Understanding the Increase in Opioid Overdoses in New Hampshire: A Rapid Epidemiologic Assessment

Abstract

BACKGROUND:

New Hampshire (NH) has had among the highest rates of fentanyl-related overdose deaths per capita in the United States for several years in a row—more than three times the national average in 2016. This mixed-methods study investigated drug-using practices and perspectives of NH residents who use opioids to inform policy in tackling the overdose crisis.

METHODS:

Seventy-six participants from six NH counties completed demographic surveys and semi-structured interviews focused on drug-using practices and perspectives, including use precursors, fentanyl-seeking behaviors, and experiences with overdose. Rigorous qualitative methods were used to analyze interview data including transcription, coding and content analysis. Descriptive statistics were calculated on quantitative survey data.

RESULTS:

Eighty-four percent of interviewees had knowingly used fentanyl in their lifetime, 70% reported overdosing at least once, and 42% had sought a batch of drugs known to have caused an overdose. The majority stated most heroin available in NH was laced with fentanyl and acknowledged that variability across batches increased overdose risk. Participants reported high availability of fentanyl and limited access to prevention, treatment, and harm reduction programs. There was widespread support for expanding education campaigns for youth, increasing treatment availability, and implementing needle exchange programs.

CONCLUSIONS:

A confluence of factors contribute to the NH opioid overdose crisis. Despite consensus that fentanyl is the primary cause of overdoses, individuals continue to use it and affirm limited availability of resources to address the problem. Policies targeting innovative prevention, harm reduction, and treatment efforts are needed to more effectively address the crisis.

1.0. Introduction

The United States (US) is in the midst of an opioid epidemic. Prescription opioid deaths quadrupled in the past 15 years, and drug overdoses—driven by prescription and illicit opioids—are now the leading cause of death in Americans under age 50 (Hedegaard et al., 2015). The dramatic spike in opioid use disorder (OUD) has also been accompanied by marked increases in injection-related infections (including infective endocarditis and Hepatitis C [HCV]; (Centers for Disease Control and Prevention (CDC), 2016, 2017b; Hartman et al., 2016; Keeshin and Feinberg, 2016), babies born with neonatal opioid withdrawal (Patrick et al., 2015), and healthcare and criminal justice costs (Rhyan, 2017). The rapid increase in overdose deaths in the US is largely due to a rise in deaths caused by non-methadone synthetic opioids, including fentanyl (Han et al., 2019). The synthetic opioid fentanyl and related analogs, full agonists at the mu-opioid receptor, are drivers of the opioid overdose epidemic given their potency (at least 50–100 times more potent than morphine and active in the tens of micrograms making precise dosing very difficult without sophisticated equipment) and widespread availability (low production costs and high profit value) (European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), 2019; Suzuki and El-Haddad, 2017; Tabarra et al., 2019). Increasingly, fentanyl is also being detected in drug supplies in Europe, Canada, Russia, and Latin America (Baldwin et al., 2018; Mars et al., 2018c; Mounteney et al., 2015) although the drug overdose death rates in the US are 3.5 times higher in comparison to most other high-income countries on average (Ho, 2019).

Several previous studies have examined motives for fentanyl use and fentanyl-related overdoses (Carroll et al., 2017; Ciccarone et al., 2017; Mars et al., 2018a; Mars et al., 2018c; McKnight and Des Jarlais, 2018). Qualitative studies examining motives for fentanyl use have reported mixed findings (Mars et al., 2018c). While people who use opioids (PWUO) frequently report seeking heroin rather than fentanyl (Carroll et al., 2017; Ciccarone et al., 2017), some PWUO have reported specifically seeking fentanyl because of its widespread availability and potency (Carroll et al., 2017; Peiper et al., 2018; Rouhani et al., 2019; Talu et al., 2010). Preference for fentanyl may be an inevitable response to an illicit drug supply dominated by illicitly manufactured fentanyl, as increases in fentanyl exposure can result in heightened phsycial dependence and greater tolerance among PWUO (Cicero et al., 2018; Macmadu et al., 2017). (Macmadu et al., 2018, Cicero et al., 2018). Fentanyl is often mixed with heroin and other drugs, leading some PWUO to report difficulties determining whether fentanyl is present, which increases the risk for overdose (Carroll et al., 2017; Ciccarone et al., 2017; Mars et al., 2018c).

