You're starting up a new healthcare-related business. You've done your research, and you know exactly what you need to do to get the ball rolling. You want to make sure that your business is up-to-date with the latest, well, everything. You've got some fine talent onboard and now they need be fed with that most precious repast: Data.
Where does all the healthcare provider data come from? Yes, frankly, most of it is "free." But it's in quotes because time – the time it takes to find and process unfamiliar data – isn't free. You need healthcare providers' information for your application, and you've got some ideas about where to get it. You want free stuff as much as possible, so you've surfed the Medicare sites, climbed the Department of Commerce mountains, tumbled into state certification board hollers, knocked on the unyielding doors of medical societies. It is a daunting task just locating the stuff, and you feel like you've just scratched the surface. Then you have to evaluate the data you've found, and figure out which components you'll have to put together to suit your particular requirements. There are hundreds of millions of healthcare provider records out there, in hundreds of datasets from dozens of data providers, both free and fee.
But reinventing the data-sourcing wheel is hardly the innovation you're aiming for, and it's an unnecessary limitation on your progress with a new development project.
That first step is a tall one.
So skip it.
Jump ahead to one of the handful of truly good healthcare provider data suppliers. These are the companies who are doing data search and integration full-time. It won't take long to survey the field. Five vendors' names come up immediately: Definitive Healthcare, IQVIA (formerly SK&A), LexisNexis, the AMA, and CarePrecise. Now it's just a matter of chatting with a data analyst at each to figure out if they have what you need (they all probably do), how they deliver the data (this varies) and what it will cost (boy, does this vary).
Depending on your specific needs, four out of five of these vendors may cost an order of magnitude more than one of them. You'll find out soon enough which is which, but there are other questions you'll want answers to.
Reliability and Permanence
Which vendor and products can be expected to remain in the same format, and offer the same accuracy and freshness, for the long haul? Ask about backward compatibility; as a data product grows over the years, will it remain compatible with my internal processes, or will I have to spend ongoing time and effort adjusting to changes, either in data layout or in the way an API operates? And, of course, you wonder how long will this company remain in business and continue offering the components you need.
Over the 15 years that CarePrecise has been working with clients involved in development of entrepreneurial products, we have learned that the shorter the development time frame, and the fewer special competencies are required, the faster a product gets to market and reaches its earnings objectives. Obviously. Less obvious are the special gotchas of developing a data-based product. Make sure that whatever data package you choose will come to you in by far the most useful of data formats: a relational database. Whatever database management system you'll be using, now or in the future, one data format is sure to be supported. Relational tables, each packaged as a separate CSV file, is far and away the most reliable and long-term successful format. But whatever format you plan to use, make sure your data supplier can accommodate you.
For developers new to healthcare, provider data can be confusing, at least until you've taken a long leisurely hike through a data portal or software product that makes it clearer. For example, an Access database or other visualization front end where you can play with the data, learning the relationships. CarePrecise offers just such a "toy" for learning the structures of U.S. healthcare provider data, while performing real work, such as creating lists of providers by specialty, or by geographic radius. An hour of playing with a tool like this can be all the introduction you need.
If your application will be accessing the data through an API, you'll have lots of questions, not least of which is uptime. You'll also want to know if the API is robust enough for the particular complexities of your use case. Or you could instead go for a data package that allows you to download the whole database and all of its components to your local environment – doable if the dataset is very cleanly and sensibly laid out). In this latter case, your application may benefit from greater responsiveness, and coders will be better able to wrap heads around the data universe.
Ongoing Data-related Costs
Will the cost of the data subscription and license be a burden down the road, once the initial cash burn has passed? Also, the data package needs to be straightforward and intuitive enough that new hires will be able to pick it up quickly, with clear documentation.
This is where a poor commitment to backward-compatibility on the part of your data supplier can create staffing strains, as the data team scrambles to accommodate some change in data layout. Or worse: A data product no longer contains what you needed! A painful example is the recent decision by CMS to drop hospital information (even its name and address!) from the Physician Compare data file, and to split it into two files in a relational format, where it had been a simple flat file. (CarePrecise customers experienced no outage from this event because it maintains alternative sources for this and similar drop outs.)
What happens when your needs change? If you have a "universal" license that covers all of a company's data offerings, you'll skip the step of having to negotiate a new license for a change in data subscription. Try to negotiate a single fee that permits you to add data components later on, paying only the data subscription fee, with no license upcharge.
A Growing Field of Healthcare Apps
Healthcare applications – both consumer-facing and B2B – are taking flight all across the industry. May the best innovators, with the best data, win!