Lameness Scores

Lameness scores are a vital clinical outcome assessment tool. The walk and trot are the gaits most commonly evaluated because the speed of the limbs during movement is easier to assess without specialized equipment, and these gaits are symmetric, making identification of a lame limb easier. The reader is referred to the chapter on orthopedic and neurologic evaluation for information regarding identification of lameness. It is important to separate the walk from the trot when scoring lameness. Dogs are generally less lame at a walk because less force is placed on the limb at this less strenuous gait. Trotting may accentuate a lameness that is mild to moderate at a walk because of the greater forces placed on the limb with increased velocity. Having separate lameness scores for the walk and trot allows finer discrimination of gait analysis and may be more sensitive for detecting subtle improvements during rehabilitation (Box 13-1).

Although lameness scales are simple, commonly used, and inexpensive, they may not be accurate in evaluating subtle or mild lameness. One study compared a numerical scale and visual analog scale (VAS) with ground reaction forces and concluded that subjective scoring scales do not replace force plate gait analysis in dogs with surgically induced lameness.⁶ Agreement of experienced evaluators with ground reaction forces was low unless lameness was severe. In addition each observer had his or her own individually unique scale. Therefore observers must stay the same during the evaluation of a patient for accurate analyses. Another study of induced hind limb lameness of dogs found that subjective evaluation of lameness varied greatly between different observers, and there was poor agreement with objective ground reaction force measurements.⁷ Similar results regarding estimating the severity of lameness using a VAS have been found in dogs with fragmented coronoid processes.⁸ However, owners may not be able to assign a valid VAS score.⁹ Therefore, this form of evaluation is better performed by people experienced in evaluating lameness and pain.

An alternative approach may be to use a questionnaire to assess pain and lameness in dogs. One study evaluated a variety of questions using a VAS and found significant correlation with peak vertical force and vertical impulse.¹⁰ Responses to questions recorded on a standard 10-cm VAS that had significant correlation with ground reaction forces included:

The Helsinki Chronic Pain Index also found mood, play, locomotion while walking and trotting, getting up, difficulty moving after rest, and difficulty moving after major activity to be valuable in helping to identify dysfunction with osteoarthritis.¹¹ Vocalization, galloping, jumping, and laying down are other activities that help identify dogs with osteoarthritis.

A Canine Brief Pain Index (CBPI) has also been developed to evaluate chronic pain in dogs with osteoarthritis, and consists of 11 questions related to pain, function, and overall impression of quality of life; a 10-point scale is used to indicate the severity of the condition for each of the 11 questions.¹² While the CBPI is able to detect improvement in pain scores of dogs with osteoarthritis after treatment with an NSAID,¹³ there is not significant correlation of CBPI scores with ground reaction forces as measured with a force platform, suggesting that owners evaluate things other than lameness in assessing improvement with treatment.¹⁴

In summary, although lameness scores are commonly used and clinically useful in some circumstances, care must be used in interpreting the scores. In general, agreement among individuals making assessments is greatest in dogs with severe lameness and no lameness. There is generally poor agreement in mild cases of lameness, and there is poor correlation of lameness scores with objective measurement of ground reaction forces using a force platform. VAS scores may offer some advantages, but experience in lameness evaluation is important in assigning values. Finally, various pain indexes may be useful in assessing treatment effects in dogs with osteoarthritis. However, there is relatively poor correlation with ground reaction forces, likely because a variety of functional items other than just lameness are assessed by owners when using indexes.

Kinetic (Force Plate) Analysis of Gait

Kinetic evaluation of gait involves the measurement of ground reaction forces with a force plate or platform (Figure 13-2). It is an objective, repeatable measure of weight bearing on limbs when proper technique is used for data collection. Lameness can be compared over a period of time without relying on memory of previous assessments because data may be stored on a computer. Although measurement of ground reaction forces is a reliable and well-accepted method of determining the degree of weight bearing on the limbs, it is an artificial situation, and some dogs may display different clinical signs in a home environment. Force platform systems are available at many veterinary colleges and some private practices. It is important that appropriate software be used for quadruped animals.

Force plates use either strain gauges or piezoelectric crystals. The force plate is either mounted on a platform or embedded in the floor so that it is even with the surface of the floor. A runway of adequate length is essential. Most systems have timer lights that are triggered as the handler and dog approach and cross the force plate to allow the calculation of mean velocity and acceleration. Control of velocity and acceleration within appropriate parameters is essential for repeatable data collection, because these greatly affect the force placed on each limb. The force plate is connected to a computer that calculates ground reaction forces. Force plates may also be embedded in treadmills so that consecutive footfalls can be recorded, with the gait being very rhythmic. However, some systems may not be able to accurately evaluate braking and propulsion forces.

