Summary
Background
Dermatitis is an important health outcome for workers whose jobs put them in contact with irritants or sensitizing agents.
Objectives
We conducted analysis of data from the Epidemiological Study on the Risk of Asthma in Cleaning Workers 2 (EPIASLI2) study to assess worksites and cleaning products as risk factors for hand dermatitis among professional cleaning workers.
Materials/Methods
We distributed 4 993 questionnaires to employees of 37 cleaning companies and used data from 818 (16%) respondents who provided information about skin symptoms and cleaning-related exposures. We assessed associations between the frequencies of worksite and cleaning product exposures and a symptom-based definition of hand dermatitis among current cleaning workers (n=693) and a comparison population (n=125).
Results
Hand dermatitis was reported by 28% of current cleaning workers, versus 18% of the comparison population, and was associated with cleaning outdoor areas and schools, and the use of hydrochloric acid (prevalence ratio [PR]: 1.92, 95% confidence interval [CI]: 1.22, 3.02) and dust mop products (PR: 1.75, 95% CI: 1.11, 2.75).
Conclusions
Professional cleaning workers may not be sufficiently protected from cutaneous disease at work. Future research should further investigate the roles of multiple product exposures and personal protective equipment.
Keywords: cleaning, dermatitis, epidemiology, occupational diseases, occupational exposure
Introduction
Professional cleaning workers keep homes, hospitals, hotels, office buildings, restaurants, schools, shopping areas, sidewalks, and other public and private spaces clean. The cleaning activities they perform range from light tasks, such as dusting, sweeping, and vacuuming, to specialized activities that require hazardous cleaning solutions, heavy equipment, and job training (1,2).
Men and women in the cleaning industry routinely come into contact with a wide range of hazards potentially responsible for causing occupational skin diseases. Professional cleaning workers clean and sanitize surfaces that put their skin in contact with biological, chemical, and physical hazards. The process of cleaning and disinfecting also exposes workers to a range of cleaning products that varies widely according to the cleaning tasks and locations. While some cleaning agents seem to be simple and relatively safe (e.g., soap and water, vinegar), many contain preservatives, solvents, fragrances, and other compounds with well-known irritating or sensitizing properties (3). In addition, cleaning may involve “wet work” that may impair the epidermal barrier, allowing for skin irritation and sensitivity (4,5). To reduce the risk of exposure to both the substances being cleaned and the cleaning products themselves, workers may use gloves or other personal protective equipment that protect the skin from hazardous exposures, but that may also exacerbate skin allergies or irritate the skin.
Previous research provides evidence of an elevated prevalence of hand dermatitis among cleaning workers in hospitals (6,7) and school buildings (8). We conducted the present study to assess a wide range of worksites, cleaning activities, and cleaning products as risk factors for hand dermatitis among men and women employed in the cleaning industry in the province of Barcelona, Spain.
Materials and Methods
Study population and data collection
Epidemiological Study on the Risk of Asthma in Cleaning Workers 2 (EPIASLI2) was a two-stage epidemiological study conducted to assess associations between cleaning work, including specific worksites and exposures to cleaning products, and dermatological and respiratory health symptoms (9). The study methods have been described previously (9). Summarized briefly, between February and December 2008, 4 993 self-administered, paper-and-pencil-style questionnaires were distributed to employees of 37 cleaning companies operating in the province of Barcelona, Spain. Nine hundred and fifty questionnaires were returned by mail; of those, 132 (14%) were excluded because they were incomplete or lacked responses to key questionnaire items and the remaining 818 respondents constituted our final study population for analyses of the cross-sectional survey data. Following the cross-sectional survey, 95 participants were recruited into a nested case-control study designed to further assess associations between cleaning work and respiratory disease (10). Data for the sub-study were collected when participants completed a clinical examination and an interviewer-administered questionnaire. For 70 of the 95 participants (74%), the clinical examination included an in-person skin health evaluation, therefore our final study population includes additional data collection for a subset of 70. The study was approved by the ethics committee of the Instituto Municipal de Asistencia Sanitaria (IMAS) and the analytical plan for this analysis was also approved by the Institutional Review Board of the Wake Forest School of Medicine.
Occupational exposures
Participants were categorized as working as cleaners at the time of their participation using the questions “have you ever worked as a cleaner?”, “what position do you currently hold in your company?”, and responses to a series of questions about cleaning worksites, activities, and products used at work. As in previous analyses of EPIASLI2 data (9), respondents who indicated that they currently work as cleaners and those with positive responses to any of the questions about cleaning worksites, activities, and products in the last month were categorized as “current” cleaning workers (n=693). Those who indicated that they had worked, but do not currently work as a cleaner were categorized as “former” cleaning workers (n=57). The remaining respondents were those who had never worked as cleaners (n=68); this population includes office workers and other employees not performing cleaning jobs. In this analysis, former cleaning workers and those who had never worked as cleaners comprise the comparison population.
Respondents with positive answers to the series of questions about worksites (e.g., hospitals, private homes, schools), activities (e.g., window cleaning), and products (e.g., ammonia, bleach, glass cleaner) used in the last 12 months were then asked to estimate the frequency of their work at these sites, performing these activities, or using these products, respectively, during the last month. Glove use was accessed using a single questionnaire item: “how often did you use rubber gloves during the last 12 months?” Whether the respondent cleaned his/her own home was assessed with the question “do you do cleaning tasks in your home?” Missing responses to these questions were re-assigned as negative responses – that is, not using the specific cleaning product, use of rubber gloves less than once per week, and not cleaning one’s own home.
