1. Why does exposure generally decrease over time? John W Cherrie, Karen Galea, Hilary Cowie, Joanne Crawford, Martie van Tongeren Other collaborators: DT While, H Kromhout, A Agostini, F de Vocht, A Bolton, M Graham, AJ Soutar, J Tickner
13. The Theory of Planned Behaviour Behavioural beliefs Attitude towards the behaviour Normative beliefs Subjective norms Control beliefs Perceived behavioural control http://people.umass.edu/aizen/tpb.diag.html Intention Behaviour Actual behavioural control
Thank you. This presentation has grown out of work we have been doing to quantify time trends in exposure.
I will show some of our earlier work on time trends, discuss some of the environmental and other factors that are responsible for educing exposure, outline a theoretical construct that may help understand the reasons for changes and, finally, I will discuss three quite different cases.
In a review of published evidence for temporal trends that we published in 2007 we identified 38 cases where there was informative data for aerosols. We analyzed the temporal trends on the log-scale assuming an exponential decline in exposure level over time. 58% of these involving aerosols there was a significant reduction in exposure, typically between 5% and 10% per year. Only one dataset (3%) showed a significant increase.
For gases and vapours there was a similar pattern, in this case with 56% of cases showing a significant decrease in exposure – none showed a significant increase. Overall the median decrease in level was 7% per annum.
In an individual plant it might be assumed that over time there are periodic changes to the production process that give rise to a reduced exposure. For example in this case there are three changes in the second, third and seventh years, in each case exposure being reduced by at most 50%, although as time goes on the reduction gets smaller to reflect the possibility that it gets more difficult to control exposures. We can further explore the impact of changes over time in a group of plants with a Monte Carlo simulation – in this case we allow the interventions to take place at random with a decision each quarter – 10% chance and on each occasion a 50% chance that it would reduce exposure. The initial level being set at random between 10 and 300 mg/m3.
With five plants you see quite a variable pattern of intervention. On average about one intervention every two years in a plant.
Ten plants
By the time you have 25 plants the average exposure declines exponentially, in this case with a decrease of about 11% per year. Clearly to achieve the level of change that we noted in the review you do not need anything other than a low level of modest interventions.
Of course not every situation presents a decreasing profile and perhaps the most striking example that we have identified is for flour dust in Britain, where we find a complete absence of any temporal trends looking at data covering about the last 25-years.
At the other extreme is data from the PVC plants in the early 1970s when the risk of angiosarcoma of the liver was first identified as a consequence of high VCM exposure. Here over a about a year the exposure dropped by more that an order of magnitude. Interestingly, although there was not systematic monitoring data for the period prior to the dramatic change the evidence is that there were improvements taking place before this which would have given a series of modest reductions in exposure and after about 1975 the rate of decline in exposure again dropped to about 6% per annum. To get this kind of dramatic reduction in exposure in the MC simulations you need to increase the inclination to 90% with a 50% success rate although still retaining a modest reduction in exposure each time.
The changes that were taking place around this time were partly improvements in the process, for example adding scale inhibitors to reduce the adhesion of PVC to the autoclave walls and partly retrofitted engineering controls, such as local ventilation, or PPE such as BA.
Now, the Theory of Planned Behaviour is a successful approach used to understand individual behaviour. It is based on the assumption that behaviour is directly linked to our expressed intention to act in a given situation, of course influenced by our abilities (the actual behavioural control) – there is no possibility that despite my intention to be a successful concert pianist that I will ever be able to play the piano. In addition, according to the Theory there are three other areas of our attitudes and beliefs that influence our intention to act – control beliefs and our perceptions of the ease or difficulty in performing that action (here control is in the sense of control over the behaviour). Subjective norms relate to our perceptions of the social pressures to act , and beliefs about the consequences of a behaviour and outcome evaluations in terms of the desirability of the behaviour .
We argue that TPB in relation to control of exposure should be targeted at those who control the process and that the behaviour we are interested in is any steps taken to improve the process, which may or may not be directly aimed at reducing exposure. For example, in the case of the VCM exposure in PVC plants the introduction of scale inhibitors and pressure washing was happening before it was clear there was a cancer risk with the purpose of making the process more efficient by cutting down the amount of time the operators would need to spend inside the vessel scraping off PVC residues. TPB relies of questionnaires to collect information about attitudes and beliefs and you can see the start of a questionnaire that we are developing.
Returning to our investigative studies, we have carried out interviews with process managers in about 30 companies where we collected data to quantify time trends in exposure. Here the average exposure change was between about 4 and 8% average decrease per year. These data were obtained by revisiting companies that had previously been investigated ten or more years previous. Note, we identified a greater reduction in toluene in our routinely collected data but this was probably as a consequence of removal of toluene from products rather than company specific changes.
In our companies there were several factors that were important including retrofitting local controls, as for example in this situation where a slot ventilation system was added to a mixing machine.
From our questions we identified several reasons for change, but we have grouped these into economic, legislation-related and safety culture related. For example, …MORE
It can be seen that legislative explanations predominated, followed by economic factors and the safety culture. However, the relative importance of these differed between sectors with legislative being most important in quarries, economic in wood and safety culture in industries using solvents containing toluene. This data is helping to guide us in the development of a TPB questionnaire to explore these factors more widely.
Although we do not yet have data from the TPB questionnaire we can use the approach to think about some of the examples we’ve described. For example, in situations where flour dust exposure occurs in Britain we believe that prevailing view is that… Wood, toluene and quarry dust are all examples of what we’ve characterized as “business as usual” where… The most extreme example of change that we have encountered is VCM exposure, where … (Also EtO, asbestos…). The most important aspect of this type of analysis is the indication of how we might best approach changing attitudes and beliefs in other situations to ensure that exposures are decreased – flour dust is an example, but we should also consider other situations where little change has been occurring, for example diesel exhaust particulate, radon, paint exposure etc. MORE