Wednesday, July 20, 2011

IHE Week and FDA announces mHealth regulatory approach

Another IHE Week, this time preparing Trial Implementation Profiles. I'm on the hook for Reconciliation, and we have another couple dozen comments to go.  One of the surprising responses from commenters is that the profile should REQUIRE that external identifiers for problems, medications and allergies be preserved.  That is going to result in a quite a bit of new text (now I have to explain what changes the identity of an entry).  I'm pleased with this response.  We made it a strong recommendation, but not a requirement because I felt that many EHR vendors would not implement the profile if it were required.  Most systems would need to alter their basic tables to record the external identifier.  But experience with prior implementations indicates that this really is the best way to handle it, and I certainly agree with the sentiment.

I spent a good bit of time with IT Infrastructure on Monday discussing Cross Enterprise Document Workflow (pdf).  This profile is critical for Patient Care Coordination, even though it is coming out of IT Infrastructure.  Here's a brief pitch I'm giving on it tomorrow for another group:
  • In Ambulatory Care, providers are desperate for Workflow Management to track referrals, orders, and manage quality of care.
  • But their workflows are loosely coupled and ill-defined.
  • These need to integrate with well defined, tightly coupled workflows in a departmental system (e.g., imaging).
  • XDW uses industry workflow standards to describe Human Tasks that can be well integrated across both settings.
The replacement of CDA with Human Task as the standard to manage the tasks is more than appropriate, and @rjhorniii has done a great job leading the discussion, and getting me and one of my colleagues to agree on an approach.  You can expect some significant changes to the public comment version to come out of this meeting.

While at the meeting, the FDA came out with a proposed regulatory approach for mHealth devices and applications.  I haven't had a change to do more than skim it once.  I'll look over it in greater detail later.  One of the tweople I follow expressed surprise at the exclusion of Mobile devices being used as an EHR.  His interpretation was that EHRs were not medical devices.  In case you are curious, they also excluded EHRs from the Medical Device Data System rule.  It's not that EHRs aren't medical devices.  It's that FDA carefully classifies things so that something doesn't fall into competing classifications.  They may be issuing separate guidance on EHR systems, so they exclude anything that can be viewed as an EHR from the other rules and guidelines.  That way, when the EHR rule comes out, it will be clear WHICH regulations and procedures apply.

Tuesday, July 19, 2011

If you want to change a culture, start with its children

My kids are brilliant. If you don't believe me, just ask my wife.

-- Anonymous
Well, they are. Let me give you an example. Weekend before last, my youngest daughter (9) woke up on Saturday morning with an earache. I heard her crying in the living room and went to see what the problem was. She told me it felt like someone put a bowling ball in her ear. She had been swimming the day before and couldn't get all the water out, so I was pretty sure she had an ear infection. I called the pediatrician, but they don't open until 9:00 am on Saturday, so I called my practice's urgent care center, and they opened at 8:00am. Since it was about a 40 minute drive, I gave her some ibuprofin to help with the pain, and off we went.

To make the ride easier for her, I engaged her in conversation. She wanted to know what was going to happen at the office, and I explained how it worked. You didn't need to make an appointment, you just showed up. They would take her history, and the doctor would take her vital signs, record them in the EMR (she knows what that is because it is what daddy does), and then look at her ear, and probably prescribe some antibiotic and maybe some ear-drops to help with the pain. Then I told her that since her doctor is a "Meaningful User" of HIT, she could get a print-out of her record. Next I explained that everyone has a right to get a copy of there medical records, and that by law, in most cases, the doctor or the hospital has to give them to her (or in her case, her parents). And then I explained that since her doctor was already in Meaningful Use, she would not have to wait a long time to get them.

"We can ask them for the record, do you want it?" I asked her. "Yes, I want it." She told me. "When we're done, you can look at it. And if there is anything wrong with it, you can tell them to fix it. They have to respond to you, and tell you at least if they are going to or not." I told her. "It's the law. There is this thing called HIPAA (not HIPPOs), that makes sure you can do that." "OK." She said.

