Suspected North Carolina counterfeit pill-involved deaths, 2020–2022

Abstract

The NC Office of the Chief Medical Examiner regularly assumes jurisdiction over deaths that are suspicious, unusual or unattended by a medical professional. In recent years, the presence of counterfeit pills is occasionally suggested by investigatory notes and/or scene findings that document reported consumption of prescription drugs, or prescription drugs on scene, which are not reflected in the final autopsy findings after toxicological analysis of the decedent’s blood samples. Counterfeit pill consumption is a major public health hazard worthy of attention from the forensic toxicology community. Seventy-five cases from January 2020 to December 2022 serve as a convenience sample of cases where prescription pills including formulations of alprazolam, oxycodone and hydrocodone were specifically referenced during the death scene investigation as recently consumed, yet an unexpected substance was found during toxicological analysis rather than the expected pharmaceutical drug. Of note, novel benzodiazepines detected included flualprazolam, etizolam, clonazolam metabolite (8-aminoclonazolam), bromazolam, flubromazolam and desalkylflurazepam. Decedents’ ages ranged from 16 to 69, across 33 different NC counties. Case notes indicated that eight of the decedents obtained pills through direct personal relationships, six decedents obtained them from “the street” and one decedent likely purchased pills online. Pills were largely consumed orally or through insufflation. Seven case reports contained indication that decedents knew or suspected the counterfeit nature of their pills. This study describes the context and characteristics of 2020–2022 suspected counterfeit pill-involved deaths in NC to further the understanding of the forensic science community, law enforcement partners, public health stakeholders and those potentially at risk through the consumption of counterfeit pills.

Introduction

The increase in counterfeit pills in recent years has added complexity and unknown risk into the US drug supply and poses a major public health and safety concern. According to an analysis of national seizure data from the High Intensity Drug Trafficking Areas program, from Quarter 1 of 2018 to Quarter 4 of 2021, the number of fentanyl-containing pill seizures increased from 68 to 635, and the proportion of fentanyl-containing pills to total seizures increased from 13.8% to 29.2% (1). In August and September 2021, the Drug Enforcement Administration (DEA) seized 1.8 million fentanyl-laced pills, and in September 2021, they published a public safety alert to communicate concerns about the illicit counterfeit pill market (1, 2). A study in Washington DC also detailed an increase in counterfeit pills; the first detection of counterfeit pills was in April 2018, and counterfeit pills increased to 62.5% of all pill exhibits by 2021 (3). Finally, a recent analysis of data from the Center for Disease Control’s State Unintentional Drug Overdose Reporting System indicates the percentage of overdose deaths with evidence of counterfeit pills consumption more than doubled between the third quarter of 2019 and the fourth quarter of 2021 (4). Deaths involving counterfeit pills in NC Office of the Chief Medical Examiner (NC OCME) casework have also been increasing during this timeframe.

Counterfeit pills are frequently embossed with markings that are consistent with traditional pharmaceutical medications such as Xanax or Percocet and often advertised as such. These pills may in fact contain unexpected substances such as novel designer benzodiazepines (unapproved pharmaceutical drug candidates, structural modifications of approved drugs or metabolites of approved drugs), fentanyl and fentanyl analogs and/or other novel opioids. Historically, counterfeiting has been a problem with non-narcotic pharmaceutical medications intended to treat medical conditions such as osteoporosis, hypertension, high cholesterol and erectile dysfunction or diseases such as malaria, tuberculosis or human immunodeficiency virus/acquired immunodeficiency syndrome, particularly in poorer nations with less stringent control mechanisms and more informal medical or pharmaceutical infrastructure (5, 6). The occurrence of counterfeit narcotics and benzodiazepines in the USA is a more modern phenomenon, with ingredients often paralleling those found in the illicit powdered and crystalline drug supply chain (1).

Fentanyl started appearing in counterfeit pills around the same time it began gaining prominence in the illicit US opioid supply. The DEA reports seizing approximately 7,000 counterfeit oxycodone containing fentanyl and/or heroin in New England from early 2014 to late 2015 (7). Counterfeit Xanax, containing fentanyl and/or etizolam, was recorded from October through December 2015 in California; eight cases were identified by the poison control system, one of which resulted in death (8, 9). In April 2016, the National Institute of Drug Abuse issued a warning about counterfeit hydrocodone, oxycodone and alprazolam, containing fentanyl, that were being sold at lower prices than the legitimate pharmaceuticals (10).

Within a year or two of fentanyl’s appearance, fentanyl analogs and other novel opioids entered the counterfeit pill supply. One study reported a 2016 CA death from counterfeit oxycodone containing furanyl fentanyl (11), and a cluster of overdoses in June 2017 in GA were the result of counterfeit Percocet containing cyclopropyl fentanyl and U-47700 (12). Another case report from MA in 2017 demonstrated counterfeit Xanax that contained U-47700 (13). A Washington, DC study from April 2018 to June 2021 found the fentanyl analogs para-fluorofentanyl and acetylfentanyl in counterfeit oxycodone, along with numerous other substances (3).

Novel benzodiazepines began to appear around the same time as illicit fentanyl started to dominate the market. Etizolam was one of the first and most common illicit benzodiazepines, and its presence has remained widespread. In the USA, counterfeit Xanax containing etizolam was first described in 2016 (9). In December 2022, the DEA published an intent to place etizolam, flualprazolam, clonazolam, flubromazolam and diclazepam temporarily under Schedule 1 control, and on 26 July 2023, they were placed under temporary Schedule 1 control (14, 15). The US National Poison Data System has tracked etizolam exposures since 2007. However, tracking exposures to other novel benzodiazepines (including adinazolam, clonazolam, cloniprazepam, diclazepam, flubromazepam, flubromazolam, meclonazepam, nifoxipam, norflurazepam and pyrazolam) did not begin until 2016 (16). Flualprazolam was first documented by the DEA in 2018, and case reports indicated its presence in counterfeit Xanax in OR in June 2019 (17, 18).

Counterfeit benzodiazepine pills may contain other substances in addition to fentanyl or novel benzodiazepines, and concentrations of those substances can vary widely. One extremely comprehensive analysis of expected benzodiazepine street pills was conducted by a community-based harm reduction program offering drug-checking services in Canada from October 2017 to March 2020. Of 139 expected Xanax pills, >25% were negative for benzodiazepines and only 33 (23.7%) contained alprazolam. Seven pills (5.0%) tested positive for fentanyl via fentanyl test strip, and other substances found in the pills included etizolam, flubromazolam, cyproheptadine, promethazine, amantadine, AB-FUBINACA, 5-fluoro-MDMB-PINACA, U-47700 and “various fentanyl analogues” (19).

While direct tablet analysis is rarely done at NC OCME, pill contents can be often hypothesized from the combination of case history details and toxicology results. Initial scene information including local medical examiner calls to the medicolegal death investigators, toxicology request form information, the final report of investigation by the medical examiner and final postmortem toxicology results can be combined to create a more comprehensive picture of the threat to public health than that available with pill data alone. Often, counterfeit pill stories appear on the internet in the form of journalism or educational resources. The purpose of this study was to describe scientific analytical findings and forensic investigatory information surrounding a series of suspected counterfeit pill-involved deaths. This convenience sample from NC casework will add to the toxicological and epidemiological understanding of the counterfeit pill phenomenon.

Materials and methods

The NC OCME Toxicology Laboratory is accredited by the American Board of Forensic Toxicology. All postmortem toxicology cases included in this study were screened for common pharmaceuticals and drugs of abuse via liquid chromatography/orbitrap mass spectrometry in the parallel reaction monitoring mode with a targeted inclusion list. This list of 86 compounds included 16 common natural, synthetic and semi-synthetic opioids as well as 35 novel psychoactive opioids such as fentanyl analogs and U-compounds. It also included 12 common benzodiazepines, 6 stimulants such as methamphetamine and cocaine metabolite, 6 synthetic cannabinoids and 11 additional substances such as gabapentin, PCP, phenibut, tianeptine, zolpidem and tetrahydrozoline. When necessary to support further investigation of cause/manner of death, certain cases received additional testing via untargeted gas chromatography–mass spectrometry with nitrogen phosphorus detection. Specimens that screened entirely negative via these two techniques but had a strong case history suggestive of illicit drug use were subjected to an additional novel/emerging drug screen via orbitrap high-resolution mass spectrometry. Toxicologically significant positive substances from these screening techniques were generally confirmed via quantitative liquid chromatography/triple quadrupole mass spectrometry or, in the case of novel substances, qualitatively confirmed using liquid chromatography/orbitrap mass spectrometry in parallel reaction monitoring mode. Ethanol was screened and confirmed using gas chromatography with flame ionization detection. Full method details and targeted screen inclusion lists are available upon request, and some method details have been published previously (20, 21).

The NC OCME still employs a largely paper-based data system, with toxicology requests and medical examiner case information first submitted on handwritten forms and stored in manila folders. Initial information, demographic details, autopsy findings and final toxicology reports are stored in an in-house developed Medical Examiner Information System or filed at the respective regional autopsy center using the Medical Examiner Information System and/or their respective systems (both digital and paper-based). The postmortem toxicology laboratory serves all 100 counties in NC; therefore, toxicology data are maintained in Raleigh, NC, and stored in a separate Microsoft Access database serving as a laboratory information management system. The NC OCME Toxicology Unit received nearly 13,000 case submissions in 2020, nearly 15,000 in 2021 and just over 15,000 in 2022, for a total of almost 43,000 case submissions over this 3-year study period.

