Recommendations for Fatal Drug Overdose Surveillance
Methodology and Data Quality Improvements:
A Follow-up to the Safes States Injury Surveillance
Workgroup on Poisoning
CSTE Overdose
Subcommittee Presentation
2016 Annual Meeting
Safe States Alliance
April 12-14, Albuquerque, NM
Presenters and Co-Authors
• Dr. Svetla Slavova, PhD, Kentucky Injury Prevention and Research Center
• James W Davis, MA, New Mexico Department of Health
• Dr. Denise Paone, EdD, New York City Department of Health and Mental Hygiene
• Barbara Gabella, MSPH, Colorado Department of Public Health and Environment
Co-Authors:
• Dr. Jennifer C Sabel, PhD, Washington State Department of Health
• Dr. Dagan A Wright, PhD, MSPH, Oregon Health Authority – Department of
Public Health
Fatal Drug Overdose Surveillance
“Levels 1 - 4”
Level 4: Toxicology and
Literals
Level 3: Literal text from death certificate
Level 2: Multiple Cause
Level 1: Basic Underlying
cause of death file
Agenda
• Safe States & CSTE (Denise)
• Levels 1 and 2 (Svetla)
• Level 3 & Epi tool for everyone (Jim)
• Level 4 (Denise)
• Lessons learned paper (Svetla, Denise, Jim)
• Future & your ideas (Barbara)
Safe States Background
• Safe States Poisoning
Workgroup – 2008
• Safe States forms Injury
Surveillance Workgroup (ISW7)
• ISW7 Publication – 2012
• Conferences of CDC, Safe
States, CSTE - 2012
CSTE Background
• Council of State and Territorial Epidemiologists (CSTE)
• CSTE Drug Overdose
– Willing and wanting to test ISW7 indicators
– Providing organizational base for effort
• Real partnership with Safe States – liaison
• Scott Proescholdbell
www.cste.orgUsing the power of epidemiology to improve
the public’s health
Why CSTE Overdose Subcommittee
chose this work
• Increased attention and need to track drug overdose
• Lack of consistent definitions and indicators
• Variations across states and local jurisdictions
• Recognition that lack of detail on death certificates
underestimates deaths for specific drugs
• Concern about shifting from opioid analgesics to heroin
→ accurate classification of opiates critical
Timeline 2012-2013
• August 2012 - Organizational call
• Creation of Poisoning Surveillance Workgroup
• Fall 2012 - Creation of tables shells and Levels I-IV (basic
to complex)
• Winter- submission of data by volunteer states and
proposal to present data
• Spring 2013 - Data results & write up of methods and
process for Special Emphasis Report
• April 2013 - CDC meeting on poisoning for Core states
Fatal Drug Overdose Surveillance
“Levels 1 - 4”
Level 4: Toxicology and
Literals
Level 3: Literal text from death certificate
Level 2: Multiple Cause
Level 1: Basic Underlying
cause of death file
Svetla Slavova, PhD
Associate Professor, University of Kentucky
Kentucky Injury Prevention and Research Center
Underlying and Multiple
Causes of Death
(Level 1 and 2)
Level 1 Background
• Underlying cause of death and multiple causes of death listed on death certificates are coded in ICD-10.
• A drug overdose death is defined as a death with an underlying cause of death in these ICD-10 ranges:– X40–X44 (unintentional),
– X60–X64 (suicide),
– X85 (assault), and
– Y10–Y14 (undetermined intent).
• ICD-10 codes T36 -T50 identify the specific drugs or drug classes contributing to an overdose deaths.
• The next three slides are examples of how the information might appear on the death certificate and in the electronic data file.
CAUSE OF DEATH (See instructions and examples) Approximate
interval:
Onset to death32. PART I. Enter the chain of events - - diseases, injuries, or complications - - that directly caused the death. DO NOT enter
terminal events such as cardiac arrest, respiratory arrest, or ventricular fibrillation without showing the etiology. DO NOT
ABBREVIATE. Enter only one cause on a line. Add additional lines if necessary.
IMMEDIATE CAUSE (Final
disease or condition resulting in death)a. ANOXIA BRAIN INJURY
Sequentially list conditions,
if any, leading to the cause
listed on line a. Enter the
UNDERLYING CAUSE (disease
or injury that initiated the events
resulting in death) LAST
Due to (or as a consequence of):
b. SECONDARY TO INGESTION OF METHADONE
Due to (or as a consequence of):
c.
