Problem Statements:
- Reduce the workload of the medical staff taking care of the patient.
- Reduce the frequent needle Blood Sugar checks.
- Ability to predict critical situations for the patient Solution:
- Building an algorithm that can estimate a patient’s likelihood of going into hypoglycemia or hyperglycemia.
- Providing a trigger warning when patient seems to go towards either of the conditions, allowing nurses and patients early warnings while allowing lower physical monitoring.
Why this solution? Using our programs priority-based mechanism and future implementation of Priority/To-do list we can provide conscious and user-friendly alerts to notify when someone might need more manual care.
Build:
-
Input period: We need basic details of the patient required to solve BMI and daily Insulin levels (1st 2-4 days).
-
We need Blood sugar levels right before meals and basic meal count.
-
Factors: Carbohydrate Intake: Daily Carbohydrate log as the food is served to patient.
Physical Activity: Since the patient is hospitalized, we can assume that they will have minimum physical activity.
Insulin: For diabetes patient, we can calculate the Basal Insulin Level + Insulin injected. Otherwise just the basal insulin level for all the others.
Metabolic Rate Factor: We can estimate this by the Harris-Benedict Equation.
-
Equation to use: Change in Blood Sugar (mg/dL) = (Carbohydrate Intake (grams) / Carb-to-Insulin Ratio) - (Metabolic Rate Factor x (Initial Blood Sugar - Target Blood Sugar)
-
Code logic: a. Read into data from info page into a datatype (most likely an array).
Estimating the change in blood sugar 15min, sending the new blood sugar value separately to get stored. This can be achieved by inputting tested blood sugar, then combining the data from the other factors to give a close estimate.
If()…else() Block: i. If the patient is diabetic, we’ll input the range differently and then go into a case switch block that will return RED, YELLOW, GREEN; based upon how the current levels are.
else for all the other patients we’ll have a standard value system. b. This is where our program comes into play, our focus on prioritizing care is what stands us apart from every other option on Medicare.
- Information Re-use: We take back the data generated by the algorithm to monitor the fluctuations over a longer period, and in the process, we refine our Metabolic Rate Factor and eventual introduce a new variable Glycosydic Index to further temper the results and increasing accuracy.
Conclusion:
By Implementing this algorithm, we can allow nurses and health care worker to spread out attention to patients at priority basis, allowing them work more efficiently.
Most importantly this will reduce the number of patients going into Hypoglycemia, which can even lead to coma and other dangerous outcomes. It will also reduce the hyperglycemia chances, reducing the chances of internal infections.
This will reduce the reliance of constant prick test of patients.
