To some extent, AI seems magical. Just think about its use cases in healthcare: It can analyze huge datasets of patient information to identify trends, which can be studied and addressed to help improve the health of different communities and populations. It can record and provide summaries of doctor/patient interactions, giving those doctors back valuable time for delivering care. And it’s already been used for years by IT teams to support and speed up back-end operations.
But let’s be honest, friends: For all AI’s wizardry, it’s also a buzzword. Though the future of AI in healthcare seems bright, and it feels like everybody wants to say they’re doing something with it, that “something” is not always easy to define. Truly new innovations are also rare. In many cases, the hype cycle and bandwagon become too hard to ignore, and FOMO sets in.
At Olah, we’re starting to thoughtfully explore AI’s applications for data management and archiving. We’ve examined a lot of the technology’s benefits and limitations, and we’ve seen many — and we mean MANY — companies discussing both on the conference circuit. But we feel that the cornerstones of success for rolling out AI-backed technology are following a strategic planning process and keeping a focus on putting your customers’ most critical needs first.
AI Data Management for Healthcare: Charting New Territory
As a tool to support data archiving, AI is still in its infancy. For all the broader industry talk, AI data management solutions are like a mirage on the horizon. We do believe there are excellent use cases for AI, however. It holds the potential to increase the efficiency of the overall archiving process, making archiving faster and potentially more cost-effective for customers as they transition to modern platforms and retire their legacy systems.
A greater possible benefit is close-to-real-time access to data. AI could help clinicians view information from the past one to two years in an easier and more digestible way so they can better understand the patient experience and act upon what they see, asking questions and getting responses back from the people they serve. This ability to surface data that the user needs could also speed up the process Olah uses to migrate a customer’s database to our , by reducing the manual labor of finding the right information to create reports. Other advantages might include being better able to understand trends in data that’s stored in different ways on multiple databases and tapping into data from smaller systems that gets lost during acquisitions.
But it's important to consider both the pros and cons of AI in healthcare. While it can do a lot, it can also hallucinate false or misleading information. You need to consider patient data security and safety during data transfer, ensure outputs are accurate, watch out for bias in the gathering and assessment of data, and, most of all, earn end-users' trust. AI is a big topic, and it can be a scary one, so it’s crucial to get to a certain comfort level with customers around the tools you offer them.
Olah’s Approach: Slow and Steady Strategic Planning
The good news with AI is you can always keep building off it. If you shoot for something really big, though, you have a good chance of missing the mark. So, it's better to take incremental steps in your explorations of AI-backed technologies while the future of AI in healthcare is still being decided.
Part of our approach is to evaluate those steps that make the most sense for us organizationally as we test this technology out, and then build the in-house knowledge that we need. Efficiency gains are one of the first building blocks we’re addressing, and later we’ll look at bigger-picture concerns like clinical trust.
We’re also planning to investigate sophisticated off-the-shelf tools that already exist. We’ll assess how to use them creatively and integrate them into EAS in interesting and unique ways, like using them to move data more easily between systems without a lot of mapping.
Enhancing our platform with AI data management rather than trying to reinvent the wheel is a strategic business decision — one that will afford us the ability to innovate while also avoiding the potentially costly investment that AI software development requires, including training, maintaining data models, and adding expert resources. At a time when the AI landscape is changing rapidly, it’s also a smart move to avoid overinvesting in solutions that could lead to minimal gains.
Advice for Healthcare Organizations Considering AI Initiatives
We try to drive a lot of our road map based on customer requests — but not necessarily what they’re asking for today. At Olah, we’re proactive and consider what they may need in a year or two years’ time.
Because the future of AI in healthcare seems bright — and the capabilities of AI-backed tools are likely to influence what our customers ask for — we’re committed to being prepared to execute with that knowledge and have plans in place to act.
Along this journey, we’ve picked up several lessons and learned just as much about the cons of AI in healthcare as the pros. Here’s our advice for others just starting out:
- Get a partner and have conversations with other organizations. Olah has learned a lot about AI implementation from industry expert Gartner in one-on-one consultations, and we've also searched out peer insights. Take advantage of webinars and forums from industry groups like HIMSS and Chime. Search out advice from CIOs experienced in the space and see what learning curves other organizations have faced in industries outside healthcare. Don’t go it alone and do have an expert by your side.
- Use existing tools. There are at least four barriers to successful AI implementation. Cost, as we’ve discussed, is one. You also need a lot of data and expertise that takes time and effort to secure, and you must comply with governance boards around how your technology will be used. These collectively can distract you from more important organizational priorities. Fortunately, many of these problems are mitigated by leveraging existing AI tools — or even better, AI built into your existing products.
- Don’t get sucked down the rabbit hole. Though it might be tempting to play the role of Alice and try experimenting with every AI tool available to see how each one enhances your offerings, take a beat first. Evaluate if a non-AI tool can meet your users’ most pressing needs. If you do need an AI solution, take an iterative approach. Map out use cases with success criteria for each. If the AI’s capabilities don’t measure up, reevaluate your approach and pivot to what makes more sense.
- Ensure your team’s AI knowledge is socialized. There’s a good chance that few people at your organization truly know what AI is, does, and is capable of. Sure, they may have an intuitive feeling for how it can be beneficial, but that’s not enough to implement a significant change in your business operations. Be ready to educate teams and elevate one key executive who is fully bought in to any AI efforts you undertake.
Balancing AI’s Potential With Practicality in Healthcare Strategy
There is immense promise for the future of AI in healthcare, but just as immense is the hype. It’s crucial to weigh whether it should be an organizational priority and set proper expectations for its internal and external use in your healthcare organization’s operations and offerings, while grounding any strategic planning in what your customers need from you today and in the future.
At Olah, our customers have two primary needs: sunsetting legacy systems and archiving vast stores of data so providers can deliver safer care, healthcare organizations remain in compliance with regulatory requirements, and systems are protected against cyberattacks on aging technology. We strive to help them with the most cost-effective and advanced approaches possible and are excited to see where AI and other innovations take us in the future. If you’re interested in further thoughtful conversations and want to see EAS in action, schedule a demo today.
 |
Nick Anderson |
 |
Mick Blackwell |
