In any organization, no matter what sector – manufacturing, supply chain, or even healthcare – the conversation around whether to invest in building or buying solutions is always a common topic. IMO’s Dr. Bhalla breaks down what factors into making this decision.
Outstanding questions when building in-house health IT solutions
When considering the creation of a solution in-house, there are numerous questions to ask:
What architecture design pattern should we go with, and which cloud vendor should we use to host? Is this a new product or a platform to build multiple products? How will the activities impact our roadmaps, and will they directly contribute to our return on investment? Who will develop and maintain it, and whom do I call when something breaks? Do we have the subject matter expertise in-house, or will we have to go on a hiring spree?
There are so many things to consider that, often, the decision-maker will take a break and do something else only to begin ruminating again.
Certainly, that last point is pivotal: Do we have the subject matter expertise in-house, or will we have to go on a hiring spree? In the health IT sector, the cost of entry is exceptionally high due to the nature of the data. It is constrained by rules and regulations and, combined with highly opinionated data, encapsulated in systems and processes which only folks with subject matter expertise know how to utilize accurately and effectively. But who are these experts?
At the center of every HIT organization is a group of people known as informaticists – and yes, there is a spectrum of them in practice.
Some have specialized in bioinformatics, which is more on the genomic side of the house, while others have specialized in larger, more population-based approaches. Some have specialized in specific domains, such as oncology, neurology, or workflow management. Some are hardcore engineers, and others are masters of data analytics, data harmonization, and building machine learning models.
There is a spectrum of skillsets and domain expertise. So, when I am asked, “How much does it cost to build a team of informaticists?” The answer is that it depends.
Calculating the cost
It would be helpful to have precise data from all the primary job and salary websites regarding the total number of individuals trained in healthcare informatics and how much they make. There are not many of us, and only a few with the skillsets to build an enterprise-grade, production-ready solution that can scale with an organization’s needs to unlock the value of healthcare data.
Let’s break down the process with an organizational use case that you can build a portfolio of products upon.
Imagine you are a “real world data (RWD) / real world evidence (RWE) organization that wants to build a data platform to aggregate disparate data sources from which you’ll derive novel insights.” For this business use case, you are going to need several groups:
- A group of informaticists who know the interoperability of clinical data to tackle the “disparate data.”
- A group of informaticists who specialize in the alignment of data to tackle the issue of developing a data model and building out the ETL logic to support the ingestion and utilization of the data.
- A group of informaticists specializing in the semantic alignment of data to ensure that all the data sources are harmonized correctly.
These are just the foundational informatics groups, not those who are required to build the platform business use cases, the platform, and support those teams. Don’t forget the other folks, too, such as architects, full stack engineers, cloud engineers, service reliability engineers, DevOps, people managers, program and project managers, and the people team supporting their emotional needs if you choose “to build.”
Keep in mind the costs of all the required tooling, such as team collaboration tools, project management tools, SDLC tools, mapping tools, document repositories, marketing repositories, and everything in between. I could do a deep dive into the number of processes and workflows you will need to develop and maintain to cover the above operations, but I think you can figure that out on your own.
Lastly, if your business also leverages unstructured data, you can add the need for those rare types of data science informaticists along with a complete data science team. Data science is trendy, and finding data scientists who know healthcare…well, that topic is best saved for another day.
But wait, there’s more.
Sadly, this is all before you even begin the second part of the use case, “derive novel insights.” With the second component, you will need domain-specific informaticists – bioinformaticists, population health informaticists, regulatory informaticists, the elusive pharmacist informaticists, and everything in between to support your overall organizational needs with the other individuals previously mentioned. The more specialized these resources are, the more expensive they become, and the harder it is to keep them motivated and engaged.
Finally, one component that people often fail to mention is time. Nothing in health IT is quick and easy, hence the scarcity of successful HIT companies out there. This is why partnering with third party solutions, who already have proven competencies in clinical informatics and terminology alignment, may be more strategic (and less headache-inducing) than choosing to build your own.