By the time the visit started, her ear only felt like it had a baseball in it (or so she reported). And during the visit, I asked the doctor if he could print out her record for her. He said he would as soon as he finished writing it up. He then dug out his iPhone and started calculating her dose. I asked him what app he was using for that. He told me, and I laughed, and said, I have that same app on my iPad. "Are you a practioner?" He asked. No, just a healthcare IT geek with an iPad I said. He computed her dose, order the antibiotics, and we talked about whether he should prescribe ear-drops for pain management. "Given that she woke up crying," I said, "let's go ahead and do that."

So he placed that order too. He then handed me a printed set of patient instructions regarding her treatment that described her problem and how it was being treated, and what we should do if it didn't work. He came back a few moments later with the completed visit report and hand it to her (not me -- which was the whole point for me). We left to go to the pharmacy to pick up her meds.

On the way there, she read through her record. She asked me questions about what stuff meant. When we got to the vital signs section, she said: "Daddy, it says here that they took the Oxygen off my right hand, but they used my left." So, I told her that she was good to spot that and that she needed to write them a letter to tell them about the mistake. I figured that this was a great way to drive home the lesson.

When we got to the pharmacy, we had to go through the "Insurance Card" rigamarole because I gave her nickname rather than her legal name to the doctor, and they didn't have her on file (or so they thought). We fixed that. Then the pharmacist told me that the insurance was going to cover the anitbiotics, but not the ear-drops. "OK, thats fine," curious, but not interested in battling with the pharmacist. He then went to fetch an "Ear-wax removal kit". I looked at it and said, "No, that's wrong." I asked my daughter to run out to the car to get her records. She ran out and back, and we showed the pharmacist what the doctor's records had said. Then he showed me the e-prescription that they recieved. There was a mismatch. The drug listed in her chart was "Ear Drops (some chemical name with percentage in solution)", but what was on the printed version of what had been recieved via the ePrescription message just ended before the parenthetical clarification, just: Ear Drops.

So my daughter learned that having her records helped with stuff like that, too.

Then I called the office, to tell them about the mixup, but they apparently don't have a procedure to deal with that kind of error report. I know who to talk to there, so I'll fix that later.

Last Sunday, while I was reading, I heard my daughter working out the letter with her mother that she is going to send to the place that saw her. I'm going to write one too and send it with her letter. We'll address it to the HIPAA compliance officer, so that I know it will get the appropriate attention.

My daughter won't forget this lesson, and she's learned something exceedingly valuable that will help her when she is responsible for her own healthcare, something that not many people are really aware of. Last Saturday, I got to spend an hour with @Lygeia and several other patient advocates at Health Foo. We talked about how to get patient engagement. I recalled parts of this story to the group, and pointed out that by the time a person leaves high school, they are able to balance a check-book and write a resume, but they are not at all trained to deal with doctors, choose an insurance policy, or manage their own interactions with the healthcare system. These are necessary life skills that we need to teach our children. Because if you want a society to change, you need to start with its children. So, you might think that its funny that ONC is going to use cartoons to teach Health IT, but I really don't. I think its brilliant, just like my daughter.

-- Keith

Update: Not to be outdone by her sister, my eldest went to the pediatrician's this morning. Diagnosed with Swimmer's ear and Otis Media. She's picking up an electronic copy of her records mid-week on disk! When asked why she wants them, "I wanna be an e-patient!" she said to them. I think we need to declare a week for e-pediatric patients.

Monday, July 18, 2011

The HIT100

After a couple of weeks of anticipation, the HIT100 tallies are finally out.  I was very grateful to finish in second place, just behind Brian Ahier in a very close race.  Michael Planchart e-mailed me his results spreadsheet so I could post it here.  Others (Nate Osit for one), have also posted the list on their blogs.