Fentanyl-related overdoses have significantly increased in certain regions of the US, including the state of New Hampshire (NH) (Centers for Disease Control and Prevention (CDC), 2019). NH has had among the highest rates of fentanyl- and opioid-related overdose (OD) deaths per capita in recent years (ranked 1^st nationally from 2014 to 2016 and 3^rd in 2017 with no rate reduction from the previous year) relative to all other states in the US (Centers for Disease Control and Prevention (CDC), 2017a). Fentanyl-related OD fatalities in NH have increased over 1,590% from 2010 to 2017, and have more than doubled since 2014 alone (Centers for Disease Control and Prevention (CDC), 2018). Fentanyl is the primary opioid contributing to overdose deaths — 86% of opioid-related OD deaths in NH in 2017 included fentanyl, only one death was from heroin alone, and less than 12% included prescription opioids (New Hampshire Department of Justice Office of Chief Medical Examiner, 2017). Non-fatal overdose is a strong risk factor for future overdose and is associated with a range of health risks including cognitive impairment and musculature dysfunction, as well as high healthcare costs (Caudarella et al., 2016; Warner-Smith et al., 2001).

Despite the high rates of opioid-related deaths in NH, funding and access to treatment, particularly medication for opioid use disorder (MOUD), has been limited in the state (Knudsen et al., 2015; Substance Abuse and Mental Health Services Administration (SAMHSA), 2017). Although MOUD treatment in NH has recently become increasingly available, NH has historically had limited funding for public health and substance use disorder treatment, including the lowest total and per capita spending on public health compared to other New England states (America’s Health Rankings Analysis of Trust for America’s Health, 2019; Center for Substance Abuse Treatment (CSAT), 2015; New Hampshire Governor’s Commission, 2015), and lower public health funding per resident than the national average (Trust for America’s Health, 2016). Further, NH has a lower rate of buprenorphine prescribers per capita compared to the national average and other New England states (Knudsen et al., 2015). In 2017, only 3.8% of outpatient, detox, and residential treatment admissions in NH were to MOUD programs (Substance Abuse and Mental Health Services Administration (SAMHSA), 2017).

To better understand factors contributing to the disproportionately high rate of opioid-related overdose deaths, the National Institute on Drug Abuse (NIDA) funded research teams at the National Drug Early Warning System (NDEWS) and Dartmouth College to conduct a surveillance study in NH in August 2016, which included meetings with a diverse array of stakeholders (e.g., state authorities, policymakers, treatment providers, first responders and community groups). Stakeholders emphasized that data from PWUO was imperative to more precisely understand and tackle the crisis (National Drug Early Warning System (NDEWS), 2016). The present mixed-methods study aimed to respond to these recommendations and systematically evaluate factors giving rise to the exceptionally high rates of overdose deaths in NH from the perspectives of PWUO, with an emphasis on their perceptions about, and experiences with, fentanyl.

2.0. Materials and methods

2.1. Design

Initial assessment measures and interview guides were compiled based on knowledge gaps identified from the NDEWS/Dartmouth College surveillance study (e.g., what are precursors to opioid use in NH?; are PWUO seeking fentanyl?; what are preferences for harm reduction models and treatment?). Guidance was also provided by expert qualitative researchers at Dartmouth College. PWUO completed a demographic survey and semi-structured interview related to their experiences with, and perspectives about, opioid use and overdoses in NH.

2.2. Participants and settings

Seventy-six active or recent (past 12 months) adult NH PWUO were recruited from November 2016 through January 2017 from six of NH’s 10 counties via advertisements (Craigslist, community-posted flyers, and newspapers) and word of mouth. Hillsborough County was heavily targeted, given that it has among the highest rates of overdose deaths and includes the two largest metropolitan areas (Manchester and Nashua) in the state. The other targeted counties were Cheshire, Grafton, Rockingham, Strafford, and Sullivan counties. Participant interviews were conducted by telephone or in person at partnering treatment facilities or Dartmouth College, based on participant preference.

2.3. Measures

2.3.1. Participant characteristics

The demographic baseline survey was an 18-item self-report form on age, gender, race, ethnicity, education level, employment, marital and housing status, substance use history, substance use and mental health treatment history, experiences with medications for opioid use disorder (MOUD), overdose history, and naloxone administration.

