ArticlePDF Available

Abstract and Figures

National statistics on drug intoxication deaths are based solely on data derived from death certificates. This study examines the variation in the manner of death classification by state, and the specificity of drugs involved in drug intoxication deaths by state and by type of death investigation system. The National Vital Statistics System Multiple Cause of Death mortality files (2008-2010) were analyzed. Drug intoxication deaths were those with ICD-10 Underlying Cause of Death of X40-X44 (unintentional), X60-X64 (suicide), X85 (homicide), or Y10-Y14 (undetermined intent). Among drug intoxication deaths, deaths involving non-specified drug(s) were those with Multiple Cause of Death (MCOD) of T50.9, and no MCOD in the range T36-T50.8. State death investigation systems were categorized as follows: centralized state medical examiner offices, county/district medical examiners, hybrid, and decentralized county coroners. In 2008-2010, there was an average of over 37,250 drug intoxication deaths per year in the U.S. The manner was undetermined for 8% for the U.S., ranging from 1% to 85% among the states. During 2008-2010, 75% of the drug intoxication deaths had at least one specific drug reported on the death certificate. States with centralized state medical examiner systems had a higher percent (92%) of drugs specified than did those with other systems. Across the U.S., there is variation in the percent of drug intoxication deaths classified as having undetermined manner of death and with specific drugs identified on death certificates. This variation has significant implications for public health surveillance and for prevention efforts.
Content may be subject to copyright.
State variation in drug intoxication certification Warner et al.
 National statistics on drug intoxication deaths are based solely on data derived
from death certicates. This study examines the variation in the manner of death classication
by state, and the specicity of drugs involved in drug intoxication deaths by state and by type
of death investigation system. The National Vital Statistics System Multiple Cause of Death
mortality les (2008-2010) were analyzed. Drug intoxication deaths were those with ICD-10
Underlying Cause of Death of X40-X44 (unintentional), X60-X64 (suicide), X85 (homicide), or
Y10-Y14 (undetermined intent). Among drug intoxication deaths, deaths involving non-speci-
ed drug(s) were those with Multiple Cause of Death (MCOD) of T50.9, and no MCOD in the
range T36-T50.8. State death investigation systems were categorized as follows: centralized
state medical examiner ofces, county/district medical examiners, hybrid, and decentralized
county coroners.
In 2008-2010, there was an average of over 37,250 drug intoxication deaths per
year in the U.S.
The manner was undetermined for 8% for the U.S., ranging from 1% to 85%
among the states. During 2008-2010, 75% of the drug intoxication deaths had at least one
specic drug reported on the death certicate. States with centralized state medical examiner
systems had a higher percent (92%) of drugs specied than did those with other systems.
Across the U.S., there is variation in the percent of drug intoxication deaths classied as hav-
ing undetermined manner of death and with specic drugs identied on death certicates. This
variation has signicant implications for public health surveillance and for prevention efforts.
 Forensic pathology, Death certication, Surveillance, Manner of death, Drug overdose
In recent years, analyses of data from the Na-
tional Vital Statistics System have documented a
dramatic increase in drug intoxication deaths (of-
ten referred to as drug poisoning deaths or drug
overdose deaths in public health literature) in the
United States (1, 2). Since 2010, when more than
38,000 such deaths occurred, drug intoxication
has become the leading cause death due to injury.
These national statistics are based solely on data
derived from death certicates. Medical examin-
ers and coroners determine the cause and man-
ner of death on the death certicate for almost all
drug intoxication deaths, and toxicologists, both
forensic and clinical, are often involved in deter-
mining the drugs involved.
Two signicant problems related to certication
limit the usefulness of the death certicate data
for public health prevention efforts. First, the
manner of death (i.e., homicide, suicide, acci-
dent) is frequently recorded as “undetermined.”
Second, many certicates do not specify the
drugs involved in intoxication deaths. Accurate
information on both the manner of death and the
drugs involved is essential for determining the
scope of the public health problem. In addition,
prevention measures will differ depending on
whether the drug intoxication deaths are unin-
tentional versus suicidal and the types of drugs
involved. This paper highlights state differences
in the completeness of death certicate informa-
tion regarding the manner of death and the drugs
involved in drug intoxication deaths. The rela-
tionship between the type of death investigation
Author Alliations: Centers for Disease
Control and Prevention - National
Center for Injury Prevention and
Control, El Paso, TX (LP), Ofce of the
Medical Investigator - Department of
Pathology, University of New Mexico,
Albuquerque, NM (KN), University of
Alabama - Pathology, Birmingham, AL
(GD), New York University School of
Medicine - Emergency Medicine, New
York, NY (LN).
Contact Dr. Warner at:
Acad Forensic Pathol
2013 3 (2): 231-237
This work was authored as part of the Contributor’s ofcial
duties as an Employee of the United States Government
and is therefore a work of the United States Government.
In accordance with 17 U.S.C. 105, no copyright protection
is available for such works under U.S. Law.
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
Original article
Volume 3 Issue 2
system and the reporting of the types of drugs
involved is also described.
The National Vital Statistics System (NVSS)
Multiple Cause of Death (MCOD) mortality les
for data years 2008-2010 were analyzed. The
NVSS is compiled from information recorded
on death certicates, which is processed and
reported at the state level, and then provided to
the National Center for Health Statistics (NCHS)
through the Vital Statistics Cooperative Program
for compilation and national reporting.
Deaths with an underlying cause of death (UCOD)
of drug intoxication were identied using the fol-
lowing ICD-10 codes: X40-X44 (unintentional),
X60-X64 (suicide), X85 (homicide), or Y10-Y14
(undetermined intent). Among deaths with an un-
derlying cause of drug intoxication, those which
specied one or more drugs were identied with
ICD-10 MCOD in the range T36-T50.8. Deaths
involving non-specied drug(s) were identi-
ed as those with MCOD of T50.9 (other and
unspecied drugs, medicaments, and biological
substances) and no mention of any other drugs
(i.e., without MCOD in the range T36-T50.8).
