Non-alcoholic fatty liver disease (NAFLD)

Nonalcoholic fatty liver disease (NAFLD) is becoming the most common cause of chronic liver disease in the developing world, found in 17-30% of the population in Western countries and 2-4% worldwide. NAFLD is diagnosed predominantly in the fourth through sixth decades of life, although the childhood obesity epidemic has caused an increase in the rate of pediatric NAFLD. The prevalence of NAFLD varies by ethnicity, affecting ~45% of Hispanics, 33% of Whites, and 24% of Blacks. Among Whites, NAFLD is more common in men than in women but not in other racial background (Angulo, P. (2007)).

Alcoholic Fatty Liver Disease (ALD):

In order to compare the genetic or clinical results between non-alcoholic and alcoholic fatty liver disease, we will also collect individuals with alcoholic fatty liver condition using available ICD-9 code:

Information
Phenotype ID: 
588
Date Created: 
Monday, August 29, 2016
Status: 
Final
Do Not List on the Collaboration Phenotypes List
Contact information
Institution: 
CCHMC
View Phenotyping Groups: 
Denny’s Group at Vandy
eMERGE CCHMC Group
eMERGE CCHMC/BCH Group
eMERGE CHOP Group
eMERGE Columbia Group
eMERGE Geisinger Group
eMERGE GHC/University of Washington Group
eMERGE Harvard Group
eMERGE Marshfield Group
eMERGE Mayo Group
eMERGE Mount Sinai Group
eMERGE Northwestern Group
eMERGE Phenotype WG
eMERGE Vanderbilt Group
PCORI CDRN/PPRN
Owner Phenotyping Groups: 
eMERGE CCHMC Group
Unstructured Data: 
NLP
Structured data: 
CPT
ICD10
ICD9
Data Source/clinical domain: 
observations
medications
Demographics
Age: 
Pediatric
Adult
Files: 
Image icon Pathology Scoring Sheet
File Component Scoring of NAS Histology
File NAFLD-inclusion-exclusion_final_06182018.docx

Suggested Citation

CCHMC. Non-alcoholic fatty liver disease (NALFD) & Alcoholic Fatty Liver Disease (ALD). PheKB; 2016 Available from: https://www.phenotyping.org/phenotype/588

Todd Lingren

Mon, 2018-01-29 10:30

Some labs report SGOT, some report it as AST

Please leave blank the one that your lab does not use.

Xinnan Niu

Mon, 2018-01-29 10:57

Questions.

1. If all units in the Labs v4 DD shoudl be  u/dl, what Bilirubin should be?

2.  For bilirubin, We need to collect?  total bilirubin? conjuated? or unconjuated? 

Thanks !

Xinnan Niu

Mon, 2018-01-29 11:00

1). If all units in the Labs v4 DD shoudl be  u/dl, how about bilirubin?

2). For bilirubin, do we need to collect total bilirubin ? conjuated ? unconjuated ?, or all of the 3?

thanks !

Xinnan Niu

Mon, 2018-01-29 15:04

Question: data defined in this sheet only has the following cases (see below) but does not include case #4 which is Alcohol dependence without any liver disorder and Control.

1=Case1_NAFLD;2=Case2_NAFLD;3=Case3_ALD

Question, If we provide data, can we denote "Alcohol dependence without any liver disorder" as "4" and Ctl as "0"?

Thanks !

 

 

Todd Lingren

Tue, 2018-02-06 12:45

Other item is correct

Alcohol dependence without any liver disorder" as "4"

Xinnan Niu

Mon, 2018-01-29 15:14

Question, there are very few patients who missing data for heights or weights, what we should put over the data sheet, using '.' or '..' because the DD does not specify this situation.

Ken Borthwick

Thu, 2018-02-01 10:16

I found a few LOINC mapping differences at our site on which I wanted to get your approval before I included the values. They have the same units (U/L) and are also from serum.