Dermatitis and eczema
Participants reported whether they had a history of eczema or other skin allergy by responding to the following question: “have you ever had eczema or other skin allergies?” The questionnaire also included a series of questions developed by Coenraads et al. (11) and Smit et al. (12) to assess the 12 month prevalence of symptoms of hand dermatitis or eczema (hereafter referred to as “hand dermatitis”). The symptoms of hand dermatitis included the following: (a) red and swollen hands or fingers, (b) red hands or fingers and fissures, (c) vesicles on the hands or between the fingers, (d) scaling hands or fingers with fissures, and (e) itching hands or fingers with fissures. Symptoms were then classified using the methods described in Smit et al.; that is, participants were classified as having hand dermatitis if they indicated that they had at least one of the symptoms in the last 12 months and that the symptom(s) lasted more than 3 weeks or occurred more than once (12).
For 70 participants who completed the in-person clinic exam portion of the study, questionnaire items included in the survey at the clinical examination were identical to those included in the cross-sectional study described above. The clinical examination also included an in-person assessment by an occupational medicine physician with specialized training in dermatology and evaluation of anonymised photographs of the hands, wrists, and forearms by a dermatologist with expertise in contact dermatitis. The in-person assessment was based upon a visual evaluation of the hands, wrists, and forearms and conversation with the participant at the time of the exam. Evaluation of anonymised photographs occurred after photographs of the fronts and backs of participants hands were taken at the time of the exam. Photographs were taken with a Sony Cyber-shot® digital camera (model: DSC-W55), from a distance of 50 cm, and the images were stored digitally and reviewed later. Each of the 70 participants was classified as positive or negative for hand dermatitis independently by the two clinicians. For cases in which the two assessments generated discrepant results, the in-person and photographic assessments were re-evaluated together to reach a final classification.
Statistical analysis
We estimated associations between each cleaning-related risk factor and hand dermatitis using Poisson regression models, specified with a log link and robust error variance estimation (13,14). Cleaning-related risk factors included employment as a current cleaner, worksite, cleaning activities, and products used at work. Results for current cleaning were generated using a single model estimating the prevalence of hand dermatitis among current cleaning workers (n=693) compared to the prevalence of hand dermatitis among former and never cleaning workers (i.e., the comparison population, n=125). Results for each worksite were generated using two models. First, we used a model estimating the prevalence of hand dermatitis among current cleaning workers who have and have not cleaned at the specific type of worksite in the last 12 months compared to the prevalence of hand dermatitis in the comparison population. In order for the estimates generated for the effect of cleaning at the specific worksite within the last 12 months to include the full study population and to retain the ability to directly contrast the estimates across worksites, the population that reported not cleaning the specific type of worksite in the last 12 months is included in each model and the results are presented. Effect estimates presented for the population not cleaning the specific worksite within the last 12 months may be interpreted as the effect of currently cleaning, but not cleaning at the specific worksite; the primary outcome of interest generated using these models is the prevalence of hand dermatitis among current cleaning workers who reported cleaning the specific worksite in the last 12 months relative to the prevalence of hand dermatitis in the comparison population. Second, we used a model estimating the prevalence of hand dermatitis with increasing frequency of work at the specific type of worksite in the population of current cleaning workers who reported such work within the last 12 months compared to the prevalence of hand dermatitis in the comparison population. Results for analyses of work activities and use of cleaning products were generated using similar models. All models were adjusted for age, country of birth (Spain versus other), sex, history of eczema or other skin allergy, cleaning one’s own home, and frequency of glove use at work (<1 day/week, 1-3 days/week, 4+ days/week). Associations are presented as prevalence ratios (PRs) with 95% confidence intervals (CIs). All analyses were performed using SAS™ version 9.1 (SAS Institute, Inc.).
In a second set of analyses, we used data collected at the time of the clinical examination (n=70) to compare the classifications of hand dermatitis based on participants’ responses to survey items about hand dermatitis with the classification based on the dermatologic assessment. We considered the dermatological assessment as the gold standard and calculated the sensitivity, specificity, positive predictive value, and negative predictive value of the questionnaire-based classification of hand dermatitis.
Results
Characteristics of the 818 study participants are shown in Table 1. Overall, the population of current cleaning workers was slightly older (median age: 45 years) than the comparison population (median age: 40 years), and included larger proportions of women (84% versus 74%) and individuals who clean their own homes (95% versus 83%). The percentage of current cleaning workers with current asthma was slightly higher (11% versus 7%) and the percentages of respondents with a history of eczema or other skin allergy were similar in the two populations (29% versus 30%).
Table 1.
