The Farmer, the Camel and the Cow with a Bad attitude.

farmer sickleFarmers in Ireland are bred tough. They don’t come in to see their Rheumatologist with a bit of muscular pain. They don’t come in with a few Heberden’s nodes. When they do arthritis, they do it properly.

The typical encounter with a farmer would be a chap who hobbles in ruefully, asking politely ‘Sorry to bother you doctor, but is there anything you can do for my knees?’. I usually rub my beard sagely and say –  ‘Hmm, let’s have a look at your hips, then’. Chances are that his knees are fine but his hips are shot to pieces. By the time they stumble into clinic you can guarantee that their hip joints are done for. It never ceases to amaze me that someone can develop severe osteoarthritis (OA) of the hips without ever having experienced pain in the hips or groin. And by the way, I still don’t know why OA of the hips is an occupational hazard for farmers. You see people from other heavy occupations who don’t seem to have the same problem. Perhaps it is something to do with the heavy work most of them did on the farm as young teenagers.

Another common example is the farmer with a seemingly straightforward shoulder capsulitis or tendon rupture. When you get down to the story, however, you find – for instance – that he has been dragged halfway up the field hanging on to the back of a tractor driven away by his 12 year old son who had found the accelerator instead of the brake. Perhaps the most memorable ‘rheumatological case’ was a farmer who hobbled painfully into clinic leaning precariously on two walking sticks with the worst knee joints I’ve ever seen. The valgus deformity on each side was at least 45 degrees and the knees had rubbed together with such violence that he had developed huge suppurating rheumatoid nodules.With each step he had to swing from one side to the other using his stick a bit like a pole vault. When I got him up onto the couch his legs seemed to be connected to the thigh by no more than half an ACL. How this man had continued working on the farm for the previous 5 years I will never know. To my amazement, he recovered well after his knees were replaced and he is probably still working on his farm to this day.

camelAnyway, I digress and really must return to the subject of my story. I could barely contain my excitement when a farmer come in to clinic and showed off his impressive knee swellings. I immediately recognised this as the little known ‘Bactrian camel’ sign. The ‘double hump’ is a dead giveaway. Given the usual yarns, I was not surprised when he told me the story of how this all started with being kicked on the shin by a mischievous young cow. It’s never something mundane like an ingrown toenail. For a farmer this is a regular hazard – but his story was tame compared to one of my patients recently who had barely escaped with his life after having been trampled over by half a dozen stampeding cows. Anyway, our farmer had developed a cellulitis in his ankle which was eventually treated in the local hospital. I say eventually because it is highly unlikely that a farmer would go to see the doctor when there was only a little redness around a scratch. He recounted how the swellings had started after the kick and now they were getting a bit painful. When he took his trousers off I was taken aback. The swelling on top of his right kneecap was like a six inch ‘bap’, and just below that he had a second fluctuant swelling above the tibial tuberosity. Both the pre-patellar and infra-patellar bursae were swollen and clearly displayed in all their glory. The profile was unmistakably that of a double hump-backed or Bantian camel. I wish I had taken a photograph for posterity. A large amount of clear fluid was aspirated from each swelling and steroid injected, to the patient’s evident relief.

As an educational aside, it is well known that infected bursitis often occurs in association with cellulitis, and a good long course of antibiotics is called for. However, this is not the first time I have seen ‘post-infective’ bursitis persisting long after the infection has cleared up. There isn’t a strong evidence base when treating bursitis, but in my experience a small dose of steroid can do the trick quite effectively. Some experts counsel weeks of inactivity, but in my experience if you tell a farmer to rest it is likely that you will be wasting your breath.

The old Irish farmer belongs to a stoical and dependable tradition, in touch with the land and in command of their over-worked but hardy frames. But are they a dying breed? I now see so many of them working on well into their 80s because their children have moved away seeking greener pastures and easier lifestyles. As a rheumatologist who sometimes tends to their ‘war wounds’, I salute them.

Would you be my Bematist?

alexander-the-great-mapIf I had been around when Alexander the Great was looking for volunteers I definitely would have looked the other way. History tells us that he employed two men – Diognetus and Baeton – as ‘bematists’ to accurately count their steps and thereby measure the distance between cities as his army marched through Asia. Quite how they managed to achieve astonishing levels of accuracy is sadly not recorded, but that is one job I would certainly not have signed up for! Strange as it may seem, but accurate step counting is back in fashion!