Deaths among U.S. residents were classied by
state based on the state in which the death oc-
curred. State death investigation systems were
categorized depending on whether the state had
a centralized medical examiner, county/district
medical examiners, a mix of medical examiners
and coroners, or county coroners (3).
In 2008-2010 in the U.S., there was an aver-
age of over 37,250 drug intoxication deaths per
year (36,450 in 2008; 37,004 in 2009; 38,329
in 2010). During this time period, the manner
of death was undetermined for 8% of the drug
intoxication deaths and determined for 92% of
the deaths [i.e. accident (78%), suicide (14%),
homicide (<1%)]. The percentage of deaths with
an undetermined manner of death ranged from
1% to 85% (Figure 1). In two states the man-
ner of death was undetermined for more than one
third of the deaths (85% in Maryland and 40% in
Utah), while in 11 states (CA, FL, ME, NJ, NV,
OH, RI, SC, TX, VA, WY), the manner of death
was undetermined for 4% or fewer of the deaths.
When the states with the two highest percentages
were excluded from the analysis, the manner of
death was undetermined for an average of 6% of
the drug intoxication deaths.
During 2008-2010, 75% of the drug intoxication
deaths had at least one specic drug reported
on the death certicate. For the remaining 25%,
the type of drug(s) involved was not specied.
For many of the deaths with no drug specied,
the text on the death certicate indicating that
the cause of death involved drugs was limited
(e.g., “multiple drug intoxication” or “drug over-
dose”). The percentage of deaths with specic
drugs mentioned varied by type of death investi-
gation system and by state (Table 1). States with
centralized state medical examiner systems had a
higher percent (92%) of drugs specied than did
those with decentralized county coroner systems
(62%). The remaining systems had between 71-
73% specied. In ten states, the drugs involved
were specied for 95% or more of the deaths. In
11 states, the drugs involved were specied for
only 33-65 % of the deaths.
The percent of drug intoxication deaths classi-
ed as having undetermined manner of death and
with specic drugs identied on death certi-
cates varies widely across the United States. This
variation has signicant implications for public
health prevention efforts. For example, to prevent
drug intoxication deaths, it is important to know
the manner of death so that targeted prevention
strategies can be developed. Interventions de-
signed to change behaviors will be different for
programs focusing on suicidal compared with
unintentional (e.g., accidental, overuse, abuse-
related) intoxication. In addition, it is important
to know the type of drug or drugs involved. For
75% of the deaths nationally, some information
about the drug involved is provided on the death
certicate. However, this leaves one quarter of
the deaths in the U.S. with no further information
other than that the death involved a drug.
For public health surveillance, the variations by
state in the availability of manner of death infor-
mation and in the specicity of drugs involved
make some comparisons among states mislead-
ing. For instance, when comparing state-specic
death rates for unintentional or suicidal drug in-
toxication deaths, the magnitude of the problem
will be underestimated in states with high per-
centages of death in which the manner is unde-
termined. When comparing drug-specic death
rates (e.g., heroin, methadone), rates will be un-
derestimated for states with high percentages of
deaths lacking specic drug information (4).
Drug intoxication deaths are among the hardest to
determine manner of death. In the late 1960s, in
recognition that for a small proportion of deaths
there may be ambiguity in the manner of death
even after a complete investigation, the revised
U.S. Standard Death Certicate included “unde-
termined” as an option (5). The U.S. mortality
data show that among all external causes of death
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
State variation in drug intoxication certification Warner et al.
South Carolina
Rhode Island
New Jersey
North Carolina
North Dakota
New Mexico
District of Columbia
New Hampshire
West Virginia
New York
South Dakota
All states
0 20 40 60 80 100
Undetermined Suicide Unintentional
Figure 1: Percent of drug intoxication deaths by manner and state, 2008-2010.
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
Original article
Volume 3 Issue 2
(i.e., non-natural deaths), the overall proportion
of deaths certied with an undetermined manner
is about 2.8%. Of the leading external causes in
2010, poisonings (including drug intoxication
and toxic substances) had the highest proportion
with an undetermined manner (7.4%), followed
by drowning (4.0%) and re/hot object (4.0%)
Figure 1 documents the wide range by state in
the proportions of death with an undetermined
manner of death for drug intoxication deaths.
The variation may be due to differences in death
investigation methods, in criteria used to deter-
mine a manner of death, or a combination of the
two. Although the type of death investigation
system may play a role, when manner of death
classication was categorized by death investi-
gation system no patterns were obvious. Over
the years, many publications have provided in-
sight into how best to determine manner of death
(7-11). Previous studies focused on manner of
death certication practices for drug intoxication
deaths have found that certication practice var-
ies among states and suggest that the ambiguity
is typically because of the inability to distinguish
between accident and suicide (12-14). Because
depression, chronic pain and medication abuse
are all frequently found in the history of drug in-
toxication deaths, determining a person’s intent
for using the drug can be a challenge for a medi-
cal examiner or coroner (15). Medical examiner
or coroners may also be challenged with limited
information about the deaths. In particular, for
some deaths involving the use of illicit or legal
substances obtained illicitly, witnesses and fam-
ily may share fewer details. In addition, bodies
may be found in locations that don’t appear to be
where the death occurred (e.g., alleys); not only
does this make death scene investigation impos-
sible, it also indicates that someone witnessed or
was aware of the death, but clearly did not want
the circumstances known.
Table 1 provides the percentage of drug intoxi-
cation deaths with drugs specied by state and
death investigation system, and shows specica-
tion varied widely among the states. The amount
 Percent of Drug Intoxication Deaths with Drugs Specied by State and Death Investigation Sys-
tem, 2008-2010
Death Investigation System State Percentage of Drug Intoxication
Deaths with Drugs Specied
All states 75.1
Centralized state medical examiner 92.1
West Virginia 99.4
New Hampshire 99.1
Vermont 98.9
Maryland 98.6
Rhode Island 97.3
Oklahoma 97.2
Massachusetts 97.0
Alaska 95.5
Utah 94.2
North Carolina 92.9
Virginia 92.7
Oregon 91.2
Maine 89.7
Delaware 79.3
Connecticut 76.8
New Mexico 68.7
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
State variation in drug intoxication certification Warner et al.