Demographic characteristics and the prevalence of hand dermatitis among current cleaning workers and members of the comparison population
| Total Population |
Comparison Population |
Current Cleaning Workers |
|||
|---|---|---|---|---|---|
| No. (%) | No. (%) | No. (%) with Hand Dermatitis |
No. (%) | No. (%) with Hand Dermatitis |
|
| Total | 818 (100) | 125 | 22 (18) | 693 | 191 (28) |
| Demographic Characteristics | |||||
| Age, in years | |||||
| Mean ± SD | 45 ± 10 | 42 ± 11 | 45 ± 10 | ||
| Median | 45 | 40 | 45 | ||
| Minimum-Maximum | 18-65 | 22-61 | 18-65 | ||
| Country of birth | |||||
| Spain | 608 (74) | 107 (86) | 19 (18) | 501 (72) | 151 (30) |
| Other | 210 (26) | 18 (14) | 3 (7) | 192 (28) | 40 (21) |
| Sex | |||||
| Female | 673 (82) | 92 (74) | 17 (18) | 581 (84) | 168 (29) |
| Male | 145 (17) | 33 (26) | 5 (15) | 112 (16) | 23 (21) |
| Cleaning | |||||
| Cleans own home | |||||
| No | 53 (6) | 21 (17) | 2 (10) | 32 (5) | 5 (16) |
| Yes | 765 (94) | 104 (83) | 20 (19) | 661 (95) | 186 (28) |
| Rubber glove use in the last 12 months? | |||||
| <1 day/week | 266 (33) | 101 (81) | 17 (17) | 165 (24) | 45 (27) |
| 1-3 days/week | 108 (13) | 4 (3) | 2 (50) | 104 (15) | 31 (30) |
| 4+ days/week | 444 (54) | 20 (16) | 3 (15) | 424 (61) | 115 (27) |
| Health History | |||||
| Current asthma symptoms1 | |||||
| No | 736 (90) | 116 (93) | 19 (16) | 620 (89) | 160 (26) |
| Yes | 82 (10) | 9 (7) | 3 (33) | 73 (11) | 31 (42) |
| History of eczema or other skin allergy | |||||
| No | 579 (71) | 88 (70) | 9 (10) | 491 (71) | 76 (15) |
| Yes | 239 (29) | 37 (30) | 13 (35) | 202 (29) | 115 (57) |
As defined in Vizcaya et al. (9): positive response to at least one of the following questions: “Have you been woken by an attack of shortness of breath at any time in the last 12 months?”, “Have you had an attack of asthma in the last 12 months?”, “Are you currently taking any medicines including inhalers, aerosols or tablets for asthma?”
Of the individual skin symptoms included in our questionnaire, the most prevalent symptoms reported were red hands or fingers with fissures (20%) among current cleaning workers and scaling hands or fingers with fissures (9%) in the comparison population; thirty-six percent (n=248) of the current cleaning workers and 22% (n=27) of the comparison population reported one or more of the symptoms (Table 2). Overall, 28% of current cleaning workers and 18% of the comparison population (former cleaners: 14%; never cleaners: 21%) met our definition for hand dermatitis and the prevalence among current cleaning workers was elevated compared to the comparison population (PR: 1.60, 95% CI: 1.03, 2.47) (Table 3). Like all our models, this estimate was generated using a model adjusted for potential confounders selected a priori, as well as for cleaning one’s own home (PR: 1.14, 95% CI: 0.56, 2.32) and the frequency of using rubber gloves at work (1-3 days/week versus <1 day/week: PR: 1.12, 95% CI: 0.78, 1.61; 4-7 days/week versus <1 day/week: PR: 0.98, 95% CI: 0.74, 1.30) (data not shown).
Table 2.
Skin symptoms in the last 12 months, as reported by current cleaning workers and members of the comparison population, and criteria used to define hand dermatitis
| Total Population (n=818) |
Comparison Population (n=125) |
Current Cleaning Workers (n=693) |
|
|---|---|---|---|
| No. (%) | No. (%) | No. (%) | |
| Symptoms | |||
| Red hands or fingers and fissures | 146 (18) | 10 (8) | 136 (20) |
| Scaling hands or fingers, with fissures | 136 (17) | 11 (9) | 125 (18) |
| Red, swollen hands or fingers | 124 (15) | 10 (8) | 114 (16) |
| Itching hands or fingers, with fissures | 102 (12) | 8 (6) | 94 (14) |
| Vesicles on hands or between fingers | 43 (5) | 5 (4) | 38 (5) |
| Criteria | |||
| ≥1 symptom | 275 (34) | 27 (22) | 248 (36) |
| ≥1 symptom that lasted >3 weeks or occurred >once1 | 213 (26) | 22 (18) | 191 (28) |
| ≥2 symptoms that lasted >3 weeks or occurred >once | 95 (12) | 6 (5) | 89 (13) |
Definition of hand dermatitis used in this analysis
Table 3.
