Quality of patch testing patient's own material in patients with suspected occupational skin diseases throughout Germany: Interim results of the German Social Accident Insurance (DGUV) research project FB 317b

Due to limited availability of commercial test preparations, patch testing patient's own material (POM) is of great importance to diagnose occupational allergic contact dermatitis.


| INTRODUCTION
Allergic contact dermatitis is one of the most common occupational skin diseases (OSD). 1 Identification of the causative allergen by patch testing is pivotal in order to implement targeted preventive measures based on allergen avoidance. 2Therefore, it is essential that the patch test reflects the current exposure to allergens.In Germany, around 269 commercial test allergens for patch testing are approved or marketable test preparations in the approval process, which are currently on an equal footing with test allergens that are actually approved under a transitional regulation. 3,4In practice, however, only around 40% of these commercial test allergens are currently available due to (supply) shortages.On the one hand, this results in fewer and fewer previously available test substances and on the other hand, no new test substances are being developed and marketed.This is especially worrying taking into account that there are around 5200 chemicals known to cause contact allergy for which information on adequate concentrations and application vehicles for patch testing is available. 5,6The current situation results in a vast diagnostic gap to which the lack of knowledge about potential contact allergens newly placed on the market also contributes. 7Further, the constant change in the work environments, which goes hand in hand with growing technical progress, has resulted in the use of new hitherto unknown allergens.
The number of chemical substances used is constantly increasing, with around 12 000 new substances being added to the register of the chemical abstracts service (CAS) every day. 8 order to meet the resulting future challenges successfully in the long term, there is a need to optimize the diagnosis of contact allergies.Accordingly, patch testing with patient's own material (POM) is a highly relevant, however, complex diagnostic measure with great practical importance in the diagnosis of OSD, 8 which contributes to closing the diagnostic gap and is highly important in detecting new allergens and new sources of exposure to known allergens (e.g.isobornyl acrylate (IBOA) in glucose sensors, 9 sodium cocoamphopropionate in an occupationally used skin barrier cream 10 and in occupationally used hand cleansers 11,12 or N-butyl-1,-2-benzisothiazolin-3-one in a cutting fluid). 13The high relevance and additive value of testing POM has already been stressed in several publications over the last decades [14][15][16] ; for such patch testing, the POM must be prepared appropriately, including the selection of a suitable dilution (i.e.test concentration) and application vehicle, 8 in order to receive conclusive test results and to avoid adverse events (e.g.strong skin irritation, active sensitization). 5,8Not only the adequate execution but also the adequate documentation of patch tests with POM is crucial, as this is the only way to understand afterwards what exactly was tested and whether the test result is plausible.In Germany, within the dermatologist's procedure of the German Social Accident Insurance (DGUV), dermatologists are encouraged to patch test POM from workplaces in the case of suspected OSD at the expenses of the statutory employers' liability insurance funds. 17e present study aims at assessing the quality of performance and documentation of patch testing POM in patients with suspected OSD throughout Germany.

| Study design
In this ongoing retrospective-prospective study with continuous systematic processing, evaluation and documentation, a sample of nationwide carried-out and documented patch tests with POM in patients with suspected OSD has been collected.The study was conducted in agreement with the principles expressed in the Declaration of Helsinki.Ethical approval was obtained by the ethics committee at Osnabrück University, Osnabrück, Germany (procedure number 37/2020).This paper presents an interim analyses of the German Social Accident Insurance (DGUV) research project FB 317b for quality management in the diagnostics of occupational contact allergy.

