CASE DEFINITIONS AND DIAGNOSTIC CRITERIA

Musculoskeletal cases have been identified by a series of acronyms -- CTDs, RSIs, OUSs, RMIs, MSDs (UEMSDs or LEMSDs), etc. While there may be some agreement that certain pathologies are included in these general descriptors, case definitions for others may be lacking or less than clear. "Symptoms," unusual, persistent pain or discomfort, or other surrogate indicators are sometimes included, sometimes excluded, but often poorly characterized.

Broadly accepted diagnostic standards and more rigorous case definitions of musculoskeletal disease conditions are important to sound, consistent medical practice, clinical and epidemiological research. The presenters at this session will review the state of the science, possible implications for the medical practitioner, and recommendations for future research.

Session Arrangers

Howard M. Sandler, MD, President, Sandler Occupational Medicine Associates
Lawrence J. Fine, MD, DrPH, Dir., Div. of Surveillance, Hazard Evaluation & Field Studies, NIOSH

Presenters

J. Steven Moore, MD, MPH, CIH, Prof. of Occupational Health Services, University of Texas Health Center
Disorders of the Muscle-Tendon Units of the Distal Upper Extremity

Bradley Evanoff, MD, MPH, Head, Sect. of Occup. & Env. Med., Washington Univ. St. Louis
Consensus Criteria for the Classification of CTS in Epidemiologic Studies

William C. Lauerman, MD, Associate Professor of Orthopaedic Surgery, Chief of Spine Surgery, Georgetown University Medical Center
Case Definitions and Diagnostic Criteria for Disorders of the Low Back

Dean Louis, MD, Professor of Surgery, American Society for Surgery of the Hand
The Challenge from a Hand Surgeon's Perspective

Discussants

Gary M. Franklin, MD, MPH, Medical Dir., Washington State Dept. of Labor & Industries
Ronald H. Gray, MD, Professor & Deputy Chairman Dept. of Population Dynamics, Johns Hopkins University

Mr. DAVID FELINSKI, AAMA

Welcome back everyone. I should like to apologize for that morning session getting a little bit away from us there. We'll try to keep that from happening again. I have installed a timer up here which shoots electric jolts into the speakers when they exceed their allotted time, so hopefully that will help. It's my pleasure to introduce the session moderator for this afternoon's session on case definitions and diagnostic criteria, Dr. Howard Sandler. Howard Sandler is President of Sandler Occupational Medicine Associates, an occupational environmental health and safety consulting firm, and he received his undergraduate degree and medical education from the University of Maryland. He served as a medical officer with the National Institute for Occupational Safety and Health where he participated in program development and evaluation of special projects such as NIOSH reproductive effects research strategies. Dr. Sandler also took the lead for occupational health care delivery programs for NIOSH, and provided peer review for professional publications. Dr. Sandler has testified on ergonomics and proposed regulations by OSHA and CAL-OSHA before the U.S. Congress and California State Legislature. He has served as principal scientist for the National Coalition on Ergonomics, and has performed a formal analysis of the literature on ergonomics risk factors and adverse health effects. Dr. Sandler, the session is yours.

Dr. HOWARD SANDLER, Sandler Occupational Medicine Associates (SOMA)

Good afternoon. It's a privilege to be here today. And I also want to thank my co-moderator and session arranger, Dr. Lawrence Fine, who is the Director of the Division of Surveillance, Hazard Evaluation and Field Studies for NIOSH. He was the Director of Occupational Health Programs in the Department of Environmental Industrial Health at the University of Michigan School of Public Health from 1985 to 1988. Prior to that, he was Assistant Professor of Occupational Medicine at the Harvard School of Public Health. His credentials go on and on. But to get to the speakers, I would just like to make a few comments. I am delighted to be here today as part of this conference to advance the science, research and policy issues of musculoskeletal disorders and their relationship to work. It is critical that we all speak the same language whether it is for research, surveillance or diagnosis and treatment. We have four excellent speakers and two outstanding discussants who will address the problems and opportunities involved in assuring that our knowledge and potential workplace solutions in this area come from standardized and consistent case definition and diagnostic criteria. Our four objectives today will be to review the range of definitions and criteria which have been used in research and medicine, identify areas of conflict and inconsistency, discuss different approaches for research, surveillance and medical care, and to present efforts underway to establish consistency and standardization.

Our first speaker is Dr. Steven Moore. Dr. Moore received his undergraduate degree in physics from the University of Texas, his medical degree from Southwestern Medical School and his Masters in Public Health from the Medical College of Wisconsin. He has worked as an occupational physician in the Navy and industry. He was at the Medical College of Wisconsin from 1988 to 1995, and he's moved to where he now resides at the University of Texas Health Center at Tyler. He and Dr. Arun Garg have been instrumental in developing various tools and worked on a number of ergonomic areas. He's developed a semi-quantitative job analysis method called the strain index. Dr. Moore is certified by the American Board of Preventive Medicine and Occupational Medicine, by the American Board of Industrial Hygiene in the comprehensive practice of industrial hygiene and by the Board for Certification of Professional Ergonomists. I'd like to welcome Dr. Steven Moore.

Dr. Howard Sandler, SOMA

Dr. Moore, thank you for that very interesting topic. It's a lot to try to tackle in a short period of time. Our next speaker, Dr. Bradley Evanoff, holds the Sutter Chair of occupational, industrial, environmental medicine at Washington University where he heads the section of Occupational and Environmental Medicine and holds the rank of Assistant Professor of Medicine. His research activities involve three main areas in occupational health: the epidemiology and prevention of work-related musculoskeletal disorders, work-related health problems in health care workers and the evaluation of occupational medical education. I shortened up so that we have more time for his presentation. Let's welcome Dr. Evanoff.

Dr. Howard Sandler, SOMA

Thank you very much, Dr. Evanoff. We're going to switch now from the upper extremities to something that's not been as well concentrated on today and that is the spine and back. Our next speaker is Dr. William Lauerman. He graduated medical school from Georgetown University. It was at Georgetown University that he also did his internship and residency. He then completed a fellowship at the University of Minnesota's Twin Cities Scoliosis Center. Dr. Lauerman is currently serving as Associate Professor of Orthopedic Surgery and Chief of Spine Surgery at Georgetown University Medical Center. I'd like to welcome Dr. Lauerman.

Case Definitions and Diagnostic Criteria for Disorders of the Low Back

Dr. WILLIAM LAUERMAN

I want to thank you for the opportunity to come and discuss case definitions and diagnostic criteria for disorders of the low back. I'm going to review several things. I'll give you a general overview of the problem as I perceive it. We'll touch briefly on anatomy, look a little bit at epidemiology, then I'll discuss a little bit about how we evaluate patients in the clinical setting, talk about the pitfalls and efficacy of screening, and then get into some of the case definitions and the difficulty with diagnostic criteria. And I think research in this field is extremely important. We have a spine research laboratory at Georgetown. As you may be aware, in Washington D.C. we have taken to getting away from using rats in the laboratory, and we now use lawyers for the majority of our research. And there are three reasons for that. One is that in Washington in particular, lawyers are more plentiful than rats. The other thing that we found is that the researchers tend not to bond with lawyers the way they occasionally do with their lab rats. And the final and most important point is that you can get lawyers to do things that no lab rat would ever do. It's an old joke, but always a good one.