As manually searching approximately 43,000 decedent folders, either paper or electronic, is prohibitively labor intensive, a thorough comprehensive search for case histories indicating possible counterfeit pill presence could not be reasonably conducted. Instead, cases were designated to be appropriate for this case series whenever data and/or case history indicating possible involvement of counterfeit pills in a death were identified during toxicology test assignment, analytical batch review, administrative review or final toxicologist certification. Cases positive for rare novel benzodiazepines are routinely tracked by the laboratory in an Excel file for surveillance efforts and were retroactively investigated for a scene history indicative of counterfeit pill consumption after qualitative or quantitative identification. Hence, this case series is not meant to be complete for all 3 years but rather includes select examples of possible counterfeit pill-involved deaths. Cases were selected for inclusion if observed to have extremely strong case history, such as direct bystander observation of pill consumption, remaining pills observed or collected on scene and/or text message history or family/friend account detailing recent pharmaceutical pill purchase, without toxicological presence of the expected drug found in the decedent.

Results

Seventy-five postmortem cases occurring between January 2020 and December 2022 were identified for inclusion in the study, including 17 from the retroactive review of the novel benzodiazepine surveillance file. Brief case histories with expected and actual toxicology results from 25 select representative cases are displayed in Table I. Decedents ranged in age from 16 to 67 years, from 33 of NC’s 100 counties, and the majority of the decedents were white (68.0%), male (60.0%), and under age 40 years (74.7%). Further demographic information is available in Table II.

Table I.

