Ask ten people what big data means and you get ten different answers. Nearly all will agree that it includes data sets that are so large in volume (terabytes to petabytes) or complex in nature that traditional data processing applications can’t be used. The one thing nearly everyone in healthcare will agree on is that data can drive not only marketing and sales, but diagnoses and outcomes as well. That said, having big data and using it are two different entities. Who does the analysis and what the outcomes of that analysis are often determines not only the future of a company, but often the end results of patient care.
In most companies, big data is used to help identify the customer’s needs. Sadly, marketing often remains separate from the rest of the enterprise and may be operating without the benefit of the available data to promote products. While this data can be used to customize the end-user experience and to eliminate the one-size-fits-all solution commonly offered today, customers are still faced with approaches by vendors that only minimally tailor solutions to them – and consumer patience is wearing thin. The information needed to address the specific needs of the customer is becoming more and more available, but getting it into the right hands remains a challenge.
In healthcare, big data remains largely unproven, although that hasn’t stopped companies and vendors from jumping on the big data bandwagon. Big data analytics was a hot topic at HIMSS 2014, and it no doubt will be again this year. The promise has been promoted that big data can provide a better quality of care and reduced expenditures, but the evidence to support those claims – at least to date – is somewhat tentative at best,
Big data analytics takes mounds of data from many disparate sources to discover patterns that could be useful in problem solving. These sources typically include clinical, financial and operational data and often work in the cloud as well. Much of it is designed to allow patient interaction by taking a proactive, or preventative, approach. Clinical data is typically normalized and validated from across the continuum of care to often include not only medications, lab results, vital signs, demographics, hospitalizations and outpatient visits, but also physician notes and lab results, taking advantage of both structured and unstructured data.
One of the more interesting areas where big data is used is pediatric cardiology, where analytics are applied to make patient-specific recommendations for treatment. The Pediatric Cardiac Critical Care Consortium (PC4) uses big data to try and improve the quality of care by collecting data on clinical practice and outcomes from each patient’s medical record and analyzing the data to provide clinicians with timely performance feedback. This fosters a culture of continuous improvement through analytics and collaborative learning. This disease-specific registry is also essential as we move towards a value-based healthcare system.
Big data in radiology is more about decision support than anything else and plays an important role in defining the way radiologists use clinical decision support systems to assist them in reading images. According to a recent survey, nearly 89% of radiologists said they always use the clinical decision support software computer-aided diagnosis (CAD), yet only 2 percent said they often change their interpretation based on CAD. The confidence simply isn’t there because the universe, while seemingly large, isn’t nearly as large as it needs to be to instill the degree of confidence required by radiologists. Rather than relying on individual studies, each clinical course and data would be saved and made available for decision support, capturing not just the data in a patients in electronic medical records but their radiology data, as well. These large data sets could be used in the future in clinical decision support systems like CAD to study patients with similar characteristics and to calculate the likelihoods of malignancies and other diseases. Putting these in a format that allows for data mining requires some additional coding and tagging but in the long run it will make the data easier to organize and search through and should improve both radiology diagnoses and ultimately patient outcomes through improved diagnostic capability – all made possible through the use of big data.