February 24th, 2013

Solve This Problem Please — Microbiology Results in Electronic Medical Records

Our hospital and affiliated practices have had electronic medical record (EMRs) of some sort for decades, so I’ve had my chance to try my hand at multiple “platforms,” both commercial and home-brew.

(Weirdly — and I kid you not on this — a version of the first iteration from the 1980s is still around, running parallel to a more modern program. That old one remains the best at displaying simple things, like quickly showing a patient’s creatinine or white blood cell count.)

But one thing all these EMRs have in common is that they are pretty lousy at rapidly displaying microbiology results. I certainly see the problem — microbiology reports are a melange of obscure words, diverse numbers, and inscrutable abbreviations. And finding important results in a complex patient with dozens (or hundreds or thousands) of cultures is a major challenge for even the most tech-savvy clinician.

With current EMRs, it’s as if all the work that went into designing text macros and special “functionality” (cringe) had thoroughly drained the programmers’ brainpower, so by the time they got to the microbiology part they just gave up. I can just imagine a dialogue between team members as they faced writing the software for microbiology, realizing that there might be more to an EMR than just fast ways to spit out boilerplate text and nifty graphs:

Software Engineer 1:  Hey, look at this — the clinician can now enter a complete normal physical examination just by typing “alt-PE”. And if they hit “control-alt-F7”, then it inserts three paragraphs that document a review of systems, patient education/counseling, and plans for follow-up — a guaranteed billing up-code.
Software Engineer 2:  Cool!  And watch this — we can graph this patient’s MCV, MCH, and MCHC going back to 1983 in 3 dimensions and using 9 colors. Nifty MP3 audio file adds a “swoosh” sound when the graphic appears.
Software Engineer 1:  Awesome, great work! (Takes a swig of Red Bull, eats a few Doritos.) Hey, did you decide what to do with this information? (Hands over a piece of paper with the following printed on it.)

Specimen: Wound
Date collected:  February 16, 2013
Date reported:  February 18, 2013
4+ PROTEUS VULGARIS GROUP      VITEK AST-GN53 CARD
      Antibiotic                      Result
      ----------------------------------------------
      Amikacin                        <=2      S
      Ampicillin                      >=32     R
      Cefazolin                       >=64     R
      Cefepime                        <=1      S
      Cefoxitin                       <=4      S
      Ceftazidime                     <=1      S
      Ceftriaxone                     <=1      S
      Ciprofloxacin                   <=0.25   S
      Ertapenem                       <=0.5    S
      Gentamicin                      <=1      S
      Levofloxacin                    <=0.12   S
      Meropenem                       <=0.25   S
      Nitrofurantoin                  128      R
      Trimethoprim/Sulfamethoxazole   <=20     S
      Unasyn                          8        S    

    4+ ESCHERICHIA COLI      VITEK AST-GN53 CARD

      Antibiotic                      Result
      ----------------------------------------------
      Amikacin                        <=2      S
      Ampicillin                      <=2      S
      Cefepime                        <=1      S
      Cefoxitin                       <=4      S
      Ceftazidime                     <=1      S
      Ceftriaxone                     <=1      S
      Ciprofloxacin                   <=0.25   S
      Ertapenem                       <=0.5    S
      Gentamicin                      >=16     R
      Imipenem                        <=1      S
      Levofloxacin                    <=0.12   S
      Meropenem                       <=0.25   S
      Nitrofurantoin                  <=16     S
      Tigecycline                     <=0.5    S
      Trimethoprim/Sulfamethoxazole   >=320    R
      Unasyn                          <=2      S

Software Engineer 2:  (Wearily.) Nah … too tired.  (Drinks some Mountain Dew. Gets excited again.) I know, let’s work on building in even more ways to guarantee that people will forget their username and password!

Maybe I’m not being fair to our friends in IT, but it seems someone should have figured out by now how to make information from the microbiology laboratory more digestible — maybe even searchable! And as our institution is in the midst of a giant shift — one could even call it an EPIC shift — to a new EMR, I’m hopeful we’ll see some innovative work in this tricky area.

Any bright ideas out there?

19 Responses to “Solve This Problem Please — Microbiology Results in Electronic Medical Records”

  1. Igz says:

    The only paradigm shift to intelligently designed (aka not solely optimized for billing) EMRs will come with the adoption of an open, standardized framework to share the same datasets across multiple platforms

  2. Ringer Lactate says:

    We’re moving to Epic also in the next few weeks. I’m anticipating a disaster of colossal proportion. And a guarantee I won’t know where to find the culture results.

    But a normal physical exam?

    Child’s play.

  3. Sous says:

    Common sense will prevail…one day

    I laughed out loud at the “swoosh”, thank you, I needed that after 4 hours of SAS, spot on as usual.

  4. Edtwisy says:

    Have you explored Caisis.org as a solution? Please feel free to contact the team for a demo

  5. robert barker says:

    Nursing notes are even worse. They have degenerated into a grid where the nurse places X’s in little boxes, conveying little or no useful information, except that the patient may still be alive.

  6. General Internist says:

    Having suffered through the conversion to Epic at our institution, resulting in non-informative and inaccurate boilerplate documentation, I wish you the best in your coming click-click-right click-select from drop down-click transition. Now, once I find a culture of interest, I have to double click on a date to see any actual results – then close and double click on another culture, using my short term memory for necessary comparisons. Not good.

  7. Robert Stark says:

    This article was forwarded on to me by my wife. She has heard me gripe about the sad state of EMR’s many times. As a dentist, our software situation is similar. And we are a smaller base than in medicine so there is even less attention for us. But being someone who is involved in dental informatics and Electronic Healthcare Records, I totally agree with Igz above. It’s only when open source initiatives that center around agreed upon standards develop, that we might see some really cool stuff. See http://openehr.org.

  8. Phil says:

    OASIS is what we used and is possibly where the software engineers are having come up with a program that I could have coded better in Word Macro and Visual Basic!

    Seriously though when I see crappy results like that… I just copy paste and zap it into MS Word/Notepad. Control – F and off I go… that is until “privacy” stepped in and disabled the copy function. WTF! Can’t win!!!

  9. Richard Kremsdorf says:

    In the early 1990’s, 3M marketed the Intermountain Health Care/LDS Hospital-developed HELP system, which had all the micro data in coded form and accessible not only for reporting, but also for alerting based on the content. For example, we got alerts if a patient’s culture grew an organism that was not susceptible to the drugs that the patient was receiving. Alas, the market wasn’t interested and the product was sunsetted.

    Even worse, you can be sure that the data exists in coded form in the lab’s system, it’s just when it is shipped out to the EMR that it is converted to a barely useful text blob!

  10. gasman says:

    Sad but very, very true!
    As has been said many times before, this is the inevitable result when highly complex software solutions are written by people who have no idea what the end user really does. Imagine if Word was designed by people who could neither read nor write- the result would be the word processing equivalent of Epic and its competitors.

    EMR’s have been designed by software engineers who have no idea about how medical records are used in highly complex medical environments and have no understanding of interface design, either. What is really scandalous is that the poor design is now being solidified and institutionalized because our government is pouring a bazillion dollars at hospitals to install these substandard systems, enriching the software companies and saddling us with software that complicates our care and diminishes the quality of medical communication.

  11. Pulmonogist says:

    I am sorry to hear that you will be on EPIC soon. We have struggled with the lab reporting of cultures since the inception of EPIC at our clinic in 2008. The results of the culture are visible just the first time they appear in your inbasket (after multiple clicks). Subsequently, the lab tab only states that a final result was published, but it refuses to show it.
    I have spent numerous hours sending e-mail complaints with no success. If you or any EPIC user ever figures out how to make it work, please announce the solution.

  12. JGM-Fuenla says:

    It’s true. Micro results can be so complex to some EMR’s. But we can complicate it even more: what about standardise all those micro records?

  13. Gillian Arsenault says:

    Do you suppose that WHONET’s approach to analysis of bacterial culture results might be a useful model for an EMR lab reporting module?
    http://www.who.int/drugresistance/whonetsoftware/en/

  14. joseph muita says:

    Thank you for sharing your list, it’s a big help for my microbiological research.

  15. Alice says:

    Very true…It’s true. Micro results can be so complex to some EMR’s. But we can complicate it even more: what about standardise all those micro records?

  16. Alice says:

    I like ….OASIS is what we used and is possibly where the software engineers are having come up with a program that I could have coded better in Word Macro and Visual Basic!

    Seriously though when I see crappy results like that… I just copy paste and zap it into MS Word/Notepad. Control – F and off I go… that is until “privacy” stepped in and disabled the copy function. WTF! Can’t win!!!

  17. Alice says:

    Very true….Sad but very, very true!
    As has been said many times before, this is the inevitable result when highly complex software solutions are written by people who have no idea what the end user really does. Imagine if Word was designed by people who could neither read nor write- the result would be the word processing equivalent of Epic and its competitors.

    EMR’s have been designed by software engineers who have no idea about how medical records are used in highly complex medical environments and have no understanding of interface design, either. What is really scandalous is that the poor design is now being solidified and institutionalized because our government is pouring a bazillion dollars at hospitals to install these substandard systems, enriching the software companies and saddling us with software that complicates our care and diminishes the quality of medical communication.

  18. Alice says:

    True…..In the early 1990’s, 3M marketed the Intermountain Health Care/LDS Hospital-developed HELP system, which had all the micro data in coded form and accessible not only for reporting, but also for alerting based on the content. For example, we got alerts if a patient’s culture grew an organism that was not susceptible to the drugs that the patient was receiving. Alas, the market wasn’t interested and the product was sunsetted.

    Even worse, you can be sure that the data exists in coded form in the lab’s system, it’s just when it is shipped out to the EMR that it is converted to a barely useful text blob!

  19. Alice says:

    All will be fine…..Having suffered through the conversion to Epic at our institution, resulting in non-informative and inaccurate boilerplate documentation, I wish you the best in your coming click-click-right click-select from drop down-click transition. Now, once I find a culture of interest, I have to double click on a date to see any actual results – then close and double click on another culture, using my short term memory for necessary comparisons. Not good.

HIV Information: Author Paul Sax, M.D.

Paul E. Sax, MD

Contributing Editor

NEJM Journal Watch
Infectious Diseases

Biography | Disclosures | Summaries

Learn more about HIV and ID Observations.