CSTE logo
This website uses cookies to store information on your computer. Some of these cookies are used for visitor analysis, others are essential to making our site function properly and improve the user experience. By using this site, you consent to the placement of these cookies. Click Accept to consent and dismiss this message or Deny to leave this website. Read our Privacy Statement for more.
CSTE Features
Blog Home All Blogs
Search all posts for:   


View all (176) posts »

Epi Tool to Analyze Overdose Death Data

Posted By James Davis (NM), Jennifer Sabel (WA), Dagan Wright (OR), and Svetla Slavova (KY), Friday, March 13, 2015
Updated: Friday, March 13, 2015

Did you know that drug overdoses are the leading cause of injury death in the nation? While we know that they are a leading cause of injury death, the ICD codes often only tell us the broad class of drugs involved and not the specific drugs. However, there is often more information about the specific drugs on the death certificate.

A new analytic tool developed by the CSTE Overdose Subcommittee treads new ground by directly analyzing the cause of death information on the death certificate, sometimes referred to as the literal text. Literal text entries from the death certificates are now either entered into electronic death certificates or transcribed paper certificate. These entries are used to classify the causes of death using the ICD-10. The literal text entries are input into the SuperMICAR program of the suite of software used to code the causes of death and thus these literal text entries are sometimes referred to as the SuperMICAR literal text.

The cause of death section from the US standard death certificate

This new program searches the electronic version of the literal text for references to specific drugs and other words of interest that are included in the cause of death statement and the “how the injury occurred” text box. The program creates additional variables in the data set to record the drug names and number of drugs.

After processing the literal text data with this program (and after a thorough manual review of the output), state and local analysts can learn more about the leading drugs involved in deaths in their jurisdictions. For example, the leading drugs involved may be oxycodone, methadone, heroin, and alprazolam. These data could be used for surveillance to monitor the number of deaths from a particular drug. Analysts could also use the search terms spreadsheet to monitor the literal text for new drug threats.

There are some limitations to the current program. Importantly, in some states these literal text data are restricted and may not be available to all analysts. The drug list is a work in progress and can be improved. Also, the program searches for mentions of a drug without considering the context. There are likely to be false positives; that is, a drug is mentioned, but did not contribute to the death. For instance, “insulin” is on the drug name list and would identify “non-insulin dependent diabetes.” Another example would be “heroin user” which does not necessarily imply that heroin was involved in the death.

Click here to download the tool
The downloadable zip file includes an Excel document, an SAS document, and a PDF document.

The tool is available here on the CSTE website. We are requesting feedback though updates will be made as times permits. Please send feedback to Nidal Kram. We hope that you will take the tool and expand on it and use it in your work reviewing other causes of death.

James Davis, MA is Substance Abuse Epidemiologist at the New Mexico Department of Health. This project was prepared by the CSTE overdose workgroup on literal text analysis, including Jennifer Sabel, Jim Davis, Dagan Wright, Svetla Slavova from CSTE; Margaret Warner, Ari Minino, from NCHS; and Len Paulozzi, and Rose Rudd from NCIPC. For more information, read about the CSTE overdose subcommittee and the CSTE Substance Abuse Subcommittee.

This post has not been tagged.

Share |
Permalink | Comments (0)