I really like this! Obviously you're not proving the direction of sampling bias, but given that this diagnosis requires an echocardiogram, it's very suggestive that only dogs on fancy grain-free diets are getting diagnosed.