Friday, December 20, 2013
PAUL CAME TO the clinic because he was feeling depressed. As the first physician to see him, I opened a new file in our electronic health record system.
After typing in "depression" as the presenting symptom, I was shown a menu of dozens of potential diagnoses, including "adjustment disorder with depressed mood," "adjustment disorder mixed anxiety and depressed mood" and "atypical depressive disorder." A competent clinician could choose one of several options and still be accurate. Unfortunately, once saved, these entries are difficult to change.
Electronic health records were once touted as a turning point in U.S. health care, expected to revolutionize recordkeeping on patient histories and to reduce the potential for medical errors. I hope that someday they will. But as the Affordable Care Act launches and, more Americans seek medical treatment, there is no standard system nationwide, and data-entry errors and inconsistencies are becoming common.
The most significant benefit of the expensive, multiyear effort to implement electronic recordkeeping may be legible medical notes.
Medical billing is a chronic war between those wanting to get paid (physicians, hospitals) and those not wanting to pay (insurance companies, government). While billing has long been computerized, the shift to electronic records among clinical practices is relatively recent. The 2009 stimulus legislation made billions in federal funds available for the transition, with more earmarked for the future. Hundreds of electronic systems now exist.
To greatly simplify, the dog-and-pony show goes something like this: A company hypes its system, emphasizing its compatibility with billing procedures. Administrators buy the systems, generally without input from the doctors, nurses and others who enter the data.
A June editorial in the Annals of Internal Medicine rightly called the universe of systems a "Tower of Babel," noting that a 2012 study found that only about 10 percent of U.S. physicians reported "meaningful use" of them.
They tend to be fantastically complex. One doesn't have to be intimately familiar with, say, Hertz or Enterprise to rent a car online. But many electronic health record systems have pull-down screens listing each of the 68,000 possible diagnosis codes in the World Health Organization's International Classification of Diseases and 87,000 possible procedure codes.
Consider what happens when I write a prescription: Every potential drug interaction or side effect listed generates a warning prompt. Inevitably, recognizing that the warnings are generally inapplicable and take time to sort out, clinicians start to bypass the alerts. Sooner or later, ignoring one will lead to a serious complication.
Compounding these issues is the fact that hospitals, clinics and offices have different systems. When these systems get overloaded or go down, more work is generated, sparking delays in care.
Like many health-care providers, I work at more than one location. Each facility has its own electronic record system, with different ways to enter data. Each program takes time to learn and requires continuing education.
Perhaps the most pernicious side effect is the erosion of the provider-patient relationship. When I first began working with electronic health records, I caught myself staring at the computer screen instead of engaging with patients, who rightly felt ignored. Like many colleagues, I've reverted to the practice of taking notes with pen and paper. After the office visit is over and the patient has left, I type in the data. This takes much more time, but it is the only way to complete a proper history and exam.
The result is decreased productivity and frustrated providers -- and a lack of meaningful data with which to manage patient care.
It's hard not to think that much confusion and duplication could have been avoided if policymakers had involved providers in every step of the transition.
(Continued on page 2)