Thursday, December 29, 2011

2011 Year End Summary


It has been an eventful year.

In January I began working with the ONC S&I Framework on the Consolidated CDA Guide. I was a co-chair of the Documentation workgroup. The Consolidated Guide was published in December 2011 and I am listed as a co-editor. This IG should be one of the base standards for Meaningful Use Stage 2, and I am honored to have contributed, however slightly, to this effort.

I attended the two ONC S&I Framework Face to Face Meetings in Washington, DC.

I changed jobs in May. I am now a Healthcare Solutions Architect for Covisint.

I worked on a state-wide HIE project, HealthShare Montana. I also work on a Michigan based ACO. I have worked on some internal projects, and will be leading an effort to deploy a terminology service to translate local terminology to standard codes before we store this data. I am also leading an effort to deploy messaging implementation guides and computable artifacts to simplify our process for on-boarding new trading partners.

I am company "standards guy", so I participate in the ONC S&I Framework, HL7, IHE and a few other initiatives.

I decided to become more sociable this year, so I joined twitter (@PeterNGilbert) and facebook.

I was recognized as one of the top contributers to the HIT Social Media conversation (#HITsm).

http://www.hl7standards.com/blog/2011/12/20/2011-top-hitsm-contributors/

Looking forward to 2012:

We expect to do work for an existing state-wide HIE in 2012 helping them to on-board their backlog of hospitals and provider offices. This project, alone, should keep me very busy in 2012.

I will be attending the HL7 Working Group Meeting in San Antonio in January. I will see if I can get to the two other WGMs.

I don't know if I will be going to HIMSS in February.

2011 was a good year, and 2012 should be even better.

Tuesday, December 20, 2011

2011 TOP #HITSM CONTRIBUTORS

I am honored to be on this list, but I feel a bit like Groucho Marx: "I wouldn't join any club that would have me as a member." :-)

The HITsm tweet chats are an interesting place where a bunch of us that are passionate about using Health Information Technology (HIT) to improve patient care get together to exchange ideas. The chats will resume on Fridays at noon eastern in 2012.

Here is a link to the story. http://www.hl7standards.com/blog/2011/12/20/2011-top-hitsm-contributors/

Wednesday, December 14, 2011

Terminology Management

Up to now, my company has performed Terminology Management using translate tables in each application. This kind of sort of works, but is cumbersome and prone to error. DocSite has a translate table for each practice, so the code mapping needs to be done over for each practice. Mirth, our CDR, has translate tables for each data source, and so the crossmaps have to be set up for each data source. Also, Mirth will not perform Standard-to-Standard code mapping, so if the code comes in in CPT, Mirth will not translate it to LOINC, for example.

I have a new project coming up where we will have perform code translation as the messages pass through our interface engine, because we will be recieving messages from trading partners (hospitals and clinics) and delivering them to a state-wide Health Information Exchange (HIE).

So, we will use Apelon's Distributed Terminology Service (DTS) and will work with Apelon to create cross maps from local lab codes to LOINC. We will then call DTS from within our interface engine to translate the codes before sending them to the HIE's component systems.

We will also use DTS to perform code set validation. For example, we will have the engine call DTS to ensure that we are actually getting a LOINC code in the field that we are expecting to receive a LOINC code.

I've done code mapping at a couple of previous jobs, so this is a process that I am familiar with. This will be Covisint's first foray into this technology, so it should be fun.

The other thing that should be fascinating is that we will be on-boarding approximately one hundred hospitals and approximately one thousand physician offices for this project. I should be very busy.

Friday, October 21, 2011

Summary of ONC S&I Framework Face to Face Meeting



I attended the ONC Standards and Interoperability Framework Face to Face Meeting in Washington, DC the week of October 17. This is a summary of the meeting and my travels.


I decided to attend the Workgroup leads meeting that started at 3pm on Monday, even though my workgroup was not meeting. The two CDA related workgroups do their face to face meeting during the HL7 Work Group meetings, and so do not attend the ONC Face to Face. The corporate travel system put me on a flight that departed Detroit at 6am, which meant that I left the house before 4am. My brother Rob came out to watch my dog, Sonny, while I was out, which was very nice of him. The flights were uneventful and I made it to the hotel at about 1pm.