Associations of cleaning worksite and major work activities with hand dermatitis
| No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Comparison population | 125 | 22 (18) | 1.00 | |||||||||
| Current cleaning workers | 693 | 191 (28) | 1.60 (1.03, 2.47) | |||||||||
| Worksite: | Hospitals | Kitchens | Laboratories | Outdoor Areas | ||||||||
| In the last 12 months | ||||||||||||
| No2 | 515 | 135 (26) | 1.59 (1.03, 2.47) | 532 | 146 (27) | 1.61 (1.04, 2.50) | 572 | 157 (27) | 1.61 (1.04, 2.49) | 525 | 137 (26) | 1.51 (0.97, 2.36) |
| Yes | 178 | 56 (31) | 1.62 (1.00, 2.62) | 161 | 45 (28) | 1.54 (0.94, 2.53) | 121 | 34 (28) | 1.53 (0.91, 2.57) | 168 | 54 (32) | 1.85 (1.16, 2.96) |
| Frequency in the last month3 |
||||||||||||
| <1 day/week | 58 | 15 (26) | 1.04 (0.51, 2.13) | 88 | 24 (27) | 1.33 (0.68, 2.61) | 54 | 17 (31) | 1.72 (0.85, 3.48) | 56 | 20 (36) | 1.97 (1.07, 3.64) |
| 1-3 days/week | 15 | 6 (40) | 1.70 (0.61, 4.73) | 32 | 10 (31) | 1.29 (0.53, 3.11) | 15 | 4 (27) | 1.32 (0.40, 4.31) | 33 | 15 (45) | 2.20 (1.20, 4.05) |
| 4+ days/week | 105 | 35 (33) | 1.19 (0.63, 2.26) | 41 | 11 (27) | 1.25 (0.56, 2.77) | 52 | 13 (25) | 1.26 (0.57, 2.78) | 79 | 19 (24) | 1.52 (0.80, 2.86) |
| Worksite: | Other Healthcare Settings | Private Homes | Residential Building Common Areas | Schools | ||||||||
| In the last 12 months | ||||||||||||
| No | 612 | 164 (27) | 1.59 (1.03, 2.46) | 480 | 130 (27) | 1.59 (1.03, 2.45) | 499 | 131 (26) | 1.53 (0.98, 2.37) | 537 | 138 (26) | 1.53 (0.98, 2.38) |
| Yes | 81 | 27 (33) | 1.64 (0.97, 2.78) | 213 | 61 (29) | 1.64 (1.01, 2.66) | 194 | 60 (31) | 1.77 (1.11, 2.84) | 156 | 53 (34) | 1.84 (1.15, 2.93) |
| Frequency in the last month |
||||||||||||
| <1 day/week | 41 | 13 (32) | 1.60 (0.75, 3.40) | 79 | 16 (20) | 1.22 (0.62, 2.41) | 66 | 16 (24) | 1.63 (0.87, 3.05) | 64 | 14 (22) | 1.28 (0.64, 2.56) |
| 1-3 days/week | 10 | 4 (40) | 1.66 (0.55, 4.97) | 78 | 32 (41) | 2.40 (1.30, 4.45) | 62 | 19 (31) | 1.74 (0.89, 3.37) | 9 | 3 (33) | 2.48 (0.98, 6.29) |
| 4+ days/week | 30 | 10 (33) | 1.50 (0.72, 3.11) | 56 | 13 (23) | 1.29 (0.61, 2.73) | 66 | 25 (38) | 2.08 (1.17, 3.67) | 83 | 36 (43) | 2.05 (1.14, 3.69) |
| Work Activities: | Construction/renovation Clean-up | Floor Cleaning | Street/sidewalk Clean-up | Window Cleaning | ||||||||
| In the last 12 months | ||||||||||||
| No | 533 | 134 (25) | 1.50 (0.96, 2.34) | 602 | 166 (28) | 1.60 (1.03, 2.48) | 631 | 173 (27) | 1.62 (1.05, 2.50) | 440 | 122 (28) | 1.65 (1.06, 2.57) |
| Yes | 160 | 57 (36) | 1.87 (1.18, 2.95) | 91 | 25 (27) | 1.58 (0.91, 2.75) | 62 | 18 (29) | 1.44 (0.80, 2.58) | 253 | 69 (27) | 1.50 (0.94, 2.40) |
| Frequency in the last month |
||||||||||||
| <1 day/week | 92 | 32 (35) | 1.54 (0.88, 2.69) | 50 | 17 (34) | 1.53 (0.73, 3.21) | 34 | 9 (26) | 1.21 (0.49, 2.95) | 67 | 17 (25) | 1.50 (0.82, 2.75) |
| 1+ days/week | 68 | 25 (37) | 1.76 (0.98, 3.13) | 41 | 8 (20) | 0.86 (0.39, 1.89) | 28 | 9 (32) | 1.30 (0.53, 3.18) | 186 | 52 (28) | 1.75 (0.99, 3.08) |
Adjusted for age, country of birth, sex, history of eczema or other skin allergy, cleaning one’s own home, and frequency of rubber glove use
Current cleaning workers who did not work at the worksite or perform the work activity in the last 12 months are included in each model. Effect estimates presented for the population not cleaning the specific worksite within the last 12 months may be interpreted as the effect of currently cleaning, but not cleaning at the specific worksite. The primary outcome of interest generated using these models is the prevalence of hand dermatitis among current cleaning workers who reported cleaning the specific worksite in the last 12 months relative to the prevalence of hand dermatitis in the comparison population.
Frequency in the last month among respondents who reported working at the worksite or performing the work activity in the last 12 month
Current cleaning workers cleaning outdoor areas (PR: 1.85, 95% CI: 1.16, 2.96), residential building common areas (PR: 1.77, 95% CI: 1.11, 2.84), and schools (PR: 1.84, 95% CI: 1.15, 2.93), and those who reported cleaning up at a construction or renovation site (PR: 1.87, 95% CI: 1.18, 2.95) each reported significantly higher prevalences of hand dermatitis compared to the comparison group (Table 3). For respondents working in residential building common areas in the last 12 months, we observed a monotonic increase in the prevalence of hand dermatitis from 24% among those who reported cleaning in a residential building common area <1 day per week to 38% among those working 4+ days per week. A similar increase was observed among respondents working in schools in the last 12 months, where the prevalence of hand dermatitis increased from 22% to 43%.
