© University of Aberdeen

The effectiveness of barrier creams in protecting against dermatitis: a pilot study

 Matthew Parkinson and John W Cherrie ^{1, 2}

1. University of Aberdeen, Department of Environmental and Occupational Medicine, Forresterhill Road, Aberdeen AB25 2ZP, United Kingdom.

2. Institute of Occupational Medicine, 8 Roxburgh Place, Edinburgh EH8 9SU, United Kingdom.

Abstract

A short study was undertaken to develop a method to evaluate the effectiveness of barrier creams in the workplace. The method used a fluorescent tracer as a marker of exposure to metal working fluids. After a period of exposure the workers hands were washed. Where the skin was protected by the barrier cream the tracer was removed. The proportion of the skin area left uncovered, i.e. showing fluorescence was used as an index of effectiveness.

Workplace assessments on two people showed that exposure occurred mainly to the palms. Tests with three barrier creams and an emollient cream suggested that at best barrier creams might protect about half of the skin over a period of two hours. Further development of the method would be advantageous.

Introduction

Occupational dermatitis is a problem affecting about 85,000 workers in the UK at any one time. Irritant dermatitis is an inflammatory reaction following damage to the skin, which can arise when certain hazardous substances permeate through the stratum corneum, denature the keratin and damage the lipid structure. Irritants can affect anyone, although the development of dermatitis depends on individual susceptibility to the substance. It is possible that this condition can occur as a result of a single acute exposure or, more likely, from chronic exposure over a period of time.

In an attempt to prevent irritant contact dermatitis in the workplace, many industries provide their workers with ‘barrier creams’, also referred to as ‘protective ointments’, ‘protective creams’ or ‘invisible gloves’. Barrier creams are often employed as substitutes for protective clothing where gloves and sleeves cannot be safely or efficiently used. The technique has been around for some time. In 1915, a general practitioner from Wigan, England, Dr R. Prosser White wrote that it was ‘necessary that men’s clothes and skin should be protected by overalls and suitable covering. Any cutaneous surfaces that have become soiled should be cleansed as soon as possible. To assist in this, some bland, insoluble ointment has been advised to be rubbed into the exposed surfaces prior to work. The quantity used should not be large but enough to block up the stomata of the skin’ (Cronin, 1985).

The choice of the name, ‘barrier cream’ has a great advantage in that it implies a benefit to the user, but how effective are they really? This study was undertaken in a medium sized engineering company to address this question. The aims of the work were to develop a method to assess the effectiveness of barrier creams and to carry out a pilot study in a small engineering workshop where the employees were exposed to soluble cutting oils. In heavy engineering environments using metal working fluids (MWF) dermatitis prevalence rates can be as high as 30%, so if they are effective barrier creams could be of great benefit to the workforce.

Development of the method

The method was based on the assumption that to protect the skin a barrier cream must prevent the contaminant coming into contact with the stratum corneum. We used fluorescent tracer added to the MWF as a marker for exposure (2500 μg/ml of Fluorescent Brightener 28 dissolved in Multicut Universal Extra metalworking fluid). Subjects applied a barrier cream and then had their hands photographed under ultraviolet (UV) light.

They then worked for a period of time before they were asked to wash their hands, which were then again photographed under UV light. We assumed that the washing removed the residual barrier cream so that where the tracer adhered to the skin there had been no protection and where the skin was free of tracer the barrier cream had been effective. Effectiveness was defined as the proportion of the hand surface free from fluorescent tracer, adjusted for any initial fluorescence. It was decided that the test apparatus should be portable to enable the work to be undertaken within the workplace. We devised a small light tight box: 435mm length, 360mm width, 270mm height and 8mm thickness. One hole was cut in the box for a worker’s hand, with a rubber shield to prevent extraneous light entering and the other for a digital camera (Figure 1). Four battery-powered UV lamps (each lamp was 15cm in length with UV emission at 352nm) were attached to the underside of the box lid with Velcro attachments. The lamps were positioned in a square arrangement, each one 5cm from the central viewing hole.

Figure 1 - Test box

The digital camera used to take photographs during the study was a ‘FujiFilm MX2700’. It was equipped with a MG-16S 16MB ‘SmartMedia’ card that electronically stored the digital images. At the end of a series of tests the images were downloaded to a computer and the area of fluorescence was analysed using the Corel Photo-Paint 8 software package. The software allows the user to convert the fluorescence in the digital image to a number of white pixels on screen. The number of white pixels can be compared to the total number of available pixels (2,160,000 pixels) and the percentage fluorescence was calculated.

