search

opensafely / ics-research

This is the code and configuration for our paper, Inhaled corticosteroid use and risk COVID-19 related death among 966,461 patients with COPD or asthma
https://opensafely.org
0 stars 1 forks source link

*OTHER* Oxygen Therapy #19

Closed annaschultze closed 4 years ago

brianmackenna commented 4 years ago

I have created a oxygen NHS dm+d medicines list here.

GPs do not prescribe/supply medicines via the normal FP10 prescription route. They are supplied via Home Oxygen Order Forms (known as HOOF) which are written by specialists but may also be written by GPs (you may hear reference to HOOF part A in this context). Camden CCG have more information here oxygen prescribing. Consequently I'm pretty sure Oxygen will be recorded in the GP record in a variety of ways 1) via a non-issued medicine (notebook abpve will suffice) 2) HOOF forms will be recorded as paperwork 3) Clinical codes CTV3

I think we will need a strategy to identify products whichever way it is recorded locally.

annaschultze commented 4 years ago

I searched both SNOMED and QoF for "Oxygen Therapy". For QoF there were no hits. For SNOMED I followed the following process:

  1. I extracted all results
  2. I removed duplicates
  3. I removed some terms that were clearly irrelevant, specifically:
    • hyperbaric oxygen therapy (for diving sickness)
    • oxygen therapy assessment not indicated
    • wound therapy related procedures (i.e. - Wound therapy oxygenator dressing (physical object))
  4. I removed children terms if all children under a given parent were considered relevant (i.e, the parent was "oxygen therapy"), so that the highest level codes possible are included only.

There were some I was not sure about. For example, I wasn't sure if "Supplier reported volume variance between prescribed and used supplemental oxygen per day (observable entity)" would be relevant (in which case, the parent term "Oxygen delivery (observable entity)" can be included). I also wasn't sure if products and/or assessments/education should be included or not.

Where I was not sure I left the codes in; these will need to be checked by a clinician as this list is likely way too broad.

@ianjdouglas is awaiting a Read2 code list for this, I believe.

Oxygen_Therapy_SNOMED.xlsx

ianjdouglas commented 4 years ago

Thanks @annaschultze. Just to confirm the list I got hold of was for products rather than Read codes so not applicable here

brianmackenna commented 4 years ago

Pragmatically we have decided to set this list aside due to data quality issues.

We may return to it for future study but closing for now.