Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Jun 1.
Published in final edited form as: Int J Drug Policy. 2017 Jun 1;44:121–129. doi: 10.1016/j.drugpo.2017.05.007

Increased in synthetic cannabinoids-related harms: results from a longitudinal web-based content analysis

Francois R Lamy a,b, Raminta Daniulaityte a,b, Ramzi W Nahhas c,d, Monica J Barratt e,f,g, Alan G Smith b, Amit Sheth b, Silvia S Martins h, Edward W Boyer i, Robert G Carlson a,b
PMCID: PMC5545681  NIHMSID: NIHMS876447  PMID: 28578250

Abstract

Background

Synthetic Cannabinoid Receptor Agonists (SCRA), also known as “K2” or “Spice,” have drawn considerable attention due to their potential of abuse and harmful consequences. More research is needed to understand user experiences of SCRA-related effects. We use semiautomated information processing techniques through eDrugTrends platform to examine SCRA-related effects and their variations through a longitudinal content analysis of web-forum data.

Method

English language posts from three drug-focused web-forums were extracted and analyzed between January 1st 2008 and September 30th 2015. Search terms are based on the Drug Abuse Ontology (DAO) created for this study (189 SCRA-related and 501 effect-related terms). EDrugTrends NLP-based text processing tools were used to extract posts mentioning SCRA and their effects. Generalized linear regression was used to fit restricted cubic spline functions of time to test whether the proportion of drug-related posts that mention SCRA (and no other drug) and the proportion of these “SCRA-only” posts that mention SCRA effects have changed over time, with an adjustment for multiple testing.

Results

19,052 SCRA-related posts (Bluelight (n=2,782), Forum A (n=3,882), and Forum B (n=12,388)) posted by 2,543 international users were extracted. The most frequently mentioned effects were “getting high” (44.0%), “hallucinations” (10.8%), and “anxiety” (10.2%). The frequency of SCRA-only posts declined steadily over the study period. The proportions of SCRA-only posts mentioning positive effects (e.g., “High” and “Euphoria”) steadily decreased, while the proportions of SCRA-only posts mentioning negative effects (e.g., “Anxiety,” “Nausea,” “Overdose”) increased over the same period.

Conclusion

This study's findings indicate that the proportion of negative effects mentioned in web forum posts and linked to SCRA has increased over time, suggesting that recent generations of SCRA generate more harms. This is also one of the first studies to conduct automated content analysis of web forum data related to illicit drug use.

Keywords: synthetic cannabinoids, web-forums, semantic web, drug abuse ontology, NLP text processing

Introduction

Synthetic cannabinoids are a large family of chemical substances designed to reproduce specific therapeutic/psychotropic properties of cannabis. Amongst these compounds, the Synthetic Cannabinoid Receptor Agonists (SCRA) mimicking Δ-9-tetrahydrocannabinol (THC) have attracted substantial attention due to their potential for abuse, their numerous emergency department presentations and increasing number of overdose-related cases reported in the US and in Europe (AAPCC, 2016; Bush & Woodwell, 2014; J. Cohen, Morrison, Greenberg, & Saidinejad, 2012; European Monitoring Centre for Drugs and Drug Addiction, 2013; Forrester, 2012; Riederer, 2016; Trecki, Gerona, & Schwartz, 2015). These substances belong to the Novel Psychoactive Substances (NPS) category, a heterogeneous group of substances temporarily not regulated by international legislation (Papaseit, Farré, Schifano, & Torrens, 2014; Zawilska, 2011; Zawilska & Andrzejczak, 2015). SCRA products have been available for purchase in “head shops” and through online websites since 2004 (Fabrizio Schifano, et al., 2009). SCRA were initially dissolved and sprayed onto inert vegetal material to be sold as “herbal incense”, “potpourri” or “legal high” labeled as “not for human consumption” under various marketing names (e.g., “Black Mamba”, “Ultimate Warrior”, “Mad Hatter”). SCRA were also commercialized directly in their powder form under their chemical designation (e.g., JWH-018, MAM-CHMINACA, UR-144).

Unlike natural cannabis products (e.g., floral cannabis, hashish, marijuana concentrates), SCRA are full agonists of the endocannabinoid CB1 and CB2 receptors inducing a stronger neuronal response (Brents & Prather, 2014; Huffman & Padgett, 2005). This enhanced neurological response is one of the main factors contributing to a large spectrum of observed adverse effects such as increased risk of psychosis (Every-Palmer, 2011; van Amsterdam, Brunt, & van den Brink, 2015), seizures and convulsions (Schneir, Cullen, & Ly, 2011), dependence (Zimmermann, et al., 2009), and kidney failure (Bhanushali, Jain, Fatima, Leisch, & Thornley-Brown, 2012; Centers for Disease Control Prevention (CDC), 2013; Kazory & Aiyer, 2013).

Despite legislative efforts to regulate new cannabinoids as soon as they reach the market (e.g., the U.S. government has scheduled various SCRA molecules five times in the last five years), banned SCRA are rapidly replaced by molecules presenting chemical variation(s) to bypass current legal interdictions (J. A. Cohen, 2014; Lindigkeit, et al., 2009). For example, the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) has reported the emergence of 134 SCRA from the end of 2008 until end of 2014 (EMCDDA, 2015). Reported emergency department cases linked to SCRA have increased in the U.S. over the past three years (Riederer, et al., 2016). According to the American Association of Poison Control Centers (AAPCC), 7,779 calls for of synthetic cannabinoid exposures were reported in 2015 exceeding both annual cases reported in 2013 (2,668) and 2014 (3,682) (AAPCC, 2016).

Except for a small number of epidemiological surveys describing SCRA use (Barratt, Cakic, & Lenton, 2013; Castellanos, Singh, Thornton, Avila, & Moreno, 2011; Caviness, Tzilos, Anderson, & Stein, 2015; Hu, Primack, Barnett, & Cook, 2011; Johnson, Johnson, & Portier, 2013; Vandrey, Dunn, Fry, & Girling, 2012; Winstock & Barratt, 2013a, 2013b), most of this literature informs either the neuropharmacology and neurophysiology of these products (Grigoryev, et al., 2011; Kavanagh, Grigoryev, Savchuk, Mikhura, & Formanovsky, 2013) or the adverse effects observed on emergency department patients (Besli, Ikiz, Yildirim, & Saltik, 2015; Bonar, Ashrafioun, & Ilgen, 2014; Johnson, et al., 2013; Tait, Caldicott, Mountain, Hill, & Lenton, 2016). Collecting information regarding SCRA patterns of use and related effects remains challenging because SCRA users are sometimes difficult to reach through conventional epidemiological surveillance methods. Furthermore, not all SCRA users who have experienced negative effects require or seek medical attention, nor were those who did receive medical attention necessarily assessed for SCRA-related symptoms (Tait, et al., 2016).

To overcome the challenges of identifying and recruiting SCRA users, we analyzed posts from web-forums dedicated to drug discussions. The last few years have seen an increased interest in the use of large-scale internet-based data as a source of information on evolving public health problems and emerging trends (Butler, et al., 2007; Daniulaityte, Nahhas, et al., 2015; Eysenbach, 2011; Lamy, et al., 2016; Lazer, et al., 2009; P. G. Miller & Sonderlund, 2010). Recent research has also demonstrated that social media analysis provides valuable information regarding new drug trends, such as the use of loperamide by opiate users to ease opiate withdrawal (Daniulaityte, et al., 2013), or changes in attitudes and behaviors of a population of synthetic cannabinoid users in Norway (Bilgrei, 2016).

