Medication amounts and units

We should maybe revisit the current modelling patterns for medication amounts and their units. Currently, we have amounts expressed as a DV_QUANTIY with units constrained to ‘1’, and an additional DV_TEXT element for the unit. This undermines the DV_QUANTITY data type, and doesn’t work particularly well when using form builders. This pattern is currently used in these archetypes (that I can find):

I suspect this modelling pattern was originally created to get around some tooling issue regarding templating units, but I think that’s been corrected now. I think we should at the very least unconstrain the units of all DV_QUANTITY elements modelled like this, and possibly also removing the corresponding DV_TEXT unit elements.

1 Like

Yes . It was due to Quantity originally not supporting non-UCUM units.


I’m glad you agree! I’ve created new revision branches for each of the archetypes:

There are additional change requests for some of these, but can we go forward with these changes, in addition to any resulting from the change requests?

1 Like

I’ve a similar question relating to units. We’ve a need to describe a quantity of Grays (Gy) in centigrays cGy) and also to a lesser extent volts (V) in kilovolts (kV). We’re using dv_quantity, but there doesnt seem to be an option for these two, or a prefix option on dv_quantity. I think Uuom notes the c (centi) and k (kilo) as prefixes. Obviously there’s a mathematical operation to get to kilovolts from volts etc. Is there a clean way to do this with dv_quantity?

1 Like

Reposting to a new thread on tooling - Adding to the DV_QUANTITY units options

1 Like

I’m going forward with these proposed changes, since there has been no negative feedback in the month that has passed since the above was posted. The INSTRUCTION.medication_order and CLUSTER.medication and CLUSTER.dosage all have open and unresolved change requests though. Could I ask those who have submitted change requests or participated in the discussions about them to go back and try to get them resolved? (@heather.leslie @ian.mcnicoll @varntzen)

1 Like

I scanned the CRs. Most seem uncontroversial. We currently only implement the medication cluster (for covid vaccin and wound treatment medications). But at first glance I don’t see any problems with this proposal or the CRs. Keeping the archetypes aligned with FHIR would be valuable to us, since our medication management uses FHIR (and wrappers around that for persistence). And we’d like to integrate with/migrate to openEHR one day.

There is already very high alignment with FHIR in themed models, and I agree we need ot keep that close, where possible.

One issue that we might want to try to resolve


My preference is to keep ‘Route’ at the current level but also add it lower down the tree to be associated with specific dose instructions, which is the gap we have now, mostly for fairly obscure Mental health examples. The vast majority of mixed-route instructions are readily handled by the current top-level Route (not supported by FHIR).

Whatever we decide, this is the time to do it!!

Do we have these examples available? Seeing them would help a lot with understanding how to model this well.

From Christine Wadsworth who isa Pharmacist and Better’s UK Clinical Lead.

Olanzapine oral vs IM may have the same dose ranges and maximum daily dose but would have different instructions e.g. If olanzapine IM is given then must wait >1 hour before administering lorazepam

Haloperidol oral vs IM – different dose ranges and different maximum daily doses

Aripiprazole oral vs IM – different doses and instructions

which FHIR supports and we do not)

In more common cases we might have Paracetamol 4-6 hourly Oral or IM i.e same dosage instructions for all routes (which we support and FHIR does not).

When this was discussed with Better meds team felt that the Mental health examples were best handled as separate medication orders. The FHIR approach is more elegant for these use-cases but quite a bit more burdensome for the more common multi-route / same instruction scenario (IMO)

There is an argument for doing nothing, of course!!

1 Like

I used paracetamol 2x500mg oral OR 1000mg iv (in case of nausea) all the time at my ward of people recovering from bowel surgery. I always hated that I had to enter them separately because it’s so easy for the nurse not to pay close attention and to give both, which is dangerous.


This is really the same dosage though, and INSTRUCTION.medication_order supports any number of routes for the same dosage. Do you really need to specify 2x500 mg (oral) vs 1000 mg (IV)? Where I’ve worked clinically, this exact example has always been done as one order, like this: “1 g paracetamol, oral/IV”.

I would suggest we postpone this decision, since there seems to still be some uncertainty? In any case, the solution would be to add a route element to CLUSTER.dosage, which is a non-breaking change?

Clinically I don’t care. But tablets are only available in 500mg and iv is only available in 1000mg. And I’m told I should prescribe only dosages that are available.

Maybe this is a jurisdiction thing. Here, doctors prescribe what they think is a good idea for a dosage, and it’s up to the nurses to figure out how to make it work in practice :woman_shrugging:

1 Like

Hahah I would love that for our jurisdiction too XD. But how does it work for tablets? It’s pretty hard to administer 546 mg of paracetamol of the tablet only comes in 500mg dosage? And how does it work if the patient (at home) is responsible for taking the medications?

1 Like