Dear Dr. Silverstein,
As you write, there is not a transaction of medical care that does not go through EHR systems.
However, these poorly usable EHR systems stifle creative and artistic thought required to link risk, benefit, and probability of diagnosis with risks and benefits of testing and therapeutics.
Assuring safety and efficacy with pre- and aftermarket surveillance will maximize the possibility of achieving the potential of the technologies.
Additionally, when I use these electronic ordering systems and libraries of medical information, they fail to keep up with the agility and nimbleness of my mind as I seek 'random access' to pieces of data to formulate and synthesize diagnoses and therapeutic strategies.
The EHRs are too slow, do not have a robust (if any) search function, randomly and whimsically store key information with ever changing formats, and generally obfuscate what should be simple. They are cumbersome and disable the ability to simultaneously and contemporaneously compare myriad data points.
They get an "F" as enablers of complex diagnostics.
Paper, since it can be organized as needed and set out on a desk to be seen and compared as quickly as the eye registers the data, gets an "A".
The EHRs are impediments and disrupters of communication.
Example: Just today, I was witness to the fact that a stat EKG was ordered by CPOE on a heart patient yesterday at or shortly after 4:30 pm. The intended recipient of the order (heart station) never got it because they close near 4:30 pm and there was no warning to the ordering health professional that was so.
Thus, the EKG was never done, and this morning, when the requisition was seen, no one did it because it was ordered stat "yesterday", and the techs asked themselves "what good would it do for a 'stat' to be done now, a day late?"
I do not know what happened to the patient. I have many other examples of such delays facilitated by the CPOE and EHR systems that I am required to use at numerous facilities.
They facilitate 'stealth' alterations in care. Also just today, a disease-critical test ordered 3 days ago was not done because it was cancelled in 'stealth' (automatically "expired") without warning to me by the lab responsible for doing it.
There is the "silent silo" syndrome as you've called it. Also just today, a disease critical test ordered 5 days ago came back with results, but the results were posted in the information 'silo' of 5 days ago. The lab screen default on the EHR only goes back 4 days (so unless I knew to look for it, it would not be seen or acted on), further obfuscating data and delaying treatment.
The EHRs lose data and orders. Also just today, I found that blood coagulation monitoring tests that were ordered to be done with kidney dialysis (3 days per week) on a patient somehow got "lost" and were not being done for 5 days, putting my patient at risk of bleeds - or stroke if the blood was not 'thin' enough.
I just walked in to examine a hospitalized patient with multi-organ failure and diabetes, on multiple meds including insulin, and recovering from respiratory failure.
The nurse anxiously informed me that the blood sugar was dangerously low. I ordered treatment stat.
I see patients in the morning before labs come back, and depend on nurses to review labs and notify me.
Turns out that the patient was hypoglycemic on yesterday morning labs that arrived in the EHR 'silo' after I left the hospital; and was also low in potassium, but the tests just laid there comfortable in their silos; and were not communicated to anyone like in the old days when a human ward clerk or other undistracted human received them and disseminated them to the appropriate professionals.
Thus, instead of getting less insulin, the patient got the usual dose with near catastrophic adversity.
Misidentifications are facilitated by EHRs. I noticed that on several critical clinically significant changes that arose on my patient that were entered as such in an EHR silo by the RN, it was stated that they called attending physician 'Dr X', which was not me...obviously a case of EHR-facilitated misidentification.
Here is a misidentification variant: yesterday, someone (non doctor but not clear who) ordered a specialist consultation on one of my patients under my name. I did not order it nor was it needed, yet it showed up as an order for me to sign.
Like you, I agree this is representative of a toxic impact of these systems on medical care and I feel like the care environment is foul, like a cesspool, compared to what has been replaced.
These systems of medical devices cannot be trusted in the care of sick patients. Perhaps, they are OK for managing hang nails.
I offer no additional comments other then if I am sick, I do not want my care interfered with in this manner by IT.
Rest assured, though - there are IT hyper-enthusiasts out there (http://hcrenewal.blogspot.com/2012/03/doctors-and-ehrs-reframing-modernists-v.html) who would see little problem with this, as any accidents that occur are "anecdotes", "learning experiences" or "bumps in the road."
That's if they don't simply blame the user.