The work group leads meeting was interesting. There were many folks that I know through HL7 in the room and it was nice to say hello to them. I met with the Transitions of Care work group leads because the CDA workgroups are part of ToC. Doug Fridsma stopped in at about 5:30 pm and thanked us for our efforts. He gave an entertaining speech about his goals for the initiative.

I went to dinner with the ToC workgroup leads. I am not a huge fish fan, but the shrimp were fine.

The first full day of the conference began with greeting more colleagues that I had not seen in a while. One friend had some good news concerning a project in his state. Two colleagues from my company were also attending, so we divided up the workgroups so that we had some coverage in most of them.

The following groups were meeting:

  • Transitions of Care
  • Laboratory Results Interface
  • Provider Directory
  • Data Segmentation
  • Query Health
  • Electronic Submission of Medical Data (esMD)

The last three groups are relatively new. I attended the Provider Directory meeting in Q1. We are implementing Provider Directory support for a customer that could not wait until the PD workgroup finalized its implementation guidance. We took our best guess at some things and know that we will have to make some changes once the PD workgroup publishes its final recommendations.  My two work colleagues attended the LRI, Query Health and ToC sessions.

I enjoyed a cigar outside during our lunch break.

I switched to the Data Segmentation workgroup for Q2 through Q4. This is a new initiative and I have sat through the introductory material that they presented in Q1 several times. This is an interesting initiative that focuses on implementing patient privacy preferences in EHRs. The initial focus in on restricting access to Substance Abuse (42CFR Part 2) and Self Pay information. We do not implement this well, so I hope to be able to learn from this initiative. We have several state-wide Health Information Exchange projects in production as well as several Beacon Initiatives and ACOs, so we can actually implement this in real world settings and provide feedback on what works well and what does not. We considered initial User Stories, and it was interesting to have experts on the regulations in the room providing feedback.

At the end of the day, many of us gathered in the hotel bar for some refreshment. Doug Fridsma stopped by and chatted with most of us. He explained his vision of moving from Templated CDA to Green CDA to support the PCAST vision of a Universal Exchange Language for healthcare. I blogged about this over the summer, but it was nice to see that I did understand it.

On day two, I met with the Data Segmentation group during the morning. We reviewed the changes that we had made to the User Stories based on yesterday’s feedback. Then folks in the room gave short presentations on how they might be able to assist in implementing the initiative. Since we have operational exchanges, I think we will be able to implement some of this fairly quickly.

I attended a presentation on the Model Driven Health Tools (MDHT) and the upcoming S&I Framework Repository during lunch. I’ve worked with MDHT since early in the year as part of the CDA IG consolidation project. It was nice to finally meet the principal author if MDHT in person after countless hours on teleconferences. It turns out that he lives in Montana and works with one of the provider groups that will be joining HealthShare Montana, which is a state-wide HIE that I am the architect of. It is a small world. The repository demo was interesting, and I look forward to actually seeing it once it goes live.

I met with the Provider Directory group in Q3. They are finalizing implementation guidance for PDs. We implemented IHE’s Healthcare Provider Directory (HPD) support into our ProviderLink product. That should be placed in to production next month. We’ll be listed as a pilot implementation by the initiative.

I left the hotel at about 4pm to get to the airport for my flight back to Detroit. It was raining pretty hard as we left DC. We were delayed slightly getting in to Atlanta, but I had plenty of time before my flight to Detroit. We were actually delayed in leaving Atlanta because our plane did not arrive on time. The flight back to Detroit was uneventful, and I was home shortly after 1am. It was a long day.

When I got to the office, I attended a meeting and discovered that we will actually be implementing  esMD for a customer.

The ONC S&I Framework loves implementers. None of this stuff is any good until we get it out into the real world and make it work.



Friday, September 30, 2011

Even More Stupid CCD Tricks

Here is another one that I just saw.

In addition to the usual missing codes, units, etc., it looks like  they tried to place the observation in the procedure reason. So I read this as we performed this unknown procedure from CPT-4 because we got a result for an improperly coded creatinine random urine test. I suspect that the procedure entry was used instead of the result observation entry, and that they really want to be reporting the urine test result. At least, that's what the narrative block seems to indicate.