Among current cleaning workers using specific cleaning products, the highest prevalences of hand dermatitis were observed among those who reported moderately frequent (1-3 days per week) use of hydrochloric acid (prevalence: 40%, PR: 1.73, 95% CI: 1.00, 2.99) and dust mop products (prevalence: 36%, PR: 2.12, 95% CI: 1.22, 3.71) (Table 4). Relative to the comparison group, elevated PRs were also observed among current cleaning workers who reported frequent (4+ days per week) use of products, including ammonia (PR: 2.22, 95% CI: 1.26, 3.91), bleach (PR: 2.02, 95% CI: 1.20, 3.39), multi-use cleaning products (PR: 2.24, 95% CI: 1.31, 3.83), and perfumed products (PR: 1.96, 95% CI: 1.18, 3.26). The use of degreasing agents was associated with hand dermatitis, regardless of how frequently it was used (<1 day per week: PR: 1.79, 95% CI: 1.04, 3.05; 1-3 days per week: PR: 1.85, 95% CI: 1.08, 3.14; 4+ days per week: PR: 1.77, 95% CI: 1.04, 3.02). When we assessed the role of using multiple products, we found a monotonic increase in the prevalence of hand dermatitis with increasing numbers of different cleaning products (Table 5).
Table 4.
Associations of cleaning product use in the last 12 months and frequency of use in the last month with hand dermatitis
| No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Comparison Population | 125 | 22 (18) | 1.00 | |||||||||
| Cleaning Product: | Air Fresheners | Ammonia | Bleach | Degreasing Agents | ||||||||
| Used in the last 12 months | ||||||||||||
| No2 | 351 | 98 (28) | 1.65 (1.06, 2.56) | 406 | 110 (27) | 1.62 (1.05, 2.51) | 160 | 38 (24) | 1.53 (0.97, 2.42) | 240 | 61 (25) | 1.59 (1.02, 2.49) |
| Yes | 342 | 93 (27) | 1.52 (0.96, 2.41) | 287 | 81 (28) | 1.55 (0.97, 2.49) | 533 | 153 (29) | 1.66 (1.04, 2.66) | 453 | 130 (29) | 1.61 (1.02, 2.54) |
| Frequency of use in the last month3 |
||||||||||||
| <1 day/week | 68 | 16 (24) | 1.28 (0.68, 2.41) | 107 | 23 (21) | 1.26 (0.68, 2.35) | 77 | 17 (22) | 1.51 (0.82, 2.78) | 110 | 29 (26) | 1.79 (1.04, 3.05) |
| 1-3 days/week | 74 | 18 (26) | 1.28 (0.67, 2.46) | 64 | 19 (30) | 2.03 (1.09, 3.80) | 89 | 25 (28) | 1.69 (0.96, 2.99) | 135 | 44 (33) | 1.85 (1.08, 3.14) |
| 4+ days/week | 200 | 58 (29) | 1.44 (0.84, 2.46) | 116 | 39 (34) | 2.22 (1.26, 3.91) | 367 | 111 (30) | 2.02 (1.20, 3.39) | 208 | 57 (27) | 1.77 (1.04, 3.02) |
| Cleaning Product: | Dust Mop Products | Glass Cleaners | Hydrochloric Acid | Multi-use Products | ||||||||
| Used in the last 12 months | ||||||||||||
| No | 347 | 86 (25) | 1.49 (0.95, 2.33) | 274 | 77 (28) | 1.68 (1.09, 2.61) | 389 | 83 (21) | 1.41 (0.90, 2.21) | 258 | 66 (26) | 1.60 (1.03, 2.49) |
| Yes | 346 | 105 (30) | 1.75 (1.11, 2.75) | 419 | 114 (27) | 1.50 (0.94, 2.38) | 304 | 108 (36) | 1.92 (1.22, 3.02) | 435 | 125 (29) | 1.60 (1.00, 2.55) |
| Frequency of use in the last month |
||||||||||||
| <1 day/week | 73 | 15 (21) | 1.46 (0.78, 2.74) | 64 | 16 (25) | 1.67 (0.90, 3.09) | 120 | 38 (32) | 1.59 (0.92, 2.75) | 66 | 13 (20) | 1.56 (0.78, 3.11) |
| 1-3 days/week | 74 | 27 (36) | 2.12 (1.22, 3.71) | 107 | 28 (26) | 1.49 (0.83, 2.70) | 87 | 35 (40) | 1.73 (1.00, 2.99) | 89 | 21 (24) | 1.53 (0.86, 2.73) |
| 4+ days/week | 199 | 63 (32) | 1.93 (1.16, 3.19) | 248 | 70 (28) | 1.75 (1.04, 2.97) | 97 | 35 (36) | 1.54 (0.87, 2.73) | 280 | 91 (33) | 2.24 (1.31, 3.83) |
| Cleaning Product: | Perfumed Products (e.g., pine) | Polishes or Waxes | Rug or Carpet Cleaners | Solvents | ||||||||
| Used in the last 12 months | ||||||||||||
| No | 310 | 77 (25) | 1.54 (0.98, 2.40) | 504 | 137 (27) | 1.56 (1.00, 2.42) | 576 | 160 (28) | 1.61 (1.04, 2.50) | 553 | 155 (28) | 1.62 (1.05, 2.52) |
| Yes | 383 | 114 (30) | 1.67 (1.06, 2.64) | 189 | 54 (29) | 1.71 (1.06, 2.75) | 117 | 31 (26) | 1.51 (0.90, 2.54) | 140 | 36 (26) | 1.51 (0.92, 2.49) |
| Frequency of use in the last month |
||||||||||||
| <1 day/week | 58 | 12 (21) | 1.48 (0.77, 2.85) | 236 | 53 (22) | 1.85 (1.01, 3.40) | 52 | 17 (33) | 1.52 (0.78, 2.97) | 77 | 24 (31) | 1.55 (0.85, 2.80) |
| 1-3 days/week | 71 | 19 (27) | 1.60 (0.88, 2.92) | 37 | 10 (27) | 1.81 (0.83, 3.94) | 27 | 3 (11) | 0.56 (0.19, 1.63) | 20 | 6 (30) | 1.43 (0.58, 3.49) |
| 4+ days/week | 254 | 83 (33) | 1.96 (1.18, 3.26) | 74 | 21 (28) | 1.69 (0.91, 3.15) | 38 | 11 (29) | 1.06 (0.49, 2.31) | 43 | 6 (14) | 0.78 (0.34, 1.81) |
Adjusted for age, country of birth, sex, history of eczema or other skin allergy, cleaning one’s own home, and frequency of rubber glove use
Current cleaning workers who did not work at the worksite or perform the work activity in the last 12 months are included in each model. Effect estimates presented for the population not cleaning the specific worksite within the last 12 months may be interpreted as the effect of currently cleaning, but not cleaning at the specific worksite. The primary outcome of interest generated using these models is the prevalence of hand dermatitis among current cleaning workers who reported cleaning the specific worksite in the last 12 months relative to the prevalence of hand dermatitis in the comparison population.