2.3.2. Qualitative interview

The hour-long, semi-structured qualitative interview consisted of open-ended questions on: (1) precursors of opioid use, (2) trafficking and supply chain, (3) formulation of heroin and fentanyl, (4) fentanyl-seeking behaviors, (5) experiences with overdose, (6) experiences with naloxone, (7) prevention, (8) harm reduction, (9) experiences with treatment, and (10) laws and policies. Examples of interview questions include: “Have you ever overdosed on drugs? Tell me about that/those experience(s).”; “Where do you think the drugs causing these overdoses come from?”; “Do you think people are seeking out drugs that might have caused overdoses?”; “Please tell me a little about your experiences getting treatment for your opioid use, if you have ever sought treatment.”; “If anything were possible, how would you prevent people from using opioids in the first place?”; and “What are your thoughts on harm reduction practices?” Additional probes were provided for many interview questions.

2.4. Procedure

Prospective participants contacted the research team via phone or email. Individuals aged 18 years or older who reported illicit opioid use within the past 12 months were eligible and scheduled at a later date. Participants provided verbal consent, completed the demographic survey and were then interviewed by one of five trained research staff members at a single visit. Visits, whether remote or in person, took approximately 1.5 hours on average. Interviews were audio recorded and transcribed verbatim. Participants received a $50 gift card as compensation.

Dartmouth College’s Committee for the Protection of Human Subjects approved the measures, consent process, and collection, analysis, and reporting of these data.

2.5. Analysis

Interview transcripts were uploaded into qualitative analysis software, ATLAS.ti (Scientific Software Development GmbH, 2013). Analysts used content analysis, a methodology used to systematically code and categorize (Cavanagh, 1997; Hsieh and Shannon, 2005) and ultimately describe participants’ perspectives and experiences by condensing transcripts into content-related categories (Elo and Kyngas, 2008; Weber, 1990). First-level codes were derived deductively (Elo and Kyngas, 2008), predetermined by the interview guide.

Analysts independently reviewed a subsample of transcripts to identify patterns, develop a refined code list, and then collectively coded several transcripts. The code list was iteratively refined as analysts reviewed additional transcripts. Once consensus was reached that the code list was complete, the remaining transcripts were coded; two analysts then conducted subtheme analyses for each code. The full research team met weekly to reach consensus regarding observed patterns and provide feedback on the ongoing subtheme analyses. Demographic data were analyzed using Stata to calculate descriptive statistics (StataCorp, 2015). Means and standard deviations were calculated for continuous variables, while proportions were calculated for categorical variables.

3.0. Results

3.1. Participant characteristics

Participants (n=76) were mostly non-Hispanic (96%), White (91%), adults (m=34.1 years), consistent with demographic characteristics cited among treatment-seeking people who use heroin in NH and across the US (Substance Abuse and Mental Health Services Administration (SAMHSA), 2017). The majority had no college degree (96%) and were unemployed (62%) (Table 1).

Table 1.

Demographic characteristics of study participants

Demographics	Participants (n=76)
Age m(sd)	34.1 (8.3)
Gender n(%)
Male	37 (49%)
Female	39 (51%)
Race n(%o)
Asian	1 (1%)
Black/African American	1 (1%)
White	69 (91%)
Other	1 (1%)
Multiracial	4 (5%)
Ethnicitya n(%o)
Hispanic and Latino	3 (4%)
Not Hispanic or Latino	72 (96%)
Education n(%)
Less than high school	5 (7%)
High school/GED	41 (54%)
Some college	16 (21%)
Associate’s	11 (15%)
College degree	3 (4%)
Employment status n(%)
Working full time	20 (26%)
Working part time	9 (12%)
Unemployed	22 (29%)
Disabled	13 (17%)
Other	12 (16%)
Marital status n(%)
Married	10 (13%)
Divorced	9 (12%)
Separated	8 (11%)
Never married	31 (41%)
Living with partner	18 (24%)
Housing status n(%)
Rent or own home	42 (55%)
Live with someone, no rent	20 (26%)
Residential/halfway house	2 (3%)
Shelter	6 (8%)
Homeless	6 (8%)
County n(%)
Cheshire	7 (9%)
Grafton	6 (8%)
Hillsborough	41 (54%)
Rockingham	6 (8%)
Strafford	8 (11%)
Sullivan	8 (11%)

^aOne participant did not report ethnicity. The percentages reflect the number out of the 75 participants who reported ethnicity.

3.2. Identified themes

Results are presented by the 10 categories targeted during interviews and the themes that emerged from each. Representative quotes from interviews are included to illustrate each primary theme (Table S1). All results are based on participants’ own perspectives and experiences.