 Continued
Death Investigation System State Percentage of Drug Intoxication
Deaths with Drugs Specied
Decentralized county or district medical examiner (physician) 71.1
Iowa 96.1
Arizona 80.2
Tennessee 77.8
Florida 68.8
Michigan 65.8
New Jersey 59.3
Hybrid system: county coroner and medical
examiners (state and/or county) 73.2
New York 94.0
Washington 92.6
Illinois 86.5
Wisconsin 85.9
Hawaii 83.1
Minnesota 82.4
Missouri 79.3
Texas 74.7
California 73.1
Ohio 71.4
Georgia 71.4
Montana 69.9
Kentucky 64.8
Pennsylvania 45.0
Alabama 45.0
Mississippi 43.4
Decentralized county coroner 62.4
Nevada 97.7
South Dakota 88.8
North Dakota 87.9
Arkansas 76.7
Colorado 70.4
Nebraska 69.4
Wyoming 64.8
South Carolina 59.8
Idaho 59.6
Kansas 58.8
Indiana 45.8
Louisiana 34.8
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
Original article
Volume 3 Issue 2
of specication on the death certicate results
from several possible factors that may vary by ju-
risdiction, including the circumstances in which
the tests are performed, the substances tested for,
and the practice of recording on the death cer-
ticate. When categorized by death investigation
systems, in states with a centralized medical ex-
aminer system, more than 90% of drug intoxica-
tion deaths had a drug specied, while in states
with a decentralized county coroner system, only
two-thirds of the drug intoxication deaths had a
drug specied. States with a mix of medical ex-
aminer and coroner systems had proportions of
drugs specied between these two. Differences
between centralized statewide and county-based
systems, and medical examiner and coroner sys-
tems, might explain the disparity in specication
of drugs responsible for death (3, 16-18).
County based systems, particularly those serv-
ing smaller populations, might perceive or ex-
perience more barriers to using toxicologic ser-
vices to determine the types of drugs involved.
The 2004 Bureau of Justice Statistics’ Census of
Medical Examiner and Coroners’ Ofces found
that ofces serving jurisdictions of 250,000 or
more conducted toxicology analyses in a greater
percentage of accepted cases (57%), compared to
ofces serving smaller jurisdictions (34%) (19).
To address the issue of toxicological testing, the
Kentucky legislature recently revised its Coro-
ner’s Statute to require coroners to test for the
presence of controlled substances in postmortem
examinations unless another cause is clearly es-
tablished (20).
In addition to differences in toxicological testing
for the drugs involved, organizational structure
of death investigation within a state may also
play a role in the approach taken to completing
the death certicate. For instance, with decentral-
ized systems there may be more certier-to-cer-
tier or ofce-to-ofce variability in comparison
to centralized systems, where the Chief Medical
Examiner is able to provide oversight or even
mandate a practice. Within county-based coroner
or medical examiner systems, there may be fewer
mechanisms to rely on for standardization. How-
ever, national and state associations do provide a
forum for these issues.
Differences in background and training between
many coroners and medical examiners may also
be a factor in some cases. Coroners are typically
non-physicians, and are sometimes elected of-
cials, whereas medical examiners are usually
physicians who have specialized in pathology
and subspecialized in forensic pathology (16-
18). Coroners typically don’t have medical train-
ing and have less familiarity with toxicological
terminology, which could explain some differ-
ences in the level of detail reported on the death
certicate concerning the drugs causing death.
Although in many cases, coroners will work with
or employ forensic pathologists to determine the
cause of death, the coroner may be lling out
the cause of death section of the death certicate
based on reports. In addition, coroners may have
more limited contact with medical and forensic
toxicologists than medical examiners, especially
in highly decentralized systems. Outreach edu-
cational efforts to the medical examiner and the
coroner communities might facilitate more com-
prehensive surveillance for these deaths.
The number of drug intoxication deaths has in-
creased six-fold since 1980 (2), and detailed
documentation of the specic causes is impor-
tant for understanding this problem. This short
report highlights the surveillance challenges and
is intended to bring awareness to issue. A recent
position paper from the National Association of
Medical Examiners and the American College of
Medical Toxicologists makes recommendations
for conducting investigations and for improving
the specication of manner and type of drug for
opioid-related intoxication deaths (21). The posi-
tion paper discusses the importance of accurate
death certication, and includes guidance for cer-
tifying deaths, regardless of the system of death
investigation. Improving the quality of death
investigation and certication will maximize
the utility of our existing national registration
of deaths for public health surveillance and re-
search, and contribute to the design of programs
to prevent drug intoxication deaths.
The ndings and conclusions in this report are
those of the authors and do not necessarily repre-
sent the ofcial position of the Centers for Disease
Control and Prevention (CDC).
The authors, reviewers, editors, and publication
staff do not report any relevant conicts of interest.
1) Paulozzi LJ, Jones CM, Mack KA, Rudd RA. Vital
signs: overdoses of prescription opioid pain relievers—
United States, 1999-2008. MMWR Morb Mortal Wkly
Rep. 2011 Nov 4; 60(43):1487-92. Available from:
2) Warner M, Chen LH, Makuc DM, et al. Drug poisoning
deaths in the United States, 1980-2008. NCHS Data
Brief. 2011 Dec; (81):1-8. Available from:
3) Scientic Working Group for Medical Death Invest-
igation Infrastructure Committee report 1, death
investigation systems and NAME, ABMDI, IACME
representation in states [Internet]. Scientic Working
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
State variation in drug intoxication certification Warner et al.
Group for Medical Death Investigation. 2011 Mar 21
[cited 2013 May 6]. 3 p. Available from:nal.pdf.
4) Paulozzi LJ, Ibrahimova A, Rudd RA, et al. A compari-
son of Florida medical examiners’ reports and death
certicates for specic drug related deaths. Acad
Forensic Pathol. 2012 Jun; 2(2):190-7.
5) Tolson GC, Barnes JM, Gay GA, Kowaleski JL. The
1989 revision of the U.S. standard certicates and
reports. Vital Health Stat 4. 1991 Jun;(28):1-34.
