Transportation Safety: DOT Has Taken Steps to Verify and Publicize Drug and Alcohol Testing Data but Should Do More
Fast Facts
Drug and alcohol testing is required for certain transportation industry employees. Federal agencies that regulate these industries collect test result data from employers and use it to set random testing rates.
The Department of Transportation (DOT) office that collects the data and the agencies that use it take steps to verify its completeness and accuracy. But these steps vary across agencies and haven't been reviewed to ensure they are sufficient.
Also, DOT publishes the data on its website. But we found opportunities for DOT to improve the transparency of the data and to inform the public about it. Our recommendations address these issues.
Highlights
What GAO Found
Federal regulations require employers in the different transportation industries to: (1) randomly test their safety-sensitive employees for drugs or alcohol, or both, and (2) certify and report the aggregate data annually or when requested to the Department of Transportation (DOT). The DOT modal administrations and the U.S. Coast Guard (USCG) primarily use these data to determine the annual random testing rate (the specified minimum percentage of these employees that employers use to calculate the minimum number of random tests they must conduct that year). For example, for drugs, if 1 percent or more of employees in an industry randomly test positive, then the following year's random testing rate will be set at 50 percent.
DOT, the modal administrations, and USCG have taken steps to ensure that drug and alcohol testing data reported by employers are reliable:
- DOT incorporated automated checks to alert employers that the data they are submitting are outside of expected ranges and may be inaccurate.
- After data are reported, the modal administrations and USCG use varying processes to verify data, e.g., some compare data employers reported to physical records during routine compliance reviews.
These data verification processes improve the reliability of the data, but are not consistently used by each of the modal administrations and USCG. Officials have the discretion to use different data reliability processes and may have valid reasons for doing so. However, officials have not evaluated the sufficiency of these various processes and therefore do not have adequate assurance the data are reliable for calculating the random testing rates.
Laboratory Urinalysis for Drug Testing
In March 2019, DOT published aggregated drug and alcohol testing data on its website, as required. This website, updated annually, follows several key actions for transparently reporting government data, but does not follow other key actions, including:
- DOT has not disclosed known data limitations that exist in the public data that would prevent users from accurately calculating the random testing rate.
- DOT has not engaged with the public to encourage data use.
DOT officials told us there is limited interest in the data based on website traffic. However, because DOT has not taken steps to inform the public of the data, DOT may be missing opportunities to identify additional users and improvements.
Why GAO Did This Study
Since 1988, DOT has regulated the process by which employers in the different transportation industries (aviation, trucking, rail, transit, pipeline, and maritime) are required to test their employees for drug and alcohol use. Employers must self-report these test results annually to DOT or when requested by DOT. In a 2018 statute, Congress required DOT to publish the aggregate drug and alcohol testing data on DOT's website and included a provision for GAO to review the website and these data.
This report examines: (1) how DOT uses drug and alcohol testing data, (2) how DOT verifies that data are reliable, and (3) whether DOT follows key actions for transparently reporting drug and alcohol testing data. GAO reviewed relevant laws and regulations, among other things. To determine how DOT verifies that data are reliable, GAO reviewed documents, analyzed data in the internal database from calendar years 2003 through 2018, and interviewed DOT officials. GAO also reviewed the public website and compared it to key actions for open government data.
Recommendations
GAO is making several recommendations, including that DOT (1) evaluate the processes the modal administrations use to verify testing data, (2) disclose known limitations in the website's publicly reported testing data, and (3) reach out to the public to promote the website and evaluate the benefits and costs of other possible improvements. DOT concurred with our recommendations.
Recommendations for Executive Action
| Agency Affected | Recommendation | Status |
|---|---|---|
| Department of Transportation | The Secretary of Transportation should direct the Administrators of FAA, FMCSA, FRA, FTA, and PHMSA to: (1) evaluate the different processes used by each modal administration to verify drug and alcohol testing data—including comparing data to records during inspections, checking data for errors manually or with software, and contacting employers that do not submit a report or submit an incomplete report—and (2) determine what, if any, additional steps should be taken to improve the reliability of the information. (Recommendation 1) |
As of May 2024, the Administrators of the modal administrations are conducting reviews of their drug and alcohol data collection processes to identify process improvements, if any, that should be taken to improve the reliability of the information. GAO will continue to monitor and report on DOT's progress to address this recommendation.