And just as he finished, theEHRGuy posted a new set of instructions for selecting the #HIT5 and #HIT1.  I've extracted them from the twitter stream below:

Just when you thought the fun was over now we have to choose our #HIT5 and #HIT1 #healthIT heroes! Instructions below:

  • Those that scored above 12 will have an opportunity to have their place in #HIT5 and #HIT1. The top #HealthIT ranks. 
  • Those that scored above 12 must abstain from voting. Please do not re-tweet votes since they will be ignored. 
  • Please do not reply to votes. 
  • The tweep that gets most votes from this round wins the #HIT1 place in history. Others fall in the #HIT5 according to their scores. 
  • Vote like this: #HIT100 My #HIT5 vote goes to @twitterhandlehere #HealthIT
  • Deadline for voting (according to a recent tweet by @theEHRGuy) is Thursday, July 21st.

I've made it easy for you.  In the list below, if you click on the link, it will take you to twitter with a correctly formatted vote.  Please follow Mike's instructions to make it easy for him to do the final tally.  I'm pretty sure Twitter folk don't fit the profile of people who like to wait.


PositionTwitter HandleNominations
76sonalpatel_HIT4
76poikonen4
76meducate4
76marxists4
76lostonroute664
76Jsamplesgmt4
76JohnSharp4
76jimtate4
76HealthIT4
76HealthEugene4
76harrygreenspun4
76GrahameGrieve4
76GovHITeditor4
76EMRDailyNews4
76EHRoutlook4
76drjosephkim4
76dirkstanley4
76danamlewis4
76cyberslate4
76cristianliu4
76CorepointHealth4
76blogbrevity4
76billians4
76Anthony_Guerra4
76anitasamarth4
63RossMartin5
63RMartinBoulder5
63richelmore5
63omowizard5
63lisacrymes5
63IngaHIStalk5
63HITNewsTweet5
63hitconsultant5
63GailZahtz5
63fierceEMR5
63erikpupo5
63ElinSilveous5
63Andrewspong5
52todd_park6
52ShimCode6
52rilescat6
52RichmondDoc6
52realworldhl76
52jkvedar6
52jbselz6
52HIMSS6
52gnayyar6
52cindythroop6
52BarbaraFicarra6
40WTeeGreen7
40MicrosoftMD7
40MedicalQuack7
40kevinmd7
40KentBottles7
40JessPKahn7
40jasoncolquitt7
40fredtrotter7
40FairCareMD7
40Egrunden7
40bobcoffield7
40acowatch7
33westr8
33tedeytan8
33pjmachado8
33ftvelasco8
33ekivemark8
33EdBennett8
33CLOUDHealth8
29PhilBauman9
29john_chilmark9
29HITExchange9
29histalk9
27jonmertz10
27HealthyThinker10
26jimmyweeks11
22VinceKuraitis12
22HITAdvisor12
22healthblawg12
22aviars12
18HealthStandards13
18faisal_q13
18Cascadia13
18boltyboy13
15susannahfox14
15Lygeia14
15Farzad_ONC14
13ReginaHolliday16
132healthguru16
12jhalamka17
10MatthewBrowning18
10ej_butler18
9Perficient_HC20
8janicemccallum21
7EMRAnswers22
6lizasisler23
4NateOsit26
4ePatientDave26
3TheGr8Chalupa31
2motorcycle_guy35
1ahier36

Based on comments by @omowizard and @ej_butler, I equalized rankings of those with equal votes.

Saturday, July 16, 2011

What's a Health Foo

This morning I saw a tweet / four square checkin from Brian Ahier (@ahier: "Health Foo" at Logan Airport (BOS) with 69 others. I responded: What is health foo and how long are you in Boston. We spent a bit of time together at HIMSS last year and the year before and since I live nearby, I thought it would be a good idea to get in touch. He responds back, "Healthcare Unconference .. leaving Sunday". I wonder to myself how a healthcare unconference happens in Boston and I don't hear about it.

Later that same day, Nate DeNiro (@unclenate asks me to follow him (I thought I already had been, so I did) so he could DM me. He also sent me a link on more information about Health Foo. I get a later invite from him to the event (it is invitation only).

So, I shifted my one call for the day up an hour, and headed into Cambridge, back to an office I used to work in more than a decade ago.

