A new job: 'Supervising dermatologist' at RIVM


CESES is an acronym for Consumer Exposure Skin Effects and Surveillance. This was a project set up by RIVM (National Institute for Public Health and the Environment; www.rivm.nl) in 2009 to monitor the undesirable effects of cosmetics in the Netherlands. Clients were the Netherlands Food and Consumer Product Safety Authority NVWA (formerly: Keuringsdienst van Waren) and the Ministry of Health, Welfare and Sport. The data obtained from the study were used to monitor the incidence and prevalence of cosmetic side effects and to identify specific products and ingredients responsible for these side effects. The knowledge acquired in this way will then, the aim was, contribute to legislation relating to cosmetics in the European Union. At the start there was a 'public route', in which consumers could report side effects of cosmetic products, and a 'clinical route', with reports by general practitioners and dermatologists. I restrict myself here to reports by dermatologists.

The project

In April 2012 I received a call from a lady working at RIVM, who introduced herself as the project leader of CESES. She had heard my name from Prof. Derk Bruynzeel, who had told her that I am an expert in the field of side effects of cosmetics. She informed me of the CESES project. It was quickly clear to me that the design of the study was as follows: a dermatologist, who participates in the project, diagnoses cosmetic allergy in a patient; the responsible product is sent to the NVWA; the NVWA  tries to acquire the ingredients from the manufacturer or importer; the NVWA produces materials for patch testing and sends them to the dermatologist; the dermatologist patch testst the patient with these materials, hopefully sees one or more positive reactions to the components of the product and then sends the results – digitally, an entire software program had been developed for CESES – to RIVM.

Does this sound familiar to you? It immediately did to me, because it was exactly the same as what the members of the Contact Dermatoses Committee and I had done almost 30 years earlier in the context of my PhD research. It is not surprising that the study design was the same, as this is the correct way to demonstrate the allergens in cosmetics and also because Derk had already very actively participated in my research in the mid-1980s. So I was immediately very enthusiastic, I let it show in our conversation and that went down well with the CESES project leader.

Would you like to be 'Supervising dermatologist'?

The project leader said that a 'supervising dermatologist' had to monitor the entire process. That process started with that the dermatologist had to fill in a - very extensive, as it turned out - digital form with all the patient's details, the product that caused the side effect, type and batch number of the product that had caused the allergy, a detailed description of the side effect with localization, how long the product had already been used and a host of other data, including where the product had been purchased. The cosmetics industry had also participated in the preparation and development of the study and it turned out that they wanted to collect all kinds of data that were not relevant to the actual purpose of the research. Well, I didn't see all that until later.

The ingredients of the cosmetic that had caused an allergic reaction could also be requested on the form. Incidentally, it was also possible to report an allergic reaction to a cosmetic without requesting the ingredients, for example when the patient refused to be tested further, or when the nature of the allergen was already established (or very likely established) based on the results of patch testing the European baseline series or a cosmetic series, in combination with data on the composition of the product. The further course of events: requests for components, receipt of components and feedback of the results of the patch tests with them, was all done via the digital forms. All in all it was a rather complicated system.

Well, the task of the 'supervising dermatologist' was to monitor this process and to ensure that the data entered by the dermatologists and the conclusions drawn were correct. He also had to decide whether the request for the ingredients of the culprit cosmetic product could be approved on the basis of the information provided. The project leader immediately asked me on the phone whether I wanted to take on this task, which Derk had done up to this moment. There was no need for me to apply, my credentials were apparently sufficient. I didn't have to think long about it and said that the project – because of my previous activities in this field and interest in the subject – very much appealed to me and that I certainly would like to become 'supervising dermatologist'. At the age of 61, De Groot got a new job!

Market-based remuneration

At that point, my business instinct came to the surface and I cautiously asked if there was any form of remuneration for the position. Indeed there was, a 'market-based fee'. When I heard the amount per hour worked, my breath caught. My skills were apparently very highly rated! There was one limitation: the 'contractor' (me) could not bill more than a maximum allowed amount per year. That  would not spoil the fun, I immediately decided. Shortly afterwards I went to RIVM, met the entire CESES team and discussed all details with the participants.

From left to right: localized allergic contact eczema on the abdomen, the metal button of the jeans responsible for it, and a positive patch test to nickel. The eczema shows a brownish discoloration. This indicates that it has existed for a while already. Not infrequently, long-standing eczema results in an increase of pigment in the skin (hyperpigmentation), especially in pigmented individuals. This can occur with any inflammatory reaction, eczema is a (sterile) inflammation. The pigmentation is called post-inflammatory hyperpigmentation (post = after; inflammatory = inflammation). Nowadays, the buttons of jeans hardly release any nickel anymore and this type of allergic contact eczema has become rare.

