I have written previously about how the Intuitive Mind affects type 1 diabetes (T1D) self-management, for example based on people's situational awareness of changes in their own blood sugar as they occur. In one prior study, our team found that Intuitive-level variables such as motivation and social perception (based on a daily survey) were related to successful daily blood sugar control (based on time-in-range [TIR], a commonly used metric from continuous glucose monitoring [CGM]) among adolescents with T1D. I also wrote about my own experience trying a CGM for 2 weeks, which did seem to result in increased situational awareness. In another study, we found that adolescents' proactive use of a hybrid closed-loop system (pictured above), which incorporates an insulin pump and a dosing algorithm together with a CGM, resulted in better TIR results than when people waited for the technology to tell them what to do. Specifically, adolescents who looked at their CGM readout less often, and those who entered carbohydrate values to more accurately dose insulin with their meals, had better TIR results.
Our latest study with this population, now available online before print in a forthcoming issue of Health Psychology, combined both sensor and survey measures as potential predictors of T1D self-management. We again found that less interaction with the glucose pump was associated with better glucose control, and we replicated the finding that entering carbohydrates with a meal predicts better TIR. Both of these findings again suggest that the closed-loop system is most helpful when used in a proactive and situationally-aware way, versus a more reactive way where the user waits for alerts to know that something has gone wrong. In this study, we found that receiving more alerts was actually predictive of worse TIR results.
A new feature of the latest study was that we asked people to fill out triggered surveys whenever they interacted with their insulin pump, in addition to a version of the single end-of-day survey about self-management that we had used in our prior survey research. These more intensive surveys allowed us to gather data about the reasons why people were interacting with their closed-loop system (which, if it works correctly, shouldn't require too much direct user interaction). We characterized a set of 11 different reasons into "proactive" and "reactive" categories. And consistent with our understanding of situational awareness in TMT, we found that the proactive interactions predicted better TIR.
Many of the users' reasons for interacting with their closed-loop system had to do with a perception that their glucose level wasn't where it should be, which we allowed them to classify as "I think my blood sugar might be too high," "I think my blood sugar might be too low," or "I just feel 'off'." The perception that blood sugar was too high predicted better TIR. But by comparing to the CGM results collected at the same time as the survey, which we also had in this dataset, we discovered that participants were actually wrong most of the time about this risk -- they thought their blood sugar was too high, when it was still within the target range. Our interpretation is that people had learned some subtle bodily cues that allowed them to predict their blood sugar was likely to become too high in the near future. Because they then took action (e.g., giving themselves a bolus of insulin), the predicted spike in blood sugar never actually happened. This implies a level of situational awareness that can't be replicated by CGM monitors -- at least not yet! Maybe an AI program could learn these subtle cues in the same way that humans do, for a better future version of a closed-loop system that would automatically deliver more insulin at these times.
Overall, our findings in this latest study replicate some important results from our prior research on T1D self-management. They also extend the prior findings with a look at people's thought process in the moment when they are actually interacting with their insulin pump. Our findings were again consistent with TMT, in that an Intuitive-level awareness of one's own blood sugar seems to permit more focused and intentional interactions with the closed-loop system. People with diabetes still need to do some self-management, even with a closed-loop system that automatically delivers insulin when their glucose levels get too high. The key seems to be interacting with the device only at times when it is most appropriate and necessary, versus checking it all the time or never using it unless an alert message comes up. That kind of subtle "appropriateness" determination is something that our participants said they were making largely unconsciously (in a parallel qualitative study with these same participants). In other words, it's something that happens at the level of the Intuitive Mind.
Comments
Post a Comment