Dermatoses determined in a population of farmers in questionnaire-based clinical study including methodology validation

Dermatoses determined in a population of farmers in a questionnaire-based clinical study including methodology validation. Scand J Work Environ Health 1995;21: 30-5. Objectives The study analyzed skin diseases in a population sample of Finnish farmers descriptively and in the process validated the question "Do you have a skin disease now?" Methods All farmers from one Finnish municipality were surveyed for dermatoses, first with a questionnaire and then with a clinical examination of those who reported dermatoses. Another population of farmers answered a set of questions in~n~ediately before a clinical examination, and the self-report of current dermatosis was validated. Results Eczema was diagnosed for 66% of the women and 53% of the Inen who had reported dermatosis in the questionnaire study 6 to 12 months earlier. Toe-web maceration, psoriasis, folliculitis, and acne were, after eczema, the most frequent diagnoses (in that order). In more than 50% of the cases, the location of clinically determined dermatoses corresponded with the skin disease areas reported 6 to 12 months earlier. In the validation study, everyone who reported a skin disease immediately before the clinical examination were found to have a skin disease. In addition 22% of those not reporting dermatosis were found to have a skin disease. Toe-web maceration was the most colnlnon dermatosis not reported by the farmers. C O ~ C ~ U S ~ O ~ ~ S Finnish farmers suffered from the same type of dermatoses as other populations. The prevalence of eczema and hand eczema was similar to that of other risk populations. A self-report of current dermatosis is probably a good indicator of the point prevalence of explicit skin diseases in populations.

Some epidemiologic studies have been done on skin diseases in populations, including surveys among inhabitants of the Faroe Islands (I), United States inhabitants (2), Londoners (3), and Finnish Lapps (4). In Finland, a few skin disease surveys have been conducted also in occupational groups [eg, bricklayers (3, shipyard workers (6), workers in the food industry (7), and reindeer herders (8)].
No standardized validated questionnaires exist for epidemiologic studies in dermatology. Some epidemiologic studies of skin disease have also included methodology validation (3,(9)(10)(11), however, or have been designed for validation purposes (12,13). There are neither uniform criteria nor standards for the definition of a case or a noncase of dermatosis or eczema in clinical studies or in postal surveys. Because, in most studies, the primary screening of the cases is an inquiry (postal or inter-view), it is critical to laow the sensitivity and specificity of the method used.
The present study, being a part of a large epidemiologic investigation of dermatoses among farmers, is a descriptive analysis of skin diseases in the farming population of a Finnish municipality (Pielavesi). In addition, the validation of a "self-diagnosis" of current dermatosis ("Do you have a skin disease now?") in a questionnaire was included in the study. Information on the validity of answering this question is important for epidemiologic surveys of skin diseases. The answer serves as a simple screening method for further examinations and also as a means of estimating the point prevalence of skin diseases in population studies.
The analysis and publication of the data were delayed because of the death of the principal researcher (LH) after the data collection.

Subjects and methods
In December 1979 a questionnaire regarding health and w o k was sent to a population of Finnish farmers including all registered farmers between the ages of 18 and 64 years in two municipalities (A and B) (14). Most of the farmers in these two inunicipalities were involved in dairy farlning. The questioinlaire contained the question "Do you have a skin disease now?" The positive answers to this question gave the point prevalence of current skin diseases. The men and women with a positive answer were asked to indicate their dermatosis areas by shading them on a picture of a person. The hand and the forearin were together in the same area.

Skin diseases in a dairy farming municipality
In municipality A, 962 farmers (5 14 men and 448 women) out of 1052 (95%) responded to the questionnaire in 1979. All of those reporting a "skin disease now" (88 men and 102 women) were invited to a clinical examination (history of skin disease and inspection always, skin prick tests, patch tests, and additional laboratory investigations when needed) at a department of dermatology in a university hospital. Altogether 73% of those reporting skin disease (64 Inen and 74 women) attended. They were examined by two derinatologists (LH and AH). The examination took place 6 to 12 months after the questionnaire survey, either in April to May or September to October 1980. The main farming operation of 85% of the surveyed population was dairy farming, and the figure was similar (86%) for the examined farmers. Thc mean age of the examined far~ners was 44.5 (10.9 SD) years for the men and 44.5 (10.1 SD) years for the women.