A number of laboratory calibration and validation checks were made, although these are not reported here.

Methods used in the workplace study

Three barrier creams and one emollient were to be used in the tests. ‘Taktosan’ and ‘Stokoderm’ were identified from a safety equipment suppliers’ catalogue. ‘Gojo’ barrier lotion was used in a local engineering company and ‘E45’ was chosen because it is a popular moisturising cream in the United Kingdom. Control tests were also carried out where the worker used no creams.

The worker applied enough barrier cream to give a complete covering of the hands and rubbed it in deeply for 1 minute. Although the barrier cream was not weighed, a similar amount was applied in all tests. The UV lights in the digital imaging dark box were warmed for 5 minutes, the digital camera was switched on and four images of the hands were taken and stored. These were the right back, right palm, left back and left palm. These images showed the amount of natural fluorescence combined with any fluorescence from the barrier cream. The worker then returned to their machine. Prior to work, the machine was sprayed with the tracer fluid at the points where the individual was likely to contact it, and where there was potential to be exposed to the metalworking fluid (it was considered impractical to add tracer to the whole MWF sump). The machines were sprayed every 30 minutes during the 2-hour tests. At the end of the sampling period the worker washed their hands for 1 minute using an unfragranced and uncoloured soap. The hands were then dried with clean, disposable paper hand towels. Their hands were placed into the dark box and photographed as before. The images were downloaded and analysed using the appropriate software.

A male worker and a female worker were involved in the study. The man worked on a boring machine and regularly forced gear levers, pulled handles and turned wheels to carry out his work. The woman worked on a lathe and turned wheels, pulled levers and touched the metal being cut. Both machines were more than 30 years old and were covered in a thin layer of grime and oil. Their jobs were different and their distinct working techniques were noted. The man was in regular contact with five areas of his machine and so each area was sprayed with the tracer fluid. The woman was in frequent contact with four areas of her machine, and each of these was sprayed as above. Both machines were sprayed in an identical manner.

Results

The results of the workplace investigation confirmed that both workers were mainly exposed on the palms of both hands. Only very small amounts of fluorescence appeared on the back of the hands. The photographs below validate this pattern of exposure.

Figure 2 - A typical image of the back of hand	Figure 3 - A typical image of the palm of the hand

The following box-plots provide further insight into where on the hands they were exposed. The central bar across the box-plot refers to the median and the whiskers refer to the largest and smallest values that are not outliers or extremes.

It is important to note that all results are normalised, to actual levels of fluorescence in relation to the maximum amount of pixels on screen. The male worker’s hand was approximately 190cm² and the female worker’s hand approximately 140cm². The total digital image area was 486cm²; therefore the normalising factors were 2.55 for the man and 3.47 for the woman.

Wilcoxon matched pairs test was used to investigate differences between hands and fronts and backs as the data was not normally distributed, outliers were present and sample sizes were small. The significance level was set at 0.05. To avoid any confusion for the reader, the outliers have been removed from the box-plots, as they were incorporated into the statistical tests.

Figure 4 - Natural Fluorescence Before Exposure to Tracer Fluid for the Male Worker

The man was left-handed and this may explain why his left palm had greater variation in fluorescence than the other three areas. The level of natural fluorescence was very low when expressed as a percentage, the maximum amount of fluorescence being 4.8% on the left palm with Gojo barrier lotion. It was clear that there was more natural fluorescence on the palms of his hands than on the back of his hands. His left palm fluorescence was significantly different from his left back fluorescence (p = 0.043), although it was not significantly different from his right palm fluorescence (p = 0.068).

Figure 5 - Fluorescence After Exposure to Tracer Fluid for the Male Worker

It was evident that this man was exposed to the Application Fluid mainly on the palms of his hands. Both the right and left palm median percentages were much higher than the back of the hand plots. They were also higher than the fluorescence levels before exposure. Although the man was left handed, he gained most exposure on his right palm. When questioned he explained that he controled the machine more often with his right hand.

For the statistical analysis of the after exposure data, the right back, right palm and left palm were significantly different from the measurements before exposure. The p-value was less than 0.05 in each case. The left back area was not significantly different after exposure (p-value 0.08) compared to before exposure. In both cases, the left and right palms were significantly different to their respective left and right backs of hands after exposure to the tracer fluid (p < 0.05). There were no significant differences between the left and right palms (p = 0.50) or the left and right backs (p = 0.14) after exposure.