Frequency in the last month among respondents who reported working at the worksite or performing the work activity in the last 12 month
Table 5.
Associations of the numbers of different types of worksites and different cleaning products used in the last 12 months with hand dermatitis
| No. | No. (%) with Hand Dermatitis |
PR (95% CI)1 | |
|---|---|---|---|
| Comparison Population | 125 | 22 (18) | 1.00 |
| Number of worksite types | |||
| 0 | 159 | 31 (19) | 1.38 (0.85, 2.25) |
| 1-2 | 327 | 97 (30) | 1.69 (1.06, 2.69) |
| 3+ | 207 | 63 (30) | 1.80 (1.11, 2.93) |
| Number of different cleaning products |
|||
| 0 | 86 | 20 (23) | 1.62 (0.97, 2.72) |
| 1-3 | 95 | 23 (24) | 1.41 (0.81, 2.45) |
| 4-6 | 225 | 58 (26) | 1.50 (0.91, 2.48) |
| 7+ | 287 | 90 (31) | 1.70 (1.04, 2.75) |
Adjusted for age, country of birth, sex, history of eczema or other skin allergy, cleaning one’s own home, and frequency of rubber glove use
When we compared the our survey-based definition of hand dermatitis to classifications based upon dermatologic assessments in the sub-sample of 70, we found that using the questionnaire-based definition, 49% of the sample was classified as having hand dermatitis, compared to 24% when the definition was based on a dermatologic exam. Using the dermatologic exam classification as the gold standard, analysis of our questionnaire-based definition generated a sensitivity of 0.82, a specificity of 0.62, a positive predictive value of 0.41, and a negative predictive value of 0.92.
Discussion
In these data, we identified groups of professional cleaning workers at elevated risk of hand dermatitis. Our results extend previous findings from epidemiological research into skin symptoms in cleaning workers (6-8) by reporting elevated risk among users of products known to affect the respiratory tract and skin (1,15,16), providing evidence of increased risk with increasing frequency of use, and generating increasing adjusted risk estimates for workers performing a variety of tasks or using multiple products.
In addition to occupational risk factors, individual susceptibility factors such as atopy also play a role in the prevalence of hand dermatitis. In our data, hand dermatitis was reported among 39% more respondents with than without a self-reported history of eczema or other skin allergy. Although our data collection was not designed to distinguish allergic from irritant dermatitis, the cleaning-related exposures we identify here should be considered risk factors for both. Indeed, the adjusted PR generated for current cleaning overall was similar to those generated when the data were stratified by self-reported history of eczema or other skin allergy (history: PR: 1.61; no history: PR: 1.57). These estimates and stratified estimates for particular worksites and cleaning products (not shown) do not provide sufficient evidence of effect modification by atopic status; improved information about atopy and a larger sample size would have allowed us to more thoroughly explore this potential effect modification. Nonetheless, any interventions aimed at reducing these exposures will likely reduce the burden of hand dermatitis among atopic and non-atopic individuals.
The validity statistics of our symptom-based definition of hand dermatitis compared to the dermatologic evaluation are similar to those published previously (17,18) and suggest that outcome misclassification may affect our results. In particular, our definition may overestimate the prevalence of current hand dermatitis in this population. This is unsurprising, given that our survey-based definition assessed 12-month prevalence, whereas the in-person dermatologic assessment was based on symptoms visible at the time of the exam. Indeed, some participants described improvement of their symptoms at the time of the physical examination. Still, these data provide strong evidence of the importance of hand dermatitis symptoms among cleaning workers. That over 41% of the respondents who indicated that they had hand dermatitis in the last 12 months had their symptoms observed by our study team suggests a high incidence of hand dermatitis or a long duration of the condition in this population.
