NTSB was launched in 1926, when the automotive industry was in its infancy and few people thought about all the fatalities that lie ahead. Over the years, it has been responsible for insisting on speed limits, seatbelts, airbags, and a host of other invaluable safety measures.
An NTSB-like organization for EHRs would at the very least provide a reporting mechanism to keep track of incidents and life-threatening consequences of misusing e-records. More importantly, it could police vendors and healthcare providers who repeatedly ignore these dangers.
During his presentation, Sittig was quick to point at that errors can occur at any stage in the life cycle of an EHR, during design and development, implementation, or post-installment evaluation. He went on to describe several areas in which one can expect to see problems, which are not limited to computer glitches. Many problems are due to the way humans interact with machines. But unless we have a NTSB-like system to track all the errors and categorize them, EHR designers and end users will have a much harder time systematically fixing them.
And statistics suggest we are doing a poor job of tracking these snafus without some sort of government or industry oversight. Over the last 18 years, for instance, the FDA has received reports of only 120 EHR-specific errors. There is no way that number can be accurate; it only tells us that such errors are grossly underreported.
One problem that needs to be tracked is missing or inaccurate clinical content within an EHR. Probably the most notorious mistake that Sittig mentioned occurred because of a mapping error in England's National Health Service. It resulted in patients who were supposed to get a prescription for Zyban, a medication to help them stop smoking, being prescribed Viagra instead.
Other reports suggest that even when an EHR churns out correct information, it's not always reaching the clinicians who need to know, or they are simply not acting on the information. In one study, of the 1,163 lab results that were sent out as electronic alerts, about one of every 10 were ignored by physicians. Worse yet, in a separate investigation, 7-8% of abnormal imaging results were never followed up by physicians.
Another pain point for IT managers and clinicians: EHR-generated alerts sometimes have unintended consequences. An experiment done at the University of Pennsylvania showed that a hard-stop alert can reduce the likelihood of ordering dangerous meds. But the investigators also found that in some cases, these roadblocks actually resulted in dangerous delays in ordering meds for some patients who really needed them.
These are only a few of the snafus one can expect to see when computers and people interact, which is all the more reason for IT departments to thoroughly measure and monitor the performance of their e-record systems. Our EHRs need seatbelts, speed limits, airbags, and a variety of still unimagined features if we ever hope to safely navigate this new highway.
As IT pervades every aspect of healthcare, complying with federal regulations on patient privacy and security is becoming increasingly difficult. Our webcast will help attendees navigate the latest HIPAA privacy and security requirements, proposals, and penalties and provide insight into what healthcare providers need to do to be compliant. It happens Oct. 27. Register now. (Free registration required.)
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?