Pressure pads or mats are also used for evaluation of canine gait. Many commercially available systems have an adequate number of sensors to detect ground reaction forces with reasonable accuracy. One advantage is that they can measure data from several consecutive footfalls. Another advantage is the ability to determine stride length, cadence, and average velocity.

The most useful forces measured are the peak vertical force (Z_Peak) and vertical impulse (Z_Impulse) (Figure 13-3). Other forces that may be useful are the peak braking (Y_{A Peak}) and propulsion (Y_{B Peak}) forces and braking (Y_{A Impulse}) and propulsion (Y_{B Impulse}) impulses (Figure 13-4). Medial-lateral forces (X_Peak) and impulse (X_Impulse) are likely too small and variable to be clinically useful. Forces may be measured during stance, walking, or trotting.

It is essential that appropriate technique be used during data collection. In addition to consistent velocity and acceleration targets, an experienced handler is necessary to minimize intertrial variability, although there is little or no difference in measured forces between experienced handlers. The handler should be between the dog’s head and shoulder, and the dog should be gaited without undue tension or pulling on the leash. The dog should not alter the gait, throw its head, turn the head, lunge, or make other sudden movements as it approaches and crosses the force plate or pressure mat. The head carriage should be in a neutral position to avoid shifting weight to the front or rear limbs.

In general, dogs bear approximately 30% of their body weight on each front limb and 20% on each rear limb while in a standing position with the limbs placed squarely under the body. Walking at a velocity of 0.7 to 1.0 m/sec results in forces equivalent to approximately 60% of body weight on each forelimb and 40% on each rear limb in a medium to large dog. Increasing the velocity to a trot of 1.7 to 2.0 m/sec results in weight bearing of 100% to 120% of body weight on each forelimb and 65% to 70% of body weight on each hindlimb in a similar size dog. Reduced weight bearing on an individual limb may result in mild weight shifts to the other limbs.

Kinematic (Motion) Analysis of Gait

Kinematic or motion analysis of gait is a powerful tool that can be used to measure flexion and extension angles of joints during gait, stride length, and other parameters of stride. It is usually combined with kinetic gait analysis. There is limited availability of three-dimensional kinematic gait analysis because the necessary equipment is sophisticated and expensive. Two-dimensional systems are less expensive, but they do not capture rotation and circumduction of limbs during gait, although simple flexion and extension of joints may be captured.

Attention to detail is critical for kinematic gait analysis. Most systems use a number of reflective devices that are attached to the dogs at very specific anatomic points to allow repeatable data collection at different times (Figure 13-5). Motion of the markers in relation to the joints is determined with a series of cameras interfaced with a computer. Software for quadruped animals allows reconstruction of a walking “stick figure” on the computer screen. The software is then used to calculate a number of measurements, including flexion and extension angles of joints during gait, angular velocity and acceleration of joints, and stride length and frequency.

Noninvasive, computer-assisted, three-dimensional kinematic gait analysis has been used to study lameness associated with cranial cruciate ligament rupture in dogs.¹⁵ Dynamic flexion and extension angles and angular velocities were calculated for the hip, stifle, and hock joints. Distance and temporal variables were also determined. Mean flexion and extension curves were developed for all joints, and the changes in movement that occurred over time after cruciate rupture were compared. Each joint had a characteristic pattern of flexion and extension movement that changed with cruciate rupture. The stifle joint angle was more flexed throughout stance and early swing phase of stride and failed to extend in late stance. Angular velocity of the stifle joint was damped throughout stance phase, with extension velocity almost negligible. The hip and hock joint angles, in contrast with the stifle joint angle, were extended more during stance phase. Stride length and frequency also varied significantly after cruciate rupture. A change in the pattern of joint movement appeared to occur in which the hip and hock joints compensated for the dysfunction of the stifle joint.

Knowledge of kinematic parameters may allow more effective rehabilitation of patients following surgery or those with chronic conditions. In particular, changes that are noted during the course of treatment may allow for stepping up the level of activity, or more important, may signal changes that occur in other joints as a result of fatigue or overuse. Knowledge of these data may help prevent injuries to other joints or limbs. Recently, studies have been conducted to evaluate joint motion during therapeutic exercises.16–21 The kinematics of various therapeutic exercises are discussed in the biomechanics of rehabilitation chapter.

Kinetic and kinematic gait analysis techniques have been combined using an inverse dynamics technique to estimate net joint moments and power. This technique has been used to evaluate tibial plateau leveling osteotomy and a lateral fabella-tibial technique to stabilize stifle joints in a cranial cruciate ligament transection model,²² Labrador retrievers with and without cranial cruciate ligament disease,²³ and forelimb joints.²⁴ This form of gait evaluation is time-consuming and requires specialized software, but it can provide valuable information.

