Questions You Should Ask When Selecting a Healthcare Analytics Platform
|Click to View|
As vice president of technology for a healthcare IT company, I’m often asked what should be considered when selecting a solution for healthcare analytics. It starts with the data warehouse infrastructure which is the foundational level of the Healthcare Analytics Adoption Model and includes other criteria around implementation that can make or break your success.
Healthcare organizations have many choices when selecting a healthcare data warehouse and analytics platforms. I advise them to consider the following fundamental criteria:
Time-to-Value Needs to be Measured in Months, Not Years
Time-to-value is the most important factor in healthcare at this critical juncture. Gone are the days of one- to two-year data warehouse implementations. Most organizations have already been through a recent, lengthy EMR implementation and need to prove the value of their investment yesterday.
Choosing a solution that provides a rapid time-to-value is critical because there are cost reduction pressures that won’t wait for complex data integration strategies. They won’t wait for data to be modeled in an ‘enterprise model,’ or comprehensive vocabulary mappings before they can be addressed. All of the pressures healthcare organizations are facing require quick access to reliable data. Healthcare data warehouse and analytics platforms that can be implemented and show value in ninety days are available today. I advise healthcare organizations to put these high on their list for consideration. These platforms represent the lowest-risk solutions. Some vendors are even willing to offer no-risk contracts to show how confident they are in creating value quickly.
Experience and Proven Success Are Predictors of Your Success
Experience and proven success in data integration and analytics is of utmost importance. Look at core competencies:
- Does the vendor have a history of successfully delivering a data warehouse and healthcare analytics that actually improve care and reduce costs?
- Ask vendors what their track record is by asking for specific examples of how they have helped to improve care with analytics.
- Ask for references. There are some major vendors in the market who don’t have a single client willing to vouch for their data warehouse and healthcare analytics platforms. The only value they have is a big name that everyone recognizes (and assumes they can’t get fired for choosing).
It can be tempting to go with an incrementally low-cost healthcare analytics solution that comes packaged with your EMR or other transactional system.
EMR vendors have a core competency – they write software that collects data, provides data entry interfaces, and looks at data on a patient-level. They generally do this very well.
But how can one expect these vendors to have the experience necessary to build successful healthcare data warehouse and healthcare analytics platforms? Would you ask a data warehouse and analytics vendor to build an EMR? Heck no!
Extensibility to Meet Your Needs
Is the solution extensible? The current regulatory environment is constantly changing and adding demands on healthcare data collection. Does the vendor (or better yet, a solution partner) understand the nuances of clinical data and regulatory requirements?
Any vendor who claims they have an out-of-the-box solution that can determine whether VTE prophylaxis has been administered has probably never sat down with a clinician to discuss the six or seven different ways this can be documented and the dark recesses of the medical record where this documentation may live. A vendor who talks to you about ‘fingerprinting’ and ‘clinical engagement’ is probably the vendor you want to be talking with.
Fingerprinting is the process of tailoring healthcare analytics to the nuances of your data. Clinical engagement is how you get to successful fingerprinting.
For example, I once worked on a project where we were first tasked with identifying a clinical cohort for asthma. There was quite a bit of debate about the cohort definition and some back and forth between the technical and clinical team members. When we were finished, one of the physicians took out $10 and said ‘I bet we have the best acute asthma cohort definition in the country.” We didn’t get him to put up 10 bucks by bringing in a pre-defined cohort definition; we did it by listening to the clinicians tell us how their patients are coded, what medications they are treated with, and validating that cohort with them. It’s a good sign if your vendor provides solutions with the data warehouse; an even better sign if they ensure the solutions work with your data. There is a lot of work that needs to be done to standardize data collection across different organizations. Until we get there, ‘black boxes’ don’t work in healthcare.
Don’t Ask For a Data Model. Ask For Results!
I get asked if there are additional selection criteria to consider. The three we’ve talked about are the primary ones to consider. Cost will always be part of your consideration set. However, cost alone can be the cause of poor decisions if it is not balanced with fundamentals. If you look at the three fundamentals — time-to-value, experience and extensibility — you’ll take into account the criteria that will lead you to the right vendor. Don’t ask for a data model. Ask for results!
What criteria have you used for selecting a healthcare data warehouse and analytics platform?