Prevention of post-bariatric hypoglycemia using a novel glucose prediction algorithm and mini-dose stable glucagon
The present disclosure provides a hypoglycemia prediction algorithm (HPA) specifically designed for the unique postprandial glycemic patterns characteristic of PBH. This algorithm can predict impending hypoglycemia by performing a series of steps. The steps can include collecting data from at least one sensor. The data can comprise a concentration of glucose in the bloodstream of a subject. The data can be processed using the HPA and impending glucose concentrations can be calculated. The method can then provide for determining whether the predicted glucose concentrations are lower than a hypoglycemic threshold parameter. In response to determining that the predicted glucose concentrations are lower than the hypoglycemic threshold parameter, the method can provide for enacting an impending hypoglycemia protocol.
1. A system for outputting an indication of an optimal time to deliver a dose of glucagon, the system comprising:
a glucose sensor configured to output glucose data related to a concentration of glucose in the bloodstream of a patient;
a glucagon pump;
a memory containing machine-readable medium comprising machine executable code having stored thereon instructions for performing a method of delivering glucagon; and
a control system coupled to the memory comprising one or more processors, the control system configured to execute the machine executable code to cause the processor to:
store, in the memory, a model for determining an optimal time to inject glucagon based on a predicted blood glucose level;
receive, from the glucose sensor, a set of glucose data;
process the set of glucose data using the model to determine an optimal time to administer glucagon to the patient wherein the model comprises:
a post-prandial hypoglycemia prediction module, designed to predict low glucose levels and identify when the patient's glucose level has fallen below a hypoglycemic threshold, upon at least a portion of the set of glucose data determined to be a meal pattern; wherein the meal pattern is determined by a positive rate of change (ROC) of the patient's glucose level followed by a negative ROC of the patient's glucose level; and
a proximity hypoglycemia prediction module, designed to predict low glucose levels and identify whether a rising or falling ROC of the patient's glucose level indicates onset of post-bariatric hypoglycemia at a predetermined future time, upon at least the portion of the set of glucose data determined to be a non-meal pattern; and
send a command to the glucagon pump to administer glucagon at the optimal time; and
wherein the post-prandial hypoglycemia prediction module predicts low glucose levels and identifies when the patient's glucose level has fallen below a hypoglycemic threshold according to the following algorithms:
(a) if the glucose level is ≥100 mg/dL, then the glucose sensor continues to collect glucose data;
(b) if the glucose level is ≥75 mg/dL and <100 mg/dL, then the ROC of the glucose level is determined,
(i) if the ROC is between −0.5 mg/dL/min and −5 mg/dL/min, then a time-to-hypoglycemia (TH) is determined based on the ROC,
(A) if the TH is <20 minutes, then the patient's glucose level is predicted to fall below the hypoglycemic threshold, and the module generates an alert, takes a blood sample, and sends a command to inject glucagon: or
(B) if the TH is not <20 minutes, then the glucose sensor continues to collect glucose data; or
(ii) if the ROC is not between −0.5 mg/dL/min and −5 mg/dL/min, then the glucose sensor continues to collect glucose data; and
(c) if the glucose level is <75 mg/dL, then the ROC of the glucose level is determined,
(iii) if the ROC is <−0.1 mg/dL/min, then the patient's glucose level is predicted to fall below the hypoglycemic threshold, and the module generates an alert, takes a blood sample, and sends a command to inject glucagon; or
(iv) if the ROC is not <−0.1 mg/dL/min, then the glucose sensor continues to collect glucose data.
2. The system of claim 1 , wherein the processor is further caused to determine whether the predicted blood glucose level will fall below a threshold within a window of time based on the model.
3. The system of claim 2 , wherein the processor is further caused to determine the window of time based on the time of a subset of the set of glucose data that indicates the patient has recently consumed a meal.
4. The system of claim 1 , wherein the proximity hypoglycemia prediction module is designed to predict when glucose level is below the predefined hypoglycemic threshold.
5. The system of claim 1 , wherein the predetermined future time is less than thirty minutes.
6. The system of claim 1 , wherein the patient has post-bariatric hypoglycemia.
7. The system of claim 1 , wherein the onset of post-bariatric hypoglycemia occurs one to three hours after a meal.
8. The system of claim 1 , wherein the positive ROC is at least 1 g/dL/min, and the negative ROC is less than 0 g/dL/min.
9. The system of claim 1 , wherein the negative ROC is between −0.5 g/dL/min and −5.0 g/dL/min.
10. The system of claim 1 , wherein the proximity hypoglycemia prediction module predicts low glucose levels and identifies whether a rising or falling ROC of the patient's glucose level indicates onset of post-bariatric hypoglycemia at a predetermined future time according to the following algorithms:
(a) if the glucose level is ≥150 mg/dL, then the glucose sensor continues to collect glucose data; or
(b) if the glucose level is <150 mg/dL, then the ROC of the glucose level is determined,
(i) if the ROC is between −0.5 mg/dL/min and −5 mg/dL/min, then a time to a low glucose level (t low ) is determined based on the ROC,
(A) if the t low is <30 minutes, then the patient's glucose level is predicted to be low and the module generates an alert, takes a blood sample, and sends a command to inject glucagon; or
(B) if the t low is not <30 minutes, then the glucose sensor continues to collect glucose data; or
(ii) if the ROC is not between −0.5 mg/dL/min and −5 mg/dL/min, then the glucose sensor continues to collect glucose data.