AI Can Help Labs Manage Data to Improve Stewardship

Stewardship is defined as “the careful and responsible management of something entrusted to one's care.” In health care, laboratory administrators are frontline stewards, providing quality, appropriate laboratory services to guide patient care.  

But laboratory administrators are increasingly in charge of a different kind of stewardship—managing the cost of laboratory spend and achievement of financial incentives for meeting lab-based quality metrics. That’s because laboratory testing drives most health care decisions, with significant downstream consequences on patient outcomes and costs. 

In recent years, laboratory stewardship programs have become a significant focus of hospitals and health systems seeking to improve care, lower risk of patient harms, and drive superior financial performance. These efforts typically focus on identifying inappropriate testing patterns, such as duplicate test orders or failure to order guideline-recommended tests. 

To be successful, lab stewardship programs must analyze current, accurate lab data. And that’s where many lab stewardship programs fall apart. Ordering of lab tests is health care's single highest-volume activity, creating a stream of data that must be curated, normalized, and tracked. Few lab directors have access to the technologies or human power to manage and glean actionable insights from such large quantities of disparate data. 

New artificial intelligence technology is enabling lab directors to manage data with greater effectiveness and efficiency. As lab administrators consider their options, an understanding of basic technologies is essential: 

• Artificial intelligence: AI, also known as cognitive computing, is making notable inroads in health care, and lab administration is no exception. For instance, machine learning, which is a subset of AI, can be used to train algorithms to make decisions about laboratory test names. This is incredibly useful to lab stewardship programs involving multiple lab compendiums. In these cases, ML can “learn” how to normalize differing naming conventions given to the same test across multiple laboratories, improving accuracy of data analysis. 
• The Cloud. Cloud computing means that data from remote servers is accessible securely through the internet. Services that allow staff to access data from diverse lab information systems and electronic medical records in a secure, online environment is critical for managing stewardship programs, particularly those involving multiple inpatient and outpatient labs or sites. The cloud allows for a bird’s eye view, so the administrator can identify and act on the biggest problem areas across their health care system. 
• Graphics: When it comes to lab stewardship, a picture really does tell a story. Look for platforms that present lab ordering trends in graphic, intuitive interfaces. A good interface will allow the user to quickly identify problems, such as sharp ordering spikes in a type of lab service, so remedial actions can be taken quickly. 

Lab stewardship is only as good as the technologies that enable it. AI, cloud-based computing, and graphic-rich analysis are hallmarks of any lab stewardship program that aims to improve patient care and lower costs.