A simple idea: Dr. EMR




Much of medicine is no harder than Mom, a Band-Aid and a scrapped knee. Flu shots save lives, give flu shots. Bleeding causes anemia, give iron or, if severe, blood. There is a fracture, fix it. A boil hurts, lance it.  This is not rocket science.

Perhaps medicine is so simple that it can be automated. Instead of a doctor at all, use an algorithm. Rule #1; If A then B. For the harder cases; If A then B unless C. Easy, flawless, idiot proof, cheap.

Therefore, I am intrigued by the following treatment protocol, which is proposed for use in my hospital. It seems rudimentary, but it is revolutionary or at least prophetic … tomorrow’s medicine today.

The idea, which came from our orthopedic service, is simple. If a patient’s blood test shows a deficiency in Vitamin D, than give that patient Vitamin D. Now, please note, the key here is that no direct order for Vitamin D treatment is written by or requested from a doctor, nurse practitioner or nurse. If the lab reports low Vitamin D, then the patient will be automatically treated. In effect, the EMR will note the lab test and order therapy.

This seems like a splendid and efficient solution to a real and common medical problem. Many studies indicate that low Vitamin D is associated with a mix of different maladies, some life threatening. The strongest data is around bone health as osteoporosis is increased with low Vitamin D. So is the risk of falls with fractures, the complications of which can be devastating.

With Vitamin D deficiency, colon and lung cancer increase, as do breast cancers in postmenopausal women. In addition, research suggests that if you do get cancer and you already have low vitamin D levels, you are more likely to die. Finally, there may be effects on heart health, immune function, and diabetes.

What is not completely clear is how much benefit is derived by replacing low Vitamin D. While bone fractures happen less, can the same be said for other illness? The research is still out. However, in patients with low Vitamin D, there is very little risk from supplements and if the only thing we achieve is less hip fractures, that is still a big win.

The question remains, are we comfortable if an automatic Vitamin D prescription is at the core of the decision? If A, then B, sans MD.   If we accept this very simple, safe plan, how far do we want to go? Is there anything wrong with an automatic test of every patient that is seen in a clinic or office for Vitamin D and giving a Vitamin D prescription if “decided” by the EMR, without the doctor ever being notified?

This apparently innocuous example is a model for future medical care, where information systems completely manage medical decisions. What is the real difference between ordering Vitamin D, and penicillin? The EMR notes a positive blood culture, checks sensitivities, measures renal function, rules out allergies, and then the computer orders an on-time, on-target antibiotic, while the doctor is asleep in bed, playing golf or, hopefully, treating another, more complex, patient?

How about computers titrating seizure medicines based on EEG results and blood levels? Drug and dose ordering home blood pressure machines? Auto-analyzing telemetry, suppressing arrhythmias by adding meds and adjusting pacemakers? Intelligent ventilators? Sentient da Vinci robots? A cell phone App, which adjusts IV pain medications for patients on hospice?

Enhanced medical informatics has a mindboggling but unknown potential for accuracy, speed, patient empowerment, reduced cost and might guarantee state-of-the-art care. However, the transformation will be complex and, to no small extent, disrupting and disturbing. How and where we apply such technology will depend not only on the development of automated systems, but on our understanding of what it means to be human. Nonetheless, wherever our journey takes us, we can be certain that what it means to be a patient or a doctor will change forever.


  • Liz
    I am sure that would work for the 100% this is always the solution, 100% of the exceptions can be programed into the computer. Not so sure about the stuff that involves a judgement call, where there is more than one option and no clear cut A>B>C exists with respect to choices, where there hasn't been treatment failure prior to this for the same thing, where you aren't going to potentially have interactions with other meds or medical conditions… where life gets complex and the answers are neither simple, clear cut nor universal.
  • Julie
    I had a family member who at 28 went to the ER [not local] with a reaction to the anti-biotic given to her for her pneumonia. In the ER one of the drugs administered was a blood thinner. I asked why a healthy 28yr old would be given a blood thinner. I was told it was because it was the pneumonia "protocol". I could imagine health concerns that might warrant this drug but not for this patient. Upon questioning my medical colleagues they agreed it was the algorithm not the MD order that triggered the drug. NOT the practice of medicine.
  • jillma
    Why is someone deficient in any area? The human body is an entire organism that has a cooperative system...since when has medicine ever been simple? If A than maybe an infinite answer for everyone rather than finite.
  • D Someya Reed
    So we will treat conditions, not patients. This is empowerment? Not by any definition yet agreed upon in the medical (or any other) field. Is this global detachment from patients where we really are going even though we say we are not? What is described here is more likely to be cost cutting (won't need all those pesky doctors or nurses with all their outrageous and costly "demands") but it certainly won't include billing reductions. All those new machines are going to be expensive to buy, program and maintain so billings will have to rise accordingly (even with less overall operational costs, including acquisitions, to cover). And, as they say, you get what you pay for. Perhaps I'll just wait for the inevitable "home version" with a copy of 'Medical Practice at Home for Dummies' and an enhanced medical informatics version of the game 'Operation' but which will, by then, be guaranteed safe to use on a family member. If medicine is not "rocket science" then each of us should already know it either instinctively or inherently. Common sense, right? With all that we don't know, what is the rush to take "the people" out of "the relationship?"
  • meyati
    My body temp is pretty low. When I was in the surgery ward, they kept coming and checking my monitor. Then some nurses came in and disconnected me, and connected themselves up- They said- Oh, that's amazing- They felt my warm feet and shook their heads. Then a couple of techs came in and made sure that i was alive, and they asked me what I did to their machine. Later they came in with a real old-fashioned oral thermometer (I was lucky that it wasn't rectal). They said that the brand new machines couldn't register my temp-because it was too low. The techs checked all of them. I told them that they'd better get it fixed before somebody came in with hypothermia. Somebody said-- I didn't think of that. We need to get new ones. Things like that give me so much faith in medical electronics. I doubt that I'd bother picking up half of the meds anyway, as they'd insist on synthetic thyroid which gives me reactions, Tylenol-ditto.
  • Kathryn Zusmanis
    Dr. Salwitz's article seems pro automation, because in many ways this is already happening. I agree, and owe my life to the fact Western medicine uses technology to enhance good patient outcomes. This is how he put it. "Enhanced medical informatics has a mindboggling but unknown potential for accuracy, speed, patient empowerment, reduced cost and might guarantee state-of-the-art care." Computers already flag out of range test results, which enhances your doctor's interpretation. To err is human. Many times human error, slow response, missed and misread results, result in less than good treatment plans. The key to using algorithms, which are written by humans, is to improve accuracy and speed by allowing it to add to, and enhance human interactions.
  • meyati
    This has nothing to do with med, but it takes about 30 hours to fly from my home to Naha, Okinawa. The last trip lasted about 50 hours getting there, my family didn't know where I was. My brother sat at Naha for over 24 hours. Because of Japanese and American privacy laws-and the airlines losing us electronically-my family didn't know where I was at. all. Hell, I didn't know where I was. Phone numbers changed because I was in a different country-I was locked out of all of my Email accounts, and I can go on and on. What was worse, the airline staff at Tokyo misread the message from Denver. It took hours to get a hotel voucher and a ticket for the connecting flight. And it was raining, and we were all wet. Rockers headed for Singapore began crying. Then a radiologist misread my reports. I kept telling them that I was only a radiology and surgery patient, not a chemo patient at all. They were scanning the wrong area, They kept saying I had tonsil cancer, I had negative biopsies-and the idiots kept saying this. I think that between my low temp---how electronics, algorithms, and people do things, that I have adequate reason to question the accuracy and validity of these things. I highlighted the anomalies on my radiology reports and took them into the boss. Setting up for the next scan-they got after me- then they read the intake papers and saw that they just didn't listen to me or bother reading-not even reading what the doctor wrote. I liked the old days, where they used a grease pencil or a sharpie and wrote-- Temp- 96. Cancer Rare- inside nose-mouth. There's lots of good people, but I think that they get blasee-blasee--not quite invested in doing their job. My Dad said that there was an exception for every rule or algorithm. He taught math at UCLA. One person, one tech can really mess things up.
    • Liz
      meyati - getting wrong information out of medical records is incredibly difficult as well. They can't go back and change them, where as in the "olden days" they could cross off the wrong stuff. While I have people put corrections in there, the next person may not read the correction and then does a copy/paste of someone else's history notes and the wrong info again pops up as current. Technology has a long way to go and isn't even close to fool proof.
      • meyati
        At my last visit, my oncologist asked me how I got the image company to remove the erroneous fact that I was a chemo patient, and properly list me as a radiation patient? Liz- they happened to accidently scan the proper area, but they thought that I had cancer in a different part of my face. The last straw was when the reader said that he couldn't completely read the scan, because my face was at the wrong angle. I visited the head of the clinic-and I asked him who set the slant? Santa Claus? Then I showed him the highlighted problems in my last 2 scans- and I told him that I was ready to complain to Medicare about their fraud----That took care of things. Like they kept saying that I had cancer in the tonsil area. My ENT biopsied it, and it was normal flesh-I asked what type of idiot can't look at multiple scans and not realise that nothing changed-the biopsies showed normal tissue. Idiots.
  • Dr. Salwitz, a terrific post! We are discussing these very issues at a physician summit and townhall we'd love for you to provide a talk or discuss some of these issues via webcast or live. The goal of the conference is to restore doctor leadership and push back against some of these burdensome regulations - EMR is one of our main areas of discussion. We are looking for EMR experts such as yourself to discuss ideas for where we could go from here and what doctors' roles should be in this process. Our website: www.letmydoctorpractice.org/conference. Dates: July 20-26, Keystone, CO.

Leave a Reply