ACKNOWLEDGMENTS

This report was prepared by the Center for Behavioral Health Statistics and Quality (CBHSQ), Substance Abuse and Mental Health Services Administration (SAMHSA), U.S. Department of Health and Human Services (HHS), and by RTI International (a trade name of Research Triangle Institute, Research Triangle Park, NC). Work by RTI was performed under Contract No. HHSS283200700002I.

PUBLIC DOMAIN NOTICE

All material appearing in this publication is in the public domain and may be reproduced or copied without permission from SAMHSA. Citation of the source is appreciated. However, this publication may not be reproduced or distributed for a fee without the specific, written authorization of the Office of Communications, SAMHSA, HHS.

RECOMMENDED CITATION

Center for Behavioral Health Statistics and Quality (2013). Guide to DAWN Trend Tables, 2011 Update. Rockville, MD: Substance Abuse and Mental Health Services Administration.

ELECTRONIC ACCESS

This publication may be downloaded from https://store.samhsa.gov. Or please call SAMHSA at

1-877-SAMHSA-7 (1-877-726-4727)
(English and Español).

ORIGINATING OFFICE

Center for Behavioral Health Statistics and Quality
Substance Abuse and Mental Health Services Administration
1 Choke Cherry Road, Rockville, MD 20857

April 2013

 

CONTENTS

1. Major features of DAWN Trend Tables, 2011 Update
1.1 Analytic groups
1.2 Workbooks containing national estimates
1.3 Workbooks containing metropolitan area estimates
1.4 Information organization and format
1.5 Special note on age categories

2. Values reported in DAWN Trend Tables
2.1 Weighted annual estimates
2.2 Rates per 100,000 population
2.3 Relative standard error (%)
2.4 Percent change (p < 0.05)
2.5 Lower and upper 95 percent confidence limit on weighted annual estimate
2.6 Suppression

List of Tables

Table 1. DAWN analytic groups
Table 2. Workbook names for national estimates
Table 3. Workbook names for metropolitan areas and divisions
Table 4. Tables in each workbook of the DAWN Trend Tables

1. MAJOR FEATURES OF DAWN TREND TABLES, 2011 UPDATE

Drug Abuse Warning Network (DAWN) Trend Tables provide estimates of drug-related visits to hospital emergency departments (EDs) for different groups of patients, different years, and different geographic locations. Each year DAWN produces an updated set of DAWN Trend Tables that includes data for 2004 through the current year. The DAWN Trend Tables, 2011 Update, includes 120 Microsoft Excel workbooks: 10 workbooks contain estimates for the Nation, and the same 10 workbooks are repeated for each of 11 metropolitan areas. Each workbook contains 56 tables (1 table per worksheet). Each table presents data for 2004, 2005, 2006, 2007, 2008, 2009, 2010, and 2011. This document is intended to help DAWN users find the workbooks, tables, and estimates of interest to them.

Further information about DAWN is available at the DAWN website, including detail on the DAWN data program and the methodologies used to collect, process, and report DAWN data.¹ Information on other sources of data on substance abuse and mental health from the Center for Behavioral Health Statistics and Quality is located at https://www.samhsa.gov/data/.

1.1 Analytic groups

DAWN analytic groups represent different groupings of visits that were developed to meet the data needs of a range of audiences. The DAWN analytic groups and their definitions are provided in Table 1.

1.2 Workbooks containing national estimates

Workbook names have three parts: a prefix that describes the geographic coverage of the workbook, a middle term that reflects the latest year of the data, and a suffix that describes the analytic group. Table 2 lists the workbook names for the 10 workbooks containing estimates for the Nation. Each worksheet contains data for 2004 through 2011.²

1.3 Workbooks containing metropolitan area estimates

DAWN prepares estimates each year for DAWN metropolitan areas that have sufficient participation to support estimates with acceptable reliability and precision. Table 3 lists the names of workbooks containing estimates for metropolitan areas and divisions. For example, the workbook containing national estimates for ED visits involving all drug misuse or abuse is named "Nation_2011_AllMA.xls." The workbook with parallel estimates for Boston is named "Boston_2011_AllMA.xls." Each of the 11 geographic areas listed in Table 3 has a set of 10 Excel workbooks, one workbook for each analytic group listed in Table 2.

