I am developing a FHIR converter to map structured data to the TW Core IG . I am currently facing a challenge with incomplete records and would like to seek a solution that allows for data retention. Scenario: Incomplete but Valuable Data In my dataset, certain rows lack mandatory elements like dates or patient identifiers. However, I cannot discard these records because the clinical values (Weight, Height, BMI) are essential for our downstream statistical analysis. date; name; weight; hight; bmi 2026/1/2 ; Joe; 50; 160; 19.5 (missing); (missing); 68; 172; 23.3 The Conflict: While the first row converts smoothly, the second row fails validation due to missing mandatory ( 1..1 ) fields required by the Composition and Observation profiles in TW Core. Since I must preserve this data for research purposes, what is the best practice to represent these missing mandatory elements in FHIR? Should I use Data Absent Reason extensions? 2. Methodology: Template-based Mapping Regarding the conversion logic, I am currently using a "Template-based" approach. For instance, I take a valid Observation-Weight example from the TW Core IG and programmatically inject my dataset's values into the JSON structure, as shown below: Template: Observation.valueQuantity.value = [Template_Value] Logic: Replace [Template_Value] with "68" Is this a correct method? I have attempted to find guidance in the official IG and forums but have not found a clear strategy.(It's possible I don't know the right keywords. :sob: ) I look forward to your kind suggestions. :blush: Thank you so much!

what is the best practice to represent these missing mandatory elements in FHIR? that very much depends on context. Take date, for instance. Unless the observation or condition is a phenotype or genotype, the value changes over time, and therefore it's important to know the date. You don't have a date... how can the value even have any use? Since this question arises, it's up to the community for which the IG represents their consensus to decide whether the data has to be thrown away, or a Data Absent Reason is acceptable. Or something else is done (cause you probably really have a implied range of date here) Should I use Data Absent Reason extensions? there's no general answer to this. You can, but just because you can, doesn't mean that you should. It's a discussion to have within the TW Core community

It may be that the intended scope of TW Core isn't intended to apply to your use of FHIR - most national cores aren't designed to support anonymized statistical analysis, they're designed to support direct patient care.

@Grahame Grieve @Lloyd McKenzie Thank you so much for your kind help and valuable recommendations! Apologies for the delayed response— I’m going to think them over and discuss them with the IG community. Thanks again! :hug: