Healthcare is the ideal market for big data. Having the ability to assimilate and analyze vast quantities of information from various sources makes it possible to prevent epidemics, cure illness and generally improve patient care and quality of life. It also helps hospitals and care facilities deal with operational challenges and improve efficiency. However, as with any big data initiative, the trick is knowing how to ask the right questions and where to find the right data.The healthcare market continues to grow and now represents 17.6 percent of the nation’s GDP, or about $600 billion. Trends in treatment have evolved to drive revenue, and treatment payments have migrated from pay-for-treatment, where providers are rewarded for treating more patients, to risk-sharing, where providers pay fees for patient outcomes. The move to reward wellness rather than treatment is being driven partly by analytics and the ability to generate reliable metrics for treatment effectiveness.
Concerns about patient confidentiality have slowed adoption of big data for healthcare, but the proven value of analytics is now starting to drive faster big data adoption.
Know Your Use Case
Because big data has so many applications in healthcare, you have to begin with a well-conceived use case. The technology can play a role in multiple areas , including:
Patient care – Using big data to analyze patient information can improve care. In addition to patient records such as medical history, family history, allergies and eating habits, caregivers are using monitoring tools to assess patient health, including new wearables. Even data from a Fitbit or smartwatch can be incorporated into analytics to get a complete patient picture. That information can then be compared to other data sets to correlate symptoms and suggest treatment options. Wellness treatment and preventative care are also becoming applications for big data analytics.
Disease management – Big data is ideal for epidemiology and assessing disease risk and spread. The Centers for Disease Control (CDC), for example, uses disease mapping and real-time analytics in order to understand disease patterns and manage outbreaks. Data can be used on a local level as well. For example, a number of hospitals are finding big data invaluable for reducing patient readmissions.
Integrated records management – Integrating electronic health records (EHRs) can improve both patient care and hospital management. Although EHRs are now mandatory, very few hospitals and care providers actually integrate EHR files. By creating a centralized picture of patient records, you can assess trends in patient care as well as in hospital caregiving and management.
Healthcare operations – Big data analysis also can streamline operations. For example, hospitals have large quantities of biohazard materials and disposables that need to be tracked and managed, both for safety and to ensure inventory is at appropriate supply levels. Using sensors provides real-time information about pharmaceuticals and inventory levels in order to optimize ordering and automate inventory management.
Insurance fraud – Security and fraud detection continue to be among the biggest applications for big data. Using correlated data, including EHRs, insurance underwriters can identify fraud such as false claims, overbilling, Medicare fraud and fraudulent prescriptions. Big data also can be applied to simplify billing strategies and insurance procedures as well as improve their accuracy.
These are just a handful of big data use cases for healthcare. When working with healthcare customers to understand their big data needs, be sure to align use cases with available data sets. For example, EHRs have to be integrated before they can be used for analytics. You also want to see which outside data resources could be useful for analysis.
The Elements of Big Data Success
The benefits that healthcare providers can realize from big data analytics can be substantial, but to maximize results, you have to have the right elements. Here are four tips that you need for big data success:
1. Pick the right people – Be sure you have the right personnel in place for any big data project, including in-house and consulting personnel. From the solution-provider side, you want a team that understands both big data and healthcare, including the basics of healthcare management, patient care and regulations such as HIPAA and the HITECH Act. The in-house team should include stakeholders who can articulate big data objectives and identify available resources so that the consulting team can design the appropriate use case and develop the appropriate analytical models.
2. Pick the right use case – Set realistic parameters for the use case. Many big data projects fail because they didn’t define the use case appropriately. Identify closely defined use-case objectives as well as data sources. Invest in discovery analysis to see if you need to gather additional data or refine use-case assumptions. It’s always best to start with smaller use cases that deliver focused results and then build on the proof of concept.
3. Pick the right data sets – Along with setting the right parameters for the use case, you also want to make sure you have access to the right data sets. Use all available data, including cleansed data archived in the data warehouse and unstructured data such as test, equipment sensors and Web logs.
4. Pick the right infrastructure – For most big data initiatives, it’s best to build out rather than up. Make optimal use of available services and enterprise resources, but consider cloud options that offer extensible computing and data storage capacity. As part of the infrastructure assessment, be sure to take into account security and regulatory compliance concerns to protect patient and healthcare data.
Collaboration is the key to big data success, and by working closely with healthcare customers in order to define their objectives, data assets and computing resources, you can work together to uncover insights guaranteed to improve operations and, more important, patient care.