<entry typeCode="DRIV">
                        <procedure classCode="PROC" moodCode="EVN">
                            <templateId root="2.16.840.1.113883.10.20.1.29"/>
                            <templateId root="2.16.840.1.113883.3.88.11.83.17"/>
                            <templateId root="1.3.6.1.4.1.19376.1.5.3.1.4.19"/>
                            <id root="36e3e930-7b14-11db-9fe1-0800200c9a66"/>
                            <code code="0" codeSystem="2.16.840.1.113883.6.12" codeSystemName="CPT-4" displayName="UNK">
                                <originalText>
                                    <reference value="#proc5"/>
                                </originalText>
                            </code>
                            <text>
                                <reference/>
                            </text>
                            <statusCode code="completed"/>
                            <effectiveTime nullFlavor="UNK"/>
                            <entryRelationship typeCode="RSON">
                                <observation classCode="OBS" moodCode="EVN">
                                    <templateId root="2.16.840.1.113883.3.88.11.83.15.1"/>
                                    <templateId root="1.3.6.1.4.1.19376.1.5.3.1.4.13"/>
                                    <templateId root="2.16.840.1.113883.10.20.1.31"/>
                                    <id root="36e3e930-7b14-11db-9fe1-0800200c9a66"/>
                                    <code code="UNK" codeSystem="2.16.840.1.113883.5.83" displayName="MERCURY, RANDOM URINE"/>
                                    <text>
                                        <reference value="#Res5"/>
                                    </text>
                                    <statusCode code="completed"/>
                                    <effectiveTime nullFlavor="UNK"/>
                                    <value unit="UNK" value="4" xsi:type="PQ"/>
                                    <interpretationCode code="L" codeSystem="2.16.840.1.113883.5.83"/>
                                    <referenceRange>
                                        <observationRange>
                                            <value xsi:type="IVL_PQ">
                                                <low nullFlavor="UNK" unit="UNK"/>
                                                <high nullFlavor="UNK" unit="UNK"/>
                                            </value>
                                        </observationRange>
                                    </referenceRange>
                                </observation>
                            </entryRelationship>
                        </procedure>
                    </entry>


Wednesday, September 21, 2011

More Stupid CCD Tricks

I got this problem observation from another trading partner.

<observation classCode="OBS" moodCode="EVN">
    <templateId root="2.16.840.1.113883.10.20.1.28" assigningAuthorityName="CCD"/>
    <templateId root="1.3.6.1.4.1.19376.1.5.3.1.4.5" assigningAuthorityName="IHE PCC"/>
    <id/>
    <code code="" displayName="" codeSystem="" codeSystemName=""/>
    <text>
        <reference value="#problem-93333"/>
    </text>
    <statusCode code="completed"/>
    <effectiveTime>
        <low value="20081119000000"/>
        <high nullFlavor="UNK"/>
    </effectiveTime>
    <value xsi:type="CD" nullFlavor="UNK">
        <translation code="466.0" codeSystem="2.16.840.1.113883.6.103" codeSystemName="ICD9CM"/>
    </value>
</observation>

The problem type needs to be specified in the Code element. This is the error from the validator.

Error: HITSP/C83 Conditions Problem Type SHALL be coded as specified in HITSP/C80 Section 2.2.3.1.2 Problem Type. The Problem Type code element SHALL contain a code attribute that identifies the SNOMED CT code for one of the following seven conditions: Finding (404684003), Symptom (418799008), Problem (55607006), Complaint (409586006), Condition (64572001), Diagnosis (282291009, Functional limitation (248536006). See HITSP/C83 Section 2.2.2.7.3, rule C83-[DE-7.02-1].

Error: HITSP/C83 Conditions Problem Type SHALL be coded as specified in HITSP/C80 Section 2.2.3.1.2 Problem Type. The code SHALL contain a codeSystem attribute that identifies the SNOMED CT codeSystem (2.16.840.1.113883.6.96). See HITSP/C83 Section 2.2.2.7.3, rule C154-[DE-7.02-1].

The Problem Type is to be taken from the Problem Type Value Set (Table 2-60).


