*DISEASE* Chronic Respiratory Disease (excluding asthma)

[sebbacon]

One ticket as I understand the approach will be very similar

• Using QoF cluster codes (which ones? There's more than one)
• and medicines e.g. inhalers where at least rx in 3 consecutive quarters (this was @brianmackenna's suggestion)

If the logic is exactly as above, this has been implemented here - but we still need to define the code lists for disease and medicines.

[CarolineMorton]

Code lists coming from Helen MacDonald for COPD and other chronic lung disease, excluding asthma. This may be better than cluster codes they think.

[CarolineMorton]

chronic_resp_no_asthma_Jul19.xlsx
codes lists with no asthma: https://datacompass.lshtm.ac.uk/1663/

issues with asthma. From Helen M:

Validated 'specific' asthma codes published by Niseen et al. However this code list alone probably ascertains asthma when used "ever

Would likely need to be combined with use of asthma medications to avoid capturing childhood asthma or wheeze when had infection in the past.

[CarolineMorton]

Cystic Fibrosis:
cystic_fibrosis_Jul18.xlsx

[brianmackenna]

Some cystic fibrosis medicines are infrequently supplied from primary care. We could use a these medicines to hopefully pick up further exclusions in practices where the READ codes are not recorded comprehensively. Let me know and I can generate a list.

[CarolineMorton]

Agreed in phone call today to use asthma medicines to identify patients with asthma. @brianmackenna @hmcd @amirmehrkar

[CarolineMorton]

@brianmackenna Shall we split Asthma into it's own issue as it has a different methodology?

[alexwalkerepi]

It's quite possible we'd want to measure it separately anyway right?

[CarolineMorton]

agreed

[alexwalkerepi]

Draft sign off (COPD only)

DEFINITION: Patients who have any COPD Read 3 code ever on their medical records held by TPP. Absence of a code on the record is taken as no presence of COPD.

CODE LISTS: Read 3 code list (when available) - mapped by TPP from combined Read 2 LSHTM validated code lists (https://github.com/ebmdatalab/tpp-sql-notebook/files/4409910/chronic_resp_no_asthma_Jul19.xlsx; https://github.com/ebmdatalab/tpp-sql-notebook/files/4409917/cystic_fibrosis_Jul18.xlsx); and sense checked
FLEXIBILITY NEEDED BETWEEN STUDIES:
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:

CLINICAL SIGN OFF & DATE:

EPIDEMIOLOGY SIGN OFF & DATE:

SHARED WITH WIDER TEAM: Yes/No

FINAL SIGN OFF DATE (and apply label)

[hmcd]

Draft

DEFINITION: Patients who have any chronic respiratory disease Read 3 code ever on their medical records held by TPP. This includes chronic obstructive pulmonary disease (COPD); and such conditions as bronchiectasis, cystic fibrosis, and interstitial lung fibrosis. Absence of a code on the record is taken as no presence of disease.

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

1. Combined Read 2 LSHTM/PRIMIS code lists

   • Chronic Respiratory Disease, no asthma included. https://github.com/ebmdatalab/tpp-sql-notebook/files/4409910/chronic_resp_no_asthma_Jul19.xlsx
   • Cystic Fibrosis: https://github.com/ebmdatalab/tpp-sql-notebook/files/4409917/cystic_fibrosis_Jul18.xlsx

2. Adding in key clusters from QOF and mapping to CTV3 (read code 3). qof-chronic-resp-disease.xlsx
   Inclusion and exclusion criteria.

   • Did not include codes on inhaler technique or spirometry as could be for asthma and does not always indicate diagnosis as inhaler technique may be measured during diagnostic spirometry.

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

   • chronic respiratory failure
   • lung transplant
   • bronchiectasis
   • Asbestosis
   • pulmonary alveolitis
   • pulmonary fibrosis
   • cystic fibrosis
   • Chronic obstructive lung disease

   NOTE: Some parents could not be used higher than the one in the document because starts to incorporate cancers, for example parent of bronchiectasis is disorder of bronchus which includes malignancy

4. Final list sense checked by clinician

• FLEXIBILITY NEEDED BETWEEN STUDIES: We may want to add in at a later date, patients who are in receipt of active treatment.

• 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: Excludes undiagnosed patients. Combines a range of different respiratory conditions, which may obscure associations of specific respiratory diseases.