|
| United States Coast Guard | The Commandant of the Coast Guard should direct the Assistant Commandant for Prevention Policy to: (1) evaluate the different processes used by USCG to verify drug and alcohol testing data—including comparing data to records during inspections, checking data for errors manually or with software, and contacting employers that do not submit a report or submit an incomplete report—and (2) determine what, if any, additional steps should be taken to improve the reliability of the information. (Recommendation 2) |
Federal regulations require employers in maritime industries to randomly test their safety-sensitive employees for drugs or alcohol, or both, and certify and report the aggregate data annually or when requested to U.S. Coast Guard (USCG). USCG primarily uses these data to determine the annual random testing rate, which determines the number of random tests that USCG must conduct that year. With rising drug use and the opioid crisis, USCG's efforts to accurately set the drug and alcohol testing rate of maritime employees and make the resulting information transparent are increasingly important. In 2021, GAO reported that while USCG had taken various steps to verify that drug and alcohol testing data are reliable, such as contacting an employer if they receive a submission but the information appeared to be incomplete, these validation efforts had been inconsistently implemented. Further, USCG did not know which processes were sufficient to ensure the quality and completeness of the data, because officials had not evaluated them. Having not evaluated the processes to discern whether the data have been sufficiently verified, or whether additional verification steps are needed, USCG does not have adequate assurance that the data are reliable for the purposes of calculating random testing rates. Therefore, GAO recommended that USCG (1) evaluate the different processes used by USCG to verify drug and alcohol testing data and (2) determine what, if any, additional steps should be taken to improve the reliability of the information. In 2023, GAO confirmed that USCG initiated a comprehensive evaluation of policies, procedures, and practices regarding drug and alcohol testing data. GAO obtained and reviewed USCG memoranda that summarized USCG's evaluation of the drug and alcohol testing data and listed the changes to USCG policies, procedures, and practices that were made as a result of the evaluation. As a result of this evaluation, USCG determined additional steps to improve its data, including (1) identifying and reporting suspect or invalid data to the cognizant Coast Guard field units to verify data, (2) organizing data reports by geographic locations, allowing investigators timely access to reports relevant to their employers, and (3) initiating audits to do a comparative analysis of the data. Collectively, these steps will provide USCG greater assurance that the data are reliable and the random testing rates are set at an appropriate level to deter drug and alcohol misuse by maritime employees and promote public safety.
|
| Office of Drug and Alcohol Policy and Compliance | The Director of ODAPC should disclose known limitations of drug and alcohol testing data on DOT's website, consistent with key actions for open government data. (Recommendation 3) |
Since 1988, DOT has regulated the process by which employers in the different transportation industries (aviation, trucking, rail, transit, pipeline, and maritime) are required to test their employees for drug and alcohol use. Employers must self-report these test results annually to DOT or when requested by DOT. DOT uses this data to set requirements for how many drug and alcohol tests employers must conduct the following year (known as the "random testing rate"). In a 2018 statute, Congress required DOT to publish the aggregate drug and alcohol testing data on DOT's website. In 2021, GAO reported that while DOT followed several key actions for transparently reporting government data, DOT did not follow the key action of disclosing known data issues, which is consistent with open government data. Specifically, DOT's drug and alcohol testing website does not inform the public of data limitations that would prevent users from calculating annual testing rates for two modal administrations-Federal Railroad Administration (FRA) and Federal Motor Carrier Safety Administration (FMSCA). Because DOT did not disclose these known limitations, the public may be confused about how the data are used to calculate the random testing rates and could view the data as unreliable if they cannot reconcile DOT's data with their own calculations. Further, not disclosing the limitations is inconsistent with the purpose of the public database, which is to provide greater transparency of drug and alcohol testing data. Therefore, GAO recommended that DOT disclose known limitations of drug and alcohol testing data on DOT's website, consistent with key actions for open government data. In 2023, GAO confirmed that DOT updated the data reported on its website to include a description of data limitations for FRA and FMCSA drug and alcohol testing data. Disclosing these data limitations provide greater transparency about drug and alcohol testing data and will aid the publics' ability to understand and use them.
|
| Office of Drug and Alcohol Policy and Compliance | The Director of ODAPC should reach out to potential users in the public to determine if there is a broader audience for the public data, consistent with key actions for open government data, and if a broader audience is identified, engage with users to evaluate the benefits and costs of adopting additional key actions for open government data and any other possible improvements to the website. (Recommendation 4) |
Since 1988, DOT has regulated the process by which employers in the different transportation industries (aviation, trucking, rail, transit, pipeline, and maritime) are required to test their employees for drug and alcohol use. Employers must self-report these test results annually to DOT or when requested by DOT. DOT uses this data to set requirements for how many drug and alcohol tests employers must conduct the following year (known as the "random testing rate"). In a 2018 statute, Congress required DOT to publish the aggregate drug and alcohol testing data on DOT's website. In 2021, GAO reported, in publishing the data on its website, DOT had not followed a key practice for open government data by reaching out to potential users in the public to encourage data use. Without this outreach, the public may not be aware of the website or the potential uses of the data. Because DOT has not identified or reached out to potential users in the public, the agency cannot be certain the public is not interested in the data. Further, DOT does not know whether the website meets the requirements for transparently providing data, or if additional improvements could make this a more valuable resource for users. Therefore, GAO recommended that DOT reach out to potential users in the public to determine if there is a broader audience for the public data, and, if a broader audience is identified, engage with users to evaluate the benefits and costs of adopting additional key actions for open government data and any other possible improvements to the website. In 2022, DOT published a Request for Information in the Federal Register to request information from potential users of the data, including (1) which entities use the data and how it is used, (2) the functionality of the drug and alcohol testing data website, and whether users have specific recommendations regarding possible improvements to the website that would enhance the user's ability to use the available data, and (3) whether users envision other appropriate uses (other than to determine the annual random testing rate for each industry) for the drug and alcohol testing data on the website. The comment period was open from June 30, 2022, through August 1, 2022. In April 2023, DOT provided GAO with the four comments received in response to the request. This limited response indicates that there is not a broader use for the drug and alcohol testing data. Even so, officials noted that they will continue to discuss the availability of the public drug and alcohol testing data on the website at industry and government conferences and training sessions. DOT's efforts to identify potential users of the data are consistent with key practices for open government data, and as a result, DOT can be assured that they are not missing a broader audience for the data.
|