Skip to Main Content
This report was prepared by the Office of Applied Studies (OAS), Substance Abuse and Mental Health Services Administration (SAMHSA), and by RTI International (a trade name of Research Triangle Institute, Research Triangle Park, NC). Work by RTI was performed under contract number 280-03-2602. The co-authors are Judy K. Ball, Ph.D., M.P.A., previously of SAMHSA, and Victoria A. Albright, RTI, with the assistance of David Skellan, OAS/SAMHSA.
All material appearing in this report 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, U.S. Department of Health and Human Services.
Ball, J., & Albright, V., Drug Abuse Warning Network, 2005: Area Profiles of Drug-Related Mortality. Rockville, MD: Office of Applied Studies, Substance Abuse and Mental Health Services Administration, 2009.
(English and Español).
Office of Applied Studies
Substance Abuse and Mental Health Services Administration
1 Choke Cherry Road, Rockville, MD 20857
Profiles and Spotlights
Table A: Metro area overview: Deaths and population by county, 2005
Figure B: Deaths by manner of death, 2005
Figure C: Top 5 drugs involved: Drug-related deaths, 2005 Top 5 drugs involved: Drug-related suicide deaths, 2005
Figure D: Death rates by gender and age: Drug-related, 2005 Death rates by gender and age: Drug-related suicide deaths, 2005
Table E: Place of death, 2005
Table F: Drug-related deaths by drug category: 2004-2005
Limitations to data
Abbreviated profiles for selected metropolitan areas
Metropolitan Area Profiles
Atlanta-Sandy Springs-Marietta, GA
Barnstable Town, MA
Buffalo-Niagara Falls, NY
Burlington-South Burlington, VT
Houston-Sugar Land-Baytown, TX
Kansas City, MO-KS
Los Angeles-Long Beach-Santa Ana, CA
Louisville-Jefferson County, KY-IN
Milwaukee-Waukesha-West Allis, WI
Minneapolis-St. Paul-Bloomington, MN-WI
New York-Northern New Jersey-Long Island, NY-NJ-PA
Oklahoma City, OK
Portland-South Portland-Biddeford, ME
Salt Lake City, UT
San Diego-Carlsbad-San Marcos, CA
St. Louis, MO-IL
Abbreviated Profiles for Areas With Few Drug-Related Deaths
Las Cruces, NM
Santa Fe, NM
Sioux Falls, SD
St. George, UT
Albuquerque, NM: Bernalillo County, NM
Atlanta-Sandy Springs-Marietta, GA: Fulton County, GA
Baltimore-Towson, MD: Baltimore City, MD
Baltimore-Towson, MD: Baltimore County, MD
Boston-Cambridge-Quincy, MA-NH: Essex County, MA
Boston-Cambridge-Quincy, MA-NH: Middlesex County, MA
Boston-Cambridge-Quincy, MA-NH: Norfolk County, MA
Boston-Cambridge-Quincy, MA-NH: Plymouth County, MA
Boston-Cambridge-Quincy, MA-NH: Suffolk County, MA
Buffalo-Niagara Falls, NY: Erie County, NY
Chicago-Naperville-Joliet, IL-IN-WI: Cook County, IL
Chicago-Naperville-Joliet, IL-IN-WI: Lake County, IL
Denver-Aurora, CO: Adams County, CO
Denver-Aurora, CO: Denver County, CO
Detroit-Warren-Livonia, MI: Macomb County, MI
Detroit-Warren-Livonia, MI: Oakland County, MI
Detroit-Warren-Livonia, MI: Wayne County, MI
Kansas City, MO-KS: Jackson County, MO
Minneapolis-St. Paul-Bloomington, MN-WI: Hennepin County, MN
New Orleans-Metairie-Kenner, LA (Pre-Katrina): Jefferson Parish, LA
New Orleans-Metairie-Kenner, LA (Post-Katrina): Jefferson Parish, LA
New York-Northern New Jersey-Long Island, NY-NJ-PA: Bronx County, NY
New York-Northern New Jersey-Long Island, NY-NJ-PA: Kings County, NY
New York-Northern New Jersey-Long Island, NY-NJ-PA: New York County, NY
New York-Northern New Jersey-Long Island, NY-NJ-PA: Queens County, NY
New York-Northern New Jersey-Long Island, NY-NJ-PA: Suffolk County, NY
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD: Delaware County, PA
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD: Philadelphia County, PA
Portland-Vancouver-Beaverton, OR-WA: Multnomah County, OR
Provo-Orem, UT: Utah County, UT
Salt Lake City, UT: Salt Lake County, UT
San Francisco-Oakland-Fremont, CA: San Mateo County, CA
Seattle-Tacoma-Bellevue, WA: King County, WA
Seattle-Tacoma-Bellevue, WA: Snohomish County, WA
Springfield, MA: Hampden County, MA
St. Louis, MO-IL: St. Louis City, MO
St. Louis, MO-IL: St. Louis County, MO
Tulsa, OK: Tulsa County, OK
Washington-Arlington-Alexandria, DC-VA-MD-WV: District of Columbia
List of Tables
