We would like to thank Dr. Valeika for pointing out some serious weaknesses in population studies done in veterinary medicine. As explained in our introduction, in our clinical work we have identified cats with conditions anecdotally thought to increase the risk for thrombosis. When some of these cats had renal infarcts identified by sonography, we wondered if there might be a connection. Based on this observation, we were interested in identifying clinically relevant concurrent conditions, to determine if we should be recommending further workup. The study was not intended to be a comprehensive examination of all clinical conditions.
While occult hypertrophic cardiomyopathy was not expressly defined in our writing, Dr. Valeika points out that it is a term which is inherently understood in the veterinary community, requiring clinical signs of the disease to be absent.
To clarify, controls were subjected to the same exclusion criteria as cases.
Dr. Valeika does an excellent job reviewing the essential aspects of epidemiology that need to be considered when conducting research, as well as when reviewing the scientific literature. Just as a point of clarification, age distributions for cases and controls were not compared with a statistical test as suggested. The raw numbers and distributions were visually examined and compared, but as all authors know, it simply isn't feasible to present all of the relevant raw data in publications. Age was found to be an effect modifier, therefore, we reported age‐specific effect measures where possible. Though we agree that transparency is the goal, perhaps the discussion should focus on space limitations, both in word count and in figures and tables, and editorial support to include more raw data. Or perhaps veterinary journals should considered more appendices that can be accessed electronically for those wishing to explore the data more thoroughly. [Editor's note – the Journal encourages publication on‐line of supplementary material when indicated.]
We also welcome the lively discussion of p values and confidence intervals and the call for better consistency within the veterinary literature. Perhaps this could be the beginning of discussions with the journal editors about reporting guidelines and an increased emphasis on confidence intervals. However, in order for such a change to be made, reviewers, as well as editors need to be on board. Often the final version of a manuscript reflects responses to specific editorial comments from reviewers.
We hope that Dr. Valeika will continue to critically review all journal articles – his enthusiasm for statistics and epidemiology is remarkable.