J Forensic Sci, November 2014, Vol. 59, No. 6
doi: 10.1111/1556-4029.12541
Available online at: onlinelibrary.wiley.com

PAPER
PATHOLOGY/BIOLOGY; TOXICOLOGY

Kristen J. Mertz,1 M.D., M.P.H.; Jennifer K. Janssen,2 M.S.; and Karl E. Williams,2 M.D.

Underrepresentation of Heroin Involvement in
Unintentional Drug Overdose Deaths in
Allegheny County, PA

ABSTRACT: Drugs contributing to overdose deaths are listed on death certificates, but their validity is rarely studied. To assess the accuracy
of “morphine” and “codeine” listings on death certificates for unintentional overdose deaths in Allegheny County, PA, investigative and laboratory reports were reviewed. Deaths were reclassified as heroin-related if documentation showed 6-monoacetylmorphine in blood or urine,
“stamp bags” or drug paraphernalia at scene, history of heroin use, or track marks. Deaths were considered morphine-related if notes indicated
morphine use, prescription, or morphine at scene, or codeine-related if the codeine blood level exceeded morphine. Of 112 deaths with morphine but not heroin listed on the death certificate, 74 met heroin criteria and 21 morphine criteria. Of 20 deaths with both morphine and heroin
listed, only one met morphine criteria. Of 34 deaths with codeine listed, only five were attributed to codeine. Consideration of patient history,
death scene evidence, and expanded toxicology testing may improve the accuracy of death certificate drug listings.

KEYWORDS: forensic science, overdose, poisoning, heroin, morphine, opioid analgesics

The recent rise in drug overdose deaths in the US has been
linked to an increase in the abuse and misuse of opioid analgesics (1,2). CDC reports that death certificate information indicates that opioid analgesics now account for more overdose
deaths than heroin and cocaine combined (3). The accuracy of
drug listings on death certificates, however, is rarely verified.
A review of unintentional drug overdose deaths from 2008
through 2010 at the Allegheny County Office of the Medical
Examiner (ACOME) in Pittsburgh, PA, indicated that heroin was
listed as a contributing cause on 171 death certificates (unpublished data). The review also revealed 132 deaths with “morphine”
listed on the death certificate, unexpected given the limited availability of pharmaceutical morphine as a drug of abuse compared
with heroin. We were concerned about misclassification of heroin
deaths as morphine deaths, given that heroin rapidly metabolizes
to morphine. Similarly, we suspected that codeine was too often
listed as a contributing cause of death (34 deaths), given that it
can be detected as an impurity in heroin or morphine (4). We
report here the results of an in-depth analysis of unintentional
overdoses with morphine or codeine on the death certificate.
Methods
The Allegheny County Office of the Medical Examiner
(ACOME) is charged with investigating all unnatural or suspicious
1
Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA 15261.
2
Allegheny County Office of the Medical Examiner, 1520 Penn Avenue,
Pittsburgh, PA 15222.
Received 20 April 2013; and in revised form 13 Sept. 2013; accepted 12
Oct. 2013.

© 2014 American Academy of Forensic Sciences

deaths. All suspected drug overdoses reported to the office are
screened for multiple drugs, including opiates by enzyme-linked
immunosorbent assay (ELISA) with confirmation and quantitation by gas chromatography/mass spectrometry. ACOME also
conducts external exams or full autopsies, on-site investigations,
and interviews with next-of-kin for all suspected overdose
deaths. The pathologist involved in the case completes the
cause-of-death section of the death certificate and includes a list
of drugs he believes contributed to death.
We reviewed toxicology results, autopsy reports, and investigative reports for all deaths investigated by ACOME in 2008
through 2010 with “drugs and/or poisoning” listed as cause of
death, “accidental” listed as manner, and “morphine” listed on
the death certificate drug list. Deaths were reclassified as heroin
deaths if reports indicated known heroin use; drug paraphernalia,
typical glassine “stamp bags,” or heroin at the scene; track
marks; or presence of 6-MAM at levels less than the limit of
quantitation (5 ng/mL) but above the limit of detection. Morphine was considered a cause of death if the decedent had a prescription for morphine or a history of morphine use or abuse, or
if morphine was found at the scene. Cases not meeting the criteria for heroin deaths or morphine deaths remained classified as
morphine deaths for lack of a better determination.
Toxicology results for all deaths with codeine listed on the
death certificate were also reviewed. If the codeine concentration
was greater than the morphine concentration, we assumed that
codeine administration was responsible (5–7). If codeine was
detectable only in urine and morphine or 6-MAM was detected,
or if the codeine concentration was less than that of morphine,
we assumed that codeine represented an impurity in commercial
morphine or heroin and that codeine did not contribute to the
death (4,7,8).
1583