3.2.1. Precursors to opioid use

Survey data indicate that almost all participants (99%) reported lifetime use of alcohol and cannabis, initiating around 13–14 years of age and generally preceding initiation of other substances (Table 2). Participants’ mean age of first prescription opioid use [21.1 (SD=7.1) years; 99% of sample] predated their first use of heroin [24.1 (SD=7.0) years; 92%] or fentanyl [28.1 (SD=7.3) years; 84%] (Figure 1). Eighty-six percent used prescription opioids at a younger age than heroin or fentanyl, and 71% used heroin at a younger age than fentanyl, highlighting the pattern of opioid initiation with prescription opioids (either medical and non-medical) followed by use of heroin and subsequently illicit fentanyl/fentanyl-laced heroin (F/FLH). Among those endorsing F/FLH use, 32% reported using F/FLH within the past week (Table 3). Thirty-four percent of participants reported using any opioid in the past week, and 54% had used in the past month.

Table 2.

Participant substance use, treatment history, and experiences with overdose

	Participants (n=76)
Lifetime overdose history
Lifetime overdose n(%)
No	23 (30%)
Yes	53 (70%)
Number of overdoses m(sd)	3.0 (3.7) (Range: 0–20)
Overdoses caused bya: n(%)
Heroin only	78 (35%)
Fentanyl only	32 (14%)
Heroin and fentanyl combination	68 (30%)
Other	48 (21%)
Received naloxoneb n(%)
No	20 (38%)
Yes	33 (62%)
Number of naloxone administrations per overdosec m(sd)	3.0 (1.6) (Range: 1–7)
Lifetime substance use
Alcohold	74 (99%)
Cannabis	75 (99%)
Prescription opioids	75 (99%)
Cocaine	71 (93%)
Heroin	70 (92%)
Fentanyl	64 (84%)
Benzodiazepines	53 (70%)
Hallucinogens	52 (68%)
Stimulants	51 (67%)
Inhalants	25 (33%)
Sedatives	24 (32%)
Other	4 (5%)
Lifetime substance use treatment history
Previous treatment for opioid use
Yes	69 (91%)
No	7 (9%)
Number of prior treatment episodes m(sd)	6.1 (7.7)
Currently on OUD treatment waitlist n(%)	11 (15%)
Naltrexone treatment n(%)
Never	68 (89%)
Previously	6 (8%)
Currently	2 (3%)
Buprenorphine treatment n(%)
Never	26 (35%)
Previously	14 (19%)
Currently	35 (47%)
Methadone treatment n(%)
Never	47 (62%)
Previously	16 (21%)
Currently	13 (17%)

OUD, opioid use disorder

^aSelf-reported and not confirmed by toxicology test; out of 228 total self-reported overdoses

^bOf those who had ever overdosed (n = 53)

^cOf those who had ever received naloxone (n = 33)

^dDenominator of 75 due to item nonresponse

Figure 1.

Mean age of prescription opioid, heroin, and fentanyl initiation among study participants

Table 3.

Recency of prescription opioids, heroin, and fentanyl use among participants

	Prescription opioids (n=75)	Heroin (n=70)	Fentanyl (n=66)
Last reported use n(%)
Past week	8 (11%)	20 (29%)	21 (32%)
Past month	12 (16%)	13 (19%)	12 (18%)
Past 6 months	16 (21%)	18 (26%)	14 (21%)
More than 6 months	39 (52%)	19 (27%)	19 (29%)

Participants identified four main precursors to opioid misuse: (1) early experimentation with substances (70%), (2) injuries, surgeries, pregnancy, or chronic pain resulting in opioid prescriptions (67%), (3) self-medication of mental health conditions (50%), and (4) intrafamilial substance use (40%). “[Doctors] weren’t really taking care of me enough, and my insurance wouldn’t cover me to get into a good pain clinic, so I was kind of flying on one wing. I was still in a lot of pain, so what they ended up making me do was look for other people that had pain meds so I could just be right… next thing I knew [heroin/fentanyl mix] was in front of me” [Participant identification number: 3002; age: 35 years; gender: M].

Sixty-two percent referenced only one or two of these precursors, while 20% referenced three and 18% referenced all four.