| Data collection
The data collection was started in October 2020 and is to be continued until September 2025.All German social accident insurance institutions covering all occupational sectors were asked to screen their files for past and current cases with patch testing of POM which were billed according to the scale of fees for physicians (UV-GOÄ) No. 379 (i.e.patch tests with POM after prior commissioning by the social accident insurance institution).After identification of cases, anonymization or pseudonymization was done by the accident insurance institutions and test protocols were made available to our department, together with other relevant documentation from the patients' files (e.g.dermatologist's reports).In case of pseudonymization, an additional written informed consent was obtained from the patient; this also counted for ongoing patient cases, but not for closed cases which were anonymized.If patch tests with POM were carried out during a legal assessment commissioned by a statutory accident insurance institution, the complete assessment was sent.If workplace visits or other exposure assessments, prompted by the results of the respective tests, were performed, their reports were also attached.
The anonymized or pseudonymized documents were sent by email to a secure email mailbox (or by mail) to the study centre in Osnabrück, Germany.Patient cases were included in the order in which the data were received.The data were recorded in an electronic database for systematic evaluation.The participating social accident insurance institutions were quarterly invited for 1-hour online meetings in which they were informed about the project's progress and asked about the current status of data transfer as well as any potential problems regarding data transfer to ensure a smooth process.

| Data analyses
Patch test data were analysed referring to the national and international guidelines and recommendations for patch testing.whilst not using the sheet itself) were used, this was counted as using the form.The test protocols are presented to experienced doctors for relevant insurance law issues (e.g.need for prevention efforts, claims for benefits in the case of allergies) and the quality of the testing/test documentation is considered in the assessment.If testing/test documentation is insufficient, further testing will be carried out accordingly.
It was assessed whether the following data were documented: a full description of the POM (e.g.full name of a glove with manufacturer information instead of only stating 'blue glove'), the test concentration (if relevant, not needed for some substances that are tested 'as is', e.g.glove materials, emollients), the test vehicle (if diluted/dispersed, e.g. in aqua or petrolatum), the pH value (if relevant, e.g.metalworking fluids) and the results at the various reading times.
In addition, the suitability of the declared test concentration (i.e.not too high or too low), the declared vehicle (i.e. the correct vehicle for the POM) and the declared pH value (i.e.not too high or too low) were evaluated.Moreover, it was assessed by an experienced occupational dermatologist whether in case of a positive test reaction to the POM sufficient patch testing of corresponding test series 3 | RESULTS

| Patient and data characteristics
For this interim analysis conducted in June 2022, 460 cases of patch testing with POM performed between 2013 and 2021 were included.
They were submitted by 17 social accident insurance institutions (Table S1).Most of them (32.8%) were submitted by the social accident insurance institution for health and welfare services (BGW), followed by the social accident insurance institution for foodstuffs and catering industry (BGN) and the social accident insurance institution for the woodworking and metalworking industries (BGHM) with 20.0% and 10.4% of the cases respectively (Table S1).The included 460 patients (40.7% male, mean age 43.2 ± 13.0 years; Table 1) worked in 60 different professions, with nursing professions being the most frequent with a share of 12.6% (Table S1

| Quality assessment
The results of assessing the quality of test performance and documentation of the patch tests conducted with POM are provided in b If relevant, not needed for some substances that are tested 'as is', for example, glove materials, in the case of undiluted liquid substances (tested 'as is'), at least the use concentration must be documented so that it is clear whether dilution was necessary or not as the use concentration is then the test concentration.
c If diluted/dispersed, for example, in aqua or petrolatum.d If relevant, not needed for some substances that are tested 'as is', for example, glove materials, in the case of, for example, used metalworking fluids in the appropriate usage concentration 'as is', the pH value needs to be determined, while this is not necessary for every diluted substance (e.g.not for soap).
e For example, positive reaction to a metalworking fluid and corresponding patch testing of at least the baseline and the 'metalworking fluid series' of the German Contact Dermatitis Research Group (DKG). f That is, at least two readings have been taken, one after 24 or 48 h and the other more important one after 72 or 96 h respectively.g For example, due to testing of the substance on the skin generally not recommended.
one POM (9 metalworking fluids and 1 dishwashing detergent  2017. 15Notably, positive patch test reactions to skin and hair cleansers, disinfectants (unclassified) and cleaning agents/detergents were frequently considered presumably false positive indicating that in many cases unsuitable test conditions may have led to overestimation of sensitizations to these irritant products and may have caused an unnecessary burden to the patients (e.g.strong irritant reactions).Some substances (e.g.epoxy resin systems) were underrepresented compared to other studies evaluating patch testing of POM, 16 which might be explained by differing approaches and cohorts in respective centres.It should be mentioned that some occupational substances were tested particularly frequently in the present collective (e.g.gloves and skin products).This is not only due to the selection by the predominantly contributing accident insurance providers and the people insured there (healthcare sector, food industry) but also because these products are easy to test (no dilution, no special risk to be considered through testing).False-positive test reactions were most often suspected with cleaning products.One could therefore point out that there is a particular need for training in the testing of irritant POM, which must be adequately prepared/diluted for testing.
Only in a small proportion of patch-tested POMs, the type of the tested substances was considered inadequate for patch testing, for example, highly skin irritating substances as methylated spirits or sanitary cleaners.The above-mentioned various deficiencies not only regard mainly to documentation but also to some extent performance of patch tests with POM lead to the conclusion that, in a high number of cases, the value of the documented test result is limited.Due to this, it cannot be ruled out that a share of the financial resources provided by the social accident insurance institutions might currently be misallocated.
Moreover, it is impossible to estimate the substantial consequential costs that arise from incorrect test results.These include the high costs associated with unnecessary allergen avoidance in the case of false-positive test reactions (e.g.unnecessary replacement of a metalworking fluid) or protracted courses of illness in the case of falsenegative test reactions and thus missing identification of a culprit allergen.
As a limitation of this study, it should be mentioned that not all tests could be allocated to either hospital or practice due to anonymization of the test protocols.This uncertainty, however, can be tolerated as also the total results are displayed which include all test results.Further, missing information on POM (e.g.name, manufacturer, article number, etc.) can also be based on non-cooperation by or incapacity of the patient to provide these details; the testing dermatologist then cannot provide this missing information in the test protocol and cannot be made accountable for this shortcoming.Also, within the framework of this study, only a sample could be examined and not all patch tests were conducted within a certain time period.
Further, the patch tests on which we found incomplete or lacking information (e.g.concentration and vehicle) actually might have been performed with good quality and the dermatologist or whoever filled in the form simply was not motivated to spend time to disclose such detailed information to an insurance institution.This could be solved by linking appropriate documentation to reimbursement of the tests.
It should finally be considered that the conclusions drawn from data obtained from the German social accident insurance institutions are not necessarily directly applicable elsewhere.