The results from the female worker showed a similar pattern and are not presented here.

The box-plots show clearly that workers were principally exposed to tracer fluid (and MWF) on the palms of their hands. Their hands need adequate protection against the fluid. The question is, do the barrier creams offer any protection? The following bar charts attempt to answer this question and identify which creams provided protection and which did not.

Effectiveness of Barrier Creams

Figure 6 below illustrates how the hand appeared under the fluorescent lighting with the individual barrier cream applied. The Gojo barrier lotion had more natural fluorescence than the other barrier creams, though all were low when expressed as a percentage of the maximum possible area.

Figure 6 – Hand Fluorescence Before Exposure to Application Fluid for the Male Worker

After exposure to the tracer fluid for two hours, Figure 7 shows that the greatest fluorescence appeared on the palms of the worker’s hands. In this man’s case, it was evident that E45 provided no protection and may have even increased his exposure, i.e. the E45 fluorescence is greater than that of the control. Gojo barrier lotion appeared to give the best protection in this case. Taktosan provided the second best protection and the Stokoderm appeared slightly less effective than Taktosan.

Figure 7 – Hand Fluorescence After Exposure to Application Fluid for the Male Worker

The following Figures, 8 and 9, indicate the effectiveness of the barrier creams as compared to the control. The ratio line shows the level of protection afforded to the hands without the application of barrier creams, i.e. the control. The bars exceeding unity illustrate barrier creams that offered no protection, and may have even enhanced a worker’s dermal exposure to metalworking fluid. The bars below the ratio line demonstrate creams that offered some level of protection against it. Therefore, the smaller the bar, the more effective the barrier cream.

Figure 8 – The Effectiveness of the Creams when compared to the Control for the Male Worker

Taktosan, Gojo and Stokoderm all offered the man varying levels of protection. E45 offered him no protection.

Figure 9 – The Effectiveness of the Creams when compared to the Control for the Female Worker

Taktosan, Gojo and E45 offered the woman some protection. Stokoderm offered her no protection.

Discussion

The best performing barrier creams in the tests were Gojo barrier lotion for the male worker and Taktosan for the female worker. It should be noted that Taktosan was the second most effective cream for the man, and provided the best overall protection of all creams in the workplace study. We were initially surprised that one barrier cream was not the most effective for both individuals. Based on previous laboratory test, this was the expected result. However, when the skin of the individuals was studied, potential reasons for the difference were discovered. The male worker had large, coarse, dry hands whereas the female worker had smaller, smooth, well moisturised hands. The variation in the skin type may have affected the absorption of the barrier creams.

We consider that the results support the view that barrier creams do not provide a high level of protection. At best they appear to be able to protect about half of the skin from contact with the fluorescent tracer. However, our tests were only carried out over two hours and a longer period of exposure might be expected to show less effective coverage. Most of the barrier creams appeared to be less effective, with only 20 to 30% of the skin being protected.

Goh and Gan (1994) undertook a study in the metal industry in Singapore to compare the effect of a barrier cream and an afterwork emollient on occupational metalworking fluid dermatitis. In an earlier study they had found that about 50% of machinists employed in a ball bearing factory developed dermatitis within 6 weeks of starting work where they were using a metalworking fluid. Fifty-four machinists with healthy skin participated in the study. Twenty-three used no barrier cream or emollient cream and were classed as controls. Seventeen workers used the barrier cream and the remaining fourteen used the emollient cream. The point prevalence of cutting fluid dermatitis for the first 12 weeks of the study was 56% for the controls (no creams), 52.9% for the barrier cream group and 28.6% for the emollient group. These findings suggest that the barrier cream was ineffective in protecting against the cutting fluid dermatitis.

It must be realised that we report the results from a small pilot study involving only two workers. It is not really possible to draw reliable conclusions about the relative protection offered by these products. However, the method does appear to offer a practical method of investigating the effectiveness of barrier creams in the workplace. Further development and validation would be advantageous.

Acknowledgement

This work was generously supported by the Colt Foundation.

References

First published on www.herox.org on 14th May 2001.

Herox.org undertakes to publish any valid comments or criticisms of the papers published on this web site. Please direct any comments to publish@herox.org.

This paper has not been peer-reviewed and the publisher takes no responsibility for the scientific quality of the work.

© University of Aberdeen

Back to newsletter index

about HEROX · register · discussion group · publishing · conferences · journal club · newsletter · links · projects · jobs · home