Table 1. Participation of medical examiner/coroner jurisdictions in DAWN, 2005
Table 2. Rates of drug-related deaths and drug-related suicides per 100,000 population, 2005
Table 3. Rates of drug-related deaths and percentage change, 2004 and 2005
Table 4. Metropolitan area profiles and county spotlights
Table 5. State profiles and metropolitan area profiles wholly in those States
The Drug Abuse Warning Network (DAWN) is a public health surveillance system that monitors drug-related deaths referred to medical examiners and coroners (ME/Cs) in selected metropolitan areas and States. Findings in this publication reflect data on drug-related deaths that occurred during calendar year 2005 and were reported by participating ME/Cs to DAWN. In selected tables, data from reporting year 2004 are included for comparison. The Office of Applied Studies (OAS) of the Substance Abuse and Mental Health Services Administration (SAMHSA), U.S. Department of Health and Human Services, is responsible for DAWN.
The mortality component of DAWN does not rely on a statistical sample of ME/Cs. Findings cannot be considered representative of ME/Cs that did not participate, and results cannot be extrapolated to the United States as a whole. DAWN mortality data for 2003 and later are not comparable to mortality data for any years prior to 2003 because of changes introduced in the 2003 reporting year.
Since 2003, a DAWN case is any death reviewed by an ME/C that was related to recent drug use. Findings in this publication pertain to drug-related deaths and drug-related suicide deaths reported by participating death investigation jurisdictions as DAWN cases.1
DAWN cases are identified through a retrospective review of decedent case files in each participating death investigation jurisdiction. A DAWN case is any death that is determined by the ME/C as being related to drug use. The relationship between the death and the drug need not be causal; the drug need only be implicated in the death. The drug use may have been for legitimate, therapeutic purposes of for the purpose of drug abuse/misuse, but in either case, the drug use must have been recent.
These eligibility criteria for a DAWN case are intentionally broad and inclusive. Since death record documentation varies in clarity and comprehensiveness across jurisdictions, broad criteria reduce the potential for judgment calls that could cause data to vary systematically and unexpectedly across reporters and jurisdictions. Broad criteria also capture a diverse set of drug-related deaths that support a wide variety of analytical purposes and interests.
For decedents under the age of 21, DAWN cases include deaths where the only drug involved was alcohol. For those 21 or older, there must be at least one other drug involved besides alcohol for the death to be a DAWN case.
The data items submitted on drug-related deaths are listed in Appendix A.
Drugs that make a death eligible for DAWN include:
Findings in this publication focus on two major categories of drug-related deaths, based on the manner of death as determined by the ME/C.
(1) Drug-related deaths (other than drug-related suicide deaths) include the following:
(2) Drug-related suicide deaths include suicide deaths with drug involvement. The determination of suicide is made by the ME/C. Because of the broad eligibility criteria for determining DAWN cases, drug-related suicide deaths include more than deaths due to overdoses. A reported drug may not be the cause of the suicide death even if only one drug was involved.
Findings reported in this publication are based on concluded investigations that were submitted by May 18, 2006, for deaths that occurred during 2005. Death investigations that were not concluded by the ME/C by the end of the data collection period are excluded.
Death rates (i.e., the number of deaths per 100,000 population) are reported to permit comparisons within or across areas or across demographic subgroups.4 This use of death rates, as opposed to counts, is important because two areas with similar numbers of drug-related deaths may have vastly different populations. Rates, which take population differences into account, standardized these comparisons.
