I recently came across a study from Chilmark Research that highlighted an important, ongoing challenge in the world of improvement – incorporating analytics into existing workflows. Essentially, the study suggests that a health care organization’s needs have gone beyond the analytic products on the market. This isn’t a call for better or more analytic products; rather, it’s a decree that more is needed for an improvement than just analytic tools. As a proud “data geek," this resonated with me because I’ve always believed there’s a difference between analytics and using the information to accelerate improvement.
Don Berwick, one of the most influential health care leaders of our time has said that the biggest issue we have in the U.S. health care system is our inability to improve. This is not due to lack of data or analytic products. The amount of data available in health care is overwhelming: administrative data, electronic health records, claims data, clinical data and patient behavior data. Adding analytics to these sources transforms the data into information, which used to be thought of as the necessary lever to motivate change in processes and drive improvement. This has turned out not to be enough. Instead, that information needs context and understanding to transform it into insights. While analytics has been driven by technology, adding context can only be accomplished by people, which have been termed data scientists.
An October 2012 Harvard Business Review article describes the data scientist as bringing expertise, curiosity and polish to the understanding and presentation of data, information and insights. The article stresses that data scientists are difficult to find because there are very few academic programs available in the area. In addition, these academic programs may not have the correct focus. Finally, the article suggests that healthcare is far behind other industries in the use of data scientists.
Too many hospitals are only hiring analysts and statisticians to build reports, predictive models and dashboards. While needed, these aren’t data scientists.
Consider these two outcomes at Vizient member hospitals:
Hospital A, a large academic medical center in an urban northeast setting, set up a data science department and staffed it with statisticians from the aircraft industry. Their role is to meet with and listen to clinicians about what is wrong with the healthcare experience, and then design ‘fixes’ to those issues. Seven months into the new department, no measureable improvements have been recorded.
Hospital B, a large academic medical center in an urban central U.S. setting, took a different approach. They hired a well-respected, data-savvy physician to work with two statisticians in a newly formed data science department. This group has experienced tremendous improvement, vaulting this hospital into the top hospitals in the Quality & Accountability Study, where Vizient ranks academic medical centers in mainly quality metrics.
The improvements Hospital B was able to realize clearly demonstrate that organizations should be looking for data-savvy, seasoned leaders who have the capacity to develop excellent relationships and a strong knowledge of the healthcare industry.
It’s time that health care follows other industries and realizes the importance of the data scientist. This can be done without adding FTEs by ensuring that the chief quality officer fits the data scientist profile. If the current CQO does not fit the profile, get them the necessary training and skills they need to be your organization’s main motivator for improvement.
If you’re interested in providing existing staff with competency-based training around synthesizing data and framing critical issues, there are certification programs available. Learn more.