Weak spots

I soon came to realize that there were some weaknesses in CESES and that the goal, which can be roughly described as 'finding important new allergens in cosmetics and doing something about it with EU legislation', would be difficult to achieve. The set-up of the study was marginally  suitable for it and we had far too few participants and reports of cosmetic allergy. Only 5 or 6 academic centers participated, including the VUMC Amsterdam and UMC Groningen, plus 4 dermatologists in private practice. Of 2 of those 4, we did not receive even a single case of cosmetic allergy. In the period October 2015 to October 2017, for example, there were only 90 reports filed by the participants, of which 70 were rated by me as 'probable' or 'very likely' (which included 'proven', but this could not be scored as such according to protocol. The chance that important new allergens will be discovered with such small numbers is of course not great. On the contrary: 63 of these cases (90%) were caused by known allergens that tested positive in the European baseline series or a cosmetic series. I knew from own experience and from literature that at least 10% of patients patch tested by a dermatologist have an allergy to cosmetics; in academic centers the percentage is probably higher. So I was sure that only a fraction of the eligible cases were reported to CESES. Therefore, several times during the period that I was a supervising dermatologist, from July 1, 2012 until the termination of the project in December 2018, I tried to entice participants with email and telephone messages to report more cosmetic allergy cases. Unfortunately, the results were rather limited.

Not enough cases of cosmetic allergy

Why didn't I receive more notifications of cosmetic allergy? It could not be due to the the remuneration for the participants, which was generous. The main reason is probably that the academic colleagues were much too busy with their (other) activities, which I completely understand. Completing the digital forms would take a lot of time and even if an employee of the allergy department were to do the work, it would still have to be checked by the participant her- or himself. A second reason is probably that the added value for the dermatologist and his or her clinic was (correctly) perceived to be limited. And a third possible explanation (but now I may be projecting) is that the participants themselves (at least those from the academic centers) received nothing from the financial compensation, which would go to the clinic.

Testing with the ingredients of cosmetics

Testing with the ingredients of the cosmetics which had caused allergic contact dermatitis also had problems and yielded little result. The NVWA participant, who had to provide the materials for patch testing, did not always perform her task adequately in CESES. I could understand that. In the years before, the NVWA had been forced to implement one reorganization (= downsizing) after the other at the behest of politics, and ever more work had to be done with fewer staff. Well, requesting the ingredients from cosmetic manufacturers or importers and preparing the test materials was a lot of work.  When a request was submitted to the manufacturer to supply materials, but the manufacturer did not respond (which was rule rather than exception), the NVWA would often not act on it and sometimes a reminder was sent only after six months. When those test materials finally arrived at the clinic 9 months after the first patch tests, many patients no longer wanted to be retested and then all work and every effort had been in vain. To improve and streamline this process, I wrote a protocol which was approved by all parties. Yes, well, that only helps if everyone adheres to it......

Cleaning up the data in the database

Over time I have cleaned up the entire database. I came across more than 75 reports that had not yet been completed. I could see that Derk had been working on some of them and, for example, had asked the participant for additional information, but had not received it. It was clear that a central figure was needed who kept a close eye on everything and who had to intervene and keep the participants engaged  if necessary. Derk, he had told me, didn't feel like doing this anymore and he thought I'd be better at it. Many of these reports concerned patients who had been tested with all ingredients of the cosmetic, to which they had previously had a positive reaction, but who now did not respond to any of those ingredients (meaning the all work had been for nothing). This should have been reported to CESES, but it wasn't and the file then remained unfinished in the database with nobody checking up on  it.

Erythematous and mildly scaly eczema in the genital region caused by miconazole in a cream for the treatment of a fungal infection of the skin

A positive patch test to a cosmetic, but a negative reaction to all ingredients can have many causes, but the most frequent cause is probably that the reaction to the cosmetic had been false positive. In these cases, the patch test reaction is interpreted as positive and allergic, but it is in fact caused by irritation and not on allergy, hence a false positive reaction. Such cases in the database were especially patients who had a 'positive' reaction to soap, bubble bath, shower foam and shampoo, which are known to cause these irritant reactions. Requesting the ingredients, which was associated with high costs and a lot of manpower at the NVWA (and at the manufacturer) in such cases is completely useless. Incidentally, this cannot be blamed on the dermatologists concerned: the distinction between an irritation reaction and a weak-positive allergic reaction is often very difficult to make. Yet, I felt we had to reduce such false positive reactions from entering the study.  

Another protocol

To reduce the risk of such situations from happening, I wrote a protocol on how the various cosmetics should be tested (e.g. pure ('as is'), 10% in water or 1% in water) and set stricter criteria for obtaining approval for requesting the ingredients of the cosmetic. In other words, I wanted to be more certain that the 'positive' patch test to the product was really an allergic one before the ingredients could be requested. That greater certainty could be obtained by a second patch test with the cosmetic (which should be positive again) or a positive use test by the patient with the product. After the adoption of the protocol, the number of these types of reports and requests for the ingredients post auto propter fell sharply.

I also arranged that from that moment on, when the ingredients had been obtained and were patch tested, the culprit cosmetic product itself would be tested along for a second time. If no positive reactions were found to the ingredients and the product itself was now negative, then we knew that it had not been an allergic patch test reaction the first time (false positive). When the product did respond positively the second time, but all ingredienst were negative, we knew that something had gone wrong with those components. e.g., testing in too low a concentration, testing in the wrong vehicle, not all ingredients provided by the manufacturer or batch variations. This, by the way, only happened once. 


