Anesthesia SIG (Special Interest Group) Meeting
7th – 8th October 2009
Doubletree Hotel, 8120 Wisconsin Ave, Bethesda, Maryland
Wednesday 7th October 2009
Present:Andrew NortonMelvin Reynolds (GoToMeeting)
Martin HurrellPat Gunter (GoToMeeting)
Terri MonkPippa Norton (GoToMeeting)
Ronald CornetTony Madden (GoToMeeting)
John BastienSteven Dain (GoToMeeting)
Richard DuttonCarol Petersen (GoToMeeting)
Dan Goulson (Q3,Q4)
Apologies:Ron GabelAlan Nicol
David ReichChristine Spisla
Thomas MarshDan Goulson
Q1
Preoperative assessment terminology and requirements
Ronald Cornet summarized the progress made in the Netherlands on preoperative terminology. The dataset of 185 items is considered to be final. Work will now focus on looking at the definitions to ensure they are adequate and appropriately worded. Finalizing this and mapping to SNOMED CT is expected to take until the end of the year.
Terri Monk and Pat Gunter at DukeUniversityHospital, Durham, North Carolina, are currently working on a Domain Analysis Model (DAM) for preoperative anesthesia. Terri suggested that they could help with checking definitions in the Dutch dataset, and with the mapping work.
Martin Hurrell suggested that the Dutch group should also be looking at producing a model which will allow properties and attribution to be structured in a sensible way and tie in with HL7 standards.
This ties in with the work Terri and Pat are involved with at Duke. Andrew Norton is quality assurance lead for preoperative assessment design work for NHS Connecting for Health in the UK. Pat is starting work on a data element list, and feels that it would be very valuable to pull together these datasets from Duke, the Netherlands, and the UK to create an international “superset” to take to HL7.
Pat, Ronald, Andrew and Terri will arrange to meet in Durham on 21st or 22nd October to further discuss this work.
Action points
- Ronald Cornet to send both spreadsheet and PDF files of the Dutch preoperative dataset to Terri Monk and Pat Gunter.
- Andrew Norton to seek consent from NHS Connecting for Health to release their content for this work.
- Pat Gunter to undertake the task of comparing the existing datasets to produce an international superset.
- Pat Gunter, Ronald Cornet, Andrew Norton and Terri Monk to arrange a meeting in Durham in October.
Q2
Update of ISO11073/SNOMED CT Medical Device Communication harmonisation and cross mapping project
Melvin Reynolds explained that the Memorandum of Understanding has been approved by the IEEE and their legal advisors. It has been sent to IHTSDO for approval by their legal advisors. If the proposal is approved, it will allow IEEE to contain some clinical concepts from SNOMED CT, relevant to aspects of device communications. It will also allow SNOMED to include and reference some of the more clinically orientated IEEE device terms. It is hoped that there will be a formal sign-off of this document by mid-November 2009.
One issue to be resolved is potential overlap and conflict between five groups: Global Medical Device Nomenclature (GMDN), Universal Medical Device Nomenclature (UMDN), GS1, SNOMED CT and IEEE. However, there are differences in purpose and scope:
- SNOMED CT aims to allow exchange of clinical information
- IEEE aims to safely exchange device-level information
- GS1 provides barcoding and ID tagging
- GMDN and UMDN aim to exchange information about classification and categorization of devices.
IEEE can carry some GMDN/UMDN information in the messages (though not in the terms themselves), which is where there is the potential for crossover and confusion.
It was agreed that the devices section in SNOMED CT could usefully be tied to another standard, but GMDN is becoming less viable as a recommended standard, and some of its definitions are now out of date. Steven Dain mentioned ISO 121 and 4135 anesthesia vocabulary and asked where this might fit into the overall scheme of things. It was felt that a liaison between IHTSDO, IEEE and ISO 121 would be useful, and Steven Dain agreed to act as the principal contact in ISO 121.
Melvin Reynolds will send Andrew Norton the terms that relate to Continua devices for standards already published. Andrew will ask Ian Arrowsmith (Chief Terminologist, NHS Connecting for Health) to let the Anesthesia SIG view any requirements coming from Continua.
