National Health and Nutrition Examination Survey

Between 2009 - 2010, the hepatic steatosis (fatty liver) was assessed by re-reviewing the archived gall bladder ultrasound video images originally obtained in NHANES III between 1988 and 1994. The examination made it possible to grade the presence of fat within the hepatic parenchyma.

The ultrasound examinations were performed on adults aged 20 to 74 years who were examined in NHANES III. Fasting was preferred but not required.

Ultrasound examinations were originally recorded using a Toshiba Sonolayer SSA-90A and Toshiba video recorder. The examinations were performed using standardized procedures (Gallbladder Ultrasonography Procedure Manual, Third National Health and Examination Survey, https://wwwn.cdc.gov/nchs/data/nhanes3/manuals/gallblad.pdf), which had been developed to ensure that each examination was performed in a consistent manner and that the results of each examination were accurate and reliable.

For the hepatic steatosis assessments, the video tapes of the NHANES III ultrasound examinations were digitized using a DVD-VHS Video cassette Recorder (SONY RDR-VX560) onto a Recordable DVD (Memorex, +R or +RW). DVD video-images were reviewed using a Dell Flat Panel Monitor (2408WFP, active matrix, thin-film transistor, liquid crystal display, 24-inch (609.6 mm) viewable area display, 1920 x 1200 resolution).

Three ultrasound readers were trained by a board certified radiologist who is specialized in hepatic imaging. Quality control and quality assurance procedures were used to standardize the reading form all readers (see below and Procedure Manual, Hepatic Steatosis). The following information was recorded on a standard paper collection form: 1) the presence of liver-to-kidney contrast 2) the degree of the brightness of the liver parenchyma, 3) the presence of deep beam attenuation, 4) the presence of echogenic walls in the small intrahepatic vessels, and 5) the definition of the gallbladder walls. Finally, an overall primary finding was given based on the presence or absence of each of the five parameters. The liver was graded as normal, mild, moderate, or severe hepatic steatosis.

For practical reasons, NHANES III Phase 2 ultrasounds were reviewed before Phase 1.

A more detailed description of the protocol is included in the Procedure Manual, Hepatic Steatosis, NCHS 2010. https://wwwn.cdc.gov/nchs/data/nhanes3/manuals/hepatic_steatosis_ultrasound_procedures_manual.pdf

A radiologist, with 21 years of experience in the interpretation of ultrasound images, trained, observed, and validated the readers. Routine monitoring of the reliability (both inter- and intra-reader) was conducted every three months, throughout the study to detect if there was any need for re-training. For the training, the following steps were taken: 

1) All readers were required to read the relevant sections on hepatic ultrasound from "The Requisites Ultrasound" (Middleton et. al., 2004.); 

2) The readers met with the expert radiologist and reviewed a minimum of 100 ultrasound exams, from randomly selected NHANES III ultrasound video tapes. The key concepts were explained and demonstrated in detail using several exams;

3) The readers then re-reviewed the sample images from a library of more than 100 NHANES III examinations and external references images which were chosen as characteristic of a normal liver, a liver with mild-to-moderate hepatic steatosis, and a liver with severe hepatic steatosis;

4) The data collection form and procedures were also reviewed;

5) Additional exams (a minimum of 10 additional exams) were reviewed with the readers, until they demonstrated a good understanding of the concepts and procedures for reading the ultrasounds, and familiarity in completing the data form. The radiologist observed the readers to identify any reader difficulties and to ensure that findings are properly identified and documented;

6) A random sample of 100 NHANES III ultrasound exams were read by each reader separately and in random order and re-read at least one week apart. The kappa coefficient for inter- and intra-rater reliability was determined. Each reader was approved if their intra- and inter-rater kappa coefficients were ≥0.6. Should a reader fail to be approved, the training was repeated.

Throughout the study, each reader re-read a 5% randomly selected sample of the tapes s/he read previously and another 5% were re-read by another reviewer. Monitoring of the reliability results (both intra-rater and inter-rater) was conducted every three months. All ultrasound assessments for hepatic steatosis were conducted by readers who had no access to any other participant data.

A more detailed description of the Training, Quality Assurance and Quality Control is included in the Procedure Manual, Hepatic Steatosis, NCHS, 2010. https://wwwn.cdc.gov/nchs/data/nhanes3/manuals/hepatic_steatosis_ultrasound_procedures_manual.pdf

The primary finding variable (GUPHSPF), is a 4 level variable. A primary finding recode (GUPHSPFR) variable was created from this variable. This is a two category variable: hepatic steatosis yes or no. The recode value of "Yes" indicates that moderate or severe hepatic steatosis was present. A recode value of "No" indicates that the liver was normal or had mild hepatic steatosis.

When analyzing these data, the MEC examination sample weight should be used (WTPFEX6). Users are encouraged to note carefully the "Instructions" describing, in brief, the various variables and suggested analysis in the codebook documentation. For more detailed information, readers should review the Procedure Manual (Hepatic Steatosis, NCHS 2010).Please refer to the NHANES Analytic Guidelines and the on-line NHANES Tutorial for further details on the use of sample weights and other analytic issues. Both of these are available on the NHANES website.

Users are encouraged to use the primary finding recode (GUPHSPFR) variable for all the analyses. This recode will help to avoid potential overlap between mild and moderate hepatic steatosis. The data file provides all the different parameters that were obtained for the assessment of hepatic steatosis: liver to kidney contrast, parenchymal brightness, deep beam attenuation, vessel walls definition, and gallbladder wall definition and the overall confidence in the assessment. These parameters should allow more experienced users to explore alternative criteria and to conduct sensitivity analyses.

GUPHSQC (Hepatic Steatosis assessment status comment) is a quality control variable that allowed NCHS to monitor the status of the assessment and the reasons readers provided why an exam was not analyzed or missing. The variable is provided in the data release file to allow analysts to have some information on missing data. Image is present, but ungradable may indicate any of the following: VHS tape was damaged, the ultrasound has a very poor quality that impairs the assessment of hepatic steatosis (e.g. presence of gas, patient examined in a wheelchair). No image reflects the fact that there was no ultrasound available for the participant.

In addition, for each of the above mentioned parameters (liver to kidney contrast, parenchymal brightness, deep beam attenuation, vessel walls definition, and gallbladder wall definition), in cases where the ultrasound technique and coverage did not include the parameter, then we marked "cannot assess" (coded as 9).