Research paper
Toxicosurveillance in the US opioid epidemic

https://doi.org/10.1016/j.drugpo.2017.05.057Get rights and content

Introduction

The opioid epidemic presents an ever-changing epidemiological, clinical, and forensic landscape that poses a significant challenge to regulators and legislators. Heroin is a long-standing contributor to opioid related death, but its use and consequences have markedly escalated over the past 20 years. US national surveillance mortality data suggests heroin deaths have more than quintupled since 2000, with 10,574 deaths reported in 2014 (Compton, Jones, & Baldwin, 2016). Changes in the heroin marketplace including increased purity, increased availability, and adulteration with synthetic opioids, specifically fentanyl and its congeners, contribute to its toxicity.

Although historically a drug used primarily in lower socioeconomic minority areas, heroin is increasingly abused across all demographic groups (Unick & Ciccarone, 2017). As a result, between 2002 and 2013, heroin overdose death rates in the US nearly quadrupled, from 0.7 deaths to 2.7 deaths per 100,000 population (Hedegaard, Chen, & Warner, 2015), totaling more than 8200 deaths in 2013 (Centers for Disease Control and Prevention, July 7, 2015). Between 2013 and 2014, age-adjusted rate of death increased by 26% (Rudd, Aleshire, Zibbell, & Gladden, 2016) and between 2014 and 2015, heroin overdose death rates increased by 20.6%, causing nearly 13,000 deaths in 2015 (Centers for Disease Control and Prevention, January 26, 2017). In 2015, surveillance through the National Survey on Drug Use and Health (NSDUH) noted that 828,000 people aged 12 or older reported heroin use within the past year. Though slightly decreased from 2014 data, this represents an increase of 82% since 2013 (Center for Behavioral Health Statistics and Quality, 2015, 2016). NSDUH, which is administrated through the Substance Abuse and Mental Health Services Administration (SAMHSA), involves in-person interviews with approximately 70,000 households regarding personal drug use habits (National Survey on Drug Use and Health (NSDUH)). Despite the methodological limitation of reporting bias, this finding is concerning because, not surprisingly, rates of heroin use and dependence have a strong positive correlation with heroin-related deaths (Jones, Logan, Gladden, & Bohm, 2015).

Heroin overdose deaths have increased in many regions across the US, specifically the Northeast and areas of the Midwest. The National Drug Threat Assessment Survey of state, local, and tribal law enforcement agencies assessed the greatest reported drug threat by jurisdiction with respect to availability, demand, transportation, and distribution. In 2007, only 8% of respondents identified heroin as the largest drug threat in their jurisdiction. By 2016, heroin led all other drugs, with 45% of respondents identifying it as the largest drug threat (US Drug Enforcement Administration Intelligence Report, June 2016). Also, nine of the 21 domestic Drug Enforcement Administration (DEA) Field Divisions ranked heroin as the largest drug threat, with nine more divisions ranking it as the second largest threat (US Drug Enforcement Administration Intelligence Report, June 2016). The distribution of heroin throughout the US is also changing. Historically, the Mississippi River was the transition zone that divided heroin by type and origin. To its west, heroin was predominantly Mexican black tar and brown powder heroin, whereas eastern markets received Asian and South American derived white powder heroin (Ciccarone, Unick, & Kraus, 2009). Mexican traffickers have expanded their markets and are gaining a larger share of the Northeast and Midwest markets (US Drug Enforcement Administration Intelligence Report, June 2016).

There has been a significant change in the demographics of heroin use over the past two decades. Previously, in 2000, non-Hispanic black persons aged 45–64 had the highest rates of poisoning deaths involving heroin (2 per 100,000 population). However, the demographic balance has changed with the movement of heroin into non-urban communities along with use by younger people. Heroin use has more than doubled in the past 10 years among those aged 18–25 years (Centers for Disease Control and Prevention, July 7, 2015). In 2013, non-Hispanic white persons aged 18–44 had the highest rates of heroin-related death at 7.0 per 100,000 population (Hedegaard et al., 2015). This trend continued in 2015 when males aged 25–44 had the highest rates of death from heroin at 13.2 per 100,000 population, an increase of 22.2% from 2014 (Centers for Disease Control and Prevention, January 26, 2017). Increased use is also being seen among groups with historically low rates of heroin use, including women, the privately insured, and those with higher incomes (Centers for Disease Control and Prevention, January 26, 2017).