Melvin will extract terms used in existing standards and pass these on to the Anesthesia SIG.
Action Points
- Melvin Reynolds to send Andrew Norton terms relating to Continua devices.
- Andrew Norton to contact Ian Arrowsmith regarding forwarding Continua term requests to the Anesthesia SIG.
Q3
Anesthesia Quality Institute
Dr Richard Dutton, recently appointed director of the Anesthesia Quality Institute (AQI), gave an overview of the background and aims of the AQI.
Quality Assurance is high on the agenda now in the US, and there is an ever-higher demand for performance data and for regulatory documentation. One of the problems is that often those who produce the measurement standards and quality indicators do not have a real understanding of perioperative anesthesia, and therefore of what is important, what can and should be measured. There is, therefore, a need for a meaningful, appropriate and comprehensive set of quality indicators. This is what the Anesthesia Quality Institute, working under the aegis of the ASA, aims to achieve.
AQI are creating a National Anesthesia Clinical Outcomes Registry (NACOR) for which individual anesthesia practices and hospitals will provide data. January 1st 2010 is the proposed date for the registry to go live and start collecting data. The goal is to extract data from AIM systems to export to the registry. There is currently no standard package for exporting outcomes data. AQI will send out the requirements, and contributors to the registry will have to find out how to get the data and send it. Templates will be developed as data is in sent in.
Terri Monk gave an overview of the Anesthesia SIG, it’s history and development, its current work and aims. The group originally came about due to the recognition of the need for a standardized terminology in order to do meaningful outcomes research. The group is currently working on a HL7 CDA-compliant schema for the anesthesia record. VA hospitals are now working with vendors towards the implementation of SNOMED CT/IOTA terminology.
Terri briefly demonstrated IOTA’s Protégé terming tool, explaining the modelling and use of properties. Some referenced definitions are also included in Protégé. Andrew Norton said that IHTSDO has now agreed to include definition fields in its international releases.
Definitions for these outcome indicators are being developed by the ASA CPOM (Committee on Performance and Outcome Measures).
Q4
Discussion of AQI outcomes
Death
It was agreed that this term is not specific enough and that terms such as “perioperative death” and “intraoperative death” are better.
The term “death” was retained in Protégé as a placeholder for perioperative, intraoperative and postoperative death:
death
perioperative death
intraoperative death
postoperative death
preoperative anesthetic death
Preoperative anesthesia death may be a child of anesthesia-related death, though this would not necessarily be part of AQI data submissions.
“Cause of death” was added to Protégé as a child of “general characteristic of patient”.
Cardiac arrest
This is already in SNOMED and Protégé. The accepted definition was added to Protégé.
Time of cardiac arrest would be dealt with by means of a temporal relationship.
A link could be made to the AACD glossary for procedural times.
.
Preoperative myocardial infarction
The concept may relate to patient history, or could be an outcome. It was suggested that “perioperative” should be dropped from the title and that time and relationship to surgery handled as attributes. The CPOM definition was agreed to be useful in terms of how the definition of perioperative myocardial is made. Further consideration may need to be given for a specific SNOMED term for perioperative myocardial infarction although this would be inconsistent with situational modelling guidance from the IHTSDO content committee.
Anaphylaxis
It was agreed that the definition needs more reference to standard definitions of anapylaxis.
Dr Dutton was asked to convey issues about definition of terms to CPOM and it was hoped that a collaborative relationship could be developed in terms of ensuring that clinical terminology is available to support adequately defined anesthesia related outcomes and quality indicators.
Thursday 8th October
Present:Andrew NortonPhil Dickaty (GoToMeeting)
Martin HurrellPat Gunter (GoToMeeting)
Terri MonkPippa Norton (GoToMeeting)
Dan GoulsonTony Madden (GoToMeeting)
John BastienSteven Dain (GoToMeeting)
Martin Rooney (GoToMeeting)
Tom Marsh (GoToMeeting)
Q1, Q2, Q3
Preoperative assessment terminology and requirements
Time was spent going through preoperative assessment screens in Synopsis IQ and design screens from the NHS. Required terminology was extracted and an attempt made to map this to SNOMED CT and Protégé.
