Last night, Martin Zand, M.D., Ph.D., director of the Rochester Center for Health Informatics and co-director of the CTSI discussed the future of big data in health care at the Simon Business School New York City Seminar Series.
Big data, defined as too large or complex to be captured, analyzed, or stored by conventional data processing methods, has captured the world’s attention in recent years. The inception of many new and inexpensive ways to collect large amounts of data (think Fitbit® and genome sequencing), has unlocked new and boundless potential to inform practices in business, government, health care, and beyond.
When it comes to health care and medical research, data abounds. Data is collected in real time by medical devices, such as EEG, and new technological devices, such as consumer fitness monitors or wearable health monitors. Millions of clinical images, laboratory results, and electronic medical records are produced every day. Massive amounts of de-identified patient data are also available from national registries and Medicare. The human genome is itself a treasure trove of information and the Precision Medicine Initiative Cohort aims to collect and analyze one million genome sequences. All of these examples show that the US has numerous resources and a preponderance of big data that could be used to improve health care in the US … and it sure could use some improvement.
It’s no secret that the US overpays and under performs in the health care sector compared to other nations. We pump much more money – most of which comes from private sources – into health care than any other nation in the world and yet we have shorter life spans and higher infant mortality than countries who spend less on health care. According to The Commonwealth Fund Executive Rankings from 2014, although the US health care system does well in terms of safety and overall efficacy of medical care, it ranks poorly overall and in terms of health care efficiency, equity, and in the health of our daily lives.
However, the boundless potential offered by big health care data will remain untapped if data analysis methods fail to keep pace with data collection. In his talk, Zand suggested that we need to remove obstacles of data sharing and use, improve data visualization and train researchers and health care professionals on how to collect and handle data. Then we can use this data to figure out who is doing things well – in terms of health care data analysis and health care outcomes – and emulate them. Using big data, we could predict when and where adverse events might occur, and intervene before they happen. Similarly, we could predict risk of disease in patients from their genetic data and implement preventative medicine. Data science can also help us figure out why quality and cost of health care varies across the nation and how to standardize it, how to get health care to remote areas efficiently, and how to leverage networking between healthcare providers.
There are many issues in the US health care system that could be addressed using big data. Zand urges researchers to clearly identify issues needing intervention, try many different methods of analyzing the data, ensure the data is accurate and reliable, and to share their results. That is one promising path to transforming US health care in the new era of big data.