CLINICAL SIGN OFF & DATE:

EPIDEMIOLOGY SIGN OFF & DATE:

SHARED WITH WIDER TEAM: Yes/No

FINAL SIGN OFF DATE (and apply label)

[CarolineMorton]

I have received the Read 3 code list from TPP (Thanks @chris-tpp). This is the code list with all the codes provided:
Chronic_Resp_No_Asthma_CTV3_Raw-REVIEWED.xlsx

I have reviewed all the codes and marked all the ones that I think we should not include by putting a '1' in the not include column and a reason. The main reasons are:

• Pneumonia due to infection (i.e. not a chronic lung infection) including TB
• Asthma related code
• Lung transplant related code which should be in the solid organ code list (*CONDITION* Fragility #31 )
• Spirometry code where there is no specific diagnosis
• Negative / borderline / test noted tests for cystic fibrosis
• Pulmonary oedema NOS (primarily another organ cause such as heart failure)

I have included long term complications of TB such as bronchiectasis. I have also included chronic pulmonary oedema.

FINAL CODE LIST (i.e. these codes removed): Chronic_no_asthma-final.xlsx

[CarolineMorton]

One issue identified by TPP:

There is no longer a COPDEXC cluster. These would have been the exclusion codes and we may have to add any old codes manually to this list.

I don't think this makes much difference to our final code list but happy to discuss

[chris-tpp]

We’ve got a solution to this remaining issue. If we remove the restriction on active QOF clusters then we pick up these codes and it doesn’t seem to expand the list too much for clinical review. All other lists have had - will have this applied; we’ll get a list of COPDEXC codes separately to you.

[CarolineMorton]

Ok great, and then would the idea be that I merge them into the file above.

@brianmackenna, to follow on from Slack chat. I am not sure that pulmonary TB should be included which is why I took it out as it is an infection and shouldn't be a chronic condition (although we know it can be if not treated). Agree with your point at chronic pulmonary heart disease - I can remove this.

[brianmackenna]

Agree with rationale for exclusion for pulmonary TB. Other codes I propose for possibly excluding are related to pulmonary heart disease which will affect low numbers (so lets not spend a lot of time thinking about)

• G41.. | Chronic pulmonary heart disease
• G41y. | Other chronic pulmonary heart disease
• G41yz | Other chronic pulmonary heart disease NOS
• XE0Un | Chronic pulmonary heart disease NOS

[CarolineMorton]

Discussed with @hmcd: agreed to remove chronic pulmonary heart disease and cor pulmonale for code list and make sure are in the chronic heart conditions list.

I will amend the definition and repaste here and then we can sign off

[CarolineMorton]

FINAL SIGN OFF

DEFINITION: Patients who have any chronic respiratory disease Read 3 code ever on their medical records held by TPP. This includes chronic obstructive pulmonary disease (COPD); and such conditions as bronchiectasis, cystic fibrosis, and interstitial lung fibrosis. Absence of a code on the record is taken as no presence of disease.

CODE LISTS: Final read 3 code list:
Chronic_no_asthma-final (v1.1).xlsx
)

Created using this method by TPP:

1. Combined Read 2 LSHTM/PRIMIS code lists

   • Chronic Respiratory Disease, no asthma included. https://github.com/ebmdatalab/tpp-sql-notebook/files/4409910/chronic_resp_no_asthma_Jul19.xlsx
   • Cystic Fibrosis: https://github.com/ebmdatalab/tpp-sql-notebook/files/4409917/cystic_fibrosis_Jul18.xlsx

2. Adding in key clusters from QOF and mapping to CTV3 (read code 3). qof-chronic-resp-disease.xlsx
   Inclusion and exclusion criteria.

   • Did not include codes on inhaler technique or spirometry as could be for asthma and does not always indicate diagnosis as inhaler technique may be measured during diagnostic spirometry.