Just a brief overview of the problem. Basically from the point of view of the clinician taking care of patients with the back, I see patients on the symptomatic end - when they've injured their back or have back pain. And a large proportion of my practice is workers with back injuries. And many times it can be a relatively difficult problem for me to sort out what their history is and what their problem is. And in general in the office setting at least or in the operating room my concern is not, Ais this is back injury, is this a repetitive overload syndrome or what, it's just trying to make the patient better and it's only a year or two down the road where I'm giving a deposition that the causality and aggravation and causative factors come to the fore. The other way that back disorders can be looked at is from the point of view of predictive value or screening value. And I'll talk quite a bit, or talk a little bit, in a few minutes about if there is any way to predict what individual and/or what worker is at risk for a back injury down the road. Some of the questions that I'll try to answer are how common is back pain, and how common are back injuries, are there any predictive signs or symptoms or x-ray or other findings such as MRI findings that suggest that someone is at risk to develop a back injury, what is causation, what is aggravation, and what is the significance of pre-existing conditions. And my prediction, unfortunately, is that I'll probably only answer two or three of those questions to your satisfaction.

Just an overview of the anatomy. It's a given throughout the animal kingdom that form follows function. That, therefore, the anatomy of a structure is custom-made to maximize the function of that structure. And the function of the back is primarily three-fold. One is to allow human beings to stand erect. The second is to protect the neurologic tissue such as the spinal cord and the nerve roots as they exit the spinal cord. And the last is to allow and to 'supervise', if you will, normal ranges of motion -- to allow us to turn and look behind us, or to bend over and tie our shoes in a physiologic or guided fashion that is neither excessive nor overly restricted. And to maximize that form we have the functional spinal unit. And the functional spinal unit is made up of the vertebrae above, the vertebrae below and then the disk in between. And this could be in your neck or in your low back. It really doesn't matter. And the bone, the bony portion, is the vertebral body which is the large weight bearing portion of the bone, and then a bony arch in back which is like the sidewalls and the roof of the house that protect the contents in the middle. And the contents in the middle are the spinal cord and the nerves. And then there are ligaments throughout the spine some of which are, probably the thickest is the anterior longitude ligament which is perhaps one-fourth the thickness of your anterior cruciate ligament in your knee which everybody's heard about. So these are relatively small, thin ligaments, but still vitally important. And then there are joints in the back. And the joints in the disk tend to supervise or define the limits of motion.

Now some of the key structures that we'll focus on just a little bit are highlighted here. The disk is the soft tissue structure between the two vertebrae. So here's a vertebral body, here's disk, and here's the vertebral body. And the disk is made up of two components. One is the thick outer casing called the annulus fibrosis, and that's made up of a very tenacious grizzly connective tissue that tends to encase the inner aspect of the disk which is the gelatinous nucleus pulposis. Now the nucleus pulposis is made up primarily of water - it's 70% water admixed with a variety of proteins. And during the course of the day, the work day or whatever day, the water is actually squeezed out of the disk and we lose height, and then when we go to bed at night water is imbibed back into the disk, we gain height, and that's why you get in the car in morning you have to adjust your rearview mirror because you're actually about a half inch taller than when you went to bed at night. See, you thought you weren't going to learn anything today.

The nerve tissue runs between the anterior column which is made up of the vertebral body in the disk and this posterior bony arch in back. And this is actually somewhat misleading because the spinal cord actually ends before the low back, that your spinal cord ends just above your waist and gives rise to a number of nerves and the nerves are paired. One nerve comes off at each level, and those nerves go down your leg and make up your sciatic nerve, ergo 'sciatica', and your femoral nerve which can occasionally cause pain down the front of the leg. And then finally there's a muscle which is probably the most important and the most often injured structure in the back, although it's not the most dramatic - it rarely if ever, hopefully, leads you into the operating room, and so it tends to be neglected somewhat. Another reason that it's neglected is that it's very difficult to image. You can get x-rays, MRI scans, bone scans, every scan known to man and nothing really tests for injury to the muscles of the low back even though probably in at least 90 if not 99% of work injuries on the job, and 90-99% of the work injuries that I see in my office, it's primarily muscular injury. Of note also is that even though you see a lot of white here, the back muscles do not give rise to rope like tendons the way muscles in the extremities do. So we looked at what the tendons of the hand looked like and how they can run through little pulleys. The back muscles insert by tendinous structures but they tend to be very broad, flat structures. And it's certainly somewhat debatable again because we have no definitive diagnosis testing modality and we never operate on muscular injuries to the back, so it's somewhat debated when somebody has a back strain which we'll look at exactly what the site of the injury is.

Now just to look a little bit at epidemiology. And this is one of the problems that makes it so difficult to determine what's normal and what's abnormal and what's work related and what's not work related, and what are you at risk for if you're a high school football player versus not being a high school football player. It's like saying, does being a truck driver put you at risk for getting a cold, catching a cold? Well of course you catch cold three or four times a year, so how can you sort it out? And back injuries are about as common as that. Sixty to eighty percent of normal healthy individuals at some time during their life will have a significant episode of back pain. Everybody has back pain on almost a daily basis. But a significant episode of back pain, meaning they've limited their activities and/or they've sought care from a health care professional. In 10-30% of individuals they will describe in a given month that they've had some back pain. So that if you pass out a questionnaire to this audience today and say, have you had back pain in the last month, 10-30% of you will say that you have. Fourteen percent of individuals in the course of a lifetime will have an episode of sciatica which is defined as pain traveling down the leg, usually below the knee and is believed to be related to some sort of nerve irritation, possibly compression, in the low back. And there are a variety of normal abnormals because the back is x-rayed so often and now studied with MRI scans so often, we know much more about the structure of the back than we really to know in the majority of cases. And there are a lot of so called normal variants that we now feel have no relationship to back pain. But, again, in the literature if you look back, there'll be a case report from the 1960's of three patients who had surgery, and two of them got better and somebody proposing that this is a vitally important possible answer to the source of back pain. Essentially if you look at the epidemiologic studies and you look at the potential for screening studies, there are basically no x-ray abnormalities that really are predictive of low back pain. And the reason I have an asterisk there is because that's not entirely true. There are a couple of very obscure findings that may have some positive correlation with the presence of back pain. But, in general, all of things that we look at, all of the things that an emergency room doctor will look at and all the things that our residents in our clinic will look at and think, boy we operated on this last week, maybe I can operate on this patient next week. All of those have very poor predictive value for whether or not the person has back pain. And this is just an example of an MRI scan which shows a lot of disk degeneration, but is actually taken in a research control subject. By way of example, 5 to 6% of patients have spondylolisthesis which is the forward slippage of one vertebrae on another. That means 5 to 6%, if we have 400 people in the audience here, somewhere between 20 and 25 of you will have spondylolisthesis. Well spondylolisthesis is a common reason that I do surgery on people's backs, but we know that the majority of people don't have symptoms from their spondylolisthesis and it is extremely difficult certainly on an individual basis to determine whether that spondylolisthesis is a red herring. First of all, it has nothing to do with the person's back pain because they have a muscular injury. Whether it's a preexisting condition that was aggravated by a work injury, or whether it is truly the cause of the injury and may even be new. We also know that if you get MRI scans, and MRI scans have sort of become the gold standard now in evaluating patients clinically, and they are wildly over used. And a very good example at least in my little sphere of medicine as to why health care costs are so high. But if you get MRI scans in asymptomatic people, if we take this whole block here, and you all say that you've never had back or leg pain of any significant degree, 20% of those of you between the ages of 20 or 40, and I assume that's most of you out there just from looking at you, 20% of you will have a herniated nucleus pulposis, a herniated disk. And if you get MRI scans in patientsYin individuals, not patients, in individuals age 60 or older, between 70 and 80% of individuals will have this type of disk degeneration which can be the source of incapacitating back pain, can cause a person to have to retire from work and can be a reason that we operate on people, but is present in 70 to 80% of 60-year-olds who've never had a lick of back pain to speak of in their lives. So as you can imagine, epidemiologically it's really a morass to try to think that any radiographic imaging study is going to help us identify people who have back conditions or back problems.

