Discussants

Good morning, ladies and gentlemen. My name is Dave Felinski and I’m with AAMA. On behalf of our other co-sponsor, the Center for Office Technology, I’d like to welcome all of you here to day three of our Managing Ergonomics Conference, and to what should be a very interesting session on exposure assessment.

A couple of quick announcements before we get started. One is just a reminder that the proceedings from this conference should be available by the end of July via the Internet. You should have received in your registration package a form which gives you an account number and a password to log into the ErgoWeb® system. If you did not receive an account number and a password, please stop by the ErgoWeb® booth and they will set you up with that information. The other thing I’d like to announce, and unfortunately we didn’t get this information in time to get it into the final program, is that if you are either a certified safety professional or a certified industrial hygienist, this conference has been awarded certification points. In the case of the CSP’s, this conference has been awarded two COC points; and in the case of the American Board of Industrial Hygiene, for certified industrial hygienists, the conference has been awarded 3.5 CM points. The certification number for that is 1336.

It’s my pleasure to introduce this morning’s session moderator, Dr. Vern Anderson. Dr. Anderson received his Ph.D. from the University of Wisconsin, Madison, with a major in human factors engineering and psychology. He is currently the Chief of Research Group in Psychophysiology and Biomechanics at the National Institute for Occupational Safety and Health in Cincinnati. The goal of this research group is the prevention of work-related musculoskeletal disorders. He has extensive experience in conducting work site investigations in different industries and in office environments. Dr. Anderson has published numerous articles in the fields of human factors, ergonomics and behavioral toxicology. In addition, he is the editor of Cumulative Trauma Disorder, a manual for musculoskeletal diseases of the upper limbs. Dr. Anderson, the session is yours.

Good morning. When I sent that in, I didn’t think they were going to read it. I’m really happy to be here this morning to be a part of what is a sort of a pivotal and ground breaking experience bringing people together to talk about one of the most important issues in this period of time. Let’s move right into our program. We have four very well known and very interesting speakers that most of you have heard about and perhaps talked with and read their materials over the years. Each presenter will have about 30 minutes for their presentation, and then we’re going to have a period of time where we hear the discussants and then entertain questions and answers. We are trying to maintain our schedule. So with that in mind, I would like to start our first presentation.

Let me say a few words about our first presentation by Dr. William Marras who is a Professor, the Director of the Biodynamics Laboratory in the Ohio State University. Dr. Marras will be talking about a prospective validation of the lumbar motion monitor low back disorder risk model. I’d like you to welcome Dr. Bill Marras.

William S. Marras,
W.G. Allread,
M.J. Jorgensen,
and F.A. Fathallah

Thank you, Vern. And good morning to all of you. This morning I’m going to be talking about the validity of the science. And I’m going to limit my comments to low back disorders. As a good human factors person we ought to have good contrast. There’s lots of controversy in the area of low back disorders and the work relatedness of them, as you know. And my point today is that if you step back and look at more than just the epidemiology I think there’s a pretty clear picture that emerges. And I think you see a lot of convergence that shows that there is a relationship between the work people are doing and the risk of a low back disorder.

If you look at the low back disorder environment, it’s fairly complex. One thing that we’re learning over the years is that it’s really interactive among several factors. Biomechanics is what we typically talk about, which is the loads on the spine and whether or not you exceed tolerance levels. Those tolerance levels could be represented by personal factors, personal tolerance levels, that are physical, as well as psychological tolerance levels. And we also have social factors that are involved in terms of the pressure to report or not to report an injury. And for years and years all of us sort of worked in our individual areas in our individual corners and we never talked to each other. And over the past 5 or 6 years it’s been apparent in the literature that we have to start considering the other parts of the system - and it truly is a system. If you look at this from a systems approach what you see is that low back disorder reporting, as well as low back disorder research, is really a sequence of events as shown here. I know this doesn’t show up too well. That’s exposure and basically that’s the biomechanical ends of things. We were looking at the loading of the joint in trying to figure out how it matches a particular tolerance. This, if you follow it up, could lead to discomfort. If you survey the workers and ask them where they hurt, you might find that before it becomes a lost time or an injury or a reported injury. Injuries you could find out by doing some active surveillance - examine the workers, seeing whether there’s any evidence of damage. And if it gets bad enough, a person might file a claim or an incident. And if it’s bad enough, you may even have some lost time or a disability associated with that. Now our research traditionally has been up here in the exposure or the biomechanics part. If you look at what the epidemiologists have done, they’ve really looked at more the yellow words here. They run around and take a look at how many people report injuries or actively surveyed them or whatever. And if you look at the psychosocial people, they look at what might cause somebody to report a injury. And we’re really all talking about the same thing; we’re just talking about different phases of the same thing. And so our research has really tried to step back and look at the larger picture. Our primary goal is to see what kinds of workplace factors expose the person to dangerous joint loadings. And we’ve also worked at the other end of the system, which is trying to see whether we can predict how many people are going to report injuries.

The way we do this, and the way we’ve traditionally done this, is shown in this simple flow chart. The workplace is our laboratory, as well as our proving grounds for our research. In terms of finding out where the risk factors are, the first thing we do is surveillance. We simply look at the workers, watch the workers, observe them over time and see what seems to be related to incidence rates, injury rates, things like that. Given that information we can build statistical models and provide feedback to the workplace that tells us, we don’t know why but if you move this way or that way or have these types of conditions, we expect you’ll get an injury. But that doesn’t tell us anything about the end of the line mechanism, so we also allow this to drive our biomechanical research. It tells us what to look at and at what levels to look at it. The first thing I’d like to talk about is just simply the surveillance block. This has been reported in the literature in both ’93 and ’95. Some of you may be familiar with the low back disorder risk model that we’ve developed. The fundamental concept behind this is to historically observe high and low risk jobs for low back disorder in industry. And typically what we’ve done is compared workplace and trunk motion conditions in a database of over 400 high and low risk jobs. So we simply go in, find out whether the jobs are high risk or low risk, and look at what kind of characteristics are associated with those jobs. Then we sit down and we mix these together in a statistical model - actually a multiple logistic regression model - to try and figure out whether we could predict membership in this high or this low risk group.

When we performed this study we knew we had one chance to bother industry and make lots of measurements, and so we wanted to make the most out of it. We went in and we looked at 114 measures on every job. And as you can see, everything varies, from motions of the spine as people are working to the traditional workplace factors such as the weight of the object, the distances through which the object is lifted, lifting frequencies, characteristics about the workers such as their anthropometry, their work history and some rudimentary questions about worker satisfaction. In order to measure the motions, I’m sure a lot of you have seen this picture. This is the lumbar motion monitor that we developed that allows us to measure motion in three dimensional space. And as you see, there is quite a lot of motion that occurs in a lot of activities you see in industry. Now when we put these together in a model, we included five variables that are shown here: lift rate, twisting velocity, the moment which is the weight of the object the person is lifting times the distance from the spine, the maximum sagittal flexion which is how far forward you bend doing your job, and lateral velocity which is how rapidly you move from side to side. When we mix these together - and I’d like to emphasize the word mix - you get an odds ratio of 10.7. And basically an odds ratio tells you how many times more likely than chance you are to pick out somebody who is going to be in the high risk group versus the low risk group. And so you can see we’re about almost 11 times better than chance when we do that. Now the way we put this together in a usable system is through this chart right here which is really our…you could call it operating system for our assessment. This shows the five different variables on the left here, and the blue columns show the actually value associated with each one of those variables. And if you take an average of these, you have a red line that goes down here and points to this bottom row which is the probability of high risk group membership. And if that arrow is all the way to the right here, you’re almost assured of being in that high risk group. If it’s all the way to the left here, you’re almost assured to be in that low risk group. So the whole idea here is to design jobs so you get this red arrow as far to the left as possible. The other thing that this chart tells you quantitatively is how much of each one of these variables is too much. You can see they’re all scaled appropriately relative to this risk level. And as I said before it also tells you how you mix the variables. For example, in this example here, we have a moment that it’s fairly high and sagittal flexion that is fairly low, and these other ones are sort of in the middle. And so you get a medium type of a risk associated with this. You can see other situations where the moment may be very low, in other words the objects that the people are lifting may not be very large but you may have high lift rates or high twisting velocity or high lateral velocity and that’ll still cause that to be all the way to the right. It’s more than just the weight. So this gives us at least an organized way to mix and match and say how much of one is too much, given what the other ones are doing. As a matter of fact if you look at the way these jobs are distributed - that’s shown on this plot here. We see two distributions of risk. Here’s our low back disorder risk value categories that go all the way from 0% to 100%. The green are all the low risk jobs in the database; the red are all the high risk jobs in the database. And this shows how probable you are to be in the high or the low risk group given a certain percentile value of that probability we talked about earlier. So for example, if you’re in between 0 and 10% risk, that means that you’re ten times more likely to be in the low risk group than you are in the high risk group. If your risk is between 10 and 20 - and, again, this is the total mix of the various variables - you’re far more likely to be in the low risk group than the high group. And, as you can see, the ratio there is about 5 to 1. Similar, if you’re 20 to 30, you’re over 25…about 27% of the jobs in that category came from the low risk group and about 12 came from the high risk. And you can see this ratio changes. At right at about 30 or 40, it gets fairly even. It’s only separated by not even a 2 to 1 ratio. And then over 40 it switches. At 40 you’re twice as likely to be in the high risk group than you are at the low risk group. A little more likely to be in the high risk group than the low risk group at 50. And then over 50…over 60, you can see you’re almost assured to be part of that high risk group.

Now we validated this model in a couple different ways, both cross sectionally, by enriching our exposure database. In other words, collecting a lot more data and see how it behaved, as well as through a prospective study. The first one I’d like to talk about is the cross sectional validation. Since we published this data in 1993 we’ve been still collecting data. A lot of you know that because we’ve been bugging you about coming into your plants and collecting data. And this shows the types of plants that we have been into. And you can see here’s the exposure in terms of person hours of exposure. We’re at 25 million person hours of exposure back in ’93. Now we’re up to about 35 million person hours of exposure which gives us about a 39% increase in our exposure database. If we look at the individual work factors, which is one way we could see whether the workplace is responsible for the risk, we see some significance. The blue shows our updated database; the red shows our original database from ’93. And what we see is there are lots of factors that are significant. A lot of these you’ll find in a lot of control mechanisms, such as the NIOSH Lifting Guide. For example, lift rate has an odds ratio of almost two. If you look at things like the average weight of the object, it’s got an odds ratio of about five which means you’re five times more likely to be in that high risk group if you’ve got a heavier object as opposed to a lighter object. You can see the ones that really stand out here are the weight of the object and moment associated with that object, or that weight times the distance. Any one of those seem to be the single most predictive factors of whether or not you’re going to be at risk of low back disorder. These are the variables that went into our low back disorder risk model. And you can see, they’re all significant. If anything, with the updated database, it’s just reinforced the idea that they are significant. For example, moment is now up here around 8 where earlier it was down around 5 - a little over 5. If you look at the calculation of the multiple logistic regression, which is the mixing of these variables again, you can see we’re still significant; originally our odds ratio was about 10.7. Now with the enriched database we’re up to about 23.7. So with this enrichment of the database with a lot more exposure, it just gives us more confidence that we’re on the right track. And the variables themselves, or the coefficients of the variables have changed very little.