What is an Unconference? Well, to answer that question, you have to think about what a conference is. A bunch of people pay to go hear other people that might be interesting speak about topics they might be interested in, and the topics are chosen by the conference organizer. An unconference twists that around. First of all, O'Reilly Media is sponsoring this event, and Microsoft is making the space available, so attendees don't need to pay to get in (but they do need to be interesting). Secondly, other than time blocks, nothing is preset.

So, we spend several hours talking, grabbed a beer from the foo bar (Yes, it is called that), and introduce ourselves to each other. Then we have a short formal session where we learn what foo is about, how it got started, and how to do it. Foo it turns out, stands for "Friends of O'Reilly (Media)", and the whole thing started during the .COM bust when O'Reilly had a bunch of unused space that they wanted to do something interesting with.

After the formal session, we went out to develop the agenda. THere were sticky notes and several large boards where you could describe a session you wanted to run, and put it in a time slot and a room. I decided to do a session called "Is there a standard 4 that?", basically to help various folks share their knowledge about IT standards and their possible uses. If you are on twitter or buzz or G+ or FB tomorrow, I might call on you to help, because we are looking to crowd source answers to questions I may know nothing about on standards. So I put my sticky up. Then @Lygea posted one at the same time opposite mine that I wanted to attend, so I moved it to a different slot. Then @ePatientDave did the same thing, so I moved it again, but this time after Fred Trotter's session (that I'll be helping him with) on What is the Health Internet.

The foo goes on from 10:00 to 8:00 pm, at which point their with be several 5 minute Ingnite sessions. I have no clue what they are, but I'm cooking something up for one anyway. I'll report on that tomorrow.

While I'd love to hang out with this crew all weekend, unfortunately, I have family committments, and so won't be able to spend Sunday morning with them. But even so, it will be very interesting. While there were probably a dozen people at Health Foo that I already know (some quite well), there are many others completely outside my usual comfort space. That's the whole point. Get interesting people and ideas together and see what flows.

On a side note: I had an interesting conversation with @Farzad_ONC at the meeting about a blog post that I basically wrote with about a three person audience. He remembered it better than I did. That rocked, since he was at the top of that list.

Thursday, July 14, 2011

News around the HL7 CDA

The CDA Consolidation project reached a major milestone today as the Structured Documents Workgroup voted to approve the ballot reconciliation for posting.  This reconciliation addressed over 500 comments since the HL7 Working Group meeting in May, just less than two months ago.  That's a major amount of work.  There's still quite a bit left to do, and another ballot forthcoming that could result in a similar volume of work.  Even so, congratulations are in order.

As part of that, one of the things we finally resolved was how to deal with how to say that the patient is on no medications, or that it is unknown if the patient is taking any medications.  We based that off of work that I reported on here in the last couple of days.  We did make one minor modification which I will correct on those pages tomorrow.  In stating that it is unknown whether a patient is on medications, I incorrectly stated that the medication is unknown, as Grahame Grieve pointed out comments two days ago.  We fixed that by replacing that unknown with a general code indicating "medication" (and the same would be true for allergies and problems).

We also developed a more clear statement around the requirements for locations of entries in a section.  The Consolidation guide requires "direct" containment of the required acts within the section, whereas previous guides required them to be in the section or in any subordinate section.  That was clarified, and we further agreed that while they must be contained within an entry contained directly within the section, we did not require them to be immediately within that entry.  The rationale for this is to support enhancements such as those found in the IHE Reconciliation profile which I've discussed previously.

A couple of days ago the ONC Transitions of Care Project developed its initial ecosystem consensus statement recommending CDA Release 2 as the standard to use to convey information necessary for Transitions of Care.  That consensus statement has yet to be finalized (it should be by next week), but I'm hopeful that it will go forward pretty much as stated (There are a few things I want to review, and one suggestion that might be contentious).

Finally, there are several things happening at the same time that should enable closer cooperation between IHE and HL7 on the development of IHE profiles.  The "Activities with other SDOs" report that I give on Sunday at the HL7 WGM as the HL7 assigned liason to IHE will still occur.  A new initiative will probably happen at the next HL7 that will help HL7 and its partners like IHE to develop goals and objectives around those collaborations.  I look forward to that discussion.

