Three Barriers to Effectively Using Information Stored in EHRs
The healthcare industry won't realize the full value of its investment in electronic health records until it finds secondary uses for all of the data being captured, such as predicting public health trends and improving patient care, according to a report by PricewaterhouseCoopers Health Industries Group.
Seventy-six percent of the more than 700 healthcare executives surveyed in June 2009 said that the information gathered in EHRs will be their organization's biggest asset in the next five years. But very few healthcare organizations are building systems and care delivery processes to effectively use the billions of gigabytes of data being collected.
"I'm surprised that more thought hasn't been given to the broader idea of using the clinical and administrative data to do continued improvement and process improvement in the industry," says Dan Garrett, head of the health IT practice at PricewaterhouseCoopers. "People are so busy doing the basic digitization of the whole industry that they haven't had time to think through what they will do with all of this data, and so it has not been taken into consideration in the deployment of some of these larger systems."
Healthcare executives should be thinking beyond implementing EHRs to how they want to use this data after the technology is in place. "If you know that you are going to try and aggregate the data and make statistical sense out of it, you are going to do it in a very different way than if you are designing a transactional CPOE," explains Garrett.
Unfortunately, there is no industry road map to follow. Instead healthcare providers are faced with three primary obstacles to the secondary use of data.
1. Data quality. Is there enough volume, depth, and breadth of data to produce statistically relevant information? For example, data that is aggregated from transactional-based systems like bill collection may not include clinically relevant information. In healthcare, after a bill is collected, "it is like that data is pushed off a cliff," says Garrett. He adds that in other industries—financial and hospitality—that doesn't happen. If people knew that data would be used by another physician, medical center, or research organization, systems would be designed differently and information would be captured differently, he says.
2. Workflow. Deciding when to physically interject that information into the clinical or administrative process is challenging, because it should be done at a point where it is relevant, the user can absorb it, and the user can take the appropriate action. "You are looking for impact, so you have to do it at the right point of the administrative and care delivery process," Garrett says.
3. Legal and policy concerns. Organizations are still struggling to determine when they can use this data and what are the liabilities associated with aggregated de-identified data. What if the data can be re-identified and traced back to patients? When can patients opt in and out?