3.2.2. Formulation of heroin and fentanyl

The majority of participants felt they could distinguish fentanyl alone (e.g., no other illicit substances) or FLH from heroin alone by appearance (F/FLH is lighter in color —”like a white beige” [5001, 31, M]), taste (F/FLH is sweeter —”cut with a sugary base” [3002, 35, M]), effect (F/FLH onset is faster, more intense, and has a shorter duration —”hits you hard, but then it seems you are dope sick quick” [3014, 33, M]) and/or cost (F/FLH is cheaper—”fentanyl is dirt cheap” [6001, 38, M]). One-third reported no ability to distinguish the difference, particularly because heroin and fentanyl were generally sold as a mixed product. “The only time you knew you had fentanyl was when it was white and it was straight fentanyl. Any other time, like the heroin mixed with the fentanyl, the heroin still had a tint of tan/brown to it, so it looked the same” [1007, 26, F].

3.2.3. Trafficking and supply chain

Participants reported F/FLH entering NH around 2014, primarily from Massachusetts because of greater profit potential in NH. “Just a quick drive to Massachusetts, and then you can sell it [in NH] for double the price” [3031, 29, F]. There was general agreement that the demand for F/FLH in NH is driven by lower cost, higher potency, and greater availability compared to other opioids.

3.2.4. Fentanyl-seeking behavior

Eighty-four percent had knowingly used fentanyl in their lifetime (Table 2). Twenty-five percent of all participants endorsed seeking fentanyl specifically. “We want whatever is strongest and the cheapest. It’s sick. I know me using, when I hear of an overdose, I want it because I don’t want to buy bad stuff. I want the good stuff that’s going to almost kill me” [3022, 29, F]. After initiating F/FLH use, heroin alone is sometimes considered “maintenance dope”[3034, 43, M]. It does not adequately address withdrawal and certainly won’t “f*** you up [the way F/FLH will]” [3034, 43, M]. The remaining 75% were resigned to using F/FLH due to lack of alternatives:”I don’t know anybody that just goes out looking for fentanyl… mainly people are looking for heroin, but now they’re hand in hand” [1001, 47, M].

3.2.5. Experiences with overdoses

Seventy percent of participants had experienced at least one overdose from any substance (Table 2). Nearly half reported having a friend or family member overdose, and many had witnessed an overdose. There was near unanimous agreement that fentanyl was the cause of increased overdose rates in NH, mostly due to F/FLH batch potency and variability making it “pretty hard to make a safe dose for yourself” [3014, 33, M].

Forty-two percent had sought batches of drugs known to have caused an overdose: “[Overdosing] is one of the best highs you’ll ever have” [6004, unknown age, F]. The majority reported that being witness to an overdose was scarier than personally experiencing one due to their inability to recall their own overdose once unconscious. Experiences with overdoses were not limited to an injection route of administration, as some reported experiencing or hearing of opioid overdoses from smoking or intranasal use.

3.2.6. Experiences with naloxone

Sixty-two percent of those who had overdosed reported receiving naloxone reversal medication. Participants reported an average of three naloxone doses per overdose (Table 2). Although 21% of participants stated that naloxone was either unavailable or they lacked knowledge on how to obtain naloxone, 14% believed access was increasing. However, participants still reported significant barriers to naloxone access in NH, including a lack of knowledge (e.g., mechanism of action and how to access), high cost, a fear of stigmatization (“I don’t think they [police/fire/hospitals] would give it to a strung-out addict who is just gonna overdose and try and bring himself back” [5001, 31, M]), and a fear of precipitated withdrawal (“It’s like your pain sensors are covered up with a nice, cozy blanket, and some a**hole comes and rips it off you” [3034, 43, M]). There were no reports of unanticipated side effects from naloxone use.

3.2.7. Harm reduction

Thirty-eight percent of participants reported current or past use of non-prescribed buprenorphine (“street bupe” [3003, 35, M]) as a harm reduction strategy, stating their primary motivation was to prevent withdrawal symptoms, often while awaiting placement in formal treatment. Participants believed buprenorphine was safe to use as a taper and that fluctuations in its availability contributed to recurrent use of illicit opioids. In addition to increasing “street bupe” [3003, 35, M] availability, there was broad support for legislative approval of needle exchange programs. Lack of available clean needles in NH contributed to “desperate measures” [3022, 29, F] for participants, like choosing a “syringe over a sandwich” [3006, 35, M]. Participants emphasized that “people are going to shoot no matter what” [3034, 33, M] and that needle exchange programs could mitigate some of the risks, including “reduc[ing] HIV/AIDS transmission and hepatitis C transmission” [3036, 40, M].