6) Murphy SL, Xu JQ, Kochanek KD, Deaths: nal data
for 2010. Natl Vital Stat Rep. 2013 May 8; 61(4):1-168.
7) Davis GG. Mind your manners. Part I: History of death
certication and manner of death classication. Am J
Forensic Med Pathol. 1997 Sep; 18(3):219-23.
8) Goodin J, Hanzlick R. Mind your manners. Part II:
General results from the National Association of
Medical Examiners manner of death questionnaire,
1995. Am J Forensic Med Pathol. 1997 Sep; 18(3):224-7.
9) Hanzlick R, Goodin J. Mind your manners. Part III:
Individual scenario results and discussion of the
National Association of Medical Examiners manner of
death questionnaire, 1995. Am J Forensic Med Pathol.
1997 Sep; 18(3):228-45.
10) Hanzlick R. Cause of death and the death certicate:
important information for physicians, coroners, medical
examiners, and the public. Northeld (IL): College of
American Pathologists; 2006. 235 p.
11) National Center for Health Statistics. Medical examin-
ers’ and coroners’ handbook on death registration and
fetal death reporting, 2003 revision. Hyattsville (MD);
2003. 130 p. Available from:
12) Donaldson AE, Larsen GY, Fullerton-Gleason L,
Olson LM. Classifying undetermined poisoning deaths.
Inj Prev. 2006 Oct; 12(5):338-43.
13) Breiding MJ, Wiersema B. Variability of undetermined
manner of death classication in the US. Inj Prev. 2006
Dec; 12 Suppl 2:ii49-ii54.
14) Sorenson SB, Shen HK, Kraus JF. Undetermined
manner of death. A comparison with unintentional
injury, suicide, and homicide death. Eval Rev. 1997
15) Toblin RL, Paulozzi LJ, Logan JE, et al. Mental illness
and psychotropic drug use among prescription drug
overdose deaths: a medical examiner chart review.
J Clin Psychiatry. 2010 Apr; 71(4):491-6.
16) Committee for the Workshop on the Medicolegal Death
Investigation System. Medicolegal death investigation
system: workshop summary. Washington: National
Academies Press; 2003. 86 p.
17) Hanzlick R. Medical examiners, coroners, and public
health: a review and update. Arch Pathol Lab Med.
2006 Sep; 130(9):1274-82.
18) Committee on Identifying the Needs of the Forensic
Sciences Community, National Research Council.
Strengthening forensic science in the United States: a
path forward. Washington: The National Academies
Press; 2009. 352 p.
19) Hickman MJ, Hughes KA, Ropero-Miller JD, Strom
KJ. Bureau of Justice Statistics special report: Medical
examiners and coroners’ ofces, 2004. Washington:
U.S. Department of Justice; 2007. 8 p.
20) Kentucky Revised Statutes Chapter 072.00 [Internet].
2012. Available from:
21) Davis GG, National Association of Medical Examiners
and American College of Medical Toxicology Expert
Panel on Evaluating and Reporting Opioid Deaths.
National Association of Medical Examiners position
paper: Recommendations for the investigation,
diagnosis, and certication of deaths related to opioid
drugs. Acad Forensic Pathol. 2013 Mar; 3(1):77-83.
Downloaded from by Kurt Nolte on 2013-06-03.
This article is intended for personal use, but may be distributed by subscribers solely for scholarly purposes.
... Consistent with these challenges, data suggest that suicide deaths are undercounted and there is significant heterogeneity in suicide rates regionally and among different groups, which may reflect variation in classification methods more so than true rate variation. For example, Maryland, Massachusetts, and Rhode Island recorded higher rates of undetermined poisoning-related deaths and lower rates of accidental poisoning-related death than other states (Warner et al., 2013). Such variation affects policy decisions, since public health policies rely mostly on national rather than state-specific trends. ...
Objectives: To improve the accuracy of classification of deaths of undetermined intent and to examine racial differences in misclassification. Methods: We used natural language processing and statistical text analysis on restricted-access case narratives of suicides, homicides, and undetermined deaths in 37 states collected from the National Violent Death Reporting System (NVDRS) (2017). We fit separate race-specific classification models to predict suicide among undetermined cases using data from known homicide cases (true negatives) and known suicide cases (true positives). Results: A classifier trained on an all-race dataset predicts less than half of these cases as suicide. Importantly, our analysis yields an estimated suicide rate for the Black population comparable with the typical detection rate for the White population, indicating that misclassification excess is endemic for Black suicide. This problem may be mitigated by using race-specific data. Our findings, based on the statistical text analysis, also reveal systematic differences in the phrases identified as most predictive of suicide. Conclusions: This study highlights the need to understand the reasons underlying suicide rate differences and for further testing of strategies to reduce misclassification, particularly among people of color.
... This study has limitations, including small sample size and the possible lack of accuracy in death certificates for determining underlying cause of death. Nonetheless, WV uses a centralized medical examiner system that reports the highest percentage of drug intoxication deaths with drugs specified across the states.11 The study excluded persons who had dual eligibility with Medicare, which accounts for about 20% of overall Medicaid beneficiaries. ...
Full-text available
Background and objectives: Compare proportion of all-cause and cause-specific mortality among West Virginia Medicaid enrollees who were discharged from infective endocarditis (IE) hospitalization with and without opioid use disorder (OUD) diagnosis. Methods: The proportions of cause-specific deaths among those who were discharged from IE-related hospitalizations were compared by OUD diagnosis. Results: The top three underlying causes of death discharged from IE hospitalization were accidental drug poisoning, mental and behavioral disorders due to polysubstance use, and cardiovascular diseases. Of the total deaths occurring among patients discharged after IE-related hospitalization, the proportion has increased seven times from 2016 to 2019 among the OUD deaths while it doubled among the non-OUD deaths. Discussion and conclusions: Of the total deaths occurring among patients discharged after IE-related hospitalization, the increase is higher in those with OUD diagnosis. OUD is becoming a significantly negative impactor on the survival outcome among IE patients. It is of growing importance to deliver medication for OUD treatment and harm reduction efforts to IE patients in a timely manner, especially as the COVID-19 pandemic persists.
