Wednesday, September 18, 2013

How One Hospital's EMRs and the "Experts" Who Designed, Acquired, Implemented And Used Them Ruined Mother's Day

This post is a personal account of a disastrous encounter with healthcare information technology (HIT) by my own mother. This "as it happened" account should put an end to doubts as to the toxic effects that poorly designed and implemented health IT can have on medical care, even on relatively basic issues such as tracking of a mere five common medications.

Today is Mother's Day.

I weep.

It's almost one year to the day when my mother suffered severe and now clearly irrecoverable cardiac and brain injuries due to an EMR-related catastrophic blunder in the ED of a large hospital.

She spent the entire night last night in an agitated delirium, which is occurring more often now, with me, her son, tending to her needs. Not even strong sedatives helped much. It is only now, this morning at 9:30 AM, that she has finally drifted off to sleep, giving me the time to write this.

To the (ir)responsible people at the hospital, I offer my sincerest ingratitude for what you did to my mother through stupidity. Ignoring my confidential April 2010 warning letter to the CEO and CMO about my observations of your organization's EHR deficiencies did not ingratiate me to your organizational culture, either.

Yet I hope your mothers are healthy and responsive to your gifts and appreciation on this day. I honestly do.

I would not wish what my mother went through (cerebellar hemorrhage) on anyone.

Here's a sampling of how such errors occur due to the toxicity of EHR's:

My mother was placed on a medication, Sotalol Hydrochloride, by this hospital's cardiologists in appx. 2001 to prevent atrial fibrillation.

An ED visit of April 2010 shows Sotalol as a "current medication" (as did multiple ED and inpatient charts dating back almost a decade):

From the ED EHR of April 15, 2010 in an admission for abdominal pain from rectal stricture, resulting in an anoplasty (widening) as an elective outpatient surgical procedure a few days later:

ED EHR of April 15, 2010


1) Cozaar: 25 milligram(s) PO Daily. [For high blood pressure - ed.]
2) Albuterol: unknown dose. [For bronchitis - ed.]
3) Restoril: 30 milligram(s) PO Daily. [Sleeping pill - ed.]
4) Humulin N: 10 units in am,8 units in pm. [Insulin - ed.]
5) Sotalol Hydrochloride: 120 milligram(s) PO Every 12 hours. [For heart rhythm, see here - ed.]


: Hernia NOS , Atrial Fibrillation [a dangerous heart rhythm that Sotalol prevents - ed.], DMII Wo Cmp Nt ST Uncntr [diabetes - ed], Bronchitis, Urinary Tract Infection NOS, Hypertension.

This is not exactly a complex or taxing medications list or medical history.

Now, from the ED EHR of May 19, 2010, after a Transient Ischemic Attack with temporary slurred speech (aphasia) and narrowed left carotid artery identified as the culprit:

ED EHR of May 19, 2010


1) Albuterol: unknown dose.
2) Humulin N: 10 units in am,8 units in pm.
3) Cozaar: 25 milligram(s) PO Daily.
4) Restoril: 30 milligram(s) PO Daily.
5) Atrovent HFA

[Note that something important's missing - the critical heart rhythm maintaining medication, Sotalol. It's simply gone - de-listed - even though WE WERE ASKED on May 19 in the ED if she still took it based on the computer's current meds listing of April 2010, and replied "yes." It then somehow disappeared ... perhaps due to a
mission hostile user interface, or quality issues analogous to these ED EHR observations from Down Under? - ed.]


MEDICAL HISTORY: Hernia NOS , Atrial Fibrillation, DMII Wo Cmp Nt ST Uncntr, Bronchitis, Urinary Tract Infection NOS, Hypertension.

anaplasty [misspelling in the original; errors like this make record searching and data analyses a challenge. "Anaplasty" and "anoplasty" are quite different. But what's a little EHR error among friends? - ed.]

No previous psychiatric history.

Denies alcohol abuse, denies tobacco abuse.

