AI and automation have tremendous potential to reduce the dollars spent on healthcare administration rather than patient care. But for every health system eager to start their AI journey in house, there are others discovering potential pitfalls of a do-it-yourself (DIY) AI strategy. Janus Health CEO Todd Doze recently sat down with Beth Friedman of FINN Partners to reminisce about how far healthcare information exchange has (and hasn’t) come. He offers some cautionary tales about DIYing complex, costly technology, particularly around:
- Investing in AI talent and infrastructure,
- Understanding the AI use case, and
- Committing to ongoing AI maintenance.
Investing in AI talent and infrastructure
Excepting some large health systems, providers are in the business of patient care, not software development. It takes substantial human and financial resources to enable successful, accurate AI model development and upkeep. Building and retaining a best-in-class engineering team — with healthcare expertise — is challenging enough for specialized software companies, let alone provider HR teams faced with overwhelming shortages in more traditional health system roles1. Add to that the complexities of maintaining distinct development, testing, and production environments, the costs of data processing and storage, and even the legal implications of potentially developing new intellectual property, and AI development is simply out of reach for many health systems.
Understanding the AI use case
When considering an AI solution, leaders must understand not only how they’ll use it, but how they’ll measure the value of it. Off-the-shelf AI solutions can accelerate initial development but configuring them for your unique workflows, technology stack, and payer mix may require more time and resources than anticipated. Too many DIY software projects end up 12 months out with no ROI to speak of, often a result of unclear requirements and use cases up front. A strategic AI partner can guide health systems to the right tools to tackle the right problems the first time.
Related reading: Operational improvement solutions help revenue cycle leaders recognize and prioritize what’s next.
Committing to ongoing AI maintenance
A long-term AI vision and investment strategy must include provisions for maintenance. AI models change, the systems they integrate with are in constant flux, regulations come and go, and security requirements evolve with each new threat. Without ongoing maintenance investment, health systems will at best be saddled with resource-draining technical debt, and at worst inaccurate or even dangerous AI models. It makes sense for many providers to outsource this responsibility to an AI partner dedicated to staying ahead of technology and industry changes.
AI partnership enables possibilities
Prioritizing AI partnerships over DIY is often the best strategy for health systems, especially if they don’t already have significant IT infrastructure and staff. By engaging an innovative partner and driving continuous improvement, providers can realize transformative changes in labor costs, denial reduction, write-off improvement, and the employee and patient experience.
Listen to the full podcast here.
- Do health systems have the IT talent to support AI rollouts? Becker’s Health IT