In contrast to other social media platforms (such as Twitter or Instagram), Web forums allow their users to post extensive comments about their drug use, share experiences or ask questions regarding new products and trends. These websites also favor dialogue between users, who can share their thoughts, advice, and techniques on similar topics. Analysis of such data can provide valuable information regarding effects, patterns of use, and opinions directly from the user's perspective (Miller & Sønderlund, 2010; Schifano, et al., 2006). Nevertheless, harnessing data on drug forums is a significant challenge due to the large amount of data displayed on Web forums. We used an advanced software platform, eDrugTrends (Daniulaityte, Nahhas, et al., 2015; eDrugTrends, 2015), to collect and analyze forum posts related to SCRA and their effects, and conducted a longitudinal analysis of drug-focused web-forum posts. The key aims of this study are to: 1) identify the most common SCRA-related effects discussed on Web forums; and 2) analyze how frequencies of SCRA and their effects mentioned in Web forum discussions changed from January 2008 through September 2015.

Methods

The eDrugTrends system is a semi-automated comprehensive platform resulting from collaboration between social scientists in the Center for Interventions, Treatment and Addictions Research (CITAR) and computer scientists in the Ohio Center of Excellence in Knowledge-enabled Computing (Kno.e.sis Center) (eDrugTrends, 2015). This platform is designed to extract, process, and analyze social Web data, with the current application directed towards understanding cannabis and synthetic cannabinoid-related trends from Twitter and selected Web forums.

Data Collection

Data were collected from three publicly available web-based sources: Bluelight, Forum A, and Forum B. Web forum posts were retrieved using a combination of source specific custom Web Crawlers for Forums A and B, and through direct data transfer from Bluelight. Only publicly available/viewable information was collected and stored in the eDrugTrends platform. We have collected data from the opening of the Web forums until September 30, 2015. Since synthetic cannabinoids emerged in 2008, for this analysis we used Web forum data posted between the January 1, 2008 and September 30, 2015.

Ethical considerations

The study was approved by the Wright State University IRB under exemption 4 (i.e., Exempt Human Subjects Research) as collected data are publically available. One of the three forums used in this study, Bluelight.org, has a research portal accessible from the front page of the website, which asserts Bluelight's ownership of the forum content and instructs researchers to contact Bluelight administrators to discuss proposals for research, including archival analyses. The researchers contacted Bluelight to initiate discussions regarding this project, resulting in a partnership approach involving contributions of Bluelight representatives to the eDrugTrends project. The other two forums have been anonymized because data were freely available and consent has not been obtained. Any direct quotations in this paper were adequately modified without altering their basic meaning to protect the anonymity of the Web forum users who have posted these comments from being re-identified through searching on these quotations (Roberts, 2015).

Data Processing and Drug Abuse Ontology

Data extraction is based on a domain expert-driven ontology, the Drug Abuse Ontology (DAO). The DAO is a domain-specific conceptual framework interconnecting sets (named “classes”) of drug-focused lexicons initially designed for the PREDOSE project which analyzed Web forum posts mentioning buprenorphine (Cameron, et al., 2013; Daniulaityte, Carlson, Brigham, Cameron, & Sheth, 2015). One of the key benefits of using an ontology-enhanced semantic approach is the ability to identify all variants of a concept in data. For example, Daniulaityte et al. (Daniulaityte, Carlson, et al., 2015) report that for every occurrence of “buprenorphine” there were twenty-nine occurrences of its variants in other forms such as slangs (e.g., “bupe”) and alternative names (e.g., “Suboxone”).

The DAO was extended and refined for the eDrugTrends project (eDrugTrends, 2015) by encompassing and classifying words and expressions specific to cannabis and synthetic cannabinoids. Hence, the first phase included identifying specific SCRA terminology such as chemical denominations, commercial names and slang terms, and updating the DAO for the new research objectives. These search terms were drawn from the scientific literature, and names of “spice” products were collected from online retail shops (e.g., Ghcmarijuana; OrganicSpiceBlend). Next, the DAO was populated with general medical terms and slang expressions representing physical/psychological effects induced by drugs. These terms were based on the medical literature describing drug use consequences and extended with common denominations and/or slang terms for each of the existing medical terms. For example, the medical terms “bruxism” has been associated with the common denomination “jaw clenching” and the slang term “teeth grinding.” Drug effect expressions frequently contain a mention of a body part or a specific organ (e.g., “head”, “legs” or “lungs”) with a combination of adjectives and/or present/past participles describing the symptoms (e.g., “heart racing”, “upset stomach”, “nasty cough”). Hence, a Natural Language Processing algorithm was developed to inductively extract such expressions in posts by using the NLTK (i.e., Natural Language ToolKit) package from the Python computational language (Bird, Klein, & Loper, 2009; van Rossum, 1990).

Based on 4000 random posts, terms referring to organs/body parts were first recognized by comparing each word in the post with effect terms listed in the DAO and “tagged” as entities of interest. Secondly, the contextual words surrounding the tagged organ/body part were collected to recreate expressions. Relevant extracted expressions were manually selected and added to the index of effect-related terms already populating the DAO. Based on the scientific literature and the content of websites selling SCRA, the DAO was populated with 189 SCRA-related terms (cf. Supplementary material 1), which brings the total of substance denotations and slang terms contained in the DAO to 497. The DAO also indexes 501 effects and associated slang terms as search terms (cf. Supplementary material 2) based on medical literature concerning drug use, the domain-expert knowledge of the CITAR team, and expressions extracted using the NLP algorithm.

Content analysis

Content analysis focused on quantitative word frequency analysis and concept cooccurrence analysis. To extract relevant posts, the following post search queries were performed: first, posts containing at least one mention of SCRA based on DAO searched terms were extracted. Next, SCRA-related posts that mentioned other types (or type) of substances were removed by processing a negative query (SCRA-related terms AND NOT other drug-related terms) to ensure that effects described in one post were linked to SCRA and not to another drug or to a combination of substances (herein designated as SCRA-only). Third, SCRA-only posts were processed further to identify frequencies of different effect mentions. This analysis was based on the search terms referring to potential drug effects indexed in the DAO. Effect-related terms belonging to the same group were encompassed in a common category; for example, “anxiety”, “anxious”, “anxiety attack”, and “panic attack” were grouped under the common category “Anxiety.” A “fuzzy” query based on Damerau-Levenshtein edit distance was used to collect misspelled effect terms (Damerau, 1964; Levenshtein, 1966). This edit distance allows misspelled words, such as “siezure” instead of “seizure”, be captured by our query.

Trend Analysis

To examine changes over time, the following data were recorded for each month of data collection: (1) the number of drug-related posts (“Posts”); (2) of these, the number that mentioned an SCRA-related term (“SCRA Posts”); (3) of these, the number that mentioned no terms related to any other drugs (“SCRA-Only Posts”); (4) of these, the number that mentioned any SCRA-related effect (“SCRA-Only Effect Posts”); (5) of these, the number that mentioned any of the seven most commonly mentioned SCRA effects based on our content analysis (anxiety, euphoria, high, nausea, overdose, seizure, and hallucination) (“Frequent Effect Posts”); and (6) the numbers of posts mentioning each of the seven most frequent effects (“Individual Effect Posts”). We have limited our results to categories representing more than 2% of the effects mentioned in posts to present the most frequent effects discussed by forum users.

The following proportions were computed for each month:

  • SCRA Posts, SCRA-Only Posts, SCRA-Only Effect Posts, and Frequent Effect Posts, each as a proportion of Total Posts

  • SCRA-Only Posts, SCRA-Only Effect Posts, and Frequent Effect Posts, each as a proportion of SCRA Posts

  • SCRA-Only Effect Posts and Frequent Effect Posts, each as a proportion of SCRA-Only Posts

  • Frequent Effect Posts as a proportion of SCRA-Only Effect Posts

  • Individual Effect Posts, each as a proportion of SCRA-Only Posts

  • Individual Effect Posts, each as a proportion of Frequent Effect Posts

Time trends for these 24 proportions were estimated using generalized linear regression (Pinheiro & Bates, 2006) without assuming a linear trend; rather, time was modeled via a restricted cubic spline function (Devlin & Weeks, 1986) with four knots. The model was fit using the RMS and NLME packages in R 3.3.1 (Harrell, 2016; Pinheiro, Bates, DebRoy, & Sarkar, 2016; R Core Team, 2016), with knots at the 5th, 33rd, 67th, and 95th quantiles of the time distribution (05/2008, 09/2010, 01/2013, and 05/2015). For each of the 24 outcomes, the null hypothesis of no time trend was tested using a 3 d.f. Wald test comparing the spline fit to a flat trend. Adjustment for multiple testing across the 24 tests was done via the Hommel procedure (Hommel, 1989) in SAS PROC MULTTEST (SAS, 2010) to preserve a familywise α = 0.05 Type I error rate. Individual p-values are presented as Hommel-adjusted (HA) p-values.

