2

I am working on mobile research platform, which is meant to be FHIR-interoperable. The idea is to map conventional 12-lead ECG data (in fact, any data set comprised of monitoring data) to FHIR resources but after reading official documentation I still can not figure out some problems:

  1. Concerning multidimensional SampledData in Observation resource, what is the meaning of "If this value is greater than one, then the dimensions will be interlaced - all the sample points for a point in time will be recorded at once."? Does it mean that multidimensional data points are flatenned in space separated list and repeated in some fixed sequence or what?

  2. How is it supposed to handle annotated continuous data sets? Use Extensions/custom Resource type? For now Observation seams to be the only Resource type capable of carrying device data.

For me FHIR looks like the only possible way to bring interoperability in EHR-centered data flows. The question is, whether it can provide this clever data homogeneity and serve as a medium all the way to user with his rapidly evolving tools for self-analysis and control.

Robert
  • 21
  • 2

1 Answers1

1

  1. Yes. So if you have 12 leads, the first 12 values would be the measurements for those leads at time point 1, the next 12 values would be the measurements at time-point 2, etc.

  2. There's no defined mechanism to annotate sampled data, so yes at the moment it would require an extension. (Presumably a complex extension with a timing offset, a lead sequence number and an annotation.) It would make sense to propose this as a standard extension. Continuous feeds must be transmitted in chunks with a separate Observation for each chunk. You can choose the granularity of the chunks.

Lloyd McKenzie
  • 6,345
  • 1
  • 13
  • 10
  • Thank you! Annotation type may require period extension, but for now we can live with it. – Robert Aug 17 '16 at 10:09
  • Is this still the case with Observation that continous AnnotatedECG needs extensions, or has it been updated since 2016? – Larsie May 10 '22 at 06:36