Next I’d like to talk about the prospective study. A prospective study is one where you simply follow jobs over time and see how they’re related to the risk of injury. And here’s our general approach for the prospective study. Basically what we’ve done is knock on the doors of industry and said, please let us look at your jobs that you are considering changing. We’ve observed these jobs, typically over a 3-year period. We’ve looked at the incident rate and how that incident rate has performed over a 3-year period. And then we go in there with our low back disorder risk model that we just talked about and see how well that predicts what that risk is at that point in time. Then we wait for them to change the job. Sometimes we recommend things based on this. Sometimes they don’t want to hear it, and they just go and change the jobs themselves. We’ve worked with a lot of ergonomic committees that have made these changes. But the real key is something’s got to change in the job. Then we go back and we wait another couple years, two to three years after that change is made. We do the same thing again in terms of our low back disorder risk model, we put the LMM on the workers again in the job and this is usually done with many workers in the job, up to 25 workers at a job. And we want to see whether the risk model predicts something that coincides with what the changes are with the incident rate. And then we simply see how they coincide. So that’s the general approach. In this database we’ve been following 30 jobs for the past six years. And here, the person years of exposure, total pre-change exposure is 3,017 person years of exposure - that’s the total amount of data before the people changed the job. And after they changed the job, we’re at 1,041. So you can see it’s a pretty healthy database. An example of this. This fellow is taking a tire and mounting it on this truck you see here. And the tires come down here. He’s got a little roller system. He rolls them up there. And then he mounts them on the wheel of the truck. And you can see just from observing what he has to do in this job. He’s got to bend to the side a little, he has to twist a little; he probably has a pretty high moment as he’s lifting that. And when we did the LMM analysis we saw that was indeed the case. As you can see, the factors that are high on here - maximum moment is way up there. That’s almost at 100%. Twisting velocity is high; lateral velocity is high. He’s doing a lot of moving. And sagittal moment, he doesn’t bend forward very much because of the little ramp there. And lift rate is medium. Overall the risk was about 60%. Remember when we looked at the ratio of high risk to low risk jobs, that’s an area where we’re almost assured to be in that high risk group. That job had an incident rate of about 15. So 15 injuries per 100 workers doing the job, and these are back injuries.

They changed the job, and the way they changed it was by giving a person a lift assist device. And you can see this thing automatically grabs onto the tire, grabs under the hub and it automatically allows the person to guide this thing into the wheel and run the nuts on the bolts. We looked at the risk associated with the job after that. And you can see the moment decreased tremendously. Twisting velocity decreased tremendously. Lateral velocity went down to a medium level. Lift rate increased. He was able to do more work now with this lifting device. Overall that gave us a risk of about 36 or 37%, incident rate fell to zero. So these are the types of comparisons that we were considering over the six-year period. Types of job changes we saw people make are shown here. A lot of hoists and lifts were added - 11 jobs there. We saw a lot of lift tables incorporated, total redesign of the work station or new equipment incorporated into the jobs. Several jobs there. Automation - a couple of them were totally automated. Flip tables and one included an exercise or a back safety program. And we didn’t care what they did. We just wanted to see that they changed something. And we didn’t care whether it made a big effect or not. As a matter of fact, here’s the results of 30 jobs that we’ve been looking at over the years. And what you see here is the risk category after the job was changed. Okay, so this is how it ended up. The red and the blue columns relate to incident rate. The yellow and the green columns relate to our risk model - our low back disorder risk in terms of percent. And so what we see here if we look at the high risk group, the incident rate was high before they made the change - obviously, or else they wouldn’t be considering a change. And out of these, I think there were 11 jobs in this category, 11 jobs ended up remaining in that high risk category afterwards. Okay. And if you look at what the LMM risk model predicted - it predicted that things would not change. And this is kind of amazing to me that over a third of the jobs people recognized there was a ergonomic problem with, they did the wrong thing. Okay. Ergonomics does not always work, and it’s not because ergonomics doesn’t work, it’s because people choose the wrong solution sometimes. The medium risk category - these are jobs that were originally high risk and they made a change - and you can see the incident rate showed it went down considerably, down to what we’d call a medium risk of injury. And you can see, if we look on the back here, the risk model predicted that it was high to begin with and predicted it would go to a medium level. And indeed it did. Same thing with the low risk. These were high risk jobs to begin with. They went down to low risk jobs. The LMM risk model predicted that would happen. So what’s very comforting about this is you see a lot of parallels. The green columns have the same trend as the blue column. The yellow columns have the same trend as the red column. Okay, now on average what we see is we have a tool that can predict whether or not you’re choosing the right solution, and you don't have to wait around a few years to see whether your incident rates are going to go up or down.

Now that’s the validity that we have looked at for the surveillance. As I said, my whole goal here today is to step back and look at more than just the epidemiologic data. I want to look at the literature as a body of knowledge and look at the preponderance of evidence that may point to the association with low back disorder risk. So the next thing we’re going to look at is the biomechanics. As we said before, part of the objective of the surveillance was not only to build this risk model, but it was to figure out the role of biomechanics in all this. And so we let the surveillance system guide our biomechanics. If we saw that twisting at 10 degrees per second is risky, then we wanted to look at twisting the 10 degrees per second and see what happens biomechanically. If we saw that lateral bending at 30 degrees per second was risky, we wanted to see that. And so we knew exactly what levels to set our biomechanical evaluations at. Here’s the five variables we just talked about on the risk model - lift rate, twisting velocity, moment, sagittal flexion and lateral velocity. On the right here in this column are all the journal articles that we’ve published over the years that have explored the biomechanical significance of each one of these. Okay. And these, as you see, are all in very good journals. They’ve been through scientific peer review. And I’d just like to share with you how we do this very briefly. The tough thing about biomechanics is figuring out what kinds of loads there are internal to the body. I mean, I think we all assume that the state of knowledge is far beyond what it is. But the way we’ve done it is we spent about 15 years developing a model that is shown here, at least, geometrically. We assume that in the torso there’s a plate in the trunk, and in the pelvis there’s a plate in the trunk. And the spine is the bony structure that holds those two plates apart. Connecting in the edges of these plates are all these vectors which are essentially muscles. So if we know exactly how hard each one of these muscles is pulling, we can work backwards and figure out what the compression and sheer loads are on the spine. And that’s exactly what we’ve done. Now the way we figured out the orientation of those plates is with our back monitor that we talked about earlier. And so as we move and twist and turn, the back monitor is keeping track of this top plate relative to the bottom plate. Also on here, it doesn’t show up real well, but this person has electrodes on their trunk muscles, all ten of those trunk muscles, and we modulate that electrical activity to figure out what the force is in the muscle. And I don’t want to get too technical here, but basically what we need to know about the person and about the situation to figure out the force in the muscle is, first, the capacity of the muscle to generate force, we need to know something about their electromyographic activity level. And so this is just a normalized activity level. The muscle relative to the maximum they could produce. We need to know something about the person. We need to know how big their muscles are. And we can get this either through MRI scans or by some regression equations. Then we have to know how fast the muscle is moving and how long that muscle is at the point of interest. And if we modulate the signal by all these factors we can do a pretty good job of figuring out what the force is in that muscle. And working backwards like that from the force of the muscle, we can figure out what the loads are on the spine. So we could figure out compression, which is how the loading of the spine occurs crushing the vertebrae together. Lateral sheer with is side-to-side motion of the spine, and forward and back motion of the spine. And each one of these has a different tolerance level. And these tolerance levels, for example side-to-side and forward and back, it’s only about a 1,000 Newtons. It’s really not that high. For compression, it’s much, much greater. And this is something that came from the NIOSH lifting guide in ’81. As you can see from some cadaver research you’d expect vertebral implant microfractures to occur right around 3,400…and as they begin to occur for people under 40, by the time you get up to 65 or 6,400 newtons of compression about 50% of the people will get vertebral implant microfractures. So you can think of this as like a normal distribution. And if we know what level of compression is associated with the job, we can tell you what percentage of the people we’d expect to have back problems. So that’s the logic there.

How do we know the model works? Well, as you can see there’s a lot of math in here; there’s a lot of assumptions in here. And so we need some validity check. And the way we do that is by looking at the torque or the moment that a person is producing outside the body as they’re doing this and using the EMG to predict that also and seeing how well they match up. And so here’s an example of an exertion like this. This is part of our computer model. You can see a video of the person lifting here. The person’s on a force plate. We can predict the moment. And you can see the blue and black lines here show the correlation between the predicted moment from the EMGs and the actual moment that we’re reading from the force plate. And this particular example is at 91% agreement. So that’s how we know whether this is working or whether it’s not working.

Let’s take a look at the evidence that we see for the that these various variables are actually loading the spine in the way we thought. In terms of the lift rate, we’re about to publish an article, which should occur this year in Spine. And we see a couple things happen. The blue line here shows hip angle changes over time. The red line shows trunk angle changes over time. And what we’re showing here are 12 standard tests during a five-hour period. These are workers who are lifting an average of 50 pounds for 125 repetitions per hour, and they’re doing it for five hours straight. Okay. And every once and awhile we stop them and have to do the standard test and see how their mechanics change. And so what we see is, things do change over time. Repetition is important. We can see they rely a little less on the trunk motion and a lot more on the hip motion. What does this mean in terms of loading? Well, it’s pretty significant. What we see is a slight drop in the compression, okay? And notice the compression is not anywhere near that 6,400. It’s above the 3,400. So this would be moderate levels of compression. You would expect some people to have problems at this level. But look what happens to the shears. Shears are over here. Especially the anterior/posterior shear which is right here. What you’re doing is you’re inching up towards this 1,000 newton compressive or shear load which does start to get dangerous. So one of the things that happens is you shift from compression to shear over time with repetition. What about twisting velocity? As we’ve said before, what we found if you look at the LMM risk model is even small amount of twisting velocity tends to put a person at risk. As small as 10 degrees per second. And look what happens when we had people twist at 10 degrees per second. The red is the compression; the blue and the green are the anterior and lateral shears. Here’s a static exertion, and you can see that’s the level of compression. And look what happens once you introduce any type of velocity into it - even 10 degrees per second which is very slow twisting - it doubles. Okay? And you also see slight increases in the shears.