For AST, we use 30239-8   Aspartate aminotransferase [Enzymatic activity/volume] in Serum or Plasma by With P-5'-P

For ALT, we use 1743-4 Alanine aminotransferase [Enzymatic activity/volume] in Serum or Plasma by With P-5'-P

For LDH, we use 14804-9 Lactate dehydrogenase [Enzymatic activity/volume] in Serum or Plasma by Lactate to pyruvate reaction

If I use only the LOINC values that you provided I only find a very small amount of labs for these types. Is this ok?

Todd Lingren

Thu, 2018-02-01 13:50

If your LOINC codes don't match, its ok. Those are just a guideline. If you have alternate mappings for AST labs, for example, please use what you have. 

 

Zi Ye

Tue, 2018-02-06 12:16

Hello,

We are unable to find any NAFLD activity score in any cases, either in the clinical notes or in the pathology reports, using terms provided in the DD. Do you have any suggestion or alternative terms that we shall use for the algorithm?

Thanks,

Zi (Carol)

Todd Lingren

Tue, 2018-02-06 12:47

So you looked at the terms in the algorithm? Table three can help find the score (which is NASH or NAFLD Activity Score).

Does that help?

Are you looking in all notes, or just pathology notes?

Zi Ye

Wed, 2018-02-07 09:45

Hello,

When we intended to calculate LOS, we found 196 types of encounters in our EMR. Do you have a general guidance that which type of encounter we shall exclude?

Thanks,

Zi (Carol)

Todd Lingren

Wed, 2018-02-07 09:49

LOS is intended for inpatient encounters. Does that help? I doubt all 196 types will be present in the patient data for this phenotype.

Ning Sunny Shang

Wed, 2018-04-04 10:41

In our data, the score is recorded as (for example): The NAFLD activity score (NAS) is 3 (2+0+1). 

Key words we used for finding relevant NAS score  are  "NAFLD" OR  "NAS" OR "activity score” OR "NAFL" OR "NASH" OR "fatty liver" OR "activity score".

Fibrosis stage is recorded as (for example): stage 2 fibrosis.

David Carrell

Tue, 2018-04-24 12:59

There has been a lot of good discussion here.  Are the pseudo code document and data dictionary currently posted on PheKB up to date as of 4/24/2018?  In other words, if we follow those documents can we ignore the discussions posted above?

Thanks!

David

Todd Lingren

Tue, 2018-04-24 13:08

Hi David,

Yes all of the documents are current. Both the nas-histology-modified.xlsx and nafld_pathology_NAS_score.png can be used to manually grade the pathology reports and fill in the NAS scores. 

 

David Carrell

Tue, 2018-04-24 14:50

Wonderful!

Jen Pacheco

Thu, 2018-06-07 18:33

as of today, the algorithm doc does not contain all the necessary updates based on all the comments I've read on PheKB:  

i.e.,

Case 1 & 2 criteria are "AND" criteria, i.e., all 3 criteria must be met for ea., i.e. NLP is required to be case 1 or 2, correct? 

Table 1:  TPN Dx are only excluded if the occur in the 365 days before first NAFLD Diagnosis code?

Table 4:  does NOT contain the same diagnoses and procedures as the corresponding data dictionary NAFLD_DataDictionary_CovDx_v2.csv, please clarify in both the document and the data dictionary which codes to use, and which are ICD-9 & 10 diagnoses, vs. ICD-9 & 10 procedures, vs. CPT codes.   

Table 5: should not have ALP & GGT listed, & SGOT should be listed, & SGOT is the same as AST?

Please update the document and the NAFLD_DataDictionary_CovDx_v2.csv Data Dictionary so that it is clear, I need to be sure I'm doing the right things.

thank you!

Todd Lingren

Mon, 2018-06-18 12:51

Sorry for the confusion algo document is updated.

yes, AST/SGOT are the same, but there are different entries in the caDSR. Please use whatever you your system records for each lab result (either AST or SGOT) for the Covariate Lab DD. (updated to v5)

TPN within a year of NAFLD diagnosis could indicate some other nutritional deficit. So, correct, only exclude if the TPN happens in the year prior to the NAFLD diagnosis.

Current Cov_DX DD is updated to v3.

