Spanish MSD Study Highlights Challenges in Ergonomics Research

The Ergonomics Report contributor Tim Villnave set out to summarize this research article several weeks ago, but ceased the effort after some  frustration. I decided to take a second look, and quickly understood why Tim became frustrated. I’m not sure whether it is the quality of the study itself, or the way it has been reported in this article, or whether it’s all the unanswered the questions that come to mind as we review it.

Article Title: Work demands and musculoskeletal disorders from the Spanish National Survey
Authors: Martin Caicoya and George L. Delclos
Publication: Occupational Medicine, Volume 0, Number 2010
Access the full article: This study can be acquired at: http://occmed.oxfordjournals.org/cgi/reprint/kqp191v1
Author Affiliations
Caicoya: Occupational Risk Prevention Service, Hospital Monte Naranco, Oviedo, Asturias 33012, Spain
Delclos: Division of Environmental and Occupational Health Sciences, The University of Texas School of Public Health, Houston, TX 77030, USA;  Centro de Investigacion en Salud Laboral, Universidad Pompeu Fabra, Barcelona, Catalonia 08003, Spain; and CIBER Epidemiologa y Salud Publica, Barcelona, Catalonia 08003, Spain.

This study is a post-hoc analysis of observational data collected in the 5th Spanish National Institute for Safety and Hygiene at Work (INSHT) survey (V Encuesta Nacional de Condiciones de Trabajo or VENCT). The INSHT survey was conducted using two separate interviewer-administered questionnaires: one for employers and one for workers. However, for the purposes of this study, Caicoya and Delclos used only the workers’ questionnaire data, which consisted of 107 items.

Primary Findings

• Nearly 89% of workers reported musculoskeletal symptoms that they attributed to postures or efforts at work.
• However, only 19% of the workers perceived musculoskeletal risk factors at work.
• There were 513 work-related injuries, affecting 11% of the workers.
• Of these, 146 (28%) were attributed, by the workers, to efforts or postures at work, which was the second most frequently perceived cause of work-related MSDs.
• The proportion of workers reporting the use of significant force or manipulating heavy weights during part of their shifts was more than 33%.
• The proportion of workers having to use significant force for more than half of their shift was nearly 4%.

The authors then analyzed associations between self-reported duration of exposure to selected tasks and self-reported symptoms, as well as associations between self-reported duration of exposure to selected tasks and self-reported injuries.

Interpretation of Findings

The tasks the authors looked at included "using significant force," "lifting or carrying heavy weights," and "repetitive movement." This is where the confusion really starts to set in, because it’s not at all clear what their findings were, other than self-reported injuries and symptoms show a statistically significant increase with self-reported increases in duration of exposure to these tasks.

Notice that I’ve italicized quite a few key words in the above summary.

For example, self-reported. The complete data set contained both worker reported and employer reported data, yet the authors only used the worker self reports. Why? Furthermore, how reliable can we expect self-reported, subjective data to be?

For example, Symptoms vs. injuries. This is not the fault of the authors, but is a sticky question for anyone trying to understand MSD related data. When do symptoms become an injury? What is the difference between an injury and a disorder? What is the difference between an injury and an illness? The stickiest question of all might be, "how do we know if an MSD, whether classified as symptoms, disorder, injury or illness, is work-related?" Welcome to the political underbelly of the debate surrounding OSHA’s proposal to publish a definition of an MSD and add a tracking column for such on the OSHA 300 Log.

Risk factors vs. tasks. Risk factors, such as force, posture, repetition and duration, may or may not be present in a task. In a lifting task, it is the combination of these risk factors that ultimately defines risk, not just the existence of one or more of these factors (e.g., lifting posture, object weight, anthropometry, repetition, duration and other factors interact to determine risk). So, does "lifting or carrying heavy weights" automatically imply risk? What’s the difference between "lifting or carrying heavy weights" and "using significant force" and "repetitive movement?" What defines these variables?

Post-hoc analysis. This is not the fault of the authors, since it is the nature of the type of study they performed. In the context of design and analysis of experiments, post-hoc refers to looking at the data, after the experiment has concluded, for patterns that were not specified beforehand. It’s precarious to interpret cause and effect relationships using such an approach. Associations alone do not necessarily demonstrate cause-and-effect relationships.

Observational. There is an important distinction between randomized and observational studies. Observational studies, while very common, are not always a good way to investigate cause-and-effect relationships. I touch on this in more depth in this Ergonomics Roundup article.

The Bottom Line — How this Applies to Ergonomists

Professional ergonomists should rely on scientifically sound knowledge when seeking to prevent or remediate MSDs in the workplace. Our interventions should be evidence based. The body of research knowledge we rely on contains some very good evidence demonstrating cause-and-effect relationships between certain risk factors and the onset or aggravation of certain MSDs. That knowledge base also contains supporting evidence that, on its own, may not demonstrate cause-and-effect relationships, yet does further the scientific foundation that will assist future researchers as they dig for deeper understanding. This study falls in the latter category.

This article originally appeared in The Ergonomics Report™ on 2010-04-07.