Technology Alone Can't Improve Patient Care

Little is known, at this point, what technical components a physician's EHR system will need in order to qualify for a portion of the $17.2 billion in incentives set aside in the recent economic stimulus bill, but one thing is certain: E-prescribing capabilities will be one of them.

Before the economic stimulus bill even passed, the Center for Medicare and Medicaid Services began offering physicians a reimbursement bonus of up to 2% to begin e-prescribing this year, and the Obama administration's subsequent actions have made it clear that widespread e-prescribing is a top priority in its healthcare reform plans.

Fast forward five years. If all goes according to plan, as many as 75% of physicians will be prescribing electronically in 2014, and that will increase to 90% by 2018, a recent forecast projects. Only about 13% of physicians currently e-prescribe.

That's a lot of ground to cover, and the federal government is willing to spend billions to make sure it happens. But as this potentially irreversible momentum builds, it is worth taking a closer look at what we are paying for and how it might affect patient care.

The government hopes e-prescribing will save $22 billion over the next decade by preventing medical errors and increasing the use of generic drugs. The forecast, prepared by healthcare research firm Visante, claims electronic prescribing can prevent 3.5 million medication errors and 585,000 hospitalizations by 2018.

Technology alone won't make those estimates a reality, however. One of the features designed to prevent errors, for instance, is a medication safety alert that warns doctors of potential drug interactions, but a recent study suggests that system is ineffective at best and a hindrance at worst.

Researchers looked at the prescription records of more than 2,900 physicians in Massachusetts, New Jersey, and Pennsylvania, and of the 230,000 times physicians were warned of potential drug interactions via the technology's medication alert system, physicians proceeded as if nothing happened 90% of the time. That's worth repeating. Ninety percent of medication alerts did absolutely nothing to change physician behavior.

It's not that physicians are simply ignoring dangerous warning signs or unable to use the technology. A previous study found that one-third of all alerts lacked "adequate scientific basis or were not clinically useful," and even more were raising very minor or insignificant complications.

Although the onus on preventing errors and properly using e-prescribing technology will ultimately be on physicians, these flaws stem primarily from the manufacturers. Many of the alerts seem like they were written by a medical liability lawyer, rather than a clinician. It's as if someone took all the legal mumbo jumbo that comes in tiny print with every prescription and turned it into an annoying pop-up ad.

It's not that medical liability protection isn't important, but these tools that are supposed to make physicians' lives easier are “actually torturing them," said Saul N. Weingart, MD, PhD, co-author of the study.

When the majority of medication alerts aren't based on clinically-valid information, physicians get used to ignoring the warnings, and there's a greater chance that legitimate alerts will be dismissed or overlooked. A computer can only cry wolf so many times before doctors stop believing it.

Although there are many other benefits from e-prescribing and widespread adoption is still a worthwhile endeavor, this is a cautionary tale that policymakers should keep in mind while moving forward with EHR adoption and other, more expensive healthcare IT initiatives.

These technologies are intended as tools to help physicians improve how they practice medicine. It's important to keep that goal in mind and involve physicians in the process so the end result is a system that works with physicians, rather than against them. Because spending $19 billion for technology that is only effective one out of 10 times is a colossal waste of money.