Readers are cautioned, though, in drawing conclusions based on comparisons of death rates. While differences in rates may signify differences in underlying drug-related mortality (or a lack of differences may suggest similarity), other factors may confound such comparisons. For example, State laws dictate which deaths are subject to ME/C review. These laws vary by State and, within each State, by time. Within ME/C offices, toxicology testing practices vary depending on local concerns, funding, and testing technology. Such factors will affect the number of deaths determined to be DAWN cases and the number of deaths attributed to particular drugs. Small changes in the number of deaths (e.g., an increase of 5 deaths) can result in a large percentage difference if the base is small (e.g., an increase of 5 to 10 deaths is a 100% increase). Even though there is no sampling error in DAWN ME/C data, the possibility of nonsampling errors (i.e., errors in reporting, changes in testing protocols) suggests that caution should be observed in assigning importance to such findings.
DAWN relies on the voluntary cooperation of ME/Cs in selected areas of the United States to provide standardized data on drug-related deaths. For 2005, 169 jurisdictions in 51 metropolitan areas and 217 jurisdictions in 8 States submitted mortality data to DAWN.5 Metropolitan area definitions used by DAWN for 2003 and later years are those established by the Office of Management and Budget (OMB). (See Appendix A for additional detail.)
Table 1 provides information on the metropolitan areas and States that participated in 2005. It includes the following:
|Number||Percent of total||Population||Percent of
|SOURCE: Office of Applied Studies, SAMHSA, Drug Abuse Warning Network, 2005 (08/2008 update).|
|Fifty-one metropolitan areas||322||169||52%||92,443,457||72%|
|Atlanta-Sandy Springs-Marietta, GA||28||8||29%||3,248,432||65%|
|Barnstable Town, MA||1||1||100%||226,161||100%|
|Buffalo-Niagara Falls, NY||2||2||100%||1,144,796||100%|
|Burlington-South Burlington, VT||3||3||100%||205,222||100%|
|Dallas-Fort Worth-Arlington, TX||12||2||17%||1,215,920||21%|
|Houston-Sugar Land-Baytown, TX||10||1||10%||3,762,844||70%|
|Kansas City, MO-KS||15||4||27%||1,039,655||53%|
|Las Cruces, NM||1||1||100%||189,306||100%|
|Los Angeles-Long Beach-Santa Ana, CA||2||1||50%||9,941,197||77%|
|Louisville-Jefferson County, KY-IN||13||1||8%||699,051||58%|
|Milwaukee-Waukesha-West Allis, WI||4||1||25%||918,673||61%|
|Minneapolis-St. Paul-Bloomington, MN-WI||13||7||54%||1,832,939||58%|
|New Orleans-Metairie-Kenner, LA||7||3||43%||547,753||42%|
|New York-Northern New Jersey-Long Island, NY-NJ-PA||23||10||43%||11,101,883||59%|
|Oklahoma City, OK||7||7||100%||1,154,991||100%|
|Portland-South Portland-Biddeford, ME||3||3||100%||512,992||100%|
|Salt Lake City, UT||3||3||100%||1,046,685||100%|
|San Diego-Carlsbad-San Marcos, CA||1||1||100%||2,936,609||100%|
|San Francisco-Oakland-Fremont, CA||5||2||40%||948,278||23%|
|Santa Fe, NM||1||1||100%||140,801||100%|
|Sioux Falls, SD||4||1||25%||160,051||77%|
|St. George, UT||1||1||100%||119,188||100%|
|St. Louis, MO-IL||17||9||53%||2,362,217||84%|
An awareness of the extent of DAWN's coverage within a given area is essential to an accurate interpretation of DAWN mortality data. ME/C participants in DAWN are not part of a scientific sample at either the metropolitan or the national level. Within a metropolitan area, findings based on participating jurisdictions are not representative of nonparticipating jurisdictions. Reports from only a portion of jurisdictions within a metropolitan area can be extrapolated neither to the metropolitan area as a whole nor to the nation as a whole. 6
While the data do not support any representations at a national level, some generalizations can be made at a metropolitan level even if some ME/Cs do not participate. For example, while only 1 (10%) of the 10 counties that comprise the Houston, TX, metropolitan area participated in DAWN in 2005, that county is home to 70 percent of the area's total population. The important consideration is population coverage, not ME/C participation, per se.
Among the metropolitan areas listed in Table 1, population coverage exceeded 90 percent in 29 metropolitan areas, with 100 percent coverage in 25 of those areas. The remaining metropolitan areas had response rates that range from a low of 21 percent in Dallas-Fort Worth-Arlington to 85 percent in Chicago-Naperville-Joliet. Population coverage below 50 percent usually equates to the absence of large jurisdictions.