Results

Data extraction

The eDrugTrends platform collected a total of 19,823,726 drug-related posts (Bluelight: 1,558,832; Forum A: 18,170,129; Forum B: 94,765) written by 67,841 forums users. Forum users were identified by their unique ID serial for each forum. Counting users by their forum ID implies that (a) one individual could create several accounts in a web-forum and s/he will be considered as several different “forum users”; (b) the same individual will be identified as different users across several forums. Our geolocation data remain partial as drug-focused Web forum users are not generally inclined to reveal their country of residence and because Forum B does not display any location field for users. Aggregate data provide by Bluelight indicate that the majority of their users are located in the U.S (61.9%), followed by U.K (10.9%), Canada (5.6%), Australia (4.6%), Germany (1.4%), and Netherlands (0.9%). Data extracted from Forum A publically displayed user profiles also indicate that most of users are geolocated within the U.S (52.1%), followed by U.K (19.5 %), Canada (5.1%), Ireland (3.3%), Australia (2.9%), Netherlands (2.3%), and Germany (1.5%).

The extracted data covered the time period from 01/01/2008 until the 09/30/2015. Out of the total 19,823,726 posts, 45,163 (0.22%) were SCRA-related posts (Bluelight (n=8,254), Forum A (n=9,271), and Forum B (n=25,981)) posted by 5,468 users (12.88% of all Web forum users). We then isolated posts only mentioning SCRA (and no other drugs of abuse), reducing the number of posts to a total of 19,052 (Bluelight: n=2,782; Forum A: n=3,882; Forum B: n=12,388) from 2,543 users (3.7% of all Web forum users). Among the 19,052 posts, 5,160 posts (27.1%) posted by 1,281 users (50.9% of users that posted SCRA-only posts) contained at least one mention of one of the 501 effect-related terms indexed in the DAO. Feeling “High” appeared to be the most common effect mentioned in SCRA-only posts (43.2%, n=2,155), followed by “Hallucinations” (11.2%, n=560), “Anxiety” (9.4%, n=470), “Overdose” (6.2%, n=307), “Euphoria” (5.7%, n=286), “Seizures” (5.4%, n=267), and “Nausea” (4.9%, n=246) (see examples in Table 1). The effect categories, associated search terms, and results of the frequency analysis are presented in Table 2.

Table 1.

Examples of SCRA effects mentioned in web-based forums.

High:
“My Trip Report on JWH-018 is like, a good example of being super high. I'm high on JWH-018 at the moment, and I'm really good.”
Hallucinations:
“I love psychedelics and would consider myself a happy and experienced tripper but the trip I got from K2 was the worst trip that I have ever experienced. No connections with reality, I couldn't remember who I was, thought I was living in a post-apocalyptic world. Vomit everywhere. My buddy told me that my lips went blue. Visibly, I was shaking and was unable to move, kept asking who I was, who they were.”
Anxiety:
“After smoking spice, I experienced depersonalization…. I hated it and it was intense and… I ended up in the ER, having a panic attack.”
Euphoria:
“I made a very tiny rolled joint, approximatively 75mg, and I had effects much milder than last night. I had a head rush for about 2-3mins, then sober up in a relaxed state.”
Nausea:
“My friend got too high from that stuff and it was one of the scariest thing I've ever witnessed. He was screaming and begging for an ambulance, sweating alot, crying, afraid, and was vomiting for over a couple of hours.”
Overdose:
“I blacked out and overdosed while throwing up and convulsing on K2. I was stupid and changed to another type and blacked out one more time. That crap is Fucking evil. I will never do that again.”
Seizure:
“I was hooked on spice for about a year back in 2010. Lost my voice from smoking it too much and was shoplifting after I sold every games and BluRay I had. I never allowed myself to buy the pure chemical because I had 3 seizures before because of that. One of those seizures let me 3 days in a coma and I ended up with severe amnesia I waste about 10 months of my life due to that stuff.”
Open in a new tab

Table 2. Frequencies of effects mentioned in SCRA-only posts (n=19,052).

Effects Searched Terms N (Percentage)
High high*, stoned, “headrush” 2155 (43.2%)
Hallucination trip, tripping, hallucinating, visual distortions, hearing things, time dilatation, psychedelic 560 (11.2%)
Anxiety anxiety, panic attack, anxious, anxiety attack 470 (10.2%)
Euphoria euphoric, euphoria, rush, buzz, bliss 286 (5.7%)
Nausea nausea, vomiting, threw up, puking, vomit, throwing up, puke, throw up, vomited 250 (4.9%)
Overdose overdose, od, overdosed, oded, od's 307 (6.2%)
Seizure seizure, seizures** 267 (5.4%)
Open in a new tab
*

Terms such as “too high”, “high dose”, “high temperature”, and “high amount” were removed from the query to ensure more accurate results relative to this search term.

**

To avoid confusion with law enforcement seizures, posts containing mention of “border”, “custom”, or “DEA” were removed.

Longitudinal analysis

Even after adjusting for multiple testing, all but three of the 24 trends were statistically significant (cf. Table 3). For brevity, we present specific results only for trends in the following nine proportions: SCRA-Only Posts per Post, Frequent Effect Posts per SCRA-Only Effect Post, and each of the seven Individual Effect Posts as a proportion of SCRA-Only Posts.

Table 3. Results of trend analysis.

Test # Proportion Raw p-value P-value after multiple comparisons adjustment
1 Frequent Effects (% of all posts) <.0001 <.0001
2 Any Effect (% of all posts) <.0001 <.0001
3 SCRA-Only Posts (% of all posts) <.0001 <.0001 Figure 1
4 SCRA Posts (% of all posts) <.0001 <.0001
5 Frequent Effects (% of SCRA posts) <.0001 <.0001
6 Any Effect (% of SCRA posts) <.0001 <.0001
7 SCRA-Only Posts (% of SCRA posts) <.0001 <.0001
8 Frequent Effects (% of SCRA-only posts) 0.0076 0.0305 Figure 2
9 Any Effect (% of SCRA-only posts) 0.0001 0.0008
10 Frequent Effects (% of any effect posts) 0.7909 0.7909
11 anxiety (% of frequent effects) 0.0003 0.0019
12 euphoria (% of frequent effects) <.0001 <.0001
13 high (% of frequent effects) 0.0007 0.0034
14 nausea (% of frequent effects) <.0001 <.0001
15 overdose (% of frequent effects) 0.0179 0.0536
16 seizure (% of frequent effects) <.0001 <.0001
17 hallucination (% of frequent effects) 0.0002 0.0013
18 anxiety (% of SCRA-only posts) 0.0002 0.0013 Figure 3
19 euphoria (% of SCRA-only posts) <.0001 <.0001 Figure 3
20 high (% of SCRA-only posts) 0.0006 0.0030 Figure 3
21 nausea (% of SCRA-only posts) <.0001 <.0001 Figure 3
22 overdose (% of SCRA-only posts) <.0001 <.0001 Figure 3
23 seizure (% of SCRA-only posts) <.0001 <.0001 Figure 3
24 hallucination (% of SCRA-only posts) 0.0749 0.1497 Figure 3
Open in a new tab

The percentage of SCRA-only posts out of the total number of collected posts steadily decreased from about 0.425% in January 2008 to 0.055% in September 2015 (HA p < 0.0001; see Figure 1). Of these SCRA-only posts, the proportion mentioning at least one of the seven most frequently mentioned effects also decreased, from about 20% in January 2008 to 16% in September 2015 (HA p = 0.0305; see Figure 2).

