The basic pattern for reporting a problem uses the CDA <observation> element, setting the classCode='OBS' to represent that this is an observation of a problem, and the moodCode='EVN', to represent that this is an observation that has in fact taken place. The negationInd attribute, if true, specifies that the problem indicated was observed to not have occurred (which is subtly but importantly different from having not been observed).

The value of negationInd should not normally be set to true. Instead, to record that there is "no prior history of chicken pox", one would use a coded value indicated exactly that. However, it is not always possible to record problems in this manner, especially if using a controlled vocabulary that does not supply pre-coordinated negations, or which do not allow the negation to be recorded with post-coordinated coded terminology.

The specific observation being recorded must have an identifier (<id>) that shall be provided for tracking purposes. If the source EMR does not or cannot supply an intrinsic identifier, then a GUID shall be provided as the root, with no extension (e.g., <id root='CE1215CD-69EC-4C7B-805F-569233C5E159'/>). While CDA allows for more than one identifier element to be provided, this profile requires that only one be used.

The <code> describes the process of establishing a problem. The code element should be used, as the process of determining the value is important to clinicians (e.g., a diagnosis is a more advanced statement than a symptom). The recommended vocabulary for describing problems is Value set CZ-CodeProb. This value set is required when used within the CZ Problem Concern Entry

The <text> element is required and points to the text describing the problem being recorded; including any dates, comments, et cetera. The <reference> contains a URI in value attribute. This URI points to the free text description of the problem in the document that is being described.

A clinical document normally records only those condition observation events that have been completed, not observations that are in any other state. Therefore, the <statusCode> shall always have code='completed'.

The <effectiveTime> of this <observation> is the time interval over which the <observation> is known to be true. The <low> and <high> values should be no more precise than known, but as precise as possible.

While CDA allows for multiple mechanisms to record this time interval (e.g. by low and high values, low and width, high and width, or centre point and width), we are constraining Medical summaries to use only the low/high form.

The <low> value is the earliest point for which the condition is known to have existed.

The <high> value, when present, indicates the time at which the observation was no longer known to be true. Thus, the implication is made that if the <high> value is specified, that the observation was no longer seen after this time, and it thus represents the date of resolution of the problem.

Similarly, the <low> value may seem to represent onset of the problem. Neither of these statements is necessarily precise, as the <low> and <high> values may represent only an approximation of the true onset and resolution (respectively) times. For example, it may be the case that onset occurred prior to the <low> value, but no observation may have been possible before that time to discern whether the condition existed prior to that time.

The <low> value should normally be present. There are exceptions, such as for the case where the patient may be able to report that they had chicken pox, but are unsure when. In this case, the <effectiveTime> element shall have a <low> element with a nullFlavor attribute set to 'UNK'. The <high> value need not be present when the observation is about a state of the patient that is unlikely to change (e.g., the diagnosis of an incurable disease).

The <value> is the condition that was found. This element is required. While the value may be a coded or an un-coded string, the type is always a coded value (xsi:type='CD'). If coded, the code and codeSystem attributes shall be present.

The value set to be used when this template is specialized for describing adverse reaction is CZ-ReactionAllergy.

The Value Set used is CZ-IllnessandDisorder, with the !!!! OID 1.3.6.1.4.1.12559.11.10.1.3.1.42.5. !!!!

One of the concepts from the Value Set CZ-AbsentOrUnknownProblem shall be used in the <value> element to indicate the known absence of problems, or that problems are not known.

The <originalText> element within the <code> element described above is used as follows: the <value> contains a <reference> to the <originalText> in order to link the coded value to the problem narrative text (minus any dates, comments, et cetera). The <reference> contains a URI in value attribute. This URI points to the free text description of the problem in the document that is being described.

The contained entry describes a subjective assessment of the severity of the condition as evaluated by the clinician.

The contained entry describes the current status of the condition, for example, whether it is active, in remission, resolved, and so on ...

An optional <entryRelationship> may be present referencing the health status of the patient, e.g., resolved, in remission, active. When present, this <entryRelationship> element shall contain a clinical status observation conforming to the template (1.3.6.1.4.1.12559.11.10.1.3.1.3.10). The value set to be used is eHDSIResolutionOutcome, OID 1.3.6.1.4.1.12559.11.10.1.3.1.42.30. The typeCode shall be ‘REFR’ and inversionInd shall be ‘false’.

Optional <entryRelationship> elements may be present providing an additional comments (annotations) for the condition.