Occupational dermatitis is a skin disorder caused or aggravated by conditions at the workplace. An example is contact dermatitis, an inflammatory, eczematous skin reaction caused by direct contact with or environmental exposure to a substance, which may be a low-molecular weight compound or a protein.1 It is a pruritic condition that is often of multifactorial origin. Contact dermatitis initially manifests as intensely pruritic, erythematous lesions, immediately followed by vesicles, exudation, crusting, and lichenification caused by scratching. The condition is classified as acute or chronic depending on the predominant type of lesions. There are 2 types of contact dermatitis: irritant and allergic. Irritant contact dermatitis (ICD) is much more common than its allergic counterpart (80% vs 20%). Allergic contact dermatitis (ACD) is a delayed type IV hypersensitivity reaction, unlike ICD, which develops following exposure to sufficient concentrations of an irritant substance.1–3 It has been reported that 20% of the general population are at risk of developing contact dermatitis at some point in their lives.4

Occupational contact dermatitis, which is well characterized both clinically and pathologically, can be caused or aggravated by working conditions. According to data published in 2004, skin disorders account for up to 30% of all occupational diseases in industrialized countries5 and contact dermatitis accounts for 90% of these skin disorders.6

It is difficult to compare data between countries because of differences in practices in the reporting of occupational disease. Furthermore, the incidence of occupational contact dermatitis is underestimated and records tend to be incomplete.3

Occupational skin disorders have a considerable socioeconomic impact. Contact dermatitis occurring at the workplace, for example, is a frequent cause of work-related disability and can account for up to 30% of workers’ compensation costs in many countries.3 The impact of occupational skin disorders on the lives of patients and their families has not yet been studied in depth.5,7

It is difficult to establish a causal link between contact dermatitis and a specific occupation or activity at the workplace without the use of objective criteria. In 1989, Toby Mathias1 proposed 7 such criteria to assess the link between contact dermatitis and occupation. None of the criteria in isolation provides sufficient evidence of a causal link,6 but the fulfillment of 4 of the 7 criteria is considered indicative of probable causation.1

In Spain, no studies have validated the usefulness of the Mathias criteria. Our aim was to assess the value of these criteria for establishing probable occupational causation of contact dermatitis.

The aim of this study was to assess the usefulness of the Mathias criteria to establish occupational causation of skin disorders.

The characteristics of the patients based on the information collected during the Mathias criteria interview are shown in Table 2.

After analyzing the results of the interview, we divided the 103 patients into 2 groups: those with a nonoccupational skin disorder (n=90) and those with an occupational skin disorder (n=13). A number of differences were observed between the groups depending on the variable analyzed: a) age: 53.8% of patients with an occupational skin disorder were 35 years old or younger, compared to 28.9% of those with a nonoccupational skin disorder; b) occupation: 61.5% of the patients with an occupational skin disorder worked in the sales and services sector, and 40% of those with a nonoccupational skin disorder did not have paid employment; c) location of skin lesions: 76.9% of patients with an occupational skin disorder had lesions, compared to only 9% of those with a nonoccupational skin disorder (this was one of main differences observed); d) diagnosis: the main diagnosis in the occupational skin disorder group was ACD (53.8%), while in the other group it was other skin disorders (39.6%).

Of the 103 patients analyzed, 13 met 4 or more of the Mathias criteria (Table 3), indicating a probable occupational causation.

Seven of the 12 patients with an occupational skin disorder according to the Mathias criteria had a primary diagnosis of ACD, and in 5 of these the positive allergy test results were considered to be of present occupational relevance (Table 4). Patient #11, a waiter with a primary diagnosis of ACD with foot involvement, wore shoes at work that caused contact dermatitis on his feet. Patient #10 had a presumed diagnosis of ACD with involvement of the back, but the positive patch test results were classified as being of unknown relevance. The patient was a policeman, who attributed his condition to a new heat-resistant shirt he wore at work. This could not be demonstrated, however, as a patch test performed with material from the shirt was negative; nonetheless, the shirt produced an irritant dermatitis on the patient's back. It cannot, therefore, be ruled out that the initial symptoms assumedly caused by the heat-resistant shirt were responsible for the patient's dermatitis. Five patients were diagnosed with ICD. One of these (patient #1) had a primary diagnosis of ACD and a secondary diagnosis of ICD, but only ICD was classified as having occupational relevance (negative patch tests). In patient #7, who had a primary diagnosis of urticaria, occupational relevance was established on the basis of the patient's clinical and occupational history, a history of contact dermatitis of the hands, and allergy test results.

Twelve of the 13 patients who had a work-related skin disorder according to the Mathias criteria were also considered to have a skin disorder of present or past occupational relevance according to the dermatology specialist (Table 5). The present relevance was not the same for all the patients. For example, in 1 patient, the ICD but not the ACD was considered to be work-related and in another patient, there were 2 positive patch test results but only 1 of them was occupationally relevant. All the patients were analyzed on a case-by-case basis.

In brief, probable occupational causation was established in 13 patients according to the Mathias criteria and in 12 according to the dermatology expert (Table 5).