 

Jen Pacheco

Wed, 2018-06-20 18:47

2 questions: 

1) for T2DM, ICD-10 E11.* is listed which encompasses all codes for T2DM, but only 1 corresponding ICD-9 code is listed (250.00), did you intend to instead include 250.x0 & 250.x2, which would be all the ICD-9 codes for T2DM?  

2) Some of the diagnoses in Table 4 still aren't listed in the DD ...CovDx_v3.csv, should I just refer to Table 4 for that DD?  

Thanks, 

Jen

Jen Pacheco

Mon, 2018-06-25 10:45

and Q2 was do I use codes listed in table 4 or in covar DD, for that DD (table 4 has more listed)?

 

David Carrell

Tue, 2018-04-24 14:52

Hi,

Can you please describe what the term "activity score" is intended to represent (what is its meaning) in the NLP task?  Some of our reports discuss "activity" in the liver, but the term "activity score" does not appear.

Thanks,

David

Todd Lingren

Thu, 2018-04-26 07:26

Activity score is the same as a NAS score. See the nafld_pathology_NAS_score.png grading document for a full explanation.

Todd Lingren

Thu, 2018-04-26 07:23

NAS stands for non-alcoholic steatosis (NASH is non-alcoholic steatohepatitis)

Jen Pacheco

Thu, 2018-06-07 20:08

ICD-9: 

571.5, 571.8,  571.9

ICD-10:

K75.81, K76.0, K76.9

I ask b/c the document says "*571 is different from 571.0," so I'm not sure if I'm supposed to inc. 571.0 & 571 

and it says "non-specific" for K76.0, K76.9

Jen Pacheco

Mon, 2018-06-18 20:32

never mind to my last question RE: case diagnoses, I answered myself, I see now that ALD (case type 3) is diagnosis 571.0, not NAFLD case types 1 & 2

Jen Pacheco

Fri, 2018-06-08 13:14

for case type 1, it says to look for evidence "by imaging or by hiostology";  

What type of imaging should we use : abdominal ultrasound, CT scan, &/or MRI reports?

 

Jen Pacheco

Mon, 2018-06-18 11:07

2  more questions:  

1) by "clinical notes" do you mean just progress notes from in person clinical encounters, either outpatient office visits or inpatient hospital stays, or do you mean all notes in the EHR, or something else?  

  

2) We are assuming the only subjects' records that are to be in the score data dictionary data file are those dates with positive mentions of any of the phrases in Table 3, even if we can't find any scores or parts of scores (in which case we leave teh score fields as missing '.', but date & subject ID will be there so you know every time a subject had a mention of NAFLD or related phrases).  Is this correct?  

 

thanks!  Jen

Todd Lingren

Mon, 2018-06-18 12:55

1) Primarily pathology reports. But some sites have indicated other notes can be relevant. Please start with pathology report notes.

2) correct, thank you.

Jen Pacheco

Mon, 2018-06-18 17:29

I asked about "clinical notes" b/c that appears to be what differentiates between case type 1 & 2, is that correct, i.e. I see the only difference between case types 1 & 2 being the NLP in step 3 (i.e., steps 1 & 2 are the same for those 2 case types), where step 3 for both case types is to find "Evidence of hepatic steatosis" but to use NLP of different text :    

for case type 1, step 3 says NLP "...by imaging or by histology in biopsy report"  

vs. for case type 2, step 3 says NLP "... in clinical notes"  

Can you please clarify if this is the difference between case types 1 & 2, & if so, then what is meant by "clinical notes" (assuming biopsy rpt. & pathology rpt. are the same thing)? 

  OR if we are to just use pathology reports and not worry about case type 2?

Thanks!  Jen

Todd Lingren

Tue, 2018-06-19 09:08

Yes the clinical note vs biopsy report/pathology report is the critical distinction. 

Patients could overlap, of course. But Primary is Case 1. So if a patient meets part 1 and part 2 of the case definition, but doesn't have any pathology reports, then procede to case 2, part 3 (clinical notes), if those are available. 

Pages