3.2.8. Experiences with treatment

Over 90% of participants reported receiving some treatment for opioid use during their lifetime (Table 2). Among participants who received MOUD, more reported receiving prescriptions for buprenorphine than methadone or naltrexone. There is unanimous agreement that participants cannot stop using opioids without MOUD and/or counseling support and that available services are lacking in NH. “When you know someone who’s willing and able and ready and physically standing there in the halls of the [treatment program] in front of you, and you say ‘Come back in 8 weeks,’ that’s crazy. You could be dead tonight. Eight weeks is a f***ing long time in the trenches” [1006, 25, M]. Barriers to treatment in NH included lengthy waitlists, trouble navigating the treatment system, payment challenges, and the lack of long-term programs. Participants also reported that treatment referral rates after an overdose episode were low.

3.2.9. Prevention

Five main prevention strategies were suggested (in order of frequency reported): (1) initiate education/prevention campaigns before middle school, (2) address societal stigma, (3) attend to intergenerational substance use, (4) prescribe opioids more prudently, and (5) educate patients who are prescribed opioids (“I honestly never knew you could get withdrawals from it [prescription opioids] until I dealt with them” [3027, 26, F]).

3.2.10. Laws and policies

Overall, participants reported not being well informed on NH laws and policies related to opioids. They perceived the laws to be harsh (“The law’s coming down harder on the opioid user than they ever have” [6004, unknown age, F]) and expressed continued skepticism of the protections the Good Samaritan Law purports to provide (i.e., legal protection for people calling for overdose emergency assistance). Participants emphasized that jail is not treatment and described a near-total absence of substance use treatment services in state correctional facilities. Additionally, participants stated that opioid prescribing crackdowns may reduce the number of pills in circulation for diversion but will unquestionably result in increased F/FLH use.

4.0. Discussion

The fentanyl-related overdose fatality rate and high levels of concomitant fentanyl use in NH underscored the importance of conducting this study. Seventy percent of PWUO in this sample reported experiencing at least one lifetime overdose, a noted risk factor for future fatal and non-fatal overdose. The study results point to a confluence of factors contributing to the state’s opioid crisis. F/FLH was believed to be the primary cause of the increase in overdoses, due in part to the ubiquity of F/FLH and the variability in batches of F/FLH. However, participants underscored a network of other factors they believe also contribute significantly to NH’s opioid overdose crisis, including: 1) an array of precursors to opioid initiation, 2) the formulation of heroin and fentanyl available in the region, 3) barriers to accessing naloxone and limited uptake of other harm reduction strategies, 4) fentanyl-seeking behavior, 5) limited treatment infrastructure, as well as a 6) criminal justice approach to a public health crisis. Taking into account the experiences and perspectives of those at risk of overdose in NH can provide critical insights to help inform public health initiatives.

Initiation of opioid use was largely attributed to four primary precursors, namely early experimentation, legitimate opioid prescriptions, self-medication of mental health symptoms, and intrafamilial substance use. Study findings emphasize the importance of prevention strategies targeting risk and enhancing protective factors in childhood, such as the Strengthening Families Program (SFP), which is shown to reduce early experimentation with substances, including prescription opioid use by 50% (Kumpfer and Magalhães, 2018; Kumpfer et al., 2010; Spoth et al., 2013). Interventions helping to break the cycle of intergenerational substance abuse by strengthening the family and parenting skills, such as the Life Skills Training (LST) program, are also important (Crowley et al., 2014).

Legitimate opioid prescriptions – another noted precursor – were associated with the demand for heroin and synthetic opioids like fentanyl among participants. NH has historically prescribed significantly higher rates of prescription pain medications (specifically, long-acting/extended-release and high-dose opioids) relative to the national average (Centers for Disease Control and Prevention (CDC), 2014). Participants suggested improved patient education regarding the addictive potential of opioid medications, as well as more prudent opioid prescribing as preventative measures. Half the sample also reported that mental health concerns, including anxiety, depression, and abuse/trauma were related to their initiation of opioid use. Preventing opioid deaths will require a better understanding of the intermingling of opioid-overdose and mental health, especially suicidal behavior (Oquendo and Volkow, 2018).