Concept Code
Concept Name
(SNOMED Fully Specified Name)
Definition
Usage Note
404684003
Clinical finding (finding)
Not Available
Finding
418799008
Finding reported by subject or history provider (finding)
Not Available
Symptom
55607006
Problem (finding)
Not Available
Problem
409586006
Complaint (finding)
Not Available
Complaint
64572001
Disease (disorder)
Not Available
Condition
282291009
Diagnosis interpretation (observable entity)
Not Available
Diagnosis
248536006
Finding of functional performance and activity (finding)
Not Available
Functional limitation


Then the value needs to be taken from the Problem Value set. This is constrained SNOMED-CT.


Description
Identifier
2.16.840.1.113883.3.88.12.3221.7.4
Name
Problem Value Set
Source
Veterans Administration/Kaiser Permanente (VA/KP)
URL
Purpose
Problems and Diagnosis
Definition
See VA/KP Problem List Subset of SNOMED CT
This describes the problem. Diagnosis/Problem List is broadly defined as a series of brief statements that catalog a patients medical, nursing, dental, social, preventative and psychiatric events and issues that are relevant to that patients healthcare (e.g., signs, symptoms, and defined conditions)
Version
20081203
Type
Extensional
Binding
Dynamic
Status
Active
Effective Date
Unknown
Expiration Date
N/A
Creation Date
Unknown
Revision Date
20090331
Code System Name
SNOMED CT
Code System Source
National Library of Medicine UMLS


What they are saying here is that patient has a problem, Acute Bronchitis. They coded it internally using ICD9. But, they don't know the SNOMED-CT code for this concept. They need to provide the SNOMED-CT code for this, which is 10509002.

I wonder if I am the only one that works with vendors that don't get it?


Saturday, September 10, 2011

2011-2012 Hockey

I start playing hockey again this week.

Moosejaw will be playing on Monday nights, this year. The league has four teams and five goalies. They have a goalie schedule and rotation. I will not play every week. I will also play for each of the other teams, once. It will be confusing. We're playing at the Arctic Pond. We played there once before in a different league.

Here is the schedule:

http://www.allprosoftware.net/ARCTIC_45_TEAM/ls96.htm

I begin playing with the Wednesday drop-in group this week. I've played with this group of guys for many years. For a while, most of the best players would end up on one team. They beat me like a rented mule. I finally got fed up and played against them every week. When you are the first goalie on the ice, you get to pick the goal that you will play in. I got to the point where I was beating them regularly, and they asked why I never played for them. I told them that I wanted to face the tougher shots.

Last year, they started drafting teams each night, so I never knew who would be shooting at me.

It will be good to be back on the ice.

Saturday, August 27, 2011

Changing Roles

Healthcare IT is different from other IT in that our customers typically operate 24/7. Thus, our maintenance windows are usually in the wee hours of the morning. You cannot work on healthcare IT any other time.

I've been involved in these early morning sessions for eight years. In my new role as architect, I don't have any real work to do. But, old habits die hard. We performed a maintenance change to a customer's Master Patient Index (MPI) earlier this week. We were able to start that process shortly after 9pm, which is when the last clinic closed. I joined the bridge line and watched the webex as the team worked through the process. I recorded start and stop times for each of the steps in the process and produced a summary for leadership. Even though I wasn't actually at the keyboard doing the work, I felt that just being there and contributing one or two suggestions showed the team that I was part of the team.

This weekend, we are upgrading one of my sites and activating another. The window to perform this work begins at midnight on Sunday morning. I'll try to dial in and listen to progress. I work with a great team, so they will be successful.

ONC S&I Framework Face to Face Meeting in October

This arrived in my inbox yesterday.

I attended the first ONC S&I Framework Face to Face Meeting in DC back in June. I'm not sure that I will get to go to this one, but I will put in the request. I am one of the team leads for the CDA Implementation Guide Documentation Work Group.

Our next S&I Framework is just around the corner.

Make sure you mark your calendars so you don't miss these important working sessions.

See the attachment or contact us via email for more details.

We hope to see you in Arlington in October!

--
ONC October F2F Support Team


S&I Framework F2F Meeting | October 18-19, 2011
Hyatt Regency Crystal City at Reagan National Airport
2799 Jefferson Davis Highway, Arlington, VA 22202

ONC S&I Framework F2F