The heroin overdose epidemic is intertwined with the larger prescription opioid overdose epidemic (Unick, Rosenblum, Mars, & Ciccarone, 2013). Patients who abuse or who are dependent on prescription opioids are 40 times more likely to abuse or become dependent on heroin compared to those who do not use opioids for non-medical reasons (Jones et al., 2015). Among recent cohorts of heroin users entering treatment, nearly 75% report having abused prescription pain relievers within the past year (Centers for Disease Control and Prevention, January 26, 2017; Hughes et al., 2016, September). Another study reports that heroin users are 3.9 times as likely to report non-medical use of opioids in the previous year, and 2.9 times as likely to meet the criteria for abuse or dependence on opioids compared to persons who did not use heroin (Becker, Sullivan, Tetrault, Desai, & Fiellin, 2008).

Although only 4% of prescription opioid users are expected to become heroin users, the substantial number of prescription drug misusers (3.8 million with past month misuse of prescription pain relievers (Center for Behavioral Health Statistics and Quality, 2016)), highlight the significance (Compton et al., 2016, Jones et al., 2015). Between 2010 and 2012, the death rate from heroin overdose in 28 states more than doubled from 1.0 to 2.1 per 100,000 population (increase from 1779 to 3635 deaths) whereas deaths from prescription opioid pain relievers slightly decreased from 6.0 to 5.6 per 100,000 population. For prescription drug abusers who are used to using prescription opioids with known constituents and concentrations, the use of heroin with its unpredictable purity and potential for adulteration creates significant problems. Not surprisingly, the risks of death related to prescription opioid misuse compared to heroin use are not the same. In 2014, 10.3 million people used prescription pain relievers non-medically as opposed to 914,000 people who used heroin (Center for Behavioral Health Statistics and Quality, 2015). Despite a greater than 10-fold difference in number of users, the risk of death from heroin is much greater. In fact, of the 18,893 overdose deaths related to opioid analgesics in 2014, 10,574 were attributable to heroin (Compton et al., 2016).

Factors likely associated with the increased use of heroin include decreased price compared to prescription opioids and higher purity (Mars, Bourgois, Karandinos, Montero, & Ciccarone, 2014). The US DEA reports that the amount of heroin confiscated at the southwest border of the US as ≤500 kg annually during the years 2000–2008. This amount more than quadrupled to 2196 kg in 2013 (Centers for Disease Control and Prevention, January 26, 2017). Across the US, the National Seizure System reports that heroin seizures have increased by 80% over 5 years to 6722 kg in 2015 (US Drug Enforcement Administration Intelligence Report, June 2016). Heroin traffickers are also transporting heroin in much larger amounts than previously.

The most substantive factor contributing to the rising recent death rates related to heroin is its adulteration with synthetic opioids, specifically fentanyl and its analogs. Fentanyl is about 100 times more potent than morphine and 30–50 times more potent than heroin. Historically, fentanyl and fentanyl analogs were often combined with or sold as heroin, with or without the users’ knowledge, resulting in microepidemics. For example, during a 3-month period from September to November 1988, Allegheny County, PA reported a 13-fold increase in opioid overdoses due to “China White” (adulterated with 3‐methyl fentanyl) (Martin et al., 1991). In 2006, increased reports of overdoses in multiple different US states prompted the CDC to form an ad-hoc case finding and surveillance system to identify overdoses related to non-pharmaceutical fentanyl (NPF) in six states and local jurisdictions (Centers for Disease Control and Prevention, 2008). Between April 4, 2005 and March 28, 2007, they identified 1013 deaths related to NPF, leading to a DEA raid and closure of a main NPF production facility in Toluca, Mexico in May 2006 and regulation of N-phenethyl-4-piperidone, a chemical precursor used in the production of NPF in April 2007 (Centers for Disease Control and Prevention, 2008).