And therefore, we get to screening. And this again is a good example of an x-ray test. And if you bring prospective workers in, let's say, and you get x-rays and you say I've lost so much money on back injuries in the last two years that I have got to weed out people who are at risk for back injuries. And so I'm going to at least x-ray their back. Well this individual is like a poster child for instability of the spine, and back pain. And this x-ray shows disk space narrowing and nitrogen in the disk and a bone spur there, and backward slippage of the vertebrae there. And this, again, was somebody who I saw about three years ago. A week later he was once again asymptomatic, but he brought this x-ray with him, and I saw him a year later in follow-up, mostly just out of curiosity and he's never had a bit of back pain since. So you can see that every x-ray finding in the book doesn't really predict, and in fact as we talked about, much better scientific studies than this one have been done that have looked specifically at abnormal x-ray findings to see if any them had any predictive value for someone who was at risk to have back pain and in fact none of them do.

Just to talk a little bit about the mechanism of injury of occupational back injuries. And it's sort of highlighted on the front of your conference program. You can see that most of individuals that we see, and the most common cause of injury in occupational work injuries is lifting. People think very often of falling or slipping or whatever, but the most common reason that someone hurts their back on the job is indeed lifting. And it's to that, that a lot of the occupational research is going on is better lifting techniques. Another word about occupational risk factors, and I realize I've just kind of highlighted some of these things. Occupational risk factors which you would think would be relatively easy to define. If someone drives a car more often than his counterpart and develops back pain, or someone lifts more, almost all the occupational risk factors that I know of are still ill defined. And the most common, and the thing that I learned even in medical school is the risk of driving and the frequent vibration associated with driving. And it was essentially a given for a number of years that people who drove for a living were at increased risk for having back pain. And in an elegant study that was done in literally in Finnish twins, where they have in Finland a large registry for all identical twins. And they've done any number of studies looking at different medical characteristic of twins. If you pair twins, one of whom has a job but doesn't do any significant driving and other does a lot of driving, and how they define that is neither here nor there, the incidence of back pain is essentially the same. In other words, the amount of back pain that would appear to be related to the driving controlling for all other factors, is the same. So it really kind of has eliminated, or at least calls into question, the old adage that driving is a significant risk factor for back pain.

Now how do we evaluate these patients? Well, basically we do many of the same things that are done in the world of hand surgery. The most important thing is the history - listening to what the patient tells us. Unfortunately the history is strictly subjective. And I can tell you that my personal bias is that there are no malingerers in world, at least in the world of the people who have back pain. In my experience and the way I approach my practice, the furthest thing from my mind is that worker or the person who was in an automobile accident, or the person who just woke up at home with back pain, is misrepresenting his or her symptoms which is to me what I define as malingering. Now don't take from that for me to mean that there are not other functional or emotional problems that may impact in someone's perception of pain and someone's report of pain. But frank malingering per se in my experience is essentially non-existent. The history though, by its subjective nature, it makes it very difficult to rely on it specifically, at least for the back, as a significant diagnostic criteria. Without symptoms, without what the patient tells you in terms of history - Amy back hurts or my leg hurts or I can't urinate properly - without that history, there is no back disorder. But just because you have the history doesn't necessarily mean you have anything other than a back strain. We do a physical examination. Again, physical examination is a key component of our evaluation. The physical examination is composed primarily of very non-specific findings. Unlike the hand, where it is very illuminating to see exactly where the individual is tender. Are they tender over the distal forearm, are they tender over the proximal forearm, the medial epicondyle or the lateral epicondyle? Tenderness about the back is completely non-specific. And almost without exception, I know of no condition where the spot where the patient is tender really helps me much coming up with a diagnosis. And that certainly befuddles patients, but it is almost universally true. And the majority of other physical findings, such as decreased range of motion, are very non-specific. Now it is lucky in the back, at least, that we occasionally see patients who have neurologic problems. And the neurologic findings, which is a vital part of the physical examination, the neurologic findings can be very specific. They're not always specific, but they can be very specific - which reflex is abnormal, which motor group is a little bit weak. Those are relatively specific findings. Neurodiagnostic tests are what our previous speaker referred to as EDS. For us it's more the EMG, which is the motor test part rather than the nerve conduction velocity. And in general they are confirmatory at times, but actually have very little role in the routine evaluation of back patients. X-ray testing is very, very common and, in fact, is way over done. As I talked about there are very few x-ray findings that have any positive predictive value, again we heard that defined in the last talk. Very few x-ray findings that have any positive predictive value for the presence of back pain. It's been estimated that as a primary care doctor, you would have to see 2,000 patients in your office and get x-rays on each of them before you would get an x-ray that would materially impact on what you decided to do that day for that patient. And that really is the bottom line in terms of the utility of x-ray testing, or any testing for that matter, is does it materially impact on my treatment decision making. And again, you can get the x-rays or the MRI's that we looked at before, and I've only shown you abnormal ones, but if they don't help you decide what you're going to do next it really isn't worth the money. And similarly for MRI scans, they can be very helpful, they can be very specific in showing you certain conditions, but there's such a high prevalence of false positives, abnormal MRI findings in asymptomatic patients, that they really have to be used with discretion. Interpreted with discretion and ordered with discretion.

Well, let me look at a couple of case definitions and then I'll sum up. But before I get into case definitions I want to talk a little bit about what diagnostic criteria are. Now there are two diagnostic groups, or there are two potential diagnostic criteria, and they include the objective, things that are black and white and not subject to interpretation, and the subjective, things that are shades of gray and are entirely subject to interpretation. And unfortunately, there aren't that many objective diagnostic criteria in the world of back disorders. There are laboratory tests, such as what your blood count is or what your white count is if we were looking for an infection, that's as objective as they come. Your blood count is either 42 or it's 39. And it's not subject to a whole lot of interpretation, although in terms of clinical decision making it's certainly subject to interpretation. X-rays should be objective - you either have a broken bone or you don't have a broken bone. And physical findings such as weakness in a muscle group or an abnormal reflex should be very objective also. But on the subjective side, symptoms are by definition purely subjective. Again, no malingerers but what somebody says is my back hurts so much I can't work or my back hurts so much I can't enjoy sex or my back hurts so much, but I ignore it anyway and do both of those things - that's strictly subjective, strictly the patient's perception. X-ray findings are also very subjective. Your thigh bone is either broken or it's not broken, but whether you have a badly herniated disk or some compression on the nerve is a very subjective finding. And more importantly the significance of x-ray abnormalities is very subjective. And finally, physical findings can be very subjective. Someone's exquisite tenderness when you press on their back, maybe to another person they don't even know that you want them to tell you that it hurts or not. And even things like muscle weakness or sensory loss. One person may decide that muscle weakness is not really that significant in the same person. And I can tell you from patients that I've taken care of that I'll read somebody else's report where they say it was a normal neurologic exam, and I'll find things that I think are very important - abnormalities that I think are very important. So let's look at a couple of case definitions.

Well the most common back injury and the most common back condition that I see even as a referral spine surgeon, is lumbar strain. And lumbar strain is defined as a self-limited injury to the musculotendinous unit of the low back which in most cases is related to some at least minor episode of trauma. Not usually thought of as a repetitive or cumulative trauma situation, although it may well be that. And if we look now at our diagnostic criteria, there are no x-ray findings that tell us anything about that. There's certainly no MRI findings that help us. Neuro diagnostic findings wouldn't help us. Physical examination are only the non-specific things - there'll be back tenderness, there'll be some spasm which is relatively objective but not very specific. Back spasm occurs with any number of conditions. And we're really left with the history. The individual coming in and saying, I was walking along and somebody asked me to pick something up and put it on their desk and I twisted my back and I've had back pain and spasm ever since. So it's really difficult to hang your hat on lumbar strain and define what it is or how prevalent it is.