The next factor was moments and sagittal flexion. Usually we look at these together. It’s hard to separate them out. Almost all of our experiments have people starting over in a bending position. They lift something and stand upright. And if we look at the effective moment. Here’s a study we did where we looked at the effects of case weight. These are pounds of a case. And looking at compression, over here, and shear over here, and the way they change. The green is the compression; the red and blue are the shears. And you can see a monotonic increase in compression as well as the shears as people lift more weight. This is not surprising; but, again, it coincides with what our risk models predicted. This is what happens with forward bending and lateral velocity and again what we see is that motion makes a real big difference. As soon as you introduce motion into the activity, the compression essentially doubles again even in forward lifting. And that’s because of the way you recruit your muscles.

Finally we have lateral velocity, and this is going to come out in the next two weeks in the Journal of Biomechanics. This is a study where you had people simply bend laterally and we wanted to see how the compressions and shear change. This is one of our variables in our risk model. As you can see, compression - even statically - is pretty high. It doesn’t change all that dramatically given the level. But look what happens with shears - especially at the lateral shear. It’s below this 1,000 threshold when you move statically. But once you start moving dynamically, all of sudden you pop up against this threshold and that’s where people tend to have a problem. And I’d also like to emphasize that it’s really the mix of these variables that is really causing the problem.

The last thing I’d like to talk about in terms of the evidence is the interaction between the surveillance and the biomechanics. And what we’re really looking at here in biomechanical loading - there’s several things you could look at. You could look at the static load. For example, the dotted line here shows the average load. This is the load profile of a spine during a lift. So we could look at the static load. We could look at the peak load. We could look at how that load changes in terms of load rate. Well, we did that and we tried to see how it correlated with our lumbar motion monitor risk model which is right up here. These are the probabilities of risk. And, you know, this is 20%, 40%, etc. This is the dynamic compression, the dynamic loading. Here’s the correlation between the two. And so what we see is if you go from medium, or low to medium to high risk jobs, the dynamic compression seems to change the best. And you can see where you could account for about 44% of the data when we look at that. The other thing that’s noteworthy about this is if you just look at the 3,400 newtons which is right around there, that doesn’t preclude high risk. There are plenty of high risk jobs here. And, again, that’s because of the way these things mix. And we also looked at how much of the variability we could explain by looking at the way they mix. And this is all related to the dynamic loading. Here’s compression. We already saw that’s about 44% of the loading. If we add that to anterior/posterior shear, we could explain about 47% of the variability. If we add that to load rate, then we jump up to 50. And if you look at what happens…load rate and both the anterior/posterior as well as the lateral direction, as well as the compressions, this jumps up to about 52%. So we’re not explaining 100%, but we never thought we were because remember, you’ve got all those other factors in there too. Plus this is not looking at things like repetition. But what we’re seeing is that we are finding a biomechanical basis for reporting of injuries.

And so in summary, this is what we’ve concluded from all this. Low back disorder risk is multi-dimensional, and it involves trade-offs between risk factors. As we said before, you can be high on one, low on the others, or you could trade them off. If moment is not high, if that’s low, some of these other ones could balance it off to determine risk. And now we understand how much is too much of any of these. Next, if we’re going to talk about validity, we’ve got to talk about it broadly. We just can’t talk about the epidemiology alone because it’s such a complex problem. But if you look at the epidemiology along with the biomechanics, and I think Don’s going to talk about this, we start to see a picture. And if you step back and look at this big picture, you see that the epidemiologic and biomechanical evidence converge and indicate that there is validity here, there is science here and it is related to what people are doing on the job. We could measure the effects of workplace factors upon the risk of having a low back disorder in the work place and we could use these measures as benchmarks to know how to control the job. And with that, I’d like to thank you for your attention.

Dr. Vern Anderson, NIOSH: Thank you, Bill. I’d like to move right on to our next presentation. And this is by Dr. Don Chaffin. He’s The Johnson Professor and Director, the Center for Ergonomics, at the University of Michigan.

Thank you very much, Vern. It’s my pleasure to always follow Bill. He does such a nice job of setting up this topic. And the topic is to emphasize the biomechanical aspects when we look at a job in terms of what Bill has been referring to as the maximum loading of the spine, maximum moment type of loading. I want to emphasize what he has been saying. I believe that that factor, that is how we consider posture and load in a combination that creates these maximum bending moments on a column, how that creates so much havoc for us. He’s shown us several things. Each of us go away with a different spin on these talks, I’m sure. So I’m going to try out mine, Bill, on what I carried away from your message because it acts as such a nice lead in to what I’m going to do.

First of all, one of the comments he made was that if we’re not careful, we can do it wrong. What we’re practicing in ergonomics is not common sense. A lot of people would like us to believe that. It’s not common sense. There’s a lot of knowledge here that needs to be brought to bear on the topic. I’d like to think that we can go away feeling that it’s common sense, but it’s only common sense after we’ve developed that knowledge and used it and tried it out. So we do make mistakes. We make changes, and sometimes they’re wrong. We need to be as quantitative as possible about what we’re doing. And that quantitation, in fact, helps us to learn. Again, the maximum moment he pointed out was very important, and I’m going to emphasize that, and tell you more about how that can be assessed. Remember we are interested in the exposure assessment end of the business today. And I want to emphasize that we can assess the maximum moments in several ways, and we need to. And then lastly Bill left us with the idea that when we talk about validity - and I’ll definitely emphasize this - when we talk about validity it must be in the broadest context. That to just talk about validity in terms of epidemiology is not going to get us anywhere because epidemiology in fact, allows for a lot of debate because it’s basically a very soft set of data that you’re working with. And when you look at the epidemiology, you need to also apply other knowledge and that comes from the biomechanics that’s been around for over 300 years. All right. With all that as an introduction then let me go ahead.

Human simulation. I’m going to try to emphasize that we should be, in fact, capable of looking at what the human body does and, in fact, we should be able to simulate and predict the job stresses. I’m going to try to emphasize that that’s a goal that we should have designers of jobs and engineers that are so much involved in that design process, we should give them tools that they can actually simulate the stresses on the job, at least the peak stresses, and avoid those that we know are particularly hazardous for people. So, it’s this number three point that I want to stress a little bit in my talk today, and with that the idea that job designers then can anticipate the problems. When we look at these kinds of tasks we all, I’m sure at least in this room, agree that these are appalling sort of situations that we’ve created for people. This particular company, in fact, finally went out of business last year in Michigan. And I like to think that part of the reason was that they didn’t identify early on the kinds of conditions that they were asking people to work in - lifting very, very heavy objects in very awkward positions at times. But we can go outside the manufacturing sector and look at the transportation industry that continues to be problematic in this regard. We have all kinds of heavy objects that are being shipped and we do not have at this point in time a good way of handling those objects. We have to be much more creative. It’s the posture and the load, folks. It’s the posture and the load. Please carry that away. We need to look at both. Or we can go within the warehousing or distribution centers, and I’ll come back to manufacturing in a little while. Or we can go into hospitals and look at posture and load in handling patients, etc. We can go on through many different industries. And when we look at…this is an old analysis of strains and sprains by occupations, whether it be in the general labor group or whether it be in specific groups like our trash distribution system, if you will, warehousing, nursing activities or whatever. We have many, many different occupational groups that are suffering because we have not done a good job of identifying early on in the design of their jobs what would be hazardous. And so where can we go from this? We know that overexertion related to manual materials handling is associated with excessive musculoskeletal injuries. We know that. There’s no debate on it, or rather, there should be no debate on that, particularly for the low back. But furthermore, I just point to this group that when we think in terms of a societal good, whenever we have these high exertion tasks we are, in fact, discriminating. We’re discriminating against women, we’re discriminating against older workers who often times have been injured earlier so they would fall into this group of people with an impairment. But often the older worker, just because of degenerative changes, and I’m falling into that category now. The older worker simply cannot do the same sort of things that they did before, and we must think about that folks.

With all that, where does this go? Well, we’ve been working towards the idea that we should be able to provide to you, the people who evaluate work situations, tools, software tools that work on personal computers that allow you to be very specific about saying things about risk in the workplace. And like Bill, we’re not perfect in our software tools, but we’re certainly getting closer to understanding what it is in a work environment that harms people. And with that in mind, there’s a lot of different software that we’ve been working on. I’m going to emphasize software that stresses the single exertion, the high exertion kind of task. This is where the person is again handling the load, he’s in a particular posture, and that man or woman is going to be over-stressed because of the combination of posture and load. So with that in mind, what’s the basic concept then? And it comes down to looking at strength. What happens to the body when the person has to handle that load is a matter of individual strength. And how do we model that strength? I’m going to talk about the static exertions because we know a lot more about static exertions. Bill has introduced you to the idea of dynamics, which is very important. To move from static modeling to dynamics. My appeal to you today is to least do a static exertion. And if you’re not satisfied that you have enough knowledge about the job situation from that, then you move to dynamic. But, please at least do these static exertions. We can gain a great deal of insight about what is right or wrong by doing static exertions. So I’ll talk a lot about these high effort static exertions and from that then move to where the research ought to be going.