Validation study
For the validation study, both those with derinatoses and those without were needed. Therefore, from municipality B, a random sample of those reporting no der~natosis and another random sample of those reporting hand dermatosis in the questionnaire in 1979 were invited to visit the municipal health care center for a "health check," without them being informed that the examination concerned skin diseases. Fro~n the first group, 69 men and 53 women and, from the second group, 13 Inen and 28 women attended. The examination took place 16 months after the questionnaire survey. Before the examination, the farmers answered a short set of questions, including the following: "Do you have a skin disease now?" Immediately thereafter their skin was examined by a dermatologist (LH).
In analyzing the data, cross-tabulation and a chisquare test were used.

Skin diseases in a dairy farming municipality
Among the farmers of municipality A the point prevalence of self-reported skin diseases (current slcin disease) was 17.1% for the men and 22.8% for the women. The point prevalence of hand dermatoses was 4.1% for the inen and 14.1 % for the women.
The self-reported dermatosis areas on the farmers attending and on those not attending the examination are listed in table 1. There were differences in the attendance by the reported area of dermatosis in the questionnaire (table 1). Most of those with hand derlnatoses attended the examination. There were no significant differences in the farming operation, size of farm, atopy, or age between the examined and the unexamined farmer groups.
The location of the self-reported dermatoses and the later clinically diagnosed dermatoses is presented in table 2. For 70% of the farmers reporting skin disease on their hands in the questionnaire, dermatosis involving their hands was found in the clinical examination 6 to 12 Table 1. Principal area of dermatosis (questionnaire survey) on the clinically examined farmers and those who did not come to the examination.   Principal diagnoses outside parentheses, in parentheses the total number of diagnoses including secondary diagnoses. Estimation of the prevalence: questionnaire prevalence of any skin disease in the population times the number with the specific diagnosis divided by the number of all with any diagnosis. P < 0.05, * * P < 0.01, * * * P < 0.001 (all P-values between men and women).
months later. Of all derinatoses, 50% were detected on the same areas the farmers had reported previously. The diagnoses of skin diseases among the examined farmers are listed in table 3, together with the estiinated prevalences in the surveyed population. Eczema was more common for the women (83% of the diagnoses) than the men (63%) (P = 0.03). Eczeina was on the hands of 84% of the women and 41% of the men (P = 0.0006) who suffered from eczema, and it had been on the hands in either the questionnaire survey or the examination for 94% of the women and 47% of the men. Current atopic eczema was diagnosed in seven cases (5%). For five men and eight woinen who reported hand dermatoses in the survey, the symptoms had healed before the examina-tion. Eczema not involving the hands was present on the legs of nine of nineteen men and three of eight women. In the remaining cases eczema was mainly on the trunlc. Interdigital maceration was found as the principal diagnosis for one man and two women, and as a secondary diagnosis for two men (total of 3 cases) and 11 women (total of 13 cases).
There were nine women and seven men who had not stated the location of their dermatoses in the questionnaire (table 1). For five wolnen and three men the dermatosis had healed before the examination or the inen and women had not had any dermatosis. Two of these wolnen had "rough" hands. The dermatological diagnoses made for the remaining four women and four men with no reported location in the questionnaire were four cases of eczeina on the hands, and one of each of a leg eczema, folliculitis, acne, and rosacea.

Validation study
All of the farmers answering affirmatively to the question of "having a skin disease now" immediately before the clinical examination (N = 41) in the validation study were diagnosed as having dermatosis. Of those farmers answering no, 22% (27 of 122) were diagnosed as having a dermatological disease. Six of these cases of dermatoses were eczema, and three of them were on the hands (2.5%). The remaining slcin diseases, which were not considered dermatoses by the farmers, were toe-web maceration on the feet in 16 cases, paronychia on the fingers in three cases, acne in one case, and nliliaria rubra in one case. A "self-diagnosis" of having "a skin disease now" had, in this validation, high spccificity (1.00) but lower sensitivity (0.60).