Figure 1.

Figure 1

Open in a new tab

Proportion of all posts (“Posts”) that mention SCRA but not other drugs (“SCRA-Only Posts”) (January 1, 2008 to September 30, 2015) (solid line = estimated mean, dotted lines = 95% confidence interval for the mean)

Figure 2.

Figure 2

Open in a new tab

Proportion of SCRA-Only Posts mentioning at least one of the seven most frequently mentioned effects (“Frequent Effect Posts”) (January 1, 2008 to September 30, 2015) (solid line = estimated mean, dotted lines = 95% confidence interval for the mean)

The longitudinal analysis of changes over time in individual SCRA-related effect mentions reveals disparities within the seven most frequently mentioned effects (see Figure 3). Although “High” remains the predominant mentioned effect, the proportions of mentions of “High” (HA p = 0.0030) and “Euphoria” (HA p < 0.0001) each continuously decreased between January 2008 and September 2015. In contrast, mentions of “Seizure” sharply increased over the same time (HA p < 0.0001). “Overdose” (HA p < 0.0001) and “Anxiety” (HA p = 0.0013) trends displayed a sawtooth pattern, increasing through 2010, then decreasing through 2013, and finally slightly increasing again from 2014 to September 2015. Posts mentioning “Nausea” (HA p < 0.0001) increased through 2013, but sharply decreased thereafter. Mentions of “Hallucination” increased through 2012 before plateauing until the end of the data collection period; however, after adjusting for multiple testing, this trend was not significantly different from no change over the entire period (HA p = 0.1497).

Figure 3.

Figure 3

Open in a new tab

Proportion* of SCRA-Only Posts mentioning each of the seven most frequently mentioned effects (January 1, 2008 to September 30, 2015) (solid line = estimated mean, dotted lines = 95% confidence interval for the mean)

*The scales were chosen to highlight the shape of each trend rather than the magnitude of the proportions. If plotted on a scale from 0% to 100%, trends in rare side effects would not be visible. Thus, not all figures have the same vertical scaling.

These results indicate that SCRA-related effects that are perceived as positive (i.e. “High” and “Euphoria”) have decreased since 2008, while “negative” effects (i.e. “Nausea”, “Seizure”, “Anxiety”, and “Overdose”) have increased over the same period (although mentions of “Nausea” are on the decline).

Discussion

To the best of our knowledge, this study is the first to evaluate the frequency and trends of effects associated with SCRA use based on web-forum data. Our analysis shows that the proportion of drug-related posts mentioning at least one SCRA and no other drug has consistently decreased since 2008 without completely disappearing. The most frequent categories of effects mentioned in SCRA-only posts are “High,” “Hallucinations,” “Anxiety,” “Euphoria,” “Nausea,” “Overdose,” and “Seizure.” These types of SCRA-related effects are consistent with the effects described by ER patients in the literature (Derungs, et al., 2013; Tait, et al., 2016): for example, gastrointestinal symptoms such as nausea and vomiting (Banerji, Deutsch, & Bronstein, 2010; Bozkurt, Umut, Evren, & Karabulut, 2014; Fattore & Fratta, 2011; Forrester, Kleinschmidt, Schwarz, & Young, 2011; Freeman, et al., 2013; Gugelmann, et al., 2014; Heath, Burroughs, Thompson, & Tecklenburg, 2012; Heltsley, et al., 2012; Hermanns-Clausen, Kneisel, Hutter, Szabo, & Auwärter, 2013; Hermanns-Clausen, Kneisel, Szabo, & Auwärter, 2013; Hopkins & Gilchrist, 2013; McQuade, Hudson, Dargan, & Wood, 2013; Schneir, et al., 2011; Simmons, et al., 2011; Tofighi & Lee, 2012; Vearrier & Osterhoudt, 2010), anxiety (Benford & Caplan, 2011; Castellanos, et al., 2011; J. Cohen, et al., 2012; Control & Prevention, 2013; Every-Palmer, 2011; Glue, et al., 2013; Harris & Brown, 2013), hallucinations (Auwärter, et al., 2009; Castellanos, et al., 2011; J. Cohen, et al., 2012), and seizures (Harris & Brown, 2013; Hermanns-Clausen, Kneisel, Hutter, et al., 2013; Hermanns-Clausen, Kneisel, Szabo, et al., 2013; Lapoint, et al., 2011; Schneir, et al., 2011; Vearrier & Osterhoudt, 2010) have been reported in several clinical research and emergency department reports.

The decrease of SCRA-related posts echoes the decrease of SCRA consumption observed over the past few years in the countries from where most of the Web forums users are geolocated (U.S. (Johnston, Miech, O'Malley, Bachman, & Schulenberg, 2015), U.K (U.K Home Office, 2016), Australia (Sutherland, et al., 2016)). Our longitudinal analysis demonstrates that positive effect (i.e., “High” and “Euphoria”) proportions continuously decreased since 2008. In contrast, the proportion of negative effects grouped under the categories “Nausea,” “Seizure,” “Anxiety,” and “Overdose” have increased during the same period. We are unaware of another study investigating the variations of SCRA-related effects over time. However, our results are consistent with the qualitative analysis Norwegian Web forum data of Bilgrei (2016) that demonstrated how the opinion of forum users gradually changed from an “initial attractiveness” in late 2007, to “ambivalent opinion” in late 2008/early 2009, to finally become a “communal rejection” thereafter, mainly because the number of negative effects overweighed the number of positive effects over time (Bilgrei, 2016).

The “overweighing” proportion of negative effects described by Bilgrei and the increased proportions of negative effects we observed could be explained by the constant and frequent changes in the SCRA molecules available on the drug market (J. A. Cohen, 2014; Lindigkeit, et al., 2009). Over the last eight years, despite the successive administrative bans, illicit SCRA manufacturers have continuously introduced new SCRA molecules on the market (J. A. Cohen, 2014). At least three “generations” of SCRA (Fattore & Fratta, 2011; Shanks, Behonick, Dahn, & Terrell, 2013; Shanks, Dahn, Behonick, & Terrell, 2012) have succeeded each other since the first administrative ban early 2009 for several European countries (UK Statutory Instrument, 2009) and on March 1, 2011 in the U.S. (Drug Enforcement Administration, 2016). 2013-2015 drug seizure data from NFLIS (National Forensic Laboratory Information System) indicate that first generation (e.g., JWH-018, JWH-073, JWH-250) and second generation molecules (e.g., JWH-122, XLR11, UR-144) were either not seized or were on the decline, while the number of seizures of third generation of SCRA (e.g., FUBINACA, AB-CHMINACA, MAB-CHMINACA) were on the rise. These data suggest that new generation of SCRA have become increasingly available on the drug market in contrast with banned molecules which tend to recede over time (Drug Enforcement Administration, 2016). The latest generation of SCRA (e.g., CHMINACA, FUBINACA) displays increased endocannabinoid CB1 and CB2 receptor affinities compared to the previous generations (Aung, et al., 2000; Franz, Angerer, Moosmann, & Auwärter, 2016; Iversen, 2015; Labay, et al., 2016; Uchiyama, Kawamura, Kikura-Hanajiri, & Goda, 2013; Wiley, et al., 2015), which means that, at the same dose, the third generation SCRA induce stronger reactions and, in turn, increase the risks of harm. Users are more likely to use higher than necessary doses because they are not properly informed regarding the type and potency of the higher affinity SCRA, especially while smoking “herbal blends”, as the substances sprayed on inert materials have changed in composition or molecular structure following administrative bans. Thus, an unintended consequence of these administrative bans appears to have been the introduction of more potent SCRA molecules in the drug market (Iversen, 2015; Reuter & Pardo, 2016), leading the remaining population of SCRA users to experience more negative outcomes. In turn, the deterrent effects inherent in the higher potency of newer generations of SCRA may have induced a diminution in SCRA use as observed by Bilgrei (Bilgrei, 2016) among Norwegian Web forum users.