The proposal I am going to discuss in this article is that accurate monitoring of physical activity should now be adopted as a standard outcome measure for research in arthritis and sponyloarthritis.

I make no claim to originality here. When I was a trainee in Newcastle upon Tyne in the early 90s, David Walker introduced me to an activity tracker called the NUMACT monitor. Using this monitor, his group quantified the improvements in the actual daily mobility of RA patients after knee injection (70%), and after NSAID use (50%). They then showed a significant (79%) improvement in mobility in OA patients six months after total knee arthroplasty. He compared these results to questionnaires that ask about physical mobility (such as the Nottingham Health Profile) and reported significant differences. Until recently, there had been very few other such studies in patients with arthritis, and I believe that it is about time we followed his forward thinking approach. Before we take a look at these studies, let’s review the rationale for going to all this trouble.

The stated aim of the International Classification of Functional Disability and  Health (ICF) is to “put every person in a context: functioning and disability are results of the interaction between the health conditions of the person and their environment”. They suggest that a new framework for outcome measurement is required, and in the area of assessment of general physical mobility we need to look at the issue through three different ‘lenses’. The first is to use self-reported assessment of problems in daily function using questionnaires such as the HAQ. The second is to measure physical activity under ‘laboratory’ situations (e.g. a ‘six minute walk test’) and the third is to monitor physical activity in the patient’s natural environment (at home, work etc.). This has to be a positive move for patients, as it will help to ‘re-set’ our target towards one that will get our patient back participating fully in their ‘natural environment’ (work, home and leisure).

At the moment the only one of these included as a ‘core outcome measure’ by EULAR/ASAS is the questionnaire assessment. There is no doubt that this is very useful and relatively easy to collect as part of a research project. But ultimately the goal of our treatment has to be to get the person back to the optimal level of function in real life. So can a questionnaire such as the HAQ accurately tell us when the patient’s function has returned to normal? Certainly not: there is a ‘floor effect’ that will not detect changes between no disability and low-moderate disability. And how can you really take account of aids and adaptations? Does the HAQ correlate well with 6 minute walk tests or activity monitoring studies? The research here is still a bit patchy, but it appears that these three ways of measuring of function may have quite different stories to tell.

stepwatchWe should perhaps eat a little humble pie and learn from our orthopaedic colleagues. In the past the success or failure of hip replacement was mainly judged by measuring standard surgical outcomes such as mortality and morbidity but for some years now they have realised that with a younger and more active group of patients the expectations of a good outcome are higher. Surgical success rates are higher than patient satisfaction rates for both hip and knee arthroplasties, and actual improvements in physical activity can lag even further behind. This is a problem, as we have always assumed that an important goal of these operations is to return the patient to a relatively high level of physical functioning. A study by deGroot et al (2008) looked at recovery after hip and knee arthroplasties using all three methods: the SF36 questionnaire, the 6 minute walk test and monitoring daily activity using an accelerometer device. This study reported substantial improvement in self-reported physical activity whilst daily physical activity had not improved to anywhere near the same extent. This surprising result will hopefully lead to renewed efforts to improve rehabilitation programmes.

I also look across at the nearby cardiac rehabilitation clinic with some admiration. Research has shown that a good ‘medical’ outcome after acute myocardial infarction is often followed by a significant loss of mobility and participation in normal activities, and the success of cardiac rehabilitation programmes has been most impressive. What amazes me is how well motivated their patients seem to be in sticking to the rehab plan! I think that our patients who continue to suffer from pain and stiffness due to their chronic arthritis probably need a little more help and encouragement using technology such as activity monitors. In a small but important randomised study Talbot (2003) showed that in patients with symptomatic knee OA, a 23% increase in monitored physical activity was seen in the group given a pedometer with a specific rehabilitation plan whilst those who were given normal education advice showed an actual fall in performance. A striking difference between groups was also evident in tests of isometric muscle strength (the ‘laboratory’ test). This sort of study should give us hope that outcomes can be improved by employing technology in targeted rehabilitation programmes.

