All posts by TheTraumaPro

Minority Report In The OR

The movie “Minority Report” showed an interesting way to manipulate visual data using hand gestures. It required a special glove and used large transparent display surfaces. Microsoft has helped make this achievement both easy and cheap using their Kinect controller using a combination of visual and infrared imaging.

Now Siemens Healthcare has embraced this technology and developed a hands-off image manipulation system for use in the OR. The Kinect system projects an infrared grid into the room and records them using an offset camera. This allows the system to construct a 3D representation of objects in the room. The Kinect software can identify movements and objects using this data.

Siemens is using special software with the Kinect that allows it to detect and interpret fine movement of a surgeon’s hands in the operating room. The final product will allow a surgeon to browse, pan and zoom relevant patient images while they remain scrubbed and sterile, just by gesturing with their hands. This product will be tested in two hospitals in the near future.

Here’s my prediction: why will we need a big, clunky robotic system interface like DaVinci? Just have the surgeon sit in a comfortable chair, waving their hands to move the laparoscopic camera and instruments. I see especially interesting applications of this technology in military settings and in space!

Reference: Siemens Game Console Technology

EMR vs Trauma flow Sheet: The User’s Perspective

Well, you’ve read me railing against the (current) use of an electronic medical record in place of a paper trauma flow sheet for days. I’d like to share some comments from an end user (nurse) who started using the Epic Trauma Navigator last year:

I am really not impressed with the Trauma Narrator for a few reasons: 

  • Bulky & cumbersome to access during a trauma team activation. Our build team promised us that this would be an efficient model of documentation, however, that has not been the case. It takes more steps to document than before and the output is in so many different places review of the chart is extremely difficult to do. You need to know exactly where to look for this information.
  • While the rest of Epic has a feature that allows for the automatic integration of vital signs from cardiac and NIBP monitors, Epic does not allow this feature in the Trauma Narrator.  All vitals need to be entered manually which can be time consuming. Knowing this up front, I think I would have advocated for not using the Trauma Narrator at all.
  • Vital signs and GCS are not displayed within the same flowsheet in Epic. You can find VS in several places, however GCS are in one specific location and if you don’t have the secret treasure map to find them, you will be searching the high seas of frustration for a long period of time.
  • During our build, there were several requests that were not included in the build. I am told that the once Epic goes live, there is a lock-out of up to 12 months before any “optimization” occurs. My advice to you all who are going to Epic is to be adamant about what you want and ensure it is there before go-live. We are missing small things like “logroll time” and level of activation among other “simple” items.
  • Massive transfusions are difficult to document as you need to address each blood component separately and there are several steps in the process for each component. Again, not a user-friendly system at this point for that.
  • Our training was done concurrently with our build so our training was on a generic template/flowsheet within the Epic playground that did not mirror our live version. This was not at all what our production/end-user system looked like at all, so our employees had to be retrained on the job on how to document with the Trauma Narrator.
  • Order sets are available within Epic, however not all staff use the trauma order sets. This creates confusion and the incorrect items being ordered. Again, bird-dogging is required to assure compliance.
  • Once the patient is “arrived” within the Epic system (aka patient chart is noted as the patient actually being in the ED) you cannot go back and document on the EMS radio/report sheet. Staff need to be diligent to assure that documentation is completed before the patient arrives. We have had scenarios where there have been multiple trauma patients arriving and once the nurse begins the documentation of the trauma team, the ED Charge nurse could not go back and enter the radio report in the proper section. 

… remember, Epic is a documentation tool, and as a tool, it depends on the user. Some will continue to document incredibly well and others not so much.

Bottom line: I have absolutely nothing against Epic. I consider it to be one of the best EMRs out there, and I’ve been exposed to quite a few. As you can see though, even it suffers from many of the input problems I’ve written about in the past. And trauma flow products on other EMRs don’t even come close to this one. So for now, buyer beware! Wait until the input technologies and report capabilities become so intuitive that anyone can use them.

Related posts:

Yet More On The EMR / Trauma Flow Sheet Debate

My opinions about using an electronic medical record (EMR) system for recording trauma activations are well known. I received this well thought out response that I wanted to share and comment on:

My quote: “I defy any of them to come to a trauma resuscitation and rapidly and accurately transcribe all of the information presented, or try to review a PI case based on a printed EMR report.”

Their response: I’m a nurse informaticist, I’ve been a trauma nurse for 15 years, and I review PI cases using the electronic record. I work in an inner city Level 1 trauma center that’s associated with a large academic institution. We have recently implemented an ED Information System in our department and I have documented major traumas electronically. My trauma charts beat my colleagues written flowsheet in accuracy, comprehensiveness, and detail, hands down.

Your blog entry on this topic seems very close-minded. The “flowsheet” is not the silver bullet for trauma documentation. I agree that an EMR can be lengthy but it, by far, surpasses the flowsheet in thoroughness in detail. They both have their pros and cons, why be so quick to pick one? Yes, flowsheet data might be convenient for the reviewer but have you considered the effect on workflow for the frontline staff? An ED that has a comprehensive information system (CPOE, electronic tracking, physician and nurse documentation) must pull out a piece of paper and write on it so that a reviewer can find things easier retrospectively? It seems to me the priority should be the appropriate care for the patient and positive outcomes versus a reviewer being inconvenienced by having to read a long chart.

My response: My main problem with using an EMR to record a trauma activation is that the current human interface technology (keyboard, mouse) do not allow for rapid data entry and movement between different screens of input boxes. If a scribe such as yourself becomes extremely familiar with the system, it is certainly possible to overcome these difficulties with sheer skill and familiarity. However, your response implies that you are the only one capable of doing this. Your colleagues must still use the written trauma flow sheet.

The purpose of the flow sheet is to allow any scribe to record meaningful data that can be used to document patient care and to review and rebuild a complex resuscitation for performance improvement purposes. It is not designed to please trauma center reviewers. But the process the reviewers use to reconstruct a trauma activation is the same one that the hospital’s trauma program must use to dig into the events that occur in a trauma activation. If the input data is faulty because the scribe could not keep up in the EMR or had to enter it later, or if the output is dozens of pages of data that is difficult to sift through, the trauma program manager must spend an inordinate amount of time trying to figure out exactly what happened. The “thoroughness and detail” you mention in the EMR can be a hindrance if the quantity of data eclipses its quality. I have reviewed EMR records with 30 pages in the trauma flow sheet report!

The reviewers look for some kind of trauma flow data that they can use to rapidly rebuild what happened in the trauma room. If the reviewers can’t do it, then the trauma program probably can’t do it either. Neither I nor any of the other reviewers I have worked with have found an EMR trauma flow sheet that matches the utility of paper. Yet. The day will come, but it’s not here yet.

I welcome any additional opinions on this debate. Please leave a comment!

Related posts:

Once Again: Trauma Flow Sheets vs the Electronic Medical Record

There’s been renewed interest among my readers regarding trauma flow sheets and the EMR. This is an update on information from a previous post.

There is a continuing push by hospital administrations nationwide to move toward the use of electronic medical record (EMR) systems in hospitals. In the US, much of this is being driven by the Meaningful Use initiative by CMS. There are a number of benefits from using such systems, including but not limited to:

  • Comprehensive and permanent data collection
  • Easily accessed system-wide
  • Reduction in human errors
  • Increased throughput once the initial learning curve has been completed
  • Multifaceted reporting capabilities

Unfortunately, many hospitals or hospital system IT departments are insistent in moving all charting to the EMR, including the trauma flow sheet. For some, it is a revenue enhancement tool that takes advantage of improved documentation by ED physicians. For others, it is a result of the irrational urge to make everything paperless.

As a trauma center reviewer, I have had the privilege of visiting many hospitals and inspecting their trauma flow sheet charting tools. The bottom line is that I have still not found an electronic medical record system that can replace the handwritten trauma flow sheet.

A trauma team activation is a complex, fast-paced, finely orchestrated performance that does not lend itself well to being recorded electronically. There are two major problems:

  • Accurate and timely data entry (human interface issue: mouse, keyboard)
  • Intelligible reports (report organization problem)

There is so much information being transferred nearly simultaneously (vital signs, physical findings, procedures, fluid volumes given, laboratory and radiology orders, narratives) that it is not possible to record it completely and accurately using any current computer data entry interface or medical record system. Frequently, it ends up being recorded by hand on another piece of paper and is then entered later into the EMR. This is easily spotted by trauma reviewers.

The reporting features of virtually all EMRs allow for a listing of events sorted one way or another. It is rarely graphical in nature, and typically spans many, many pages of text output. Charts that I have reviewed have “reports” ranging from 8 to 20+ pages. It is virtually impossible for a human being to read through this type of output and reconstruct the flow of a trauma resuscitation. In many PI review cases, the trauma program manager is reduced to transcribing the individual data items from the EMR back onto a paper trauma flow sheet in order to conceptualize the resuscitation. Frequently, trauma reviewers identify care problems during the trauma activation that were buried in all the report output.

IT personnel may claim that the problem is an “end user failure.” It’s not. I defy any of them to come to a trauma resuscitation and rapidly and accurately transcribe all of the information presented, or try to review a PI case based on a printed EMR report.

The real bottom line: trauma flow sheets (and other similar code sheets) can not and should not be reduced to electronic data entry. It is not only frustrating, but will hamper the trauma PI process. If the reviewers find that the impact on the trauma PI program is significant, it may result in a PI criterion deficiency and can jeopardize a trauma center’s verification status!

Related post:

Why Do Trauma Patients Get Readmitted?

Readmission of any patient to the hospital is considered a quality indicator. Was the patient discharged too soon for some reason? Were there any missed or undertreated injuries? Information from the Medicare system in the US (remember, this represents an older age group than the usual trauma patient) indicates that 18% of patients are readmitted and 13% of these are potentially preventable.

A non-academic Level II trauma center in Indiana retrospectively reviewed their admissions and readmissions over a 3 year period and excluded patients who were readmitted on a planned basis (surgery), with a new injury, and those who died. This left about 5,000 patients for review. Of those, 98 were identified as unexpected readmissions. 

There were 6 major causes for readmission:

  • Wound (23) – cellulitis, abscess, thrombophlebitis. Two thirds required surgery, and 4 required amputation. All of these amputations were lower extremity procedures in obese or morbidly obese patients.
  • Abdominal (16) – ileus, missed injury, abscess. Five required a non-invasive procedure (mainly endoscopy). Only 2 required OR, and both were splenectomy for spleen infarction after angioembolization.
  • Pulmonary (7) – pneumonia, empyema, pneumothorax, effusion. Two patients required an invasive procedure (decortication, tube placement).
  • Thromboembolic (4) – DVT and PE.  Two patients were admitted with DVT, 2 with PE, and 1 needed surgery for a bleed due to anticoagulation.
  • CNS (21) –  mental status or peripheral neuro exam change. Eight had subdural hematomas that required drainage; 3 had spine fractures that failed nonoperative management. 
  • Hematoma (5) – enlargement of a pre-existing hematoma. Two required surgical drainage.

About 14% of readmissions were considered to be non-preventable by a single senior surgeon. Wound complications had the highest preventability and CNS changes the lowest. Half occurred prior to the first followup visit, which was typically scheduled 2-3 weeks after discharge. This prompted the authors to change their routine followup to 7 days. 

Bottom line: This retrospective study suffers from the usual weaknesses. However, it is an interesting glimpse into a practice with fewer than the usual number patients lost to followup. The readmission rate was 2%, which is pretty good. One in 7 were considered “preventable.” Wounds and pulmonary problems were the biggest contributors. I recommend that wound and pulmonary status be thoroughly assessed prior to discharge to bring this number down further. Personally, I would not change the routine followup date to 1 week, because most patients have far more complaints that are of little clinical importance than compared to 2 weeks after discharge.

Reference: Readmission of trauma patients in a nonacademic Level II trauma center. J Trauma 72(2):531-536, 2012.