Our study illustrates the potential to collect useful epidemiologic information by harnessing web-forum social data and how monitoring emerging NPS for harmful effects using Web forum content analysis can inform public health professionals. The Natural Language Processing techniques developed for this study allow extending inductively the number of relevant expressions describing effects.

Despite these findings, this study has several limitations. Although little is known about the population of SCRA users, the population of web-forum users is not necessarily representative of the general population of SCRA users in any country. Several newspapers articles (Doward, 2015; Karlamangla, 2016; Thompson, 2016) and a scientific paper (Nolan, Allen, Kunins, & Paone, 2016) have described an increased number of ER presentations in the population of homeless due to the consumption of SCRA in the past few years, and these populations are not likely to be represented in web-forum populations. Other limitations are inherent to the field of Internet-based research. Although our findings are consistent with trends and results of other research and reports, by focusing on keywords and not on the whole post text, only mentions of keywords were captured without the possibility to ensure direct relationships between SCRA and identified effects. Our semantic approach involving use of Drug Abuse Ontology and Natural Language Processing techniques was targeted to alleviate this limitation. Second, our current approach does not allow differentiating effects mentioned as the report of self-experienced effects, witnessed in other users, or part of a more general discussion. Because this latter possibility could potentially amplify the proportions of effects mentioned in the total number of posts, our future work will develop a classification of posts by type of communication to increase the precision of this method. Third, by limiting our analysis to posts containing only SCRA and no other drugs, we have potentially missed some frequent effects; users frequently compare the effects of one substance with another, for example, comparing the withdrawal linked to SCRA to heroin withdrawal. Fourth, this analysis did not compare the effects of SCRA in powder form versus effects induced by “herbal blend” or e-liquid as not all users specify which form of SCRA they were using. Fifth, it is unclear if these increases are related to immediate effects induced by SCRA or if long-time users are discussing their past experiences while sharing their opinions with other forum users. Sixth, our users' geolocation data are only partial making difficult to deduce accurately the correlation between negative effects and SCRA generations as the administrative bans starting dates vary from country to country. Finally, the most frequent effects described in this study, as well as in the literature, are short-term effects. Long-term effects remain unknown at the moment and require further research.

Supplementary Material

s1

Acknowledgments

This material is based in part on datasets obtained from Bluelight.org. Bluelight.org is a nonprofit online community dedicated to reducing drug-related harm. This work was conducted with support from Bluelight.org. The content is solely the responsibility of the authors and does not necessarily represent the official views of Bluelight.org.

Role of funding source: This study was supported by the National Institute on Drug Abuse (NIDA), Grant No. R01 DA039454-01 (Daniulaityte, PI; Sheth, PI). The funding source had no further role in the study design, in the collection, analysis and interpretation of the data, in the writing of the report, or in the decision to submit the paper for publication.

Monica Barratt is supported by a fellowship from the NHMRC (APP1070140). The National Drug and Alcohol Research Centre and the National Drug Research Institute are supported by funding from the Australian Government under the Substance Misuse Prevention and Service Improvement Grants Fund. We also acknowledge the contribution of the Victorian Operational Infrastructure Support Program received by the Burnet Institute.

