E-mail RSS
 
 
Home » Uncategorized » Predictive Analysis in mHealth
formats

Predictive Analysis in mHealth

In the last blog, we discussed Walmart’s successes with predictive analysis using their data-gathering RetailLink software. The company is able to dominate sales because they have the information necessary to predict needs, and market accordingly. Walmart takes external factors like the weather, internal factors from their collected data, and combine them with demographic knowledge to cultivate a sales triumph. This could be applied to healthcare, especially in the mobile realm, with great outcome. An app that recorded the daily health data of individuals, combined with EHR and external factor data, could achieve the beginning stages of analysis. This information could then be applied to every facet of the healthcare realm.

A simple example would be dietary data. Daily dietary data was recorded that revealed that in a certain time period, people drank more sweet drinks than usual. This was compared to external data which said that this was when it was cold outside and coffee companies with sugary, warm drinks were running sales. Also, this time coincided with the holiday season, which gave people more opportunity and motivation to drink sugary sodas or alcoholic drinks. This data collection is given to dentists, who are then able to send teeth cleaning reminders to their current patients, or even run deals or special marketing for dental work near the end of this time. These reminders come right when the patient is beginning to feel the effects of all the sugar on his teeth, and thus is motivated to take care of them. This cause and effect is already highly visible in the holiday season, as it is a known fact that the season often produces weight gain. So after the holidays, many more weight loss commercials are released and sales are rampant.

This analysis can get more precise though, in dietary and other fields, such as if a fast food chain were to push a certain special meal, for example a McRib. They obviously have done the analysis with their marketing team to determine that this is the right time to bring back and advertise this meal, and they are expecting big sales. The next step would be for pharmaceutical companies to push out their advertisements for blood pressure and cholesterol medicine, or other such medicine. The correlation is not huge, but it is proven, and small connections can add up. These small connections are when the demographic data and external factors align to create an obvious consequence. This information could be utilized in a community health center setting as well. If the CHC were to gather data on their patients they would begin to understand the nature of their surrounding demographic. Dietary and fitness data could illuminate the basic needs of their patients. They could then combine this with external factors: a park nearby, or a lot of young sports leagues, a retirement community or a smattering of nightclubs. Each of these comes with its own unique set of health needs. A health clinic near a youth sports hotspot, which has patients whose diet contains little dairy, would perhaps stock up more in cast and splint items than cholesterol and blood pressure regulators.

The potential of a predictive analysis system for healthcare goes off in many directions. Fitness opportunities would be numerous, and the data gathered from chronic disease patients could help tailor suggested treatment options in each individual hospital, just to start. The problem is that the data has not been gathered yet. The strength of Walmart’s system is their data. RetailLink, their data collector, holds petabytes of information, larger than most of the world’s server collections. Because of their years of sales data, Walmart can predict based on past trends. For healthcare to step into these predictive analysis possibilities, data must be available. With the app variety today, it is hard to imagine that large-scale data collection is far off. This should be the next step in mHealth innovation; health data collected for purposes of predictive analysis.

 
comments  Comments Off
© © 2011 Decide mobility Inc., All Rights Reserved