Due to (or as a consequence of):
d.
PART II. Enter other significant conditions contributing to death but not resulting in the underlying cause
given in Part I.
33. WAS AN AUTOPSY PERFORMED?
Yes No
34. WERE AUTOPSY FINDINGS AVAILABLE
TO COMPLETE THE CAUSE OF DEATH?
Yes No
35. DID TOBACCO USE
CONTRIBUTE TO DEATH?
Yes Probably
No Unknown
36. IF FEMALE:
Not pregnant within past year
Pregnant at time of death
Not pregnant but pregnant within 42 days of death
Not pregnant but pregnant 43 days to 1 year before death
Unknown if pregnant within the past year
29. MANNER OF DEATH
Natural Pending
Accident Investigation
Suicide Could not be
Homicide Determined
38. DATE OF INJURY
(Mo/Day/Yr)(Spell Month)
39. TIME OF INJURY 40. PLACE OF INJURY (e.g., Decedent’s home, construction
site, restaurant, wooded area)
HOME
41. INJURY AT WORK?
Yes No
42. LOCATION OF INJURY: State: City or Town:
Street & Number: Apartment No. Zip Code:
43. DESCRIBE HOW INJURY OCCURRED
DRUG INGESTION
44. IF TRANSPORTATION INJURY, SPECIFY
Driver/Operator
Passenger
Pedestrian
Other (Specify)
To B
e C
om
ple
ted B
y:
ME
DIC
AL C
ER
TIF
IER
X
CAUSE OF DEATH (See instructions and examples) Approximate
interval:
Onset to death32. PART I. Enter the chain of events - - diseases, injuries, or complications - - that directly caused the death. DO NOT enter
terminal events such as cardiac arrest, respiratory arrest, or ventricular fibrillation without showing the etiology. DO NOT
ABBREVIATE. Enter only one cause on a line. Add additional lines if necessary.
IMMEDIATE CAUSE (Final
disease or condition resulting in death)a. ANOXIA BRAIN INJURY
Sequentially list conditions,
if any, leading to the cause
listed on line a. Enter the
UNDERLYING CAUSE (disease
or injury that initiated the events
resulting in death) LAST
Due to (or as a consequence of):
b. SECONDARY TO INGESTION OF METHADONE
Due to (or as a consequence of):
c.
Due to (or as a consequence of):
d.
PART II. Enter other significant conditions contributing to death but not resulting in the underlying cause
given in Part I.
33. WAS AN AUTOPSY PERFORMED?
Yes No
34. WERE AUTOPSY FINDINGS AVAILABLE
TO COMPLETE THE CAUSE OF DEATH?
Yes No
35. DID TOBACCO USE
CONTRIBUTE TO DEATH?
Yes Probably
No Unknown
36. IF FEMALE:
Not pregnant within past year
Pregnant at time of death
Not pregnant but pregnant within 42 days of death
Not pregnant but pregnant 43 days to 1 year before death
Unknown if pregnant within the past year
29. MANNER OF DEATH
Natural Pending
Accident Investigation
Suicide Could not be
Homicide Determined
38. DATE OF INJURY
(Mo/Day/Yr)(Spell Month)
39. TIME OF INJURY 40. PLACE OF INJURY (e.g., Decedent’s home, construction
site, restaurant, wooded area)
