Big data is quickly expanding to a number of industries, and healthcare is no exception. With the use of big data, all kinds of medical records and studies can be digitized and easily analyzed. However, using big data in the healthcare space comes with its own set of challenges that anyone involved in the industry should be aware of.
Unreliable Data
The U.S. government and other private organizations have poured billions of dollars into digitizing medical records, but so far the data has basically just stayed where it is. Next to nothing has been done to analyze and actually use that data, in large part because the data is incredibly difficult to use and interpret. Medical data is often stored in databases, which tend to not be easily compatible with each other. Some of the best and most useful information is often added to records as freeform notes, which can be hard to digitize and interpret. Medical records also pass through multiple peoples hands, from nurses to techs and doctors, before making it to the digital world, meaning it is relatively easy for errors or discrepancies to enter someones personal information. One of the biggest pushes for big data in healthcare is to use information from scientific studies, but researchers are resistant to do that. Study data is notorious for being replaced by something new and more updated every few years, which means much of the data pulled by the digitization effort is likely already obsolete.
One solution to muddling through the sea of potentially unusable data is to test it out. If a study or database claims something is true, experts recommend creating a small study with a limited number of participants to test the claim. Researchers should then be able to decipher which information is correct and usable and what needs to be overlooked. Much of the errors in records can be lessened with the growth of online medical transcription courses to ensure data is accurately moved from paper to the computer.
Unnecessary Data
Another problem with big data in healthcare lies in the fact that it cant solve every problem. Many medical issues, especially when it comes to rare diseases and conditions, need to be addressed with a scientific study, not a computer algorithm. When it comes to saving lives, especially if there is an immediate need, patients need drugs developed by traditional clinical studies and research, not numbers pulled from a computer. However, using big data can be more cost effective and quicker, which means many research organizations and hospitals are often anxious to turn straight to a full-blown study.
The key to solving this problem is to have data analysts and scientific researchers and doctors work closely together. When a full study is warranted to solve a problem, doctors should be alerted. Just like with testing existing data in a database, the studies dont necessarily need to be very large, especially for rare conditions. In some cases, studies can even be consolidated to combine tests for multiple related issues into one series of studies. Once the study results are finalized, the data can then be put into a large database.
No Time for Studies
The issues of unreliable and unnecessary data both warrant more scientific studies, which is easier said than done. Many health researchers and doctors say they simply dont have time to put together and execute a study for every issue that arises. According to FDA commissioner Robert Califf, its the top problem facing big data usage in healthcare.
There isnt necessarily an easy answer to this problem, and it could have a large affect on big data usage in the coming years. The solution could be a combination of additional research resources, a large medical care staff, and centers dedicated solely to studying certain ailments.
Big data in healthcare could be the start of something new, but it needs to be combined with traditional research to truly be effective. There may be growing pains during the transition process, but the final result could be beneficial to the overall health of people all over the world.