 1584

JOURNAL OF FORENSIC SCIENCES

FIG. 1––Unintentional drug overdose deaths involving heroin, morphine,
or codeine according to death certificates and after reclassification, Allegheny County, 2008-2010.

Results
Of 112 unintentional poisoning deaths with morphine but not
heroin listed on the death certificate, 74 (66%) met our criteria
for heroin involvement. Of these 74 decedents, a history of heroin use was noted for 52, drug paraphernalia or stamp bags
(often both) were found at the scene for 33, heroin verified by
drug chemistry was found at the scene for 7, and track marks
were noted on 5; 6-MAM was detected in blood or urine of 17
although at a level <5 ng/mL for 11. Almost half (47%) of the
74 deaths involved more than one of these criteria.
For the 112 deaths with morphine but not heroin listed on the
death certificate, toxicology results indicated that morphine was
detected in all of these cases; however, only 21 of these deaths
satisfied our criteria for morphine-related deaths, based on a history of morphine use/abuse, having a prescription for morphine,
or morphine found at the scene. Another 19 with morphine
detected did not meet our criteria for implicating heroin; thus we
left them classified as morphine.
Of the 20 deaths with both heroin and morphine listed on the
death certificate, all 20 met our criteria for heroin involvement,
but only one met our criteria for morphine involvement.
For 34 deaths with codeine listed on the death certificate, only
5 (15%) met our criteria for codeine involvement with codeine
level greater than morphine level. For 28 deaths, codeine was
most likely detected as an impurity in morphine or heroin. One
death had no evidence of codeine.
Our review led to revisions in the number of unintentional
overdose deaths attributed to heroin, morphine, and codeine in
Allegheny County in 2008 through 2010 (Fig. 1). According to
drugs listed on death certificates, heroin was involved in 171
deaths, morphine in 132, and codeine in 34. After this review,
we attributed a total of 245 deaths to heroin, including the 171
originally identified from death certificates plus 74 identified by
this review, 41 to morphine, and five to codeine.
Discussion
Over 27,000 persons per year die from unintentional drug
overdoses in the United States (9). Public health officials rely on
death certificates for assessing the contribution of specific drugs
to overdose deaths, but the validity of death certificate information is rarely evaluated. In Allegheny County, we found that morphine was overreported as a cause of death on death certificates;
heroin was most likely responsible for a majority of deaths attributed to morphine. Other jurisdictions suspect that some heroin
deaths are misclassified as morphine deaths (10), but the extent

