Evaluating evidence-based occupational health

How useful is the occupational health evidence base in practice? Nic Paton asks the experts.

What should be the evidence base for occupational health interventions? What, in the context of occupational health, should even constitute “evidence”? What sort of an evidence base do occupational health practitioners need, or should they even be desiring, to do their jobs more effectively?

For OH nurses and physicians who have come through and been educated in the conventional medical and healthcare establishment, and for whom the medical journal probably remains one of their primary sources of information, the unthinking, knee-jerk response might well be: “The same as for anyone else, surely?”

Why should OH be any less reliant on the best possible available evidence than any other medical speciality? Why should it practise to any lower evidential standards? And the best possible available evidence, the “gold standard” of evidence gathering, it is widely recognised, is the randomised controlled trial (RCT).

Well, yes and no. The difficulty for both OH and public health professionals is that, when it comes to interventions, there may be little or no evidence to back up what they are doing, even if there is a strong basis around custom and practice for the effectiveness of a particular approach.

Or it may be that journal-based evidence is scant or contradictory or perhaps because the causation is to do with HR, management, work culture or environmental failings, there is nothing “medical” to base itself against.

Reality of evidence bases

The creation of a robust base of evidence and data on the effectiveness of OH interventions is at the heart of the Dame Carol Black reforms for the profession. But, of course, saying a proper evidence base is needed is much easier than creating one in reality. What the profession is therefore increasingly having to debate is: does this evidence base need to be focused around a conventional, RCT or evidence review approach, is such an approach even appropriate for occupational health. Or do practitioners need to be looking more innovatively at this question?

According to Occupational Health columnist and consultant occupational physician Dr Richard Preece, the answer is definitely the latter: “It is interesting to try to decide what is evidence and within this it is important to capture the nursing view. A lot of the ways we have historically looked at this have been based on a medical science model. But we may need to come up with an alternative.

“How, for example, do you show the value of a consultation with an OH nurse against a similar consultation with someone with no OH training? No one, to my knowledge, has shown the value of the different kinds of nurse interaction.

“So there may be an argument for putting together more studies, which might be a bit weak in the conventional sense, but which will, nevertheless, add to the evidence base,” says Preece.

Evidence reviews and guidance

In the past few years, Preece says, an increasing number of evidence reviews and guidance have been produced from different sources or have included a wider range of literature.

NHS Plus and the Health and Work Development Unit (HWDU) at the Royal College of Physicians in London (formerly the Occupational Health Clinical Effectiveness Unit), among others, have produced OH evidence-based recommendations drawing on a wider hinterland, he argues.

What’s more, if OH needs a new, different base of evidence, OH nurses have a real opportunity to be pushing the agenda and leading in its creation, suggests Preece.

“It is easy to say they are too difficult, there is no time, or they are too expensive to do. But we do need people to take up the vanguard on this. In my own studies of articles over the past decade I have found next to no articles produced by OH nurses in peer-reviewed journals.”

OH time and funding

But the reality is that most OH nurses do not have the time or, just as importantly, the funding to carry out much in the way of research, contends Anne Harriss, OH course director at London’s South Bank University. “OH nurses tend to be providing a service for employers. They might like to be able to do research but do not have the time because if they do not provide the service, their employer is going to come down on them like a ton of bricks,” she says.

“Especially in this economic climate, academics are finding it increasingly harder to fund research. Budgets are being cut across the country, as are staff numbers.

“Quantitative research may not necessarily be the most appropriate for OH. Qualitative research, too, can be useful or sometimes even better. The holy grail for many researchers – and too often the focus of medical papers – is the number crunching, but sometimes qualitative research can give a better response. Quantitative is not right for every kind of research.”

Back to basics in OH

The profession needs to get back to basics and debate what it is trying to do and achieve, what it is trying to say, and what it wants its evidence and outcomes to show, argues OH consultant and honorary fellow of the Faculty of Occupational Medicine Kim Burton. “If you are trying to show that if you have a bad back and going to a chiropractor will do x, y or z is better than taking a pill or potion, then the answer to a question like that is probably going to be best answered by an RCT,” she argues.

“But if you are trying to show whether having OH nurses working on, say, pre-employment health screening works, then that is an area where there is perhaps less need for specialised research. Or, for example, what can occupational health nurses do to help people with common health problems to start work or return to work? There is probably not that much work going on to capture or collect data regarding this.”

Workplace changes

Burton suggests an alternative route. “It might be just a question of recording data in a more systematic way and then comparing it before and after with other sites. It will be a perfectly legitimate data-set but it will not be RCTs.

“People may not think it is worth doing because it will not be taken notice of, or have less notice taken of it. But as long as it is smart and interpreted intelligently, and not over-interpreted, then it is probably still interesting and valid research; it is still a serious issue.”