Participants in this study were largely aware of their potential exposure to fentanyl. Consistent with other research exploring fentanyl use and overdose among PWUO in the US (Carroll et al., 2017; Ciccarone et al., 2017; Mars et al., 2018a, b), some participants reported using the visual appearance, taste, cost, and subjective effect to determine whether their heroin was adulterated with fentanyl. Others felt these cues were inadequate and could not detect the difference between heroin and FLH. This uncertainty contributed to the risk of overdose by making it challenging to use a safe dose. Though fentanyl test strips have been suggested as a harm reduction strategy and some PWUO in other regions have reported a willingness to use such test strips and have reduced risky fentanyl use when using them (Krieger et al., 2018; McKnight and Des Jarlais, 2018; Peiper et al., 2019), participants in this study expressed mixed attitudes toward them, and some even reported they would use test strips to confirm that fentanyl was present rather than to avoid it.

Results underscored that the expansion of harm reduction resources in NH is necessary. During data collection, NH had no needle exchange programs despite consensus from participants that they are a critical harm reduction strategy. In June 2017, NH enacted legislation to allow the operation of syringe service programs (SSPs) (New Hampshire State Legislature, 2017) and since then, five programs have launched (Moon, 2019; New Hampshire Division of Public Health Services, 2018). SSPs typically provide access to tools for the prevention and reversal of opioid overdose, including fentanyl test strips and naloxone (Moon, 2019; New Hampshire Division of Public Health Services, 2018). SSPs often provide multiple benefits to those who participate beyond accessing tools for preventing and reversing overdoses, including referalls for mental health services (Carroll et al., 2018). Individuals who participate in SSPs are less likely to spread HIV and Hepatitis C, and are more likely to seek treatment for a substance use disorder than those who do not participate (Carroll et al., 2018). Although naloxone is available in NH, participants cited substantial barriers to its use, including concerns about side effects, fear of legal prosecution, and uncertainty about how to access it, when to use it, and laws surrounding its use (Bessen et al., 2019). These barriers can be reduced through increased access to SSPs and other distribution networks staffed by a knowledgeable, non-judgemental workforce. As one example, NH received additional funding focused on the opioid crisis, including a grant by the First Responders: Comprehensive Addiction and Recovery Act (FR-CARA) and SAMHSA to launch NH Project First in April 2019. Project First is a Mobile Integrated Healthcare (MIH) program for first responders targeting reduction of opioid-related overdoses, linkage of those with OUD to recovery and treatment services and increased training in the administration of naloxone (New Hampshire Department of Safety, 2019).

Approximately one-fourth of participants reported seeking fentanyl because of its potency, availability, and low cost. Most were resigned to using fentanyl, reporting its effects were aversive but it was challenging to avoid. Some PWUO reported seeking batches of drugs known to have caused an overdose while others would avoid these batches. This variation is consistent with other research exploring drug-seeking behaviors (Carroll et al., 2017; Ciccarone et al., 2017; Mars et al., 2018a; McLean et al., 2019; Rouhani et al., 2019). One study found that indivduals who reported preferring fentanyl were less likely to report using harm reduction strategies (e.g., using less of the drug, using more slowly or doing a tester shot) to reduce overdose risk than those who did not prefer fentanyl (Rouhani et al., 2019). Another recent study found that overdose history interacts with fentanyl preference. Compared with those who had never overdosed, PWUO with a recent overdose were less likely to prefer fentanyl while those who survived an overdose at least one year ago were more likely to prefer fentanyl (Morales et al., 2019). These findings suggest early intervention with people who have overdosed to capitalize on a window of time during which PWUO and have overdosed may be more receptive to outreach and education about harm reduction strategies. And, overall, more outreach and education on harm reduction strategies to reduce the risk of overdose death may be particularly important for individuals who are seeking fentanyl, including more widespread distribution of naloxone and encouragement to not use drugs alone.

A limited treatment infrastructure compounded by long waitlists at the time this study was conducted was also found to contribute to high rates of overdose according to study participants. PWUO overwhelmingly agree that access to MOUD, buprenorphine in particular, is an effective and preferred treatment. Efforts to expand access to MOUD are critical to reduce rates of overdose deaths in NH. Since this study was conducted, on January 1, 2019, NH was awarded the State Opioid Response Grant by the Substance Abuse and Mental Health Services Administration (SAMHSA), which has been used to fund The Doorways, nine entry points across the state for residents with OUD (New Hampshire Department of Health and Human Services, 2019). The Doorways offers increased access to MOUD, peer recovery support and prevention programs. Among the many evidence-based strategies for preventing overdose is the elimination of prior-authorization requirements for MOUD (Weber and Gupta, 2019). Notably, the NH State Medicaid program has removed prior authorization requirements for MOUD in early 2019 (New Hampshire Department of Health and Human Services (NH DHHS), 2019).