Though these steps taken were effective at curbing NPF in the mid-2000s, it was accurately predicted that given the ease of production and availability of obtaining precursors for production, NPF was likely to be a recurring problem. In late 2013 and 2014, the DEA National Heroin Threat Assessment Summary noted spikes in overdose deaths related to fentanyl and its analog, acetyl-fentanyl, most predominantly in the eastern US, where white powder heroin is used (Jones et al., 2015; US Drug Enforcement Administration Intelligence Report, June 2016). NPF was often mixed with white powder heroin or sold disguised as heroin. In March 2015, the US DEA issued a nationwide alert regarding the dangers of fentanyl and fentanyl analogs and identified them as significant threats to public health and safety (US Drug Enforcement Administration, March 2015). Between 2012 and 2014, the National Forensic Laboratory Information System (NFLIS) reported a more than 8-fold increase in the total number of fentanyl drug seizures (CDC Health Alert Network, October 2015). The NFLIS, part of the US DEA, systematically collects and analyzes the results from law enforcement chemistry laboratories across the country (Drug Enforcement Administration & Diversion Control Division). The observation that increased numbers of drug seizures positively correlates with opioid overdose deaths has been reemphasized in 27 states in the US. An increase in drug products obtained by law enforcement that tested positive for fentanyl increased by 426% (from 1015 in 2013 to 5343 in 2014) and deaths from synthetic opioids increased 79% (from 3105 in 2013 to 5544 in 2014) over the same time period (including fentanyl and other synthetic opioids such as tramadol—although these numbers are largely driven by NPF as identified by supplementary data in eight high burden states) (Gladden, Martinez, & Seth, 2016). Law enforcement encounters where the drug is submitted for analysis that have tested positive for NPF have sharply increased between 2014 and 2015, from 5343 to 13,882 (Centers for Disease Control and Prevention, August 24, 2016). In addition, since 2014 US law enforcement agencies have been seizing more ‘counterfeit medications’ containing fentanyls (US Drug Enforcement Administration Intelligence Report, July 2016). This further supports that NPF represents an ever-growing issue.

A recent study assessed samples believed to be heroin obtained in 2015 from cryptomarkets of the dark web and found the presence of ocfentanil, a short acting opioid with roughly the same potency as fentanyl (Quintana et al., 2016). Previous fentanyl adulterated heroin was localized to specific regions, however, expansion into the dark web creates the possibility of widespread exposure. Synthetic opioids represent a moving target, not unlike that for the synthetic cannabinoids and cathinones, as new compounds are continuously added to the marketplace.

There are many challenges encountered with attempting to accurately monitor the incidence rates and toxicity from heroin and heroin adulterants. National statistics are derived largely from death certificate data and there exists significant state variation in determining manner of death and the specific drugs involved in drug intoxication deaths (Warner, Paulozzi, Nolte, Davis, & Nelson, 2013). In one study comparing different types of medical examiner systems, states with a centralized medical examiner system specified a drug involved in more than 90% of drug intoxication deaths, versus only 62% in states with a decentralized county coroner system (Warner et al., 2013). This difference in drug identification creates difficulties for establishing informed public health surveillance and focusing preventative efforts. Even for states in which drugs involved are specified following drug intoxication deaths, the validity of the reporting is sometimes questionable. A review of unintentional overdose deaths listed as “morphine” in Allegheny County, PA from 2008 to 2010 suggested that heroin deaths were underrepresented in the group (Mertz, Janssen, & Williams, 2014). Heroin overdose deaths may be mislabeled as morphine-related since heroin is rapidly metabolized to 6-monoacetylmorphine (6-MAM) then to morphine, and only morphine may remain to be identified post mortem. Reevaluation of the 112 deaths that listed morphine, but not heroin, on the death certificate suggested that two-thirds met the heroin classification criteria based on the patient drug history, evidence from the death scene, and expanded testing for 6-MAM (Mertz et al., 2014).