1.4 Information organization and format

Workbooks and worksheets

With minor exceptions, each Excel workbook in the DAWN Trend Tables is organized the same way. Each workbook represents one geographic area and one analytic group. Each workbook contains 56 tables (worksheets), with each table representing visits for a single demographic or visit characteristic. Examples of a demographic characteristic are "male patients" or "patients aged 0 to 5." An example of a visit characteristic is "the patient was discharged home." Table 4 lists the tables that appear in each workbook. Unless otherwise noted in Table 4, each table in each workbook has the same arrangement of rows and columns.

Rows

The rows of the tables represent drug categories and drugs. Each drug and each drug category appear on the same row in each table (e.g., the estimates of ED visits involving cocaine appear on row 15 in every table). The classification of drugs used in DAWN is derived from the Multum Lexicon, © 2012 Lexi-Comp, Inc., and/or Cerner Multum, Inc. The Lexicon was slightly modified to meet DAWN's need to report on illicit drugs using street names.

Columns

The 39 columns in each table provide the following information:

• drug/drug group name;
• weighted annual estimates of ED visits for 2004 through 2011;
• rates of ED visits per 100,000 population for 2004 through 2011;
• relative standard error (RSE) of estimate and rate, expressed as a percentage, of the visit estimates for 2004 through 2011;
• tests for statistically significant differences between visit estimates for select years (e.g., in the DAWN Trend Tables, 2011 Update, estimates for 2011 are compared with those for 2004, 2009, and 2010);³ and
• lower and upper 95 percent confidence intervals (CIs) for visit estimates for 2004 through 2011.

1.5 Special note on age categories

The age categories used for reporting ED visits in the DAWN Trend Tables reflect critical junctures in drug use. For example, patients aged 5 and under are reported separately to facilitate study of visits involving accidental ingestion. Patients aged 12 to 17 are considered to be in their formative years, and understanding the nature of their drug use is important for prevention efforts. Patients under the age of 21 are reported separately to facilitate study of topics such as underage drinking. Patients aged 18 to 20 are reported separately from those aged 21 to 24 to isolate drug-taking behaviors before and after the critical age of 21. Older patients are grouped in wider categories where age differences are not as critical to intervention and treatment. As a consequence, the age categories are not evenly sized—for example, the age group 30 to 34 covers 5 years, whereas the age group 35 to 44 covers 10 years. The size of an age group is an important consideration when comparing estimates of ED visits for different age categories. 

2. VALUES REPORTED IN DAWN TREND TABLES

The following values appear in the cells of the DAWN Trend Tables. The order in which values are listed here corresponds to the order in which they appear in the standard table format, going from left to right across the table. The section headings below match the overarching headings that appear in row 6 of the standard table format.

2.1 Weighted annual estimates

Estimates of drug-related ED visits are calculated by applying weights and adjustments to the data provided by the sampled hospitals participating in DAWN. The primary sampling weights reflect the probability of selection, whereas separate adjustment factors are included to account for nonresponse, data quality, and the known total of ED visits delivered by the universe of eligible hospitals, as reported by the most current American Hospital Association survey.  

Estimates of the number of visits are reported to the fifth decimal place, though they are formatted to appear as whole numbers. An asterisk (*) appears if a value is suppressed (see Section 2.6). All estimate fields are populated with a value or an asterisk; blanks do not appear and are not valid.