HOME
41. INJURY AT WORK?
Yes No
42. LOCATION OF INJURY: State: City or Town:
Street & Number: Apartment No. Zip Code:
43. DESCRIBE HOW INJURY OCCURRED
DRUG INGESTION
44. IF TRANSPORTATION INJURY, SPECIFY
Driver/Operator
Passenger
Pedestrian
Other (Specify)
To B
e C
om
ple
ted B
y:
ME
DIC
AL C
ER
TIF
IER
X
G93.1 Anoxic brain damage, NEC
T40.3 Methadone
T50.9 Other and unspecified drug
X42 Accidental poisoning and exposure to
narcotics and psychodysleptics, NEC
Part I Part II Underlying Cause Multiple Causes of Death
How injury occurred Manner Line a Line b Line c Line d
Significantconditions
contributing to death
ICD-10 Code Text
1st in ICD-10 1 in Text
2nd in ICD-10 2nd in Text
3rd in ICD-10 3rd in Text
PRESCRIPTION DRUG OVERDOSE ACCIDENT ANOXIC BRAIN INJURY
ACUTE OXYCODONE TOXICITY X42
Accidental poisoning by and exposure to narcotics and psychodysleptics [hallucinogens], NEC G931
Anoxic brain damage, not elsewhere classified T402
Poisoning, other opioids T509
Poisoning, Other and unspecified drugs
INJECTED HEROIN ACCIDENT HEROIN INTOXICATION X42
Accidental poisoning by narcotics and psychodysleptics [hallucinogens], NEC F119
Unspecified mental and behavioral disorder T401
Poisoning, heroin
MULTIPLE DRUG INTOXICATION ACCIDENT
MULTIPLE DRUG INTOXICATION (OXYCODONE AND DIAZEPAM) X44
Accidental poisoning by and exposure to other and unspecified drugs T402
Poisoning, other opioids T424
Poisoning, Benzodiazepines T509
Poisoning, Other and unspecified drugs
USED TOXIC LEVEL OF BATH SALTS ACCIDENT
ACUTE ALPHA-PVP TOXICITY (BATH SALTS)
HISTORY OF DRUG ABUSE X44
Accidental poisoning by and exposure to other and unspecified drugs T509
Poisoning, Other and unspecified drugs F191 Harmful use
COMBINED DRUG TOXICITY ACCIDENT
COMBINED DRUG TOXICITY X44
Accidental poisoning by and exposure to other and unspecified drugs T509
Poisoning, Other and unspecified drugs
SUBJECT WAS FOUND UNRESPONSIVE IN BED BY A FAMILY SUICIDE
ACUTE COMBINED DRUG TOXICITY (LORAZEPAM, BUPROPION, VENLAFAXINE) X61
Intentional self-poisoning (suicide) by and exposure to antiepileptic, sedative-hypnotic, antiparkinsonism, and psychotropic drugs,NEC T424
Poisoning, Benzodiazepines T432
Poisoning, other and unspecified antidepressants T509
Poisoning, Other and unspecified drugs
DECEDENT INGESTED DRUGS
UNDETERMINED
ACUTE FENTANYL INTOXICATION Y12
Poisoning by and exposure to narcotics and psychodysleptics[hallucinogens], not elsewhere classified, undetermined intent T404
Poisoning, other synthetic narcotics T509
Poisoning, Other and unspecified drugs
Level 1 Example Text & Codes
Used “underlying cause of death” only
Classification by intent and by drug type
11 states (KS, KY, MA, MD, MI, NC, NYC, OK, OR, UT, WA)
produced the Safe States ISW7 recommended indicators
Drug Type Unintentional Suicide Undetermined Homicide
Nonopioid analgesics,
antipyretics and anti-
rheumatics
X40 X60 Y10
X85
Sedative-hypnotic,
psychotropic
X41 X61 Y11
Narcotics and
psychodysleptics
(hallucinogens), NOS
X42 X62 Y12
Other drugs acting on
CNS
X43 X63 Y13
Other and unspecified
drugs
X44 X64 Y14
Level 1 Method
Level 1 Results(Underlying Cause of Death only)
Comparing states/jurisdictions on drug overdose rates by
intent could be misleading!
0.0%
10.0%
20.0%
30.0%
40.0%
50.0%
60.0%
70.0%
80.0%
Percentage of Drug Overdose Deaths with "Undetermined" Intent, By State, 2014
Level 1 Results
Level 1 Results
CSTE Overdose Subcommittee
Recommendations
• It is better to report all drug overdoses, regardless of
intent, for state-to-state comparisons.
• Categorization by intent is still useful, for example, to
examine the proportion of the drug overdose deaths
by intent over time within a given state/jurisdiction.
Fatal Drug Overdose Surveillance
“Levels 1 - 4”
Level 4: Toxicology and
Literals
Level 3: Literal text from death certificate
Level 2: Multiple Cause
Level 1: Basic Underlying
cause of death file
Level 2 Method
• Use underlying cause of death and multiple causes of death to identify drugs contributing to overdose deaths.
• When no specific drugs/classes of drugs are listed on the death certificate, the overdose is attributed to– “Other and Unspecified Drugs”, coded with ICD-10 code T50.9.
• An overdose death could be attributed to “Other and Unspecified Drugs” when – it involves drug that cannot be classified under any other existing
category (e.g., “bath salts”).