Given the inconsistency in death certification related to heroin and other opioids and the implications for public health surveillance and preventative strategies, the American College of Medical Toxicology and the National Association of Medical Examiners made joint recommendations for the investigation, diagnosis, and certification of deaths related to opioids. Strict adherence to these guidelines, which suggest a complete autopsy that is performed in consideration with the circumstances surrounding death, the medical history, and scene findings, could improve the accuracy and precision of death certificate data (Davis, 2014). Regarding heroin specifically, detection of 6-MAM alone can ascribe the intoxication to heroin because humans cannot acetylate morphine to create 6-MAM. In the absence of 6-MAM, since both heroin and codeine are metabolized to morphine, interpretation of serum concentrations can be complex. The use of blood morphine:codeine ratio can help to distinguish heroin from morphine or codeine (Davis, 2014). Determination of a morphine:codeine ratio greater than 1 is indicative of heroin use rather than morphine or codeine (Ceder and Jones, 2001, Jones and Holmgren, 2011).

Another difficulty complicating toxicosurveillance involves confirmatory opioid testing in patients presenting to the emergency department (ED) with opioid toxicity. Due to the logistics of clinical laboratory testing for substances not directly related to patient care, advanced analytical testing in patients whose presentation is consistent with simple opioid intoxication is not performed. The cost of utilizing a reference laboratory for such testing is prohibitive and only performed in select cases, generally those of academic or public health interest. This represents a huge loss of surveillance data, since the vast majority of patients who overdose on opioids and present for healthcare do not die, yet analytical testing is more routinely performed by medical examiners.

Although local health departments and federal agencies actively track hospital ED visits, such as through National Hospital Ambulatory Medical Care Survey (NHAMCS) (The National Bureau of Economic Research, April 11, 2017), and poison centers rely on passive case reporting, the data they collect on patients with opioid exposures and overdose are very limited in granularity. In order for health departments, EDs, and poison centers to track these patients and to identify the incidence and prevalence of specific opioids, emergency providers would have to judiciously send serum and urine samples for testing. However, even if samples are reliably sent, there are multiple shortcomings with routine urine drug screens that use morphine-specific immunoassays. Although hospital laboratories can generally assay urine for morphine-based opioids, and would be positive in the setting of heroin use, these assays do not detect fentanyl and its analogs or many of the prescription opioids (e.g., oxycodone) (Moeller, Kissack, Atayee, & Lee, 2017). Detection of synthetic opioids generally requires the use of more advanced immunoassays that are not readily available at most institutions. During outbreaks, the health department or law enforcement may procure the necessary advanced analytical testing through a forensic or reference laboratory.

Even with advanced analytical testing, as novel opioids are introduced into the drug market, their presence remains unknown until they can be identified and added to the screening library. An investigation of non-targeted screening for novel psychoactive substances in patients presenting to an ED due to substance abuse noted this approach was effective at identifying novel psychoactive substances, including opioids, which were confirmed by acquiring the reference standard (Lung, Wilson, Chatenet, LaCroix, & Gerona, 2016). This prospective identification of novel psychoactive substances may be useful to help identify new and emerging synthetic opioids.

Currently, many patients that present with toxicity related to opioids are managed according to local guidelines and the offending opioid is never identified. Suspicion for the presence of fentanyl, fentanyl analogs, or a novel opioid should be increased if an experienced drug user concedes that their drug experience was somehow different from normal (Armenian et al., 2017). In addition, a significant increase in the number of opioid overdoses, especially if routine urine drug screens are negative for opioids, suggests that fentanyl, a fentanyl analog, or a novel opioid may be implicated (Martin et al., 1991).

How can we prevent heroin’s contribution to the opioid public health crisis from worsening? A multi-faceted approach that includes harm reduction, opioid substitution programs, increased availability of out-of-hospital naloxone, and improved toxico-surveillance is required to address the worsening opioid epidemic. Some regions, including Vancouver, have also adopted a controversial harm reduction strategy that includes prescribing heroin to users so that it can be used safely at supervised injection sites (Gartry, Oviedo-Joekes, Laliberte, & Schechter, 2009; Oviedo-Joekes et al., 2008). This will help to assure heroin is pure and unadultered with fentanyl and other compounds. Increased availability and access to substance abuse treatment centers and opioid substitution programs can also help to lessen some of the harms seen with heroin abuse. Preventing opioid use disorder through judicious opioid prescribing and careful clinical monitoring of those receiving opioid pain therapy may be the optimal, though incompletely effective, solution. In addition, improved detection and reporting by recruiting surveillance through poison centers, local EDs, medical examiners, and law enforcement may help to further define the scope of the problem.

