The Issue of Joint Laxity and Stress Radiography

OFA does not normally respond to the various opinions expressed by individuals on Internet web sites and/or chat lines. Instead OFA maintains a web site (http://www.offa.org) to provide information that may be of value to breeders and veterinarians. However, a response to the opinions expressed by many people is prompted, as the opinions appear to have deteriorated to the point of becoming non-productive. OFA stated its position on any testing method, including PennHIP, that involved stress radiography to the breed clubs in 1994. This posting is a review of that position.

Contrary to some Internet postings, OFA, a not-for-profit organization, does support and encourage research on joint laxity and its meaning. The fact that joint laxity plays a role, but is not the only factor to be considered in development of hip dysplasia and its secondary changes of degenerative joint disease, has been recognized for over 30 years. This fact is not in dispute. The issue has been, and remains to be, the relationship of laxity that is demonstrated by forcing the heads of the femurs away from the acetabula by palpation or a fulcrum/stress device (i.e., a distraction device) to abnormal laxity (functional laxity that occurs in hip dysplasia.) Since 1972, when an independent panel of scientists classified the techniques for demonstration of joint laxity by use of an externally applied force as experimental, OFA has financially supported three research projects, recommended by external review, to answer the basic question. Dr. Belkoff, et.al. (VCOT 1: 31-36 1989) developed a device that measured the amount of force applied to the hips and noted that some dogs that demonstrated abnormal amounts of laxity were free of hip dysplasia at necropsy. These authors questioned the meaning of joint laxity as demonstrated by force. The other two projects supported by OFA are ongoing.

PennHIP is another technique for demonstration of forced (passive) laxity that is also attempting to answer the basic question of the relationship of passive laxity to functional laxity. OFA encourages their research efforts; however, OFA takes exception to the marketing techniques and claims used to promote the PennHIP testing method for clinical use, as the use of this method appears to be premature. In other words, commercialization (marketing) of the method has outreached the science.

OFA feels that general use of PennHIP as a mass screening test method for hip dysplasia is premature because:

1. The primary basis for marketing PennHIP was reported by Dr. Smith, et.al. (Am J Vet Res, July 1993) using a modification of a previously described positioning, stress/fulcrum technique. The study was a survey type involving 142 dogs (105 of which were German Shepherd Dogs). The results of the study were questioned by Dr. Susan Shott of the Biostatistical Unit, Rusk Cancer Institute (Am J Vet Res, December 1993) who challenged the analysis of the data and stated: "The data does not support the author's conclusion that the DI was the most important and reliable phenotypic factor for determining susceptibility of hips to degenerative joint disease ... and that this determination could be made with an acceptable degree of accuracy as early as 4 months of age."

2. Dr. Lust, et.al. (Dr. Smith was a coauthor) in a report involving 42 Labrador Retrievers (Am J Vet Res, December 1993) concluded that a DI of <0.4 at 4 months of age correctly predicted normal hips in 88% of the cases and a DI of >0.4 correctly predicted hip dysplasia in 57% of the cases. The authors further concluded that: "Distraction indices between 0.4 and 0.7 and at either 4 or 8 months of age were not associated strongly enough with evidence of disease to be clinically reliable in predicting, on an individual basis, the outcome for dysplastic hip conformation when the dogs were older."

3. No breeding data based on PennHIP has been reported. Dr. E. A. Leighton (JAVMA, May 13, 1997) reported on genetic progress in improving the hip quality in German Shepherd Dogs and Labrador Retrievers in the Seeing Eye closed colony of dogs. In less than 5 generations the percentage of hip dysplasia was decreased from 55 to 24% in the German Shepherd Dogs and from 30 to 10% in the Labrador Retrievers using the hip extended position and a modified OFA evaluation procedure. PennHIP DI measurements were also made but the mean DI across generations did not change. It should be pointed out that DI was considered experimental and breeding selection criteria did not include the DI. It will be interesting to see the results when DI is included as a selection criterion.

With the above reservations, plus experience with the issue of joint laxity, OFA would be remiss in its responsibility to either endorse or reject the PennHIP testing method. In other words, the jury is still out! This leaves the breeder in a dilemma as to which testing method to use, OFA or PennHIP or both, as they are entirely different test methods for the same disease.

There is a great economic advantage to breeders for determination of the hip status at a young age and to assess the risk for development of hip dysplasia at a later age. OFA reported (Vet Clinics of No Am, May 1992) on a study of 3,369 dogs from 25 breeds. Reliability of the preliminary evaluations ranged from 71.4% in the Chesapeake Bay Retriever to 100% in the Welsh Springer Spaniel. The preliminary evaluation appeared to be breed dependent and dependent on the evaluator's experience with the skeletal development of that breed at the age of evaluation.

When faced with the problem of comparing entirely different test methods for determining dysplasia, scientists evaluate the results of reported values for false negative (probability of diagnosing a dysplastic dog as normal), false positive (probability of diagnosing a normal dog as dysplastic), specificity (probability of a normal dog receiving a normal evaluation), and sensitivity (probability of a dysplastic dog receiving a dysplastic evaluation). These values for OFA preliminary evaluations by age and hip ratings, in a different population of dogs than previously reported (Vet Clinics of No Am., May 1992) have been reported (JAVMA, November 1, 1997). The false negative and false positive values for PennHIP were reported by Dr. Smith et.al. (Am J Vet Res, July 1993). No report of selectivity or sensitivity values for PennHIP were given. There were 2,332 dogs in this OFA study and 142 dogs in the PennHIP study. The limited number of dogs resulted in a larger confidence interval for the PennHIP values. Confidence intervals (CI) are determined so that one can be 95% confident that the true value lies within the calculated range. The false negative values for OFA evaluations were 8.9% (CI=5.9 to 12.9%) at 3-6 months, 6.0% (CI=4.4 to 8.0%) at 7-12 months and 3.8% (CI=2.6 to 5.4%) at 13-18 months of age. The false negative values for PennHIP evaluations were 12% (CI=1.5 to 38.3%) at 4 months and 0% (CI=0.0 to 15.4%) at 12 months of age. It appears that the probability of retaining a dysplastic dog in the breeding pool is the same for either test method.