Our findings should be interpreted with particular attention to the low survey response rate in the EPIASLI2 study. Limitations of our study design and response rate as well as a comparison of demographic characteristics of study participants and non-responders in two cleaning companies have been described elsewhere (9). If participation in our study was associated with symptom status differently among current and non-current cleaners, then our results may be affected by a response bias. A comparison of demographic characteristics at two cleaning companies by responder status did not reveal major differences between the two populations with respect to the characteristics evaluated (9). Because of laws protecting personally identifying information in Spain (19), our study did not include any additional data collection from non-responders and we are not able to draw further conclusions about differences between responders and non-responders (9). The prevalence of hand dermatitis reported by participants in our study is within the range of those reported among cleaners (5,8,20) and general population-based samples (21,22), suggesting that despite the low response rate, the external validity of our data is not markedly compromised. In fact, enrollment of 818 participants into a research study focused on the health of workers in large and diverse industry in which investigator and other public health personnel have limited workplace access is a noteworthy strength of our study.
Our findings may also be affected by the healthy worker survivor effect. If individuals who are susceptible to hand dermatitis or who have a history of hand dermatitis left their cleaning jobs, then the workplace-based recruitment in the EPIASLI2 study may have excluded individuals with cleaning-related dermatitis. If symptomatic individuals reduced the frequency of their use of specific cleaning products, then our analysis may have incorrectly attributed their symptoms to a category of less frequent use. Similarly, if symptomatic individuals moved into administrative jobs, then our analysis incorrectly attributes their cleaning-related symptoms to the comparison population. If our study were affected in this way by the healthy worker effect, then our results may underestimate the actual burden of professional cleaning on skin health.
Symptoms of hand dermatitis may have long-term consequences for employment, economics, and quality of life (23-26). Our findings indicate potential opportunities for reducing hand dermatitis among professional cleaning workers; namely, the feasibility of performing cleaning work using fewer cleaning products and less hazardous products should be evaluated. Other aspects of cleaning work that may affect the risk of hand dermatitis include cleaning techniques, product mixtures and dilutions, methods used to mix and dilute products, and safety training. Improved information about the types of gloves used, hours of use, and number of glove changes would provide valuable information about the role of gloves in protecting workers during specific activities and from individual products and product mixtures.
In conclusion, our findings support the hypothesis that cleaning work is a risk factor for hand dermatitis and that professional cleaning workers may not be sufficiently protected from dermal hazards at work. Occupational medicine and dermatology specialists, and others who provide healthcare to professional cleaners, should be aware of the health risks of performing cleaning work and handling specific cleaning products and the possibility that symptoms of cutaneous disease are underreported and underdiagnosed in this population. These results justify further evaluation of primary prevention methods to reduce cleaning workers’ contact with dermal hazards.
Acknowledgments
Funding: This research was funded by a Fondo de Investigación Sanitaria grant from the Instituto de Salud Carlos III (Grant Number FIS PI06/1378). The authors also acknowledge partial funding from CIBER Epidemiología y Salud Pública (CIBERESP), Spain. M. Mirabelli received funding from the US National Institutes of Health (Grant Number 1F32ES014142).
Footnotes
Conflicts of Interest: The authors have declared no conflicts of interest.
Author Contributions: Each author participated sufficiently to take responsibility for the work. The study was conceptualized and designed by JM Antó, E Barreiro, MC Mirabelli, R Orriols, D Vizcaya, and JP Zock. The data were collected and managed by D Vizcaya, L Arjona, and A Martí Margarit. Data analyses were performed by M Mirabelli (statistical analysis), and A Martí Margarit and A Gimenez-Arnau (dermatologic images). The manuscript was drafted and revised by M. Mirabelli. All authors contributed to the interpretation of the data, reviewed and revised the manuscript for intellectual content, and approved the final manuscript.
References
- 1.Zock JP. World at work: cleaners. Occup Environ Med. 2005;62:581–4. doi: 10.1136/oem.2004.015032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bureau of Labor Statistics U.S.Department of Labor . Building Cleaning Workers. U S Department of Labor; Occupational Outlook Handbook, 2010-11 Edition. [Google Scholar]