Wednesday, July 13, 2011

Starting new Religion?

I spent some time yesterday on a call regarding yet another government agency's interest in interoperability in Healthcare IT.  It seems, with the advent of meaningful use, interconnected healthcare IT applications and devices, that Interoperability has become the new religion.  As with any new "fad", a lot of previously uninvolved organizations are trying to see how they can capitalize upon it and the existing stakeholders are looking at how they should do so also.

There are a few things that I would hope by now we all realize, but they bear repeating, because quite honestly, I don't think some of the newcomers are aware of this yet.

  1. The industry doesn't need another definition of interoperability.  There is a standard one (actually several).  If your organization doesn't know what it is, it should definitely reconsider getting involved in a leadership position, because it hasn't been paying attention.
  2. The industry doesn't need another organization in the US to work under (unless it is an umbrella organization like the Canadian Standards Collaborative).  On the International front, there is already the Joint Initiative Council
  3. There are already more initiatives than any one person can reasonably keep track of.  If you cannot fit what you are trying to do in one of the existing initiatives, see what can be done to change that with the organizations sponsoring the existing work.  Many of them provide opportunities to create projects, including HL7 and IHE which are pretty easy to initiate.  You just need to find the right champions.
  4. In healthcare IT, the number of "experts" in the field is pretty limited (and thus, so is bandwidth).  This is evidenced by how you see most of the same faces over and over again at Healthcare IT standards (and related) meetings.  It's a pretty tight community, and most of us are connected either directly or indirectly through just a few hops.  Little goes on that we don't find out about pretty quickly.  The number of little birds that tweet (or mail, or plus, or call or ...) about this stuff just makes it impossible to keep anything under wraps for too very long.
Some strategies to think about for newcomers into Healthcare Standards:

  • Partner with existing players in the space.  Find out who they are and what they are doing.
  • Do some research and find the thought leaders already in the space.  Get them engaged and through them, get your projects moving forward. If you don't know who to ask, ask me.
  • Don't be afraid to enter into the development process with existing organizations.  If you truly have leadership potential, you will soon find yourself in that position. 
  • If you have critiques about the organization, share them directly with thought leaders in that organization in a constructive fashion.

Some strategies to think about for the existing stakeholders (e.g., IHE, DICOM, HL7, ASTM ...)

  • When you hear about a new "initiative", engage (and encourage them to engage) in a way that takes advantage of existing organizations, processes and expertise.
  • Be ready, willing and able to form coalitions with other organizations, and to engage with them in ways that meet their needs.
  • Listen to outside criticism, whether it is constructive or not (organizations that don't listen to criticism eventually don't survive).

Forming a new initiative is costly in time and effort.  You need to develop processes, governance, and infrastructure to make it work.  It takes time to develop these, and that's time that could be better spent solving the real problems. Starting a new project under already well established processes and governance with an organization that has the infrastructure to make it work is much easier, and you will find a ready pool of engaged volunteers from which to draw expertise.

Tuesday, July 12, 2011

How to say no...

Please see the updated Bestiary, the information on this page is outdated.


Yesterday I posted a long write-up describing how to say I don't know.  Following up on some questions from yesterday, I'll explore how to say no.

First, let's summarize the bestiary we need to address:

  1. Patient is not on THIS drug.
  2. Patient is not on ANY drugs.
  3. Patient does not have THIS ailment.
  4. Patient does not have ANY ailment.
  5. Patient does not have THIS allergy.
  6. Patient does not have ANY allergy.
Two patterns here:

  1. Patient does not have / is not on a specific thing.
  2. Patient does not have / is not on any known thing.
Now, I didn't address this when dealing with unknown, because this is knowledge.  Forget the hedge implicit in "Patient has no known allergies."  No known is NOT the same as unknown, and is there to remind the reader that what is known does not necessarily cover every possibility that could occur.  I didn't know I was allergic to bee stings until I was exposed to them.  A close friend developed an allergy to wheat products later in life, possibly as a result of some other illness.