... ICD-10 codes are used at the national and local levels to calculate OD mortality statistics, to monitor trends in the drugs involved in ODs, and for epidemiological analyses. Specifically, ICD-10 codes assigned as an underlying cause-of-death are used to identify drug OD deaths and those assigned as supplemental cause-of-death codes (up to 20) are used to identify the drug(s) involved in the OD (1,3,6). This process is central to the surveillance of drug OD mortality and is the primary way in which OD mortality information is reported to communities. ...
Surveillance of drug overdose deaths relies on death certificates for identification of the substances that caused death. Drugs and drug classes can be identified through the International Classification of Diseases, 10th Revision (ICD-10) codes present on death certificates. However, ICD-10 codes do not always provide high levels of specificity in drug identification. To achieve more fine-grained identification of substances on a death certificate, the free-text cause of death section, completed by the medical certifier, must be analyzed. Current methods for analyzing free-text death certificates rely solely on look-up tables for identifying specific substances, which must be frequently updated and maintained. To improve identification of drugs on death certificates, a deep learning named-entity recognition model was developed, which achieved an F1-score of 99.13%. This model can identify new drug misspellings and novel substances that are not present on current surveillance look-up tables, enhancing the surveillance of drug overdose deaths.
... Types of medicolegal death investigation system were distinguished as centralized (state) medical examiner, decentralized county or district medical examiner, medical examiner and coroner hybrid, or decentralized county or district coroner. 10,11 We compared states by selected system type with all others. Comprising multiple categories, both system type and major geographic region were represented as dummy variables. ...
Full-text available
Importance: Self-injury mortality (SIM) combines suicides and the preponderance of drug misuse-related overdose fatalities. Identifying social and environmental factors associated with SIM and suicide may inform etiologic understanding and intervention design. Objective: To identify factors associated with interstate SIM and suicide rate variation and to assess potential for differential suicide misclassification. Design, setting, and participants: This cross-sectional study used a partial panel time series with underlying cause-of-death data from 50 US states and the District of Columbia for 1999-2000, 2007-2008, 2013-2014 and 2018-2019. Applying data from the Centers for Disease Control and Prevention, SIM includes all suicides and the preponderance of unintentional and undetermined drug intoxication deaths, reflecting self-harm behaviors. Data were analyzed from February to June 2021. Exposures: Exposures included inequity, isolation, demographic characteristics, injury mechanism, health care access, and medicolegal death investigation system type. Main outcomes and measures: The main outcome, SIM, was assessed using unstandardized regression coefficients of interstate variation associations, identified by the least absolute shrinkage and selection operator; ratios of crude SIM to suicide rates per 100 000 population were assessed for potential differential suicide misclassification. Results: A total of 101 325 SIMs were identified, including 74 506 (73.5%) among males and 26 819 (26.5%) among females. SIM to suicide rate ratios trended upwards, with an accelerating increase in overdose fatalities classified as unintentional or undetermined (SIM to suicide rate ratio, 1999-2000: 1.39; 95% CI, 1.38-1.41; 2018-2019: 2.12; 95% CI, 2.11-2.14). Eight states recorded a SIM to suicide rate ratio less than 1.50 in 2018-2019 vs 39 states in 1999-2000. Northeastern states concentrated in the highest category (range, 2.10-6.00); only the West remained unrepresented. Least absolute shrinkage and selection operator identified 8 factors associated with the SIM rate in 2018-2019: centralized medical examiner system (β = 4.362), labor underutilization rate (β = 0.728), manufacturing employment (β = -0.056), homelessness rate (β = -0.125), percentage nonreligious (β = 0.041), non-Hispanic White race and ethnicity (β = 0.087), prescribed opioids for 30 days or more (β = 0.117), and percentage without health insurance (β = -0.013) and 5 factors associated with the suicide rate: percentage male (β = 1.046), military veteran (β = 0.747), rural (β = 0.031), firearm ownership (β = 0.030), and pain reliever misuse (β = 1.131). Conclusions and relevance: These findings suggest that SIM rates were associated with modifiable, upstream factors. Although embedded in SIM, suicide unexpectedly deviated in proposed social and environmental determinants. Heterogeneity in medicolegal death investigation processes and data assurance needs further characterization, with the goal of providing the highest-quality reports for developing and tracking public health policies and practices.
... 1. Learn more about existing studies, methods and tools  A recent study by Warner, et al. (2013) provides a baseline state-to-state comparison of drug overdose deaths by manner of death and by percentage of drug overdose deaths lacking information on the specific drugs involved. This baseline comparison can be very informative for states as they review the quality of their drug overdose mortality data. ...
Technical Report
Full-text available
State and local epidemiologists can play an important role in improving the quality of drug overdose mortality data in their jurisdiction and consequently improve the quality of drug overdose mortality surveillance at the national level. This document, prepared by the CSTE Overdose Subcommittee (, outlines various strategies that epidemiologists and staff at state and local health departments can use to evaluate the quality of their drug mortality data and to collaborate with vital registrars and medical examiners and coroners to improve the drug specific information reported on death certificates (for background, see Hanzlick, 2006). Recommendations and lessons learned provide concrete examples that may be applicable in other jurisdictions.
... Dr. Robert Anderson, chief of the Mortality Statistics Branch of the NVSS, has been quoted as saying that before the current coronavirus pandemic, one in every three death certificates was 'wrong' and that things were about to get worse [14]. This percentage of error is in keeping with the findings of previous studies of the accuracy of death certificate source data [33,34]. ...
