Blog

ANALYTICS FOR DIABETES PREVENTION

2017-06-20/ Bratati Mohanty

Analytics for Diabetes Prevention

Globally, there are over 415 million people living with diabetes today and more than 640 million of us may be living with diabetes by 2040 as estimated by International Diabetes Federation1. It is growing alarmingly in the world, specifically, India. India ranks among the top countries in the world with high diabetic population. One in every two adults with diabetes goes undiagnosed. If the growth in the number of diabetics is not intervened and prevented, the prediction of diabetes becoming one of the most leading causes of death is inevitable.

While PredPol’s predictive boxes predict that a crime will happen in the prediction area, there is no guarantee that an incident or arrest will occur. The presence of police officers in the prediction areas creates a deterrence and suppression effect, thus preventing crime in the first place.

PredPol does not collect, upload, analyze or in any way involve any information about individuals or populations and their characteristics – PredPol’s software technology does not pose any personal privacy or profiling concerns. The algorithm uses only three pieces of data – type, place, and time – of past crimes.

The Chicago Police Department Take Predictive Policing One Step Further

As with PredPol, the approach in predictive policing seeks to forecast where and when crime will happen; another focuses on who will commit crime or become a victim…

Predictive analytics for diabetes prevention

IBM Watson Health and the American Diabetes Association, ADA, have joined hands to create new digital tools that will ultimately envisage how diabetes is prevented, identified and managed. They aim to leverage the cognitive computing power of Watson and the association’s repository of diabetes clinical and research data to create digital tools for patients and providers. ADA has a repository of 66 years of data, which includes aggregated data about self-management, support groups, health activities and diabetes education3. The project includes training to understand diabetes data to identify potential risk factors and create recommendations for health decisions. The goal of the collaboration is to develop solutions that enable the diabetes community to optimize clinical, research and lifestyle decisions, and address important issues that influence health outcomes, such as social determinants of health4.

US researchers have earlier used the popular statistical modelling method called “proportional hazards model” to predict an individual’s risk of diabetes. These types of models predict the time that passes before some event occurs (in this case, the occurrence of diabetes).In a proportional hazards model, the unique effect of a unit increase in a covariate is multiplicative with respect to the hazard rate.

For example, taking a drug may reduce one's hazard rate by 50% for occurrence of diabetes, or, having higher than average carbohydrate intake may double its hazard rate for failure. The model identified a list of 7 factors/variables that are highly predictive of diabetes risk.The subjects were then scored using the trial data based on these factors.

The results revealed that people with scores in the top 25% were at highest risk of getting the disease. The trial population was divided into quarters of pre-intervention risk on the basis of model predictions and assessed. Patients at extreme predicted probabilities of developing diabetes should have a more straightforward decision about the benefit of treatment.

The aim was to use data analytics to predict which pre-diabetic patients would gain the most from which treatment approach, treatment with a drug that prevents diabetes, or from a lifestyle change such as weight loss or regular exercise.

This approach is known as the “precision medicine” approach in the healthcare sector. Participants were classified into risk pools based on the model’s prediction.

Effect of Lifestyle and drug (Metformin) on hazard risk7

(Referring to the picture above)

High risk participants:

- Were highly benefitted by the use of the drug, which reduced their risk of diabetes by 21%

- Lifestyle interventions reduced their chance of developing the disease by 28%

Low risk participants:

- No benefit from the drug

- Same intensive lifestyle change brought down their risk by only 5%

All participants:

Exercise and weight loss, with guidance from a health coach, benefited all to some extent, irrespective of their risk scores.

Many patients receive treatments unnecessarily with very low benefit. This issue can also be prevented by such analysis, thus reducing the healthcare cost involved. Customized tailoring of treatment for pre-diabetics and diabetics can improve the lives of all significantly. Doctors can be well-informed to determine the best treatment path for each patient as well as identify potential risk factors for each individual. Most importantly, the accuracy of the model is the key to determine the outcome of any project.

Many patients receive treatments unnecessarily with very low benefit. This issue can also be prevented by such analysis, thus reducing the healthcare cost involved. Customized tailoring of treatment for pre-diabetics and diabetics can improve the lives of all significantly. Doctors can be well-informed to determine the best treatment path for each patient as well as identify potential risk factors for each individual. Most importantly, the accuracy of the model is the key to determine the outcome of any project.

Various other organizations are trying to leverage big data and analytics to manage the disease by proper monitoring and controlling, so that optimal care is provided to the patients at reduced costs.

For example, Glooko, founded in 2010, provides a diabetes management platform which is sold directly to the healthcare units and insurance providers.8Patients can use the Glooko mobile app on their smartphones to enter information about their food intake or physical exercise to make appropriate decisions. Healthcare professionals can track and analyse a patient’s real-time progress to provide optimal care to the patient.

Comments