Let's get into some more specific problems. Lumbar disc disease is defined as degeneration or loss of water content of the disk. And you sometimes will hear the term degenerative disk disease. Well, it's not a disease because it's a normal part of the aging process as we saw. Seventy- to eighty-percent of asymptomatic 60-year-olds will have lumbar disk degeneration. And it probably does a disservice to everyone involved to label somebody at age 42 with some abnormality in their disk on MRI scan or x-ray as having disk degeneration. Lumbar disk disease can be the source of instability of the back, abnormal motion in the back causing back pain. The disk itself can be source of pain as it degenerates, and it's sometimes a source of ongoing incapacitating back pain and the source of injury. But, if you look again at the diagnostic criteria, the history is helpful - usually back pain, pain in the buttock. Physical examination is totally non-specific. Again, loss of motion sometimes, tenderness about the back, occasionally spasm, but no specific physical abnormalities. And certainly no objective findings that are irrefutable such as reflex abnormalities or neurologic findings. Neurodiagnostic studies don't help. X-ray and MRI findings are usually there, in fact by definition they are there if the MRI is not abnormal. The MRI is a very sensitive test looking for disk degeneration, but it is so non-specific as almost not to be helpful. Herniated nucleus pulposis or herniated disk, or slipped disk is a condition that's extremely common, usually self-limited. The nucleus pulposis, the gelatinous inner aspect of the disk squirts through the annulus and presses either here, sort of in the mid-line against a bunch or nerves, or off to the side, presses against one of the spinal nerves as it exists the spine to become the sciatic nerve. And it's manifested by back pain, pain down the leg, numbness and tingling down the leg and occasionally on physical examination, weakness or reflex abnormalities. And it is probably the easiest diagnosis to firmly be able to tell a patient who comes in with clearly associated symptoms, not just back pain but with pain down the leg in the appropriate distribution into let's say the outer border of foot for the lowest nerve, to be able to tell that person that there are associated physical findings that confirm that and there's an x-ray or an MRI test that makes a definitive diagnosis. So I frequently tell patients that 88% of patients who come into my office, I can't tell them in a definitive fashion what the source of their pain is. This diagnosis falls into the one of the 12% that I can usually say, this is the source of your problem. But again the history is fairly is very specific actually when it's a positive case. The physical examination is relatively specific. Neurodiagnostic studies, although I don't usually get them, is usually pretty specific although there are some false negatives, tests that come back normal even though there is a truly herniated disk with an associated nerve or compression. Plain x-rays are not helpful, but the MRI scan is very sensitive and very specific for this. So of all the conditions that we can correlate the patient's history, their complaints, the physical findings, the mechanism of injury and the special studies, this is probably the best.

Just to go briefly through a couple. You heard about stenosing tenosynovitis. Stenosis refers to narrowing of any tubular structure. This is a CAT scan of the spinal canal and you can see arthritic overgrowth here and compression on the nerves. And this is common is a somewhat older population, usually starting in the 40's, 50's, 60's, 70's and manifest with back pain and pain down the legs. And although it's a relatively rare condition to see in the worker, it is certainly sometimes something that the older worker will come in with a lifelong history of, let's say being a manual laborer, and finally worn out with back pain and now having pain down their legs. And the contribution that their life of manual labor has played is difficult to sort out. AgainYwell, I'll sort of skip through this because we're running low on time. It's the same old story basically.

Spondylolisthesis, again we talked before about the prevalence of spondylolisthesis as being 5 to 6% of the normal population. Anybody out there over the age of 18, 5 to 6% of you will have spondylolisthesis. And it refers to the forward slippage of this vertebrae on this one, and the nerve that comes out here can be compressed causing pain down the leg. So in that setting it's very difficult to say when you know that one out of twenty people has this to start with, and when you know almost without exception that even the worker who has a clear cut back injury never had back pain before, and now has pain down the leg and has spondylolisthesis, that person almost invariably, that spondylolisthesis, spondylo as we call it, was present before the injury. Usually these occur between the ages of 4 and 7. So that what we see in the office is just the last straw - maybe a slightly greater slippage, or maybe a disk has bulged out or maybe this preexisting condition has started hurting for reasons that no one can explain. And I've seen any number of patients who have very bad looking spondylos who come in and never had pain until the age of 50 and you get an MRI scan and you can't believe that they've not had symptoms with this until then. But they give a very clear cut history of having been in an automobile accident, let's say, 6 months ago and have been painful ever since. And, you know, fortunately it's not my job necessarily to sort out the significance of the preexisting condition versus the aggravating factor, it's just my job to try to help that person and I have enough trouble just trying to do that. So, let me wrap up here a little bit because I'm already well over my time. I don't want to get zapped by any lightning.

In conclusion, we've seen that screening is really almost impossible for risk factors for back pain in general without getting into a whole gamut of job satisfaction and other things that are a little too complicated to discuss today. But in terms of thinking that there's an x-ray test or an MRI test, it's virtually impossible to screen for the individual at risk for back pain. The risk factors, particularly the occupational risk factors, are really still ill-defined. And it's really too bad for such a common condition with a cost to society of between 50 to 75 billion dollars a year and the percentage of the insurance dollars that go to it, as you heard this morning. It's too bad that there haven't been better epidemiologic studies done to finding what are the risk factors and how can we improve on them, more importantly, for back pain. Preexisting conditions are common in sorting out causation and aggravation. It is really almost a cellamonic task. And if I can just put in a plug here for my own personal bias which is I think, that what I've seen in Washington D.C., and before that I was at the University of Pittsburgh, is that importance of the independent medical examination in sorting through some of these cases for injured workers. And I would just make a plea for redefining the concept of the so called independent medical examination. Because the independent medical examination, as many of you know, is far from that. It's I get paid $200 or $400 or $600 by either the insurance company or the plaintiff's lawyer to give the opinion that they want to hear. And no doctor would get up here and say that they lie or that they don't tell the truth when they give that IME, but I can tell you that, you know who you're working for. And everybody who does an IME knows who they're working for. And it would be a far better system to have truly independent individuals, clinicians, willing to give IME's appointed by some neutral party such as a worker's comp referee, not worrying where the money's coming from, and giving a true independent IME. Thank you.

Dr. Howard Sandler, SOMA

Please remain seated. We have one more speaker. Thank you very much. Anybody have some Napercin? Our last speaker is Dr. Dean Louis. He graduated from the University of New Hampshire and received his M.D. from the University of Vermont Medical School. His internship was done at the Maine Medical Center, and he then served in the public health service. He completed an orthopaedic surgery residency at the University of Michigan, and went on to do a fellowship in surgery of the hand at Columbia Presbyterian. He joined the faculty of orthopaedic surgery at the University of Michigan, and is currently a Professor of Surgery in Orthopaedic Surgery at the University of Michigan and Chief of the Orthopaedic Hand Service. In 1995 he was elected to the office of President-Elect of the American Society for Surgery of the Hand, and is currently the President of this organization. We want to welcome Dr. Dean Louis.