Footnotes

Conflict of Interest Statement: Monica Barratt is the Director of Research at Bluelight.org, a not-for-profit harm-reduction website.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. AAPCC. Synthetic Cannabonoid Data. American Association of Poison Control Centers 2016 [Google Scholar]
  2. Aung MM, Griffin G, Huffman JW, Wu MJ, Keel C, Yang B, Showalter VM, Abood ME, Martin BR. Influence of the N-1 alkyl chain length of cannabimimetic indoles upon CB 1 and CB 2 receptor binding. Drug and Alcohol Dependence. 2000;60:133–140. doi: 10.1016/s0376-8716(99)00152-0. [DOI] [PubMed] [Google Scholar]
  3. Auwärter V, Dresen S, Weinmann W, Müller M, Pütz M, Ferreirós N. ‘Spice’and other herbal blends: harmless incense or cannabinoid designer drugs? Journal of Mass Spectrometry. 2009;44:832–837. doi: 10.1002/jms.1558. [DOI] [PubMed] [Google Scholar]
  4. Banerji S, Deutsch C, Bronstein A. Spice ain't so nice. Clinical Toxicology. 2010;48:632. [Google Scholar]
  5. Barratt MJ, Cakic V, Lenton S. Patterns of synthetic cannabinoid use in Australia. Drug and Alcohol Review. 2013;32:141–146. doi: 10.1111/j.1465-3362.2012.00519.x. [DOI] [PubMed] [Google Scholar]
  6. Benford DM, Caplan JP. Psychiatric sequelae of Spice, K2, and synthetic cannabinoid receptor agonists. Psychosomatics. 2011;52:295. doi: 10.1016/j.psym.2011.01.004. [DOI] [PubMed] [Google Scholar]
  7. Besli GE, Ikiz MA, Yildirim S, Saltik S. Synthetic cannabinoid abuse in adolescents: a case series. The Journal of Emergency Medicine. 2015;49:644–650. doi: 10.1016/j.jemermed.2015.06.053. [DOI] [PubMed] [Google Scholar]
  8. Bhanushali GK, Jain G, Fatima H, Leisch LJ, Thornley-Brown D. Clinical Journal of the American Society of Nephrology. CJN; 2012. AKI associated with synthetic cannabinoids: a case series; p. 05690612. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bilgrei OR. From “herbal highs” to the “heroin of cannabis”: Exploring the evolving discourse on synthetic cannabinoid use in a Norwegian Internet drug forum. International Journal of Drug Policy. 2016;29:1–8. doi: 10.1016/j.drugpo.2016.01.011. [DOI] [PubMed] [Google Scholar]
  10. Bird S, Klein E, Loper E. Natural language processing with Python. O'Reilly Media, Inc; 2009. [Google Scholar]
  11. Bonar EE, Ashrafioun L, Ilgen MA. Synthetic cannabinoid use among patients in residential substance use disorder treatment: prevalence, motives, and correlates. Drug and Alcohol Dependence. 2014;143:268–271. doi: 10.1016/j.drugalcdep.2014.07.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bozkurt M, Umut G, Evren C, Karabulut V. Clinical characteristics and laboratory test results of patients admitted to outpatient clinic for synthetic cannabinoid usage. Dusunen Adam. 2014;27:328. [Google Scholar]
  13. Brents LK, Prather PL. The K2/Spice Phenomenon: emergence, identification, legislation and metabolic characterization of synthetic cannabinoids in herbal incense products. Drug Metabolism Reviews. 2014;46:72–85. doi: 10.3109/03602532.2013.839700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Bush DM, Woodwell DA. Update: drug-related emergency department visits involving synthetic cannabinoids. Network, Drug Abuse Warning. 2014 [PubMed] [Google Scholar]
  15. Butler SF, Venuti SW, Benoit C, Beaulaurier RL, Houle B, Katz N. Internet surveillance: content analysis and monitoring of product-specific internet prescription opioid abuse-related postings. The Clinical Journal of Pain. 2007;23:619–628. doi: 10.1097/AJP.0b013e318125c5cf. [DOI] [PubMed] [Google Scholar]
  16. Cameron D, Smith GA, Daniulaityte R, Sheth AP, Dave D, Chen L, Anand G, Carlson R, Watkins KZ, Falck R. Predose: A semantic Web platform for drug abuse epidemiology using social media. Journal of Biomedical Informatics. 2013;46:985–997. doi: 10.1016/j.jbi.2013.07.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Castellanos D, Singh S, Thornton G, Avila M, Moreno A. Synthetic cannabinoid use: a case series of adolescents. Journal of Adolescent Health. 2011;49:347–349. doi: 10.1016/j.jadohealth.2011.08.002. [DOI] [PubMed] [Google Scholar]
  18. Caviness CM, Tzilos G, Anderson BJ, Stein MD. Synthetic cannabinoids: use and predictors in a community sample of young adults. Substance Abuse. 2015;36:368–373. doi: 10.1080/08897077.2014.959151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Centers for Disease Control Prevention (CDC) Acute kidney injury associated with synthetic cannabinoid use-multiple states, 2013. MMWR Morbidity and Mortality Weekly Report. 2013;62:93. [PMC free article] [PubMed] [Google Scholar]
  20. Centers for Disease Control Prevention (CDC) Notes from the field: Severe illness associated with synthetic cannabinoid use-Brunswick, Georgia, 2013. MMWR Morbidity and Mortality Weekly Report. 2013;62:939. [PMC free article] [PubMed] [Google Scholar]
  21. Cohen J, Morrison S, Greenberg J, Saidinejad M. Clinical presentation of intoxication due to synthetic cannabinoids. Pediatrics. 2012;129:e1064–e1067. doi: 10.1542/peds.2011-1797. [DOI] [PubMed] [Google Scholar]
  22. Cohen JA. Highs of Tomorrow: Why New Laws and Policies Are Needed to Meet the Unique Challenges of Synthetic Drugs, The. JL & Health. 2014;27:164. [Google Scholar]
  23. Damerau FJ. A technique for computer detection and correction of spelling errors. Communications of the ACM. 1964;7:171–176. [Google Scholar]
  24. Daniulaityte R, Carlson R, Brigham G, Cameron D, Sheth A. “Sub is a weird drug”: A web-based study of lay attitudes about use of buprenorphine to self-treat opioid withdrawal symptoms. The American Journal on Addictions. 2015;24:403–409. doi: 10.1111/ajad.12213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Daniulaityte R, Carlson R, Falck R, Cameron D, Perera S, Chen L, Sheth A. “I just wanted to tell you that loperamide WILL WORK”: A web-based study of extra-medical use of loperamide. Drug and Alcohol Dependence. 2013;130:241–244. doi: 10.1016/j.drugalcdep.2012.11.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Daniulaityte R, Nahhas RW, Wijeratne S, Carlson RG, Lamy FR, Martins SS, Boyer EW, Smith GA, Sheth A. “Time for dabs”: Analyzing Twitter data on marijuana concentrates across the US. Drug and Alcohol Dependence. 2015;155:307–311. doi: 10.1016/j.drugalcdep.2015.07.1199. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Derungs A, Schwaninger AE, Mansella G, Bingisser R, Kraemer T, Liechti ME. Symptoms, toxicities, and analytical results for a patient after smoking herbs containing the novel synthetic cannabinoid MAM-2201. Forensic Toxicology. 2013;31:164–171. [Google Scholar]
  28. Devlin T, Weeks B. Proceedings of the 11th Annual SAS Users Group International Conference. Vol. 646. SAS Institute Inc; Cary, NC: 1986. Spline functions for logistic regression modeling; p. 51. [Google Scholar]
  29. Doward J. Plague of cheap ‘spice’ strengthens its grip on homeless and jobless in northern towns. The Guardian 2015 [Google Scholar]
  30. Drug Enforcement Administration. Schedules of Controlled Substances: Temporary Placement of Five Synthetic Cannabinoids Into Schedule I. Springfield, VA: 2016. [Google Scholar]
  31. Drug Enforcement Administration. Special Report: Synthetic Cannabinoids and Synthetic Cathinones Reported in NFLIS, 2013–2015. Springfield, VA: 2016. [Google Scholar]
  32. eDrugTrends. Project Information. Retrieved 11/05 2015 from http://wiki.knoesis.org/index.php/EDrugTrends.
  33. P. O. o. t. E. Union, editor. EMCDDA. New psychoactive substances in Europe An update from the EU Early Warning System (March 2015) Luxembourg: European Monitoring Centre for Drugs and Drug Addiction; 2015. [Google Scholar]
  34. European Monitoring Centre for Drugs and Drug Addiction. Perspectives On Drugs: Synthetic cannabinoids in Europe. Luxembourg: Publications Office of the European Union; 2013. [Google Scholar]
  35. Every-Palmer S. Synthetic cannabinoid JWH-018 and psychosis: an explorative study. Drug and Alcohol Dependence. 2011;117:152–157. doi: 10.1016/j.drugalcdep.2011.01.012. [DOI] [PubMed] [Google Scholar]