Allergies
There seems to be adequate terminology for this area in SNOMED, although SNOMED does allow specific allergies as precoordinated terms (e.g. “allergy to nuts”, “allergy to pollen”). Open EHR are currently working on an archetype for allergies.
Intubation events
Both SNOMED and Protégé have “failed intubation” and “difficult intubation”. SNOMED also has the term “failed or difficult intubation”, which was added to Protégé as a parent term:
tracheal intubation event
failed or difficult intubation
failed intubation
difficult intubation
Postoperative problems
SNOMED has “pain in throat”, but no specific “postoperative throat pain”. This is a contextual term, so it would probably fall foul of the pre-coordination rules, but it was agreed that it would be a useful term for preoperative assessment to distinguish it from current sore throat, throat infection, etc.
It was noted that the SNOMED term “tracheal intubation morbidity” is a useful term and should be added to Protégé. Some remodelling of the “complication of procedure” area in Protégé needs to be done first though.
“Postoperative nausea and vomiting” exists in SNOMED and Protégé, as does “awareness under general anesthesia”.
Terms for breathing problems are largely there, though there is nothing for opiate-induced apnea. This was added as a new term to Protégé, as a child of “drug-induced respiratory depression”, which was agreed to be a useful “placeholder” term which could hold other drug-induced breathing problems if necessary:
finding of respiration
abnormal breathing
respiratory depression
drug-induced respiratory depression
opiate-induced respiratory depression
Unplanned admission
SNOMED uses “unexpected” (with the synonym “unanticipated), rather than “unplanned”. It was agreed that “unplanned” should also be there as a synonym in each of these terms. Most of the terms were already in Protégé but some rearrangement was made. The “unplanned” synonym was added to the relevant Protégé terms.
Problems with heart or heart rhythm
There was discussion as to whether these terms should be “history of” terms. It would need to be clear that they were due to surgery and anesthesia, as opposed to general “history of”. It was agreed that this could be done using a temporal context value, so that all relevant “history of” terms have a “surgical context” attribute. There is also an issue of how to distinguish between past history that is now treated and controlled, and past history that is still a problem. The NHS screen has a “Further details” text field where this can be specified.
There are already quite a lot of cardiac “history of” terms in SNOMED and Protégé. “H/O: dysrhythmia” was added as a new term to Protégé as a child of “H/O: heart disorder”. It was suggested that some of the descendant terms in SNOMED should be moved.
SNOMED has no specific term for “H/O: heart valve disorder”, although there are lots of “heart valve disorder” terms. This was added to Protégé as a new term.
The NHS screen has “frequency and severity of chest pain” and contains the “angina, class I, II, III, IV” grading. These angina class terms are in SNOMED and should be defined. It was pointed out that having these gradings probably makes the NHS screen’s question about the character of the chest pain redundant.
Synopsis includes the New York Heart Association (NYHA) and Canadian Cardiovascular Society (CCS) grades for heart failure and angina. NYHA grades have been modelled in Protégé; CCS is there but the grading has not yet been modelled. All these terms need definitions.
Terms for “structural disorder of heart” are well covered in SNOMED. It was suggested that “mitral valve prolapse” should be added to the drop-down list on the NHS screen.
The difficulty of dealing with time range was discussed. Terms like “within 6 months” or “in the last year” should generally be avoided as these will only be true in relation to the time at which the data is recorded. Instead, dates should be used, from which a time range can then be calculated. This is the approach used by Synopsis.
Headache after spinal/epidural
SNOMED and Protégé both have “post dural puncture headache”. There is also a new term in SNOMED, “history of headache after dural puncture” and this was added to Protégé as a child of “H/O: disorder”.
Synopsis’ “fainting after spinal/epidural” was felt to be unclear and not a suitable concept to attempt to map.