The continued criminal justice approach to the public health overdose crisis highlighted by participants is not unique to NH. Specifically, mistrust of law enforcement poses a serious threat to overdose prevention. The Police Executive Research Forum recently published ten standards of care after convening noted experts on the intersection of policing and the opioid epidemic (Del Pozo et al., 2018). Beyond arrests, and drug/weapon/drug-related funding seizures, police departments have a role to play in working collaboratively to reduce overdose deaths. Several of the standards speak directly to the issues raised by study participants, such as educating the public on addiction and stigma, referring individuals to treatment, and supporting good Samaritan laws. An additional factor noted by participants contributing to the opioid crisis in NH is access and proximity to the opioid supply chain via interstates (Seelye, 2016). On October 6, 2019, NH State Police announced it would receive more than $600,000 in federal funds to help law enforcement agencies fight the opioid crisis (Associated Press, 2019). State police representatives suggested this money would support the state’s anti-heroin task force and be used to disrupt cross-border drug trafficking between NH and MA.

There are an abundance of public health and policy implications from the findings of this study, consistent with several recommendations over recent years (Frank and Pollack, 2017; Latkin et al., 2019; Saloner et al., 2018). There is an opportunity for a diversity of systems and stakeholders, including health care delivery systems, medical examiners, departments of corrections, and law enforcement agencies, to bolster surveillance and early warning systems for rapidly predicting, preventing, detecting, and responding to overdose outbreaks by improving screening for subthreshold opioid use problems, inclusion of synthetic opioids in routine toxicology tests, increasing low-barrier access to naloxone and MOUD (e.g., through emergency departments, corrections, and first response systems), and identifying and responding to the distribution and manufacturing of synthetic opioids (Del Pozo et al., 2018). Furthermore, state media campaigns to target stigma, education (i.e., for PWOU, families, communities and providers), naloxone distribution and resources and services is warranted (McGinty et al., 2018).

4.1. Limitations

This study recruited NH PWUO and inquired specifically about participants’ experiences with opioid use in NH, a small rural state. The experiences and perspectives of participants may not represent those of PWUO in other regions. Several recent studies have explored causes of increased F/FLH overdose in other states and found similar patterns with regard to F/FLH variation and challenges accessing treatment and harm reduction resources (Bode et al., 2017; Carroll et al., 2017), suggesting that some drivers of increased F/FLH may be similar across geographic regions. Of the counties experiencing above-average overdose death rates in 2015, two-thirds were in rural America. Less populous and lower-density communities have experienced some of the sharpest upticks in their drug poisoning death rates since 2000 (Lenardson et al., 2016; Palombi et al., 2018). Although NH’s rate of fatal synthetic opioid overdoses has remained constant over the past two years, other states, including Ohio and West Virginia, have seen drastic increases (Centers for Disease Control and Prevention (CDC), 2018). It is important to similarly conduct rapid epidemiologic assessments of the opioid crisis in these other rural states to understand similarities and differences in the complexion of this epidemic across settings. Study participants comprised a volunteer sample recruited purposively; recruiting a probability sample of NH PWUO was infeasible. Additionally, this study relied on self-reported data without verification of F/FLH use through objective measures (e.g., toxicology screens). It is possible that not all participants were able to accurately report on whether they had used F/FLH, as cited in other research (Griswold et al., 2018; Kenney et al., 2018; Macmadu et al., 2017; Palamar et al., 2019). Despite this, many study participants reported being able to differentiate between F/FLH and heroin alone. The study also did not focus on other data sources, such as Drug Enforcement Administration, drug trafficking or medical examiners’ data.

5.0. Conclusions

Despite consensus that fentanyl is the primary cause of overdoses, NH residents who use opioids continue to use it and affirm limited availability of resources to address the problem. As with other public health crises in this country, the common denominators are social and structural determinants (e.g., government policies, service systems, laws) that combine with individual-level factors. Policies targeting innovative prevention, harm reduction, and treatment efforts are needed to more effectively address the opioid crisis.

Highlights.

• Respondents reported high availability of fentanyl in New Hampshire (NH).

• Consensus that fentanyl is the presumptive primary cause of overdose death in NH.

• Majority of interviewees had knowingly used fentanyl.

• Confluence of social, structural, and individual factors were associated with use.

• Lack of prevention, treatment and harm reduction resources were emphasized.