  36. Eysenbach G. Infodemiology and infoveillance tracking online health information and cyberbehavior for public health. American Journal of Preventive Medicine. 2011;40:S154–158. doi: 10.1016/j.amepre.2011.02.006. [DOI] [PubMed] [Google Scholar]
  37. Fattore L, Fratta W. Beyond THC: the new generation of cannabinoid designer drugs. Frontiers in Behavioral Neuroscience. 2011;5 doi: 10.3389/fnbeh.2011.00060. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Forrester MB. Adolescent synthetic cannabinoid exposures reported to Texas poison centers. Pediatric emergency care. 2012;28:985–989. doi: 10.1097/PEC.0b013e31826c9a97. [DOI] [PubMed] [Google Scholar]
  39. Forrester MB, Kleinschmidt K, Schwarz E, Young A. Synthetic cannabinoid exposures reported to Texas poison centers. Journal of Addictive Diseases. 2011;30:351–358. doi: 10.1080/10550887.2011.609807. [DOI] [PubMed] [Google Scholar]
  40. Franz F, Angerer V, Moosmann B, Auwärter V. Phase I metabolism of the highly potent synthetic cannabinoid MDMB-CHMICA and detection in human urine samples. Drug Testing and Analysis. 2016 doi: 10.1002/dta.2049. [DOI] [PubMed] [Google Scholar]
  41. Freeman MJ, Rose DZ, Myers MA, Gooch CL, Bozeman AC, Burgin WS. Ischemic stroke after use of the synthetic marijuana “spice”. Neurology. 2013;81:2090–2093. doi: 10.1212/01.wnl.0000437297.05570.a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ghcmarijuana. Retrieved 10/27/2016 from http://www.ghcmarijuana.com/herbal_incence.html.
  43. Glue P, Al-Shaqsi S, Hancock D, Gale C, Strong B, Schep L. Hospitalisation associated with use of the synthetic cannabinoid K2. The New Zealand Medical Journal (Online) 2013;126 [PubMed] [Google Scholar]
  44. Grigoryev A, Savchuk S, Melnik A, Dzhurko J, Ershov M, Nosyrev A, Vedenin A, Izotov B, Zabirova I, Rozhanets V. Chromatography–mass spectrometry studies on the metabolism of synthetic cannabinoids JWH-018 and JWH-073, psychoactive components of smoking mixtures. Journal of Chromatography B. 2011;879:1126–1136. doi: 10.1016/j.jchromb.2011.03.034. [DOI] [PubMed] [Google Scholar]
  45. Gugelmann H, Gerona R, Li C, Tsutaoka B, Olson K, Lung D. ‘Crazy Monkey’Poisons Man and Dog: Human and canine seizures due to PB-22, a novel synthetic cannabinoid. Clinical Toxicology. 2014;52:635–638. doi: 10.3109/15563650.2014.925562. [DOI] [PubMed] [Google Scholar]
  46. Harrell FEJ. rms: Regression Modeling Strategies. R package version 4.5-0 2016 [Google Scholar]
  47. Harris CR, Brown A. Synthetic cannabinoid intoxication: a case series and review. The Journal of Emergency Medicine. 2013;44:360–366. doi: 10.1016/j.jemermed.2012.07.061. [DOI] [PubMed] [Google Scholar]
  48. Heath TS, Burroughs Z, Thompson AJ, Tecklenburg FW. Acute intoxication caused by a synthetic cannabinoid in two adolescents. The Journal of Pediatric Pharmacology and Therapeutics. 2012;17:177–181. doi: 10.5863/1551-6776-17.2.177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Heltsley R, Shelby MK, Crouch DJ, Black DL, Robert TA, Marshall L, Bender CL, DePriest AZ, Colello MA. Prevalence of synthetic cannabinoids in US athletes: initial findings. Journal of Analytical Toxicology. 2012;36:588–593. doi: 10.1093/jat/bks066. [DOI] [PubMed] [Google Scholar]
  50. Hermanns-Clausen M, Kneisel S, Hutter M, Szabo B, Auwärter V. Acute intoxication by synthetic cannabinoids–four case reports. Drug Testing and Analysis. 2013;5:790–794. doi: 10.1002/dta.1483. [DOI] [PubMed] [Google Scholar]
  51. Hermanns-Clausen M, Kneisel S, Szabo B, Auwärter V. Acute toxicity due to the confirmed consumption of synthetic cannabinoids: clinical and laboratory findings. Addiction. 2013;108:534–544. doi: 10.1111/j.1360-0443.2012.04078.x. [DOI] [PubMed] [Google Scholar]
  52. Hommel G. A comparison of two modified Bonferroni procedures. Biometrika. 1989;75:383–386. [Google Scholar]
  53. Hopkins CY, Gilchrist BL. A case of cannabinoid hyperemesis syndrome caused by synthetic cannabinoids. The Journal of Emergency Medicine. 2013;45:544–546. doi: 10.1016/j.jemermed.2012.11.034. [DOI] [PubMed] [Google Scholar]
  54. Hu X, Primack BA, Barnett TE, Cook RL. College students and use of K2: an emerging drug of abuse in young persons. Substance Abuse Treatment, Prevention, and Policy. 2011;6:1. doi: 10.1186/1747-597X-6-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Huffman JW, Padgett LW. Recent developments in the medicinal chemistry of cannabimimetic indoles, pyrroles and indenes. Current Medicinal Chemistry. 2005;12:1395–1411. doi: 10.2174/0929867054020864. [DOI] [PubMed] [Google Scholar]
  56. Iversen L. Oral evidence: Psychoactive substances, HC 361. Sect. London, UK: Home Affairs Committee House of Commons; 2015. [Google Scholar]
  57. Johnson LA, Johnson RL, Portier RB. Current “legal highs”. The Journal of Emergency Medicine. 2013;44:1108–1115. doi: 10.1016/j.jemermed.2012.09.147. [DOI] [PubMed] [Google Scholar]
  58. Johnston LD, Miech RA, O'Malley PM, Bachman JG, Schulenberg JE. Use of ecstasy, heroin, synthetic marijuana, alcohol, cigarettes declined among US teens in 2015. Ann Arbor, MI: University of Michigan News Service; 2015. [Google Scholar]
  59. Karlamangla S. LA Times. Los Angeles, CA: 2016. ‘The cheapest buzz you can get on skid row’: Officials try to stop homeless from smoking spice after dozens sickened. [Google Scholar]
  60. Kavanagh P, Grigoryev A, Savchuk S, Mikhura I, Formanovsky A. UR-144 in products sold via the Internet: Identification of related compounds and characterization of pyrolysis products. Drug Testing and Analysis. 2013;5:683–692. doi: 10.1002/dta.1456. [DOI] [PubMed] [Google Scholar]
  61. Kazory A, Aiyer R. Synthetic marijuana and acute kidney injury: an unforeseen association. Clinical Kidney Journal. 2013;6:330–333. doi: 10.1093/ckj/sft047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Labay LM, Caruso JL, Gilson TP, Phipps RJ, Knight LD, Lemos NP, McIntyre IM, Stoppacher R, Tormos LM, Wiens AL. Synthetic cannabinoid drug use as a cause or contributory cause of death. Forensic Science International. 2016;260:31–39. doi: 10.1016/j.forsciint.2015.12.046. [DOI] [PubMed] [Google Scholar]
  63. Lamy FR, Daniulaityte R, Sheth A, Nahhas RW, Martins SS, Boyer EW, Carlson RG. “Those edibles hit hard”: Exploration of Twitter data on cannabis edibles in the US. Drug and Alcohol Dependence. 2016;164:64–70. doi: 10.1016/j.drugalcdep.2016.04.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Lapoint J, James L, Moran C, Nelson L, Hoffman R, Moran J. Severe toxicity following synthetic cannabinoid ingestion. Clinical Toxicology. 2011;49:760–764. doi: 10.3109/15563650.2011.609822. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Lazer D, Pentland AS, Adamic L, Aral S, Barabasi AL, Brewer D, Christakis N, Contractor N, Fowler J, Gutmann M. Life in the network: the coming age of computational social science. Science (New York, NY) 2009;323:721. doi: 10.1126/science.1167742. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Levenshtein VI. Binary codes capable of correcting deletions, insertions and reversals. Soviet Physics Doklady. 1966;10:707. [Google Scholar]
  67. Lindigkeit R, Boehme A, Eiserloh I, Luebbecke M, Wiggermann M, Ernst L, Beuerle T. Spice: a never ending story? Forensic Science International. 2009;191:58–63. doi: 10.1016/j.forsciint.2009.06.008. [DOI] [PubMed] [Google Scholar]
  68. McQuade D, Hudson S, Dargan PI, Wood DM. First European case of convulsions related to analytically confirmed use of the synthetic cannabinoid receptor agonist AM-2201. European Journal of Clinical Pharmacology. 2013;69:373–376. doi: 10.1007/s00228-012-1379-2. [DOI] [PubMed] [Google Scholar]
  69. Miller PG, Sonderlund AL. Using the internet to research hidden populations of illicit drug users: a review. Addiction (Abingdon, England) 2010;105:1557–1567. doi: 10.1111/j.1360-0443.2010.02992.x. [DOI] [PubMed] [Google Scholar]
  70. Miller PG, Sønderlund AL. Using the internet to research hidden populations of illicit drug users: a review. Addiction. 2010;105:1557–1567. doi: 10.1111/j.1360-0443.2010.02992.x. [DOI] [PubMed] [Google Scholar]