Static exertion. Picking up a battery, placing it into a car. In this particular posture the individual isn’t particularly stressed. But, indeed, those batteries come in on pallets that are close to the floor. And so I could show another photograph that would have the person in quite a different posture. Posture and load. The load in this case may vary from a fairly light battery in a small car - 25 pounds or so - up to a very heavy battery - approaching 60 pounds - in a very large car or diesel vehicle. Okay? Load and posture. The size of the person also comes into play. If we want to look at a job in the abstract, we must in fact consider different anthropometric characteristics of individuals, men and women, that might be performing that job. But that’s not nearly as important as the first and the third, posture and hand force. How do we evaluate the posture? Well, Bill showed you a way that you can get not just postural information but movement information. And if you’re doing dynamic analysis and the job exists, that’s a great way to go. We can also do it from a video analysis. If we’re…and I’ll use a simple example here and then move to a more difficult three dimensional example. But the simple way to evaluate the individual would be to take a video, stop frame the video and pick off the angles of the body, the major joint angles, if you will. That’s not difficult to do. In fact, there are frame grabbing technologies on PC’s that will actually pick up the image, trace it and give you those body angles. So that even exists. Or you can just simply draw on a piece of acetate over the top of your monitor the general configuration of the body and pick off the angles from that simple stick figure that you might create. Or you can use the power of the PC. If you’re in a design mode, for instance, the engineer wouldn’t know what the posture would be. But, hopefully, the designer of the job might have a pretty good idea about where the hands would be in space. And at that point what we’ve been able to do is by studying a large number of people picking things up and pushing and pulling, we put some intelligence into the front end of the computer program and it actually chooses postures that we think most people would use. And we’re continually trying to refine that posture prediction technique. So by just putting in where are the hands relative to the feet, it picks postures. And if you don’t like that posture then you can go in and click and drag and move the person around on the computer screen. So there are many different ways that you can get postural information into the analysis procedure that I’m going to outline. Basically it comes down to this when you’re analyzing a task. Sir Isaac Newton, several hundred years ago, told us all about these sorts of things. If there’s a load at the end of set of levers, at each one of the articulations of the lever, each one of the joints, it will create not just a force but a torque, a tendency to bend at that joint. And it’s, in fact, that tendency to bend that drives all of the skeletal muscles to contract. Every skeletal muscle reacts to the tendency to bend at a joint. So we must know when we look at this exertion the configuration of the levers and the amount of force that’s involved. Not just at the end, by the way, because we have masses distributed we also must know the weight of the levers if you will, the segments of the body. Given that information we can calculate by simple Sir Isaac Newton equations. Not high calculus, but just simple algebra. We can multiply two numbers together and come up with an indication of how much bending there would be at the elbow, at the shoulder and so forth. So what we’re doing is computing then what we call the moment, or torque at each one of the joints. And, by the way, at this point I’d like to emphasize we’re not just looking at the back, we’re looking at the whole articular structure. But more importantly, that’s not enough. Some people would say it is enough, but I would claim it’s not enough. We must compare those torques at each one of the joints with the strength of the population because in that comparison we then learn about how bad that particular torque is. And then list from that comparison what percent of the population, men and women, we think could perform the task. Just to dramatize that calculation, it means that you have to take load, you take the posture into account, you have to draw that stick figure, if you will, and figure out some angles, and you put that into the computer program and it calculates this bending moment, or torque, at each one of the joints of the body, and it carries that all the way down to the ankles. More importantly though again, for each one of those moments it then compares it to what it thinks the population can do. The program must have a way to do that comparison. And so what we’ve done, back in the ‘70’s we went out and we measured almost 2,000 people in different postures who worked in industry. They came into the medical department, we set up strength testers and we did a lot of strength testing back in the ‘70’s and the early ‘80’s. And all of those strength norms have been tabulated, they’ve been published, and then put into this computer program. And so we have these distributions. So what it means is that if we had the person picking up the battery and the battery was close to the body, the torque at the elbow, the torque at the shoulder and so forth would be relatively low. That’s task one. It would compare that to the strength of the population, men and women, in this case. And it would tell you, in fact, that when that 50 pound battery is close to the body, maybe 90% of women could lift it, and close to 100% of men could do that. But if we put the battery down on the floor or we move it away from the body - that’s task two - then it says you’re going to have much more difficulty with women performing that task. That’s where that discrimination factor starts to creep into our argument. And, secondly, even for men this is going to be a high demand for many men and, in fact, many men could not lift the battery. Well, that’s been computerized and it’s been around since 1984. It’s not a new thing on the market. It’s called the two dimensional static strength prediction program. And it allows you then to manipulate the input. The input variables are, again, posture, the angles of the body, the load on the hands, and the size of the individual. And you can go in and just play with those variables if you want. Move the hands around, etc., etc. It shows a little stick figure of the posture that you’re simulating. And it comes down then and it looks at the elbow, the shoulders, the low back, the hips, the knees and ankles, and it says what percent of the population would have the capability to do that. But in addition to that, over on the right it does a very specific analysis of the back. And I’m going to get to that in a minute. How good is it? Let’s go back to the validity argument. Well, the only way you can prove validity here is to take a large number of people, have them perform various isometric tasks and compare with the prediction for the percentage of the population that you think could perform those tasks using the model. And that’s been done. And basically what you come out with is that in fact, it predicts mean strengths and it also predicts the extremes of the population relatively well, accounting for a large percentage of the variation in the population as a whole. So the logic makes sense. It’s using good mechanics. It’s got good data underlying it. And from two dimensions we’ve gone to three dimensions. Now it’s much more difficult to attempt to do the hand calculations you have to do it on a computer, but there’s a three dimensional model that allows us particularly to look at shoulders. We’re going to talk about shoulders later on. But it looks at the shoulder exertions. And, in fact, I was very happy that we even had some consensual validity with another group from the Vrije University in Amsterdam, looked at a model from the Delft University that is for shoulder loading and compared it with ours. And they just issued a paper on this and said that indeed this analysis that we came up with of the three dimensional static strength program in fact correlates very well with the Delft model of the shoulder. So we have consensual validity. But is that enough? Well back in the ‘70’s and ‘80’s we were not just strength testing, we were looking at what people were doing on their jobs. And we were predicting how many people could perform certain types of tasks that would be well within their strengths versus how many people would not be able to perform the task very well. What we were doing was using the model to predict the maximum strength required. We went out, photographed the jobs, broke the jobs down into posture and load combinations, figured out what percentage of the population could perform those various exertion tasks. And then we had those strength data and we looked at what the strength was of the people that were on those jobs and we compared and we followed those people. We compared them retrospectively in terms of their injury histories, and we followed them for a year to 2 years to see what their injury history would become in the future. And out of that 551 people working on all kinds of jobs, in fact when you found people that were mismatched, in other words, the strength requirements predicted by the model exceeded the kinds of average strengths that you had in the workers, some of those workers on those jobs were having problems and in fact there was this difference - 3 to 1 in terms of complaints - of injury and illness. And on an incident rate basis, and we’re talking about a lot of incidents, that’s over 1 out of 4 workers every year that were complaining about significant, in this particular case, back problems. We went in another series, another group of people - 500 people ended up in this study in the middle ‘70’s. And here again, instead of looking at complaints of back problems we actually looked at the medical visits and the follow-up diagnosis for back problems that these people had. Not as many people had serious back problems, but the serious back problems that we recorded were in fact much more associated with high exertion tasks. Those tasks where posture and load combinations in fact ended up exceeding their strength capabilities and limitations. So there’s your early field validation.

Now let me turn to the low back issue a little bit more. Bill has introduced us to these risk factors. There’s two things on this slide that I think are important. You talked about some of the severity issues. We’ll talk more about those. When it comes to back problems, most back problems are in fact self limiting. Most people today feel, in fact, that they’re acute, uncomplicated and, in fact, after a period of five to six weeks of reduced physical activity with light loading on the structure, the structure will heal, the symptoms will go away and the person will be fine. Now there’s some serious suggestion, by the way, that just came out this past week in Spine from a large study in Great Britain that those ratios may be wrong. In fact, more serious back problems may be much more prevalent in the working population than indicated here earlier. So we may find, in fact, down here at the bottom that more than 5 to 10% of people are, in fact, converting over particularly if they’re working on jobs that have maximum exertions. My feeling on that topic is if you have people where after they have had an incident of low back problems, you put them back into a job that requires a great deal of high physical loading on the column, you will, in fact, have one out of three of them convert to a chronic, serious, permanently disabling back problem. So I’m in agreement with what the people in Great Britain are starting to report about these problems.

Okay, real quickly, the biomechanics of the back. You got it from Bill. I’m just going to resummarize it. It starts again with realizing that posture and load end up creating moments at the back, just like all other joints. I’m in particular, still concerned about compression forces. Bill introduced the idea that there are other forces on the column that are of concern, but the high compression forces are created by the fact that the muscles of the back, when stabilizing the column against bending moments, the moment has no intrinsic capability not to bend. The only thing that stops it from bending and breaking are the muscles. And when those muscles contract they work very close to the spinal rotation centers, which are the disks. And there are four of those muscles that have to contract with very, very high forces, or the ligaments, either one. Those high forces manifest themselves in a number of ways. We have now very good models, and they’re built into these personal computer simulation models which allow us to look at and predict what the muscle force pattern would be. It’s not easy to do. We’re continually refining these models. But I really believe that we’re at a point where 70% of the variation in the muscle patterns for different exertions can be well predicted - 70%. The models are fairly elegant in the sense that they model the whole column and allow the person who’s simulating the task to look at very strange postures and predict the loading configuration on the column. And from that, come up with a good indication, as indicated by the EMGs, of which muscles are turning on and turning off, as Bill indicated. And by carefully instrumenting people in the laboratory just as Bill was showing, we’ve been able to compare the model outputs with the EMG levels and again, come up with laboratory validation that we’re on the right track. We can predict how those muscles are responding. And once we know how the muscles are responding, we know what kind of compression forces are going to be created on the spinal disk. What happens, as Bill indicated, when you have high compression force is that it breaks the column down. The compression force is directly related to your overall posture. If you keep the load close, the moments are low, the muscle forces are low, the compression force is low, as shown here. If I handle 75 pounds and I hold it right against the front of the body, the compression forces are well below that 3,400 Newton, 770 pound kind of limit that NIOSH agreed to back in 1981 and reaffirmed in their most recent revision. However, if I move that same 75 pounds away from the body, the compression force on the column will go up dramatically as indicated here and approach a limit that I don’t think any of us should be debating as highly hazardous. Any time you have loads that are up in the 1,400 pound range on that column, you’re going to have a large percentage of people that will be at very high risk of serious back problems. Now why do I say that? Because look at the data. When you do compression force failure testing of columns - and it’s not just a study that I was participating in back in the early ‘70s’ - this has been done not by two or three investigators but by 25 different investigators around the world that produced these data. And they all show that the compression forces will break down the column over time. The disk and the structure around the disk is very vulnerable. The limits vary a little bit between men and women. A woman has a little bit smaller column than the man in general. So the compression forces will, in fact, create breakdown of the tissue at lower levels for men. But the point being that if we were to talk about limits, as we have been talking about limits of compression force, the 770 pound limit down there at the bottom of that curve, would still end up from these data, and these are younger spines, by the way. These data look much worse if you look at the data from older spines that have been tested - 770 pounds will still end up with some of those spinal columns showing significant failures when they’re tested. So the limit that we have - 770 pounds - is roughly equivalent in the knee of those curves to saying that some people, as Bill indicated - not a lot would be at risk - but enough. But by the time we get up to 1,430 pounds of compression force on the column picking up the 50 pound battery 20 inches away from the person on the floor with an average sized individual, 1,430 pounds, now we’ve got some problem, folks. And it’s not just the cadaver data, it’s the epidemiology. When we looked at our data in terms of maximum compression forces, those people that were subjected to peak compression forces of over 1,400 pounds, eight times as many back problems in that group as those people that never had that kind of compression force loading on their columns at work.

So what am I leaving you with? We can look at all kinds of tasks today. We can photograph those tasks, we can get postural data, we can measure loads. And if we do that correctly, we can feed it into computer models, personal computer models, that will allow you to simulate that kind of task and get a very good, indication of how bad that task is. We can predict the percentage of the population, men and women, that could perform the task. We can look at different parts of the body and how much they’re stressed. And particularly we can look at the back and make a very good prediction of how badly the back is going to be stressed in a large number of work situations. And we should be doing that regularly. The data come out very easily from the computer program. There is absolutely no reason in my mind, with the personal computers we all enjoy so much, that we shouldn’t be doing this regularly. So the last message is simply, when we talk about validity, there’s a lot of different kinds of validity. Construct validity; the biomechanics are real. We’re talking about mechanics of the tissue and we know it fails mechanically. We can predict that failure. We know to a large extent how it fails. Consensual validity; different kinds of models exist. We’re comparing all the time between Bill’s lab, my lab, Stu McGill’s lab at Waterloo. We work together. And as you can start to see from Bill’s presentation, and from this presentation, we’ve been out there in the workplace, we’ve been doing the field epidemiology. It tells us what’s right and what’s wrong. Thank you.

