This certainly matches our experience with the diabetic alert dogs in our community. Lovely, well mannered dogs, but they don’t replace the need for a meter or a sensor.
One of my son’s teachers had an alert dog. She said that the dog was rewarded for alerting her, so of course the dog eventually figured out how easy it was to get a reward!
That seems to fit from the article:
They also had false positives. Only 12% of the dogs’ alerts happened during actual low blood sugar events.
Our friends dog was really helpful for a neighbor who had un-diagnosed Type 2 diabetes and was running high blood sugar, and the dog wouldn’t get off her feet (his alert). Led her to go to the doctor and get diagnosed earlier than normal. I think the dogs are better are sussing out high blood sugar.
That’s interesting. I didn’t know they were trained for that. I always thought the emphasis was on low BG.
Not our friends dog. He had a quite unmanageable A1c, and his primary alerts were for the highs. Because he could take the dog to school, he was able to get his A1c from the 10’s to the 8 range. Worked well for him during the day for a few years. After the dog got older he wasn’t as reliable, especially at night.
There is little to no un-biased scientific evidence that dogs are good at detecting abnormal blood sugar levels. The studies that have been performed were paid for or performed by companies that sell service dogs.
I think the CGM capability is the better way to go for us diabetes.
The one positive note about service dogs is that apparently the owners do monitor their diabetes more pro-actively when they have a dog than without. Also, the owners have a more general overall positive outlook.
Obviously these are important attributes.
Nevertheless, these dogs should not be marketed as a monitor unless it is true regardless of the other benefits derived.