Table 2 reports the rates of drug-related deaths and drug-related suicide deaths per 100,000 population for metropolitan areas and States that participated in DAWN in 2005. Table 3 compares the rates of drug-related deaths in 2005 with those found for 2004 and reports the percentage change. (Comparisons are not made for drug-related suicide deaths because of their small numbers.) Table 3 is limited to those areas where the same jurisdictions participated in 2004 as in 2005. Both tables include indicators of the population coverage in DAWN for 2005.
|Metropolitan area or State||Rate of drug-related deaths per
|Deaths||Suicide deaths||Population||Percent of area
|* Drug-related deaths exclude drug-related suicide deaths.
SOURCE: Office of Applied Studies, SAMHSA, Drug Abuse Warning Network, 2005 (08/2008 update).
|Atlanta-Sandy Springs-Marietta, GA||6.2||0.6||3,248,432||65%|
|Barnstable Town, MA||14.6||1.8||226,161||100%|
|Buffalo-Niagara Falls, NY||9.3||1.1||1,144,796||100%|
|Burlington-South Burlington, VT||15.6||3.4||205,222||100%|
|Dallas-Fort Worth-Arlington, TX||6.5||1.6||1,215,920||21%|
|Houston-Sugar Land-Baytown, TX||11.7||1.9||3,762,844||70%|
|Kansas City, MO-KS||9.0||1.9||1,039,655||53%|
|Las Cruces, NM||7.9||2.6||189,306||100%|
|Los Angeles-Long Beach-Santa Ana, CA||10.9||1.3||9,941,197||77%|
|Louisville-Jefferson County, KY-IN||10.9||1.3||699,051||58%|
|Milwaukee-Waukesha-West Allis, WI||17.9||1.9||918,673||61%|
|Minneapolis-St. Paul-Bloomington, MN-WI||6.0||1.1||1,832,939||58%|
|New Orleans-Metairie-Kenner, LA||26.3||2.2||547,753||42%|
|New York-Northern New Jersey-Long Island, NY-NJ-PA||8.7||0.7||11,101,883||59%|
|Oklahoma City, OK||6.4||1.1||1,154,991||100%|
|Portland-South Portland-Biddeford, ME||10.5||2.9||512,992||100%|
|Salt Lake City, UT||18.3||2.4||1,046,685||100%|
|San Diego-Carlsbad-San Marcos, CA||10.7||2.0||2,936,609||100%|
|San Francisco-Oakland-Fremont, CA||10.2||3.1||948,278||23%|
|Santa Fe, NM||14.9||1.4||140,801||100%|
|Sioux Falls, SD||0.6||2.5||160,051||77%|
|St. George, UT||10.9||1.7||119,188||100%|
|St. Louis, MO-IL||10.1||1.7||2,362,217||84%|
|Metropolitan area or State*||Rate of drug-related
deaths per 100,000
2004 to 2005
|* The table is limited to those areas where the same jurisdictions participated in both years.
† Drug-related deaths exclude drug-related suicide deaths.
SOURCE: Office of Applied Studies, SAMHSA, Drug Abuse Warning Network, 2005 (08/2008 update).
|Buffalo-Niagara Falls, NY||8.7||9.3||7.8%||1,144,796||100%|
|Houston-Sugar Land-Baytown, TX||9.5||11.7||23.3%||3,762,844||70%|
|Los Angeles-Long Beach-Santa Ana, CA||10.5||10.9||4.4%||9,941,197||77%|
|Louisville-Jefferson County, KY-IN||8.2||10.9||33.3%||699,051||58%|
|Milwaukee-Waukesha-West Allis, WI||16.0||17.9||11.8%||918,673||61%|
|Minneapolis-St. Paul-Bloomington, MN-WI||4.6||6.0||29.9%||1,832,939||58%|
|New York-Northern New Jersey-Long Island, NY-NJ-PA||9.5||8.7||-9.1%||11,101,883||59%|
|Portland-South Portland-Biddeford, ME||9.8||10.5||7.6%||512,992||100%|
|Salt Lake City, UT||16.9||18.3||8.6%||1,046,685||100%|
|San Diego-Carlsbad-San Marcos, CA||9.6||10.7||10.9%||2,936,609||100%|
|St. Louis, MO-IL||9.0||10.1||12.7%||2,362,217||84%|
DAWN mortality data are reported for metropolitan areas with 30 or more deaths and for all participating States in six figures and tables that span two or more pages. These are referred to as "full profiles." Metropolitan areas with fewer than 30 deaths or areas with less than 50 percent coverage receive an "abbreviated profile" that includes just one of the six tables. Large, individual jurisdictions that are part of a multijurisdictional area and reported 60 or more deaths receive a full profile that is referred to as a "spotlight."