[Obviously repetitious of the above, but this is exactly how the record appears - ed.] Hernia NOS, Atrial Fibrillation, Dmii Wo Cmp Nt St Uncntr, Bronchitis NOS, Urin Tract Infection NOS, Hypertension, bronchiectasis, TGA in 88, post-herpetic neuralgia, possible WPW variant.

Denies alcohol abuse, denies tobacco abuse, lives alone with son nearby. [Also duplicative; a little cut 'n paste action? - ed.]


Note that the duplicate versions of MEDICAL HISTORY and SOCIAL HISTORY at the bottom contain critical information not in the versions at the top, such as "possible WPW variant" and "lives alone with son nearby."

WPW syndrome (Wolff-Parkinson-White) is a syndrome that predisposes to heart rhythm disturbances such as supraventricular tachycardia (SVT) and atrial fibrillation. I offered that history to ED staff as a reason my mother took Sotalol. I also offered that I once had the WPW syndrome myself, and had recurrent episodes of both SVT and atrial fibrillation, until intervention via the 'catheter ablation' technique in the mid 1990's. The "WPW variant" made it to her chart. The medication did not.

This ED EHR report is a major information presentation faux pas, showing sloppy report generation and poor presentation of information, since people in a hurry -- such as in an ED, or in an ICU where my mother was admitted -- may not scan the duplicate versions below for critical data not in the top version. The repetition with added items in the lower-down versions is inexcusable.

This is the type of IT work done by amateurs. In fact the shabby ED EHR output could be considered an accident of omission waiting to happen. I covered this type of flaw here: .


So what meds actually were ordered after she left the ED and entered the ICU?


Click to enlarge. Note the remarkable resemblance - or should I say identicality - to the 5/19/10 ED EHR medication list above. These were then entered into the floor CPOE system.

No Sotalol.

Several days later, after my mother was transferred out of ICU to the neuro floor in prep for a carotid stenting, my mother went into uncontrolled atrial fibrillation. In front of my eyes at around dinnertime. Sotalol has a half life of just about 12 hours. After a few days of missing it, it's effectively gone, at a level so low in the blood as to be ineffective.

I immediately asked nursing to call the physician, and asked how this was even possible on Sotalol. All I got back was a puzzled look.

Sotalol had never been ordered.

Long story short, in correcting the iatrogenic atrial fibrillation (the arrhythmia itself a stroke risk), Sotalol was restarted and IV heparin and electrical cardioversion (shock to the chest) were employed.

While initially successful in restoring normal heart rhythm, the IV heparin then precipitated a massive cerebellar hemorrhage around 2 AM the next morning (in an area of the brain entirely different from the cerebral language center supplied by the carotid). IV heparin is a dangerous drug you especially want to avoid in the elderly.

In what I thought might be her last words to me, all my mother could say when I arrived as they were taking her for surgery was a very agitated and frightened-out-of-her-wits "Scot-headache-headache-headache-I'm sick-I'm sick-I'm sick." The headache pain from the bleed must have been absolutely excruciating and horrendous.

She had the pallor of death about her.

The sheer horror of that moment, seeing my mother like that, sticks with me even though I am a physician.

Her CT looked dreadful, much like the stock image below:

Click to enlarge. When I saw her CT scan, I prepared to say my goodbyes.

Emergency craniotomy (brain surgery) was then performed by a neurosurgeon to save her life, followed by months of medical complications and agitated delirium resistant to most medications.

The brainstem sits anterior to the cerebellum, and brainstem compression from the bleed caused paralysis of the muscles of swallowing. A surgically placed stomach feeding tube through the abdominal wall was thus required, which in her delirium she yanked out, thus forcing total parenteral nutrition (TPN) via a central (deep) IV line until the site healed. (She was lucky she did not do herself a major injury by pulling the through-the-abdomen gastric feeding tube, which should have been far better protected with a delirious patient.)

The first ingredient I noted in the initial TPN bag was a medication my mother was allergic to, famotidine, but another EMR defect I uncovered prevented the clinicians from realizing this. It was only my personal knowledge that prevented administration. I ended up filing my own FDA MAUDE report on this EHR my knowledge the hospital did not, and has not.

