ARDI Announcements

ARDI Information

back to: Alcohol-Related Disease Impact (ARDI) Home

April 25, 2022

Updated data were released in ARDI on April 19, 2022; however, system glitches were identified and all glitches were resolved on April 25, 2022. System glitches included the following programming issues:

Inability to upload CSV data sets in the Custom User Portal.
Estimates in the state reports for deaths from motor vehicle traffic crashes were shown based on the national-level alcohol-attributable fractions instead of the state-specific alcohol-attributable fractions. This resulted in imprecise estimates of deaths and years of potential life lost from motor vehicle traffic crashes, and the total estimates, being posted in the state reports. The estimates in the national-level reports were not affected.

We apologize for this inconvenience for users who accessed the state reports in ARDI between April 19–25, 2022.

April 19, 2022

The ARDI application has been updated, including:

New estimates of alcohol-attributable deaths and years of potential life lost
- National and state estimates of average annual alcohol-related deaths and years of potential life lost are now based on the 2015–2019 mortality data from the National Vital Statistics System.
ARDI methodologic updates and programming changes
- The 2015–2019 Behavioral Risk Factor Surveillance System (BRFSS) data are used for estimating the prevalence of average daily alcohol consumption. Updated methods that adjust for the underreporting of self-reported alcohol use in the BRFSS using per capita alcohol sales data are now used in ARDI, which affects estimates that are calculated using indirect alcohol-attributable fractions, including the number of alcohol-attributable deaths and years of potential life lost for most of the chronic conditions, as well as the totals. This method for adjusting the prevalence of alcohol consumption is described in detail in Esser et al. (2022) and summarized on the ARDI Methods webpage. It involves conservatively accounting for a portion (73%) of per capita alcohol sales when calculating the prevalence of low, medium, and high average daily alcohol consumption. The prevalence estimates generated using the previous prevalence adjustment method, referred to as “indexing” — a survey-based adjustment to account for binge drinking — are available for 2015–2019 data upon request if ARDI users are interested in performing custom analyses using those estimates.
- ARDI now captures deaths from the International Classification of Disease (ICD)-10 code K70.0 for estimating deaths and years of potential life lost from alcoholic fatty liver, as part of the alcoholic liver disease condition. Due to a statistical programming error, deaths coded as K70.0 were not captured in earlier iterations of ARDI. The average annual number of deaths in the United States coded as K70.0 during 2011–2015 was 379.
- The national and state alcohol-attributable fractions for deaths from motor vehicle traffic crashes are now based on 2015–2019 data from the Fatality Analysis Reporting System. These updated alcohol-attributable fractions are available by selecting “alcohol-attributable fractions” from the drop-down menu of report options in ARDI. In addition, the estimated average annual number of alcohol-attributable deaths from motor vehicle traffic crashes now includes people of all ages. The 2011–2015 estimates of average annual alcohol-attributable deaths from motor vehicle traffic crashes included people in the following age groups: <1, 15–29, 35–39, 45–49, 55–59, and 65–69 years.
- The alcohol-attributable fraction for “pancreatitis, acute” is calculated using a single relative risk estimate for all levels of average daily alcohol consumption. Therefore, when calculating deaths from acute pancreatitis from excessive alcohol use (medium and high average daily alcohol consumption levels), the reference group for the relative risks is non-drinking rather than low average daily alcohol consumption.
- The alcohol-attributable fractions for esophageal varices and portal hypertension were updated based on the relevant 2015–2019 death data, as explained in the alcohol-attributable fractions report.

With these updates to ARDI, estimates of alcohol-attributable deaths or years of potential life lost generated in the current version of ARDI should not be compared with estimates that were generated using the default reports or analyses in the Custom Data Portal prior to April 19, 2022.

New features
- The updated ARDI Custom Data User Portal is now live, along with the accompanying Custom Data User Manual.
- The 2015–2019 ARDI data are available in the CDC’s National Center for Chronic Disease Prevention and Health Promotion’s Open Data Portal.

September 3, 2020

On September 3, 2020, corrections were made to the alcohol-attributable fractions for five acute causes of death: drownings, fall injuries, fire injuries, firearm injuries, and homicide. Overall, these corrections resulted in increases to the estimated average annual number of alcohol-attributable deaths and years of potential life lost for 2011–2015. The nationwide average annual number of deaths increased by 1,862, from 93,296 to 95,158, and the number at the state-level increased by 37, on average (range: 0 to 215 deaths). The nationwide average annual number of years of potential life lost increased by 79,844, from 2,683,211 to 2,763,055, and the number at the state-level increased by 1,574 years, on average (range: 11 to 9,495 years).

July 30, 2020

All of the scientific information in the Alcohol-Related Disease Impact (ARDI) application has been reviewed. The following updates were made to ARDI:

Average annual estimates of alcohol-attributable deaths and years of potential life lost were updated from 2006–2010 to 2011–2015 for the United States, states, and DC. Data updated include total deaths, the prevalence of alcohol consumption by level of consumption, and average life expectancies.
The list of 58 acute and chronic causes of death includes new conditions that were added (e.g., cancers of the stomach and pancreas) and some name modifications (e.g., “ischemic heart disease” is now labeled as “coronary heart disease”). Some conditions that were previously included in ARDI were removed based on updated scientific information (e.g., spontaneous abortion).
The ICD-10 codes for defining several causes of death (e.g., liver cirrhosis unspecified, atrial fibrillation, and poisonings) were revised.
The relative risks and alcohol-attributable fractions were updated to reflect more recent scientific literature from more than a dozen different sources, including peer-reviewed meta-analyses on the specific relationship between alcohol and particular causes of death (e.g., cancers, coronary heart disease, stroke) and multi-state surveillance systems (e.g., the National Violent Death Reporting System and the Fatality Analysis Reporting System). The source of the relative risk or alcohol-attributable fraction for each condition included in ARDI is described on the updated Methods page.