Pedometers

Most outcome assessment measurements are performed in a clinic in which equipment and personnel are located. However, in many cases, the animal temporarily becomes more active in the clinical environment as a result of traveling, exposure to other animals, and being in a strange environment. In these situations the dog often appears to be better than has been reported in the home environment. A pedometer may give some indication of a dog’s activity level in its home environment. Most pedometers have a pendulum that swings back and forth with the normal gait cycle and indicates the number of steps taken. In our experience pedometers may be up to 85% to 90% accurate with regard to counting the number of steps taken by a dog. One study confirmed that pedometers are reasonably accurate, with the exceptions that at a walk, pedometers overestimated the number of steps by approximately 17% in large and medium dogs and underestimated actual number of steps by approximately 7% in small dogs.²⁵ Although these researchers found that no significant differences between pedometer-recorded and actual number of steps were detected when dogs trotted or ran, our experience has been that the step count is artificially increased when dogs run or climb stairs.

Proper placement of the pedometer is crucial. We have found that placing the pedometer just proximal to the elbow most accurately indicates the number of steps taken (Figure 13-6). In the study by Chan et al., however, pedometers were worn around the dog’s neck by means of an adjustable lightweight chain.²⁵ However, we have found that when dogs shake their head or scratch their neck, the step count on the pedometer is artificially increased. Because pedometers may not be not accurate in all dogs, it is recommended that the dog be walked 100 steps; the number of steps indicated by the pedometer is compared to see if the pedometer is reasonably accurate to record a dog’s activity level. An additional test to determine accuracy is to place the pedometer and then allow the dog to have free activity while 100 steps are counted. Because the pedometer may double-count when the dog is trotting, running, or going up and down stairs, it is used only as a semiquantitative indicator of home activity. Environmental conditions are another major factor in the activity of dogs. Dogs that spend some time outdoors may not be as active if it is raining or cold, or the weather is otherwise inclement. Therefore comparative assessments should be made on days when the weather is as similar as possible.

Recently, accelerometers have been used to estimate activity in dogs. Because of the differences in technology between pedometers and accelerometers, placement at various sites on dogs can give acceptable results. One study found that placement on a collar in the ventral neck region gives acceptable results for home activity monitoring.²⁶ In addition, accelerometers were able to distinguish the intensity of activity in dogs.²⁷ Another group found that the use of accelerometers was able to detect an increase of 20% in home activity after treatment for osteoarthritis, while placebo-treated animals had no increase in activity.²⁸ Activity and location may also be monitored relatively inexpensively using GPS technology for pets.²⁹

Joint Laxity

Assessment of joint laxity in dogs is generally limited to qualitative evaluation, although some quantitative evaluations have been performed. In general joint laxity may occur as a result of developmental conditions (e.g., hip dysplasia), trauma, or pathologic degeneration of ligaments such as the cranial cruciate ligament.

Damage to collateral supporting structures is most easily assessed by placing varus and valgus stresses on the affected joint. The joints most commonly affected with collateral ligament damage include the hock, stifle, carpus, elbow, and shoulder. The digits are occasionally involved. Assessment is most accurate when the joint is placed in full extension to prevent inadvertent internal or external rotation of the joint that may be mistaken for varus or valgus movement. Breed and age differences exist regarding normal varus and valgus motion, especially in young animals, so comparison with the contralateral normal limb is helpful.

Hip joint laxity is most commonly assessed in young, growing dogs as an indication of early hip dysplasia. Subjective assessment is commonly performed using the Ortolani maneuver to create subluxation of the hip joint. A more quantitative method of assessing hip joint laxity is the use of PennHIP, in which controlled pressure is applied to the hip joint to create subluxation. A radiograph is made with the coxofemoral joint in this position, and measurements are made to quantify the degree of laxity, known as the hip distraction index. The hip distraction index has correlation with the development of hip dysplasia in some breeds of dogs.

Stifle joint laxity is clinically assessed by palpating for cranial drawer motion. There should be no drawer motion in normal dogs, with the exception of puppies, which have some drawer motion that comes to an abrupt stop at the end of drawer. Methods to quantify drawer motion, which are not commonly used in dogs, include measurement of direct cranial drawer, quantitative cranial tibial thrust, and use of instrumented devices.³² Direct cranial drawer is determined by simply marking the position of the tibial tuberosity with the stifle in reduction on a piece of graph paper, and then remarking the position of the tibial tuberosity with full cranial drawer. Caution must be used to prevent motion of the femur and to keep the marking device perpendicular to the graph paper. Quantitative cranial tibial thrust is measured in a similar fashion, except that the stifle is kept in a fixed position (usually standing angle) while the hock is flexed. This results in cranial displacement of the tibial tuberosity if the cranial cruciate ligament is ruptured. Marks are made on graph paper with the tibia in a reduced position and with the tibia in full cranial tibial thrust. Care is taken to keep the caudal aspect of the femur in contact with a solid surface so the only motion that occurs is the cranial displacement of the tibia.

