Page tree

The Medication domain includes a number of related resources

MedicationRequestAn order for both supply of the medication and the instructions for administration of the medicine to a patient.
MedicationDispenseProvision of a supply of a medication with the intention that it is subsequently consumed by a patient (usually in response to a prescription).
MedicationAdministrationWhen a patient actually consumes a medicine, or it is otherwise administered to them
MedicationStatementThis is a record of a medication being taken by a patient or that a medication has been given to a patient, where the record is the result of a report from the patient or another clinician, or derived from supporting information (for example, Claim, Observation or MedicationRequest). A medication statement is not a part of the prescribe->dispense->administer sequence, but is a report that such a sequence (or at least a part of it) did take place, resulting in a belief that the patient has received a particular medication.

Taken from


Vaccines may not be covered (see immunization resource, in which case immunization substances could be removed from the medication codes valueset).

Should we draw the distinction between products and the more abstract substances and say that only products can be administered?

Jim Case noted that vaccine substances are going to be retired (they also exist as products).

Possible use case for using substances for TPN (total parenteral nutrition). Also for post coordinated expression used in "code", however this could still be a type of product if used as an ingredient.

Can we simplify this then and cover all use cases with a binding of:

< 763158003 |Medicinal product (product)|


421967003 |Drug dose form (qualifier value)| is no longer a sensible binding. It only has 15 descendants.

< 736542009 |Pharmaceutical dose form (dose form)| is a preferable binding.

Options for dealing with conflict (from another resource)

See SNOMED on FHIR Meeting (20171017).pptx :

1.Restrict bodySite to [0..0] and require finding site in code
2.BodySite can only be populated if code has no finding site
3.BodySite (if exists) must be a specialization of finding site
4.BodySite must always be a specialization or self of finding site
5.Only allow conditions with no finding sites and include bodySite
6.Any condition and any bodySite


< 105590001 |Substance (substance)|

JR noted that this field could also allow recursive composition with further Medication resources eg for a combination pack. Is this link still valid?

It is not clear what this element refers to - the pack, the amount

Other supporting elements

Dosage.AdministrationSiteAnatomical StructureThis precludes acquired structures, so a more general binding of << 442083009 |Anatomical or acquired body structure (body structure)| might be more helpful.

Note: Interesting situation with anesthesia where the quantity might be - for example - a rate of flow of medication and the denominator would be time.


Refusal reason is a wrong binding.   Those codes in SNOMED relate to the actual medication that was refused.   Perhaps better option to use << 182895007 |Drug declined by patient (situation)|

Note that 242990004 |Drug not available for administration (event)| is another item in that valueset, but it is concerning that we're picking from different sub-hierarchies here.

  • No labels


  1. Here's my summary presentation from the Zoom session on June 5th where the Medication, Medication Request and wider family of Medication-related (and immunization related) resources and their bindings was first discussed.

    1. By way of further clarification of what is meant on Slide 2 by 'repeat prescribing', and the distinction between this and 'repeat dispensing':

      The issue relates to how the law is structured in different countries in respect of controlling prescribing and supply to patients of medications that are be taken for prolonged periods of time. The current FHIR 3.01 MedicationRequest resource - and specifically its MedicationRequest.dispenseRequest.numberOfRepeatsAllowed subelement - appears to assume one model and may not fully support others.

      Archetypal case: patients diagnosed with high blood pressure who are as a consequence required to take an ACE inhibitor drug like lisinopril every day for the rest of their life.

      In most countries the actual supply of such long term medication by a pharmacist is staged into multiple drug issue events, each providing only enough medication for a period of a month or two. This is partly because medicines have a shelf-life and so you can’t safely provide the patient with e.g. a giant box of medication sufficient for the next decade. Its also because restricting the supply to much shorter durations is routinely used by clinicians as a lever to encourage/force patient compliance with whatever monitoring is also required to ensure that the prescription remains safe and appropriate for them: in the case of antihypertensives, for example, that they have a biannual BP check to prevent overtreatment, and an annual blood test to check they have acceptable renal function.

      Of course, for some medication there’s really no practical reason why you couldn’t safely supply patients with much longer supplies e.g. 1 year’s worth of emollient for a chronic eczema sufferer would be entirely safe in very, very many instances. And so the default practice of indiscriminately dispensing equally short supplies of all medications is perhaps also partly because, in many countries, the pharmacist earns an item of service fee for each supply event. There is therefore potentially a financial incentive in play that favours more frequent supply events designed to maximise pharmacy income rather than safety or patient convenience. In the UK, for example, the standard supply interval has gradually dropped over the last two decades from 3 months to 1 month medication. The case for this happening has always been argued as based only on safety but, had their opinion been considered in the decision process, many patients would probably have made a different safety-convenience trade off.

      In most countries the pharmacist can’t supply any medication without a legal prescription - and so the key issue with the MedicationRequest resource concerns this question: how many prescriptions are required to enable more than one supply event?

      In the UK and also, I think, NZ the answer is that each legal prescription document typically - except in limited cases - authorises exactly one dispense/supply event, whereas in the US and OZ a single prescription can authorise multiple supply events: they include an implicit authority to ‘repeat dispense’, whereas until fairly recently there was (almost) no notion of 'repeat dispensing' in the UK legal frameworks.