| CONCLUSION
The presented interim results indicate an urgent need for action to improve the quality of patch testing POM.Inappropriate patch testing causes not only unnecessary distress to the patients but induces considerable (consequential) costs for the social accident insurance institutions.Improving the quality of both performance and documentation of patch testing POM is therefore of great importance.This finding should result in considerations how to enhance awareness and training of dermatologists.This should be accompanied by actions to tackle the increasing diagnostic gap due to the decreasing number of approved commercial test allergens available on the market.
(e.g.positive reaction to a metalworking fluid and corresponding patch testing of at least the baseline series and the 'metalworking fluid series' of the German Contact Dermatitis Research Group (DKG)) and/or consecutive patch testing of all ingredients of the POM was conducted to identify the culprit allergen.It was further assessed whether the reading times of the patch tests were suitable (i.e. at least two readings have been taken, one after 24 or 48 h and the other more important one after 72 or 96 h respectively) and whether substances were tested that should not have been tested at all (e.g.due to testing of the substance on the skin generally not recommended).The results presented thus cover two levels: (i) deficiencies in documentation of patch test procedures and (ii) insufficient quality of patch testing.The evaluation of the patch test data was carried out by experienced (occupational) dermatologists.Data analysis was done by means of descriptive statistics.
Displayed are results before (n = 161) and after removing the presumably false-positive reactions (n = 128).in 36 (1.2%) of 3004 POM patch tests; none of 1594 tests in hospitals and 35 (3.8%) of 915 tests in practices (one could not be allocated).Six (16.7%) of 36 POM patch test substances which should not have been tested at all were categorized as that due to patch testing of native plants (e.g.apple peel inside and outside, kiwi fruit peel inside and outside, apple pulp, kiwi fruit pulp) and 30 (83.3%) due to the fact that the substance should not at all be put on the skin (e.g.methylated spirits, sanitary cleaners, special cooling lubricants, petroleum, separating agents, dishwashing detergents, stainless steel spray, zinc spray, silicone remover).4| DISCUSSIONTo the best of our knowledge, this is the first study assessing systematically the quality of patch testing POM in patients with suspected OSD.The presented interim analysis revealed deficiencies in the quality of performing and particularly of documentation of patch tests with POM.Based on the evaluation of the quality criteria presented in this paper, the value of the documented results was severely compromised in a considerable amount of the patch tests with POM.The most frequently tested POMs in this interim analysis were protective gloves, skin care products, disinfectants (including hand disinfectants, surface disinfectants and unclassified disinfectants), metalworking fluids, work clothing, skin and hair cleansers and cleaning agents/detergents and corresponds with the most common professions of included patients (nurses, cleaners, food handlers, metalworkers).The highest number of positive patch test reactions (before and after removing the most likely false positive reactions) were caused by protective gloves, metalworking fluids and skin care products.These results were similar to a study by Schubert et al. who analysed data of 654 patients with suspected OSD patch tested between 2013 and 22[18][19][20][21]Data on the submitting social accident insurance institutions, the location in which the patch tests were executed (i.e.medical practice or hospital) as well as the kind (i.e. products or their single ingredients) and category (e.g.glove, soap, cooling lubricant, etc.) of the tested POM were extracted.Patient characteristics were summarized and primarily described according to the MOAHLFA index.22Further, the patients' professions were recorded.
times).Using this form is not strictly mandatory (e.g. for reimbursement of the patch test), but highly recommended by the statutory accident insurances in Germany.The form is meant to be used as a protocol for patch testing POMs.If comparable documentation sheets (i.e.providing the information of the sheet by the German Social Accident Insurance

Table 2 .
A full description of the tested POM, adequate to surely iden-Central criteria for evaluating the quality of test performance and documentation in 3004 recorded patch tests of patient's own material (POM) of 460 patients with suspected occupational skin diseases.
(rest could not be allocated).The test concentration was suitable (i.e.not too high or too low) in 574 (91.0%) of 631 POM patch tests with a declared test concentration; 367 (99.2%) of the 370 tested in hospitals and 145 (84.8%) of the 171 tested in practices (rest could not be allocated).For the 57 cases in which the test concentration was not suitable, a test concentration which was too high was present in 14 (24.6%) and a test concentration which was too low was present in 43 (75.4%)cases.(restcould not be allocated).The pH value was suitable (i.e.not too high or too low) in all of 89 POM patch tests with a declared pH value, regardless of the POM patch test being conducted in a hospital or practice.In 3004 patch tests of POM, 161 positive reactions were documented.The 161 positive reactions corresponded to 151 individual T A B L E 2 n/n total (%) n/n total (%) n/n total (%) n/n total (%) a Due to anonymization.
The documented reading times as well as the declared test vehicles, test concentrations and pH values were considered suitable in the majority of the POM patch tests.It should, however, be noted that the documentation of patch tests with POMs is often inadequate.In about 30%, a full description of the tested POM was missing.This may hamper the attribution of a positive test reaction to a being expensive.Moreover, this requires profound knowledge and often time-consuming research in choosing the correct test concentrations and test vehicles.In line with this, only in a very few cases, positive reactions to POMs were followed by testing of individual substances of the substance mixtures in our study.This, of course, not only hampers avoidance of the culprit allergen (e.g., by substitution) in the individual case, which might help to remain in the job but also compromises identification of new allergens.A comparison between hospitals and practices shows that the documentation quality with regard to all parameters considered in this study (i.e.full description of the POM, declaration of the test concentration, declaration of the test vehicle and declaration of the pH value) was better in hospitals than in practices.patients in practices, is much worse than by the statutory accident insurance.Therefore, many practices perform only a limited number of patch tests and hold less test substances available.However, the results clearly indicate a great need for further training of particularly in practices in performance and documentation of patch testing POM.