One of the 13 patients (patient #4) had a false positive result as the positive patch test reaction was of past nonoccupational relevance. Nonetheless, the answers to the Mathias questionnaire indicated probable occupational causation. The patient was diagnosed with contact urticaria.

The Mathias criteria had a sensitivity of 100%, a specificity of 98.90%, a positive predictive power of 92.31%, and a negative predictive power of 100%; the prevalence of occupational contact dermatitis was 11.65% (Table 6). These results show that the Mathias criteria have a high validity and diagnostic yield in occupational contact dermatitis.

The validity of the Mathias criteria according to the different measures and variables analyzed was reduced by the false positive result for patient #4 (Table 6).

Our findings show that the likelihood of establishing occupational causation in a patient with occupational contact dermatitis using the Mathias criteria was 100% (sensitivity, 100%), with a specificity of 98.9%, a positive predictive power of 92.31%, and a negative predictive power of 100%. These results confirm that the Mathias criteria have a highly validity and diagnostic yield in occupational contact dermatitis.

Of the 103 patients analyzed, a link between disease and employment was found in 13 patients according to the Mathias criteria and in 12 according to the judgment of the dermatology specialist. In other words, the application of the Mathias criteria yielded just 1 false positive result.

In a similar study to ours, Ingber and Merims6 concluded that the Mathias criteria were useful for assessing occupational causation and aggravation of contact dermatitis, but they had reservations about the first criterion. According to Mathias, all the criteria in the questionnaire should be analyzed, even if the answer to the first question is no. While Ingber and Merims agreed that a patient might have a work-related skin disease other than contact dermatitis (e.g., atopic dermatitis), they considered that it would be better to have a questionnaire where a negative answer to the first question meant that the rest of the questions did not have to be answered.

Similarly to Ingber and Merims,6 we found that the Mathias criteria were useful for establishing a causal link between occupation and skin disease.

Several aspects should be taken into consideration with regard to our findings: 1) the first question in the Mathias questionnaire (“Is the clinical appearance consistent with contact dermatitis?”) was answered using the information available during the patient interview, where not all patients had visible symptoms; 2) for the second question (“Are there workplace exposures to potential cutaneous irritants or allergens?”), it was considered that workplace exposure did not occur in students, homemakers, or pensioners; 3), question 5 (“Are nonoccupational exposures excluded as likely causes?”) was difficult to assess because it is impossible to exclude nonoccupational exposure as workers are in contact with suspect allergens both at and outside the workplace5; and finally 4) question 7 (“Do patch tests or provocation tests implicate a probable causal agent?”) had to be answered after the interview as it was necessary to review the patients’ clinical history to assess the allergy test results.

Patient #4 had a false positive result that influenced the overall sensitivity, specificity, and predictive values for the Mathias criteria. Based on the application of the criteria, there was a probable link between the patient's condition and his occupation, but the primary diagnosis according to the dermatology specialist was contact urticaria without occupational relevance. None of the Mathias criteria in isolation provides sufficient evidence to establish occupational causation of contact dermatitis.6 The questionnaire contains 7 criteria and likely causation is established on the basis of a positive answer to any 4 items. As Ingber and Merims reasoned,6 if the answer to the first question is no (as was the case with patient #4 in our series), there would be no reason to continue with the other 6 questions considering that a causal relationship between contact dermatitis and employment has been ruled out.5 In our study, we applied the Mathias criteria exactly as proposed by Mathias.1 Accordingly, all 7 criteria were tested, despite the negative answer to question 1 (the patient had urticaria and not dermatitis). The answers to the remaining 6 questions established the probability of a causal relationship between the patient's skin condition and employment (4 positive answers).

This study has several limitations: the time allocated to the study was short; a specialist was needed to answer question 7; and not all patients had symptoms during the interview, meaning that question 1 had to be answered on the basis of information gathered during the questioning of the patient. Not all patients have symptoms when seen in routine clinical practice, as many only see the specialist months after they have first consulted for their symptoms, and these patients will often have received several treatments. Nevertheless, a substantial number of patients with severe disease are seen and diagnosed when they have visible symptoms. A final limitation of our study is the possible presence of selection bias as our data are based on a hospital sample.

The study has numerous strengths: we studied a large number of patients, the dermatology department where the study was conducted has experience in occupational disease, and the Mathias criteria were applied without knowledge of the patient's clinical history.

Very few studies have investigated the causal link between contact dermatitis and employment, despite the obvious interests for patients, workers’ compensation agencies, and society as a whole. While the Mathias criteria were designed to establish the likelihood of occupational causation of contact dermatitis, they may also be useful for guiding referrals to specalists in occupational dermatology. Although more studies are necessary, our findings suggest that the Mathias criteria are useful for assessing occupational causation, although this does not obviate the need for additional, necessary measures such as a visit to the patient's workplace and an analysis of what his or her job entails.

In summary, we can conclude that the Mathias criteria are useful for establishing probable occupational causation of contact dermatitis. Finally, we believe that it is important to disseminate these criteria, because, as has been suggested by other authors, greater knowledge and application of these criteria may improve the diagnosis of occupational contact dermatitis.

The authors declare that they have no conflicts of interest.