[by]

From personal experience, this is absolutely correct and still a major problem which needs solving. The nurses in hospital will keep a fluids-in fluids-out chart constantly logging their best estimates. Outside hospital this will not happen. Lack of fluids can be fatal. I would not be surprised if this was proven to be a very common cause of fatality for people with dementia. Even in hospital fluids-in fluids-out may not be recorded accurately - the measurements require more human effort and continuity of care than is likely to be available. (It is always best for people to be enabled to do things, for example drinking and eating, for themselves if possible, and this can take a long time.)

My guess is that fluids intake could be tackled to some extent by a measured drinking device. The difficulties are spillage being counted as consumption, simple and safe operation for people with lower cognitive, physical and sensory abilities, the need for hospital-level cleanliness of the device, and of course cost. I do not know why we do not have devices like this when people can die from lack of fluids.

A less obvious way technical people can help is to test the accessibility of their products. A device with a black button on a black background, or which in some other way obscures its controls and functions, is a failure in this respect. Make all the functions visible, easy to understand and easy to operate. Test websites with automated tools like https://try.powermapper.com/Demo/SortSite