- 3.Magnano M, Silvani S, Vincenzi C, Nino M, Tosti A. Contact allergens and irritants in household washing and cleaning products. Contact Dermatitis. 2009;61:337–41. doi: 10.1111/j.1600-0536.2009.01647.x. [DOI] [PubMed] [Google Scholar]
- 4.Jungbauer FH, van der Harst JJ, Schuttelaar ML, Groothoff JW, Coenraads PJ. Characteristics of wet work in the cleaning industry. Contact Dermatitis. 2004;51:131–4. doi: 10.1111/j.0105-1873.2004.00421.x. [DOI] [PubMed] [Google Scholar]
- 5.Nielsen J. The occurrence and course of skin symptoms on the hands among female cleaners. Contact Dermatitis. 1996;34:284–91. doi: 10.1111/j.1600-0536.1996.tb02201.x. [DOI] [PubMed] [Google Scholar]
- 6.Gawkrodger DJ, Lloyd MH, Hunter JA. Occupational skin disease in hospital cleaning and kitchen workers. Contact Dermatitis. 1986;15:132–5. doi: 10.1111/j.1600-0536.1986.tb01312.x. [DOI] [PubMed] [Google Scholar]
- 7.Hansen KS. Occupational dermatoses in hospital cleaning women. Contact Dermatitis. 1983;9:343–51. doi: 10.1111/j.1600-0536.1983.tb04426.x. [DOI] [PubMed] [Google Scholar]
- 8.Lynde CB, Obadia M, Liss GM, Ribeiro M, Holness DL, Tarlo SM. Cutaneous and respiratory symptoms among professional cleaners. Occup Med (Lond) 2009;59:249–54. doi: 10.1093/occmed/kqp051. [DOI] [PubMed] [Google Scholar]
- 9.Vizcaya D, Mirabelli MC, Antó JM, Orriols R, Burgos F, Arjona L, Zock JP. A workforce-based study on occupational exposures and asthma symptoms in cleaning workers. Occup Env Med. 2011 May 10; doi: 10.1136/oem.2010.063271. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 10.Vizcaya D, Arjona L, Orriols R, Barreiro E, Perez M, Burgos F, Gómez F, Antó JM, Zock JP. Characteristics of asthma in cleaning workers: A case-control study. Annual Congress of the European Respiratory Society Meeting Barcelona, Spain Eur Resp J. 2010;36:457s. [Google Scholar]
- 11.Coenraads PJ, Nater JP, van der Lende R. Prevalence of eczema and other dermatoses of the hands and arms in the Netherlands. Association with age and occupation. Clin Exp Dermatol. 1983;8:495–503. doi: 10.1111/j.1365-2230.1983.tb01816.x. [DOI] [PubMed] [Google Scholar]
- 12.Smit HA, Coenraads PJ, Lavrijsen AP, Nater JP. Evaluation of a self-administered questionnaire on hand dermatitis. Contact Dermatitis. 1992;26:11–6. doi: 10.1111/j.1600-0536.1992.tb00861.x. [DOI] [PubMed] [Google Scholar]
- 13.Thompson ML, Myers JE, Kriebel D. Prevalence odds ratio or prevalence ratio in the analysis of cross sectional data: what is to be done? Occup Environ Med. 1998;55:272–7. doi: 10.1136/oem.55.4.272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Behrens T, Taeger D, Wellmann J, Keil U. Different methods to calculate effect estimates in cross-sectional studies. A comparison between prevalence odds ratio and prevalence ratio. Methods Inf Med. 2004;43:505–9. [PubMed] [Google Scholar]
- 15.Jaakkola JJ, Jaakkola MS. Professional cleaning and asthma. Curr Opin Allergy Clin Immunol. 2006;6:85–90. doi: 10.1097/01.all.0000216849.64828.55. [DOI] [PubMed] [Google Scholar]
- 16.English JSC. Current concepts of irritant contact dermatitis. Occup Environ Med. 2004;61:722–6. doi: 10.1136/oem.2003.010710. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Vermeulen R, Kromhout H, Bruynzeel DP, de Boer EM. Ascertainment of hand dermatitis using a symptom-based questionnaire; applicability in an industrial population. Contact Dermatitis. 2000;42:202–6. doi: 10.1034/j.1600-0536.2000.042004202.x. [DOI] [PubMed] [Google Scholar]
- 18.van Wendel de Joode B, Vermeulen R, Heederik D, van Ginkel K, Kromhout H. Evaluation of two self-administered questionnaires to ascertain dermatitis among metal workers and its relation with exposure to metal working fluids. Contact Dermatitis. 2007;57:14–20. doi: 10.1111/j.1600-0536.2007.01127.x. [DOI] [PubMed] [Google Scholar]
- 19.Ley Orgánica 15/1999, de 13 de diciembre, de Protección de Datos de Carácter Personal. 1999.
- 20.Singgih SI, Lantinga H, Nater JP, Woest TE, Kruyt-Gaspersz JA. Occupational hand dermatoses in hospital cleaning personnel. Contact Dermatitis. 1986;14:14–9. doi: 10.1111/j.1600-0536.1986.tb01146.x. [DOI] [PubMed] [Google Scholar]
- 21.Thyssen JP, Johansen JD, Linneberg A, Menne T. The epidemiology of hand eczema in the general population--prevalence and main findings. Contact Dermatitis. 2010;62:75–87. doi: 10.1111/j.1600-0536.2009.01669.x. [DOI] [PubMed] [Google Scholar]
- 22.Diepgen TL, Coenraads PJ. The epidemiology of occupational contact dermatitis. Int Arch Occup Environ Health. 1999;72:496–506. doi: 10.1007/s004200050407. [DOI] [PubMed] [Google Scholar]
- 23.Meding B, Swanbeck G. Consequences of having hand eczema. Contact Dermatitis. 1990;23:6–14. doi: 10.1111/j.1600-0536.1990.tb00076.x. [DOI] [PubMed] [Google Scholar]
- 24.Meding B, Wrangsjö K, Järvholm B. Fifteen-year follow-up of hand eczema: persistence and consequences. Br J Dermatol. 2005;152:975–80. doi: 10.1111/j.1365-2133.2005.06494.x. [DOI] [PubMed] [Google Scholar]
- 25.Mälkönen T, Alanko K, Jolanki R, Luukkonen R, Aalto-Korte K, Lauerma A, Susitaival P. Long-term follow-up study of occupational hand eczema. Br J Dermatol. 2010;163:999–1006. doi: 10.1111/j.1365-2133.2010.09987.x. [DOI] [PubMed] [Google Scholar]
- 26.Cvetkovski RS, Zachariae R, Jensen H, Olsen J, Johansen JD, Agner T. Prognosis of occupational hand eczema: a follow-up study. Arch Dermatol. 2006;142:305–11. doi: 10.1001/archderm.142.3.305. [DOI] [PubMed] [Google Scholar]