There was some discussion as to whether “unsuccessful” or “failed” terms should be precoordinated or whether to use the procedure term with a “success of action” attribute. The latter was felt to be preferable. SNOMED has “successful” and “unsuccessful” in “modifier mainly for procedure”. All procedure terms in Protégé already have these attributes. However, it was agreed that it is acceptable to have some precoordinated terms (such as “failed intubation”) where these have historically been used and have particular significance.
Some significant “failed” anesthesia terms were added to Protégé:
failed regional anesthesia (with the synonym “failed neuraxial block”)
failed spinal anesthesia
failed epidural anesthesia
Waking delayed
SNOMED and Protégé both have “delayed emergence from general anesthesia”.
Serious blood loss
SNOMED has “massive hemorrhage” (synonym “massive blood loss”). This will be added to Protégé, but some remodelling will need to be done first.
Extreme postoperative pain
This should be a term with an attribute. Appropriate terms from SNOMED were added to Protégé:
pain
posttreatment pain
postoperative pain
Then “pain” HAS A severity attribute.
Orthopnea
SNOMED and Protégé both have adequate terms for this area. The NHS design screen includes a field for number of pillows, which it was agreed should not be coded- merely recorded as an indicative numeric value.
History of cardiac intervention
Synopsis doesn’t includes questions about coronary artery stents. The options on the NHS screen should include “coronary artery stent insertion” and “coronary angioplasty”. There should also be options for stent types: “drug eluting stent” (in SNOMED) and “non drug eluting stent” (not in SNOMED).
Terminology for Vertebral Canal Haematoma
Andrew Norton discussed some problems in SNOMED, which were highlighted by the Intensive Care National Audit and Research Centre (ICNARC) diagnostic coding system in the UK. “Hematomyelia” appears as a child of “traumatic spinal cord hemorrhage”, but it tends not to be traumatic. Spinal hematoma terms in SNOMED can be confusing (e.g. “traumatic spinal extradural hematoma). Also there is no term for “vertebral canal hematoma” in SNOMED. A suggested model was created in Protégé:
disorder by body site
disorder of spinal region
disorder of spine
spinal hematoma (synonym “vertebral canal hematoma”)
epidural hematoma (synonym “extradural hematoma”)
non-traumatic extradural hematoma
traumatic extradural hematoma
hematomyelia
Q4
SNOMED CT classification of anesthesiology – specialty and occupational terms
SNOMED currently has “anesthesiologist” as a child of “specialized physician”; “anesthetist” as a child of “healthcare professional”, and “certified registered nurse anesthetist” (CRNA) as a child of “anesthetist”.
There was discussion as to whether “anesthesiologist” and “anesthetist” are synonymous, but it was agreed that they are not, and that the SNOMED hierarchy is fine here. The new term “anesthesia assistant” was added to Protégé as a sibling of CRNA, together with a definition.
“Anesthetics” is currently a child of “surgical specialty” in SNOMED. It was agreed that this should not be the case: “anesthetics” should be moved to become a child of “clinical specialty” and a sibling of “surgical specialty”.
“Pain management (specialty)” is also a child of “surgical specialty” in SNOMED, but was added to Protégé as a “medical specialty”. It was noted that it may also belong under “anesthetics”.
Dates of future meetings
There will be two joint HL7/IHTSDO Anesthesia SIG web conferences before Christmas. Proposed dates for these are:
Friday 20th November2.00 to 4.30 pm US Eastern time
7.00 to 9.30 pm UK time
Wednesday 15th December2.00 to 4.30 pm US Eastern time
7.00 to 9.30 pm UK time
Andrew Norton will post these on the collaborative space. He will also arrange for Pat Gunter to have access to the collaborative space and be invited to future meetings.
Action points
- Andrew Norton to submit the proposed new terms and changes in hierarchy to IHTSDO
- Andrew Norton to post proposed meeting dates on the collaborative space
- Andrew Norton to request access to the collaborative space for Pat Gunter
1
IHTSDO Anesthesia SIG October 2009