Big data in healthcare is smaller than it should be. While big data in manufacturing, insurance and finance seem to get all the glory, its use in the healthcare industry extends beyond just monetary benefit: it can also be used to help prevent disease and illness, provide long-term benefits and save lives.
Big data in healthcare is growing though, and the advancements are not only in predictive analytics and machine learning that is accelerating its use but also how companies and institutions are using these advancements.
However, real concerns about privacy, fraud and security, threaten the use of data at its full potential. A non-uniform set of data and constraining laws and regulations also are hindering further adoption.
What can big data do?
Using healthcare data can help predict disease so doctors can take a targeted approach and focus on preventative patient care when it’s easier and less-costly to manage. With the benefit of a centralized data set, artificial intelligence can analyze the data, which includes patient family history and DNA, and employ predictive analytics to identify high-risks.
Here are four companies already taking advantage of access to this information.
Optum Labs, a research collaborative, is using big data to better achieve individualized treatment. Optum Labs has collected over 30 million patient electronic health records (EHRs) and uses this information to build and maintain a database for predictive analytics tools.
What about big data in a geographical situation? Asthmapolis relies on mobile technology and data to pinpoint where people are using their rescue inhalers. Not only is this information sent to the patient to help them manage their asthma, but the data is shared with public authorities such as the Center for Disease Control and Prevention so they can think differently about public health surveillance and see where people are experiencing these asthma attacks. From there, data can be used to determine if there are pollution or workplace environmental triggers exacerbating the problem. Accessing that information allows Asthmapolis to send the collected finding to the patient and local authorities.
Radiology is prone to errors and discrepancies averaging 3-5 percent daily, sometimes more. Carestream, a medical imaging provider, believes the route to better analysis and fewer errors will be using numbers associated with pixel data. With the use of big data, algorithms are established by analyzing hundreds of thousands of images and evolve through artificial intelligence to create a comprehensive analysis that radiologists can study faster and diagnosis quicker with greater accuracy.
A recent study from Bayer indicated a single cancer patient can generate nearly one terabyte of biomedical data from their routine diagnostic data and clinical data. Hiding within those mounds of data, according to Atul Butte, director at the Institute for Computational Health Sciences, UCSF., is the knowledge that could change the life of a patient or change the world. Researchers are using information from big data to analyze cancer on a cellular, patient and population level. The goal is to provide oncologists with the weapons to make a customized drug treatment plan specific to a patient’s cancer cells.
Even the use of wearable athletic technology, such as Fitbit, is providing users with metrics designed to tell them the numbers of steps, speed, heart rate and other health information through phone apps.
There are countless reasons why the medical industry should adopt big data and the above are just a sample of its lifesaving possibilities. However, there are reasons that technology is stalling.
Why big data isn’t moving fast enough
There are a host of obstacles and perhaps the biggest technical challenge is incompatible data systems.
Disconnected systems. Data resources are scattered in different data pools, including hospital medical records, settlement and cost data, academic medical research data and population and public health data of government surveys. There is not much connection between these data sets.
Actionable data. Turning the healthcare data into something actionable is a major challenge. Electronic health records are often incomplete, inaccurate or not even in a usable format. They are a step in the right direction, but it’s a matter of finding how to collect the data and store it in a centralized location without compromising its security— a problem highlighted in the May 2017 ransomware attack on hospitals in the United States and the United Kingdom.
Patient confidentiality. Another key area is patient confidentiality. Laws vary from state to state and country to country, determining what patient information can be released with or without consent. The Health Insurance Portability and Accountability Act (HIPAA) requires that patients give written consent before their information is disclosed; it does allow for the release of data for research purposes but has a stringent list of technical safeguards that organizations need to consider.
Lack of sharing. Research institutes and pharmaceutical companies pour millions into conducting clinical trials and developing their own intellectual property; many do not want their findings in the public domain where competitors can take advantage of their work.
Where big data in healthcare begins
In general, the current research on medical data is not yet mature, and there are many problems to resolve. To take full advantage of the profound patterns contained in these massive amounts of data, it will be essential to have big data storage, mining and analysis. When that data is used to its fullest potential, it can support technologies and talents that benefit research and further serve a wide range of medical applications such as public health, medical care and medical insurance.
The current standards and technologies are inadequate to meet the requirements of the integrative applications of big data in healthcare. The difficulties are two-fold. First, the data lack uniform standards, consistent description format and presentation methods. Not having a uniform standard makes data comparison, analysis, transfer, sharing, and other processes difficult. Second, different levels of structured, semi-structured and unstructured data integration are difficult. At the same time, each database uses different software and formats.
Standards must be conformed to code so data can be shared, if necessary. Just as cars must be designed so the gas pump nozzle fits into the gas tank of any car, and outlets must be uniform to accommodate any plug.
Standardizing medical data is possible with government intervention and standards, but governments need to be on-board. Laws and regulations for making use of big data in the healthcare industry need to be enacted. Research subsidized with government funding should also be made public and shared with other institutions to maximize its potential.
The more the healthcare industry can access interactive and responsive software capable of managing big data, the better the doctors can improve their decisions, and researchers can make greater advancements in cures.