7,577
8,297
0 2000 4000 6000 8000 10000
Y14 (Poisoning by and exposure to other and unspecified drugs,medicaments and biological substances, undetermined intent)
Y12 (Poisoning by and exposure to narcotics and psychodysleptics[hallucinogens], not elsewhere classified, undetermined intent)
X85 (Assault by drugs, medicaments and biological substances)
X64 (Intentional self-poisoning by and exposure to other andunspecified drugs, medicaments and biological substances)
X62 (Intentional self-poisoning by and exposure to narcotics andpsychodysleptics [hallucinogens], not elsewhere classified)
X44 (Accidental poisoning by and exposure to other and unspecifieddrugs, medicaments and biological substances)
X42 (Accidental poisoning by and exposure to narcotics andpsychodysleptics [hallucinogens], not elsewhere classified)
U.S. Drug Overdose Deaths by Selected Underlying and Multiple Causes of Death, 2014
Cocaine (T40.5) Heroin (T40.1) Pharmaceutical opioids (T40.2, T40.3, or T40.4)
Level 2 Example Text & Codesfor underlying cause of death and multiple causes of death
Part I Part II Underlying Cause Multiple Causes of Death
How injury occurred Manner Line a Line b Line c Line d
Significantconditions
contributing to death
ICD-10 Code Text
1st in ICD-10 1 in Text
2nd in ICD-10 2nd in Text
3rd in ICD-10 3rd in Text
PRESCRIPTION DRUG OVERDOSE ACCIDENT ANOXIC BRAIN INJURY
SECONDARY TO INGESTION OF METHADONE X42
Accidental poisoning by and exposure to narcotics and psychodysleptics [hallucinogens], NEC G93.1
Anoxic brain damage, not elsewhere classified T40.3
Poisoning, methadone T50.9
Poisoning, Other and unspecified drugs
INJECTED HEROIN ACCIDENT HEROIN INTOXICATION X42
Accidental poisoning by narcotics and psychodysleptics [hallucinogens], NEC F11.9
Unspecified mental and behavioral disorder T40.1
Poisoning, heroin
MULTIPLE DRUG INTOXICATION ACCIDENT
MULTIPLE DRUG INTOXICATION (OXYCODONE AND DIAZEPAM) X44
Accidental poisoning by and exposure to other and unspecified drugs T40.2
Poisoning, other opioids T42.4
Poisoning, Benzodiazepines T50.9
Poisoning, Other and unspecified drugs
USED TOXIC LEVEL OF BATH SALTS ACCIDENT
ACUTE ALPHA-PVP TOXICITY (BATH SALTS)
HISTORY OF DRUG ABUSE X44
Accidental poisoning by and exposure to other and unspecified drugs T50.9
Poisoning, Other and unspecified drugs F19.1 Harmful use
COMBINED DRUG TOXICITY ACCIDENT
COMBINED DRUG TOXICITY X44
Accidental poisoning by and exposure to other and unspecified drugs T50.9
Poisoning, Other and unspecified drugs
SUBJECT WAS FOUND UNRESPONSIVE IN BED BY A FAMILY SUICIDE DRUG OVERDOSE X61
Intentional self-poisoning (suicide) by and exposure to antiepileptic, sedative-hypnotic, antiparkinsonism, and psychotropic drugs,NEC T50.9
Poisoning, Other and unspecified drugs
Centers for Disease Control and Prevention, National Center for Health Statistics. Multiple Cause of Death 1999-2014 on CDC WONDER Online Database, released 2015.
Data are from the Multiple Cause of Death Files, 1999-2014, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative
Program. Accessed at http://wonder.cdc.gov/mcd-icd10.html on Mar 30, 2016 1:51:15 PM
Level 2 Results
Centers for Disease Control and Prevention, National Center for Health Statistics. Multiple Cause of Death 1999-2014 on CDC WONDER Online Database, released 2015.
Data are from the Multiple Cause of Death Files, 1999-2014, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative
Program. Accessed at http://wonder.cdc.gov/mcd-icd10.html on Mar 30, 2016 1:51:15 PM
Level 2 Results
Centers for Disease Control and Prevention, National Center for Health Statistics. Multiple Cause of Death 1999-2014 on CDC WONDER Online Database,
released 2015. Data are from the Multiple Cause of Death Files, 1999-2014, as compiled from data provided by the 57 vital statistics jurisdictions through the Vital
Statistics Cooperative Program. Accessed at http://wonder.cdc.gov/mcd-icd10.html on Mar 30, 2016 1:51:15 PM
Drug Overdose Deaths: Let’s Get Specific.
