2019 ACVS Keynote Highlight: 20+ Years of Measuring Lameness in Horses

Laurie Tyrrell-Schroeder, DVM, Equinosis Director of Veterinary Services | DOWNLOAD PDF

At this year’s ACVS Annual Surgery Summit, Kevin Keegan, DVM, MS, DACVS, Professor of Equine Surgery at the University of Missouri, and co-inventor of the Equinosis Q inertial sensor-based lameness measurement system, presented the equine keynote “20+ Years of Measuring Lameness in Horses: The Practical Significance to Equine Veterinarians”.  As an active equine clinician, board certified surgeon, university professor and researcher, Dr. Keegan has devoted much of his life and professional career to studying equine lameness and developing a method to measure it with high sensitivity, accuracy, repeatability and practicality for the practicing equine veterinarian.  As an overview of his presentation, we share in this article what has been learned by measuring lameness over more than two decades, and a glimpse at what the future holds.

Dr. Keegan began the discussion with the central concept that lameness is a clinical sign, not a disease. As described by Percival as early as 1873, “lameness is but a symptom… not a disease itself. Lameness may exist independently of disease”.  This concept is important to keep in mind as lameness measurement in horses becomes more widely used in the veterinary profession, both for the diagnosis of lameness and for performance issues, as well as athletic horse monitoring and management.

Dr. Keegan reminded us that measuring lameness is only the beginning of the process. Determining what may be bothering the horse, if anything, is dependent upon the skills and knowledge of the evaluating veterinarian. However, measuring lameness allows you to identify things that you may not realize without the ability to do so. Using objective methods to measure lameness is like using a microscope, allowing one to “see things better”.

Detection of lameness has historically been determined subjectively through visual observation. Dr. Keegan cited the poor agreement between even experienced veterinarians using subjective evaluation of lameness.  Why is this?  There are many reasons, which include the use of unreliable or deceiving signs, our limited temporal resolution (how fast our eyes can process movement), bias, and the known inherent stride to stride and trial to trial variability of lameness in horses.  While objective measurement of lameness is not new, it has mostly been limited to research settings due to expensive and time-consuming methods, such as the stationary force plate and treadmills with cameras and markers.  The only practical method to measure lameness in the field today is with body mounted inertial sensors.

While the clinical sign of lameness can be caused by many things, pain is the most common. Pain in the limbs during weight bearing will result in the horse either landing with less force or pushing off with less force in that limb.  Force = mass x acceleration. Decreased force on the limbs is most closely approximated (better than any other movement parameter) by the vertical acceleration of the head and pelvis. Because we don’t see acceleration very well, acceleration can be converted to position, so that changes in position can be measured and reported. If you want to know if lameness exists, the limb or limbs involved, and the amplitude and timing of lameness (impact or push off), this can all be determined by information contained within the vertical trajectories of head and pelvic movement. 

The detection of lameness using head and pelvic trajectories is based on a method of fault detection. Because vertical head and pelvic movement are periodic signals embedded in the overall random movement seen or measured, you can separate movement that is periodic from that which is not. The periodic components are the horse’s normal vertical movement and vertical movement due to pain. This is particularly useful (and vital), as filtering out random movement, unrelated to lameness, can be achieved. Inertial sensors also hold a great advantage over the stationary force plate due to the ability to collect multiple contiguous strides, allowing for better handling of inherent stride to stride variability. 

Dr. Keegan went on to describe some of the most clinically important lessons gleaned from measuring lameness in horses. 

1. Lameness is frequently unstable

Lameness can spontaneously change in limb, amplitude and timing.  This is problematic for subjective evaluation as it can interfere with our interpretation of diagnostic blocks or treatments.  Lameness may get worse because of exercise exacerbation or simply because the horse is getting more comfortable with the evaluation process. Lameness can also get better due to warming out of the lameness with some exercise or due to some other stressful stimulus masking the pain causing the lameness. 

Takeaways:

• Unstable lameness is often not appreciated with subjective evaluation alone but is identifiable with objective measurement.
• Do not launch into a diagnostic blocking evaluation without confirmation of a stable lameness.

2. Interesting observations of diagnostic blocks

Gradual improvement in a lameness with successive blocks may be an indication that the lameness is secondary (overloading a second limb due to a primary lameness elsewhere). Secondary lameness is difficult to localize. Additionally, blocking a normal horse will cause it to measure as a little lame in that limb. This has ramifications for the veterinarian in the blocking investigation as complete resolution of a lameness, especially when several distal limb blocks have been performed, may not be expected.

Takeaways:

• When small incremental improvements in lameness are seen with successive blocks, be sure you are not overlooking what may be a primary lameness elsewhere. 
• When a lameness gets worse with a block, it is a good sign you are working on the correct limb.
• Avoid over-blocking. Chasing subtle residual lameness may be unrewarding and difficult to attribute as clinically significant.

3. Lunging may complicate your evaluation

When horses are lunging, they are tilted toward the center of the circle, so asymmetry of head and pelvic movement is expected. There is much more horse to horse variability in lunging horses.   Additionally, horses, whether lame or not lame, lunge differently one direction compared to another.  As an example, in a 200+ horse sample of horses, 96% of horses that measured with no lameness on the straight line measured with hind limb lameness (i.e. asymmetric pelvic movement) at the lunge in soft ground. Not all of these horses had something wrong with them. It is also known that specific patterns of asymmetry are identified in normal horses, depending upon the surface characteristics.

