*DISEASE*: osteoarthritis and rheumatoid arthritis

[sebbacon]

Population 2 of research question 2:

All patients in NHS England primary care who have received a previous diagnosis of RA or OA

[CarolineMorton]

RA code lists from @hmcd:

Rheumatoid_arthritis_Read_Pujades_Rodrigues_2016.xlsx From Paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4792375/#pone.0151245.s001

In their primary analysis they included only patients with supporting information for diagnosis (at least one diagnosis code for RA in both primary [CPRD] and secondary [HES] care, a DMARD or biologic therapy prescription, or a visit to rheumatology clinic within 6 months of the diagnosis code; or a positive rheumatoid factor test or anti-cyclic citrullinated peptide antibody test after diagnosis). But the diagnosis code list has been previously validated within CPRD and analogous medical databases with positive predictive values between 74% and 81%, so it would seem reasonable to me to use the diagnostic Read code list.

[CarolineMorton]

For Research Question 1, we need RA patients to be added to autoimmune conditions.

For Research Question 2, we need RA and OA together so we can look at NSAID use.

[CarolineMorton]

Rheumatoid Arthritis Sign Off

DEFINITION: Patients who have rheumatoid arthritis code ever on their record

Example output:	patient_id	ra_bin	condition	date
123	1	Rheumatoid arthritis	1/2/2009
332	1	Rheumatoid nodule	2/4/2016

CODE LISTS: Read 3 code list (when available). Created using this method by TPP:

1. Read 2 LSHTM code list Rheumatoid_arthritis.xlsx

OA code list from Swain et al 2020 codelist_OA.xlsx

2. Adding in key clusters from QOF and mapping to CTV3 (read code 3) - added by Caroline Morton (@CarolineMorton) qof-ra.xlsx

   • searched for rheumatoid arthritis and RA
   • manually checked for extra codes

3. Adding in high level snowmed codes and mapping to CTV3. Key Terms searched for in CT SNOWMED BROWSER: snowmed-ra.xlsx

   • rheumatoid arthritis
   • caplan's syndrome
   • Juvenile seropositive polyarthritis
   • NOTE: I have included childhood RA which was not originally included in the LSHTM list

4. Final list sense checked by clinician

• FLEXIBILITY NEEDED BETWEEN STUDIES: We may want to restrict to patients with certain medications, on treatment currently.
• EFFECTS ON COHORT SELECTION: Patients must have been registered for at least 12 months before the index date, in order to allow for recording of disease codes.

POTENTIAL BIASES: We may not have good data on patients who are managed by secondary care; although I hope we still have a diagnosis code. This may be more relevant for those studies where we are combining with active treatment, prescribed by secondary care physicians

.

CLINICAL SIGN OFF & DATE:

EPIDEMIOLOGY SIGN OFF & DATE:

SHARED WITH WIDER TEAM: Yes/No

FINAL SIGN OFF DATE (and apply label)

[hjforbes]

Reviewed RA TPP list here: I've added a variable, "Include", when 1=Yes and 0=No. RA_CTV3_Raw_HF_checked.xlsx

Caroline and I discussed over slack we should include RA monitoring invite codes, as patients must have RA to be on a monitoring list, and may help capture the less compliant of the RA patients.

@CarolineMorton do you want to scan and check, then I'll upload the final file with just the codes to be included?

[CarolineMorton]

Hi Harriet, I have combined RA into RA / SLE / Psoriasis. See here: https://github.com/ebmdatalab/tpp-sql-notebook/issues/49#issuecomment-611081601

Would you be able to check it and sign off if you are happy?

thank you

[hjforbes]

OA

DEFINITION: Current, or a history of, OA (generalised or localised).

STEP 1: READ list.
I used Swain's list posted above, but excluded the code "Arthritis" as it is too non-specific.

I have found an additional 62 OA codes (hf_include==1 and swain==0)

@CarolineMorton , can you review the additional codes? OA_Read_v2_HF.xlsx

Thanks!

[HenryDrysdale]

Hi all,

I'm helping out by doing some list reviews from a clinical perspective. Here's the OA lists:

1. I've reviewed the above OA read 2 list, added a column for include or not (1 or 0), and a column giving a reason. See below. For OA, the reason for excluding codes is mainly that they are too vague and could be another kind of arthritis. OA_Read_v2_HF_HD.xlsx

2. QOF - can't find any relevant clusters for OA!

3. SNOMED - there's a parent code for "Osteoarthritis (disorder)", which seems to include all relevant codes. SNOMED_OA.xlsx

Let me know if any questions / issues :)

[CarolineMorton]

thank you @HenryDrysdale

I am just removing the excluded Read 2 codes you mention so I can pass it onto TPP.

Step 1: tidy-read-2 code.xlsx Step 2: None Step 3: SNOMED_OA.xlsx

[CarolineMorton]

OA TPP CTV3 codes to be reviewed: OA_CTV3_Raw.xlsx

[hjforbes]

Reviewed here, with flag "include" (1=include, 0=exclude) OA_CTV3_Raw_TPP_HF.xlsx

@henrydrysdale and @carolinemorton, can you check?

[HenryDrysdale]

@hjforbes @CarolineMorton

I've checked the OA list and added an "include" list, with reasons.

• I think spondylosis is basically vertebral osteoarthritis, so I've included those. Certainly the treatment is very similar (NSAIDs, physio). This is quite a few codes - @CarolineMorton would you agree?

• (I've excluded a couple of codes for Charcot, which is a neuropathic arthropathy. I think these may have been included by accident - Charcot is mostly already excluded in the list)

OA_CTV3_Raw_TPP_HF_HD.xlsx

[CarolineMorton]

Hi, Agree with all @HenryDrysdale comments and have not changed anything so not added further file (one above can be used). I am happy to sign off.

[brianmackenna]

Please can @CarolineMorton @HenryDrysdale or @hjforbes have a final check of product on OpenCodelists

https://codelists.opensafely.org/codelist/opensafely/osteoarthritis/

[HenryDrysdale]

Here is the OA list above, but with excluded codes deleted, rather than just having excluded codes with a "0" next to them.

OA_CTV3_Raw_TPP_HF_HD_includedcodesonly.xlsx

[brianmackenna]

Confimed with @HenryDrysdale (on slack) all good to sign off

https://codelists.opensafely.org/codelist/opensafely/osteoarthritis/

[alschaffer]

I am in the process of updating this codelist. I note that the codelist linked to the Pujades-Rodriguez (2016) paper (which is about RA), but it should have been the Swain et al (2020) paper (OA) so this has been fixed.

[speed-vm]

Link to new issue for osteoarthritis here