Dr. Vern Anderson, NIOSH: Thank you, Don. In this past hour we’ve been treated to two rather comprehensive views and complimentary approaches to issues in biomechanics that represent about 20 years plus of experience. It’s really quite a treat to be able to listen even though it’s early in the morning. I’m just glad that we don’t have a test on this later, because it’s a lot of information. Now, we’re going to move on. Next we have sort of a transitional type of a presentation. Tom Bernard is going to talk about evaluation tools for short cycle tasks. Tom is from the College of Public Health at the University of South Florida. Welcome Tom Bernard.

Dr. Vern Anderson, NIOSH: Thank you, Tom. That is a nice and illuminating presentation looking at some different methods. Again, particularly with respect to some of the upper extremities. We’re going to move on and talk about another type of upper extremity assessment approach. I’d like to introduce Dr. Kurt Hegmann. Dr. Hegmann is the Assistant Professor in the Department of Preventive Medicine at the Medical College of Wisconsin. He’s going to talk to us about the application of the strain index, an advance in exposure assessment and analysis. Please welcome Dr. Hegmann.

Dr. Vern Anderson, NIOSH: Thank you, Dr. Hegmann. We’re right on time here, and I thank the speakers for keeping to our schedule. We’ve had a lot of really interesting and thought provoking information about exposure assessment and validity limitations. We’re going to take about a short 30-minute break and come back so we can all have a little more interaction. The discussants will be reviewing the presentations and then we will have a period of questions and answers from the audience.

We have three discussants that are going to talk about some of the issues that we heard about this morning. Each discussant will have about ten minutes to speak, which will leave about 30 minutes for questions and answers. Certainly you know many of these people - Dr. Barbara Silverstein, Research Director, Washington State Department of Labor and Industries; Dr. Suzanne Rogers, a consultant in ergonomics, and Joseph D’Avanzo, Partner of D’Avanzo/Morreale law firm. So let us begin this segment. We’ll begin with Dr. Silverstein.

Thank you very much. I was really very encouraged by listening to the speakers that we had this morning. I think that we have a number of tools for exposure assessment to be used both in the design of new jobs, as well as evaluating existing jobs, and after those jobs have been changed to see whether or not we’ve really reduced the exposure. I have just a few slides I really feel compelled to show. I was very happy to hear that Doug is going to be looking much more at the shoulder and after Don’s comments about epidemiology I really had to tie this back in that the costs for rotator cuff disorders are extremely high and the direct costs in terms of workers’ compensation is around $20,000 per case. I’m really very happy that, in fact, we’re looking much more at the shoulder. In addition to that, Frank Mirer made a comment a couple of days ago to another speaker in terms of needing to have a lot of variety in the exposures in any study that you’re doing in order to be able to see the effect of those exposures on disease. I would just like to suggest that some of the high risk areas for exposure that Doug certainly should take into account have to do with wallboard installation. These are relative risks, based on Washington State workers’ compensation data. We have relative risks of 11 for wallboard, garbage collection, roofing, beer distributors, nursing homes and so on. I wish you well in that study and look forward to seeing some results.

We actually heard about some other exposure assessment methods earlier in the week. We heard about Wendy Latko and Tom Armstrong’s observational method. We heard a little bit about Rob Radwin’s method as well. All of these methods are available to us for use in the workplace depending on what it is we’re trying to do. But they all basically boil down to asking, looking and measuring in one way, shape or form. There are always tradeoffs between these different kinds of techniques and we’ve heard about that today. Most of the panelists, I believe, would use all three. I think it’s pretty important to include not only posture and load, which are the critical initial factors, but also the duration, the intensity and the frequency of the exposure at the same time in terms of a potential poor health outcome. Again, looking at what it is you’re trying to do in the workplace in terms of your exposure assessment you’re going to have different options available to you. As you increase in precision and get closer to direct measurement, you also potentially increase the cost and the training requirements of those people who are going to be doing the analysis. On the other hand, you usually reduce precision as you go down to the lower end of this pyramid. But there’s very few ways that you can get huge numbers of people involved or deal with past exposures very well unless you go down to the lower end of pyramid. And, again, depending on what you are looking for and why you’re doing an exposure assessment whether it’s for research or whether it’s for identifying potential problem jobs or jobs of concern and figuring out how to fix them, you will use a variety of these approaches. I’m really thrilled that we have those available and that we have been able over the last 10-15 years to take the experience in the laboratory, move it into the field, test it, refine it, go back to the laboratory and refine it some more. This is a continuous improvement that you’re all familiar with.

The context that I think we need to look at this in is not only on research but also on public policy. The context for looking at job analysis or exposure assessment strategies has to be looked at in terms of the society as a whole. Basically if what we want to do is to protect all workers so that when they go home at the end of the day they’re in the same shape as at least when they came in, there are 6.2 million workplaces in the United States that at least are covered by OSHA jurisdiction and there are not enough ergonomists. So this poses certain kinds of questions and tradeoffs about what you do. This is a political policy decision. This is not necessarily a scientific decision. The goal of the OSHA draft checklist of 1995 has been mentioned a number of times by the speakers this morning. Basically it was never intended in its developmental stages, to be a job analysis tool, which came to an abrupt halt anyway. It was intended to assist folks in the workplace in triaging jobs, and it was intended to be sensitive, rapid, easy to use and easily generalized within different industries. And, of course, it had to be repeatable, reliable and valid in its final form which has yet to emerge. I’m very happy to hear that people everywhere are using this in testing other methods against it because all that can really do is improve whether it’s that particular tool or any other tool that’s going to help us meet these goals, it can only help us improve that process. So I thank you for your participation in that. I just wanted to point out that parts of the draft checklist are in Japanese. That’s to let you know that it’s not only being tested in the United States, but it’s been published in many parts of the world. And the draft standard and all of the appendices are actually published in Japanese and selling quite well. So maybe this will be another Deming experience.

The other thing that I wanted to say about this "ask" - meaning talking to people who know the jobs the best . . . . the "ask," "observe" or "look and measure" is something that I found last night in the restaurant in Cincinnati and I would just like you to bear this in mind because I think it applies to exposure assessment, and these wise words are from a wise sage - "you can observe a lot just by watching." This is from Yogi Berra. I’d like you to look at one of the many studies that are going on around the world that look at the draft OSHA checklist. One such study is by Laura Punnett. In this study she looks at the upper extremity disorders based on physical exam in two automobile plants with 1,400 workers. She looked at a variety of potential risk factors and personal factors when she was doing this study and she translated them into visual analog scales and then recoded them into a zero, one, two, three which is very similar to those things that are on the draft OSHA checklist. What I really want to point out is that in some of the regression models that she used when controlling gender, whether there was a previous upper extremity injury, and whether there were any relevant diseases, and then controlling for all of those. When she looked at the OSHA checklist in terms of a score of five or less being a zero score. And looking at the other quartiles, this is…within the exposures that she looked at, what she saw was a very nice exposure response relationship looking at upper extremity disorders on physical exam and this triage instrument. That also is very encouraging. We need more studies like this. We need to refine it. More research is needed in all of these methods. I think Don, Bill, Tom and Doug all pointed out some of the limitations in the methods that they’re using and also the methods that need further validation in the workplace. A lot of the issues are those of precision. We know what most of the risk factors are. And actually, now there are some very nice models for the psychosocial factors and work organization factors, in particular both observational methods and asking methods that are being used in prospective studies in Denmark.

I really want to point out that we do have models, we know that they can work. We need to increase their precision because that will really help us to better hone in on our interventions. I also want to say that the draft OSHA checklist does not, nor was it intended to, identify the cause for these risk factors to be present. That’s critical. I think a lot of what both Don and Bill are talking about identify what the causes are and what things we really do need to look at and change. It does not lead you to a solution directly. It does not prioritize the risk factors themselves for reduction. I think the job analysis or exposure assessment methods we’ve heard about help you to do that. So I’m thrilled that they’re out there and being tested. It does not separate risk by specific body part; it’s only upper extremity or by low back and lower limb. And at this point there is no provision of industry-specific examples within it. So I don’t want to go into the future of the checklist, but I do want to say congratulations to the four speakers because I think that you have shown us a way forward and we should seize it. Thank you very much.

Dr. Vern Anderson, NIOSH: Thank you. I believe our next speaker will be Dr. Suzanne Rogers.

One of the fun parts of this is that the person in front of me just gave most of what I was going to say. So now I can talk about some other things. Thanks, Barbara. What I thought maybe I would do is use three examples to show what those of us who are trying to look at jobs in industry are looking at, how some of the data that have been discussed today can help us in analyzing that, and where there may be some gaps.

The first one I’d like to take as an example would be a nurse in a hospital trying to handle patients. Which tools would we use to study that task? Don’s model and Bill’s model are two of the tools you could use to evaluate the stress but I think all of us agree that if we simply use the NIOSH guidelines as a start, we know that we are probably going to have a patient weighing more than 51 pounds. I’ve suggested to people sometimes that for a safe lift you ought to leave them in bed until they weigh 51 pounds but I don't think that’s a practical solution. So the question is, we know we’ve got a problem and when we’re trying to help the nurses in the hospital or the handlers in the hospital, we have to be able to evaluate what other things we can do. I’m trying to cross this over with some of what Steve Sauter talked about yesterday as well. What can we do? Well, we have had some excellent studies by Arun Garg and Beatrice - what was her name, Sweeney? But on handling aides we’ve had a lot through NIOSH and OSHA supporting this kind of work, trying to find better ways of gripping the patient. In reality, one of the problems is that when you go into a typical hospital - not necessarily the ones we’ve done research in - what we’ll find is they’re not using those or they can’t use them. They say they can’t use them. And the reason they can’t is they don’t have time. And why don’t they have time? Because they don’t know when to predict when somebody has to be handled. There’s not extra help around anymore the way there used to be and the problem is that they are unable to find the equipment when they want to find it. There are all those factors in the global sense that say, "we know we shouldn’t do this, but we’re going to have to do it anyway because we have to get the job done."