  71. Nolan ML, Allen B, Kunins H, Paone D. A public health approach to increased synthetic cannabinoid-related morbidity among New York City residents, 2014-2015. International Journal of Drug Policy. 2016;34:101–103. doi: 10.1016/j.drugpo.2016.05.014. [DOI] [PubMed] [Google Scholar]
  72. OrganicSpiceBlend. Retrieved 10/27/2016 from http://www.organicspiceblend.com/
  73. Papaseit E, Farré M, Schifano F, Torrens M. Emerging drugs in Europe. Current Opinion in Psychiatry. 2014;27:243–250. doi: 10.1097/YCO.0000000000000071. [DOI] [PubMed] [Google Scholar]
  74. Pinheiro J, Bates D. Mixed-effects models in S and S-PLUS. Springer Science & Business Media; 2006. [Google Scholar]
  75. Pinheiro J, Bates D, DebRoy S, Sarkar D. R Core Team (2016) nlme: linear and nonlinear mixed effects models. R package version 3.1-117 2016 [Google Scholar]
  76. R Core Team. R Foundation for Statistical Computing. Vienna, Austria: 2016. R: A language and environment for statistical computing. [Google Scholar]
  77. Reuter P, Pardo B. New psychoactive substances: Are there any good options for regulating new psychoactive substances? International Journal of Drug Policy. 2016 doi: 10.1016/j.drugpo.2016.10.020. [DOI] [PubMed] [Google Scholar]
  78. Riederer AM. Acute Poisonings from Synthetic Cannabinoids—50 US Toxicology Investigators Consortium Registry Sites, 2010–2015. MMWR Morbidity and Mortality Weekly Report. 2016;65 doi: 10.15585/mmwr.mm6527a2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Roberts LD. Ethical issues in conducting qualitative research in online communities. Qualitative Research in Psychology. 2015;12:314–325. [Google Scholar]
  80. SAS, I. I. SAS Version 9.3. Cary, NC: SAS Institute Inc; 2010. [Google Scholar]
  81. Schifano F, Corazza O, Deluca P, Davey Z, Di Furia L, Farre' M, Flesland L, Mannonen M, Pagani S, Peltoniemi T. Psychoactive drug or mystical incense? Overview of the online available information on Spice products. International Journal of Culture and Mental Health. 2009;2:137–144. [Google Scholar]
  82. Schifano F, Deluca P, Baldacchino A, Peltoniemi T, Scherbaum N, Torrens M, Farre M, Flores I, Rossi M, Eastwood D, Guionnet C, Rawaf S, Agosti L, Di Furia L, Brigada R, Majava A, Siemann H, Leoni M, Tomasin A, Rovetto F, Ghodse AH. Drugs on the web; the Psychonaut 2002 EU project. Progress in Neuropsychopharmacology & Biological Psychiatry. 2006;30:640–646. doi: 10.1016/j.pnpbp.2005.11.035. [DOI] [PubMed] [Google Scholar]
  83. Schneir AB, Cullen J, Ly BT. “Spice” girls: synthetic cannabinoid intoxication. The Journal of Emergency Medicine. 2011;40:296–299. doi: 10.1016/j.jemermed.2010.10.014. [DOI] [PubMed] [Google Scholar]
  84. Shanks KG, Behonick GS, Dahn T, Terrell A. Identification of novel third-generation synthetic cannabinoids in products by ultra-performance liquid chromatography and time-of-flight mass spectrometry. Journal of Analytical Toxicology. 2013:bkt062. doi: 10.1093/jat/bkt062. [DOI] [PubMed] [Google Scholar]
  85. Shanks KG, Dahn T, Behonick G, Terrell A. Analysis of first and second generation legal highs for synthetic cannabinoids and synthetic stimulants by ultra-performance liquid chromatography and time of flight mass spectrometry. Journal of Analytical Toxicology. 2012:bks047. doi: 10.1093/jat/bks047. [DOI] [PubMed] [Google Scholar]
  86. Simmons JR, Skinner CG, Williams J, Kang CS, Schwartz MD, Wills BK. Intoxication from smoking “spice”. Annals of Emergency Medicine. 2011;57:187–188. doi: 10.1016/j.annemergmed.2010.08.039. [DOI] [PubMed] [Google Scholar]
  87. Sutherland R, Peacock A, Whittaker E, Roxburgh A, Lenton S, Matthews A, Butler K, Nelson M, Burns L, Bruno R. New psychoactive substance use among regular psychostimulant users in Australia, 2010–2015. Drug and Alcohol Dependence. 2016;161:110–118. doi: 10.1016/j.drugalcdep.2016.01.024. [DOI] [PubMed] [Google Scholar]
  88. Tait RJ, Caldicott D, Mountain D, Hill SL, Lenton S. A systematic review of adverse events arising from the use of synthetic cannabinoids and their associated treatment. Clinical Toxicology. 2016;54:1–13. doi: 10.3109/15563650.2015.1110590. [DOI] [PubMed] [Google Scholar]
  89. Thompson T. Synthetic drugs: Battle shifts from DC stores to streets [Internet] Retrieved 16/12/2016 from http://www.nbcwashington.com/investigations/Synthetics-Battle-Shifts-From-Stores-to-Streets-332172012.html.
  90. Tofighi B, Lee JD. Internet highs—seizures after consumption of synthetic cannabinoids purchased online. Journal of Addiction Medicine. 2012;6:240–241. doi: 10.1097/ADM.0b013e3182619004. [DOI] [PubMed] [Google Scholar]
  91. Trecki J, Gerona RR, Schwartz MD. Synthetic cannabinoid-related illnesses and deaths. New England Journal of Medecine. 2015;373:103–107. doi: 10.1056/NEJMp1505328. [DOI] [PubMed] [Google Scholar]
  92. U.K Home Office. Statistical Bulletin. Second. Vol. 2. UK Home Office; 2016. Drug misuse: findings from the 2015 to 2016 CSEW. [Google Scholar]
  93. Uchiyama N, Kawamura M, Kikura-Hanajiri R, Goda Y. URB-754: a new class of designer drug and 12 synthetic cannabinoids detected in illegal products. Forensic Science International. 2013;227:21–32. doi: 10.1016/j.forsciint.2012.08.047. [DOI] [PubMed] [Google Scholar]
  94. UK Statutory Instrument. No. 3209. Dangerous drugs. The Misuse of Drugs Act 1971 (Amendment) Order 2009 2009 [Google Scholar]
  95. van Amsterdam J, Brunt T, van den Brink W. The adverse health effects of synthetic cannabinoids with emphasis on psychosis-like effects. Journal of Psychopharmacology. 2015;29:254–263. doi: 10.1177/0269881114565142. [DOI] [PubMed] [Google Scholar]
  96. van Rossum G. Python (langage) 1990 [Google Scholar]
  97. Vandrey R, Dunn KE, Fry JA, Girling ER. A survey study to characterize use of Spice products (synthetic cannabinoids) Drug and Alcohol Dependence. 2012;120:238–241. doi: 10.1016/j.drugalcdep.2011.07.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  98. Vearrier D, Osterhoudt KC. A teenager with agitation: higher than she should have climbed. Pediatric Emergency Care. 2010;26:462–465. doi: 10.1097/PEC.0b013e3181e4f416. [DOI] [PubMed] [Google Scholar]
  99. Wiley JL, Marusich JA, Lefever TW, Antonazzo KR, Wallgren MT, Cortes RA, Patel PR, Grabenauer M, Moore KN, Thomas BF. AB-CHMINACA, AB-PINACA, and FUBIMINA: Affinity and Potency of Novel Synthetic Cannabinoids in Producing Δ9-Tetrahydrocannabinol–Like Effects in Mice. Journal of Pharmacology and Experimental Therapeutics. 2015;354:328–339. doi: 10.1124/jpet.115.225326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  100. Winstock AR, Barratt MJ. The 12-month prevalence and nature of adverse experiences resulting in emergency medical presentations associated with the use of synthetic cannabinoid products. Human Psychopharmacology: Clinical and Experimental. 2013a;28:390–393. doi: 10.1002/hup.2292. [DOI] [PubMed] [Google Scholar]
  101. Winstock AR, Barratt MJ. Synthetic cannabis: a comparison of patterns of use and effect profile with natural cannabis in a large global sample. Drug and Alcohol Dependence. 2013b;131:106–111. doi: 10.1016/j.drugalcdep.2012.12.011. [DOI] [PubMed] [Google Scholar]
  102. Zawilska JB. “Legal highs” new players in the old drama. Current Drug Abuse Review. 2011;4:122–130. doi: 10.2174/1874473711104020122. [DOI] [PubMed] [Google Scholar]
  103. Zawilska JB, Andrzejczak D. Next generation of novel psychoactive substances on the horizon–A complex problem to face. Drug and Alcohol Dependence. 2015;157:1–17. doi: 10.1016/j.drugalcdep.2015.09.030. [DOI] [PubMed] [Google Scholar]
  104. Zimmermann US, Winkelmann PR, Pilhatsch M, Nees JA, Spanagel R, Schulz K. Withdrawal phenomena and dependence syndrome after the consumption of “spice gold”. Deutsch Arzteblatt International. 2009;106:464–467. doi: 10.3238/arztebl.2009.0464. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

s1
NIHMS876447-supplement-s1.docx (24.9KB, docx)

RESOURCES