2.2 Rates per 100,000 population

All rates in the DAWN Trend Tables are visits per 100,000 persons. Rates are standardized measures that are helpful when comparing levels of drug-related ED visits for different years, types of visits, and demographic groups. DAWN rates for years and types of visits are based on the whole population; for example, there were 790.4 ED visits involving drug misuse or abuse per 100,000 population in 2011 compared with 552.9 visits in 2004. For specific age groups and sexes, the denominator is limited to the population in that age group or sex. For age in particular, rates are important because the size of the underlying population differs considerably across DAWN age groups (see Section 1.5). For example, in 2011 there were 13.5 million individuals aged 18 to 20 in the United States and 40.6 million aged 35 to 44. Accordingly, the estimated number of ED visits involving drug misuse or abuse was 239,281 visits for those aged 18 to 20 and 432,957 visits for those aged 35 to 44. To allow comparison between the two age groups, standardized rates are calculated as the number of ED visits per 100,000 persons in that age group. In this example, while the estimated number of visits was much lower for persons aged 18 to 20, the rate of visits was substantially higher: 1,771.5 for persons aged 18 to 20 compared with 1,065.7 visits for persons aged 35 to 44. Because they are reported as percentages, the RSEs provided in DAWN tables apply equally to the estimates and the rates.

Rates are reported to the ninth decimal place, though they are formatted to display only the first decimal (e.g., 123.4). Rates based on suppressed estimates are likewise suppressed, with an asterisk (*) appearing instead of a value (see Section 2.6). DAWN does not produce population-based rates for race/ethnicity categories because race/ethnicity information in ED records is often missing or is very limited. Three dots (…) appear in the rate fields in the race/ethnicity tables. All rate fields are populated with a value, asterisk, or three dots; blanks do not appear and are not valid.

2.3 Relative standard error (%)

Because DAWN relies on a sample of hospitals, each estimate produced from the DAWN ED data is subject to sampling variability, the variation in the estimate that would be observed naturally if different samples were drawn from the same population using the same procedures. The sampling variability of an estimate in this publication is measured by its relative standard error (RSE). The precision of an estimate or rate is inversely related to its RSE. That is, the greater the RSE, the lower the precision.

RSEs are reported to the ninth decimal place, though they are formatted to display only the first decimal (e.g., 12.3). The RSE values reported are percentages (e.g., 12.3 = 12.3%). Because it is reported as a percentage, an RSE measure applies to both the estimate and the rate. RSEs based on suppressed estimates are likewise suppressed, with an asterisk (*) appearing instead of a value (see Section 2.6). All RSE fields are populated with a value or an asterisk; blanks do not appear and are not valid.

2.4 Percent change (p < 0.05)

The DAWN Trend Tables assess between-year changes by comparing estimates as follows:

• most current year to first year,
• most current year to year before last, and
• most current year to last year.

In the DAWN Trend Tables, 2011 Update, 2011 estimates are compared with those for 2004 (first year), 2009 (year before last), and 2010 (last year). The underlying formula is of the form: ((estimate for earlier year - estimate for later year) / estimate for earlier year). 

The resulting values are reported to the seventh decimal place, though they are formatted to display a whole number that represents a percentage difference (e.g., 12 = 12% increase in the number of visits). Declines in percentage difference appear as negative numbers (e.g., -12 = 12% decrease in the number of visits). The tables report percentage differences between years only if they are statistically significant at the p < 0.05 level; otherwise, a blank appears. A blank also appears if either estimate in the percentage difference formula is suppressed (see Section 2.6). Percent change measures for rates are not provided.

2.5 Lower and upper 95 percent confidence limit on weighted annual estimate

The DAWN Trend Tables include the lower and upper boundaries of the confidence intervals (CIs) for all estimates at the 95 percent confidence level. For example, the estimate of the number of ED visits involving any type of drug misuse or abuse in 2011 was 2,462,948 visits. A 95 percent CI means that if repeated samples were drawn from the same population of hospitals using the same sampling and data collection procedures, the number of ED visits reported (2,462,948 visits) will fall between the lower boundaries (2,112,868 visits) and upper boundaries (2,813,028 visits) 95 percent of the time.

The confidence limit estimates are reported to the eighth decimal place, though they are formatted to appear as whole numbers. An asterisk (*) appears if a value is suppressed (see Section 2.6). All CI fields are populated with a value or an asterisk; blanks do not appear and are not valid.

2.6 Suppression

DAWN estimates with RSE values greater than 50 percent or estimates based on fewer than 30 ED visits (weighted or unweighted) are considered too imprecise for publication and are not shown. An asterisk (*) is displayed in the place of a suppressed estimate or any value based on a suppressed estimate (i.e., rate, RSE, percent change, lower CI, upper CI).