Among the 51 metropolitan areas, full profiles are provided for 36 metropolitan areas. Abbreviated profiles are provided for 12 metropolitan areas that submitted 30 or fewer drug-related deaths and for 3 metropolitan areas with less than 50 percent population coverage. Spotlights are provided for 39 individual jurisdictions.
|Type of profile||Metropolitan area profile||County spotlight(s)|
|* Due to Hurricane Katrina, sufficient data were only available to prepare a pre-Katrina abbreviated profile for the New Orleans-Metairie-Kenner, LA, metropolitan area and to prepare pre- and post-Katrina spotlights for Jefferson Parish, LA|
|Full||Albuquerque, NM||Bernalillo County, NM|
|Full||Atlanta-Sandy Springs-Marietta, GA||Fulton County, GA|
|Full||Baltimore-Towson, MD||Baltimore City, MD
Baltimore County, MD
|Full||Barnstable Town, MA||None|
|Full||Boston-Cambridge-Quincy, MA-NH||Essex County, MA
Middlesex County, MA
Norfolk County, MA
Plymouth County, MA
Suffolk County, MA
|Full||Buffalo-Niagara Falls, NY||Erie County, NY|
|Full||Burlington-South Burlington, VT||None|
|Full||Chicago-Naperville-Joliet, IL-IN-WI||Cook County, IL
Lake County, IL
|Abbreviated||Dallas-Fort Worth-Arlington, TX||None|
|Full||Denver-Aurora, CO||Adams County, CO
Denver County, CO
|Full||Detroit-Warren-Livonia, MI||Macomb County, MI
Oakland County, MI
Wayne County, MI
|Full||Houston-Sugar Land-Baytown, TX||None|
|Full||Kansas City, MO-KS||Jackson County, MO|
|Abbreviated||Las Cruces, NM||None|
|Full||Los Angeles-Long Beach-Santa Ana, CA||None|
|Full||Louisville-Jefferson County, KY-IN||None|
|Full||Milwaukee-Waukesha-West Allis, WI||None|
|Full||Minneapolis-St. Paul-Bloomington, MN-WI||Hennepin County, MN|
|Abbreviated||New Orleans-Metairie-Kenner, LA*||Jefferson Parish, LA (Pre-Katrina)
Jefferson Parish, LA (Post-Katrina)
|Full||New York-Northern New Jersey-Long Island, NY-NJ-PA||Bronx County, NY
Kings County, NY
New York County, NY
Queens County, NY
Suffolk County, NY
|Full||Oklahoma City, OK||None|
|Full||Philadelphia-Camden-Wilmington, PA-NJ-DE-MD||Delaware County, DE
Philadelphia County, PA
|Full||Portland-South Portland-Biddeford, ME||None|
|Full||Portland-Vancouver-Beaverton, OR-WA||Multnomah County, OR|
|Full||Provo-Orem, UT||Utah County, UT|
|Full||Salt Lake City, UT||Salt Lake County, UT|
|Full||San Diego-Carlsbad-San Marcos, CA||None|
|Abbreviated||San Francisco-Oakland-Fremont, CA||San Mateo County: CA|
|Abbreviated||Santa Fe, NM||None|
|Full||Seattle-Tacoma-Bellevue, WA||King County, WA
Snohomish County, WA
|Full||Springfield, MA||Hampden County, MA|
|Abbreviated||Sioux Falls, SD||None|
|Abbreviated||St. George, UT||None|
|Full||St. Louis, MO-IL||St. Louis City, MO
St. Louis County, MO
|Full||Tulsa, OK||Tulsa County, OK|
|Full||Washington-Arlington-Alexandria, DC-VA-MD-WV||District of Columbia|
|State profile||Metropolitan area profiles||Type of profile|
|Portland-South Portland-Biddeford, ME||Full|
|Massachusetts||Barnstable Town, MA||Full|
|New Hampshire||Manchester-Nashua, NH||Abbreviated|
|New Mexico||Albuquerque, NM||Full|
|Las Cruces, NM||Abbreviated|
|Santa Fe, NM||Abbreviated|
|Oklahoma City, OK||Full|
|Salt Lake City, UT||Full|
|St. George, UT||Abbreviated|
|Vermont||Burlington-South Burlington, VT||Full|
At the end of this publication, the profiles and spotlights appear in separate sections in alphabetical order by state and metropolitan area name. The Contents to this publication lists the profiles and spotlights in the order in which they appear.