Discussion
The prevalcnces and estimated prevalences of all dernlatoses and some specific slcin diseases in previous population studies and the present study can be sccn in table 4. It is difficult to malce comparisons between the studies since the criteria for slun discases vary. In 1963, Loinholt (1) registered all self-rcported dermatoses needing medical attention in a population of 1 1 000 in the Faroe Islands. This is probably the most thorough epidemiologic survey of derinatoses in a spccific population. In Loinholt's survey, the prevalence of eczema was only somewhat higher than that of psoriasis (3.2 versus 2.5%). The prevalences of all slcin discases or psoriasis have been fairly consistent during the last decades according to the studies listed in table 4. The prevalence of hand eczema scems to have been much higher in the surveyed occupational groups (5-8) than in general populations (1,2).
The only statistical difference found in the examined group of farmers, when compared with the group not attending the clinical examination, was in the area of a All eczema, including atopic eczema, seborrheic eczema, contact dermatitis and neurodermatitis b Examined immediately after self-report, cases needing treatment. Cases of all grades. Moderate to severe cases. "Significant skin conditions." Skin complaints ("conditions of concern to the individual"). Q Those reporting dermatoses in a survey were examined.
dermatosis (table 1). Of those with self-reported hand dermatoses, 87% came to the examination. On the other hand, 46% of those not attending the clinical exanlination had not specified their derlnatosis area in the questionnaire. These facts speak in favor of the concept that those who had a well definable derrnatosis attended the examination, and those with more vague problelns were not prone to attend. The diagnoses of farmers with hand eczema, and also the skin test results, have been published elsewhere (15). According to that study, 93% of the hand der~natoses were eczema, and 70% were considered work-related. Allergy to cow dander, either in skin prick tests with commercial extract or in a patch test with cow dander as is, was found for 39% of the tested farmers (15). In the present material, hand eczema was more cornmon for the wornen than for the men. The finding agrees with the results of many other population studies (9,10,(16)(17)(18). Eczema had involved the hands or the forearms of most of the women but of only about half of the men with eczema. No significant differences between the genders were found in other diagnoses. The prevalences of eczema (14.5%) and hand eczema (9.8%), as well as the proportion of hand eczema of all diagnoses (50%), were high in this study compared with those of earlier population studies (1-3), but similar to the studies on risk occupations (5-7) (table 4). The prevalence of psoriasis (1.8%) was similar to the results of other population studies outside Finland (1-3), but higher than in some earlier Finnish studies (5)(6)(7)(8) (table 4). In this population, the prevalence of toe-web maceration was low among the Inen when compared with that of the women (table 3), and also when compared with the prevalence of 'tinea pedis' in some other studies (2,5,8) (table 4). This difference is partly explained by the fact that, in the validation portion of our study, most of the unreported dermatoses (16 of 27 cases) were toe-web maceration.
In the present study, 113 out of 138 farmers were found to have a dermatological diagnosis in the clinical examination 6 to 12 months after a self-report of a skin disease in the questionnaire survey. As seen from the data in table 2, the self-report of the location of the dermatosis corresponded fairly well with the findings in the clinical examination 6 to 12 months later. Of the dermatoses, three-fourths of the cases were found on the same or larger skin areas than those reported in the questionnaire.
In the validation portion of the present study the question of "having a skin disease now" (self-diagnosis) showed that the affirmative answer was very reliable (high specificity), while some of those with a negative answer also had derlnatoses (lower sensitivity). This finding agrees with that of other studies (9,10,12,13). In our study, 2.5% of those who did not report "a skin disease now" immediately before the examination were diagnosed as having hand eczema by the dermatologist. Toeweb maceration was the no st common dermatitis not reported by the farmers. It is probable that unreported derlnatoses are of minimal concern to respondents.
In Agrup's population study from 1969, 93.5% of those reporting hand derrnatosis in a questionnaire and 1.7% of those not reporting it were clinically found to have skin changes on their hands (9). Most of the unreported cases of hand derlnatoses in Agrup's study were lnild eczema. According to Johnson's study (tables 2 and 3), the prevalence of clinically diagnosed dermatoses was higher than the prevalence of skin complaints of the individuals. Meding (10) inquired about the respondent's opinion of having had hand eczema during the previous 12 months, and the affirmative answer was confirmed by a dermatologist in 89% of the cases. Berg & Axelson (12) clinically evaluated a questionnaire for skin cornplaints and found high confirmation rates for atopic eczema (0.92), seborrhoeic dermatitis (0.86), and hand dermatoses (0.83), while the rates were lower for acne, rosacea, and skin cancer. A recent Dutch study on nurses (13) validated questions for detecting hand dermatitis in epidemiologic studies. The "self-diagnosis" of hand der~natitis was a good indicator of the presence of it, but also 8% of those with a negative report were considered to have hand dermatitis by a dermatologist. In the same study, a symptom-based diagnosis of hand eczema detected all of the hand eczema cases (high sensitivity), but for those with a positive diagnosis, there were many without clinically confirmed hand eczema (low specificity).
It is not known how the perception of what a disease of the skin is varies in different populations, times, and cultures. A self-report of a skin disease is probably a good method with which to detect the derlnatoses of importance to the respondent. According to this study, shading the dertnatosis areas in a picture of a human body seems to be a fairly reliable self-report of the location of a skin disease in questionnaire studies.
Internationally accepted criteria for skin disease epidemiology need to be formed. For diagnosing hand eczema, a combination of a self-diagnosis and a symptombased diagnosis is probably the most reliable. The criteria must be validated in different populations and countries before final approval for standard use is given.