| | December 20208UTLOOKHealthcare Tech Tools and Capabilities Required for Value-Based CareBy Mark Weisman, Chief Medical Information Officer, Peninsula Regional Medical Center"We are moving towards value-based care" is a great set of buzz words. We have all used them, especially if we need to impress the board. A line such as "We are positioning ourselves to be at the forefront of value based care through the utilization of artificial intelligence, telehealth and blockchain" contains enough buzz words in it to guarantee a promotion. Try it out and send me a thank you note when they make you CEO. However, the technology that is really important in value-based care is complicated and costly. Doesn't mean we shouldn't do it but the financial incentives are still very much in favor of fee for service in most parts of the country. I believe that is why we continue to see very few areas making a meaningful change in the cost of care. I am providing this list of tools required to perform in a value-based care model so there is an understanding about the upfront costs to jumping in. · Access to data: A well run Health Information Exchange (HIE) can be incredibly helpful, but doesn't solve all our interoperability issues. In your area, if the major lab and independent imaging centers are not on your EHR then there can be significant time and money needed for integration efforts. Then there are claims data and social determinants of health data to bring in. Managing patients without a more complete understanding of their conditions and utilization patterns is very challenging.· Data aggregation capabilities: If you are successful in getting access to all this data, it now has to be normalized and ideally collated into a meaningful package that a provider can quickly ingest. No duplicate patient records and you can't put data in to the wrong chart. Many health IT shops shy away from this activity because it is hard, can be high risk, and frequently requires human labor to verify the data machine learning is not confident in its assessment. Google showed a demo of the tools they are working on in response to the Ascension article in the Wall Street Journal and their capabilities are worth watching. The ability to get data out of pdf files and other unstructured parts of the medical record will be a huge leap forward but is not commonplace today. · Registries: To do effective population health analytics, it is most efficient to have all that data from #1 and #1 above populate registries to track the chronic conditions. Usually available in the EHR but most have to be configured and maintained. To effectively deploy clinical decision support in real time to providers it is necessary to know if the patient in front of you has the disease in question and that data needs to come from your registries. In theory, that data could come from the HIE but more commonly comes from your registries.· Data visualization: Now that you have your data nicely organized in registries, the end users need to gain insights. Most systems that are serious about population health have data analysts who are subject matter experts to drive these tools because they are not as intuitive as the vendors make it seem. Then you have to get these insights in front of clinicians to impact the delivery at the point of care. A common mistake is to underestimate the challenges of getting the data in front of the right people for action.· HCC coding tools: Health systems that are successful at value based care have figured out the HCC game. How to play this game is beyond the scope of this article, but very smart systems underperform simply because they are not capturing the severity of illness of their patient population. They appear to be spending more money than they should IN MY OPINION
< Page 7 | Page 9 >