Full-text available
Mortality data in most countries are reported using the International Classification of Diseases (ICD), managed by the WHO. In this paper, we show how the ICD is ill-suited for classifying drug-involved deaths, many of which involve polysubstance abuse and/or illicitly manufactured fentanyl (IMF). Opioids identified in death certificates are categorized according to six ICD T-codes: opium (T40.0), heroin (T40.1), methadone (T40.3), other synthetic narcotics (T40.4), and other and unspecified narcotics (T40.6). Except for opium, heroin, and methadone, all other opioids except those that are unspecified are aggregated in two T-codes (T40.2 and T40.4), depending upon whether they are natural/semisynthetic or synthetic opioids other than methadone. The result is a system that obscures the actual cause of most drug overdose deaths and, instead, just tallies the number of times each drug is mentioned in an overdose situation. We examined the CDC’s methodology for coding other controlled substances according to the ICD and found that, besides fentanyl, the ICD does not distinguish between other licit and illicitly manufactured controlled substances. Moreover, we discovered that the CDC codes all methadone-related deaths as resulting from the prescribed form of the drug. These and other anomalies in the CDC’s mortality reporting are discussed in this report. We conclude that the CDC was at fault for failing to correct the miscoding of IMF. Finally, we briefly discuss some of the public policy consequences of this error, the misguided focus by public health and safety officials on pharmaceutical opioids, their prescribers and users, and the pressing necessity for the CDC to reassess how it measures and reports drug-involved mortality.
Background: There is a striking geographic variation in drug overdose deaths without a specific drug recorded, many of which likely involve opioids. Knowledge of the reasons underlying this variation is limited. Objectives: We sought to understand the role of medicolegal death investigation (MDI) systems in unclassified drug overdose mortality. Methods: This is an observational study of 2014 and 2018 fatal drug overdoses and U.S. county-level MDI system type (coroner vs medical examiner). Mortality data are from the CDC's National Center for Health Statistics. We estimated multivariable logistic regressions to quantify associations between MDI system type and several outcome variables: whether the drug overdose was unclassified and whether involvement of any opioid, synthetic opioid, methadone, and heroin was recorded (vs unclassified), for 2014 (N = 46,996) and 2018 (N = 67,359). Results: In 2018, drug overdose deaths occurring in coroner counties were almost four times more likely to be unclassified (OR 3.87, 95% CI 2.32, 6.46) compared to medical examiner counties. These odds ratios are twice as large as in 2014 (difference statistically significant, P < .001), indicating that medical examiner counties are improving identification of opioids in drug overdoses faster than coroner counties. Conclusions: Accurate reporting of drug overdose deaths depends on MDI systems. When developing state policies and local interventions aimed to decrease opioid overdose mortality, decision-makers should understand the role their MDI system is playing in underestimating the extent of the opioid overdose crisis. Improvements to state and county MDI systems are desirable if accurate reporting and appropriate policy response are to be achieved.
Full-text available
Importance Despite high rates of drug overdose death among people experiencing homelessness, patterns in drug overdose mortality, including the types of drugs implicated in overdose deaths, remain understudied in this population. Objective To describe the patterns in drug overdose mortality among a large cohort of people experiencing homelessness in Boston vs the general adult population of Massachusetts and to evaluate the types of drugs implicated in overdose deaths over a continuous 16-year period of observation. Design, Setting, and Participants This cohort study analyzed adults aged 18 years or older who received care at Boston Health Care for the Homeless Program (BHCHP) between January 1, 2003, and December 31, 2017. Individuals were followed up from the date of their initial BHCHP encounter during the study period until the date of death or December 31, 2018. Data were analyzed from December 1, 2020, to June 6, 2021. Main Outcomes and Measures Drug overdose deaths and the types of drugs involved in each overdose death were ascertained by linking the BHCHP cohort to the Massachusetts Department of Public Health death records. Results In this cohort of 60 092 adults experiencing homelessness (mean [SD] age at entry, 40.4 [13.1] years; 38 084 men [63.4%]), 7130 individuals died by the end of the study period. A total of 1727 individuals (24.2%) died of a drug overdose. Of the drug overdose decedents, 456 were female (26.4%), 194 were Black (11.2%), 202 were Latinx (11.7%), and 1185 were White (68.6%) individuals, and the mean (SD) age at death was 43.7 (10.8) years. The age- and sex-standardized drug overdose mortality rate in the BHCHP cohort was 278.9 (95% CI, 266.1-292.3) deaths per 100 000 person-years, which was 12 times higher than the Massachusetts adult population. Opioids were involved in 91.0% of all drug overdose deaths. Between 2013 and 2018, the synthetic opioid mortality rate increased from 21.6 to 327.0 deaths per 100 000 person-years. Between 2004 and 2018, the opioid-only overdose mortality rate decreased from 117.2 to 102.4 deaths per 100 000 person-years, whereas the opioid-involved polysubstance mortality rate increased from 44.0 to 237.8 deaths per 100 000 person-years. Among opioid-involved polysubstance overdose deaths, cocaine-plus-opioid was the most common substance combination implicated throughout the study period, with Black individuals having the highest proportion of cocaine-plus-opioid involvement in death (0.72 vs 0.62 in Latinx and 0.53 in White individuals; P < .001). Conclusions and Relevance In this cohort study of people experiencing homelessness, drug overdose accounted for 1 in 4 deaths, with synthetic opioid and polysubstance involvement becoming predominant contributors to mortality in recent years. These findings emphasize the importance of increasing access to evidence-based opioid overdose prevention strategies and opioid use disorder treatment among people experiencing homelessness, while highlighting the need to address both intentional and unintentional polysubstance use in this population.
Full-text available
Background: Overdose deaths involving opioid pain relievers (OPR), also known as opioid analgesics, have increased and now exceed deaths involving heroin and cocaine combined. This report describes the use and abuse of OPR by state. Methods: CDC analyzed rates of fatal OPR overdoses, nonmedical use, sales, and treatment admissions. Results: In 2008, drug overdoses in the United States caused 36,450 deaths. OPR were involved in 14,800 deaths (73.8%) of the 20,044 prescription drug overdose deaths. Death rates varied fivefold by state. States with lower death rates had lower rates of nonmedical use of OPR and OPR sales. During 1999--2008, overdose death rates, sales, and substance abuse treatment admissions related to OPR all increased substantially. Conclusions: The epidemic of overdoses of OPR has continued to worsen. Wide variation among states in the nonmedical use of OPR and overdose rates cannot be explained by underlying demographic differences in state populations but is related to wide variations in OPR prescribing. Implications for Public Health Practice: Health-care providers should only use OPRs in carefully screened and monitored patients when non-OPR treatments are insufficient to manage pain. Insurers and prescription drug monitoring programs can identify and take action to reduce both inappropriate and illegal prescribing. Third-party payers can limit reimbursement in ways that reduce inappropriate prescribing, discourage efforts to obtain OPR from multiple health-care providers, and improve clinical care. Changes in state laws that focus on the prescribing practices of health-care providers might reduce prescription drug abuse and overdoses while still allowing safe and effective pain treatment.