The Challenge from a Hand Surgeon's Perspective

Dean S. Louis, M.D., University of Michigan

Taking care of patients and finding solutions to health problems and returning workers to productive lives has always been a joy. Over the course of my career as an active hand surgeon I have seen many changes. None of these has been so dramatic as the response of workers and physicians to musculoskeletal symptoms. The traditional interaction between physician and patient has been forever altered. It has become polarized and in many ways contentious. Unfortunately this has occurred as a result of a concentrated effort on the part of some to continue to promote the concept that work causes disease.

We all know that bodily tissue loss may occur in any industrial setting. I'm not referring to that. That truly represents injury. Compensation should apply, and rehabilitation should begin early. I am referring to musculoskeletal symptoms of uncertain etiology or diagnoses that have been labeled cumulative trauma disorders (CTD's) or repetitive strain injury (RSI). The proliferation and transport of these terms by the lay press and television have created an awareness among the public of the supposed problem. The power of words and the nuances of meaning can create behavioral changes in individuals who are convinced that they have a Adisease, Adisorder or an Ainjury.

Patients now present convinced that they have a CTD or RSI instead of coming to find out what is really the matter. They come convinced because they have been led to believe they have a disease process. Further, this issue is complicated by the fact that physicians are now asked to make a decision regarding causation. If a treating physician should dare to say that the problem is not caused by work, the doctor-patient relationship is further disrupted. These new behaviors have made it much less pleasant to interact with the symptomatic working population. There are hidden agendas. They come in convinced that they have a disease process that is caused by their work. Attempting to convince a person that they do not have a disease is far different from working with a patient who just wants to know what is wrong. If the patient's' agenda is directed towards illness behavior and the compensation pot at the end of the rainbow, then the problem is almost insoluble by simple measures.

There is also the issue of who pays for it. Under managed care programs the enrolled physician is encouraged to say that the worker's problem is work-related, thus shifting the cost to their compensation carrier. Treating physicians are thus faced with conflict, and moral dilemmas that ideally should not be part of the doctor-patient interaction. Of course there are always agendas, some are just easier to deal with, like ADoctor, what's wrong - I need to get better so I can go back to work.

It is my belief that the responsibility for this state of affairs is due to a massive failure of communication. Some of us in the medical profession have been slow to respond B only lately realizing that the ergonomic experts have created a language which is pejorative, implies causality, and has no basis in medical fact. Terms are used loosely; diagnosis is based upon self-reporting symptoms and a spurious epidemiological literature has arisen that has influenced public policy and private behavior.

The issues of individual variation and individual tolerances, aging, as well as other personal variables, have been largely ignored by those who are attempting to sell ergonomics. Age, conditioning, anatomical variation, personal habits, lifestyle choices, avocational pursuits, all influence an individual's ability to respond to a defined task. The upshot of this is that patients are convinced that force and repetition and their job are responsible for their symptoms, and that they have a disease process. Most upper extremity symptoms will respond to non-surgical management. Unfortunately, some surgeons have become very aggressive surgically, just adding to the problem. Additionally, one of the issues that we do not have an answer to is what are the outcomes of surgery in selected populations.

We have heard today the misuse of the terms CTD's, RSI, work-related musculoskeletal disorder, etc. Such terms imply causation and have a negative ripple effect upon the person so labeled. We need to decide upon a common specific language and to use it appropriately. Until this happens, and until there is agreement, these issues will remain unresolved.

The persons who are most affected by these problems are the worker/patient, who we as health professionals are devoted to helping. The state of affairs now is one of confusion and polarization. We can all do better. A recent conference was held in Ann Arbor, Michigan, sponsored by the American Academy of Orthopaedic Surgeons. Several of the principal players in this arena accepted and then later declined to come, because they were convinced that nothing fruitful would come of such a small consensus meeting. I believe we must do better than that. To retreat from inquiry is the antithesis of science.

Dr. HOWARD SANDLER, SOMA

Thank you Dr. Louis. We're going to take a break now, but we'll start promptly at 4:00 with two Discussants and then open the floor to comments, questions and discussion.

[Please note: following the preceeding break, the audiovisual contractor did not begin recording the audiotape until some point during Dr. Ronald Gray's remarks as a Discussant (also lost is Dr. Larry Fine's introduction of the Discussants). We have twice attempted to have Dr. Gray >recreate' the missing portion of his remarks, however he has declined due to other priorities.]

Dr. RONALD GRAY, Johns Hopkins University

[FIRST PART MISSING]Yin the general population. We clearly also have to have criteria that we can apply to the work place where there may be some particular exposure that puts people at risk. Or equally, it may be a situation where people withdraw from a certain job because it does cause problems for them. You can get the selection out for people who might be at risk. And finally we wanted to consider the way we might diagnose these disorders in a clinical setting. And there the issue is that people who come to see a doctor, if they've got any sense, have already got symptoms. They already have something to complain about. And they're different for the general population or the workplace population. Now a key question that arose in our workshop is with the validity and the reliability for these diagnoses. The validity means our ability to truly measure what we think we measure. It's a bit vague but that's what it means. And reliability means that if we measure the same thing over again, do we get the same answer? And for many of these diagnoses those two questions cannot be resolved. And I'll come back to that. But another very important question that arose at our workshop was that because it is difficult to diagnose these conditions, it's very important to measure the degree of disability that might be associated with them. And therefore a lot of the energy was put into scales of severity, scales of functional limitation, etc. I want to turn to a brief discussion of why do we need valid diagnoses? And I think the best way to address this is to ask what are the disadvantages of having invalid diagnoses? And one of the obvious disadvantages of an invalid diagnosis is that if we can't diagnose something, we're going to miss it. We're going to miss possibly serious treatable illnesses. So we don't want to do that. A corollary to that is that if we don't know how to diagnose a condition we could over diagnose it. And there's some real penalties to over diagnosis. Unnecessary interference, unnecessary surgery, for example. There are real costs. Medical costs, clearly psychological and social and occupational costs of incorrect over diagnosis. From a research perspective, a really important problem is that if we can't precisely make a valid diagnosis then all our work is subject to either a diagnostic bias or a reporting bias. Now diagnostic bias could arise if doctors were making highly subjective and not very accurate diagnoses, or if patients were over or under reporting symptomatology. Either way, if we're trying to look for the cause of these disorders so as to develop preventive strategies, we will miss them. And therefore our methods of prevention will be limited. And I think another very important reason why we need valid diagnoses is that if we do have preventive or therapeutic interventions we need to able to measure endpoints. We need to be able to say that people got better or they stayed the same or they got worse. And that is very difficult in this field. And there is a real deficit of good clinical trials in this area unlike virtually all other areas of medicine. And I think this is a very important issue that might warrant further discussion here.

Now the specific problems with musculoskeletal disorders have been touched upon by [some of the earlier presenters.] One of the reasons we have problems with diagnosis is we don't have good external tests. We heard that, for example, nerve conduction velocity is not a good way to diagnose carpal tunnel syndrome in the absence of very specific symptoms. We heard from Dr. Steve Moore that many of the syndromes that we look at are very localized regional pain syndromes. They tend to be very non-specific. Many of these conditions are multifactorial. They may be related to job stress, but they're also related to psychologic stress and a variety of other factors. There is background frequency of these conditions that we can't ignore. And particularly they increase generally with age. In addition, not only do they increase in frequency with age but susceptibility and response to injury can vary with age and other conditions.

