Perspective: Electronic Health Records (EHRs): Can we trust them?

July/August, 2013
By Charles A. Pilcher MD FACEP

We’ve all seen the hoopla about the benefits of the EHR: standardization, easy communication between providers, reduction in paper, no need to re-enter the same data, better tracking of medications, etc. The EMR or EHR (as the Department of Health and Human Services prefers) has become the standard of the land.

But, while touted for it’s clarity, efficiency, portability, and cost saving potential, the EHR has also created problems such as higher patient bills and and decreased physician efficiency. Yes, illegible handwriting is no longer an issue, and dangerous drug interactions can be nipped in the bud. But it’s still possible to write the correct order on the wrong patient, give the wrong dose of a medicine because of confusion over on-screen tabs and menus, and deliver test results to the wrong chart, or not at all, due to both human and system errors. Worst of all, actual communication between providers has decreased. Even the credibility of the individual entering the documentation becomes an issue.

So let’s take a look at some of the downsides.

Does an EHR tell the truth?

From a medical-legal perspective, a major downside of the EHR is the truth of the document itself. Each member of the medical care team is independently entering information into this record. What is entered is a matter of both perception and veracity. It’s much like the Indian parable of the blind men describing an elephant: all may be right but none may be right. There can be a range of truths and fallacies, all relative. Facts become opinions. Example: A nurse records that a patient’s abdomen is “tender in all 4 quadrants” and “the patient walks slowly, bent over and holding his stomach.” The physician’s entry records “abdomen slightly tender” and “gait and balance are normal.” Both can be correct, but one is certainly more descriptive than the other. Another example: the HPI records certain “positives,” but later the ROS records the same data points as “normal.” In each case the entries can be made with a few keystrokes or checked boxes,  Who and what are we to believe when the patient is discharged, only to return 24 hours later with a ruptured appendix?

EHR: Patient record or billing document?

Most EHRs use computerized templates which contain “check boxes” for a physician or nurse to complete. Seeing the array of check boxes, each with an accompanying drop-down box for amplification of any positive finding, caregivers are tempted to check “normal” for 2 reasons: 1) they will not have to further document the nature of any positive finding, thus saving time, and 2) they have proven to the “chart auditors” that they have examined that system. Why is that important? Because physicians are paid on “piecework,” and the more work documented, the higher the “visit level.” Some systems can even prompt physicians to add elements to maximize the billing code.  [Editor’s note: Conversely, as some argue, the EHR may only be documenting work that was done but previously not being recorded for billing purposes.] This unintended consequence was discussed in a meeting convened by CMS on May 3, 2013, where the topic was “upcoding,” (documentation that inflates the amount of work actually done).

Most EHRs use computerized templates which contain “check boxes” for a physician or nurse to complete. Seeing the array of check boxes, each with an accompanying drop-down box for amplification of any positive finding, caregivers are tempted to check “normal” for 2 reasons: 1) they will not have to further document the nature of any positive finding, thus saving time, and 2) they have proven to the “chart auditors” that they have examined that system. Why is that important? Because physicians are paid on “piecework,” and the more work documented, the higher the “visit level.” Some systems can even prompt physicians to add elements to maximize the billing code. [Editor’s note: Conversely, as some argue, the EHR may only be documenting work that was done but previously not being recorded for billing purposes.]

For example, how many charts of patients with an ankle sprain record “7 systems reviewed” and document  an exam of the heart and lungs? How many patients with simple conjunctivitis have a neuro exam recorded? Unfortunately, the answer is “all too many,” because in 99% of cases that information is irrelevant. An April, 2013, article in Annals of Emergency Medicine  (“Upcoding Under Fire”) reports that half of  all “Evaluation and Management” billing codes were for the highest level, 99285. Caral Edelberg of Edelberg Compliance Associates  recently told ACEP News that, in fact, one of the most common reasons for a Medicare chart audit is the finding that a hospital, department or physician “has achieved better than average provider documentation.”

In the physician office, a common problem is “chart cloning,” copying and pasting a prior note into the record of the current visit. Renee Stanz noted in Medical Economics, November 25, 2012, that “the Office of Inspector General (OIG) continues to focus on ‘identical notes’ as an area of concern.” [Editor’s Note: I have personally seen 17 consecutive monthly visits recorded with less than 5 words changed in each visit.]