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

   • chronic respiratory failure
   • lung transplant
   • bronchiectasis
   • Asbestosis
   • pulmonary alveolitis
   • pulmonary fibrosis
   • cystic fibrosis
   • Chronic obstructive lung disease

   NOTE: Some parents could not be used higher than the one in the document because starts to incorporate cancers, for example parent of bronchiectasis is disorder of bronchus which includes malignancy

4. Final list sense checked by clinician. I have received the Read 3 code list from TPP. This is the code list with all the codes provided.
   Chronic_Resp_No_Asthma_CTV3_Raw-REVIEWED.xlsx

   I have reviewed all the codes and marked all the ones that I think we should not include by putting a '1' in the not include column and a reason. The main reasons are:

   • Pneumonia due to infection (i.e. not a chronic lung infection) including TB
   • Asthma related code
   • Spirometry code where there is no specific diagnosis
   • Negative / borderline / test noted tests for cystic fibrosis
   • Pulmonary oedema NOS (primarily another organ cause such as heart failure)
   • TB has been excluded as infective, and chronic nature unknown (see discussion https://github.com/ebmdatalab/tpp-sql-notebook/issues/21#issuecomment-609792488). I have included long term complications of TB such as bronchiectasis as this represents chronic disease process.
   • Chronic pulmonary heart disease / Cor pulmonale have been removed after discussion with @hmcd and @brianmackenna and will be included within the CVD (*PATIENT*: ethnicity #7) list.
   • After discussion at meeting (8-4-2020), we decided to include transplants in this list as well as the transplant list

• FLEXIBILITY NEEDED BETWEEN STUDIES: We may want to add in at a later date, patients who are in receipt of active treatment.

• 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: Excludes undiagnosed patients. Combines a range of different respiratory conditions, which may obscure associations of specific respiratory diseases.

CLINICAL SIGN OFF & DATE: Caroline Morton (@CarolineMorton) 8.4.2020 20:12

EPIDEMIOLOGY SIGN OFF & DATE: Brian MacKenna (@brianmackenna ) 10.4.20 15:23
(note we may wish to rename the checks)

SHARED WITH WIDER TEAM: Yes

FINAL SIGN OFF DATE (and apply label)

[brianmackenna]

@annaschultze has had a look at the code while reviewing ebmdatalab/tpp-sql-notebook#81 and spotted occupational does. You can view the code list on OpenCodelist @CarolineMorton I remember we discussed occupational codes one day but don't think it was related to this issue. Should we remove?

[hmcd]

Think we agreed we should include any chronic nonreversible lung disease so it would include occupational pneumoconiosis etc. (Not asthma triggered by dust though). Is there a reason we would want to exclude these?

…

On Sun, 26 Apr 2020 12:42 Brian, ***@***.***> wrote: @annaschultze <https://github.com/annaschultze> has had a look at the code and spotted occupational does. You can view the code list on OpenCodelist <http://smallweb1.ebmdatalab.net:8001/codelist/opensafely/chronic-respiratory-disease/> @CarolineMorton <https://github.com/CarolineMorton> I remember we discussed occupational codes one day but don't think it was related to this issue. Should we remove? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <https://github.com/ebmdatalab/tpp-sql-notebook/issues/21#issuecomment-619535833>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI5K5A4PZJEH6ERFCQLJX6DROQM23ANCNFSM4LXK42SA> .

[brianmackenna]

While reviewing ebmdatalab/tpp-sql-notebook#81 @annaschultze has highlighted the inclusion of the four codes below for a double check to see if they should be removed. (see here on OpenCodelists)

ID	term description	source
X1025	Occupational asthma	High_Level_SNOMED
X1026	Baker's asthma	High_Level_SNOMED
X1027	Colophony asthma	High_Level_SNOMED
X1028	Grain worker's asthma	High_Level_SNOMED

[brianmackenna]

I have updated the file and removed four occupational asthma codesaafter discussion with @CarolineMorton
Chronic_no_asthma-final.1.2.xlsx

To be clear for other studies asthma was removed and defined elsewhere as it was not included here in Immunisation risk group - chronic respiratory disease without asthma , which is discussed at started of issue.

[cminassian]

@hmcd @CarolineMorton I've removed the cystic fibrosis screening code from the chronic respiratory disease codelist and saved an amended copy here:
opensafely-chronic-respiratory-disease-version2.xlsx

The original version remains unchanged.

@inglesp suggested the amended codelist could be uploaded with a different name (e.g chronic respiratory disease version 2) as an interim measure until there's a process in place for making multiple versions of the same codelist. So I can do that if useful - though I assume will need signing off @CarolineMorton?