Questions on statistical methods
Regarding the report by Rovel et al1 titled “Evaluation of standing low-field magnetic resonance imaging for diagnosis of advanced distal interphalangeal primary degenerative joint disease in horses: 12 cases (2010–2014),” I was disappointed to see what I believe to be clear critical flaws in the statistical methods. First, t tests were performed on data that were not collected from a random sample. Second, ordinal grade data were treated as continuous measurements.
Specifically, t tests should be reserved for data collected from random samples, whereas this study was a case series, representing a convenience sample of horses. In addition, t tests should only be used for quantitative data with a normal distribution, and ordinal data such as the subjective grades used in this study do not meet these criteria. Finally, summary statistics such as mean, SD, and SEM should not be calculated for ordinal data.
Robert L. Larson, DVM, PhD
Department of Clinical Sciences, College of Veterinary Medicine, Kansas State University, Manhattan, Kan
1. Rovel T, Audigié F, Coudry V, et al. Evaluation of standing low-field magnetic resonance imaging for diagnosis of advanced distal interphalangeal primary degenerative joint disease in horses: 12 cases (2010–2014). J Am Vet Med Assoc 2019;254:257–265.
Do you have the right kind of auto insurance for a house-call business?
I enjoyed the article “Homeward Bound” in the February 15 issue of JAVMA.1 Dr. de Jong's suggestion to keep a record of mileage and vehicle expenses so they can be written off as business expenses compelled me to respond. As a prior hospital owner, both I and veterinarians I employed had, in the past, used our personal cars to travel to clients' homes when performing in-home euthanasia. My manager had also used her personal car when performing hospital business. Imagine my shock and surprise when I recently learned that personal automobile insurance policies that provide property and liability coverage are void if the vehicle is used for business services. Damage to cars and, more critically, injuries to people will not be covered, leaving the employer exposed. If personal vehicles are used to provide business services, please talk to your insurance agent to ensure that the vehicles and their occupants are adequately covered by business automobile insurance coverage.
Wendy Hauser, DVM
Parker, Colo
1. Mattson K. Homeward bound. J Am Vet Med Assoc 2019;254:450–457.
Questions regarding prednisone treatment in a dog
The recent Pathology in Practice article1 concerning a 9-month-old Great Dane with presumptive autoimmune skin disease (epidermolysis bullosa acquisita) raised some questions in my mind. According to the article, after the diagnosis was made, the dog was treated with prednisone at a dosage of 4 mg/kg (1.8 mg/lb), PO, every 12 hours “for an extended period before transitioning to a tapering regimen.” Given that the dog's reported weight was 47 kg (103.4 lb), this would mean that it received 188 mg of prednisone twice daily. I am wondering whether the authors could indicate how long the dog received prednisone at this dosage, whether any adverse effects associated with prednisone administration were reported, and whether any follow-up diagnostic testing was performed.
Robert W. Reuther, DVM
Pittsburgh, Pa
1. Stiver SL, Fisher KR, Tobias JR, et al. Pathology in Practice. J Am Vet Med Assoc 2019;254:599–601.
The authors respond:
On behalf of the authors, I want to thank Dr. Reuther for his thoughtful response to our Pathology in Practice contribution.1 The prednisone dosage published in the article is incorrect; prednisone was initially prescribed at a dosage of 4 mg/kg (1.8 mg/lb), PO, every 24 hours, rather than every 12 hours as indicated in the article. In this case, the prednisone administration regimen was selected following discussions concerning case management and treatment with members of the Dermatology Service at the North Carolina State University Veterinary Hospital. Prednisone was administered at a dosage of 4 mg/kg/d for 21 days, after which time the dog's clinical response was evaluated and determined to be favorable. The prednisone dosage was then progressively decreased over the ensuing 12 weeks, at which time prednisone administration was discontinued. Adverse effects reported by the pet owner included polyuria, polydipsia, and polyphagia. Follow-up diagnostic testing was not performed. The patient was reevaluated multiple times during the initial 6-month treatment period and subsequently on an annual and as-needed basis. Currently, the owner reports that the dog is doing well, with no signs of disease recurrence or apparent adverse effects following the course of prednisone treatment.
Shane Stiver, DVM
Westbrook, Me
1. Stiver SL, Fisher KR, Tobias JR, et al. Pathology in Practice. J Am Vet Med Assoc 2019;254:599–601.
Current accreditation standards fail to address instructor quality
At its core, accreditation is intended to ensure that educational programs provide students a quality education. To accredit colleges of veterinary medicine, the AVMA Council on Education (COE) has developed a set of standards that it uses to evaluate veterinary medical degree programs and with which accredited colleges must comply. These standards cover a variety of facets, but, we contend, do not address one of the most critical components: instructor quality.
For > 40 years, education research has found that the single greatest influence on student learning and achievement is instructor quality.1,2 Yet, most veterinary medical training programs focus on curriculum and factors other than instructor quality.
Currently, COE standard 8 states that “Faculty numbers and qualifications must be sufficient to deliver the educational program and fulfill the mission of the college. Participation in scholarly activities is an important criterion in evaluating the faculty and the college. The college must provide evidence that it utilizes a well-defined and comprehensive program for the evaluation of professional growth, development, and scholarly activities of the faculty.”3 But, even though the standard refers to the qualifications of faculty members and their commitment to professional growth, it does not specifically address competence or growth in the realm of student instruction.
In much the same way that assessment drives learning, accreditation drives institutional behaviors and practices. Given the well-documented influence of instructor quality on student outcomes, it is imperative that accrediting bodies incorporate instructor quality into its standards. Without this requirement, institutions will likely be slow to change.
The COE's standard 6, which focuses on students, offers a potential framework for revising standard 8 to include reference to instructor quality. Standard 6 states, among other things, that the “college or parent institution must provide information and access to counselling services regarding financial aid, debt management, and career advising.”3 We believe that standard 8 should similarly state that the “college or parent institution must provide faculty with information on and access to training related to effective teaching, learning, and assessment.”
Colleges could then demonstrate compliance with this new standard by establishing formal faculty development programs, educator academies, or offices of faculty development. This could not only help energize faculty to improve instructional quality but also help colleges document and quantify the variety of professional development opportunities offered, track participation rates, measure faculty growth, and much more.
In conclusion, we implore the veterinary education community to rethink what matters most and revise accreditation standards accordingly.
Kenneth D. Royal, PhD, MSEd
Department of Clinical Sciences
Kate M. Bailey, DVM
Department of Molecular Biomedical Sciences, College of Veterinary Medicine, North Carolina State University, Raleigh, NC
1. National Council on Teaching and America's Future (NCTAF). What matters most: teaching for America's future. 2012. Available at: nctaf.org/wp-content/uploads/2012/01/WhatMattersMost.pdf. Accessed Jan 2, 2019.
2. Hecker K, Violato C. How much do differences in medical schools influence student performance? A longitudinal study employing hierarchical linear modeling. Teach Learn Med 2008;20:104–113.
3. AVMA. COE accreditation policies and procedures: requirements. Available at: www.avma.org/ProfessionalDevelopment/Education/Accreditation/Colleges/Pages/coe-pp-requirements-of-accredited-college.aspx. Accessed Jan 2, 2019.