Select postmortem case histories and toxicological results

Case No.	Age/sex/race/year	History and scene findings	Expected toxicology findings	Actual toxicology findings
1	43/M/W/2020	Per girlfriend they had been drinking, the decedent took an unknown number of oxycodone, they had sexual intercourse, and then he went to sleep. The girlfriend was in bed reading when she noticed that he was not breathing. No drugs or prescription bottles found on scene.	Oxycodone	Fentanyl 9.1 ng/mL (I) Ethanol 90 mg/dL (A) Doxylamine <0.25 mg/L (A) Lidocaine present (A)
2	40/M/B/2020	Decedent’s son reported that a friend gave him two 30 mg Roxicodone tablets on the date of death and he crushed and snorted them. Decedent found unresponsive on bathroom floor with snoring respirations. 7.6 cm straw on scene and white residue in both nares which tested positive for opiates.	Oxycodone	Fentanyl 3.6 ng/mL (F) Bupropion present (A) Threo bupropion <0.25 mg/L (A) Metoprolol present (A) Naloxone present (A)
3	39/M/W/2020	Decedent and friend went to get Xanax and heroin and snorted it. Decedent fell asleep supine on bed with legs over the side of the bed, found 7 hours later in the same position and unable to be resuscitated by emergency medical services.	Alprazolam Heroin	Fentanyl 53 ng/mL (F) Flualprazolam present (F) Methamphetamine <0.25 mg/L (F)
4	27/M/AI/2020	A friend went to the decedent’s house to do some work with him. The decedent offered the friend Xanax, warning him to take only ¼ of a bar, and then the decedent put 4 bars in a 590 ml bottle with soda. The friend who only took ¼ of a bar began feeling unwell and got someone else to take him to the emergency department where he was treated for an opiate overdose. The friend was discharged and went home to find the decedent unresponsive in his vehicle, ½ of the soda bottle now empty. The friend suspected that the Xanax was laced with fentanyl.	Alprazolam	Fentanyl 20 ng/mL (I) Methamphetamine 1.9 mg/L (I) Cocaine none detected (I) Benzoylecgonine 0.35 mg/L (I) Amphetamine 0.19 mg/L (I)
5	17/M/W/2020	Mother of the decedent found and got rid of Xanax which she stated she had done multiple times this year, stated son was loopy and out of it. Mother drove him to a friend’s house. Friend called mother and stated that his breathing became shallow and then stopped, cardiopulmonary resuscitation (CPR) was started and 911 called. Urine drug screen positive for benodiazepines, decedent took Xanax per reporting doctor.	Alprazolam	Flubromazolam present (AM S)
6	52/F/W/2020	Decedent suffered a fall 2–3 days prior to death, self-reportedly fell down stairs, with bruise on head and abrasion on thigh. Did not seek medical care but according to family did not “seem right” afterward. Decedent had Xanax on scene that was prescribed to her, but it was in excess of the fill amount. 40–50 oxycodone pills found in purse not prescribed to decedent.	Alprazolam Oxycodone	Flualprazolam present (F)a Oxycodone 0.030 mg/L (F)a Gabapentin 45 mg/L (F)a Trazodone <0.50 mg/L (F)a Isopropanol 22 mg/dL (F)a Methanol 150 mg/dL (F)a
7	24/M/A/2021	Found unresponsive on floor of secured bedroom. Decedent reportedly buys Xanax bars and other pills off the internet. Large amount of green/brown powder substance on the nightstand.	Alprazolam	Etizolam 0.079 mg/L (I) Mitragynine 0.12 mg/L (I)
8	21/F/W/2021	Found lying in bed, open bottle of lamotrigine, Xanax bars in plastic bag (no prescription) and rolled up dollar bill nearby.	Alprazolam	Etizolam 0.009 mg/L (F) Fentanyl 3.6 ng/mL (F) Codeine 0.094 mg/L (F) Benzoyecgonine 0.052 mg/L (F) Morphine present (F)
9	33/F/W/2021	Per decedent’s mother, a friend of the decedent came over and then the mother went to take a shower. When she got out, she found the decedent unresponsive on the floor next to the couch. Law enforcement found a text message from the decedent to the friend who was briefly at the house inquiring about buying some “roxy 30’s”. No drugs, paraphernalia or alcohol found near the decedent.	Oxycodone	Fentanyl 17 ng/mL (F) Ethanol 70 mg/dL (A) Naloxone present (A)
10	48/M/U/2021	The decedent and a friend each took ½ of a Percocet 10 before going to another’s house. Shortly after arrival, the decedent became unresponsive and fell to the floor, striking his head on a dresser on the way down. He was given naloxone and resuscitated before being transported to the hospital where he was later pronounced due to anoxic brain injury. While on scene investigating, law enforcement noted that the other friend who also took ½ a Percocet collapsed during the interview and also required treatment for overdose.	Oxycodone	Fentanyl 8.7 ng/mL (AM P) Ethanol 170 mg/dL (AM P) Naloxone present (AM P)
11	32/F/W/2021	Decedent found unresponsive in bed by husband who stated that she had taken an “oxy or Percocet” that she obtained illicitly to help her sleep. She does not have a prescription for fentanyl or gabapentin. She does have a prescription for clonazepam.	Oxycodone	Fentanyl 41 ng/mL (I) 7-Aminoclonazolam 0.050 mg/L (I) Gabapentin <1.0 mg/L (I)
12	17/M/B/2021	Decedent was at a party with friends where bystanders witnessed him taking up to 8 Xanax bars. He then went to Waffle House with friends, and they then left and drove back to another person’s house. The decedent was “passed out” in the back seat and was left there to sleep. Friends came to check on him the next morning and found him unresponsive in the back seat. Foam cone, foam and vomit on car floor.	Alprazolam	8-Aminoclonazolam present (F)
13	46/F/W/2021	Several straws found in the home, and a cut-up bar of alprazolam was found in the bathroom. NC OCME tested the “broken white substance, possible pill” and found it to be flualprazolam.	Alprazolam	Flualprazolam present (F) Buprenorphine 7.0 ng/mL (F) Norbuprenorphine 12 ng/mL (F) Gabapentin 13 mg/L (F) Mirtazapine 1.7 mg/L (F) Propranolol present (VC) Hydroxyzine present (VC)
14	34/M/B/2021	Per girlfriend, the decedent bought 1 Xanax pill from a friend and took 1/2–1 pill (story varied) before quickly becoming unresponsive.	Alprazolam	Fentanyl 3.1 ng/mL (F)
15	57/M/W/2022	Prescription bottle of Xanax prescribed to the decedent in his car, and another bottle of Xanax not prescribed to the decedent in the living room. Suicide note said took “22 Xanax”.	Alprazolam	Bromazolam present (I) Ethanol 20 mg/dL (A)
16	20/F/W/2022	Took homemade “roxie” with boyfriend, pretty confident it was fentanyl. Known alive at 0200 hours, found unresponsive in bed next morning.	Oxycodone	Fentanyl 23 mg/L (F) Fluorofentanyl present (S) Naloxone present (S)
17	46/F/W/2022	Found down on floor near door fully dressed. Per husband uses cocaine, has had heroin overdose in the past. Fell 3 days ago injuring her hip and has been taking hydrocodone that she has gotten from the street as she has no primary doctor and no prescription.	Hydrocodone	Fentanyl 1.9 ng/mL (F) Fluorofentanyl present (S) Cocaine 0.10 mg/L (F) Benzoylecgonine 0.79 mg/L (F) Gabapentin 2.4 mg/L (F) Naloxone present (S)
18	19/M/W/2022	Decedent’s brother gave decedent street Xanax the previous evening. Also gave him a “perc 30”. District attorney pursuing 2nd degree murder charge against brother.	Alprazolam Oxycodone	Fentanyl 13 ng/mL (F) 4-ANPP present (VC)
19	67/F/W/2022	Decedent seen in emergency room for fracture of right humeral head. Discharged home. Decedent took roommate’s Percocet for pain. Roommate found her apneic, unresponsive on bedroom floor.	Oxycodone	Tramadol 2.2 mg/L (F) O-desmethyltramadol 0.31 mg/L (F) Gabapentin 14 mg/L (F) Topiramate present (F)
20	40/M/W/2022	Reportedly shot up “roxies 30” in hand, went to shower, then had seizure-like activity. Nasal naloxone (4 mg) given by law enforcement. Emergency medical services worked code. Pronounced on scene. Blue tablets and bindle with unknown substance submitted as physical evidence. NC OCME tested one of the “blue tablets”, which had the “M” imprint characteristic of oxycodone, and found it to contain fentanyl.	Oxycodone	Fentanyl 7.1 ng/mL (F) Amphetamine 0.40 mg/L (F) Methamphetamine 8.7 mg/L (F) Naloxone present (A)
21	16/F/W/2022	Found deceased in bed at friend’s residence. Reportedly using Molly (MDMA) and Percocet for the past 2 days.	Oxycodone	Fentanyl 8.3 ng/mL (F) 3,4-MDA 0.083 mg/L (F) 3,4-MDMA 1.2 mg/L (F)
22	23/M/W/2022	Found unresponsive with girlfriend in bedroom by girlfriend’s grandmother. Girlfriend also dead. Multiple fake (street-made) Xanax pills discovered in the room that were suspected to be laced with fentanyl. Five were found in the girlfriend’s wallet, 6 in a baggie by the bed and 3 were in the decedent’s belongings. History of polysubstance abuse.	Alprazolam	Fentanyl 11 ng/mL (I) Naloxone present (VC)
23	19/F/W/2022	Found unresponsive by grandmother. Pronounced on the scene after CPR. Medical examiner stated that multiple fake (street-made) Xanax pills were discovered in the room that were suspected to be laced with fentanyl. Five pills were found in her wallet, 6 in a baggie by the bed, and 3 in the boyfriend’s belongings. History of polysubstance abuse.	Alprazolam	Fentanyl 32 ng/mL (I) Naloxone present (VC)
24	45/M/W/2022	At party with 5–6 others, witnessed to take loose pills, possibly Percocet. Found unresponsive about 45 minutes later. No foam cone. History of illicit Percocet abuse.	Oxycodone	Fentanyl 7.4 ng/mL (I) Ethanol 30 mg/dL (VC)
25	48/F/W/2022	History of peptic ulcers. Reportedly had stomach issues and obtained “roxi” pain pills from friend. Found unresponsive later.	Oxycodone	Fentanyl 22 ng/mL (F)