Takeaways:

• Expect asymmetric head and or pelvic movement in lunging horses.
• Expect that horses may lunge differently each direction.
• Look for patterns of asymmetry that are uncommon for that surface characteristic.
• If lameness is measurable in the straight line, it is uncommon for the lunge to measure significantly different.

4. Evaluating horses under saddle – rider activity has an influence

As part of a National Science Foundation study, a sensor was added to the rider that detects sit from posting trot so that the effects of posting could be assessed. A few key findings were realized. First, sitting the trot had the least artefactual effect on lameness. When the rider was sitting the trot, compared to a baseline straight evaluation without a rider, only about 10% of the time was a significantly different lameness measured (different limb). Posting, on the other hand, had a predictable effect on pelvic movement in many horses. Approximately 80% of horses exhibited an increase in push off asymmetry on the inside hind limb when the rider was posting on the outside forelimb.

Takeaways:

• Sitting the trot has the least artefactual effect on lameness.
• Posting can create or exacerbate contralateral hind limb push off lameness.
• Posting can mask ipsilateral hind limb push off lameness.
• The ridden evaluation is likely most useful when lameness is not measured in the straight line without a rider, or when straight line results do not fit with clinical picture or owner/rider complaint or history.

5. Evaluating multiple limb lameness – patterns of asymmetry provide clues to the primary problem

There are three common causes of multiple limb lameness in the horse: 1) compensatory lameness, 2) secondary lameness, and 3) two (or more) primary lameness foci. Compensatory lameness is head or pelvic movement asymmetry created as the result of weight shifting off of a primary lame limb, giving the appearance of lameness in the other half (fore or hind) of the body. Compensatory lameness has often been described as the Law of Sides.  Ipsilateral lameness (e.g. RF, RH) is often the result of a primary hind limb lameness. This is true, happens often, and can be problematic for subjective evaluation, as the hind limb lameness can be quite mild and create a greater compensatory forelimb lameness.  Contralateral lameness is often the result of primary forelimb lameness. This is also true, however incomplete. It is now known that the timing of hind limb lameness matters in the determination of primary and compensatory lameness. For instance, primary forelimb lameness may cause a contralateral hind limb push off lameness. It can also cause a compensatory bilateral hind limb lameness – which is always a contralateral hind limb lack of push off and ipsilateral hind limb lack of impact. It is also known that there can be a special case of primary forelimb lameness that only shows an ipsilateral compensatory hind limb impact lameness.   As you can imagine, it is quite difficult to subjectively assess push off versus impact lameness. But by measuring, we can readily identify potential compensatory patterns, helping guide us to what is mostly likely the primary problem.

Clinicians at the University of Missouri have completed a study reviewing over 1200 cases that presented for lameness or poor performance. The results of this study will appear soon in The Journal of the American Veterinary Medical Association.  Of these cases, 56% exhibited a multiple limb lameness when trotting in a straight line. Dr. Keegan used the adage of Ockham’s Razer to approach the seemingly complex presentation of a multiple limb lameness: “all things being equal, the simplest solution tends to be the best one”.

Some interesting findings:

1. The highest definitive diagnostic rate was in horses that did not measure with lameness in the straight line. Over half of these horses were determined to be normal (not lame).
2. Forelimb lameness with contralateral hind limb push off lameness (according to the Law of Sides, a common compensatory pattern for a primary forelimb lameness), had a 70% definitive diagnostic rate, with 90% of cases localized to a forelimb lameness (77% to FL only, 13% to FL and HL).
3. Forelimb lameness with only a contralateral hind limb impact lameness (i.e. without a push off component, a pattern fitting either primary hind limb or primary forelimb lameness) had lower definitive diagnostic rate, but most were found to have a forelimb lameness (75%). Knowledge that most horses in this group were diagnosed with a forelimb lameness may improve success for future diagnoses of horses with this pattern. 
4. Of horses with an ipsilateral lameness when the hind limb lameness is push off type only (i.e. without an impact component, therefore fitting the ipsilateral component of the Law of Sides), 87% were definitively diagnosed with a hind limb lameness (66% hind limb only and 21% hind limb and forelimb).
5. An ipsilateral lameness when the hind lameness is pure impact type can be tricky. While it achieved a fairly high diagnostic rate, the primary lameness is just as likely primary forelimb as primary hind limb.
6. Lastly, forelimb lameness with contralateral impact lameness and ipsilateral push off  lameness (not a compensatory but a secondary lameness pattern) had a fairly even distribution of localization across the limbs and had the highest percentage (approximately 1 in 4) of “other”,   meaning not localized to the limb, but rather axial skeleton, metabolic, neurologic, etc.…

Takeaways:

• There is a logical decision process that can be applied to the 8 possible combinations of multiple limb lameness (considering timing of hind limb lameness as impact or push off type).
• Most pattern combinations are useful to guide us to the primary problem.
• A few patterns help us recognize that this case might be more difficult, or unusual.
• Only by measuring what the horse is doing can we recognize these patterns.

Big Data – where are we going?

Tens of thousands of horses have now been evaluated by veterinarians using inertial sensor measurement around the globe.   This data can be matched up with good clinical data and, using new data mining techniques, it is perhaps possible to find patterns useful to practicing veterinarians.  For example, it may be possible to provide differential diagnosis lists in order from most to least likely, that would be helpful to veterinarians planning how to proceed in a lameness investigation.