More Patient Information Can Lead To Better Care. Here's How To Get It Without Paying Too Much.

By Vinay Seth Mohta, Manifold

Healthcare providers have historically gathered only clinical data, collecting information when patients needed health services. But collecting information across a broader spectrum provides a more holistic view of a patient and can improve strategies for providing care. Moreover, it can be an incredibly useful indicator of a patient’s long-term health.

One poignant illustration of this is a story that was shared a few years ago by the CEO of a healthcare IT company at a panel discussion. It was about a mother who was reducing her child’s asthma medication in August. It turned out the mother was trying to save money for back-to-school supplies. Traditionally, care providers wouldn’t have been able to figure this out based on their limited data. But because more data gave them a wider field of view, they were able to intervene before an attempt at saving money led to increased costs — and potential health problems — down the road.

More healthcare organizations are taking on “fee for value” payment contracts. In fact, 90 percent of Medicare fee-for-service payments are expected to be tied to quality by the end of this year. This approach provides incentive to care for patients by intervening before major health crises occur, which saves money in the long run. Data is a key ingredient when it comes to identifying which individuals need help, so collecting background and lifestyle information can improve care.

Still, there are numerous obstacles to this approach. The prevalence of data breaches among healthcare organizations has aggravated concerns about data privacy, and many of these organizations lack the infrastructure or expertise to manage a large collection of data sets. Plus, aggregating and cleaning data is a prerequisite to any analytics efforts.

Gathering external data to add to your own healthcare data platform can be an expensive proposition. Fortunately, healthcare IT leaders can adhere to these five strategies to minimize costs and maximize results:

1. Start with a well-defined use case. Avoid the temptation to build an AI platform that can handle any data set. Ideally, you should find an appropriate partner such as a care manager, a healthcare provider, or another leader in the operational side of the organization to identify a compelling use case. One of the most common opportunities to achieve ROI centers around physician and nurse staffing. A system that can analyze factors such as weather, vacation schedules, maternity and paternity leave, and graduation rates from medical schools can optimize staffing and meet needs without understaffing or overstaffing.
2. Work with experts to identify risks early. In large health systems, internal expertise is everywhere. Still, it’s often spread out across the organization and can be hard to find. Reach out to clinical or academic researchers who may have already built data infrastructure around the questions you’re pursuing — they will be acquainted with the risks that you’ll need to mitigate. If you can’t find what you’re looking for, searching resources like PubMed can yield useful results, or you can hire third-party experts to identify risks. The Food and Drug Administration also launched a preparedness and response playbook that can help you address any threats related to medical device cybersecurity.
3. Explore publicly available software and data sets. Healthcare researchers have developed a number of open-source software packages, which are a great source of knowledge on working with various data sets. Combine the software with publicly available data sets, which could be from the Centers for Medicare & Medicaid Services, the Centers for Disease Control and Prevention, and others. Wherever you look, you’ll find communities of users who have worked with these types of data, and some of them will have published research or shared their own findings in public forums. Learn from these collective experiences, and you’ll be well on your way to achieving your data-processing goals and delivering a better patient experience.
4. Coordinate with other local organizations. Data collection and analysis can be a costly undertaking, but collaborating with healthcare organizations locally can help. You might find that other forward-thinking organizations have already begun collecting the data you need; you can partner with these providers to spread out the burden of cost and allow all participants to get the data they need to improve patient care. Other organizations could also have valuable information about the requirements of legal and regulatory frameworks with respect to aggregated data sets.
5. Purchase a software solution. Some software vendors and cloud providers offer solutions for aggregating and cleansing data. Large data providers have joined forces with software platforms to facilitate the use of their external data with other data sets. Watson Health is a well-known example, and IBM purchased Truven, the Thomson Reuters health data division, to expand the scope of its services. The Google Cloud Platform is also engaging healthcare delivery organizations, and the company has experts who can help integrate, aggregate, cleanse, and prepare data for analytics and machine learning use cases on their platform. The two other major cloud providers, Microsoft Azure and Amazon Web Services, are making similar investments in healthcare.

Data always has been an important part of healthcare, but providers are increasingly learning how to leverage it to deliver a better patient experience. As you undertake this transformation in your own organization, focus on these five strategies to keep costs low without compromising results. You’re not the first to walk this path, so learn from the mistakes of those who have gone before you, and you’ll end up with a better product.

About The Author

Vinay Seth Mohta is CEO at Manifold, an artificial intelligence product development studio with offices in Boston and Silicon Valley.