Instrumented devices for knee-drawer tests have been used in humans to measure shifts of the tibial tuberosity relative to the patella.33–35 The total anterior and posterior displacement produced by anterior and posterior loads of 20 lb is measured from a reference position. Studies have evaluated the effects of variables on the accuracy and reproducibility of anterior-posterior drawer measurements. Reproducibility is principally affected by deviations in subject positioning. In addition to drawer motion, inadvertent knee flexion and tibial rotation may occur and result in incorrect measurements, despite attempts to prevent these. The effects of different observers, time sequences, different days, and muscle relaxation have been studied and contribute to error rates of 5% to 15%. Stress radiography, a relatively sensitive method of measuring drawer motion, has been used to compare instrumented systems in humans with tears of the anterior cruciate ligament (ACL). In one study stress radiography was superior to instrumented arthrometer testing for determining cruciate ligament status.³⁶ Similar studies have not been performed in dogs, but instrumented knee-drawer devices are difficult to use in dogs and require practice to obtain reproducible results. Pediatric devices may be easier to use because the smaller size may allow a better fit of the instrument to the dog’s limb.

Muscle Mass

Muscle mass measurements are a useful outcome measure in veterinary rehabilitation. Muscle mass indicates limb use and is associated with muscle strength. Muscle mass may be estimated with limb circumference measurements, ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and dual-energy x-ray absorptiometry (DEXA).

Measurement of limb circumference is an indirect method of assessing changes in muscle mass and is inexpensive, quick, and easily performed on clinical patients. Acceptable results depend on using standard, repeatable methods of measuring limb circumference. A measuring tape with a spring tension device is useful in measuring limb circumference to improve consistent tension on the tape when making measurements (Figure 13-8).

One study evaluated the effect of limb position, clipping hair, sedation, and different evaluators on thigh circumference measurements at two different locations before and after stifle stabilization surgery.³⁷ Thigh length was determined by measuring from the tip of the greater trochanter to the distal aspect of the lateral fabella. Circumference was determined at points equal to 50% and 70% of thigh length, measuring from the greater trochanter distally. Measurements were technically easier when made at the 70% location because the skin of the flank did not impede measurements (Figure 13-9). Clipping had little effect on thigh circumference measurements, but dogs with short hair coats were used (Figure 13-10). Measurements may be affected to a greater extent in dogs with long hair.

Limb position is important. Full flexion of the stifle results in greater thigh circumference measurements as compared with measurements made with the limb placed at a functional standing angle or with the stifle fully extended (Figure 13-11). Extension of the joint causes the muscle bellies to elongate, whereas flexion results in shortening and bunching of muscle bellies, creating greater thigh circumference. Measurements made in awake dogs with the stifle extended are not significantly different from those made following heavy sedation, as long as the dogs are not overly tense, probably because the muscles are in an elongated position when the stifle is extended (Figure 13-12). Thigh circumference measurements are sensitive enough to detect muscle atrophy 2 weeks after stifle surgery (Figure 13-13). Similar measurements may be obtained by independent evaluators as long as standard technique is used and the evaluators practice before obtaining actual measurements.

The association of limb circumference with actual muscle mass is important if it is to be useful in the evaluation of rehabilitation patients. Human male cadavers were subjected to comprehensive anthropometry and weighing of all skeletal muscle.³⁸ Limb circumference had good correlation with total skeletal muscle mass for the forearm (r = .96), mid-thigh (r = .94), calf (r = .84), and midarm (r = .82).

Thigh circumference also has significant correlation with actual muscle mass in dogs (Figure 13-14). Even greater correlation may be obtained when thigh length is also considered and the volume of a cylinder is calculated. Although this method is simplistic because the thigh is not a true cylinder, the volume of a cylinder is easy to calculate and this estimate of muscle mass may be more clinically useful. Similar principles likely exist for circumference measurements of other limbs. Currently we prefer performing thigh circumference measurements with the hair short or clipped, the limb held in an extended position, at 70% of the length of the thigh, and with the animal relaxed, but not necessarily sedated. In the forelimb the length is measured from the proximal aspect of the olecranon to the distal aspect of the lateral styloid process; the limb circumference is measured at a point that is 20% of the limb length as measured from the proximal aspect of the lateral epicondyle of the humerus.