Estimated Observations in Chart

[stephen775]

Hi,
Got another clinician request for you...

Apparently they can enter estimated height/weight observations in our patient records for various reasons.

What there asking for. is if an observation is estimated could that be marked on the chart?
So you could have a mixture of actual/estimated values displayed in the chart with different dots.

I'd pass across whether the observation was estimated with the observation value.

Thanks
Stephen

[eatyourpeas]

Thanks for this @stephen775
I will add this to the list of feature requests for version 2 but this involves alterations to the API as well.

[pacharanero]

Thanks @stephen775 - this is an interesting one.

How 'estimated' is 'estimated'? What sort of accuracy do they think they have? I'd be wary of putting data into a chart that wasn't measured since this could potentially mislead.

Clearly we could implement this as some kind of flag which labelled some Measurements as 'estimated'. The display could be in a different colour, use a different shaped marker, and the tooltip could indicate that the measurement was estimated.

This would require a significant change to how the Measurement objects are currently passed into the React component. Currently you pass in an array of the raw Measurement objects as received from the API (or as persisted locally after receiving from the API). We would have to implement either a potentially breaking change to the structure of the data passed into the React component, or we would have to have a 'second pass' of objects which can be added to the chart after the measured Measurement dots are added, similar to how we do Events.

From a clinical safety point of view, I am a bit concerned with letting people put stuff in a chart that is 'estimated' anyway. This sounds like we would be formalising a local Salford workaround in the RCPCH standard national charts. I'd be keen to hear further clinical discussion on this though, and so perhaps @stephen775 you could point the local clinicians to this thread here on GitHub? Issues are very easy to use, it's just like having a conversation on a forum. We can bring the clinical debate here.

Speaking from a clinical informatics point of view, thinking not only about a local installation but of the global picture, if we made such a change it would become essential for ALL instances of the charts, everywhere implemented, to differentiate between estimated and measured. This is because In the future we will likely be seeing seeing persisted data from Measurements being moved around the system, as patients move from one hospital to another. (This sort of thing already happens in GP records with GP2GP and it does cause issues with variances between GP systems and their handling of certain data items)

The clinical hazard here is of Estimated measurements being shown as Measured if some sites didn't implement EXACTLY the same chart display options, or if the 'estimated' flag somehow got dropped in the data transfer process.