      In the UK, both doctors and patients are however understandably keen to avoid the routine inconvenience of the patient schlepping to the surgery to see a doctor every 4 weeks purely in order to procure a new signed legal document entitling them to secure one more month’s supply of life-long medication they've already been taking for 5 years, and so an elaborate process of repeat prescribing was set up years ago to reduce the friction as much as is possible within the confines of a centuries old legal framework: the local prescribing record can be annotated with an authorisation that the next, say, 6 prescribing events can be via the practice's ‘repeat prescription’ mechanism.

       Operationally, what this meant until very recently was:

      1. Before their current supply runs out, the patient ticks a box on a tear-off counterpart half of the previous paper prescription document (which they have retained from the last supply event), or they write a hand written note requesting another legal prescription for their medication(s). They drop this off at or post it to the doctor’s surgery.
      2. On receipt of this request for a new prescription, the practice back office staff check on the EPR that there is a repeat prescription authorisation still in force.
      3. If there is, they print one off (but no doctors involved so far…)
      4. If there isn’t a repeat prescribing authority for the requested item(s) - either because there never has been or because the maximum number of 'repeats' authorised has already been issued - then the request is diverted to a doctor for consideration. They will typically check e.g. that the patient still has the condition for which the drug is indicated, and is up to date with the relevant monitoring blood tests or what have you and then either recall the patient for a consult or – much more commonly but provided everything checks out - create a new repeat prescribing authorisation (typically covering the next 6 months) and send the request back to the back office for processing as usual
      5. Each day’s pile of prescriptions that have been printed out under this authorised repeat prescribing process (typically 100 or so pieces of paper) are then presented to that day’s duty doctor to sign individually, with pen and ink
      6. In signing each prescription document at all, the signing physician du jour takes ultimate medicolegal responsibility in law for that prescription, even though they will usually know nothing about the patient on each prescription they’re signing or whether the medication is definitely clinically safe or appropriate. But the repeat prescription authorisation itself is also medicolegally attributable to the physician who granted it (via their electronic system login), and so introduces a grey area in which the signing clinical colleague takes on trust that it is safe for them to prescribe, because the repeat authorising clinician will presumably have done all the necessary checks (and so is also vicariously liable if they have not).
      7. As a result of this bending of the medicolegal rules, the signing clinician can sign 100 legal prescriptions in about 5 minutes, and the patient gets the piece of paper that they need to procure a further supply event within 3 business days of asking for it, and without ever seeing a doctor.
      8. The patient takes the prescription to the pharmacy, and they get one further 'dispense event'. However, that dispense event (for which a fee is claimable by the pharmacy) may in fact be achieved by means of several partial 'dispense supply' events, for example when medication is provided weekly to patients in the form of a 7-day dosette box. Or when the medication has such a short shelf-life that it isn't possible to provide the entire course in one supply (e.g. certain antibiotic liquid formulations) 
      9. Repeat steps 1-8 ad nauseam

      So in the UK there has historically at least been a need to track separately the number of 'repeat prescribing' events remaining before a higher ceremony process will be needed to obtain a single further prescripiton that in turn authorises a single 'dispense event', and the number of 'dispense supply'Tevents remaining before a claimable 'dispense event' has been fulfilled.

      The above process was always a horrible kluge - particularly medicolegally  - and so in 2005 or thereabouts the UK changed its legal framework to explicitly support for the first time a notion of 'repeat dispensing' that's at least potentially much closer to the US sense. Combined with digital prescribing signatures, it opens the door for clinicians to write (and more clearly take medicolegal responsibility for) single prescriptions that authorise many months of claimable 'repeat dispense' events by the pharmacy. But this mechanism remains not yet widely used in the UK, at least not yet: we're in transition to whatever new way of working is best supported by the exciting new world of paperless ePrescribing and digital signatures.

      But digital signatures have also created an entirely different mechanism for skinning the same cat: instead of annotating the prescribing record with a 'repeat prescribing authorisation' of uncertain medicolegal status, or issuing a single 6 month prescription that authorises multiple repeat dispense events, the responsible treating physician can now also simply issue 6 digitally signed prescriptions at the same consultation, but where each is appropriately post-dated. These will then sit electronically in the system until the relevant dates, at which point they beam themselves electronically to the nominated pharmacy for fuilfilmment, and so the patient gets more pills without anybody needing to do anything. And again the UK potentially avoids the requirement for any notion of 'repeat dispensing'...

      1. NZ actually has a hybrid situation. The most common supply period is 3 months, in which case there is only a single dispense event (e.g. ACE inhibitors) and a fresh (repeat) prescription has to be obtained every 3 months. However, for some medications the maximum supply period is restricted to a single month, in which case a single prescription may will permit 2 repeat dispense events. There are also variations for some controlled drugs and those that can be supplied for periods greater than 3 months (e.g. oral contraceptives). No digital signatures permitted at this stage, and the law still states that a prescription must be printed even though we have a National ePrescription Service (NZePS).