Big Data Pitfalls Still Plague Healthcare Industry

Incomprehensible amounts of data are collected by all sectors of the healthcare industry on a daily basis. It is important for organizations to not only understand how to handle and collect this data, but also to translate it into actionable information that can help transform healthcare delivery. On Tuesday afternoon at the AHIP Fall Forum 2012 conference, Ashish Jha, MD, MPH, associate professor of Health Policy and Management, Department of Health Policy and Mangement, Harvard School of Public Health, spoke about the need to improve the use of data and analytics to support healthcare delivery.

Dr Jha began his presentation by saying that we as a society don’t understand the social tradeoff that occurs because of inefficiencies in our healthcares system. He illustrated this point by talking about how his hometown of Newton, MA, recently reset their budget and had to lay off teachers because of rising healthcare costs. “Laying off teachers to pay for more MRIs is not a good social tradeoff,” said Dr Jha. He added that he is a researcher, analyst, and physician, and that, while he believes he is a better researcher and analyst than he was 10 years ago, he is unsure as to whether he is a better physician. His explanation was that, as a physician, he doesn’t receive nearly the feedback he does as a researcher or analyst. “The feedback loops are lacking,” he said, and mentioned that if he submits a research study to a journal and it is not accepted, he is given feedback as to why. Conversely, when he treats a patient for pneumonia, he may not receive any follow-up feedback at all.

Complicating things even further is the fact that much of the big data that exists in the healthcare industry today is not “good” data. Despite the fact that over a billion patient visits have been recorded via an electronic health record (EHR) this year, there are pitfalls that make it difficult to analyze this data in order to produce information that is clinically useful. There are a number of variables that are the cause of this issue. Interoperability exist when a hospitalist sees a patient that was previously treated at another hospital; depending on whether the technologies between these hospitals sync up, the patient’s information could either be readily available or completely nonexistent. Some hospitals are more technologically advanced than others, whereas many still rely on nurses to input patient information manually, which leads to human errors. In many cases, healthcare data are incomplete, and this is many times caused by the overwhelming amounts of EHRs that have different templates and several different methods to enter information. All of these variables can lead to massive amounts of data that aren’t helpful.

Dr Jha mentioned some solutions for making big data more relevant. He explained that there needs to be a change in the data requirements necessary for billing. In order to “fill in the gaps,” payers must become more involved, as they are often the source for incomplete data, according to Dr Jha.

It may be another 10 to 15 years before issues like interoperability are improved to the point where many of the aforementioned pitfalls can be avoided. However, in the meantime, if providers and payers can work together to come up with more consistent ways to collect data, the result will be more clinical relevance and improved healthcare delivery.