*DISEASE*: Chronic Liver Disease

[sebbacon]

Code for identifying binary variable for "presence of chronic liver disease"

this may be on codes or lab result

[CarolineMorton]

We have code lists from LSHTM:

chronic_liver_Jul19.xlsx

These are all on read codes. If we feel confident about just using these, we could avoid using lab results; which is preferable as some patients will have one abnormal LFTs but not necessarily have chronic liver disease.

[CarolineMorton]

DRAFT

DEFINITION: Patients who have a chronic liver disease code ever on their records

• CODE LISTS: Chronic liver disease codes provided in Read Code v2 by LSHTM (need mapping into Read V3).
  1. Read 2 LSHTM validated code list chronic_liver_Jul19.xlsx
  2. QoF cluster codes QoFClusteres_CTV3Codes-liver.xlsx - added by Alex Walker (@alexwalkercebm)

3. Snomed codes: snowmed-liver.xlsx - added by Alex Walker (@alexwalkercebm). Searched for:

   • liver disease
   • viral hepatitis
   • liver transplant
   • oesophageal varices
   • portal hypertension
   • liver cirrhosis

• 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: We may miss patients who have chronic liver disease and are under investigation but do not have a diagnosis. This is likely to be small in numbers. We are not using lab results (specifically LFTs) so may miss undiagnosed, mild disease

CLINICAL SIGN OFF & DATE:

EPIDEMIOLOGY SIGN OFF & DATE:

SHARED WITH WIDER TEAM: Yes/No

FINAL SIGN OFF DATE (and apply label)

[CarolineMorton]

FINAL SIGN OFF

DEFINITION: Patients who have a chronic liver disease code ever on their records

CODE LISTS: Chronic liver disease codes in Read 3: ChronicLiver_final.xlsx

. Created by the following process.

1. Read 2 LSHTM code list chronic_liver_Jul19.xlsx

2. Addition of relevant QoF cluster codes QoFClusteres_CTV3Codes-liver.xlsx - added by Alex Walker (@alexwalkercebm)

3. Addition of relevant high level Snomed codes: snowmed-liver.xlsx - added by Alex Walker (@alexwalkercebm). Searched for:

   • liver disease
   • viral hepatitis
   • liver transplant
   • oesophageal varices
   • portal hypertension
   • liver cirrhosis

4. Final code list reviewed by @CarolineMorton. See document here: ChronicLiver_CTV3_REVIEWED.xlsx

   Reasons for inclusion or exclusion explained in Column D, with a '1' in 'non-include' column if not included.

   • Excluded obvious non liver disease such as varicose veins
   • Excluded genetic diseases such as glycogen storage diseases where there can be liver disease associated if untreated but there is not always.
   • Included primary genetic liver disease such as primary bilary cirrhosis
   • Included all viral hepatitis which were chronic in nature
   • Included all signs of cirrhosis such as oesophageal varices

• FLEXIBILITY NEEDED BETWEEN STUDIES: None
• 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 miss patients who have chronic liver disease and are under investigation but do not have a diagnosis. This is likely to be small in numbers. We are not using lab results (specifically LFTs) so may miss undiagnosed, mild disease

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

EPIDEMIOLOGY SIGN OFF & DATE: Alex Walker (@alexwalkercebm) 7/4/2020 16:20

SHARED WITH WIDER TEAM: Yes

FINAL SIGN OFF DATE (and apply label) 7-4-2020 17:39

[alexwalkerepi]

The current codelist excludes liver transplants (the same is true for the cardiac and lung transplants). We were concerned about double counting, as there is also a solid organ transplant disease definition. We perhaps need to think a little more about the implications of this.

from Slack:

Though most would also have another liver disease codelist, a few might not. The PPV for liver disease in people who've got a liver transplant code should be pretty high.

[laurietomlinson]

Hi I had a look through in view of the slightly strange results According to the list above it looks like 17 of the codes relate to liver transplant and it does seem to me that they should be listed under organ transplants. It may be that they were removed as per the discussion above - I'm not clear. We know kidney transplant is well coded at least. This however wouldn't explain why the HR is low for ITU admission. There are a few where you could argue that liver disease may not be present or was reversible, or should be attributed elsewhere, but I doubt this is very important in explaining the result. They main thing seems to be the number of codes related to having very severe cirrhosis eg with oesophageal varices where you would anticipate poor survival regardless of COVID. It would be interesting to see the number of people in code groups. I guess the key question is - would we expect these people to be listed under 'liver disease' and we probably do, but it would be good to indicate somehow that this is a very wide range of severity of disease. My thoughts here:

ChronicLiver_CTV3_REVIEWED_LT.xlsx

[CarolineMorton]

Hi Laurie,

Thanks for that. We decided to keep transplant codes in for all of the major groups (respiratory, liver, heart) to keep it consistent. There is also a separate solid organ transplant list , and patients should be flagging in both categories if they have had a transplant. There was much debate about double counting but ultimately that was the decision we came to.

Agree that some of the liver codes may indicate mild disease but these were kept in because it was not clear.

@krishnanbhaskaran it would be good to get a crosstab out for liver disease, i think?

[krishnanbhaskaran]

Added crosstab of liver disease and organ transplant in an_checks for next run