|Multiple improvements have caused a permanent disruption in trends for the Drug Abuse Warning Network (DAWN). As a result, comparisons cannot be made between old DAWN (2002 and prior years) and new DAWN. Changes to DAWN affect all of the following areas:
In January 2003, the Substance Abuse and Mental Health Services Administration (SAMHSA) launched its redesign of the Drug Abuse Warning Network (DAWN). The new design brought needed improvements to the quality and utility of DAWN data. A consequence of these sweeping changes is that comparisons between old DAWN and new DAWN cannot be made.
This issue of The DAWN Report reviews the numerous changes to DAWN and summarizes why any comparisons between old and new DAWN would produce erroneous results and conclusions.
Although the new DAWN was born in 2003, the redesign began in 1997.1 Every aspect of DAWN was evaluated in terms of its contribution, potential benefit for users, and feasibility. The evaluation generated recommendations for a new design.
Before any changes were adopted, the recommendations were deliberated by external advisors and major stakeholders.1 When disruption of trends became a possibility, the consensus was to focus on achieving DAWN's future potential, rather than preserving links to its past. Therefore, no data were collected to map old data with new data.
No single improvement is responsible for the lack of comparability between old DAWN and new DAWN. Instead, it is the collective impact of many changes (Table 1), which include the following:
National estimates. A new sample of hospitals supports true national estimates. The prior sample produced estimates for only the coterminous United States, excluding Alaska and Hawaii.
Metropolitan-area estimates. New DAWN has added metropolitan areas that old DAWN did not cover. Also, metropolitan area boundaries have been updated to reflect the 2000 census. These boundary changes, which affect 13 of the original 21 metropolitan areas in old DAWN, prevent comparisons with any previous estimates.
New case criteria. DAWN now collects data on all types of drug-related emergency department (ED) visits. This includes many cases not captured before, excludes others, and overlaps with some (Table 2). Although DAWN continues to collect drug abuse cases, even those cases have changed and cannot be compared.
Information about each case. Changes in data items prevent new DAWN data from being mapped to the old data. Data items that were not useful or obtainable were eliminated if they could not be improved. New data items were added to better describe DAWN cases and distinguish among different types of drug-related visits.
New case finding methods. DAWN now requires a review of every ED chart to find DAWN cases. This method identifies cases more systematically, consistently, and completely than ever before. It also compromises any comparability between the new and old data.
Quality assurance. Performance measurement, systematic training, proactive problem identification, and on-site interventions are some of the approaches now used to improve the quality of DAWN data. These new quality protocols also render the new DAWN incomparable with the old DAWN.
|For more information about new DAWN, see Substance Abuse and Mental Health Services Administration, Office of Applied Studies. Drug Abuse Warning Network, 2003: Interim National Estimates of Drug-Related Emergency Department Visits. DAWN Series D-26. DHHS Publication No. (SMA) 04-3972. Rockville, MD, 2004.
This and other DAWN publications are available online at http://DAWNinfo.samhsa.gov
The DAWN Report is published periodically by the Substance Abuse and Mental Health Services Administration (SAMHSA), Office of Applied Studies (OAS). All material in this report is in the public domain and may be reproduced or copied without permission from SAMHSA. Citation of the source is appreciated.
The author of this issue is Judy K. Ball, Ph.D., M.P.A. (DAWN Project Director, SAMHSA/OAS). Other major contributors were Lori Ducharme, Ph.D., James Green, M.A., and other staff of Westat who were responsible for the original disruption of trends assessment under Contract No. 283-02-9025.