of the problem nationwide is unknown. In Allegheny County, we
also found that codeine was most likely overreported as a cause
of unintentional drug overdose deaths. If other jurisdictions are
similarly overreporting morphine and codeine, national data on
unintentional poisoning may overestimate the contribution of prescription opiates and underestimate heroin involvement.
Medical examiners (MEs) and coroners are responsible for
investigating poisoning deaths and for completing the cause-ofdeath section of the death certificate. When determining which
drugs contribute to death, pathologists must consider multiple
factors, such as drug combinations, levels and ratios; drug history of the decedent; evidence at the scene of death; and time
lapse from overdose to autopsy (11,12). Merely listing drugs
detected by toxicology on the death certificate is misleading;
thorough investigation is needed to supplement toxicology
results. Given that morphine is a metabolite of heroin and
codeine, its detection could indicate the use of one of these
drugs or morphine sulfate. Also drug impurities may give misleading results, as codeine did here. In some situations, additional toxicology studies may help classify cases. Detection of
6-MAM confirms heroin involvement but it rapidly breaks down
in the blood and may not be detectable even when heroin is
involved; 6-MAM is more stable in urine or vitreous fluid (13).
ACOME does a basic drug screen and targeted 6-MAM testing
on urine when available but does not test vitreous fluid for 6MAM, so some heroin cases may have been missed. Targeted
testing of the vitreous for 6-MAM may be useful when urine is
not available or when blood tests suggest but cannot confirm
heroin involvement (13,14).
National surveillance of the drug overdose epidemic is often
based on death certificate data which is not always accurate or
complete. National guidelines to encourage more thorough
review and documentation of the cause of death on death certificate may help to standardize practices and improve data quality.
For opiate overdose deaths, consideration of the decedent’s drug
history, evidence from the death scene, and expanded testing for
6-MAM in the vitreous humor may correctly identify more heroin deaths. A more accurate picture of the drug death epidemic
can strengthen medical, legal, and rehabilitation programs
designed to prevent these premature deaths.
Acknowledgments
We thank the staff of the ACOME for maintaining the data
management system and providing data for this investigation.
We also thank Valerie Alstadt of the University of Pittsburgh for
her review of deaths attributed to morphine, including review of
investigators’ reports and toxicology results.
References
1. Paulozzi LJ, Budnitz DS, Xi Y. Increasing deaths from opioid analgesics
in the United States. Pharmacoepidemiol Drug Saf 2006;15:618–27.
2. Warner M, Chen LH, Makuc DM. Increase in fatal poisonings involving
opioid analgesics in the United States, 1999–2006. NCHS Data Brief
2009;22:1–8.
3. Centers for Disease Control and Prevention (CDC). Vital signs: overdoses of prescription opioid pain relievers – United States, 1999–2008.
MMWR Morb Mortal Wkly Rep 2011;60:1487–92.
4. Ceder G, Jones AW. Concentration ratios of morphine to codeine in
blood of impaired drivers as evidence of heroin use and not medication
with codeine. Clin Chem 2001;47:1980–4.
5. Quiding H, Anderson P, Bondesson U, Boreus LO, Hynning PA. Plasma
concentrations of codeine and its metabolite, morphine, after single and
repeated oral administration. Eur J Clin Pharmacol 1986;30:673–7.

 MERTZ ET AL.

.

UNDERREPRESENTATION OF HEROIN IN DRUG OVERDOSE DEATHS

6. Shah JC, Mason WD. Plasma codeine and morphine concentrations after a
single oral dose of codeine phosphate. J Clin Pharmacol 1990;30:764–6.
7. Tee TT, Chai SL, Anantharaman V, Foon WT. Ratios of total morphine
to total codeine in urine of subjects consuming medicinal preparations
containing morphine or codeine and in drug abusers. Ann Acad Med
Singapore 1979;8:160–3.
8. Jones AW, Holmgren A. Concentration ratios of free-morphine to freecodeine in femoral blood in heroin-related poisoning deaths. Leg Med
(Tokyo) 2011;13:171–3.
9. Centers for Disease Control and Prevention (CDC). CDC grand rounds:
prescription drug overdoses – a U.S. epidemic. MMWR Morb Mortal
Wkly Rep 2012;61:10–3.
10. Sternfeld I, Perras N, Culross PL. Development of a coroner-based surveillance system for drug-related deaths in Los Angeles county. J Urban
Health 2010;87:656–69.
11. Cone EJ, Fant RV, Rohay JM, Caplan YH, Ballina M, Reder RF, et al.
Oxycodone involvement in drug abuse deaths. II. Evidence for toxic
multiple drug-drug interactions. J Anal Toxicol 2004;28:217–25.

1585

12. Drummer OH. Recent trends in narcotic deaths. Ther Drug Monit
2005;27:738–40.
13. Wyman J, Bultman S. Postmortem distribution of heroin metabolites in
femoral blood, liver, cerbrospinal fluid, and vitreous humor. J Anal Toxicol 2004;28:260–3.
14. Pragst F, Spiegel K, Leuschner U, Hager A. Detection of 6-acetylmorphine in vitreous humor and cerebrospinal fluid – comparison with urinary analysis for proving heroin administration in opiate fatalities. J Anal
Toxicol 1999;23:168–72.
Additional information and reprint requests:
Kristen J. Mertz, M.D., M.P.H.
University of Pittsburgh
512 Parran Hall
130 DeSoto Street
Pittsburgh, PA 15261
E-mail: mertzk@edc.pitt.edu