Bolded values are the substances of interest to this study. Abbreviations: VC, vena cava; I, iliac; A, aorta; F, femoral; S, subclavian; H, heart; Unk B, unknown blood; AM P, antemortem peripheral; AM S, antemortem serum; F, female; M, male; W, White; B, Black; AI, American Indian; A, Asian; U, unknown.

a

Specimen embalmed.

Table II.

Demographics of counterfeit pill-involved deaths, NC sample, 2020–2022, n = 75

	n	%
Sex
Male	46	60.0
Female	29	40.0
Age
<20	15	20.0
20–29	30	40.0
30–39	11	14.7
40–49	12	16.0
50–59	5	6.7
60–69	2	2.7
Race
White	51	68.0
Black	20	26.7
American Indian	2	2.7
Asian	1	1.3
Unknown	1	1.3
Year
2020	21	28.0
2021	16	21.3
2022	38	50.7

While the actual contents of many pills may be assumed from toxicology results, the NC OCME laboratory is almost never the recipient of extra pill material from death scenes for further analysis; therefore, pill contents cannot be definitively determined as part of this study. Per American Board of Forensic Toxicology accreditation standards, NC OCME analytical methods are only validated for the purpose of biological specimen analysis, so on the rare occasion that pill material is received, it is generally not tested unless necessary for research or to assist with directing further biological specimen testing for cause and manner of death determination. Results for pill material are not reported on the final toxicology report. Only two of the cases in this dataset included pill material submitted as evidence. One pill, associated with Case 13 (Table I), was tested as a research specimen and found to contain only flualprazolam. A second pill, associated with Case 20 (Table I), was tested and found to contain fentanyl. For the other cases, logical inferences are made.

Of the 75 total postmortem cases, 42 (56.0%) involved counterfeit opioids, all but one of which were described as a formulation of oxycodone (see Table III for a summary of case findings). Of these, none of the decedents had oxycodone present in their blood, and all but one (98.7%) tested positive for fentanyl (five decedents additionally tested positive for fluorofentanyl). The final opioid pill, described as “Percocet”, was hypothesized to contain tramadol due to toxicology results (Case 19 in Table I). Thirty-eight (50.7%) of the 75 cases involved an indication of alprazolam use via case history and/or appearance of pills found on scene. Of those 38 cases, 24 (63.2%) decedents tested positive for fentanyl, 10 (26.3%) tested positive for flualprazolam, 8 (21.1%) tested positive for etizolam, 4 (10.5%) tested positive for the metabolite of clonazolam (8-aminoclonazolam), 2 (5.3%) tested positive for bromazolam, and 1 each (2.6%) tested positive for flubromazolam, desalkylflurazepam and lorazepam. Twenty-two decedents had a simple case history indicative of only alprazolam consumption where only one likely substitute substance was found, as shown in Figure 1. Forty-one percent of these 22 alprazolam-only cases involved decedents that tested positive for fentanyl, while the rest involved decedents positive for a novel benzodiazepine.

Figure 1.

Pie chart showing percentage breakdown by drug of postmortem cases with history and/or scene investigation indicative of counterfeit alprazolam consumption where only one likely substance was found.

Table III.