Just as the workplace has changed rapidly in the past few decades, so OH nurses need to think about their work and the basis of how they work in a different ways, says Dr Siân Williams, consultant in occupational medicine at the Royal Free Hampstead NHS Trust and clinical director of the HWDU.

“The occupational health research base has traditionally been around occupational diseases, identifying cause and effect and treatment. But those skills are less needed now. We now have a much more complex situation with a cause-and-effect model that is less straightforward,” she says.

For example, what are the influences on musculoskeletal and psychological disorders and how do they influence each other? Society has a big influence too. It is not just about the organisation you are working for. But how do we research that?

“It needs to be different, it needs to be thought about more broadly. We need to have not necessarily a better, but a different approach, to look at our research needs, who needs to be involved and who is asking the questions,” says Williams.

Creating an OH dialogue

For Williams, the question of who leads this – nurses or doctors – is less important than that somebody takes leadership. “It must be led by the people who are interested and have the skills to do it. You are talking about more complex situations and interventions to try to work out what works and what does not,” she emphasises.

It might be, too, that there is a need to include psychologists or public-health professionals, “to create a dialogue with them to ask the right questions”, she adds.

Dr Paul Nicholson, chairman of the British Medical Association’s Occupational Medicine Committee, says: “For any clinician involved with prescribing drugs or carrying out surgical or other procedures, the RCT is the gold standard when it comes to measuring the relative effectiveness of an intervention.

“But it is not the gold standard method for most areas of interest for occupational physicians or even public-health physicians. For occupational health and public health, the RCT could probably be abandoned.”

There may be, too, an argument for looking for evidence beyond medical journals, so for example, HR or management publications. “The problem is that the available hierarchies all have strengths and weaknesses and it is being debated at an academic level rather than at a practical level. We need guidance that is practical and readily applicable for patients,” says Nicholson.

“There is still a need for more primary research, more practical research and a greater need for a systematic review of the evidence base. But it is also important to be engaging with all stakeholders, so OH nurses, but also occupational physicians, employers, patients and workers’ representatives plus representatives from the relevant clinical specialties. It should not be something being done solely by nurses or occupational physicians,” he adds.

The way ahead for OH

Ultimately, the debate that the profession needs to be moving away from reliance on RCTs has, by and large, already been won, argues Dr John Harrison, clinical director at Imperial College Healthcare NHS Trust and chairman of NHS Plus’s interim network board and clinical governance group. “Most people who have got involved in this area have recognised that it is only in very few cases that RCTs are applicable for workplace issues,” he says.

“You can use RCTs in some areas and where you can use them effectively then you should, because they tend to lead to a higher quality of study and design, even though they are not always perfect.

“The problem for occupational health is that the RCT approach does not usually lend itself to the workplace. It is not like a drug trial with drugs A, B and C and a placebo. You could, perhaps, have a cluster of randomisation in a large company or particular sites, but one of the problems you tend to get in a workplace is contamination: people will speak to each other.

“So, most people now accept that you can do RCTs in certain circumstances but, more often, it is recognised that there may well be value in a well-designed cohort study or a case control study, particularly one where you are looking at behaviours, action research or doing something more qualitative. These sorts of studies can be informative.

“A lot of the time the evidence that we have is so scant that it can be interesting to look at outliers as much as the mean. So it is about looking at what we can learn from people who are unusual,” he adds.

Reviewing the grading systems

The dominance of RCTs in medical research – with their status as the gold standard of the research world because of their ability to limit bias – may serve many areas of the medical profession well.

But for those areas where RCTs are less applicable, over-emphasising their importance can create an overly rigid, less-effective research environment, a recent study has suggested.

Research in the Royal College of Physicians journal Clinical Medicine (Vol.10, No.4: 358-63) has argued that overly rigid research grading systems can be “misinterpreted in that users may assume that an absence of evidence means there is evidence against a recommendation, when in reality it means that there is no evidence available for or against a clinical action”.

Grading systems can also have the unintended effect of giving priority to research methodologies that have the highest grading of evidence of effectiveness rather than research that it is perhaps clinically more important for that specialty or field, but backed by “weaker” evidence.

“Appraisal systems have to balance simplicity with clarity while providing scope for flexibility and explicit judgments. Such a balance is difficult to maintain. The decision on which grading system should be used for specialist society guidelines depends on the research area to which the guideline questions pertain,” the researchers, led by Dr Ira Madan, consultant and honorary senior lecturer in occupational medicine at Guy’s and St Thomas’ NHS Foundation Trust and King’s College, London, concluded.

“If the research field and study designs for a guideline are largely homogenous, then one system need only be used. If, as is often the case, the study designs are heterogeneous, the specialist society will need to carefully consider the options for critical appraisal systems.

“While it is possible to consider using differing appraisal systems for different study designs, this is likely to be confusing and impractical in reality. Specialist societies would be better advised to select the one that will most effectively address the predominant type of study design being appraised,” Madan added.

Comments are closed.