Slavova, S., Bradley O’Brien, D., Creppage, K., Dao, D., Fondario, A., Haile, E., Hume, B., Largo, T., Nguen, C.,
Sabel, J., Wright, D.
Public Health Reports.
July-August 2015. Volume 130, pg.339-342
CSTE Overdose Subcommittee
Recommendations
• The use of nonspecific language to identify specific drugs
on death certificates can result in undercounting various
drug classes.
• Ranking jurisdictions by specific drug types identified on
death certificates might be misleading and should be
accompanied by analysis of the level of completeness and
specificity of the jurisdictional data.
CSTE Overdose Subcommittee
Recommendations
Epidemiologists and other public health practitioners:
• Need to be aware of the quality and limitations of the death
certificate data in their jurisdiction,
• Evaluate when possible the level of completeness and
accuracy of their Multiple Cause of Death data, and
• Interpret the reported counts and rates with caution when
the proportion of deaths with ‘Other and Unspecified Drugs”
is considerable.
CSTE Overdose Subcommittee
Recommendations
At the state and local jurisdictions, where possible:
• Identify and address factors that contribute to differences in
how states record drug overdose deaths.
• Compare with medical examiner/coroner and state
toxicology laboratory records, exactly which drugs were
involved in deaths coded nonspecifically to provide better
estimates of the contributions of individual drug types to
their overdose problem.
Fatal Drug Overdose Surveillance
“Levels 1 - 4”
Level 4: Toxicology and
Literals
Level 3: Literal text from death certificate
Level 2: Multiple Cause
Level 1: Basic Underlying
cause of death file
Literal Text Fields
and
Beyond ICD-10
(Level 3)
Jim Davis
Drug Epidemiologist
New Mexico Department of Health
Level 3 Background
• In overdose deaths, text literals often contain:• Toxic effects of heroin [or oxycodone, or methamphetamine, …]
• Multiple drug (heroin, methadone and alcohol) intoxication
• Atherosclerotic cardiovascular disease and methamphetamine toxicity
• ICD-10 coding collapses some categories of interest
• Particularly the prescription opioids and benzodiazepines
• Text recognition not a new thing, but no general tool existed
for drug recognition in death data
Level 3 Methods & Epi Tool
• SAS program adapted from NM work
– Uses macro variables to define the variable names
– Can scan one or more of:• 4 variables for cause of death text
• 1 variable for “other factors” (part II)
• 1 variable for injury description
• Simple word-by-word search of specified literal text variables
• Common drugs found on death certificates listed by NCHS
– Includes generic and trade names
– Includes common abbreviations and misspellings
Example of Drug List
Terms found on
death certificates Drug
Misspelling
or Metabolite
Substance
Type
Pharmacologic
Type
ICD T Code Drug
Classification
DARVOCET PROPOXYPHENE Brand opioid
T40.2-T40.4 (opioid
analgesics)
DARVON PROPOXYPHENE Brand opioid
T40.2-T40.4 (opioid
analgesics)
DEXTROPROPOX
YPHENE PROPOXYPHENE Metabolite
Unspecified
brand or
generic opioid
T40.2-T40.4 (opioid
analgesics)
PROOXYPHENE PROPOXYPHENE Misspelling Generic opioid
T40.2-T40.4 (opioid
analgesics)
PROPOSYPHENE PROPOXYPHENE Misspelling Generic opioid
T40.2-T40.4 (opioid
analgesics)
PROPOXIPHENE PROPOXYPHENE Misspelling Generic opioid
T40.2-T40.4 (opioid
analgesics)
Level 3 Methods (continued)
• Generates a list (array) of drug names and classifications
• Plus a slash-separated list of drug names (e.g.
OXYCODONE/ALPRAZOLAM)
• Additional programming is needed to create variables for
analysis or reporting
• Available on the CSTE web site (in SAS) :
http://www.cste.org/members/group.aspx?id=87615
• R version also available (Chris Ryan – Binghamton, NY)
• Has been used in several states (NM, KY, OR, . . .)