In one hospital I looked at, it turned out that when we kept asking why they had to rush that they were rushing because X-ray didn’t want to schedule the chronic care patients down in X-ray. They wanted to use them to fill in when people didn’t appear for appointments so that they could get maximum utilization of their equipment. So our way of solving the back problem was to tell them they had to schedule the patients, put a volunteer on to be sure the lifts and assist devices were in the right places, and schedule the lifts wherever they could. We didn’t get rid of the lift; the lift is still there but at least it allowed people to do things. And when you see the incidence rates a lot of times you say, "well, you know, you can’t do anything about that, that’s the way you have to do it." But in reality using these psychosocial factors, you can very often reduce the stress and make it better.

Now the second example I want to mention is warehousing. Here we are having the usual numbers of problems with people doing lifting tasks all day. As you know it’s a very select population that ends up in those jobs. They tend to have natural selection. And what you see in this case was a pick rate of 3.6 per minute and the bosses couldn’t see why people were in trouble. What’s wrong with that thinking? Well number one, the pick rate was 3.6 per minute by taking the total number of items and dividing them by the total minutes that they worked, not by watching what was happening out in the field. What was happening in the field is if you took the time from when they got to a stop to the time they got back to that stop to go to the next stop, the effective rate, not the real rate, of the pick rate would be more like 11 per minute. And then if you looked at from the time they got to the pallet to the time they actually let go of the load, in other words, the actual handling time, it was 22 per minute. And this is where I use Bill’s data to say it’s the acceleration of velocity that is probably most of the problem. So again if we can take the other factors that make them rush and reduce those, then we can reduce the velocity of the acceleration. That will hopefully get us using your kind of information and we’d like to have a number that we should shoot for on that, as you’re doing. That will help us know how we can reduce the stress enough to reduce the risk for handlers out in the plant…or in the warehouse.

The final example I want to show you relates to data we don't have at this point. This was a washing machine assembly manufacturing operation. It’s one where the agitator is already in the washing machine. The washing machine’s coming down a conveyor and your job is to take a rubber boot, pick it up, push it down over the top of the agitator. The next step is to take a hose clamp which is smaller than the boot, push that over the boot and then shoot the screw and tighten up the whole thing. If you don’t do this right you get water on the floor in your laundry when you try to do your wash. The job entails leaning over the height of the washing machine when you’re doing this which is about 40 inches. Which means it’s above waist height for a lot of people so they’re up on their tiptoes. The force is straight down with the boot but it’s a high resistance because it has to be tight. Remember that they are following the conveyor while they are doing this which means they are also walking. Every 14 seconds a new one comes in and they’re back there starting again. When you get to the hose clamp, because it’s smaller than the boot, you’re still in this same posture you started in. So as you’re now trying to get the hose clamp over the boot in place, you’re reaching down about 30 inches into this. Which means your whole upper body is there but you can’t lean on it because you’ve got to keep walking. And what you’ll see is what they had to do to get the clamp over was to put one side down, then grab the agitator to get the other part down underneath.

Now how would you measure that stress? That’s going to be biomechanical, shear forces, etc. And it would be done after you have been in that position for 10 seconds. Okay, so they have got a fatigued back. They’ve got, on top of the force to get it down, they now have to jerk. This is what was actually breaking the seal on the bottom of a lot of these machines. Now what industry is worried about is how much is it going to cost to fix this? How else can we do it? And before I got to this particular plant, they’d all looked at how they could change the machine on the conveyor in order to make it easier to do that task. I looked at it and said, you know, there’s no way you’re going to do that. There’s just no way you can do it. They’ve got other things on the conveyor; you can’t change the conveyor. What I got them to think about was time; how long they had to be there. And the goal, again, as I was saying in my question to Tom Armstrong, the goal was how can I shorten up the time I have to be there. Because if I can take the fatigue out of the postural problem and can take the jerk out of the band placement, I’ve reduced two serious parts of the stress of this job, right? And what we found, with a great mechanic, was that there was a one inch narrow screw that held the two parts of the hose clip together. That was as wide as it could get. The mechanic went back to the shop and made a 2-inch screw and tried it out, and it just fell right over the top of the boot, shot the screw, he got three seconds back and he had no jerk. This was a mechanic who came up with this. I didn’t come up with this. And this is why I want to say that by looking at the whole problem, identifying the problem in terms of time, frequency and intensity and looking at all the ways you can fix it, this was a real fix. They had had three people a year going out with back problems on this job. There was a question of it being a problem job. Their total cost was over $250,000 a year because they had to retrain, rebid the job, etc. It became very expensive. And nobody wanted it. Everybody was smart enough to know that they didn’t want it. So they had to keep getting people from other departments because nobody else would do it. So the point I want to make is that we can analyze that with biomechanical models, with motion models, and all these good things, and using Bill’s back monitor. But the real solution came more from the time than just the actual position and the jerk.

What I’ve tried to do with the kind of data which is critical that the scientists are developing is to look at forces. I’ve gathered this data mostly from the literature, largely from Kaiserling’s work and then from some of the Air Force literature and some of our own studies at Kodak. I think if I can make a plea to the biomechanists, the next phase I’d love to see us get into is force exertion, independent of lifting, because I feel quite comfortable with what we have on the lifting, but I don’t feel very comfortable at this point with force exertion, and what kind of upper limits we ought to be looking for in the job.

And just one final comment. I want to thank Ford Motor Company for summarizing my work as a simple tool. I’m happy to say that as I get older I get simpler. And perhaps that’s the best description of the fatigue model. But I’m pleased to see that some of the overall job analysis systems are trying to get into some of these issues of time, as well as issues of force and frequency. Thank you.

Dr. Vern Anderson, NIOSH: Thank you. Joe D’Avanzo will be our last discussant.

You’re probably all wondering what a lawyer is doing here in what was a purely scientific presentation this morning. In the defense of my clients who have been sued by various individuals making claims for repetitive stress injuries, I’ve been reading the literature from back since 1988. I read Dr. Silverstein’s thesis and followed the work that she and her colleagues at the University of Michigan have done over the years and the others around the world. And I have to say that the science, and it is a science, has come a long way from simply trying to look at performance and comfort and getting beyond that and integrating biomechanical models which seem very intricate from a lay person’s point of view, and trying to come up with some way of quantifying what takes place in the human body at work. I was hoping to give some controversial kinds of remarks and comments, one of which I had thought of as I was reading the abstracts, was "how come nobody integrates into their models what people do outside of work?" We don’t check our arms and legs at the door when we punch out on our way home. So what are we doing to try and measure the stresses on the back by the couch potato who doesn’t sit up straight on his couch as he’s watching the evening news? What do we do to quantify the stresses that people exert on their bodies when they change their oil, or rotate their tires or garden or take care of a child at home, or cook dinner, or shop? Anybody who’s bent down into a grocery cart knows that there’s stress being placed on your back as you’re doing that. I think the next step, and one of the things I was asked to do is to give some seeds, or plant some seeds in you so that you can ask questions of these people who presented this morning. And I hope that my remarks will.

I would like to see some integration into these models of what we do in the other 16 hours of our lives outside of work. I think we’ve come a long way in coming up with measures and how to observe and analyze jobs. I think we are at the point where we can actually talk intelligently about accommodation for the injured worker, talk about redesigning a job or maybe even designing a job from its inception based upon the work that has been going on in this area for the past couple of decades. But I think we’re a long way off, and I think Tom Bernard agrees with me on this, we are a long way off from predicting who will be injured and who will not. I went to Australia and Japan to try and unearth from their rich history in this area why the epidemic that took place over on the other side of the world occurred, how did it subside and what measures were taken to control it. One of the things that I was told from the Australian investigators was that no matter how much they threw in new equipment and new desks and the like to improve the ergonomics of the office setting, the number of claims did not abate. So in my mind that militates towards our integrating some of the psychosocial models that I’m sure you’ve heard over this past week with these biomechanical models because two people working side by side, doing the same job with the same work environment doesn't necessarily mean that both of them are going to complain of pain nor come away with a diagnosable injury or disease. And so the question is, what are the differences between the two, and what are we doing to get measures, with respect to those differences, into our models for the design of jobs. That’s really all I had to say. I think that this is your opportunity to ask questions of these researchers who have spent a lot of worthwhile time, in developing these measures. And I would hope that I’ve given you some seeds for thought. Thank you.

Dr. VERN ANDERSON, NIOSH: This is the time for the questions and answers portion. I was wondering if the presenters want to come up here. There’s plenty of chairs, and I could move a little bit off to the side here and that way if they want to be involved in this question and answer they’d be a little more accessible to a microphone. I guess we could probably get started. We have someone already queued up there by the microphone. Please introduce yourself and your affiliation and state your question or comment. We’re trying to keep this to about a minute in terms of questions, so we can really talk to a lot of people.

Mr. Jim Mair: My name is Jim Mair, Rohm and Haas. I’m following up on your comment, "why aren’t we looking at the non-occupational or the home activities?" One of the things is I have tried to do - especially with those people who work on a computer, people in sales, or people who are on the road a lot - is to conduct workstation evaluations on those folks whose have their office at home. It’s remarkable how they don’t call us or they don't want us to go to their home and look in on their operation. I would like to ask, what is it you and maybe others of the panel could suggest that we do, what we should look at? How do we, and do you suggest that we, go into or do evaluations of home offices? I personally am not very comfortable with that and my experience is that folks do not appreciate it. I think that one of the things we can do is certainly make clear the messages about occupational safety and try to have employees take home what they have learned at the office about the importance of being safe in the workplace. My questions are what would you suggest or what would you have us do in regard to home activities?

Mr. JOSEPH D’AVANZO: I agree that when you start talking about trying to figure out what people do at home you’re getting into an accessibility issue that is difficult now even in the workplace, but even more difficult when you have privacy interests involved. But it would seem to me that there are other ways to skin a cat other than setting up a laboratory in somebody’s living room. You can probably get some questionnaires out to these people which…I know that when I take a deposition of a plaintiff, I have a laundry list that goes on for many pages of their personal activities outside of the workplace. Maybe get some information from them by questionnaire and then as Rob Radwin did with some of his work in developing his model, you get people into the laboratory to simulate some of the activities they do and get some measurements as to those activities. My problem is that we don’t know, and we have not developed a comparison between one hour of knitting versus, you know, one hour of lifting the 50 pound box. And we don’t know how to compare those two things. So that would be my suggestion. And of course, there are people much more qualified than myself that could develop that methodology, I’m sure.

Dr. SUE ROGERS: I’ll just make a quick comment. I think we ought to design our jobs so that people can live a normal life outside of work. I don’t think there’s any way that we can know how much is occupational, and how much is non-occupational particularly with repetitive strain injuries from my own experience anyway. I desperately wanted to use my worker’s comp when I had carpal tunnel and I couldn’t think of any good reason to do it. So all I’m saying is I think we’re never going to be able to put a number on that. I know you’re being asked to do it from the medical side. But the main thing, I think, is if we make our jobs good then it’s easy to get back to work when you have these problems. It’s easy to continue working when you have these problems. And that’s our ultimate goal.

Mr. D’AVANZO: I just wanted to add one comment beyond that. Dr. Silverstein mentioned that this is more than just science, this is also policy. And one of the issues, I think, that has to be decided by employers and employees alike is, even though it’s a worthwhile goal for the ergonomist to study the mechanics of work with the proposition that you go home as refreshed as you did when you started your job in the beginning of day, I wonder whether or not that is something that should be mandated, and whether that is something that should be a policy adopted in this country.

Mr. Terry Mandzene: My name is Terry Mandzene, I work for a division of General Motors, i.e., slash ergonomist. Like Dr. Silverstein, I feel compelled to come up here because of a 26-year association with workers on the floor. Having been one and having been associated with them for many years. I have a comment or two, and I have a question. My question will seem to contradict my comment, but I guess that’s the nature of the beast. My comment is that this seems to be a debate about whether repetitive motion causes injury to workers. I say if you pound your head against a wall 400 times in 8 hours you are bound to have contusions that are probably not genetically induced. That’s my comment. Now, along with Mr. D’Avanzo’s approach, I too have noticed that workers standing side by side do not have the same occurrence rate of repetitive motion injury. His approach I believe is to find out what, besides repetitive motion injuries or repetitive motion, might be causing it. I think it is more important to find out what’s not happening, and why is it not happening. I would issue this as a challenge, and I’m sure work has been in this vein. Fifteen to 30% of the people in the workforce have CTD’s. That leaves 70 or 80 or however you define it, percent that nothing happens to. Why isn’t it happening to them? Why do the same people standing side by side, same stature, same fitness not develop the same injury? There’s genetics, there’s nutrition, there’s work method. I think we need to develop and come up with some common denominators for people who this is not happening to. There are studies going on to investigate sharks as to why they don’t get cancer. So I think we need to find out why some people do not get RMIs. I think it is extremely important and there may be some common denominators that can help.

Dr. DON CHAFFIN: I’ll rise to that one. First of all, I think you should be congratulated on your insight. There is absolutely no doubt individual differences are very large. The genetics involved in all of us is yet to be discovered in this field. We’re just starting to see now through the genetic research that’s going on, what kind of musculoskeletal risk factors we’re carrying from our mothers and fathers and grandmothers and grandfathers, etc. Being that still aside, I think there’s a lot we’re going to learn in the next ten years about that aspect of us. So I do think that there’s a lot that we will become concerned about, and this will also be policy. In my own work on the back, I’ve always advocated that a strength testing procedure that is related to the strength requirements of the job can assist in controlling back problems as part of an overall comprehensive program. That’s all ADA and EEOC would allow anyway. But the point is that we are different. We need to study those differences a lot more, but at the same time we have to still acknowledge that there are a large number of job conditions that will injure a large percentage of individuals regardless of their genetics, and regardless of their strength and endurance characteristics. So, I’ll just have to leave it that we have a lot to know about individual differences in this game. We also have to concentrate on these extremely high exertion kinds of situations we put people in, where there’s just no debate - or should not be debate - about the fact that these are highly hazardous tasks to almost everybody.

Dr. BARBARA SILVERSTEIN: I think that this is no different than what we see with tobacco smoke and lung cancer. You can see folks who have been smoking a pack a day for 20 years who do not develop lung cancer, it doesn’t mean that tobacco smoke has no impact on the development of lung cancer. I think genetics needs to be looked at. I think what really constitutes a healthy workplace needs to be looked at, and build on those models. We do know enough now to work on improving what we can in the jobs that we have, and the ones that we’re developing for the next century.

Ms. Peg Seminario: Peg Seminario from the AFL-CIO. I just have a couple of comments. Mr. D’Avanzo raised the question, at least in his mind, as to whether as a matter of public policy we should indeed mandate that people be able to go to work with the expectation that they can return home at the end of day the same shape that they went. That already is mandated. It was in 1970, and for 27 years we’ve taken action to try to identify those hazards at work that put people at risk and take steps to intervene and indeed have mandated that by law. Having been involved in that process for some time, I keep scratching my head as to why we are putting a much higher standard of proof on this particular set of injuries than we have for the last 27 years as far as taking action to control these particular problems. The issues that have been raised with respect to non-occupational risk, with respect to genetic factors - that’s no different than anything else that we’ve faced with respect to other hazards in the workplace. Barbara raised it with respect to lung cancer from tobacco smoke. None of this is new. I keep grappling with why is it that we have to answer every single issue, every single question that anybody can raise with respect to this set of injuries before we take action. And I would assume that we shouldn’t and we don’t have to do that. The question I would ask of the panel on somewhat of a different line has to do with the kind of evaluation tools which you have been developing using the laboratory, taking it out to the workplace, to what is your view right now with the broad applicability of those kind of tools given what Barbara said that we’ve got lots of different workplaces out there. What kind of tools are really available to take into the workplace to identify the kinds of jobs that are causing people to be at risk, and then to take action with respect to controls?

Mr. JOE D’AVANZO: I’ll respond to the first part of the comment. I think one of the reasons why this area has been so controversial as opposed to other areas, like reducing exposure to hazardous substances and regulation of factories, is that with respect to hazardous substances, the intervention is fairly clear - it’s reduce the inhalation or exposure to that by either breathing apparatus or just make sure that it’s just not in the air that you breathe. With respect to factory floors, things like pinch points and other hazardous conditions are well recognized and the outcomes and interventions are well recognized. I think what we’re struggling with here is trying to define precisely what movements and exertions, many of which seem to be almost normal and every day occurrences throughout our entire life, not just our working life, but our every day living, are causing problems. Where do we draw these thresholds and where do we draw the line? It does have important implications about how we accomplish work. I think somebody mentioned Newton and physics, and Newton also talked about work being force times mass as I understand it. But there’s an exertion there and I don’t think that you want a policy which prevents workers from doing work. I think one of the controversies in this is where do we draw the line so that work still gets accomplished yet nobody is getting injured.

Dr. DON CHAFFIN: There’s no doubt we want to make sure we have a work force that can be productive. I don’t think anybody in this room would refute that. And so the issue is really how do we determine what is injurious about that work. Peg, I guess all I can say from my experience, and I’ll turn to the measurement part of Barbara’s tripartite approach, look, see and measure. The measurement side today with the advent of all of us having some type of personal computer, it seems to me is now something that everybody can do. And the camcorder technology is there. You can hook your camcorder up to your personal computer. Everybody is doing this. We’re going to have digital photography that everybody will have at very moderate prices. The measurement is what we need to move to. This is not common sense. We have to go from the screening tools which are more observational, to the measurement tools. Because it’s through measurement that you get the precision. This was exactly, I thought, Barbara’s argument and it’s so wonderful the way she put that argument. Precision is what we need to make change. Precision is what the legal people are going to want. We need to be precise. And we can be precise. And to get to your question, I think the proliferation of these measurement tools is just beginning in terms of our software, for instance, there’s over I think 2,500 licenses that have been released on that. But I know that it’s probably double or triple that in other ways. So we’re just starting. We’re seeing the beginning of this across all work sites. And I think it’s wonderful. We have the technology. There’s no reason why we shouldn’t be much more precise than we have been.

Dr. SUE ROGERS: Just a quickie. I think when we get into talking about regulation, we don’t want to forget that we can still do ergonomics without having to be regulated at certain levels. And the way I come down on this now after many years of looking at it is, let’s define those points where we know we have a clear hazard, and I think that’s what the science is doing for us, and regulate those. And then let the rest of it be work practices. And the other thing I’d love to see you do is like ADA did - is to try to say that all future design is going to stay within ergonomic guidelines. Our problem is we’re reworking the horse all the time. We’re having to keep coming back and doing it again because we’re not getting the design characteristics in enough places. So I think if we separate them that way and look towards the science to tell us where that line is where we’re quite sure that the risk is high, and start there, then if we need to go further we can go further later on.

Dr. Ron Schopper: Ron Schopper with CSERIAC. My question pertains to a couple of questions about your model. One pertaining to the information you did provide, one pertaining to the information that you are researching. Rather academic questions, the first one pertaining to the velocity data beyond your design limits. In contrast to the continuing linear relationship between force and injury potential or load (in other words, hand force), the velocity data seemed to exhibit a ceiling effect and the relationship broke down after the design limit.

The second was pre-empted in the talk by Sue Rogers, and that has to do with the absence of the incorporation of higher dynamic effects in your model, from your model, acceleration and jerk. Are they absent from your model because you examined them and found them not to be significant, or an insignificant finding, or does your database still permit you to go back and address these?

Dr. BILL MARRAS: Well, first it depends on what model you’re talking about. Acceleration is certainly taken into the EMG assisted model that we talked about. In terms of the LMM risk model, after examining the variables and seeing which ones were associated with risk, acceleration and jerk did not come in there as a strong variable. It seemed like a lot of that variability was explained by the velocity variables. And you have to understand the interactive nature of a multiple logistic regression model like this. Those factors that are in there often incorporate a lot of other factors that may not be apparent. In other words, in the underlying structure of the model it incorporates those types of factors. Now I’m not sure I understood your first question.

Dr. Schopper: Your velocity effects in a couple of your graphs seemed to indicate a decrease. Can you explain this?

Dr. MARRAS: Okay. You’re talking about the biomechanical model where we went to 10 and 20. Well it seems that it deals all with very technical parts of the biomechanical model which is the co-contraction of the muscles. And what we think is happening, especially with twisting, the reason that becomes dangerous is because with any twisting at all you’re trying to stabilize the trunk, you co-contract the muscles which increases the compressive force on the spine. Now if you go beyond that from 10 to 20 degrees per second, you’re getting a little more compression but, the fact that all these kicking at the same time is already taken care of by that initial motion.

Mr. Mike Morris: My name is Mike Morris and I’m the Industrial Engineering Manager for Morton Salt Division of Morton International, so I’m the management person. I’m also entrusted with making job layout changes and general ergonomics policy at Morton. I have a question for Dr. Marras. I would have perhaps held this comment until tomorrow but it coincides with Dr. Marras’ remarks. Doctor, I’m familiar with the case picking study that BioDynamic Labs did a few years ago in the grocery industry. Your results parallel our own experiences at Morton. I believe you said your case lifting study showed that position of the case being fit ideally at waist level or chest level was far more important than the case weight itself. In fact a 40 pound case gets easier to handle than a 60 pound case at waist level or chest level and put it on a conveyor or a pallet, than a 40 pound case that you’re picking up off the floor. And, again, that parallels everything I’ve ever lifted. Is that still correct?

Dr. BILL MARRAS: That’s correct.

Mr. Morris: Okay. Then my question is regarding regulations, or the eventual regulations. A NIOSH lifting equation would have outlawed the 60 pound movement but it would have accepted the 40 pound case lift. I mention this just to caution everybody that setting eventual regulations is going to be difficult.

