At NowPatient Diabetes, we believe that everything we put into the app should be safe and well researched. However, we also believe that we should do as much as we can to help people now. We don't think we should spend 20 years researching behind closed doors as nobody sees any benefit of what we are doing for 20 years. This sometimes means we have to put features into the app that we know could be better but will give a lot of benefit to people in the meantime.
The NowPatient Diabetes graph is one of those features. What we have in the app now is well researched and providing benefits to some people using CGMs. And we know how good it will be in the near future. But at present, there are some limitations. These are as follows:
No Previous Records
If you haven't recorded the food, correction dose or hypo treatment before in NowPatient Diabetes then the record will show as not influencing the graph. This is because NowPatient Diabetes is personalised to you and so requires your previous actions recorded to start personalising the graph. As soon as you've got previous records to work from, these will show as influencing the graph.
Insulin taken with food
Insulin taken with food does not directly influence the graph but indirectly in combination with the food that it is recorded alongside. Therefore, the food record is shown as the primary factor affecting the graph. Once the food is no longer influencing the graph, any remaining active insulin will also no longer show as influencing the graph. We aim to improve this in future once the data research indicates that it is safe to do so.
Multiple Foods
If multiple foods are recorded close to each other, the graph will revert to having only the time of day influence the graph. We are still researching the effects of combining food records. An alternative is to record all food eaten around a meal in one record - e.g. Eggs on Toast with Yoghurt. We can then directly compare this Eggs on Toast with Yoghurt record with other Eggs on Toast with Yoghurt records but not Eggs on Toast alone or Yoghurt alone.
Multiple Correction Doses
If multiple correction doses are recorded close to each other, the graph will revert to having only the time of day influence the graph. We are still researching the effects of combining correction doses.
Combination of Actions
Our research shows that the last action taken of either food, correction doses or hypo treatments often is the most relevant to the NowPatient Diabetes graph. This is because the graph takes into account both what the action is but also the human behaviour behind the action. Therefore, currently, the graph will use the last action recorded to influence the graph.