Specific Information Unveils General TrendsWed, 09/06/2017 - 16:02
No one can predict the future but Big Data provides professionals with the kind of specific information needed to make fairly accurate projections. In the healthcare sector, Big Data can help detect dangerous trends and provide the necessary knowledge to allow for decisive action that could save lives. How effective it is relies on the velocity, volume, variety and veracity of the information collected.
“Big Data is mobile, analytics, cloud computing and social networks,” says Antonio Carrasco, CEO of Grupo PLM, which specializes in Big Data for the healthcare sector in 13 Latin American countries.
The company, which started 75 years ago as an editorial house for the medical segment, collects medicinal information from manufacturers and publishes the data on its website, which visitors can search. In 2016, over 100 million health professionals clicked on Grupo PLM’s Mexico website, Carrasco says. The site can handle over 2,500 searches per second and receives a variety of information over its 35 digital channels.
“When talking about Big Data, it is necessary to include most users in the sector, which is what PLM does. We have over 200,000 physicians working with our information on a daily basis,” says Carrasco. The sheer number of doctors searching the site’s resources gives the company ample data to analyze and detect trends. “We are like a small Google because people search for very specific medical information through us.”
PLM is also working with artificial intelligence or semantic analytics. “We teach IBM’s Watson to understand what is written on paper. This is semantic analytics. Artificial Intelligence (AI) can begin reading a page and give you the remaining information,” says Carrasco. Among its applications, AI can recommend a dosage, inform a doctor if the prescribed drugs will interact with each other or if there is a certain type of food or environmental element the medicines will react to.
By detecting trends, Big Data can be used to improve healthcare. Carrasco points to the 2017 flu H1N1 season as an example. During the 2016 flu season, Grupo PLM noticed a hike of 1,113 percent in searches for Tamiflu in February 2016 in comparison with the previous month. “We knew there was an epidemic because general doctors were desperately looking for Tamiflu. Normally, this trend should be relatively steady but the spike in searches was due to thousands and thousands of patients coming down with the flu,” says Carrasco, adding that in 2016, there was a scarcity of Tamiflu. “It was sold out in all drug stores in Mexico because they were not expecting an epidemic. The increase in searches for the drug was atypical.” Carrasco adds that H1N1 was responsible for over 6,000 deaths in Mexico in early 2016. Sharing such data could improve the detection of trends and ultimately improve healthcare for patients. It would also enable companies to improve treatment and their cost-efficiency.
Big Data helped identify the epidemic and action was taken, Carrasco says. The patent for Tamiflu was expired by COFEPRIS in March 2016, leaving the way open for other companies to use the medicine’s active substance, oseltamivir phosphate, to create generic versions of the drug. In addition, the Mexican Ministry of Health ran preventive campaigns throughout the country in winter 2016 to remind people to get their flu shot.
The flu is not the only trend Big Data can reveal. Carrasco says PLM’s data has uncovered a number of surprising results. One example illustrates his point: the majority of specialists searching for erectile dysfunction drugs are gynecologists. “Erectile dysfunction is a couple problem, not a man problem,” Carrasco says. “That is the advantage of Big Data: discovering something you never would have thought of.”