Patterns 1 and 2 above correspond very closely to patterns 2 and 3 from yesterday.

Pattern #1
A specific thing is not present.

Specific Problem Not Present
<entry>
<observation classCode="OBS" moodCode="EVN" negationInd="true"> <code code="64572001" displayName="Disease" codeSystem="2.16.840.1.113883.6.96"/> <value xsi:type="CD" code="38341003" displayName="Hypertensive Disorder" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT"/> </observation> </entry> Specific Medication Not Present <entry> <substanceAdministration moodCode="EVN" classCode="SBADM" negationInd="true"> <consumable> <manufacturedProduct> <manufacturedLabeledDrug> <code code="81839001" displayName="anticoagulant drug" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT"/> </manufacturedLabeledDrug> </manufacturedProduct> </consumable> </substanceAdministration> </entry> Specific Allergy Not Present <entry> <observation classCode="OBS" moodCode="EVN" negationInd="true"> <code code="106190000" displayName="Allergy" codeSystem="2.16.840.1.113883.6.1"/> <value xsi:type="CD" code="300916003" displayName="Latex Allergy" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT"/> <participant typeCode='CSM'> <participantRole classCode='MANU'> <playingEntity classCode='MMAT'> <code code="111088007" displayName="latex" codeSystem="2.16.840.1.113883.6.96" codeSystemName="SNOMED CT"/> </playingEntity> </participantRole> </participant> </observation> </entry>

Pattern #2
Nothing is present.

No Known Problems
<entry>
 <observation classCode="OBS" moodCode="EVN" negationInd="true">
  <code code="64572001" displayName="Disease" 
   codeSystem="2.16.840.1.113883.6.96"/>
  <value xsi:type="CD" code='64572001' displayName='Disease'
   codeSystem='2.16.840.1.113883.6.96' codeSystemName='SNOMED CT'/>
 </observation>
</entry>

No Known Medications
<entry>
 <substanceAdministration moodCode="EVN" classCode="SBADM" negationInd="true">
  <consumable>
   <manufacturedProduct>
    <manufacturedLabeledDrug>
     <code code='410942007' displayName='Drug or Medicament' 
codeSystem='2.16.840.1.113883.6.96' codeSystemName='SNOMED CT'/>
    </manufacturedLabeledDrug>
   </manufacturedProduct>
  </consumable>
 </substanceAdministration>
</entry>

No Known Allergies
<entry> <observation classCode="OBS" moodCode="EVN" negationInd="true"> <code code="106190000" displayName="Allergy" codeSystem="2.16.840.1.113883.6.1"/> <value xsi:type="CD" nullFlavor="NA"/> <participant typeCode='CSM'> <participantRole classCode='MANU'> <playingEntity classCode='MMAT'> <code code='413477004' displayName='Allergen or Pseudoallergen' codeSystem='2.16.840.1.113883.6.96' codeSystemName='SNOMED CT'/> </playingEntity> </participantRole> </participant> </observation> </entry>

The close relationship of these negated assertions to yesterday's unknowns is no accident. Negation is one end of the scale of knowledge about a thing, a positive assertion is the other end, and unknowns fit right in the middle along an axis of uncertainty.

Tommorow, I won't be writing about how to say yes, or at least I hope I won't. To do that, you'd simply follow pattern #1 above, and use negationInd="false" (which is what the default value SHOULD be in CDA, but isn't.) Grahame's assertion that lack of negationInd specifies that negationInd is NULL, while technically correct, is A) not obvious, and B) dangerous, especially in the presence of assumed defaults. The most common way of reporting observations in CDA doesn't require the use of negationInd at all, and the assumption is that it is set to false. This should be added as an errata for CDA Release 2.0. We'll fix that in CDA Release 3.

Keith


This post was updated on July 18th to reflect adjustments made by the HL7 Structured Documents Workgroup based on feedback from Grahame Grieve. The key change was in using a general code rather than unknown for pattern 2. A similar change was made for Smells like Unknown...