Limitations
• Assumes knowledge of SAS (or R)
• Specificity issues/completeness of the input data– “Mixed drug intoxication” is not specific
– Specificity varies by jurisdiction and over time
• Consistency over time – Different jurisdictions code differently
– Approaches to coding change over time
• Limitations of the list– Infinite number of ways to misspell drug names
• Does not try to resolve morphine vs. heroin– Morphine is the primary metabolite of heroin
An Example from NM Data
0
2
4
6
8
10
12
14
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
20
13
20
14
Death
s p
er
100,0
00
po
pu
lati
on
Drug Overdose Death rates for Selected Drugs, NM 1990-2014
Rx Opioid Heroin Methamphetamine Cocaine
Drug categories are not mutually exclusive2014 rates preliminaryRates age-adjusted to the US 2000 standard populationSource: Office of the Medical Investigator; UNM/GPS population
Top Rx Drugs in Overdose Death, NM 2014
0 20 40 60 80 100 120
oxycodone
hydrocodone
alprazolam
morphine
methadone
fentanyl
diazepam
Overdose death involvements
Deaths may involve more than one drugSource: NM Office of the Medical Investigator
An Example from Kentucky Data
0
50
100
150
200
250
300
2011 2012 2013 2014
Invo
lvem
en
ts
Selected Drugs Commonly Listed as Contributing to Kentucky Resident Drug Overdose Deaths, 2011-2014
Alprazolam Oxycodone Hydrodocone Fentanyl Heroin
2011-2014 data are provisional and subject to changeSource: http://www.mc.uky.edu/kiprc/programs/KVIPP/drug_overdose_deaths_00-14.pdf – Table 2Fentanyl increase in 2014 primarily due to illicit fentanyl
Level 3 Wrap up
• Literal text in death data can be useful for a variety of problems.
• Used experimentally with other data sources (ED, EMS)
• Epi tool available from CSTE website for SAS:
http://www.cste.org/members/group.aspx?id=87615
• Further development is possible to suit various needs.
Fatal Drug Overdose Surveillance “Levels 1 - 4”
Level 4: Toxicology and
Literals
Level 3: Literal text from death certificate
Level 2: Multiple Cause
Level 1: Basic Underlying
cause of death file
Multistage Analysis
(Introduction to Level 4)
Denise Paone, EdD
Senior Director of Research and Surveillance
New York City Department of Health and Mental Hygiene
Bureau of Alcohol and Drug Use Prevention, Care, and
Treatment
Issues
• 23% of drug overdose deaths in 2012, there were
not any drugs listed on the death certificate (wide
range).
• Morphine is used on death certificates. Does this
mean heroin or pharmaceutical morphine?
• Medical examiners have toxicology data that could
help improve accuracy.
• Death certificate (literals)
• Toxicology results
• Electronic records
• Chart review• Requires capacity
• Coroner/Medical examiner office
• Confirm they test for drug metabolites
• Familiarity with the drug metabolites
What do you need for Level 4?
Level 4 Approach
Compare toxicology results to drugs listed on the death
certificates and corresponding T codes across three
jurisdictions participating in CSTE Overdose
Subcommittee
Level 4 Objectives
• Determine the sensitivity of death certificates for
identifying drug-specific poisoning deaths in three
jurisdictions (NYC, OR, King County WA)
• Is there variation by jurisdiction?
• Next steps - How to improve sensitivity
Level 4 Methods
• Limited to unintentional drug poisoning (overdose) deaths
– Using death certificate (underlying, contributing)
• Limited to decedents 15-84 years old
• Linked death certificates to toxicology files for 2012
Level 4 Methods
• Included: benzodiazepines, cocaine, heroin, methadone,
other opioid analgesics, and methamphetamines
• All metabolites were included.
• NYC did not report methamphetamine.
Level 4 Methods
• For toxicology reports indicating morphine, relied on literal
text
– Pharmaceutical morphine = opioid analgesic
– Probable pharmaceutical morphine = opioid analgesic
– All other morphine = heroin
– Morphine + codeine = heroin
– Includes cases where the source of the morphine was
listed as unknown on death certificate
T-Codes
Drug Category ICD-10 code
Benzodiazepines T42.4
Cocaine T40.5
Heroin T40.1
Methadone T40.3
Methamphetamine (Psychostimulants) T43.6
Other opioid analgesics
Natural and semi-synthetic opioid analgesics
(aka “Other opioids”)T40.2
Synthetic opioid analgesics, excluding
methadone (aka “other synthetic narcotics”)T40.4
Sensitivity
• Calculated the sensitivity of the death certificate to detect drug-specific poisoning deaths, using the toxicology results as the gold-standard:
• The sensitivity of death certificates were calculated separately for benzodiazepines, cocaine, heroin, methadone, and opioid analgesic-involved deaths.