The full profile is composed of six exhibits plus a map and demographic information on the State or metropolitan area and their constituent counties. Figure 1 shows the general layout of the full profile. All profiles observe the following conventions:
Each profile begins with a map displaying the boundaries of the metropolitan area or State and its component counties. In this publication, the terms "death investigation jurisdiction" (or, simply, "jurisdiction") and "county" are used interchangeably because ME/Cs' offices are typically organized by county. The one exception occurs in Niagara County, NY, which is divided into four districts. For reporting purposes, the four districts that make up Niagara County, NY, are treated collectively as a single jurisdiction.
Both participating and nonparticipating jurisdictions are shown in the map. Jurisdictions that provided mortality data for 2005 are colored white. Jurisdictions in the area that did not provide data are shaded light blue. Areas outside of the metropolitan area or State are shaded darker blue.
Metropolitan area definitions used in this publication are those established by OMB, based on the 2000 decennial U.S. Census and updated in 2003.7 Changes in metropolitan statistical area (MSA) boundaries since 2003 have not altered the metropolitan areas reported on in this publication. By OMB convention, the names of MSAs reflect the largest population centers (i.e., cities) in the MSA. If the relative population size of cities changes (i.e., the second largest becomes the largest), OMB changes the name of the MSA to reflect the new order of cities by size (i.e., the name of the larger city will appear first). This publication uses the name of the MSA that was current at the end of the data collection year.
Next to the map, the following items appear:
Below the map, Table A lists each of the component jurisdictions for the area. Each jurisdiction is numbered to correspond to the numbers shown on the area map. In metropolitan areas that cross State borders, jurisdictions are ordered first by State and then alphabetically by county name. Nonparticipating jurisdictions are included in the list with a shaded background to distinguish them from participating jurisdictions.
Information in Table A for each jurisdiction includes the following:
The top row of the table totals this information for just the participating jurisdictions.
Rates, because they are population adjusted, can be compared across jurisdictions, metropolitan areas, and States. This standardization does not take into account, however, the differences in applicable laws that specify which deaths are subject to ME/C review or other factors that may confound comparisons.
The subsequent tables and figures (B through F) are based on data aggregated across the participating jurisdictions in each metropolitan area or State.
Figure B is a pie chart that displays manner of death for drug-related deaths and drug-related suicide deaths. Solid-colored slices are reserved for drug-related deaths other than suicides; the patterned slice shows the suicide deaths. Reading clockwise, the manners of death are identified as follows:
Separate bar charts show the five most common types of drugs (e.g., opiates/opioids, benzodiazepines) reported to DAWN for drug-related deaths and drug-related suicide deaths across the participating jurisdictions. The number shown above each bar is the number of deaths reported for a specific drug type. The name of the drug type is printed below each bar. Each bar is partitioned to display separately the portion of deaths involving a single drug type (solid blue area in bottom portion of bar) versus multiple drug types (striped area in top of bar). A bar is not printed if there are fewer than four deaths associated with a drug type, and therefore, fewer than five bars may appear. The top 5 drug types are identified from among 17 different drug types, as listed in Table F (see below).
A single death that involved two drugs of different types (e.g., cocaine and heroin) would be counted in two bars (e.g., cocaine and heroin, respectively). As a result, summing the number of deaths reported in each bar will double-count deaths that involved multiple types of drugs. A death that involved two drugs of the same type (e.g., multiple opiates/opioids, such as methadone and heroin) will be counted once (e.g., in the bar for opiates/opioids).
Grouping drugs by drug type eliminates double counting due to the following causes: redundant drug reports (e.g., "cocaine" and its metabolite "benzoylecgonine" being reported for the same death); redundant reports from nonspecific terms (e.g., "heroin" and "opiates" being reported for the same death); and drug reports that may be indistinguishable (e.g., "heroin" and "morphine").
Figure D displays the gender and categorical age of decedents in drug-related deaths and drug-related suicide deaths, in terms of deaths per 100,000 population. Only population in participating jurisdictions is considered in the calculation of these rates. Taking population size into account enables comparisons to be made across age and gender subgroups.
Table E reports the place of death for drug-related deaths and drug-related suicide deaths. Deaths in emergency departments and other health care facilities have been combined into the single category "Health care facility."