The complications of the bleed and craniotomy surgery also forced the re-started Sotalol to be discontinued, as bleeds in the head adversely affect the heart's electrical intervals, and a drug like Sotalol can kill under those conditions (e.g., see Torsade de Pointes).

The forced discontinuation of Sotalol resulted once again in sudden return of atrial fibrillation that is now permanent. Correction or even appropriate anti-stroke treatment would require anticoagulation such as heparin and coumadin, and these are contraindicated due to the bleed.

My mother has been painted into a very bad corner, with irreparable cardiac injury putting her at increased stroke risk, and a severely damaged brain that has resulted in her being an invalid who frequently is in an agitated delirium, and/or no longer recognizes her own son. While her swallowing returned after several months, her mind never recovered.

In effect, the EHR toxicity caused the following personnel (at the very least) to miss my mother's Sotalol de-listing, which apparently propagated from ED to ICU to floor due to lack of any discernible fail-safes or meaningful reconciliation:

  • ED triage nurse
  • ED physician [who stunningly notes in the record "She had a prior history of atrial fibrillation. However, per her report this has resolved and she is not currently taking any anticoagulation." A true statement - except he forgets to address the first-year medical student-level question: "atrial fibrillation resolved - how?" The answer that I'd related to him was "her atrial fibrillation resolved on Sotalol"; Sotalol is a take-for-life drug. On the NIH page about Sotalol: "Sotalol controls your condition but does not cure it. Continue to take Sotalol even if you feel well. Do not stop taking sotalol without talking to your doctor."]
  • ED staff nurse
  • Two neuro-interventionalist physicians
  • Multiple medical residents
  • ICU physicians
  • ICU nurses and staff
  • Floor nurses and physicians after transfer from ICU to neuro floor

I further observe: it is nearly unbelievable to me that not one person out of all the clinicians who saw my mother during this admission, even when transferring her from ED to ICU to floor, detected this fundamental, major medical error via ED and inpatient histories dating back almost a decade. Is interfacing between the floor and ED EHRs an issue? Are the EHR's too mission hostile for busy clinicians to use? What in hell was going on here?

You'd think that for the $25+ million dollars spent by this organization on HIT, the IT could have included failsafe features on meds and other life-critical data. Has the computer become deified in this culture; its outputs, a Testament that nobody challenges?
I tragically note that one Sotalol pill, worth perhaps a few cents, probably could have prevented the catastrophe if the error had been detected even as late as floor transfer from ICU.

To add insult to injury, the following was added to the ICU H&P sheet some time after the catastrophic brain bleed:

Click to enlarge. A very unwelcome discovery when I asked the neuro floor RN to see the chart a few weeks after the accident. Entry #8 was not present the day the Sotalol error was realized; I was shown the chart at that time.

Undated, untimed additions to medical charts with illegible signatures violate Joint Commission, Medicare and State Medical Professional standards of conduct, among others.

I leave it to the reader as to why this chart alteration might have been attempted.

When I saw this, needless to say, I was very upset. I demanded an immediate printout of the eMAR (electronic medication administration record), lest that be altered as well. The eMAR fortunately showed Sotalol had indeed not been ordered prior to the onset of the A. fib. I do not know what it might have showed a week ... or a month ... later, had I not been a nosy and medicine/medical informatics-educated patient advocate for my mother.

As I did not have my Power of Attorney documents for my mother with me, I demanded the eMAR printout and copy of the altered ICU H&P be sealed in an envelope whose glue flap I signed, to be held until I retrieved my POA documents from the nearby bank allowing me to take custody of copies of both these documents.

This is how toxic these wonderful non-FDA approved or vetted, Obama HITECH-pushed Cybernetic Revolutionizers of Medicine can be, when not "done well."

It should be noted that further errors of this type at this hospital system would very likely amount to criminal negligence.

Happy Mother's Day.

-- SS


Bill said...

I am sorry for what happened to your mother, especially on this day.