Sensitivity =Ʃ Condition positive (Death certificate)
Ʃ True positive (Toxicology)
Comparing toxicology results and
death certificate
Drug Type # of Deaths w/ Drug
Present in Toxicology
Results
# of Deaths w/
Appropriate T
Code
Sensitivity
Benzodiazepine 431 262 60.8%
Cocaine 412 302 73.3%
Heroin 594 414 69.7%
Methadone 289 250 86.5%
Methamphetamine** 83 77 92.8%
Other Opioid
Analgesics 308 295 95.8%
** Methamphetamine compared in OR & King County, WA only.
Sensitivity of death certificate compared
to toxicology by jurisdiction
88%
95%92%
100%97%
100%
68%71%
61%
82%
91%
23%
71%
82%
91% 90%
107%
0%
20%
40%
60%
80%
100%
120%
Benzodiazepine Cocaine Heroin Methadone Methamphetamine Other OpioidAnalgesics
Perc
en
tag
e
WA (King County) NYC OR
Number of drug deaths with toxicology
missing from death certificates by jurisdiction
72 6
0 1 0
90
102
150
34
18
72
6
24
5 5 -5
-10
10
30
50
70
90
110
130
150
170
Benzodiazepine Cocaine Heroin Methadone Methamphetamine Other OpioidAnalgesics
Dif
fere
nc
e b
etw
ee
n t
ox
& d
ea
th c
ert
ific
ate
s
WA (King County) NYC OR
Level 4 Summary
• Benzodiazepines had the lowest sensitivity in the
three jurisdictions.
• Heroin sensitivity was second lowest.
• Sensitivity varied most widely for benzodiazepines.
• King County, WA had highest sensitivity of three
jurisdictions across all drug categories (except 107%
in OR – opioid analgesics).
Level 4 Conclusions
• Using death certificates alone results in underreporting
of drug-specific mortality rates.
• Important to note that using death certificates alone will
provide the number of drug poisoning deaths.
• Death certificate sensitivity varies by drug type and
jurisdiction.
Level 4 Conclusions
• Discordance between toxicology findings and literal
on death certificate, thus T-codes.
• Impact of F-codes on reporting drug poisoning
deaths– Why does “Acute and Chronic” not always get X(acute) code?
• Potential misclassification of Opioid Analgesic and
Heroin deaths
• Increasing concern about transitioning to heroin,
critical not to under report
Level 4 Implications
• Underestimation of drug-specific mortality on death
certificates can compromise effective targeting of
resources and public health interventions.
• Medical examiner/coroner information on drug-related
deaths could be systematically included in death certificate
cause-of-death documentation.
• There is no national standard to report drug poisoning
deaths. National standards could provide guidance for
drug-specific reporting and interpretation of toxicology
results.
Lessons Learned
Paper by members of CSTE Overdose
Recommendations & Lessons Learned for Improved
Reporting of Drug Overdose Deaths on Death
Certificates
Jennifer Sabel1, Ellenie Tuazon2, Denise Paone2, Svetla Slavova3,
Terry Bunn3, Dan Dao4, David Nordstrom5, Holly Hedegaard6
Affiliations: 1 Washington State Department of Health2 New York City Department of Health and Mental Hygiene3 Kentucky Injury Prevention and Research Center4 Kansas Department of Health and Environment5 Wisconsin Department of Health6 National Center for Health Statistics
Recommendations for epidemiologists
• Learn more about existing studies, methods and tools – Baseline state-to-state comparison of drug overdose deaths1
– Evaluation of the completeness and accuracy of the multiple cause-of-
death fields for drug overdose deaths2
– CSTE Overdose Subcommittee Epi tool
• Review your drug overdose data– CDC National Center for Injury Prevention and Control’s State Special
Emphasis Report (SER) Instructions for Drug Overdose Death Data
– Assess data quality • % of drug overdose deaths with unspecified drugs (T50.9, T40.6)
• % all deaths (not just drug overdose deaths) with an underlying cause of R99
• Collaborate with state vital statistics registrar • Share observations about the quality of the data from death certificates
on drug overdose deaths with their state vital statistics registrar
1. Warner M, Paulozzi LJ, Nolte KB, Davis GG, Nelson LS. (2013). State Variation in Certifying Manner of Death and Drugs Involved in Drug
Intoxication Deaths. Acad Forensic Pathol, 3(2), 231-237.