However, the false positive values for PennHIP were significantly greater (48% at 4 months, 57% at 6 months and 38% at 12 months) than those for OFA evaluations 17.6% at 3-6 months (CI 10.8 to 26.4%), 10.0% at 7-12 months (CI 5.7 to 15.9%) and 8.5% at 13-18 months (CI 4.8 to 13.6%). It appears that the probability for removing a normal dog from the breeding pool is less with the OFA evaluations.

Dr. Adams, et.al. (JAAHA, 1998, 34: 339-47) reported (using palpation, OFA method, PennHIP, and Norberg angle measurements) on results of a study of hip laxity, in 32 dogs from 4 breeds (12 Greyhounds, 4 Labrador Retrievers, 12 Irish Setters, and 4 hound-mix) at 6-10 weeks and 16 to 18 weeks that were compared to detection of degenerative joint disease at 52 weeks of age. Five hips with evidence of subluxation but no evidence of degenerative joint disease on the OFA type evaluation of the hip extended view were eliminated from analysis. The authors concluded that DI and Norberg Angle measurements at 6-10 and 16-18 weeks were the most reliable predictors of hip dysplasia, at 52 weeks of age, with DI being more reliable than Norberg. The OFA and palpation methods at 6-10 or 16-18 weeks were not reliable predictors. This is not surprising as reliability of OFA preliminary evaluations has been shown to increase with age of evaluation. The OFA report (JAVMA, Nov. 1997) included 380 dogs evaluated at 3 to 6 months of age. The reliability was 89.6% (CI=85.4 to 92.9%) for normal evaluations and 80.4% (CI=71.4 to 87.6%) for dysplastic evaluations. The mean age was 4.8 months (19.2 weeks) and the median age was 5 months (20 weeks) which means that over half of the dogs in the OFA study were older than in the study reported by Dr. Adams.

OFA data and PennHIP data are not representative of the general population of dogs because the programs are voluntary, most dogs are in pet homes and are not radiographed, and not all radiographs of dogs radiographed are submitted for evaluation by either program. For example; if an attending veterinarian determines a dog to be dysplastic, by either method, the radiograph(s) may not be submitted to save the owner money. PennHIP collaborators may take the hip extended view first and if the radiograph shows evidence of dysplasia the DI views may not be taken or the owner may not allow submission of an obviously large DI measurement.

Breeders are aware of the economic value of early screening of dogs for determination of the hip status. They should also be aware that both OFA and PennHIP use the radiographic evaluation of the same hip extended projection as the standard for comparing with the results of the early evaluations. The OFA standard represents the consensus of 3 independent evaluations at >24 months of age by board certified veterinary radiologists. It is not clear who evaluates a radiograph submitted for PennHIP determination, but the original study reported the standard to be Dr. Smith's evaluation. This evaluation at >24 months of age has approximately 5% false negative finding as reported by Dr. Jessen (Proceedings of a 1972 symposium on hip dysplasia) and by an internal OFA study of dogs evaluated at 24 months that were re-evaluated at an older age. This is why OFA requires the 24 month certification age. Voluntary submissions to PennHIP will provide information on the range, mean and median of the DI measurements for the various breeds. The meaning of the measurements remains unclear and will require repeat studies, on the same dogs, at >24 months of age.

Breeders must be aware of the cost, strengths, and weaknesses of the test methods available for evaluation of the hip status before making the choice of a specific testing method. Once the choice is made, it must be followed for generations before progress in improving the hip status can be evaluated. OFA data has demonstrated marked improvement of the hip status in the Portuguese Water Dog (AKC Gazette, Nov 1991) and the Chinese Shar Pei (Barker, Mar/Apr 1995). OFA data on all breeds was independently evaluated and reported by Dr. Kaneene (JAVMA, Dec 1997) an epidemiologists from the Population Medicine Center at Michigan State University. The study compared OFA evaluations on dogs born between 1972 and 1980 with dogs born between 1989 and 1992. The population consisted of 270,978 dogs. The authors, having acknowledged the fact that submissions are voluntary and that there is bias due to prior screening, concluded:

We do not believe that this is the most likely explanation, because the increase in the percentage of dogs classified as having excellent hip joint phenotype (+36% [7.82 vs. 10.64%]) was substantially larger than the decrease in the percentage of dogs classified as having canine hip dysplasia (-21.% [17.39 vs. 13.82%]). If better screening of radiographs prior to submission to the OFA was the cause of the increase in percentage of dogs classified as having an excellent hip joint phenotype, then because it is easier to differentiate dysplastic hips from hips with normal phenotypes than it is to differentiate hips with excellent, good and fair phenotypes, we would have expected that the decrease in percentage of dogs classified as having canine hip dysplasia would have been larger than the increase in percentage of dogs classified as having an excellent hip joint phenotype.

Unfortunately, PennHIP has not been available long enough to accumulate the data necessary to evaluate the effect of this test method over time.