So, if we’re all agreed that this is a laudable thing to try and measure, what then should we be measuring and what is already known in this area? ‘Physical Activity’ has a technical meaning in Sports Medicine that many readers may not be aware of: it is all about energy expenditure in METs rather than just measuring steps or miles. If we are talking about achieving high intensity in daily physical activities we need to measure more than the number of steps. In my next post we’ll have a closer look at the approach chosen by a couple of recent studies. I’ll also be having a look at some of the technology out there with an emphasis on the scientific validation of the different monitors. And of course, I must emphasise that at the moment I’m just looking at general function rather than specific joint movements of relevance to the specific disease in question (such as the electronic goniometric glove, for instance).

And by the way, I didn’t know what a Bematist was either – but I thought it might get your attention!



Towards a Dataglove for Arthritis…

As rheumatologists we spend a lot of time examining the hands of people with arthritis. We examine them carefully for swelling and tenderness and use this information to assess disease activity. The effort to make this a more scientific assessment has paid off in the success of ‘treat to target’ protocols. This is fine for early arthritis, but there are other important dimensions of the problem we are only beginning to understand. Historically, rheumatologists have not been particularly interested in the scientific measurement of joint movement and function. This is a major gap in our understanding of arthritis and how it affects our patients from day to day. If we could better understand how stiffness affects joints, we could help our patients optimise their function and perhaps allow them to remain in work by testing different functional strategies in the workplace. Stiffness is also the forgotten member of the trio of symptoms that are widely regarded to reflect disease activity in rheumatoid arthritis. Technology has now reached the point where it should be possible to get accurate biometric data to record joint angles, movement and even touch. And technically it is now possible for these measurements to be recorded at home or in the workplace with minimal interruption to the normal routine. So enter stage – the ‘dataglove’. So perhaps it’s time to let our patients’ fingers do the talking…

The history of datagloves

The first wired electronic glove was patented by Thomas Zimmerman in 1982, and in 1989 Jaron Lanier patented ultrasonic and magnetic motion tracking technology to create the Powerglove. Lanier will also be remembered for coining the phrase ‘virtual reality’. The optical flex sensor used in the dataglove was invented by Young L. Harvill who scratched the fiber near the joint to make it locally sensitive to bending. The Power glove was designed for gaming but sadly for Lanier he was to lose control of the company VPL research. Interestingly, one of his VR predictions has been fulfilled: “Medical students could practice surgery on virtual cadavers that spurt virtual blood after a misplaced incision. Such uses are speculative so far, but few people doubt the technology’s potential”. Potential virtual reality applications in rheumatology include virtual homes or workplaces where the patient can explore functional problems and possible solutions. However, Lanier was well ahead of his time, and viable medical VR applications are still few and far between.

The Technology of Datagloves

The state of the art in current dataglove technology is represented by the 14-sensor 5DT dataglove ‘Ultra’ and the 22-sensor Cyberglove II. You can watch videos online of the 5DT and Cyberglove systems. At the moment these are powerful but expensive gloves which often require considerable effort to calibrate and customise. Working in conjunction with a team from the University of Ulster’s Integrated Systems Research Centre in Magee, we have done some work on programming and customising the 5DT Dataglove Ultra for patients with arthritis. Our work so far has focused on improving the repeatability and ease of calibration of the glove. We have also carried out work on the effect of using an thin inner glove – important in the healthcare setting to avoid problems with cross infection between patients. The 5DT dataglove uses optical fibre bend sensors. This technology is already fairly accurate, but has its limitations. We have therefore developed a new type of multi-functional dataglove using no fewer than 47 sensors. Our dataglove is designed to be used by people with arthritis by incorporating features to enable it to be easily put on and taken off. This will be tested in a group of patients with arthritis. We have also developed a user interface that will help patients calibrate the glove independently and use it accurately in the home setting.

project ‘digit-ease’

Our aim is to test the feasibility of using a sophisticated dataglove to take detailed measurements of joint position and movement in people with rheumatoid arthritis. We will test how closely the measurements match how our patients are feeling. We will be testing how much discomfort patients have in using the dataglove and how easy or difficult they find the visual interface. We hope that if the initial tests prove successful that we will be able to test the dataglove’s ability to detect changes in function after treatment.

For those of you who are interested in a little more detail, I have posted a copy of my recent presentation at the .med conference in Dublin on Dec 7, 2012 (#dotmed on Twitter). My co-workers James Connolly, Kevin Curran, Joan Condell and I have also recently submitted a paper for publication on the use of neural network theory to improve the accuracy of the data glove. I will provide a link if and when it is accepted for publication as this paper contains a lot more detail about our results so far.