According to the estimate of WHO, occupational diseases and injuries constitute still a considerable burden on workers, employers and societies at large around the world. The occurrence of occupational diseases and injuries depends, to a large extent, on the working conditions to which workers are exposed. In many countries these conditions have undergone enormous changes in the past decennia leading to new occupational health and safety problems. This implies that there is still a lot to do for occupational health professionals in most places in the world. Occupational health physicians and nurses treat and counsel workers for a variety of occupational and public health issues. In many countries, there are multidisciplinary occupational health services as a support for employers and workers to better tackle work related health and safety problems. In addition, there are professionals at work to enforce legislation and to make new regulations. In short, there is a plethora of activities going on to maintain and improve health and safety at work. In general, over the past decades, this has led to a reduction in fatalities and exposures that are hazardous to health. Studies of time trends of exposure within industries show that enormous reductions in exposure have been realised (1). However, it is difficult to tell what exactly was at the basis of these reductions because many things happened at the same time. No further reductions in exposure have been achieved in the past decade because it is probably more difficult to make progress at lower levels of exposure.

Therefore, it is important to find out what works well to be able to maintain and further improve healthy and safe workplaces for all workers.

How can we decide about what works?

For individual practitioners it is not easy to be sure what is effective in practice. It is well known that we are subject to many cognitive biases that blur our observations. We like to believe that what we do is beneficial and we are therefore inclined to look for and find confirmation for this belief.(2) However, unless the effects of what we do are very big, which usually they are not, it is difficult to be sure about what works solely based on individual observations in practice. The approach of evidence-based medicine has been developed to better underpin decisions about what works with evidence from research that is guarded against the biases mentioned above.(3) This type of evidence helps to make our decisions more transparent and rational, it will improve the quality of our services and in the end it will result in better health and more safety for workers.

The approach would be hopeless if every practitioner would have to search for their own research evidence. However, a lot of work has been done already and it is compiled in evidence-based guidelines and overviews of the literature. In the past 20 years, numerous authors from around the world have put remarkable effort in reviewing the literature on what works in health care as part of the Cochrane Collaboration. The Cochrane Collaboration is an international network of health professionals that aims to find out what works in health care (www.cochrane.org). The Cochrane Occupational Safety and Health Review Group now hosts about 80 reviews of what works in our field (osh.cochrane.org). The topics range from regulation in construction industry to vaccination for hepatitis B and drug therapy to improve return to work in depressed workers.

The following provides some examples of these Cochrane Reviews. Annika Parantainen from Finland and colleagues from the US concentrated on the use of blunt needles for surgeons compared to the usual sharp suture needles for preventing needle stick injuries.(4) They concluded that the risk of sustaining a needle stick injury more than halves when surgeons make use of blunt needles. Even though the use of this type of needle requires some adaptation, most surgeons found them more than acceptable. Apparently, this is an intervention that works and should be implemented. With another team of authors, we looked at what has been reported on the effect of training workers in lifting techniques.(5) We found nine randomised and nine cohort studies that compared the effect of training to either no intervention or a minor control condition. None of the studies found an appreciable difference in back pain between the group that received training and the control group. The likelihood that this was due to chance was small, because the total of the studies included more than 20.000 workers. It seems that this is an intervention that does not work and that we have to develop other measures if we want to prevent back pain. Health care staff is especially prone to the consequences of stress as their work load is often high and the possibilities to control this work load are few. In a review of stress management interventions, cognitive-behavioural interventions reduced the reporting of stress symptoms with about 25% especially when combined with relaxation techniques.(6) Also here, we can conclude that this is an intervention that works. Yet another review team from Nigeria, Finland and Canada was interested in the effect of behavioural interventions for preventing risky sexual behaviour as is prevalent in occupations like truck drivers and military personnel.(7) They found eight randomised studies that showed that the influence of worker peers positively influenced preventive behaviour and that risky sexual behaviour was decreased to a moderate degree. We can conclude that these programmes are feasible and probably work but that their efficiency should be improved.

Challenges in finding out what works

One of the challenges is to cope with the evidence that shows no effect. Especially when one has been engaged in a certain professional practice for a long time, it is difficult to accept that this is not effective and is better to be abandoned. This can easily lead to nihilism and the idea that then nothing is worth doing. However, with a more constructive approach it is also possible to see that this opens up possibilities for new and innovative interventions that should be evaluated before being implemented into practice.

Even though all occupational safety and health professionals have a common interest in finding out what works, a common language and understanding of how to do this is still lacking. For example, the method of systematically reviewing the literature is less well developed in safety science and occupational hygiene.

Since there is no central funding available for Cochrane groups or authors, we are dependent on local national funding and voluntary contributions.

Conclusions

Authors affiliated with the Cochrane Occupational Safety and Health Review Group have done a tremendous job in locating and summarising evidence on what works to make workplaces healthier and safer. Since the work has only recently started, there are good prospects that there will be more systematic reviews available in the years to come. These reviews will help us to efficiently tackle the problems in occupational health and safety also in the 21st century.