Summary of case findings, counterfeit pill-involved deaths, NC sample, 2020–2022 (n = 75)

	n	%	Example case(s) from Table I
Pill description/supposed substancea
Xanax	31	41.3	3–8, 12, 14, 15, 18
Percocet/Perc 30	15	20.0	10, 18, 19, 21, 24
Blue pills stamped with M/M30	9	12.0	Not included
Oxycodone	8	10.7	1
Roxicodone/Roxi/Roxy 30	6	8.0	2, 9, 20, 25
Alprazolam/Alprazolam bar	4	5.3	13
Fake/street Xanax	3	4.0	18, 22, 23
Homemade Roxie (Roxicodone)	1	1.3	16
Dirty 30	1	1.3	Not included
“Oxy or Percocet”	1	1.3	11
Hydrocodone	1	1.3	17
Actual toxicology findingsa
Fentanyl	60	80.0	1–4, 8–11, 14, 16–18, 20–25
Flualprazolam	10	13.3	3, 6, 13
Etizolam	8	10.7	7, 8
Fluorofentanyl	5	6.7	16, 17
8-Aminoclonazolam	4	5.3	12
Bromazolam	2	2.7	15
Flubromazolam	1	1.3	5
Desalkylflurazepam	1	1.3	Not included
Lorazepam	1	1.3	Not included
Tramadol	1	1.3	19
Method for obtaining pills
Personal relationships	8	10.7	2, 4, 9, 14, 18, 19, 25
“The Street”	6	8.0	17
Online	1	1.3	7
Unknown	69	92.0	1, 3, 5–6, 8, 10–13, 15–16, 20–24
Scene findings/witness testimony
Witness testimony of consumption	28	37.3	1–4, 10–12, 14, 16, 19–21, 24
Suspected laced/ fake pills	7	9.3	4, 26, 22, 23
Surviving acquaintance	6	8.0	3, 4, 10, 16
Likely oral consumption	21	28.0	1, 4, 10–12, 14–17, 19
Likely insufflation	15	20.0	2, 3, 8, 13
Likely injection	4	5.3	20
Consumption motivation
Pain	3	4.0	17, 19, 25
Recreation—party setting	2	2.7	12, 24
Sleep	1	1.3	11
Suicide	1	1.3	15
Divorce response	1	1.3	Not included

a

Categories are not mutually exclusive, so numbers will not sum to 75/100%.

Seventeen (22.7%) of the 75 cases contained a combination of both benzodiazepines and opiates in the case history and/or toxicology results. For those cases, an assumed link between pill contents and toxicology results is impossible—it is unknown which of the drug products contained which of the substances found in the toxicology results. For instance, as described in Table I, Case 3 documented that the decedent bought both Xanax and heroin and tested positive in postmortem blood for both fentanyl and flualprazolam. While the substance reported as heroin likely contained fentanyl, the counterfeit Xanax could also have contained fentanyl. Likewise, flualprazolam could have been present within the counterfeit Xanax and/or the supposed heroin. In Case 8, alprazolam was the only substance on scene described by the medical examiner, but the decedent tested positive for both etizolam and fentanyl. The counterfeit alprazolam may have contained both, or one of the substances may have been separately consumed. In Case 18, the decedent was described as receiving both a perc 30 and a street Xanax, but only fentanyl was found. It is possible that both the alleged oxycodone and the alleged alprazolam contained fentanyl.

Eight (10.7%) of the suspected counterfeit pills in the study were reported to have been obtained through personal relationships such as friends, family or roommates, and six (8.0%) were reportedly obtained from “the street” (examples referenced in Table III). Two decedents had medication bottles on scene prescribed to them for the relevant substances. In Table I, Case 6, the decedent had a bottle that indicated a prescription for Xanax, but it contained a quantity of pills in excess of the number indicated by the prescription; the toxicology results suggested that some of those pills may have been counterfeit. The decedent in Case 15 had a bottle of Xanax prescribed to him in his car but also a bottle of Xanax not prescribed to him on scene. As the pills on scene were not tested, it is not known if any of the alleged Xanax present was legitimate product. It is not uncommon for a patient to consume prescription pills ahead of schedule and then seek additional medication through other means, resulting in multiple medication sources. Only one decedent (Case 7) had a case history indicating the prior purchase of pills, including Xanax, from internet sources.

Of the 75 cases, 28 (37.3%) involved direct witness testimony of pharmaceutical pill consumption, although two decedents or witnesses seemed to suspect that their medications were “laced”, one indicated that the pill was “homemade” and likely fentanyl, and four had case histories indicating that the pills appeared fake. Six cases involved acquaintances that took the same medication as the decedent and survived, although the acquaintances in Table I, Cases 4 and 10 required medical assistance. Twenty-one of the case histories indicated oral consumption, 15 cases were suggestive of insufflation, and 4 had evidence of injection, indicating typical routes of administration for substances in pill form.