And lastly as being noted by our last speaker in particular, this is a very controversial environment. There are worker's compensation, legal and other factors that muddy the water and affect the rationality. So what can we do? I don't think I have the answer. I think that my clinical colleagues certainly have much more expertise in these matters than I do. But I think we can go from one extreme to another. We can use very general screening tests, there are things like symptomatic questionnaires that you could apply in this room and get some overall measure of symptomatology among us. But that's very non-specific, and it has very poor predictive value in terms of real disorder. At the other extreme we have highly invasive tests, like the electro diagnostic studies or MRI which are very invasive, they're expensive, and actually they're not necessarily all that precise in the absence of fairly specific symptomatology. Because as we heard earlier on, you can have a back x-ray that looks >horrific' and the person has no symptoms. Or you can have the symptomatic person with nothing on the back x-ray. So what you end up with is that I think we will never get away from clinical judgment. A combination of tests, physical exams, history, etc. And the problem there is that it's very difficult to get replicability, both with one doctor over time or between doctors. So I have no resolution to these problems other than to say that I'm sure they're going to keep us busy for a long time. Thank you.

Dr. Larry Fine, NIOSH

Thank you Ron. And now I'd like to turn to Gary Franklin, who as I was saying is both a research professor at the University of Washington and is Medical Director of the Department of Labor and Industry, and I think is one of the physician's who spent probably more time than any other really studying analytically the worker compensation system, so he has some unique insights to provide.

Dr. GARY FRANKLIN, State of Washington

Thanks, Larry. I just wanted to hear you introduce me twice. I think we've all flunked this red light test here. Here it is. I think the thing that we can all agree on from the earlier talks on case definition is that we need to get more specific. And I think you heard some very good talks on increasing specificity of case definitions. And this slide just kind of points out that the more sensitive and less specific use of symptom questionnaires, for example, or adding a physical examination to symptoms can be good for epidemiologic studies, case finding studies, surveillance and even some studies on causation. But if you want to do a clinical trial or do a very strong etiologic study, you probably have to have a more specific if not very sensitive set of tests. So in the case of carpal tunnel, as Dr. Evanoff pointed out, you would want a positive hand pain diagram, and maybe on examination, numbness in digits one and two, but also you'd want an abnormal nerve conduction study to be certain that somebody had the condition. And all of this relates to how much money you have to spend on doing this study, and that you have to look at the resources you have to find the cases and what the objective of the study is. If the objective is simply to define population base rates, you can use very sensitive, but not necessarily very specific, case definitions. But if you want to do surgery in somebody, or be certain that somebody has a condition, you'd want to use a more specific, but maybe less sensitive case definition. Just to speak for a moment about some of the studies and case definitions that have been used in the literature, Peter Nathan has frequently used his maximal latency difference in the carpal tunnel to define carpal tunnel syndrome. And as Dr. Evanoff mentioned earlier in his study of Japanese furniture makers, using that one nerve conduction test to find 17% prevalence, but only felt that 2% of the workers actually had carpal tunnel syndrome. And that, his articles, his studies are those studies that are used primarily by folks like Dr. Louis to argue that carpal tunnel syndrome does not exist. Well, Barbara Silverstein, on the other hand, who did one of the seminal studies on the etiology of carpal tunnel syndrome used a combination of symptoms and physical exam findings, but very detailed, structured interview and examination. In my view, far better than Dr. Nathan's overly sensitive and not very specific nerve conduction tests. And she has taken an enormous amount of heat for not using nerve conduction tests for that study. But in my view, in fact, it falls somewhere in the middle here of being a moderately sensitive and moderately specific way of doing things in the field.

But the key here in talking about how you apply case definitions is to think about how useful and practical the use of the case definition is in whatever your task is. So if you're trying to figure out which exposures you want to target, which industries you want to target, it's okay to use a very sensitive but not very specific case definition to do case finding and to generate epidemiologic rates to define exposures that might be causing the disorder for the purposes of primary prevention. You can also use case definitions once a person has a disorder to create treatment guidelines about what kind of treatments might be most useful to help that patient get back to work or to prevent further disability. And you can call that secondary and tertiary prevention. A good deal of disability that occurs in worker's compensation contrary to some of what's been said today, comes from inappropriate medical care. So we also need very strong case definitions, maybe the strongest, in trying to decide which treatments to do to workers - which invasive treatments particularly should or should not be done. These are some carpal tunnel rates from Washington State, again, similar to what Barbara Silverstein showed earlier today. We used a case definition of a physician's diagnosis on an ICD code on a submitted bill to worker's comp to count our cases. And we were able to use that kind of methodology, probably very sensitive but not very specific, to come up with the fact that there was a dramatic difference in rates between industries. And you could not explain away that difference in rates by saying that psychosocial factors alone or other things alone caused carpal tunnel syndrome. The rate in oyster, crab and clam packing, in the meat and poultry dealers and others, was 25 per 1,000, whereas the overall industry-wide rate was about 1.8 per 1,000. And it went all the way down. And the rates were much lower in keyboarding in government workers. Now it's true, when you use that kind of a case definition that is an administrative database that is relying on a bill or an ICD code on a bill, where you really can't examine the patient and you're not really reviewing all of the medical records, you're going to have some misclassification. And the key is that you do not have any misclassification that varies by industry because you're trying to demonstrate industry-wide differences. And we were able to demonstrate industry-wide differences. But also, by reviewing a random sample of medical records, we were able to demonstrate there was no misclassification by industry. And so we were able to conclude that the wide variation in rates was more due to the exposure in the various industries than in any misclassification or selection bias that we may have had. And we also found that there are a lot of coding errors in this kind of a database. And you've got to be doing some sort of reliability and validity study when you use large databases and have a case definition that relies on large databases. We found as much as a 25% rate of coding that was in error. But in the end we were able to conclude that we under counted by about 20%, and we over counted by about 20%. So in the end we concluded that, really, on balance we were finding about the right number of people that are probably out there. I just wanted to point out that Larry Fine has actually written about some of this. And this is a slide from a 1986 paper that he published talking about the advantages and disadvantages of using preexisting data like the worker's comp databases versus actually doing examinations and structured interviews in person. I think you do have to take into account cost benefit relationships when you're trying to decide what kind of methodology to use and what kind of case definition you want to apply.

Now I'm just going to finish up here very quickly by moving to the side that talks about how case definitions are useful in applying appropriate treatment to injured workers. We did an outcome study of thoracic outlet surgery in Washington State. And we were able to compare patients that had the surgery with a thoracic outlet diagnosis to those that had no surgery, but still had a thoracic outlet diagnosis. And there was very little difference between the two groups other than the fact that they had surgery. And you can see here that the mean medical costs were much higher in those that had the surgery. A dramatically greater proportion of workers who were still disabled one year and two years after the surgery. And if you want to know where all the money goes in worker's compensation, the old saw is that 5% of the cases take 85% of the dollars. These 158 cases that went through thoracic outlet surgery cost the State ten million dollars over a number of years. This is a long-term follow-up study. So the key in worker's compensation in terms of cost is trying to prevent the disability. In my view, much of the disability is preventable. And to the extent that the disability is being added to or caused by inappropriate or unnecessary surgery, it is very important to have appropriate case definitions for whether somebody actually has, for example, thoracic outlet syndrome or not. So for example, we create treatment guidelines with the Washington State Medical Association. These are community based guidelines. And in 1989 we met with all the thoracic surgeons in the Seattle area and created a treatment guideline for when somebody might get the surgery for thoracic outlet syndrome. And the and's and the or's in this algorithm really is the case definition for the presence or absence of thoracic outlet syndrome. And this particular one includes numbness and tingling in the arm, the presence of provocative signs. You raise the arm up, you turn the head to the left and his pulse goes down. But not really any objective test of abnormality like a nerve conduction or an EMG that really would tell you that the brachial plexus was being compressed in the thoracic outlet. And this case definition which was extremely non-specificYit may be hard to see this. These are numbers of thoracic outlet surgeries that occurred per year between 1986 and 1991. The guideline was created here with that case definition. And you can see that that guideline led to no change whatsoever in the thoracic outlet surgery rate in Washington State. So we did this thoracic outlet study, outcome study, and found these terrible outcomes among these workers. And went back to the drawing board with the Washington State Medical Association, and added to the case definition and therefore the guideline for who would have thoracic outlet surgery, the fact that besides the pain and the numbness and the tingling and provocative tests which were all non-specific, the patient now had to have abnormal EMG or nerve conductions that would tell you that there is actually some damage or some ongoing irritation of the brachial plexus as you would expect to see in true classical thoracic outlet syndrome. And lo and behold, the more specific version of our thoracic outlet guideline led to a dramatic decline in the thoracic outlet surgery rate - 50% decline or more over a two-year period from the time the thoracic outlet surgery guideline was introduced.