Context.—Traditionally, the emphasis of work done by medical examiners, coroners, and the death investigation community has been viewed as serving the criminal justice system. During the last several decades, however, an important role for these 3 groups has emerged within public health. Objective.—To provide important background information on death investigation systems, the evolution and framework of public health entities that rely on information gathered by medical examiners and coroners, and the role of medical examiners and coroners in epidemiologic research, surveillance, and existing public health programs and activities. Data Sources.—Previous articles on epidemiologic aspects of forensic pathology and the role of medical examiners and coroners in epidemiologic research and surveillance; a review of the Web sites of public health and safety agencies, organizations, and programs that rely on medical examiner and coroner data collected during medicolegal investigations; and a review of recent public health reports and other publications of relevance to medical examiner and coroner activities. Conclusions.—The role of medical examiners and coroners has evolved from a criminal justice service focus to a broader involvement that now significantly benefits the public safety, medical, and public health communities. It is foreseeable that the public health role of medical examiners and coroners may continue to grow and that, perhaps in the not-too-distant future, public health impact will surpass criminal justice as the major focus of medicolegal death investigation in the United States.
The American College of Medical Toxicology and the National Association of Medical Examiners convened an expert panel to generate evidence-based recommendations for the practice of death investigation and autopsy, toxicological analysis, interpretation of toxicology findings, and death certification to improve the precision of death certificate data available for public health surveillance. The panel finds the following: 1. A complete autopsy is necessary for optimal interpretation of toxicology results, which must also be considered in the context of the circumstances surrounding death, medical history, and scene findings. 2. A complete scene investigation extends to reconciliation of prescription information and pill counts. 3. Blood, urine, and vitreous humor, when available, should be retained in all cases. Blood from the femoral vein is preferable to blood from other sites. 4. A toxicological panel should be comprehensive and include opioid and benzodiazepine analytes, as well as other potent depressant, stimulant, and anti-depressant medications. 5. Interpretation of postmortem opioid concentrations requires correlation with medical history, scene investigation, and autopsy findings. 6. If death is attributed to any drug or combination of drugs (whether as cause or contributing factor), the certifier should list all the responsible substances by generic name in the autopsy report and on the death certificate. 7. The best classification for manner of death in deaths due to the misuse or abuse of opioids without any apparent intent of self-harm is “accident.” Reserve “undetermined” as the manner for the rare cases in which evidence exists to support more than one possible determination.
Death certificates are a source of information on deaths caused by specific drugs. However, the completeness of such information has been questioned. This study compared counts in the Florida Medical Examiner Commission's drug related death reporting system (ME) with those from death certificates (DC) for overdose deaths involving selected drugs during 2005-2008 to assess the completeness of death certificate information. DC data indicated 2,401 deaths with benzodiazepines as a cause, 2,182 for cocaine, 2,055 for methadone, and 416 for heroin. The ratios of ME deaths to DC deaths were 1.19 (95% CI 1.13-1.26) for benzodiazepines, 1.38 (95% CI 1.31-1.46) for cocaine, 1.37 (95% CI 1.29-1.44) for methadone, and 0.96 (95% CI 0.83-1.09) for heroin. For each drug, ratios were similar for unintentional (accidental), suicide, and undetermined intent deaths. Ratios varied across the 24 ME districts, but few districts reported significantly fewer deaths in the ME system than in the DC system. Results suggest that surveillance based on death certificate data underestimates specific drug caused deaths as counted by medical examiners in a state with multiple, independent medical examiners. Death certificate data might therefore underestimate drug overdose mortality nationwide.
More than 700 physician medical examiner/coroners (ME/Cs) were surveyed to assess differences in manner of death classifications for typical but often controversial death scenarios: 198 physicians participated by choosing the manner of death (homicide, suicide, accident, natural, undetermined) for 23 such scenarios. Sixteen questions related to death certificate training, work location, and manner of death issues were also asked. The classification of manner of death by ME/Cs was highly variable. For some challenging death scenarios, majority agreement was lacking. Agreement was ≥80% for only 11 of the 23 scenarios and was 100% for only 1. Manner of death classification method was not influenced by forensic pathology board certification status, by whether or not the physician actually completed death certificates, or by previous threats of lawsuits over manner of death classification. However, there were some differences by state. No textbook or individual was widely recognized as authoritative on manner of death issues. Few ME/Cs had formal death certification training in medical school or residency. The data lend credence to the practice of the National Center for Health Statistics (NCHS) of classifying manner of death for statistical purposes by using coding and classification rules and selection criteria rather than solely on the basis of the classification of manner chosen by ME/Cs. The data also indicate that caution is in order when one compares manner of death statistics of one ME/C with those of another. Published guidelines and more uniform training are needed so that ME/Cs may become more consistent in their manner of death classifications. Further information is presented in Part I (history of manner of death classification) and in Part III (individual death scenarios and their analysis) companion articles in this issue of the Journal.
Every death is unique, but deaths also share similar features that allow them to be grouped into categories. Since its initial description over 800 years ago, the position of coroner has been charged with the determination of manner of death. This determination has been made by examination into the circumstances surrounding death and of wounds on the surface of the body. Over the years, physicians have gained sufficient understanding of the body such that the autopsy became an important part of a death investigation. With additional time, laws were changed so that individuals charged with the determination of manner of death were required to have appropriate training. Death certification is the means by which deaths are grouped together according to similar characteristics. The practice of death certification has led to effective public health programs and the advancement of medical science. The addition of manner of death to the death certificate is an American contribution to vital statistics registration. The purpose of the autopsy report differs from that of the death certificate; the report fully addresses the unique aspects of a death, while the certificate captures the essence of the circumstances surrounding death in a few words.