Allergic contact eczema of the ears, neck and back due to paraphenylenediamine (PPD,  in hair dye. Often, there are fierce allergic reactions to the face with massive oedema of the eyelids

As I expected from the start, the research did not reveal any new important allergens. By far the most identified allergens in the products were fragrance raw materials, methylisothiazolinone (in those years a major cosmetic allergen) or methylisothiazolinone/methylchloroisothiazolinone (Kathon CG) (preservatives), cocamidopropyl betaine (an emulsifier in shampoo) and paraphenylenediamine in hair dye. There were some reactions to (meth)acrylates in artificial nails among workers in nail salons, but there was already quite a lot of literature about this, so what else is new? We also saw quite a few reactions to the UV filter (sunscreen agent) octocrylene. That was reason for the then project leader to ask the EU-commission in Brussels to take actions against this chemical. I advised her against taking this road, because shortly before I had written a review article about octocrylene published in Contact Dermatitis. Although the agent does indeed not infrequently cause allergic reactions, it is also used in a large variety of cosmetics and is an essential substance in combination with another UV filter to block waves from the UVA spectrum of ultraviolet radiation, thereby forming the only available effective UV-A filter.

In other words: yes, octocrylene regularly causes allergies, but few in relation to how much it is used and, moreover, it is an important ingredient. My advice was disregarded and the project leader got zero on her request in Brussels.

Published findings

We have identified one new allergen (capryloyl salicylic acid) and found some (relatively) rare allergens, some of which we have published as case reports, but no really important discoveries have been made in CESES.

  • De Groot AC, Rustemeyer Th, Hissink D, Bakker M. Contact allergy to capryloyl salicylic acid. Contact Dermatitis 2014;71:185-187
  • De Groot AC, Rustemeyer T, Hissink D, de Wit-Bos L. Contactallergie en fotocontact-allergie voor de UV-filter octocrylene. Ned Tijdschr Derm Venereol 2014;24:378-381
  • Jagtman BA, de Groot AC, Bakker M. Allergisch contacteczeem door propolis in een cosmeticum. Het nut van aanvullende epicutane tests met een cosmeticareeks. Ned Tijdschr Derm Venereol 2016;26:339-341
  • De Groot AC, Tupker R, Hissink D, Woutersen M. Allergic contact cheilitis caused by olaflur in toothpaste. Contact Dermatitis 2017;76:61-62
  • De Groot AC, Jagtman BA, Woutersen M. Contact allergy to neem oil. Dermatitis 2017;28:360-362
  • Jagtman BA, de Groot AC, Woutersen M. Allergisch contacteczeem door joodpropynylbuylcarbamaat in een styling wax voor het haar. Ned Tijdschr Derm Venereol 2017;27:403-406
  • Volkering RJ, de Groot AC, Woutersen M, Schuttelaar M-LA. Allergisch contacteczeem door het gebruik van massageolie. Ned Tijdschr Derm Venereol 2019;29:14-17

Allergy to paraphenylenediamine is not always caused by the use of hair dye, but also by henna tattoos. Henna itself rarely causes contact allergy; the allergen in 'henna' is the dye paraphenylenediamine, which is added to the henna solution to make the color more intense and to make the substance easier and faster to apply. In certain cultures, henna tattoos are frequently used in religious ceremonies and festive occasions (photo left). On the photo right a severe allergic contact eczema due to paraphenylenediamine in henna. People who become allergic to the dye in a henna tattoo (actually it is a pseudo-tattoo, because no dye is injected into the skin, but painted on it) will also develop allergic reactions when using hair dye and vice versa.


De Groot is definitely not holier than the Pope

When you've read all this, you may think: 'De Groot, if all of this is true and you have known it more or less from the start, why did you stay on as a supervising dermatologist for 6.5 years?' Well, I had three reasons for it. The first is that I hoped that through all my work for CESES the intended goal could still be achieved or at least that the results could become better and more meaningful. And indeed, I have done a lot to try to take the result to the next level with protocols, lectures for the participants, stimulating approach to the participants, efforts to attract new participants, (small) publications, proposals for improvements to the infrastructure and software, regular consultations with the project leaders and a careful search of outstanding reports in the database. The second reason is that I enjoyed getting articles on new or rare allergens to my name and I knew there would be some. The third reason I stayed all this time, I don't want to deny it and I'm not proud of it, is that the remuneration was generous. I don't think I should feel guilty about it. I cannot be expected to be holier than the Pope and I made it clear to all three project leaders with whom I worked that I had my doubts about the success and usefulness of the whole project. But if it turns out that the train rides on and will keep doing so, be it with or without me, then it can be better with me, for the project and for myself. Capice?

The end

On December 20, 2018, the CESES project ended with a meeting at the RIVM in Bilthoven. We all agreed (yeah, well, I didn't want to spoil the party) that the project, while not perfect, had been useful. We thanked each other for the pleasant cooperation and I went home with two fine bottles of wine.

The latest CESES report can be viewed and downloaded here .

Suggested next chapter: