Wednesday, February 22, 2012

Rethinking #Healthcare management with #SimHospital

So I'm at the HIMSS12 conference in Las Vegas this week, where it's been very inspiring - Meeting lots of #HealthIT and #Informatics people with whom I share a lot of the same passions and thoughts. Breakfast started with Eric Dishman (@ericdishman) of Intel talking about "Big Data", the greying of the world population driving the need for technology solutions to healthcare and importance of disruptive innovation. Biz Stone (@biz) of Twitter opened up the keynote today, talking about opening up information flows and the modern information revolution. The common theme at this conference seems to be : A need for disruptive innovation to help drive development of a more efficient, more personalized, and more distributed healthcare model.

For example, yesterday at the #CMIO forum, after listening to all of us share stories, it became clear that two of the important roles of a #CMIO are to be a thought leader and to be a cautious steward of disruptive innovation that really helps patients

Of course, since I'm passionate about creating #Informatics platforms and well-designed tools that clinicians and administrators can use to work together, I thought : Could I look a step higher, and develop something that helps more than just the clinicians? Could we make something that helps healthcare efficiency on a national level?

And so for inspiration, I turned to a common theme in engineering and development : The three worlds of development.

A. THE THREE WORLDS
Engineers and software engineers are very familiar with this concept - Clinicians, administrators, and politicians generally aren't. 

One of the fundamental tenets of good Informatics is understanding the way ideas come to fruition in a safe and organized way - By moving an idea through three different stages of organized development :
  1. The DEVELOPMENT stage
  2. The TESTING stage
  3. The LIVE (sometimes called "PRODUCTION") stage
Huh? What does this mean, exactly? 

To help develop things in a safe, controlled, and predictable way, most engineers think long and hard about:
  1. How can I DEVELOP an idea safely?
  2. How can I TEST an idea safely?
  3. How can I make something go LIVE safely?
And so, most engineers and informaticists are familiar with these three different worlds and how to use them : 
  1. "DEV" - (the "development" world) - Typically, used by people who help build/develop the idea
  2. "TEST" - (the "testing" world) - Typically used by those end-users who test the idea
  3. "PROD" - (the "live" world) - Typically used in real-life by those people who actually use the idea
So if you train your brain to think about these different stages of organized, controlled development, you will actually be better at developing things in an organized and predictable way.

The funny thing is that often these three worlds are used by engineers and Informaticists, but the rules actually apply to many, many other things we develop in real life, whether we realize it or not. For example, many people live in a house :
  1. "DEVELOPMENT PHASE" of House : Point where builder is building the house according to specifications
  2. "TESTING PHASE" of House : Point where safety inspector looks at house design and tests for adequate safety
  3. "PRODUCTION/LIVE PHASE" of House : Point where person moves into house
Or you might send an email :
  1. "DEVELOPMENT PHASE" - Point where you are writing and drafting the email
  2. "TESTING PHASE" - Point where you are proof-reading the email and checking the spelling
  3. "PRODUCTION/LIVE PHASE" - Point where you click "SEND" and make the email a reality
Or you might send a paper letter to someone :
  1. "DEVELOPMENT PHASE" - Point where you are writing and drafting the mail
  2. "TESTING PHASE" - Point where you proof-read the letter before putting it in the envelope
  3. "PRODUCTION/LIVE PHASE" - Point where you drop the letter in the mail to make it a reality
So I usually recommend to new Informaticists that they should become familiar with these three worlds, and :
  1. Who uses which world for what?
  2. What process will you use to transfer ideas from one world to the other?
It's also why I personally feel that some older healthcare change concepts like 'Test of Change' are kind of outdated - They only encourage crossing over from one world to the other without a formal process. 

Again, a good Informaticist understands these three worlds, how to use them, and helps an organization define the standards by which tools will be moved through these three stages of development. Generally, with regards to Health IT development :
  1. Software engineers/analysts live in the "DEV" world (often with help of an Informaticist)
  2. Owners/End Users live in the "TEST" world (often with the help of an Informaticist), and 
  3. Real-life people live in the "LIVE" world.
B. THE CONCEPT : SimHospital

So to help HIMSS and the ONC drive some really innovative thinking about bending the healthcare cost curve, I wondered - Could we actually use these common engineering/informatics principles to help more than just software engineers and informaticists? In other words, could we use these principles to help patients, administrators, clinicians, and politicians understand healthcare better? How would we do that? 

So it dawned upon me, a great tool that could help improve healthcare management and delivery would be a robust TESTING GROUND for healthcare change. Enter the idea : SimHospital.

SimHospital would be a computer-modeled, virtual hospital where all of the basic characters in healthcare could live in a safe, virtual environment that allows for testing. Just like the popular  SecondLife world or TheSims series, it would be a virtual hospital with virtual-reality avatars that are built to behave much like their real-life counterparts - E.g. virtual patients, doctors, nurses, pharmacists, respiratory therapists, couriers, and other hospital staff could all be designed to behave in fairly predictable manner, based on certain variables like :
  1. Education/training
  2. Allowed tasks
  3. Predicted compliance with tasks
  4. Clinical tools and communication to facilitate interactions between team members
  5. Contracts and Policies to guide behaviors
And I think in this virtual, simulated world, we could allow better testing of ideas like :
  1. How will changing a policy or regulation impact care?
  2. How will changing a clinical or administrative tool impact care?
  3. How will changing a workflow impact resources?
  4. How will adding/removing a department impact workflows?
This virtual world would also be an amazing training tool for clinicians, administrators, and politicians - If we commonly ask pilots to train hours in a flight simulator, maybe this SimHospital could be used to train healthcare leaders to understand their environments better.

It could also, if developed, be used as a tool to help do predictive modeling for healthcare outcomes - If the ACO movement is going to make organizations responsible for both the delivery and outcomes of healthcare, then SimHospital could be a very useful tool to predict the outcomes of a particular intervention.

And if we wanted to expand beyond the boundaries of the hospital, we could also develop SimHealthcare to model the hospital and outside PCPs and specialists, again, to help predict how a change in one or more variables will probably lead to what results.

I think it could be a pleasantly disruptive way of improving education for healthcare leaders and simultaneously help with the predictive modeling that will be required for ACOs to succeed. 

As with many of my posts, I'd like to throw the idea out there, and would be interested in hearing comments. (Do I have any readership from SecondLife or TheSims programmers who want to use their skills to help reform healthcare?) :)

Remember : This blog is for education/discussion and brainstorming only. Your mileage may vary. Always interested in hearing your thoughts and comments!

Sunday, February 5, 2012

Could a new note, the Football, help improve healthcare?

As I've mentioned in previous posts, the American government cannot set up a national patient identifier. So projects like NHIN Direct generally rely on a document push mechanism which, essentially, allows one healthcare provider to push an authorized, HIPAA-secure document to another healthcare provider.

I'll admit it - I wish we could have pull.

The reason I want pull? A centralized, patient-centric medical record (like in the #SpeakFlower model) would make it much easier for various providers to pull and update information in a virtually central location. Pushing documents is going to have its workflow challenges, and leave some with the question, "Where is the patient's real chart?".

So since I recently became involved in our Massachusetts discussion on Health Information Exchange, I'm struggling with the question of how to implement a state-wide HIE system that will allow providers, at least initially, to push documents to eachother.

So my first informatics question, on being given this challenge, is : What will people push? Who will push it? And to whom? And when?

To try to answer these questions, I invited folks to our last Interstate 91 Informatics dinner in January to discuss "Can we do better than SOAP?" by asking everyone :

  1. What documentation do people want?
  2. Could we develop any group standard templates for standardized documentation, to save us all development costs?
  3. Could we develop any rudimentary, area-wide clinical governance so we can share documentation easier, and thus all benefit from a common language?
  4. Ultimately, who will push what documentation to whom, and when?

And after a rousing discussion, the answer I heard was this : Everyone has a different opinion.

I guess it's entirely understandable... ICU docs, PCPs, surgeons, specialists, hospitalists, and everyone else has a common goal - making the patient healthier - but they have different training and thus they all have different needs. This is why when I hear docs say "I just need the important information!", I smile because ultimately, all of the information in a chart is important - It just depends on your context and clinical needs.

So I'm left with the ultimate Informatics challenge - How can we get the right information to the right person in the right place in the right time in the right way? Especially when everyone has a different opinion on what the right information is?

And is there any way we can develop a standard lingua franca that all doctors speak?

Is there something that all docs would know how/when to use, in a standard way?

THE CHALLENGE

So to better understand the challenge here, I looked to the most common issues I hear doctors, nurses, and administrators talking about :

  1. Med Reconciliation (at virtually every stage of care)
  2. Handoffs inside a hospital
  3. PCPs wanting notification that their patient has been admitted
  4. PCPs wanting discharge summaries when their patients are discharged
  5. Quality
  6. Waiting times
And given the push mechanism it looks like we are going to get, at least initially, how are we going to set any standards?

There is one thing issues #1-6 above share : They are mostly all caused by the lack of a common, portable, #SpeakFlower-type, patient-centered chart, which we currently lack in modern private healthcare. (Note : I say private healthcare because the Veteran's Administration/VA VistA system actually has a pretty seamless, continuous, portable patient chart that only works inside the VA system for various political and cultural reasons...) 

But in a private, push world, is there any way we could we start to approach some kind of a portable, patient-centered chart?

In other words, is there any way we could leverage our push system in a way that actually simulates a patient-centered chart?

And how would we implement this?

THE CURRENT STATE

Looking at the current buffet table of documentation, it's no wonder that every doctor has a differrent opinion of what they need. There aren't really any hard standards for clinical documentation. As I've mentioned in previous posts, most doctors learn about documentation from things like the Washington Manual Internship Survival Guide. So as a result, most physicians are familiar with things like :
  1. Admission H&P
  2. Progress Note
  3. Discharge Summary
  4. Transfer Note
  5. Encounter Note
  6. Procedure Note
  7. Visit Note
  8. Consult Note
And so when our Interstate 91 Informatics group got together, it's no wonder every doctor had a different opinion of which note they would want to get, and when.

So to look for inspiration on how to build a standardized document that every doctor would know how to use, and when to push to whom, again I thought : Could we make a standardized push document that approaches the portable, patient-centered chart we all want?

THE INSPIRATION
It dawned upon me that to solve this problem, we will need a new type of note. And so if it's something that's not in the Washington Manual Internship Survival Guide, it would have to be something that was so useful, so intuitive, and so desirable - like McDonalds French Fries - that every doctor would *want* to use this note, update it, and push it to the right person at the right time.

So then I thought - We are really asking for a portable, mini-chart that we can push around to the next provider.

And then I wondered, "What will we name it?" The "Mini-chart"? The "Patient Summary"?

What we're really talking about here is a "Patient Handoff Note" - The 'mini-chart' - And to make it extra-intuitive, I've decided to nickname it "The Football".

(Interestingly - "The Football" is also the nickname given to the "Nuclear Football" which the President of the United States carries around at all times, which according to Wikipedia is designed to be "a mobile hub in the strategic defense system of the United States" - A portable, role-centric tool for making important decisions... Huh! Talk about portable documentation!)

Also by nicknaming it "The Football", it gives users a visual clue about how to use it and when to punt it to the next physician.

THE PATIENT HANDOFF NOTE ("FOOTBALL")
The Patient Handoff Note ("Football") is basically a patient mini-chart, designed to be used in handing off care from one physician to another. In other words, physicians could think of the Patient Handoff Note ("Football") as a document that they update and push to the next physician expected to see the patient.

Who is the next physician expected to see the patient? Whoever is expected to see or cover the patient next. If you're a PCP expecting a specialist to see your patient, you'll update the football and send it to the specialist. If you're a specialist done with the consult, expecting the PCP to see the patient next, you'll update the football and send it back with the patient to the PCP

Of course, the key word here is expected - What if a patient has an unexpected trip to the ED?

I thought the note should be of such high value that, on arrival, the ED physicians would request the Football from the PCP. (By doing this, they would ensure the PCP knew about the visit.) And when the ED doc decides to admit the patient to the Hospitalist, they would update the football and push the patient and football to the expected Hospitalist.

And the admitting hospitalist could update and push the football to the expected hospitalist the next day. 

And the daytime hospitalist could update the football and push it to the expected overnight covering staff. 

And the overnight covering staff, if needed, could update the football and push it to the daytime hospitalist.

And the daytime hospitalist, on discharging the patient, could update the football and push the patient and football back to the PCP.

(To the patients reading this, I apologize - This is really referring to document management, not patients - I am definitely not endorsing pushing patients around!) :)

So anyway, back to the football - what could this Patient Handoff Note ("Football") look like?

Here's my first draft - As an example, I'll show how it could be used at the point of discharge :

[ DRAFT ] PATIENT HANDOFF NOTE ("FOOTBALL")
PATIENT NAME   :   VADER, DARTH A.
DATE OF BIRTH   :   Jun 06 1966

Emergency Contact :
Tarkin, Emperor
Relationship : Father
Cell (914) 555-1212

CODE STATUS : 
Full Code (last verified by Luke Skywalker, MD, PCP, Internal Medicine, Oct 30 2009)

DATE OF HANDOFF :
Feb 03 2012

HANDOFF FROM :
Dirk Stanley, MD (Hospitalist, Internal Medicine)

EXPECTED HANDOFF TO :
(   ) Overnight Coverage 
(X) Other : Luke Skywalker, MD (PCP, Internal Medicine)

AUTHOR : (Note : This is who is pushing the football today)
1. Feb 03 2012 - Dirk Stanley, MD (Hospitalist, Internal Medicine)

CO-AUTHORS : (Note : this is essentially everyone who has pushed the football in the past, with last date they pushed it, in reverse date order)
2. Jan 28 2012 - Han Solo, MD              (Attending, Emergency Medicine)
3. Sep 22 2005 - Beru Whitesun, MD    (Attending, Gastroenterology)
4. Apr 02 2004 - Luke Skywalker, MD (PCP, Internal Medicine)
5. Apr 01 2002 - Ben Kenobi, MD        (PGY-1, Internal Medicine)
6. Feb 22 2002 - Owen Lars, MD         (Attending, General Surgery)
7. Jan 11 1996 - Leia Organa, MD        (Attending, Cardiology)

ALLERGIES :
1. Mar 29 2002 - Bactrim (Rash/Hives)

PMHx/PSurgHx : (Note : This has all problems/history identified in reverse date order)
1. Feb 03 2012 - Aspiration Pneumonia
2. Feb 25 2002 - Cholecystitis, s/p cholecystectomy
3. Sep 22 2005 - Colonoscopy, s/p benign polyp removal
4. Jan 11 1996 - CAD s/p NSTEMI, no catheterization, medical management
5. Oct 12 1994 - Hyperlipidemia
6. Apr 03 1992 - HTN

SIGNIFICANT STUDIES : (Note : This is noted by docs, again in reverse date order)
1. Jan 28 2012 - 2-view Chest X-ray = (R)LL patchy infiltrate
2. Jan 04 2004 - PSA=0.06

WHAT I DID :
Patient admitted to Mos Eisley Hospital on 1/28 with cough, fever, purulent sputum approx 3d after being found asleep and intoxicated at a party. Chest X-ray showed (R)LL infiltrate, WBC=21k, PMNs=80%. Started Zosyn IV and after 3d patient improved. Changed to oral Augmentin on 2/2/2012. Now ready for discharge today 2/3/2012.

ACTIVE MEDS (AT TIME OF HANDOFF) :
1. Lisinopril 5mg PO daily
2. ASA 81mg PO daily
3. Metoprolol 25mg PO 2x/daily
4. Simvastatin 40mg PO daily
5. Augmentin 875mg PO 2x/day x7d, to complete on Feb 09 2012

TO-DO LIST :
1. Feb 15 2012 - PCP to follow-up with patient for routine follow-up visit
2. Mar 01 2012 - PCP to repeat Chest X-ray to ensure resolution of pneumonia
3. Apr 01 2012 - PCP to repeat lipid panel and LFTs to monitor Simvastatin dose
4. Apr      2015 - Gastroenterologist to repeat colonoscopy to follow-up benign polyps 
5. Jan       2020 - PCP to give repeat Tetanus vaccination 

SIGNED : __Dirk Stanley, MD_(Hospitalist, Internal Medicine)______ Date : Feb 03, 2012     
                                     
(My apologies to George Lucas - I'm obviously a big fan - Hope you don't mind me using characters to demonstrate this new medical note...!)

Anyway, I think the advantages of this drafted Patient Handoff Note ("Football") are this :
  1. It would be a very high-value note that docs would look and ask for (like McDonalds French Fries!) when receiving a patient :)
  2. After receiving the football from another physician, it makes creating your local documentation much easier.
  3. After receiving the football from another physician, it makes it very easy for you to update the football for the next provider.
  4. By making it something all doctors expected, it would drive ownership of the note by all physicians, so...
  5. ... It encourages docs to own, review, and continuously update the full med list, problem list, to-do list, allergy list, etc
  6. It makes med reconciliation easier for everyone.
  7. It could virtually replace notes involved in the expected transfer of care such as the transfer note, overnight coverage signout, discharge note, and consult referral
  8. Nicknaming it "The Football" makes it fairly intuitive about its importance and who to push it to and when
  9. In a push environment, in an unexpected transfer of care, an ED doc or Hospitalist requesting this from the PCP would pretty much ensure the PCP was notified about the admission in a timely basis.
It's definitely an off-of-the-beaten-path idea, but I'm going to suggest it to my fellow physicians here in Massachusetts, as we start to warm up our state-wide HIE and get it running. Will let you know the results!

Is this note wishful thinking, or just crazy? Always interested in feedback and questions! Send me your thoughts and ideas! Love the discussion just for education's sake!