Discussion

The danger of encountering illicit counterfeit medication is not a new phenomenon, but recorded instances are increasing, and counterfeit pill contents are shifting. Substances reported in known counterfeit pills in other areas of the USA between 2016 and 2017 included fentanyl analogs such as cyclopropyl fentanyl, furanyl fentanyl and U-47700, all of which first appeared at NC OCME in 2017 and have not been seen since 2019. During that timeframe in NC, 390 total cases tested positive for cyclopropyl fentanyl, 121 cases tested positive for furanyl fentanyl and 101 cases tested positive for U-47700, indicating a significant level of decedent positivity for these fentanyl analogs and novel opioids. Since that time, these three substances and similar fentanyl analogs, with the exception of para-fluorofentanyl, have largely disappeared from postmortem toxicology samples in NC. The substances found in suspected counterfeit pill-involved deaths appear to mimic those present in the larger drug supply in the USA over time, and that trend continues in this study today.

In contrast, the novel benzodiazepines detected in this dataset have all surfaced in NC within the last 7 years and many continue to be detected in deaths currently. Etizolam and clonazolam/8-aminoclonazolam were first detected at NC OCME in 2017, desalkylflurazepam was first detected in 2019, flualprazolam and flubromazolam were first detected at NC OCME in 2020 and bromazolam was first detected in 2021. Etizolam, 8-aminoclonazolam and bromazolam have all been detected recently in 2023.

Due to resource and staffing limitations, comprehensive testing for novel benzodiazepines is not routinely done in all suspected overdose cases by NC OCME at this time. Targeted testing for current novel substances including benzodiazepines is done if novel benzodiazepines are suspected or if no other lethal substances are detected during regular testing, but a drug-involved death is indicated. This means that comprehensive surveillance statistics for novel benzodiazepines in NC decedents are not available. The limitations of government forensic laboratories are important to recognize when attempting to glean epidemiological, population-level insights from regions not yet supported for comprehensive surveillance testing in postmortem casework.

Specifically, the epidemiological utility of this dataset is limited by the conservative approach to testing. Ideally, physical pill evidence would be submitted and tested to compare with postmortem results. The system currently lacks sufficient scene investigators to be deployed for all drug deaths, physical storage space for increased pill evidence retention and staff for pill or powder analysis method development and validation. The authors would prefer to do surveillance testing for novel opioids and benzodiazepines for all decedents and to quantify the novel substances that are found. However, current mission and staffing does not provide the resources necessary for this quantity of testing and method development. Publicly funded government postmortem toxicology laboratories could be an incredible resource for understanding polysubstance use trends including drug combination patterns and elucidating related public health hazards. They could be used for drug surveillance and the identification of emerging drug-related dangers if appropriately funded and staffed for more than cause and manner of death determination (22). In order for NC postmortem data collection to expand, the data must be valued and its collection must be materially supported.

A wide variety of reasoning has been proposed for the increasing prevalence of counterfeit benzodiazepines and opioid pills, and likely a complex set of factors is at play. In general, states have limited the prescribing of pharmaceutical opioids and both doctors and prescriptions are being more tightly monitored (23). In NC, the Strengthen Opioid Misuse Prevention Act took effect on 1 January 2018 instituting 5- and 7-day prescribing limits following initial consultation and treatment of acute pain (24). It is well established that the reformulation of legitimate pharmaceutical opioids in the past has resulted in major drug market shifts and an influx of synthetic opioids in order to meet the demand (25, 26). More recently, rogue online pharmacies and sales via social media have proliferated, and individuals have reported accessing them in order to self-medicate a variety of health conditions (such as chronic pain, depression, generalized anxiety, post-traumatic stress disorder or substance use disorder) due to lack of health insurance or healthcare cost, difficulty accessing physical and mental healthcare, high medication costs, feeling that their doctor did not prescribe them enough medication or the desire for anonymity (27, 28) in addition to generally understood dependence or desires for recreation. A 2021 survey of people who use drugs reported that 40% of participants who were denied prescription analgesia by physicians acquired diverted prescription opioids and 33% acquired illicit opioids (29). While the motivations of many of the NC OCME decedents are not known, the dataset contains evidence of pill consumption in response to recent divorce, to deal with pain, to help sleep, for recreation at a party and for suicide.

The level of awareness that consumers have about the counterfeit nature of their pills may vary geographically or socially and has likely changed over time. On the west coast of the USA, vaporization of counterfeit oxycodone containing fentanyl among regular opioid users has been highlighted as a behavior shift with the potential to reduce overdose risk when compared to injection (30, 31). The Center for Disease Control’s recent analysis of State Unintentional Drug Overdose Reporting System data offers further evidence of the presence of this behavior shift, as >50% of counterfeit pill-involved deaths in Western jurisdictions contained evidence of smoking (4). While pharmaceutical pill vaporization has not been noted within this 2020–2022 NC death dataset, case records provide data congruent with the hypothesis that some decedents were nonetheless aware of pill contents. In Case 16, the boyfriend of the decedent indicated that they were fairly confident that their pills were fentanyl, indicating known consumption of counterfeits.