Section snippets

Declaration of interest

The authors have no conflicts of interest to disclose.

First page preview

First page preview
Click to open first page preview

References (36)

  • Centers for Disease Control and Prevention

    Nonpharmaceutical fentanyl-related deaths—multiple states, April 2005–March 2007

    MMWR: Morbidity and Mortality Weekly Report

    (2008)
  • Centers for Disease Control and Prevention

    Reported law enforcement encounters testing positive for fentanyl increase across US

    (2016)
  • Centers for Disease Control and Prevention

    Injury prevention and control: Opioid overdose; heroin overdose data

    (2017)
  • Centers for Disease Control and Prevention

    Vital signs: Today's heroin epidemic

    (2015)
  • W.M. Compton et al.

    Relationship between nonmedical prescription-opioid use and heroin use

    New England Journal of Medicine

    (2016)
  • G.G. Davis

    Complete republication: National Association of Medical Examiners position paper: Recommendations for the investigation, diagnosis, and certification of deaths related to opioid drugs

    Journal of Medical Toxicology

    (2014)
  • Drug Enforcement Administration, & Diversion Control Division. National Forensic Laboratory Information System (NFLIS)....
  • C.C. Gartry et al.

    NAOMI: The trials and tribulations of implementing a heroin assisted treatment study in North America

    Harm Reduction Journal

    (2009)
  • Cited by (11)

    • Strengths and weaknesses of existing data sources to support research to address the opioids crisis

      2020, Preventive Medicine Reports
      Citation Excerpt :

      While national data on drug seizures, drug testing, and illicit drug prices that could be used to examine trends and geographic variation in illicit opioid markets exist in the National Forensic Laboratory Information System (NFLIS) or System to Retrieve Information from Drug Evidence (STRIDE) (National Academies of Sciences Engineering and Medicine, 2017; Secora et al., 2014), we identified few empirical analyses using these measures (Rosenblum et al., 2014; Stein et al., 2015a), and found local or state law enforcement databases to be more common sources of drug seizures and arrest data (Bujold et al., 2012; Piper et al., 2016; Ray et al., 2017). While mortality microdata help monitor drug overdose mortality and polysubstance involvement in fatal overdose (Jalal et al., 2018; Kandel et al., 2017), concerns about its use for public health surveillance have been raised due to state variation in procedures used by medical examiners and coroners to record manner of death and specific drugs involved in overdoses (Davis et al., 2014b; Lucyk and Nelson, 2017; Ruhm, 2017, 2018; Warner et al., 2013). Alternative data sources that have been used to examine trends, geographic “hot spots,” and product-specific characteristics for opioid-related overdose include Drug Abuse Warning Network (DAWN) emergency department data (Bau et al., 2016; Secora et al., 2017), opioid-related toxic exposures through RADARS or the National Poison Data System (NPDS) (Bau et al., 2016; Coplan et al., 2016; Coplan et al., 2013; Davis et al., 2014b; Mowry et al., 2016), detailed and timely information on fatal and nonfatal overdose through the Enhanced State Opioid Overdose Surveillance (ESOOS)/State Unintentional Drug Overdose Reporting System (SUDORS) (Mattson et al., 2018; Seth et al., 2018; Vivolo-Kantor et al., 2018), and information about opioid-related overdose from state hospital discharge databases (Cerda et al., 2017) or emergency department syndromic surveillance systems (Albert et al., 2011; Daly et al., 2017; Tomassoni et al., 2017).

    • The opioid epidemic and men’s sexual health

      2019, Effects of Lifestyle on Men’s Health
    View all citing articles on Scopus
    View full text