Table F reports, by drug type or drug, the count of drug-related deaths and drug-related suicide deaths for 2004 and 2005. The first row of Table F summarizes deaths across all drug categories; the subsequent rows provide detail for 17 specific drug types or drugs of particular interest.
Data for both 2004 and 2005 are reported when the same jurisdictions participated in both years. If comparable data for 2004 are not available (e.g., due to nonparticipation) or are not comparable to those shown for 2005 (e.g., in the Chicago metropolitan area, DuPage County, IL, participated in 2005 but not in 2004), the columns are left blank.
Counts of drug-related deaths and drug-related suicide deaths include deaths that involved both single and multiple drugs. Summing these deaths across drug types or drugs could result in double counting deaths associated with multiple drug types. To help provide a better understanding of single versus multidrug involvement, counts of single-drug deaths are reported. Single-drug deaths involve the listed drug type or drug and no other, and they are a subset of the total count of deaths.
The 17 drug categories shown in this table are derived from DAWN's standard drug classification scheme and include the following:9
The next six rows in Table F pertain to illicit drugs:
The remaining rows in Table F are devoted to prescription and over-the-counter pharmaceuticals. For this table, heroin is categorized and reported on as an opiate/opioid. Low-frequency drugs have been aggregated into higher-level categories:
Not every reported substance (drug) is, by itself, the cause of death or even a contributor to the death. DAWN's broad definition of drug involvement requires only that the drug is related to the death. Therefore, even in single-drug deaths, reported drugs may not be a direct cause of death. Furthermore, incidental reporting (i.e., reporting of drugs unrelated to the death) is unavoidable due to ambiguities and insufficiencies in the ME/C's records.
The total number of deaths in some drug categories is often quite small and of limited significance. The intent in reporting small counts is primarily to indicate the relative occurrence of deaths in different drug categories.
Numbers less than four but greater than zero are suppressed.
To warrant a full profile, the participating jurisdictions of a metropolitan area in combination must have reported more than 30 drug-related deaths or drug-related suicide deaths, and the area's population coverage must exceed 50 percent. If either of these two conditions was not met, an abbreviated profile is provided for the area. In contrast to full profiles, abbreviated profiles include only a map and Table A (see above).
Reporting on the New Orleans-Metairie-Kenner, LA, metropolitan area presented a special situation in 2005 due to Hurricane Katrina and its devastating impact on the community, its population, and the normal operation of public agencies. With available data, it was only possible to prepare an abbreviated profile for the New Orleans-Metairie-Kenner, LA, metropolitan area for the 8-month period prior to Hurricane Katrina (using U.S. Census estimates as of June 1, 2005). Death rates reported in this profile were annualized to facilitate comparison with the pre- and post-Katrina spotlights for Jefferson County, LA (see below), and other metropolitan areas.
Spotlights are produced for individual jurisdictions in which 60 or more drug-related deaths were reported. The purpose is to distinguish findings for a single location from those of the metropolitan area as a whole. Spotlights may appear for jurisdictions even if the metropolitan area itself had less than 50 percent population coverage. In some instances, even if a jurisdiction has 60 or more deaths, a spotlight may not be needed. Such is the case when a metropolitan area contains only one county or had only one county participating in DAWN.
Spotlights have essentially the same format as the full metropolitan area profile. Spotlights include the map; Figures B, C, and D; and Tables E and F, as described above. Because of the small numbers, drug-related suicide deaths have been removed from all exhibits except the jurisdiction summary and Figure B.
Two spotlights were prepared for Jefferson Parish, LA, the largest jurisdiction in the New Orleans-Metairie-Kenner, LA, metropolitan area. One spotlight corresponds to the 8-month period before Hurricane Katrina, and the other corresponds to the 4-month period following Hurricane Katrina. Population estimates for the calculation of rates for the two periods are from the 2005 American Community Survey Special Product for the Gulf Coast Area.12,13 Rates reported in these spotlights were annualized to facilitate comparisons with the New Orleans metropolitan area profile and other spotlights and profiles.
Eight statewide ME/C systems participated in DAWN in 2005. A full profile is provided for each of the following States:
Any metropolitan areas that fell wholly within the eight statewide ME/C systems and reported more than 30 drug-related deaths or drug-related suicide deaths to DAWN in 2005 received a full profile (Table 5).