2. Slavova S, Bradley O’Brien D, Creppage K, Dao D, Fondario A, Haile E, Hume B, Largo T, Nguen C, Sabel J, Wright D. (2015). Drug
Overdose Deaths: Let’s Get Specific. Public Health Rep, 130, 339-341
Kentucky Experience:Collective effort of multiple stakeholders in the state (Office of the Chief Medical Examiner,
Kentucky Coroner’s Association, Office of Vital Statistics, Office of Drug Control Policy,
Kentucky Injury Prevention and Research Center, legislators, and others).
Source: Drug Overdose Deaths, Hospitalizations, and Emergency Department Visits in Kentucky, 2000-2012. KIPRC. January, 2014.
NYC: Collaborate with local vital
statistics registrar Issues we have discussed with local medical examiners:
1) Deaths coded as “acute and chronic substance use,” which resulted in these deaths being
assigned an underlying cause of death as ‘related to mental health or behavior’, and
coded with an F-code
– Follow-up with OCME on this issue resulted in improved fewer F-codes and an
increase in drug overdose deaths
2) Heroin deaths with coding as “morphine”
– Follow-up with OCME on this issue to consistently include 6-monoacetlylmorphine on
the death certificate as positive toxicology for heroin
– OCME started testing for 6-monoacetlymorphine in urine and vitreous samples to their
standard protocols, resulted in improved specificity of heroin-involved overdose deaths
3) Lack of drug specificity in literals: heroin deaths as “opiate” or “opioid” deaths, use of
“multiple drug toxicity” or “multiple drug involvement”
– Medical examiners acknowledged this issue, and many were committed to improving
this in their work.
– Other issues were also brought up including cocaine-involved and benzodiazepine
involved
Continued collaboration is key
• Reporting on drug overdose surveillance requires
collaboration and communication with medical examiner
• Presentation of aggregate data useful
– Medical examiners look at cases, not aggregate
– Highlights trends
– Describes drug specificity
– Generates discussion
• Check with medical examiners on testing practices
– Fentanyl
New Mexico: Working with
the Medical Investigator
• Central statewide Medical Investigator
• Long relationship with DOH/Epi – providing death data
• Very high level of specific drug coding – in a separate table
• “Multiple drug toxicity” in the text fields, but specifics
available
• Changed their database system in mid 2010, dropping that
table
• New group of pathologists and new leadership
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
*2015 based on January-June
Source: Office of the Medical Investigator, UNM/GPS population estimates
Percent of Drug Overdose Deaths with
Specific Drugs Coded, NM, 1990-2015
NM: Working with
the Medical Investigator
• Annual data transfer, so problem not noticed until
mid 2011
• Initial failure to understand what had happened
• Meeting in early 2012 with agreement to list drugs
in the text fields
• Lessons:
– Pay Attention!
– Maintain relationships – keep in touch
– Ask questions
Barbara Gabella, MSPH
Injury Epidemiologist
Colorado Department of Public Health and
Environment
Future projects?
Future Project?
• Several states test BRFSS questions on chronic pain,
opioid use, or misuse
– Utah presented on their results
– Search and update voluntary database on state
added BRFSS questions
– NM will measure chronic pain
– Recommend question set for multiple states to field?
Future Project?
• Survey states about testing
– Able to distinguish source of the morphine?
• Heroin, fentanyl, and heroin/fentanyl overdose deaths
– Drafting methods
– Similar to studies described today
– Volunteer states analyze deaths in 2015
– Preliminary results for CSTE pre-conference
workshop mid-June
Significance to the Field
• Need: Accurate data to evaluate effectiveness of the laws, policies, and system changes intended to prevent drug overdoses.
• Improved drug overdose surveillance methodology and data quality
• Increased collaboration among epidemiologists Come join us or tell your staff!
• Goal: Increased capacity to address emerging drug overdose surveillance problems.
– Heroin
– Fentanyl-laced heroin
Questions? Ideas?
• CSTE Drug Overdose Subcommittee
meets monthly on the 2nd Thursday
• 1:00pm to 2:00pm Eastern time
• Dial in number: 877-626-7137
• Pass code: 695041#
• Contact: Nidal Kram at [email protected]