We adhere first to the Veterinarians Oath https://www.avma.org/KB/Policies/Pages/veterinarians-oath.aspx and the veterinarians client-patient relationship https://www.avma.org/KB/Resources/Reference/Pages/VCPR.aspx. These are part of our assessment of a case and documented on our submission form. Because, in the consulting/research business if it isn’t recorded and validated then it didn’t happen!
An opinion is based on education and experience. Our opinion continually evolves by our research and the research published by others. And then very carefully analyzing what we can apply for each case. Our evaluations are a lengthy process resulting in information that is relayed back to veterinarians, one-on-one.
Initially, our opinion is based on evaluation of a blood or cerebrospinal fluid sample. There is a lengthy process of evaluation for each sample we receive. We review and validate the data (and data entry) that we get from the veterinarian, a minimum of 5 reviews , at least one documented by the QA officer, and this is before the samples are tested. Each sample is tested against known values and the results of testing are documented. The results are first captured by computer, these data are evaluated, recorded in the consultation record, and then interpreted. The validation process for each result involves another five reviews starting with the equipment measuring the reactions to the last step, result interpretation. The interpretation draws information from the case history that was recorded on the submission form by the veterinarian and a review of the prior data. Did the gait score change? Was the horse treated and with what? Is there other information that is useful?
At this point, we have a basis for the consult. The consulting report is sent to the veterinarian. If comments are needed-- something is obvious to us but may not be to the busy veterinarian, we send an email. We use the data that was entered for each parameter into many algorithms that are programed into our computer. These are the algorithms that spit out reports for field studies, FDA forms, or new research. In house, we tag some samples for more testing or more analysis. When the veterinarian calls we quickly call up the complete consultation record on the animal and discuss our view of the case and listen to what questions they may have. This may result in more testing, sending peer reviewed literature, or sending summary tables of data under review by our statistician. We may have more discussion after the information is reviewed by the attending veterinarian.
Another review of the data for each horse is made by Dr. Benedetti. She selects cases that are more complex and would benefit from more testing, or at least a more detailed discussion with the clinician. At each point notes are made to the file if there are any significant values for the data that we track. One horse may change our algorithm, that’s how new ideas happen.
We find interesting things. For example, unexpected results prompt new queries of our data base. We build in more parameters into our new query as appropriate. That is the power of a data base. Here is a recent case, a 20 year old untreated horse presented acutely with an abnormal gait. The veterinarian requested testing for auto-reactive antibodies against myelin and a CRP. The anti-myelin protein antibodies were present but the CRP was 0 micrograms per ml of serum. That’s unexpected! Generally we expect untreated horses with MP2/MPP antibodies to have elevated CRP levels.
Dr. Benedetti queried 19,161 records and found 487 records that were similarly tested for a minimum of MP2/MPP. Of those, 403 had detected MP2/MPP antibodies and an elevated CRP while 26 had a CRP of 0. Our result is unexpected, only 5% of samples we tested are similar! Other parameters that we captured from submission forms are age and treatment history, the start of our understanding of this case. The group of samples with this result (antimyelin antibody positive and tested for CRP, CRP=0) come from horses that are teenagers, and 12 of them had no history treatment. We can do more evaluations for region of the country, time of year, or how the samples were shipped. We may take the 12 horse data set and design a questionnaire that would go to many veterinarians, targeted to a specific case to get more data for our query. The case may be one they saw years ago, we would be looking for outcome, or a recent submission that would spur additional testing. Based on the results of the investigation we evolve our algorithms. We expect veterinarians to call us and ask and discuss the case. More discussion will most likely lead us to the correct interpretation or lead us to another investigation.
If we find it appropriate, we may design a survey. For horse owners we may make the survey a quiz. Our consultation program is designed to analyze a unique case against a large data base giving veterinarians the wisdom of the crowd.