Live it or live with it said...

Hugs for you Scot. Nothing I can think to write will comfort you. I am sorry this happened to your mother. May some good come from it someday.

Anonymous said...

What a horrific experience from EHRs and CPOE. These devices are defective, unusable, error promoting, and obviously, DANGEROUS.

They should be recalled.

InformaticsMD said...

These devices are defective, unusable, error promoting, and obviously, DANGEROUS.

It's not so much the devices but the people who design, implement and kvell about today's ill conceived and markedly deficient systems.

-- SS

Marc said...

I'm sorry for what happened - my own mother has just gone through a major procedure here in the UK where we've had a chequered history so far of healthcare IT and it's been a difficult time.

But it seems from your account that you can't tell how the 'system' broke down. I write about health systems in Europe - I'm aware that the medical error rate in the US has been higher than here for some time, and not just a function of new IT. I'm wondering if there is evidence of errors rising in the US with electronic records and other admin systems.

InformaticsMD said...

Marc writes:

I'm wondering if there is evidence of errors rising in the US with electronic records and other admin systems.

Sorry to hear about your own mother.

You ask a good question. With the HIT systems exempt from the federal reporting requirements that other medical devices are subject to (see my post here), it's hard to tell exactly what is going on. Even the FDA admits its data on injuries and death related to IT is likely just the "tip of the iceberg" and the Joint Commission admits there is a paucity of data on HIT-related medical errors as well (in aforementioned JC Sentinel Events Alert on health IT of 2009).

I can say the literature is not strongly and uniformly supportive of cybernetic improvements (e.g., see "HIT Reading List.")

On a personal note, I do not believe the error that happened to my mother would have occurred in the paper era.

How can I make such a claim?

I worked in that very same ED for several years prior to computerization. The error mode would have been very unlikely to have occurred, unless someone was using disappearing ink.

-- SS

InformaticsMD said...

Re: Marc

I can also add that the "system" broke down locally when hospital officials ignored my warning letter of observed deficiencies, one month prior to my mother's injuries from the type of deficiencies I warned of (e.g., impaired clinician-clinician communications).

On a larger scale, the "system" has broken down when the advice of the pioneers and other bona fide experts is ignored by the HIT industry, e.g., the National Research Council's 2009 report on HIT which is also in the reading list links.

-- SS

Marc said...


Why are the meds handwritten on the chart? Surely that's an obvious source of error? I peeked at my mother's charts - in fact a whole binder at the end of the bed - and they were all computer printouts from a brief look. Apart from a few large and obvious notes on a board at head of the bed.

One strength we also have in the UK is a tight integration between primary and acute care with IT in many regions.

Linda MacDonald Glenn said...

So sorry to hear this, Scott. It is particularly frightening to hear this story from a physician.
I hope your story prevents this from happening to someone else.

Afraid said...

Dear Scott, Do you have any fear that you and your family may not get the best treatment at hospitals you may have criticized?

Not do you think that worse treatment is or was administered, but do you fear it, do you feel it might happen, do you worry about it?

InformaticsMD said...

Marc writes:

Why are the meds handwritten on the chart?

I do not know for sure. Perhaps in transition or interface issues between the ED EHR and floor EHR (they are from different vendors).

However it does not take Sherlock Holmes to note the identicalness to the ED list. Once the error occurred, it propagated itself and was missed by innumerable clinicians. I was present every day of the hospitalization as well. Nobody asked me about her meds.

Having worked in that ED, I would never have expected such an error. I also had trust in this hospital from my past experience. That trust was misplaced; it was based on memories of a past that clearly no longer exists.

-- SS

InformaticsMD said...

Afraid said...

Do you have any fear that you and your family may not get the best treatment at hospitals you may have criticized?

As I've not identified the hospital, I doubt it.

However, remember that hospital walls have ears, and still have people with integrity working at them.

Such retaliation could likely end up with people spending time behind bars.

-- SS

Deborah said...

My sincerest condolences and sympathy for your tragic loss. Thank you for sharing a deep and personal tragedy.