In 2008, the number of poisoning deaths exceeded the number of motor vehicle traffic deaths and was the leading cause of injury death for the fi rst time since at least 1980. During the past three decades, the poisoning death rate nearly tripled, while the motor vehicle traffic death rate decreased by one-half. During this period, the percentage of poisoning deaths that were caused by drugs increased from about 60% to about 90%. The population groups with the highest drug poisoning death rates in 2008 were males, people aged 45–54 years, and non-Hispanic white and American Indian or Alaska Native persons. The vast majority of drug poisoning deaths are unintentional (see Appendix table). Opioid analgesics were involved in more drug poisoning deaths than other specified drugs, including heroin and cocaine. Opioid analgesics were involved in nearly 15,000 deaths in 2008, while cocaine was involved in about 5,100 deaths and heroin was involved in about 3,000 deaths (data not shown). Deaths involving opioid analgesics may involve other drugs as well, including benzodiazepines (2). In addition to an increase in the number of deaths caused by drug poisoning, increases in drug use, abuse, misuse, and nonfatal health outcomes have been observed. In the past two decades, there has been an increase in the distribution and medical use of prescription drugs, including opioid analgesics (3). From 1999 to 2008, the use of prescription medications increased (4). In 2007–2008, 48% of Americans used at least one prescription drug in the past month and 11% of Americans used five or more prescriptions in the past month. Analgesics for pain relief were among the common drugs taken by adults aged 20–59 years (4). In 2009–2010, over 5 million Americans reported using prescription pain relievers nonmedically in the past month (that is, without a doctor’s prescription or only for the experience or feeling they caused), and the majority of people using prescription pain relievers nonmedically reported getting the drugs from friends or family (5,6). From 2004 to 2008, the estimated rate of emergency department visits involving nonmedical use of opioid analgesics doubled from 49 per 100,000 to 101 per 100,000 (7). Government agencies and other organizations joined together to achieve great reductions in the number of deaths from motor vehicle crashes in the past three decades (8,9). A comprehensive approach, including improvements in the safety of vehicles; improvements in roadways; increased use of restraint systems, such as seat belts and child safety seats; reductions in speed; and also efforts to reduce driving under the influence of alcohol and drugs, contributed to the decline in motor vehicle related deaths (8,9). Using a comprehensive, multifaceted approach, it may be possible to reverse the trend in drug poisoning mortality.
Between 1999 and 2006, there was a 120% increase in the rate of unintentional drug overdose deaths in the United States. This study identifies the prevalence of mental illness, a risk factor for substance abuse, and chronic pain among prescription drug overdose deaths in West Virginia and ascertains whether psychotropic drugs contributing to the deaths were used to treat mental illness or for nonmedical purposes. In 2007, we abstracted data on mental illness, pain, and drugs contributing to death from all unintentional prescription drug overdose deaths in 2006 recorded by the West Virginia Office of the Chief Medical Examiner. Decedent prescription records were obtained from the state prescription drug monitoring program. Histories of mental illness and pain were documented in 42.7% and 56.6% of 295 decedents, respectively. Psychotropic drugs contributed to 48.8% of the deaths, with benzodiazepines involved in 36.6%. Benzodiazepines contributing to death were not associated with mental illness (adjusted odds ratio [AOR] = 1.1; 95% CI, 0.6-1.8), while all other psychotropic drugs were (AOR = 3.9; 95% CI, 2.0-7.6). Of decedents with contributory benzodiazepines, 46.3% had no prescription for the drug. Mental illness may have contributed to substance abuse associated with deaths. Clinicians should screen for mental illness when prescribing opioids and recommend psychotherapy as an adjunct or an alternate to pharmacotherapy. Benzodiazepines may have been used nonmedically rather than as a psychotropic drug, reflecting drug diversion. Restricting benzodiazepine prescriptions to a 30-day supply with no refills might be considered.
This report examines the procedures followed in the 1989 revision of the U.S. Standard Certificates of Live Birth and Death; License and Certificate of Marriage; Certificate of Divorce, Dissolution of Marriage, or Annulment; and Reports of Fetal Death and Induced Termination of Pregnancy. It outlines the history and basic principles of the standard certificates and reports and describes the principal additions, modifications, and deletions of items. In addition, it discusses changes in the format of the standard certificates and reports as well as the implementation of the new certificates and reporting forms.
In 1995, a questionnaire was distributed to the > 700 physician medical examiner/coroners (ME/Cs) who are members of the National Association of Medical Examiners (NAME, St. Louis, MO, U.S.A.). The questionnaire consisted of 23 death scenarios for which individual responders were asked to assign a manner of death (homicide, suicide, accident, natural, or undetermined); 198 questionnaires were completed and analyzed. The distribution of manner of death responses was tabulated. In addition, a nosologist from the National Center for Health Statistics was provided with a cause-of-death statement based on each scenario and was asked to assign an International Classification of Diseases (ICD) code for the underlying cause of death, from which a manner of death was inferred from the ICD code's literal text description. Overall, agreement on a given manner of death in a single scenario was > 90% in only 4 of 23 scenarios and > 70% in only 12 of 23 scenarios. However, in 21 scenarios, the most common response comprised a majority. The manner of death inferred from the ICD code that was assigned by the National Center for Health Statistics (NCHS) matched the most common response of participants in 18 of the 23 scenarios. The questionnaire results show that there is substantial disagreement among experienced MEs concerning the manner of death classification that is preferred for selected types of death. Encouraging, however, is the fact that the manner of death coded for statistical purposes generally agreed with the most common classification of manner made by ME/Cs. Highlights from the discussion of each scenario that occurred during the NAME interim meeting (Nashville, Tennessee, February 1996) are also included. Other portions of the program including history of manner of death concepts and results of questions regarding responder training and characteristics are published separately in this issue of the Journal. Information derived from the questionnaire should be useful to those planning strategies to improve the consistency of manner of death classifications by ME/Cs.