1 DAWN uses the terms "death investigation jurisdiction" (or, simply, "jurisdiction") and "county" interchangeably because ME/Cs' offices are typically organized by county. The one exception occurs in Niagara County, NY, which is divided into four districts. For reporting purposes, the four districts that make up Niagara County, NY, are treated collectively as a single jurisdiction.
2 To be reportable, a nonpharmaceutical substance must be consumed by inhalation, sniffing, or snorting and must have a psychoactive effect when inhaled. A death involving inhalation of a nonpharmaceutical, psychoactive substance and no other drug qualifies as a DAWN case. Carbon monoxide is excluded from the inhalants reportable to DAWN.
3 Accidental death, a manner of death assigned by an ME/C, is not synonymous with accidental ingestion, a category used by DAWN to classify drug-related deaths from 2003 to 2005.
4 Beginning with the 2005 data year, rates are expressed as the number of deaths per 100,000 population, the metric most commonly used in public health surveillance. In prior years, rates were scaled according to the number of deaths per 1,000,000 population.
5 There is overlap between the metropolitan areas and States. In total, usable reports were received from 318 jurisdictions: 101 are only in metropolitan areas, 149 are only in States, and 68 are in both.
6 Recruitment efforts to increase participation by ME/Cs are ongoing. However, there are no plans to make the mortality component of DAWN national in scope.
7 Office of Management and Budget, Revised Definitions of Metropolitan Statistical Areas, New Definitions of Micropolitan Statistical Areas and Combined Statistical Areas, and Guidance on Uses of the Statistical Definitions of These Areas, Bulletin No. 03-04, June 6, 2003. (Available at http://www.whitehouse.gov/omb/bulletins/b03-04.html.)
8 Population estimates for 2004 and 2005 were obtained from the U.S. Census Bureau County-Level Population Estimates (CPOP file), Vintage 2006, released August 2007. (Available at http://www.census.gov/popest/estimates.html.)
9 The classification of drugs used by DAWN is derived from the Multum Lexicon, © 2008, Multum Information Services, Inc. The classification has been modified to meet DAWN's unique requirements (2008). The Multum Licensing Agreement governing use of the Lexicon is provided in Appendix C. (Also available at: http://www.multum.com/.)
10 Overall, 99.5 percent of drugs in the category "Heroin (specified)" were reported to DAWN as "heroin" or its metabolite "monoacetylmorphine." The remaining 0.5 percent was reported as acetylmorphine, diacetylmorphine, acetylcodeine, monoacetylcodeine, heroin dope, or black tar heroin. Morphine and unspecified opiates are not included in the "Heroin (specified)" category. Morphine is not classified as heroin because it is not possible to differentiate morphine, the metabolite of heroin, from morphine itself. Overall, the term "morphine" or "free morphine" accounted for 98 percent of reports classified as "morphine," and the term "opiates" accounted for 99 percent of the unspecified opiates.
11 Some examples may assist readers in interpreting this classification. A death that involved heroin and methadone would be counted once in the "Opiates/opioids" category, once in the "Heroin (specified)" row, and once in the "Methadone" row. A death that involved morphine (prescription morphine or the metabolite for heroin) would be counted in the "Opiates/opioids" category and in the row for "All other opiates/opioids." A death that involved both morphine and codeine (a possible indicator for heroin) would be counted once in the "Opiates/opioids" category and once in the row for "All other opiates/opioids."
12 U.S. Census Bureau, 2005 American Community Survey Special Product for the Gulf Coast Area, accessed 8/22/2008. (Available at: http://www.census.gov/acs/www/Products/Profiles/gulf_coast/index.htm.)
13 Pre- and post-hurricane population figures were unavailable for St. Charles Parish and St. John the Baptist Parish, which also reported to DAWN for the entire year. Data on drug-related deaths during 2005 from Orleans Parish and Plaquemines Parish were too incomplete to be included in this publication.
14 The classification of drugs used in DAWN is derived from the Multum Lexicon, © 2008, Multum Information Services, Inc. The classification schema was modified to meet DAWN's unique requirements. The Multum Licensing Agreement governing use of the Lexicon is provided in Appendix C and can be found on the Internet at http://www.multum.com.
15 Office of Management and Budget, Revised Definitions of Metropolitan Statistical Areas, New Definitions of Micropolitan Statistical Areas and Combined Statistical Areas, and Guidance on Uses of the Statistical Definitions of These Areas, Bulletin No. 03-04, June 6, 2003. (Available at http://www.whitehouse.gov/omb/bulletins/b03-04.html.)