General Points

Several items in Consolidation require elements not required in the EMS dataset, e.g., patient address, which is not always ascertainable prior to arrival at the ED. 1.8.5 & 1.8.8 suggest these can be null, (e.g., “NAV).
Nullable confirmed on SD call
Several items in Consolidation are close to the EMS specification, but require IDs. Should EMS providers generate identifiers (system identifiers, timestamps, or UUIDs)to support this requirement, even though they will not be tracked or referenced by source systems, in order to support wider recognition of the information section by template ID? Or should Null Flavors be used instead?
As above, nullable. Don’t require nulls, as implementers may wish to use the fields.
OK by implementer call 2/14/13
Several Consolidation sections (e.g., History of present illness) contain little or no structure. Is there value in specializing them for more constrained EMS sections?
SD call suggests “yes”; rationale not clear
Would use of the designated LOINC provide value, or would it only create confusion
LOINC values are not hierarchical. Can we assert relationships?
Implementation: see no problem with using extensions.
If an EMS section template specializes a Consolidation template, the new template must use the same LOINC code. LOINC feels that, in most cases, EMS specializations require EMS-specific LOINC codes, prohibiting the specialization of Consolidation section templates. See Medical History, History of past illness, Procedures performed.
If LOINC judges them semantically different, we have a new template
Same issue for questions
As in 3c, can we assert a relationship?
Or, where the difference is only that we have answers, can we make our answers examples rather than normative?
EMS models negation of collections (medications, allergies) in individual questions. E.g., the question “Is the patient on any medications” supports unambiguous understanding of the answer “no.” Consolidation creates a notional medication that is then negated. Should EMS adopt this more standard but less clear convention?
Use both paradigms to avoid confusion.

Header

Some fields required by Consolidation will be null.

Consolidation requires a patient address; NEMSIS does not. Patient address is not always ascertainable prior to delivery to ED.
Consolidation requires Name to have at least one given and exactly one family; NEMSIS names are “recommended.” Patient name is not always ascertainable prior to delivery to ED.
Consolidation requires one author; NEMSIS requires the device, with “crew member completing this report” optional.
Consolidation requires address and telecom for device as well as human author; NEMSIS requires address only for the human author.
Consolidation requires custodian telecom; NEMSIS does not (can be added).

Questions

Document text is generated from structured data. The entry relationship type code is “DRIV” (derived), meaning that the source act (Section) text is derived from the target act(s) (entries), without inversion.

EMS notes:

Add D-side cross references for custodian bits; e.g., dAgency.02 EMS Agency Number for eResponse.01; dAgency.03 EMS Agency Name for eResponse.02.
Guide needs to represent constraints from Consolidation even when not further constraining them in order to cross-reference NEMSIS element codes.
Add general requirement for Text in all sections, with guidelines for generating de-referenced text from entries.
templateID: add root = 2.16.840.1.113883.10.20.22.1.1 to assert header
Assert “DRIV” in section text, except “Narrative”

Section Digest

Advanced Directive / No – unfamiliar structure, different classification
Allergies / No – but we may be able to use drug allergy observation(not environmental due to UNII). Possibly use the same structure without asserting conformance.
Injury Incident / Unstructured – don’t specialize without reason (History of present illness)
Cardiac Arrest / Unstructured – don’t specialize without reason (History of present illness)
Current medication / Probably, if LOINC code is not an obstacle
Past medical history / Unstructured – don’t specialize without reason (Medical (General) History)
Social History / No – question set does not include ‘indicators’ of drug or alcohol use. Also uses SCT in code property.
Vitals / Probably, if LOINC code is not an obstacle
Physical Assessment / Probably, if LOINC code is not an obstacle, and if ‘clinician’ ok (OK, 2/28/13)
Med administered / Probably, if LOINC code is not an obstacle
Procedures performed / Unstructured – don’t specialize without reason (Procedure Description)
Narrative / No analog
Dispatch / No analog
Response / No analog
Disposition / No analog
Billing / Possibly. No overlap now; in future. (Payers)
Exposures / No analog
Protocol / No analog
Scene / No analog
Situation / No analog
Times / No analog

Advanced Directives

“Consolidation Advance Directives Section”and “EMS Advance Directives Section” use the same section LOINC code.

Consolidation

Uses a LOINC code with a typo (“Advanced Directive”)
Uses an observation rather than an act
Constrains the Code property rather than the Value property, leaving the semantics undefined
Uses the HITSP list, consisting of 7 procedures (providing detail far beyond the capability of one code to communicate unambiguously) and one observable entity (ditto), whereas the EMS document captures a directive classified by source type (e.g., “Living Will”); some values also indicate the directive itself (e.g., “Family/Guardian request DNR”).

Options

Do not use this template
Use the template, but leave the Observation Null, recording our own Act using CONS (Consent) act class and NEMSIS type code

While we wish to support interoperability, the divergence between these structures is dramatic. Further, the LOINC code explicitly states its concern is the directive itself rather than the document, the document being the focus of both the NEMSIS value and the underlying regulatory requirement.

Decision: do not use Consolidation template

Question: Can we nullifyConsolidated code value and add EMS modality as dependent observation?

No: they are completely different.

Question: Should the EMS section use the same LOINC code as the Consolidation guide?

No: the Consolidation code refers to the directive itself, not the document.

Implementer call 2/14/13: Seems more straightforward to have our own section. EMS allows multiple (see GForge).

