Moderated by Tom Sabulis

The spread of electronic medical records has at least one Atlanta doctor concerned about the accuracy of information on patient files and the effect of faulty info on potential malpractice. A top Emory doctor and administrator says the system is not perfect, but patients are far safer with complete  medical histories available on computer.

Commenting is open below William Bornstein’s column.

By Melody T. McCloud

Because someone rammed her SUV into my car, this physician-surgeon has, of late, been a patient at physicians’ offices and outpatient surgical centers. While medical sights and sounds don’t scare me, something about electronic medical records does.

I love my profession. I am a stickler for accuracy, and I like reviewing patient charts. I’ve always secured a copy of my own medical records: I like being an informed patient. It was sometimes difficult to decipher doctors’ scribble, but I could note the accurate documentation of my “HPI” — history of the present illness — and that of exams performed.

In this world of electronic medical (or health) records (EMR/EHR), however, I’m no longer sure that is always the case.

EMRs are helpful: physicians can review typed patient information that’s accessible with a mouse click.

But the physician’s entries of patient information must be as accurate as possible. By securing copies of my records, and seeing records of others, I am very concerned that, with the implementation of EMRs, not everything that is documented is actually being done, especially in emergency rooms. A few times, when reading my records, I said, “He didn’t examine or check for that.”

Fortunately, I understand what I’m reading. I know what should be done during exams. The layperson doesn’t have this benefit.

I fear that clinicians may rely on the EMR to literally “fill in the blanks” of examinations not done. This can promote laziness and minimized attention to detail, as well as inaccuracies concerning the doctor-patient encounter.

Automated EMR fill-ins of examination findings might benefit doctors in malpractice cases because “if it’s not documented, you didn’t do it.” So EMRs may wrongly afford protection to some undeserving physician. Plus inaccurate documentation can potentially harm patients by not giving a true account of their condition.

Under the Department of Health and Human Services, the Office of the National Coordinator for Health Information Technology (ONC) oversees the national rollout and implementation of EMRs. Their mission includes securing and protecting patient health information and “coordination of care among hospitals, labs and physicians.” But I did not see any mention of assuring accuracy of the information that is so safely transmitted. What good is it to assure the records are safe and easy to share if the medical information itself is erroneous?

The ONC’s subcommittees include a Standards Committee, with its “Clinical Quality” subcommittee. I encourage them to add “accuracy” to their national mission; and perhaps changes in EMR software are needed.

I also strongly recommend that physicians always provide a narrative documentation of the patient’s chief complaint and findings. Medical records must be accurate in every way, every day. The profession and the patients deserve no less.

Dr. McCloud is an obstetrician-gynecologist and founder and medical director of Atlanta Women’s Healthcare.

Benefits outweigh risks

By William Bornstein

Physicians understand that all good health care decisions must take into account whether expected benefits will exceed anticipated risks. The use of electronic health records (EHRs) also includes benefits and risks.

The information available to physicians about individual patients, as well as the related base of scientific evidence, has grown dramatically over recent decades. Because of this expansion of information, we have much more to offer patients to improve and individualize their care. This growing complexity of information brings risks, however, including a greater possibility of error, and potential harm created by the failure to have all the relevant information accessible at the points of care.

The traditional tools of physician information management have been memory and pen and paper. When patients had one doctor for much of their lives, that doctor took care of most of their health care needs, and there were limited therapeutic options. Today, patients see multiple providers in multiple specialties, often in multiple geographic regions.

The current level of health care information and the needs for patient and physician access cannot be met without modern information technology, any more than current aviation needs could be achieved by pilots flying aircraft by the “seats of their pants” without modern aviation information systems.

When a physician prescribes a new medication for a patient who is already on six medications, how is he or she to know about the possibility of an interaction between the new medication and one of the current medications?

It is nice to think that the physician carries all the relevant information in his or her head, or painstakingly looks up each possible interaction in a reference text before prescribing the new one. However, in reality, neither approach is feasible. This is exactly the type of cross checking that computers and EHRs do well to support physician decision-making.

Over time, EHRs will become even better, and we will improve our ability to implement them. When used appropriately, EHRs can facilitate better personal interactions among physicians, patients and families. Much of the depersonalization in health care in recent years has resulted from information overload. Typically, physicians carry a great deal of routine information in their heads, including medications, vaccination status and other elements of preventive care. If this basic information management is facilitated by a computer, physicians can devote their attention to the unique considerations of the individual patient, while still maintaining ultimate responsibility for the accuracy of the information.

The question should not be, are EHRs perfect, because they are not. Clearly, there are challenges and risks in implementing these systems, and traditional pen and paper are better tools in some cases. The question should be, are patients generally safer and better off when their care is facilitated by well-implemented EHRs? The overwhelming evidence says, “yes.”

Dr. Bornstein is Chief Quality Officer of Emory Healthcare.