With these scientific updates to ARDI, estimates of alcohol-attributable deaths or years of potential life lost generated in the current version of ARDI should not be compared with estimates that were generated using the default reports or analyses in the Custom Data Portal prior to July 30, 2020.

March 18, 2019

From March 1, 2018 to March 18, 2019, incorrect estimates of alcohol-attributable deaths, years of potential life lost, and alcohol-attributable fractions for some chronic conditions (e.g., acute pancreatitis, breast cancer, ischemic heart disease) were posted on the ARDI website. This issue has now been resolved, and the incorrect data have been removed and replaced with the data that were available on the ARDI website before this systems error occurred. However, we recommend that any ARDI user who downloaded or otherwise used ARDI data during the timeframe noted confirm that that information matches the data currently available on the ARDI website, and if not, discard the old data or reports and replace them with those currently available on the ARDI website.

We apologize for this inconvenience. Please feel free to contact us with any questions, using the Contact Us About ARDI button at the top of the ARDI webpages.

June 30, 2017

ARDI data are now included in the CDC Chronic Disease and Health Promotion Open Data Site. This site allows you to create filtered datasets and customize visualizations, as well as quickly access a wide range of chronic disease data, risk factor indicators, and policy measures. Learn more at the Chronic Disease and Health Promotion Open Data site.

May 26, 2016:

One of the ICD-10 sub-codes for Alcoholic Liver Disease (K70.0) was incorrectly denoted in the SAS code that was used to analyze death data for the Alcohol-Related Disease Impact (ARDI) application. As a result, from 2006 to 2010, 331 average annual deaths due to Alcoholic Liver Disease with an underlying cause of death coded as K70.0 were inadvertently excluded from the calculation of total deaths due to this condition. This error has been corrected, resulting in a slight increase in the estimated average annual alcohol-attributable deaths (AAD) and years of potential life lost (YPLL) for these conditions in the U.S. and in states.

December 12, 2013:

The Alcohol-Related Disease Impact (ARDI) application has been updated to provide average annual estimates of alcohol-attributable deaths (AAD) and years of potential life lost (YPLL) for the United States and for states for 2006–2010. ARDI also includes new estimates of AAD and YPLL for all 54 conditions that are assessed in the application, including all chronic (e.g., alcoholic liver disease) and acute (e.g., alcohol-attributable motor-vehicle traffic crashes) conditions.

The indexing procedure, which is used in the ARDI application to assure that self-reported information on binge drinking episodes is included in the calculation of average daily alcohol consumption, has also been updated. Specifically, instead of using a national estimate of binge drinking intensity (i.e., the number of drinks consumed per binge drinking episode), the indexing procedure used in ARDI now uses state-specific estimates of binge drinking intensity based on the average largest number of drinks consumed reported by binge drinkers in state Behavioral Risk Factor Surveillance System (BRFSS) surveys. The use of these state-specific estimates of binge drinking intensity in the ARDI application will improve the accuracy of state estimates of average daily alcohol consumption and thus the accuracy of state-specific estimates of AAD and YPLL as well.

July 25, 2012:

2005 life expectancy data has been added to ARDI, resulting in new life expectancy averages for all ages by sex for 2001–2005. Previous life expectancy average included average for 2001-2004. This update resulted in a slight change to the estimated years of potential life lost (YPLL) overall and by state.

To calculate years of potential life lost (YPLL), the alcohol-attributable fractions (AAF) for a specific cause of death are multiplied by the number of deaths in a given category and then multiplied by the average number of years of life remaining based on life expectancy of the decedent. Data on life expectancy are obtained from the National Vital Statistics System managed by the National Center for Health Statistics. Life expectancy data are stratified by age and sex using standard 5-year age groupings.

October 13, 2011:

The average grams of alcohol per drink − which is used in ARDI to convert the average number of grams of alcohol consumed per day, as specified in some of the meta-analyses used to obtain risk estimates, into the average number of drinks per day – was changed from 13.7 grams per drink to 14.0 grams per drink, consistent with current recommendations from the National Institute on Alcohol Abuse and Alcoholism (NIAAA). The prevalence of drinking at the new average daily alcohol consumption levels (e.g., low, medium, and high average daily alcohol consumption, as defined by the meta-analyses used in ARDI) was then assessed using national and state-specific data from the Behavioral Risk Factor Surveillance System (BRFSS).

This change in the number of grams of alcohol per drink – and, in turn, in the prevalence of drinking at specified alcohol consumption levels − has slightly changed the estimated number of alcohol-attributable deaths (AAD) and years of potential life lost (YPLL) for the U.S. and for states that are reported in the ARDI application.

May 10, 2011:

ARDI gets a makeover: The redesigned Alcohol-Related Disease Impact (ARDI) application was launched today. The estimates in the application have not changed.

The Alcohol-Related Disease Impact (ARDI) online application allows users to access national and state estimates of alcohol-related health impacts, including deaths and years of potential life lost (YPLL), for 54 acute and chronic causes using alcohol-attributable fractions, by age and sex. ARDI was recently upgraded to a more user-friendly platform.