Dr. MARRAS: I really doubt if the NIOSH lifting equation would call a lift from a low level acceptable, because remember there is a 51 pound constant, but you mediate that according to things like the vertical height and the distance, etc.

Mr. Morris: That’s the revised NIOSH lifting guide.

Dr. MARRAS: And the original.

Mr. Morris: That’s fine. But automatically, a 61 pound case would be void altogether or a 60 pound movement would be void altogether. To me it doesn’t make much sense.

Dr. MARRAS: Well, hopefully that’s answered your question. What we found is that these jobs are very complex and there’s a lot of different risk factors. For example, pulling versus lifting. Up here you’re pulling, at a low level you’re lifting and so there is a very different effect.

Dr. SUE ROGERS: I think you want to be sure you look at the horizontal distance as well when you’re lifting that 40 pounder off the floor because I think it’ll throw it out for that reason. So probably neither of them are going to be acceptable.

Dr. DON CHAFFIN: There’s no doubt that if you do any kind of three dimensional analysis the movement of that kind of load at waist height is going to be more of a pull in and up which makes all the difference in the world. By the way, Sue, the three dimensional analysis doesn’t assume the load is down. It’s a push/pull analysis too so you can put any kind of hand factors on - left or right - or you could simulate that kind of thing you were doing. That’s not a problem. So you really need to look at the hand force vector in that kind of analysis that you’re talking about and not just assume that it’s a dead weight.

Dr. Frank Mirer: I’m Frank Mirer, UAW. I have two scientific questions. The first question is directed to the upper extremity group. Do you believe that all these risk assessment tools for upper extremity are basically converging on the same fact with different levels of sensitivity as opposed to measuring different things. In other words, are they all basically ways of ranking jobs with whatever threshold for what’s a permissible exposure? And then secondly, for the two back guys, could we accomplish most of what’s been developed in the new science by simply throwing a multiplier into the NIOSH lifting guide for a lift and twist situation? Would a simple multiplier, as opposed to this complex analysis, enable us to get most of the additional stress that comes with a lift and twist?

Dr. KURT HEGMANN: I’ll take the question on upper extremity first. I don’t think that there is necessarily convergence. I think that what we have are a couple of methods that look to be highly sensitive but not specific. In the context of taking care of patients, we have two step testing on a lot of different disorders. You have a highly sensitive one initially, followed up by a specific one. So I think that that is one possible outcome. A second thing is that I don't think that we are going to have a generic model for the whole upper extremity because I think we will ultimately find when the science is in that there are different levels of risk with different specific disorders. And the third aspect is that the shoulder has not really been developed to any degree. And so we’ve been talking about distal upper extremity.

Dr. DON CHAFFIN: Frank, your suggestion that we need a multiplier for asymmetric twisting kind of lifts is one that a lot of us have been very concerned about as you heard. The NIOSH lifting equation in its revised form tried to deal with that. It does have an asymmetric factor in it. And it was the fifth factor that was added. It is a linear simple multiplier in that sense. So there’s already a simple multiplier for asymmetric lifting in the new lifting guide. Let me just tell you that my sense, and hopefully Bill will join me here, my sense is that that particular factor is not, biomechanically, well justified. It’s perhaps justified well on the strength and psychophysical side. But as Vern knows, we had some long discussions about this early on - a little bit of twist isn’t bad, biomechanically. We could argue about whether it’s 15 degrees, 20 degrees, or 25 degrees. A lot of twist is really bad. In my mind, biomechanically it’s not going to be a simple multiplier. It is a simple multiplier now, but I would wish that it would become a multiplier that would acknowledge that the risk of a little bit of a twist isn’t bad. But when you get all the way around to try to do something, that’s really hazardous.

Dr. Mirer: Is 90 degrees a lot? How much is a lot?

Dr. CHAFFIN: Yes, 90 degrees starts to become very, very bad.

Dr. BILL MARRAS: If I could amplify that. Interestingly enough, our laboratory just submitted a paper this past week to the Ergonomics Journal on this very issue. We looked at the effects of asymmetry and how our biomechanical evaluations with different levels of asymmetry correspond to things such as in the NIOSH lifting guide. One of the things we found, as Don just mentioned, is not linear. In some areas the modulation - according to the lifting guide - is too much. In others, it’s too little. And also one thing we’re finding is it also depends on whether you’re twisting to the right or to the left. We have very different modulating factors for each direction. It also depends on whether you’re lifting with one hand or two hands. So as I was saying before, it’s a complex issue, it’s not that simple.

Dr. TOM BERNARD: The thing to also track in all of this, I think, is what’s the depth of knowledge that you want to get out of this. In other words, there’s some things that are more a recognition tool to get to triage. And as Don has mentioned at least a half a dozen times, the computers are getting cheap and powerful, and these tools will become readily accessible and usable with a great deal of knowledge in them.

Ms. Tashlyn Chase: My name is Tashlyn Chase and I’m the CAW International Ergonomic Coordinator at Ford. I just have a couple of comments. My first comment is that I think we need to make a distinction between what is required and what is leisure. When I go to work, I’m required to go to work to take home a paycheck. When I go home, what I do at home is leisure and I have a right to say no if I don’t want to do it, or if it becomes too much. It’s not that I feel that our leisure activities are not significant, I’m sure they play a role. But at home if I feel that I don’t want to do something, or that it’s too much, I have the right to say no. At work we don't have that option. I would like to thank Sue Rogers and the others for their support in stating that jobs should be ergonomically designed so that we can go home and participate in life’s daily activities because often workers don’t have that option. When they get home, they’re too tired. They can’t pick their kids up and can’t play with their children. They can’t get the gardening done. And I just want to remind you that we as workers, we go to work, we go to work hurt, depending on our financial situation and regardless of how hurt you are, you continue to go to work. No one wants to go on workers’ compensation. For many it’s just not economically feasible for us to be off work because we’re injured. I just need to make the comment that, consider the workers. The onus has been placed on us for far too long and we need to start looking at other directions, and ergonomics is supposed to be an intervention that takes some stress off the worker and the employer. I think we should continue to think about that.

Mr. JOE D’AVANZO: My only reaction is that I’ve never taken the position, either personally or on behalf of my clients, that ergonomics and discussing how to improve the office environment from a comfort point of view is verboten. In fact, I look at it as similar to what went on in the factories many decades ago to try to make them safer. I think it’s very healthy that we have as many people interested in this subject and participating and doing research on it to try to make the office environment comfortable, and to improve people’s performance without necessarily making them unable to live their normal lives. My concern comes when people point the finger at my client or my client’s products and say that they are legally responsible for causing their carpal tunnel syndrome or the like, and that’s where I think some people take what knowledge has been gained in ergonomics and stretch it until it’s transparent. And that’s where I take issue. But I think it’s been very healthy to recognize that work in the office is indeed hard work. And that workers should be accommodated so that they’re comfortable. I know I appreciated that there’s now ergonomic furniture and chairs available for me when I spend those many, many hours late into the evening in front of my computer. But I don’t necessarily believe we’ve come to the point where we can point to one specific tool in the workplace and say, this is the culprit and this thing has caused me harm.

Dr. KURT HEGMANN: I would like to make one other comment. We’ve come back again to the person risk factor issue. And I think we need a ranking of what are the job related risk factors from top to bottom with odds ratios associated with them. And that’s the employer’s responsibility. We then need the non-occupational risk factors ranked top to bottom with an odds ratio associated with them, and that’s the worker’s responsibility.

Ms. Pat Bertsche: Pat Bertsche with the Ohio State University Institute for Ergonomics. This question is directed to Dr. Chaffin and Dr. Marras. On the first day of the conference, I understood Dr. Lauerman, an orthopaedic surgeon, to make a statement that the occupational risk factors for back problems are ill-defined. I was wondering what you two gentlemen would have to say about that comment.

Dr. DON CHAFFIN: I hope that if I did anything this morning, at least I defined a couple risk factors and convinced the majority of you that they certainly aren’t ill-defined. They’re measurable. They are real risk factors when you talk about some of these postures and load combinations. Those are real serious risk factors. The biomechanics are there to say they are. The epidemiology is there. The physiology’s there. What more do you want? As Peg pointed out, what more do we want? Now when we get into some of the dynamics - and Bill can speak to those - then clearly that’s much more research oriented, but I would certainly agree 100% that the faster you move, the more that risk factor’s going to be there. I agree with that not only from his epidemiology which is very convincing to me, but also from the biomechanics and the Newtonian mechanics, if you will, involved in that kind of movement and now going beyond that to the time domain. For the back we have much more to talk about because the physiology of the muscles involved, how fast - as Tom was pointing out - how fast fatigue develops, what role does fatigue play in all of this. Muscles get sore. Well, so what? That’s a physiological, that’s not a pathologic issue, that’s normally a physiological issue. It’s just the body saying, "hey, slow down, give me some rest." And if we didn’t have that kind of feedback, we could really injure the tissue. So in that regard we have to now start thinking much more broadly, and much more research, in my mind, still has to go on to define the time domain, the repetition, the endurance aspect of the soft tissues in particular - the ligaments and the muscles involved - as well as the hard tissue. The repetitions compressing on the tissue of the disk definitely, as we have seen in the German studies that show that those disks, if you just keep on squishing them, repetitively become much more susceptible to significant damage over time. We still need to do a lot more on that.

Dr. BILL MARRAS: I concur with almost everything that Don said. One item I do take issue with is that I think we do have the knowledge of dynamics that we can apply now to the workplace and there are ways to do that. I guess I would follow up the response that anybody who would make a comment like you suggested Pat, hasn’t been reading the literature and doesn’t know what strong evidence there really is these days.

Mr. John Amell: I’m John Amell, and I’m an ergonomist with Boeing Company in Seattle. And to preface this I could use examples from major league baseball. Some of the guys at Boeing don’t understand a lot of the issues that we’ll be dealing with in the next ten or fifteen years, but they do understand major league baseball and they understand pitching. I guess my challenge for the biomechanistic modelers is if you could come up with a predictive model, probably not major league pitching but in particular, with our sales managers.

Dr. DON CHAFFIN: Before we lose everybody just on that note, a lot of people draw from the sports world into the work world all kinds of inferences, whether it be from the back belt world of weight lifting to situations that you’re describing. All I can say is we have to be very careful in those kinds of extrapolations. The sport world is full of survivors. You don't get to be a major league pitcher easily. And so your genetics are good, you’ve been trained all your life to do that one thing, maybe, whether it be heavy power lifting or throwing balls, or whatever. And so there are different types of people, with different limits to what they can do. And the last thing I would just point out is, I hope your young workers don't have to learn the hard way how difficult it is to be old.

Dr. VERN ANDERSON, NIOSH: On that note, thank you very much for attending the session. Thank you for your participation. I thank this panel and the presenters. And I guess you’ll be back here at 1:00.