanation is that the aerial H2O2 vapour concentrations were also extremely closely matched. Otherwise we have to construct a complex artificial scenario where a mysterious “factor X” just happens to compensate exactly for the much lower aerial H2O2 concentration from the Deprox. Statistically very improbable, and there are no viable candidates for “factor X”. Experience, intuition and common sense tell us that the same concentration of chemicals will produce the same results, all other things being equal. Therefore, obviously, identical results indicate that the same concentration of chemical was present.
Henry’s law (also called the “air/water partition coefficient”) assumes a situation where there is an enclosed space containing both a dilute solution of some chemical in water, and a “headspace” of air above it. The simplest example is a half empty bottle of wine, with the cork in. In this situation, some of the dissolved alcohol will evaporate in to the air of the headspace until it reaches equilibrium – i.e. the molecules of ethanol are being redissolved in the wine at the same rate as they are evaporating from it. Once this point is reached, the proportion of alcohol in the air and the proportion in the wine stay constant. Henry’s law says that the concentration of alcohol in the air is proportional to the concentration of alcohol in the water. So, for example, a wine that is 20% alcohol, might have an equilibrium alcohol vapour concentration in the air above it of 2%. Brandy, which is 40% alcohol might have a vapour concentration of 4%. A 5% beer might have a vapour concentration of 0.5%. If a bottle is opened and some vapor escapes and is replaced by fresh air, more alcohol will evaporate until equilibrium is restored. Equally, if alcohol vapour was added to the air in the headspace, some of it would be dissolved until the concentration was again proportional to the concentration in the liquid.
This is all intuitive, and common sense. This is how breathalyzers work – because the concentration of alcohol in exhaled breath is consistently and repeatably proportional to the concentration of alcohol dissolved in the blood despite all the physiological variations between different individuals.(2)
This is why brandy has a very strong smell of alcohol, and beer has a weak smell of alcohol. This is why we can burn brandy on a christmas pudding, because the air immediately over the brandy has more than 3.3% alcohol vapour in it, which makes it a flammable fuel/air mixture. Beer will never catch fire, because the proportion of alcohol vapour in the air never gets close to the critical 3.3% which (for ethanol) is the lowest fuel/air ratio that can be ignited. The point is this – it makes no difference how much beer you pour on your christmas pudding – it is still 5% solution and the air above will still be only 0.5% vapour.
So if Mr Fentiman was to demonstrate to S. Ali et al a novel kind of beer he had invented, and he insisted that the only ingredients were 5% ethanol in water, and he was to demonstrate that he could pour his beer over a christmas pudding and set fire to it, would S. Ali et al write a paper, endorsing this newly discovered phenomonon, without expressing any curiosity as to how this could be possible? Obviously not. They would know before they had even done any calculations, that the data presented to them was not correct.
As with alcohol in beer, so with H2O2 in Deproxin. Regardless of the means by which it is vapourised (which is simply a way to increase the surface area, so as to make it reach equilibrium faster) the maximum sustainable aerial vapour concentration is governed by the concentration of the original liquid. Broadly speaking, for a 4.9% solution, this limits the aerial concentration to about 40ppm. A system using a 30% solution can achieve about 6 times this, i.e. about 240ppm – and indeed, this is the level of aerial H2O2 concentration that the Bioquell process uses in order to achieve a log6 reduction in 90 minutes.
There is an additional important limitation to low concentration systems. In order to evaporate 5 units of H2O2, they must evaporate 95 units of water. Once the humidity in the room reaches 100%, the fluid will no longer evaporate, and the H2O2 can no longer be released. Thus in practice, the 40ppm H2O2 level is only attainable in conditions of low starting humidity, i.e less than 50%. In the UK, water vapour saturation is almost always the limiting factor.
This is why all 5% systems are limited to a maximum efficacy of log 4, (i.e one hundredth of the efficacy claimed by S. Ali et al.) as every other published paper on these systems will demonstrate.
“The results of this study come as a big surprise to me. As the authors acknowledge, previous comparative evaluations of a high-concentration (30%) hydrogen peroxide system (Bioquell) with a low-concentration (5%) hydrogen peroxide system (ASP Glosair) by Fu et al., Holmdahl et al. and Beswick et al. have identified stark differences in biocidal capacity between the two systems. For example, Fu et al. found that the low-concentration system achieved a A. baumannii, and made little impact on MRSA when dried in 3% or 10% BSA. Both Fu et al. and Holmdal et al. found that the low-concentration system was consistently unable to kill 6-log spore biological indicators. Beswick et al. found that the low-concentration hydrogen peroxide system was, I think it’s fair to say, dramatically ineffective against C. difficile spores and Mycobacterium fortuitum. To add to this, I have my own experience to relate. As a follow-up to this study, we performed the same experiments but using 5% hydrogen peroxide rather than 30%. The results were very poor indeed using 5% hydrogen peroxide, with limited log reductions achieve on the pathogens studied, which included MRSA, A. baumannii and C. difficile spores.”
- The aerial concentration of H2O2 was the same for both machines tested, and was in the region of 240ppm.
- It is physically impossible to generate anything like this aerial H2O2 concentration from a 4.9% solution as it will reach equilibrium at about 40ppm.
Both the Deprox and the Bioquell machine were using the same concentration of H2O2 solution, which was in the order of 30%.
Specialist Hygiene Solutions Deprox salesman Tom Lister was given opportunity to explain the surprising results of the UCLH tests, when I met him at the IPS 2016 Conference in Harrogate. Here is his explanation…
There are some inexplicable statements in the paper in this respect:
“…while the HPS2 module was operated by hospital staff following training by a dedicated
member of the issuing manufacturer (i.e. Hygiene Solutions).The generator module recorded the concentration of hydrogen peroxide produced, room temperature and relative humidity during the decontamination cycle.”