Allergies & Adverse Reactions

LOINC code is for a Document, not a section

Has code, title text: OK

Contains 1:* allergy problem act

Has id, code, status, time: OK (generate, loinc question, active, nullable)

Contains 1:* allergy obs

Has id, status, time, value has type (active, allergy, adverse reaction, drug, food):

OK if we can infertype from child value

Or use top node “Propensity to adverse reactions (disorder)”

Has reaction observation 0:*: omit

Has status observation 0:1: omit

Has severity observation 0:1: omit

Has participant (1:1 to substance CSM)

Has role MANU, has entity MMAT

Drug: brand name or clinical drug fromRxNorm

Environmental or food: UNII

UNII does not contain all needed values

But SCT is wrong axis: can we put child codes in Observation?

Also: missing ant, wasp stings

Use wrong axis for now; request addition

NEMSIS splits this into two observations: drug and environmental; we can fit this in

Consolidation has the following requirements that would require modification of EMS:

Deeper structure, including problem act pattern to capture adverse event observations as well as attested concerns
IDs required for act and observation
Add problem status code = active 2.16.840.1.113883.3.88.12.80.68
Add times to act and observation
Observation value = 20134006Propensity to adverse reactions (disorder)
Unless we adopt more granular SCT codes, if allowed
Add associated manufactured material for allergen

Consolidation has the following requirements, incompatible with EMS:

For non-medicine allergies, the value set is FDA UNII. This set is too granular: it includes, e.g., venom from honeybees, several species of wasp, and yellow jackets, but no general “bee venom” concept.
If we adopt the more granular Observation codes, we can leave these null where necessary
Problem: two codes (ant and wasp sting) are taken from different axes.

EMS has the following requirements, incompatible with Consolidation:

EMS models “no known allergies” as a separate question rather than as a negated observation. We don’t seem to be prohibited from including this additional observation.

Implementation call: no problem with going our own way

Biggest issue: use of UNII for environmental allergies

Q: We could still adopt the structure to support reuse, even if the codes are different

Injury Incident

Similar to Consolidation “History of Present Illness,” but more specific, and structured.

Decision: see general question 3.

Cardiac Arrest Event

Similar to Consolidation “History of Present Illness,” but more specific, and structured.

Decision: see general question 3.

Current Medications

Consolidation “Medications” section has the same requirements as “Medications Administered” that would require modification of EMS:

Medication activity requires an ID, status (fixed), start and stop times

These could be provided: ID as a random UUID (not designed to be reused), status as a fixed value (e.g., current), both time interval range values as the value of the EMS time.

Status = Active

Not clear what effective time means for current medications: recommend Null.

Decision: see general question 2.

SD call: can’t use medication section due to LOINC constraint, but do adopt Consol pattern for negation & null. (Or we could ask LOINC to reconsider.)

Clarify “LOINC constraint”—LOINC code does not suggest Hospital context

I.e., question 4. (2/14/13)

Note: can we encode complications as SCT to fit constraint

Past Medical History

Consider extending Consolidation Medical (General) History, which has minimal constraints.

(Note: Consolidation “History of Past Illness” includes only Problems)

Decision: see general question 3.

Social History

Consolidation Social History observation types are more deterministic than EMS “indicators”; e.g., “Details of drug misuse behavior” rather than “Drug Paraphernalia at Scene.”

Decision: don’t specialize Consolidation template. 2/14/13: correct; follow Advance Directive pattern.

Vital signs

“Consolidation vital signs” and “EMS Vital Signs” use common LOINC codes for common observations.

“Consolidation vital signs” has the following requirements, incompatible with “EMS Vital Signs”:

ID required for each observation (CONF:7300)

“EMS Vital Signs” has the following requirements, incompatible with “Consolidation vital signs”:

Additional observations, outside the scope of 2.16.840.1.113883.3.88.12.80.62 (HITSP Vital Sign Result Type)
Grouping to associate vital signs sets with “prior aid”

The EMS template is structurally similar to the Consolidation version, but it allows multiple organizers—each associated with a “prior care” indicator—and it allows additional observations, some of which also have structure (e.g., Glasgow Coma Score, cardiac rhythm).

2/14/13: recommend reuse, as things that seem to overlap really do.

Physical Assessment

Consolidation Assessment “represents the clinician's conclusions and working assumptions that will guide treatment of the patient,” and is not appropriate.

Consolidation General Status is another candidate, as it contains what might be termed “evidence of distress;” however, it seems to intend more holistic observations.

Consolidation Physical Exam may contain General Status and Vital Signs. Does Consolidation Physical Exam section imply diagnosis?

Does “clinician” exclude EMS personnel?

2/14/13: no position on ‘clinician’; ‘assessment’

Perhaps physical Exam: it also refers to a list of LOINC codes (check congruence) (2/28/13)

Medication Administered

Consolidation Medications Administered has the following requirements that would require modification of EMS:

Medication administration requires an ID, status, times
Change from one entry to one entry per medication
Several additional EMS questions – response, complications, authorization
Otherwise, looks useable

2/14/13: reuse.

Procedures Performed

Consider extending Consolidation Interventions, which has minimal constraints, subject to LOINC code constraint.

Decision: see general question 3.

Patient Care Report Narrative

EMS section is narrative only.

Option: Use history of present illness.

Recommendation: separate dedicated EMS narrative

Dispatch