LAB 3 WRITE-UP
Descriptive Stats and Graph
| || Gold Standard || Spree
| Average || 96.65269939 || 95.5308642
| Standard Deviation || 1.918462545 || 0.870378299
| Count/Endpoint # || 326 || 324
| Standard Error || 0.106253813 || 0.04835435
| || Gold Standard || Spree
| Average || 98.08976898 || 98.94078947
| Standard Deviation || 23.03054395 || 24.83748725
| Count/Endpoint # || 303 || 304
| Standard Error || 1.32307018 || 1.424527591
|Pearson's Correlation||0.192798089||No Positive Correlation
Summary of Results:
Based on the data, it can be concluded that when measuring temperature, the Spree device was very inaccurate as opposed to the gold standard. As shown by the calculated Pearson’s Correlation Coefficient of 0.19 which is relatively close to 0, it can be concluded that there is little to no correlation between the gold standard measurements and the Spree headband as shown by no linear trend between both measurements. Furthermore, through analyzing the P-Value through conducting a t-test, with a result of 1.85E-20, it can be understood that there is a significant difference between the measurements of the Spree headband and the gold standard, showing that the Spree headband is not capable of accurately measuring temperature.
|Pearson's Correlation||0.779956521||Strong Positive Correlation
Summary of Results:
Based on the data, it can be concluded that when measuring heart rate, the Spree device was quite accurate as compared to the gold standard. As shown by the calculated Pearson’s Correlation Coefficient of 0.78 which is relatively close to 1, it can be concluded that there is a strong positive correlation between the gold standard measurements and the Spree headband as shown by the linear trend of both measurements. Furthermore, through analyzing the P-Value through conducting a t-test, with a result of 0.66, it can be understood that there is not a significant difference between the measurements of the Spree headband and the gold standard, showing that the Spree headband is in fact capable of accurately measuring heart rate.
Design Flaws and Recommendations
The headband is unattractive and would be too bulky for exercise.
The headband could become uncomfortable for long periods of use due to squeezing the head.
The sensor that is used to calculate heart rate and temperature seems to be too loose, resulting in possible breakage of the device.
The sensor may have trouble calculate heart rate and temperature due to being located on the forehead.
Move the device to the wrist or upper arm to increase comfort.
Moving the device to one of these locations should hep in making more accurate readings of temperature and heart rate.
Decrease sensor size and try to secure it better to prevent the breaking of the device.
Provide multiple design options to customers to fulfill their design specifications.
Experimental Design of Own Device
Diabetes mice patients and mice patients without diabetes:
2000 total mice.
Experimental Group= 1000 mice with diabetes.
Gold Standard=1000 mice without diabetes.
1. Make a much smaller version of the bar and try testing it on mice.
2. Create a control group of mice who do not have diabetes (gold standard), and then make a group that does have diabetes (experimental group).
3. Import the smaller version of the diabetes bar into the group that has diabetes.
4. Take data calculations of both groups and see if their glucose levels are similar (using inferential statistics and descriptive statistics).
5. Perform numerous trials to see if the bar is effective.
6. Make necessary improvements if the bar is not effective, and if it is effective, submit a request to the FDA to perform trials on human patients.
We decided our patient group on the basis of mice because we wanted to be able to test our device safely (not on humans first) and conduct many repeated trials. Our mice research will allow us to see how effective the bar is because if the bar is helping the mice with diabetes, their glucose levels should be similar to the levels of mice that do not have diabetes. We will be able to take the data from the mice and perform many inferential statistics like t graphs and pearsons coefficient to determine the effectiveness.
Diabetes human patients:
Type 1 or Type 2, males or females, ages 18-60( Adults) who are in good physical and mental health.
The sample size will be about 100-200 patients.
1. Find people to participate in the diabetes bar trial.
2. The device will be implanted for about a year in each patient.
3. Scheduled blood glucose level tests will be conducted about every seven days by professionals (doctors).
4. Patients will be told to use other diabetes control methods when they start to feel low or high on blood sugar.
5. Patients will be asked to take daily recordings of their glucose levels, how they feel (health wise), and their day-to-day thoughts on the device.
6. Use the data that is taken from the patients and use inferential statistics to test for the effectiveness of the diabetes device (T tests and/or pearsons coefficient).
7. Make conclusions based on the results and compare it to normal glucose levels.
8. Return to the design stage if necessary to make improvements.
9. If the device is found to be effective and safe, move towards the marketing of the device.
Rationale: We decided the patient group for part two of our experiment based on the study done on the birth control device IMPLANON. They had a successful trial of their device and our device will be very similar, except for the controlling of a different health issue. In their study, they chose to take patients ages 18-60 that were in good physical and mental health. Their sample size was 108 people, so we decided we would make ours 100-200 people. The only thing we have done differently is that we will also include male patients, as diabetes occurs in males and females. The implanon device was for females only so naturally they only took female patients. They chose to conduct their trial for three years, but we will do ours for a year because we believe we will not be able to conduct our trial for that long. We also decided to allow patients to use other diabetes control methods when the diabetes device seems to not be working.
A Bioequivalence Study of IMPLANON and Radiopaque IMPLANON (345280(P05720).(2008). ClinicalTrials.gov. Retrieved from CLinicalTrials.gov website https://clinicaltrials.gov/ct2/show/NCT00620464