“Upon completion of the decontamination cycle, hydrogen peroxide in the room was purged to a safe level (>1 ppm; monitored by the hydrogen peroxide control unit) before the test room was deemed safe to enter.”
The Deprox unit has no such monitor, and has no means whatever of measuring H2O2 concentration. Also, if the “generator recorded the concentration” as claimed, it would surely be helpful to publish this data…
For neutral compounds, at dilute solution concentrations in pure water, the air-water distribution ratio is referred to as the Henry’s Law constant KH (or Kaw). For real aqueous solutions (i.e., solutions that contain many other chemical species), we use the term “air-water distribution ratio” which, for practical purposes, we approximate by the Henry’s Law constant. The Henry’s Law constant KH can be approximated as the ratio of a compound’s abundance in the gas phase to that in the aqueous phase at equilibrium. This is nicely demonstrated by the blood alcohol v. breath alcohol graphs below.
The dose–response relationship, or exposure–response relationship, describes the change in effect on an organism caused by differing levels of exposure (or doses) to a stressor (usually a chemical) after a certain exposure time. This may apply to individuals (e.g.: a small amount has no significant effect, a large amount is fatal), or to populations (e.g.: how many people or organisms are affected at different levels of exposure).
Studying dose response, and developing dose–response models, is central to determining “safe” and “hazardous” levels and dosages for drugs, potential pollutants, and other substances to which humans or other organisms are exposed.
For the tests analysed in the TNO report, the delta RH setting of the machine was 40% (TNO report p.16) The delta RH sets the level of concentration of disinfectant in the air of the room being treated.
The machine being tested by TNO was not a Deprox, it was an IC-4, which is a similar but more powerful machine made by a Dutch company. However, the Deprox is not even capable of being set to delta RH 40% – the highest setting on the Deprox is 30%. In practice, because of insurmountable problems with residual gas at the end of the process, all Deprox units are set to between delta RH 5% and delta RH 15% – consequently the disinfectant concentration is much lower than it should be and the efficacy of the process is between log0.2 and log2. How to verify? Just press and hold the “secret” buttons on the Deprox front panel, and the delta RH setting will be displayed – see How to test your Deprox.
The rate of reaction is proportional to the surface area of the catalyst and the concentration of the solution. The video below shows a 3% H2O2 solution being decomposed by a solid silver bar. The 6% solution used in HPV would be decomposed very rapidly indeed by finely divided “colloidal” silver.
The significance of this is that concentrated H2O2 solution very rapidly attacks the piezo discs – experiments at Hygiene Solutions had shown that 15 cycles at high concentration was the maximum lifetime. I have no doubt that all the discs had to be replaced again after the UCLH test.
Also, the deactivation catalyst has a very short lifecycle, quickly becoming poisoned by the silver nitrate from the Deproxin.
Why didn’t Rick operate the machines himself? This was so he could claim the tests as being fully “independant” – it gave him an alabi.
Hygiene Solutions have a couple of Bioquell machines which they use for internal comparitive testing – no doubt they have stocks of 30% solution on hand for this purpose.
Dr Jon Otter postulates that the Deproxin could have been adulterated with peracetic acid to boost the potency. This I think unlikely as the strong vinegar smell of peracetic acid would have given the game away.
This is a more common occurrence than you would imagine. After lobbying it’s the second largest corrupt practice used to influence government policy and regulators. Personally I’ve directly encountered the practice several times in the Chemicals and materials markets, and at least twice in the environmental remediation market. On one occasion the individual involved perjured himself to the UK Government Advisory Committee on the safety of chemicals. Such was my concern that I saw fit to write about such events in: Dirty Business!: Or… The Strategic Implications of Criminality, Unethical Conduct and Black Economic Practice on Legitimate Business.
Thanks Peter, That is very interesting – I had no idea that it was the #2 corrupt practice. We expect scientific reports to be impartial and accurate, especially in the case of peer-reviewed papers in highly reputable medical journals. In this case the perpetrator has used a completely fictional qualification to gain the trust of the researchers, followed by what can only be described as a conjuring trick to dramatically change the outcome of the tests in his favour. The paper has already been cited and quoted multiple times, so the damage goes on and on… Freedom of speech is basic and essential in keeping science objective and unbiased. I find it very disturbing that the corresponding author of this paper has been gagged for 6 months by the UCLH while the Journal of Hospital Infection continues to uphold and distribute this pseudoscience through its publications. I can only assume that the UCLH and JHI policy is to stick its head in the sand and hope it all blows over. I can assure them that it will not. A full scale HSI investigation is in progress, and only this morning I had an email from a former Deprox engineer who had just been interviewed at length by the lead investigator, Mr Martin Ball.
Richard In my experience, the crux of this form of corruption, and it is corruption, is where you have a technical professional who has a conflict of interest and or commercial motivation to provide information that is false or biased. The conflict of interest is not disclosed. This has occurred frequently in the past and more especially so in technical expert reports to governments, quangos and trade bodies.
My company was adversely impacted by just such an individual who had a commercial interest with one of my competitors, and having never evaluated our technology submitted false information in a technical evaluation to EPRI – which they repeated to one of my prospective clients. Its a very difficult area to close down on, especially where you have professional bodies that don’t like to admit one of their own has gone rogue, and organisations that don’t want to admit they’ve been duped and will do all possible to sweep the issue under the carpet.