And the last example I'll tell you about is lumbar fusion. And I wanted to ask Dr. Lauerman in his response ifYI was surprised that being a spine surgeon he didn't speak about mechanical instability of the low back and about a case definition for diagnosing mechanical instability since the most invasive surgery done by a spine surgeon is lumbar fusion and it is based on the presence or absence of mechanical instability. Now maybe he's one of those fortunate few spine surgeons that doesn't do many lumbar fusions for internal disk derangement. And we actually think that this isYI personally think this is a horrible procedure. Our own outcome studies have demonstrated that patients who receive this procedure in worker's compYand injured workers are twice as likely to get the procedure as non-injured workers - don't do well. In fact, many of them do a lot worse. This was a guideline that was created in 1988 by the Washington State Medical Association. And this is just an example of how many doctors do things to patients based on case criteria that do not exist. There aren't any clear case definitions for the presence or absence of mechanical instability. But if you got a bunch of spine surgeons around a table to ask them, Awhen do you think a person has mechanical instability; when should you do a lumbar fusion?, they would tell you, well, they should have pain for a certain period of time that should be intractable and maybe they have a positive diskogram or some other findings on x-ray. Most of these patients did not have measurable mechanical instability with translation of the spine as you would normally see with high grade spondylolisthesis. So this is an example of a very non-specific treatment guideline based on a non-specific case definition, but it was strong enough to lead in our state to a 20% immediate decline in lumbar fusion rate in Washington state, and overall a 40% decline in the state over the past five years. So we have used case definitions both for identifying appropriate industries for targeting, and that's why the kinds of data that was presented to you this afternoon is really important. And the kind of data that would help with targeting such as presented by Silverstein this morning, are really critical for focusing resources both on a regulatory side and on the employer side.

But I'd just like to finish up by saying that I was astounded with Dr. Louis's remarks this afternoon. He seems to have the Norton Hadler approach to upper extremity cumulative trauma disorders, and that is that if work is customary and comfortable it probably could not cause these disorders. I don't know if Dr. Hadler or Dr. Louis have ever done this kind of work that leads to these disorders, but listening to him is one thing, reading the last paragraph in the handout he left outside is quite another thing. He says, It is also entirely appropriate that we assist in changing an archaic workers' compensation system that supplies rewards for lassitude and sloth rather than rewarding wellness. And I think we've heard a lot this morning that would help us move along the track, increase our knowledge, come together on all this. I think that this kind of an attitude can only serve to separate us. And if I were the medical director of a workers' comp network, I would probably send my patients to Dr. Louis for surgery, but I'd have an occupational medicine doctor see them first to talk about returning to work issues and causality.

Dr. LARRY FINE, NIOSH

Before we take questions I'd like to invite the presenters from the afternoon session to come up to the stage here. We'd like to start off with questions first directed at the presenters. And we do have several microphones throughout the hall. So why don't we start over here, David?

Mr. David Sarvadi: Larry, while they're sitting down why don't we go ahead and start with something that maybe will make things a little lighter in here. My name's David Sarvadi with the Coalition on Ergonomics. Just as an aside to Dr. Lauerman's lawyer jokes, since I'm one of the few lawyers in the room, I guess I get to say this. But there is a fourth item in your list of things reasons why lawyers are now being substituted for lab rats, and it's a down side. The problem with using lawyers instead of rats is that we're not sure we can extrapolate the results to humans. As a lawyer I get to say those things, I guess.

Dr. Evanoff, this is a question for you. On one of the slides that you had up in talking about specificity, sensitivity and PPV. The very first line had a predictive value of 44%, 0.44, is that correct.

Dr. BRADLEY EVANOFF: Yes, again in a population with a prevalence of 10%. And

Mr. Sarvadi: Well, I guess my question is, am I correct in concluding that the combination of symptoms and pain then is predictive in fewer than half the cases?

Dr. EVANOFF: You have to be cautious when you interpret the statistic called a positive predictive value. That's the number of people with a positive test who actually have the disorder. It's critically dependent on the underlying prevalence of disease. And so that gave a prevalence of disease of only 10%. So if you did no test at all, you could assume that 10% have the disease and 90% didn't. So in those circumstances, even very good tests, and if you really want to get a fright, look at the positive predictive values for things like a cardiac stress test. But I mean, the predictive values for things that we use every day in clinical medicine can be fairly low in instances where the disease prevalence is low.

Mr. Sarvadi: I accept that. And I think you're making my point. My question is, you've suggested that that kind of test is suitable for an epidemiological study looking at etiologic causes. And if the test that you're describing is only accurate half of the time, then I'm questioning how can you say that that's the kind of test that we ought to be using in that kind of study?

Dr. EVANOFF: Well, I think this deals with the issue of misclassification in epidemiologic studies. And if there is some misclassification in a study, which there is in all medical studies, if it's not a differential misclassification then actually what it does is it leads your study to find no difference or no result. So if you look at a study that shows a five-fold difference in a disease between an exposed and an unexposed population, and if you have that degree of misclassification, in all likelihood the real difference is much higher, on the order of eight- or ten-fold difference in the study. So this addresses the issue of misclassification which probably takes a longer answer than we can give in this forum.

Dr. FINE: Another question over here on this side of room?

Mr. Marc Freedman: I'm Marc Freedman with the Coalition on Ergonomics. This is for Dr. Evanoff please. Is this maybe the difference between the medical perspective and the layman perspective, but if I heard you correctly, I think there's a similarity, or would you agree with Dr. Moore's statements about carpal tunnel syndrome being heavily over-diagnosed. When I looked at your slide on the ordinal values, it seemed to me that after the first two categories there was great debate as to how you characterize carpal tunnel syndrome, leading me to believe that there's a lot of cases that may or may not qualify. Some of the other material that you presented suggested that as well.

Dr. BRADLEY EVANOFF: Right. Again, let me emphasize before I answer the question, the point that these are meant as research definitions, not as diagnostic tools that you would use in a clinical or a legal setting. I think Steve's point probably brought up two issues. One is that, as someone who goes out to companies who have had a problem with a number of upper extremity injuries, carpal tunnel syndrome often gets used as a lay person's rubric for any pain distal to the shoulder. And when you get out to some companies you may find that, yes indeed workers are having a lot of problems, although it may not be carpal tunnel syndrome that they're having the problem with. And it may be other painful problems of the distal extremity. And so, that's one issue.

I think tendonitis or other conditions in the upper extremity are probably under-reported relative to carpal tunnel syndrome. And then the last issue is, I think there certainly is in some settings over-diagnosis of carpal tunnel syndrome that you see clinically. I see patients coming to me with that diagnosis who I don't agree have carpal tunnel syndrome. So I think it was a combination of all three of those factors.