Finally, even the United States Senate has weighed in on this, with a report titled “Reboot: Re-examining the strategies needed to successfully adopt health IT.” The report noted that EHR’s have not resulted in promised savings for Medicare patients because of “code creep.” The latter is what happens when physicians can quickly document more – perhaps too much – on an EHR and increase the value of the service.

EHR: Opportunities for mistakes

Two types of mistakes are common to EHRs.

  1. The worst mistake is choosing the wrong template. Most emergency physicians have had this experience: A 40 year old male patient tells the Triage Nurse “I’ve been feeling nauseated for 2 hours and just started vomiting.” She dutifully records the Chief Complaint, begins the “GI template” for the physician and takes the vital signs. The physician picks up the chart, goes through the checklist, finds little of interest, decides the patient has gastroenteritis, and begins the discharge process. As he is getting dressed, the patient asks the doctor “So you don’t think I’m having a heart attack?” Taken aback, the doctor asks, “What makes you think it could be a heart attack?” The patient replies, “Well I had a bit of chest pain just before I started vomiting.” By picking a wrong template, one is led down the wrong path to a wrong diagnosis and a potentially catastrophic outcome.
  2. The second mistake is erroneously checked boxes. Boxes may be marked as normal, not checked at all even when the information is important, or  not being present to check because the template is designed for a problem the patient doesn’t have (see above). The EHR has thus become our master rather than our servant.

EHR: The forest and the trees

The result of all these inputs is a chart where one finds repetitive, conflicting, and misleading documentation such as “patient resting comfortably, side rails up, call light at bedside” on one page. Less than 5 minutes later, the MAR (Medication Administration Record) record says “Pain 8/10, 2 mg Dilaudid IV.” Automatically downloaded vital signs and rhythm strips may fill the EHR with information of little value (unless it later is shown that an abnormality was missed.)

EHR “protocols” also cloud the thought process involved in medical decision making. Any connection between the patient’s symptoms and signs, the differential diagnosis, and the tests ordered becomes unclear. When every chest pain patient, whether an 18 yo athlete or a 75 yo cardiac patient, gets the same workup “by protocol,” one may well ask “What was the doctor thinking?”

Actual “human communication” in the form of free text in an EHR is scant because of the extra effort required, time for which is often absent on a busy inpatient floor. The result: Bushels of chaff and very little wheat. It’s hard to make bread of that.

EHR: Meta-data and audit trails

This is data that is usually only available with a special request of the IT Department. Here you will find who entered data, when it was entered, who later accessed it and when. This can be extremely useful for the defense in a malpractice case to prove what the defendant knew. It can also be useful for the plaintiff to show that the information may never have been seen. A caveat here… It’s not unusual in a busy ED or inpatient unit that a caregiver failed to log out before another caregiver entered or viewed data. But that is becoming less common with faster login/logout identifiers and procedures.

Who actually wrote this? The doctor or the computer?

Many systems have software that  turns all of the checkboxes into free text. However, if one reads carefully, one quickly realizes that  only computers talk like that. The importance of this is that it may be helpful in a med-mal case to ask an IT department or vendor for a copy of the “bare bones” or “blank” template from which the computer generated what may appear to be a free text narrative.

Credibility matters

The whole purpose of this analysis is that credibility matters. While most physicians are diligent and honest, an EHR can be too much temptation for some, like bringing Jack Daniels to an AA meeting. Cutting corners becomes too easy. Discrepancies in the record may be hidden too deep to be addressed. If the “medical decision making” is anything but straightforward, the attentive physician will always explain this  in a supplemental dictated note (which if done using Dragon or a similar platform, should always be verified… perhaps the subject of another essay.) Too often, this important step is skipped.

The advantage of the old-fashioned paper chart is that only pertinent and credible information was recorded, exactly the type of communication that helps multiple caregivers understand each other.

For further reading:

{ 1 comment… read it below or add one }

Gerry Oginski July 30, 2013 at 9:41 AM

Here is what I see every time I review a case involving electronic medical records. I see templates with the same information repeated over and over again. That tends to be the default setting. That tells me the physician simply checks off boxes with the same information without entering new information for this particular visit.

I find that many physicians using EMR tend to copy and paste and not pay attention to what really matters. Unfortunately, the electronic medical records divert the doctor’s attention away from the patient and what they are telling them.

Leave a Comment

*