A 2016 commentary letter theorized that counterfeit Xanax containing 3 to 4 mg of fentanyl may have resulted from mislabeling or confusion in the supply chain, as this is a fentanyl potency much higher than that needed for recreation yet matching the 3 to 4 mg range of alprazolam often found in legitimate tablets (32). The creation of pills containing substances of higher potency that are expressly intended to be consumed piecemeal as would be facilitated by the four-segment Xanax bar design (i.e., the consumer intends to purchase this specific, counterfeit product) was also stated to be a possibility (32). Evidence for this theory may exist in Case 4 in Table I, as it was reported that the decedent offered a friend a Xanax but advised him to only take one-fourth of the pill.

Individuals unaware of the counterfeit nature of the pills may prefer prescription opiates to powders due to a desire for lower overdose risk from known contents and potency (29) as well as the greater stigma associated with powdered drug use. Counterfeit benzodiazepines that contain fentanyl, such as in Cases 4, 14, 22 and 23 in Table I, pose a unique risk, as experienced benzodiazepine users may still be opioid naïve. Even when counterfeit benzodiazepines contain a substance within the benzodiazepine class, they may contain the expected substance at an unexpected concentration or a novel benzodiazepine with both an unknown potency and an unknown concentration. User tolerance may vary widely by body type, genetics or use history, and the risk of overdose increases when a drug consumer is unable to accurately dose themselves in relation to their tolerance. Risk is also increased for naïve drug users who may be less skilled in safely navigating illicit drug markets, more trusting of the substances they are purchasing or less familiar with harm reduction strategies such as fentanyl test strips or drug checking. Public health messaging is needed to increase awareness of the prevalence and dangers of counterfeits and to continue to educate the public, especially those who use drugs, on harm reduction strategies.

As the proliferation of fake pills shows no signs of slowing, strategies to mitigate harm are needed. Illegitimate internet pharmacies and social media sellers are often very hard to physically find or shut down, particularly when located in other countries (28). Due to the inspection and safety systems that exist for regulated drug markets, pharmaceuticals obtained directly from legitimate pharmacies will continue to be by far the safest option. For the unregulated drug market, sanctioned drug-checking services, overdose prevention sites and naloxone distribution are becoming more widespread as strategies to combat overdose, although naloxone cannot reverse the effects of benzodiazepines. The support of law enforcement in stigma reduction and promotion of these harm reduction strategies cannot be overstated. Additionally, law enforcement at the national, state and local levels regularly seize and test drugs for criminal investigations. Documentation of the substances found during this routine testing and sharing of these data with public health officials and postmortem toxicology labs may give additional, and perhaps early, insights into changing trends in the illicit drug market.

Conclusion

This convenience sample of 75 postmortem cases from the NC OCME serves to describe personalized characteristics and toxicological data from suspected counterfeit pill-involved deaths in 2020–2022. Alprazolam, oxycodone and in one case hydrocodone were the pharmaceutical products presumed counterfeited in this dataset, due to their reported consumption in the scene investigation and subsequent lack of detection in toxicological analysis. Cases with histories indicative of counterfeit oxycodone consumption involved postmortem toxicology results positive for fentanyl in all but one case, with the final case positive for tramadol. Investigatory information indicating possible counterfeit alprazolam consumption involved more varied postmortem findings including fentanyl and the novel benzodiazepines flualprazolam, 8-aminoclonazolam (the metabolite of clonazolam), etizolam, desalkylflurazepam, bromazolam and flubromazolam. The decedents spanned a wide range of ages and motivations for drug use, and routes of administration were typically oral ingestion or nasal insufflation. Medical examiner information indicated that decedents obtained counterfeit medications from direct interpersonal relationships, “the street” and online, and some case histories indicated that individuals may have been knowledgeable or suspicious of the counterfeit nature of their pills.

Medical examiners, death investigators, law enforcement and pathologists should be aware of the possibility of counterfeit pills so that the relevant evidence can be documented or collected beginning at the death scene. This should include detailed information about drugs and/or paraphernalia found on scene, which can yield invaluable epidemiological information for understanding death trends. Scene information indicating substance type or class can help direct targeted toxicological analysis in an environment of limited resources, where expanded toxicological surveillance testing is not possible for every decedent. Instances of counterfeit pill-involved deaths will be under-counted in forensic casework when pill material, evidence of purchase or bystander accounts are lacking. Efficient and timely sharing of information between law enforcement, medical professionals, forensic practitioners, public health leaders and people who use drugs will continue to be critical for guiding policy and reducing the harms associated with drug use.

Data availability

The data underlying this article were accessed from the NC OCME Toxicology Laboratory. A subset of the data is available here, and the full dataset is available upon request.

Funding

Our office has no research funding support and no conflicts of interest to declare.

Acknowledgments

We greatly appreciate all the hard work contributed by the staff of the NC OCME, especially the production team in the Toxicology Laboratory. Thank you as well to Dr Jennifer J. Carroll for her editing assistance.