Mr. Freedman: If I may, just one more quick question. You just said that your use of the terms is for research purposes, not for diagnostic purposes. If that's the case, when are we going to be using the same language, or at what point can we rely on the term to be meaning the same thing in different contexts?

Dr. GARY FRANKLIN: Well, just on carpal tunnel syndrome, for example. In Washington State we don't allow surgery to proceed on a patient with carpal tunnel syndrome unless they have a nerve conduction abnormality. So that would fit into the definite case on the research side. There was a survey done, American Society of Hand Surgeons some years ago that found that about a third of all hand surgeons almost never do nerve conductions before they operate on people with carpal tunnel syndrome. So part of the problem is how clinicians approach these cases. We certainly see some of the worst cases in workers' comp where surgeons are literally marching up and down the patient's arm. I call it the syndrome with the spreading diagnosis. That's where the specificity of these individual diagnoses would come in handy, to be able to tell whether in fact somebody really has one or not. But this marching up and down the arm thing leaves the patient completely disabled after several years of inappropriate operation.

Dr. EVANOFF: I think there is overlap between the research definitions and clinical definitions. We didn't set out to create clinical definitions. But, there is significant overlap. I think, clinically though, you do different things based on either higher or lower standards of proof. I'm quite willing to start conservative treatment on someone who I think has carpal tunnel syndrome based on symptoms in a physical exam. Quite willing to make workplace changes, but I'm not going to do things expensive or invasive until we have a more certain diagnosis based on nerve conduction studies.

Dr. FINE: Suzanne Rogers.

Dr. Sue Rogers: Yes. If I could just make a comment. The ratio of carpal tunnel to the tendonitis and so forth, I wondered if, we've found at Kodak one carpal tunnel to nine back in'79Y.to nine actual tendonitis or related forearm kinds of problems. I'd like to know from all the panel members if you could tell us what you think the background population value is outside of taking the specific work force. I've seen this run from Barbara's point six, I think it is, up to 10 percent of the population. Do you have a figure for us?

Dr. FINE: Are you asking for a carpal tunnel syndrome, both work-related and not, or only work related?

Dr. Rogers: I'm asking for forearm problems and then carpal tunnel as a separate issue.

Dr. FINE: But, work-related, or all?

Dr. Rogers: Well, I'm looking for background rates.

Dr. FINE: Okay. So, all.

Dr. Rogers: Everything.

Dr. DEAN LOUIS: The one study that's been repeated twice is the study by Stevens in Rochester, Minnesota. And they showed a background, if I remember, it was not stratified in any way, was in the baseline population, 1%. I think that'sYdo you remember that?

Dr. GARY FRANKLIN: I think it was one per thousand.

Dr. LOUIS: One per thousand. Okay, one per thousand - point one percent. Very, very low.

Dr. Rogers: And that was just people who came in there, is that correct?

Dr. LOUIS: It was a population based study.

Dr. Rogers: But it was just people coming to the clinic, is that right?

Dr. LOUIS: No, not people coming to the clinic. People from Omstead County, Minnesota.

Dr. FRANKLIN: It was one per thousand, per year, and then in workers' comp in Washington, it's more like 1.8 to 2 per thousand within the worker population. But the background rate is one per thousand.

Dr. LOUIS: Let me just make one comment just to clear something up. I congratulate you on your study, Dr. Franklin, about thoracic outlet syndrome. That was long overdue. And it certainly shows that there is a lot of unnecessary surgery. And what you said about the upper extremity, I couldn't agree with you more. And I mentioned that in my comments, but perhaps it escaped you. There is too much surgery being done without diagnostic criteria, and what you described is something which we all have experienced. I'm not personally active in doing those things myself, but I certainly am aware of the problem. And this is one of the difficulties in terms of dealing with the patients in the workplace. Some physicians take great latitude in how they make a diagnosis without specific tests, and use that authority to proceed with all sorts of surgical adventures. And this is an unfortunate and a reprehensible practice, and I neither practice it nor condone it.

Dr. FINE: Returning to Suzanne Rogers. Thank you, Dean. Dean, do you want to Steve, do you want to comment about your view of prevalence, population based prevalence?

Dr. STEVE MOORE: The only estimate that I'm aware of for a musculotendinous disorder comes from the Nordic countries from a study by Allander, which I think demonstrated that the prevalence of epicondylitis was generally less than 1% of the population, unless you hit the decade of 40 to 50 years of age when it goes up to about 10%, and it comes right back down again which was suggestive, and also consistent with Goldie's observation that this is not a degenerative condition related to, let's call it maturation, seeing that I'>m in the target age group. When you look at trigger finger and deQuervain's disease and peritendonitis there are no studies available to estimate the population incidence or prevalence of those conditions. All you have are case series which are highly dependent upon referral basis and you can't find the denominator to go with it. I would also reemphasize this point. It's been very frustrating for me, as a muscle/tendon kind of guy, that people will debate about the work-relatedness of carpal tunnel, et cetera, and suddenly ecologically expand that to include every disorder in the upper extremity. I think it's been a real problem to throw the baby out with the bath water, whether or not the bath water even needs to go out. But I think we have to be very careful in making sure that carpal tunnel is one thing, nerve conduction impairment's another, but tennis elbow, deQuervain's, those are completely different distinct entities. And just because you think you've disproven or shown a lack of association with CTS, doesn't mean that you throw out the whole arm issue.

Mr. Eric Frumin: Dr. Louis. I got the impression from what you said that you've never seen an association in an epidemiologic evaluation of a group of workers, an association between the exposure and some increased rate, incidence or prevalence of these disorders that you could accept. And, I wanted to ask you whether, first, whether my impression is correct? And then secondly, if it is, whether you've had the opportunity in your clinical work to see an association between the exposure of any of your patients to some work related risk and the effect that it's had on them such that you would actually recommend that they ceased exposure to some or another hazard. You know, in other words, looking at the exposure, the job related problem either on a group basis or on an individual basis, have you even seen either an individual or group of people who you thought their problems were as a result of the work that they did?

Dr. DEAN LOUIS: Collectively, we have fallen into the trap that concludes that because a person has symptoms after work, that work is responsible for the symptoms. This is an insufficient and limited view of the event. Let us consider human tolerance and capabilities. Why not place the responsibility on the worker who smokes, is overweight, and out of shape, and whose physical tolerance for activity (any activity) is diminished as a result. Because a given individual is incapable performing a specific task does not tell you why that is so. Few of us could ever perform the decathlon. Some people can after intense training. This, of course, is an extreme example only to point out the variety of human tolerance. We have a long way to go in understanding why one person develops symptoms and another doesn't. To consider just one aspect of this problem as responsible is short-sighted at best.

[The Audiotape ends here; our sincere apologies to Conference Registrants - we have no means to recover the remaining questions and answers to the end of this session, we can only acknowledge what followed here:]

Question by Dr. Brad Joseph, Ford

Response by Dr. Dean Louis

Follow-up question by Dr. Joseph

Responses by Dr. Gary Franklin, Dr. Bradley Evanoff, and Dr. Steve Moore

Question by Mr. Tom Slavin, Navistar

Responses by Dr. Gary Franklin, Dr. Bill Lauerman, and again by Dr. Franklin

Question by Mr. Eric Frumin, UNITE

Responses by Dr. Dean Louis and Dr. Ron Gray

Question by Mr. Denny Ankrum, Nova Solutions

Comment by Dr. Larry Fine

Question by Ms. Joanne Wojcik, reporter.

Responses by Dr. Gary Franklin, Dr. Bill Lauerman, and Dr. Steve Moore

Comment by Dr. Steve Lamm, Consultants in Epidemiology and Occupational Health

Comment by Ms. Peg Seminario, AFL-CIO]