Data on self-reported THC-containing EVP use behaviors among adults (aged ≥18 years) are from a web-based panel survey conducted from February 25, 2020 and February 29, 2020. Respondents (n = 3,980) were selected from the US YouGov panel, a proprietary opt-in internet panel survey of 1.8 million U.S. residents. U.S. YouGov panel members are recruited through several methods to help ensure diversity in the panel composition, including web advertising campaigns, partner sponsored solicitations, telephone to web recruitment, mail to web recruitment, and traffic to the YouGov website for polling content. YouGov survey respondents are not paid to join the panel but they receive incentives through a points-based loyalty program for taking individual surveys (Ashley Grosse, YouGov, personal communication, September 24, 2020).

The inclusion criteria for the THC-containing EVP survey were: (1) aged ≥18 years; (2) used nicotine-containing and marijuana or THC-containing EVP, or reported dabbing, (i.e., using a highly concentrated form of THC) in the past 3 months; (3) no diagnosis of probable or confirmed EVALI in the past year; and (4) resident of 1 of 18 selected states (California, Colorado, Florida, Illinois, Michigan, Minnesota, New Jersey, New York, North Carolina, North Dakota, Ohio, Oregon, Pennsylvania, Tennessee, Texas, Utah, Washington, and Wisconsin). States were selected to capture geographic diversity, burden of EVALI patients (CDC) and status of THC/marijuana legalization (National Conference of State Legislators) (medical only use legalized, medical and non-medical use legalized, and not legalized).

To protect the privacy of respondents’ and to ensure non-disclosure of reported information, the THC-containing EVP survey and the associated data were protected by an Assurance of Confidentiality stating that the information obtained will be held in strict confidence, will be used only for the purposes stated, and will not otherwise be disclosed or released without the individual’s consent. This work was determined to be public health practice by CDC during human subjects review.

Survey questions primarily focused on the use of THC-containing EVPs in the 3 months before the survey; most response options were ‘select all that apply’, therefore, percentages could add up to more than 100%. Questions included “Which THC substance(s) did you use in an e-cigarette, vaping device, vaporizer, or dab rig in the past 3 months? (Answer choices were: Marijuana herb (flower or leaves),THC oils, Butane hash oil, THC concentrate (e.g., wax, badder/budder, crumble, shatter, pull and snap), THC powder form (e.g., dry sift), or Other); “What type of device(s) did you use to vape or dab THC-containing products in the past 3 months?” (Answer choices were: Disposable e-cigarette or vape, E-cigarette or vape with prefilled cartridges, E-cigarette or vape with a tank that you refill with liquids (including sub-ohm, mod or modifiable systems), E-cigarette or vape with prefilled or refillable “pods” or pod cartridges (e.g. JUUL, Suorin), Dab rig, Vaporizer (for dry herbs, etc.), Other); and “What brand of THC-containing cartridge(s) were used with device(s)”? (Answer choices were: Rove, Dank Vapes, Golden Gorilla, Smart Cart, other).

Sources for THC-containing EVPs were assessed by the following question “Where did you obtain these THC-containing products? (Answer choices were: Medical dispensary, Recreational dispensary (retail cannabis/marijuana shop), Vape or smoke shop, Pop-up shop, Grocery store/Drug store/Convenience store, Family or friend, Illicit dealer, Online, Other). “Informal” sources include accessing from a family member or friend, illicit dealer, or online. All others were categorized as ‘commercial’ sources.

Frequency of THC EVP use was determined by the following question: “Approximately how frequently did you vape THC-containing products in the past 3 months?” (Answer choices were: Monthly or less, a few days per month, a few days per week, daily) and length of time using THC-containing EVPs was assessed with the following question :“How long have you been vaping or dabbing THC-containing products?” (Answer choices were:< 3 months, 3–6 months, 7–12 months, >1 year).

Basic demographic characteristics were also assessed on the survey including sex (male or female), age (subsequently grouped into 18–24, 25–34, 35–44, 45–64, and 65–86 years), and race or ethnicity (“What racial or ethnic group best describes you”; answer choices were: White, Black, Hispanic, Asian, Native American, Mixed, Other or Middle Eastern). Racial/ethnic categories were mutually exclusive and were collapsed into the following groups (White, Black, Hispanic, Asian, and ‘other’ which included all other categories).

Descriptive statistics of the sample were calculated. Prevalence was calculated, along with 95% confidence intervals (CIs) for the THC-containing EVP use behaviors assessed in the survey. A dichotomous “daily THC-containing product use” variable was created by using responses to the THC-containing EVP frequency of use question. Respondents who selected “a few days per week,” “a few days per month,” and “monthly or less” were classified as non-daily users.

Multivariable logistic regression was used to examine the association between demographic characteristics and daily versus non-daily THC-containing EVP use. Multivariable logistic regression, adjusting for age, race or ethnicity, and sex, was used to examine the association between specifics of THC-containing EVP use behaviors and living in a state with legalized adult nonmedical marijuana use (i.e., state-wide law allows for personal possession and consumption of marijuana for all adults; California, Colorado, Illinois, Michigan, Oregon, Washington) versus living in a state without legalized adult nonmedical marijuana use (Florida, Minnesota, New Jersey, New York, North Carolina, North Dakota, Ohio, Pennsylvania, Tennessee, Texas, Utah, Wisconsin). Because the response options were select all that apply, separate multivariable logistic regression models were run for each response option comparing yes to no responses (e.g., respondents reporting yes to the use of disposable EVPs versus those who did not report such use). Results for the multivariable logistic regression models are presented as adjusted odds ratios (aOR) and associated 95% CIs. Unweighted data are analyzed and reported because the sampling frame only included selected states and is not representative of state or national populations. Analyses were conducted using SAS 9.4 (Cary, NC).

Among the 3,980 adult respondents, 53.5% were female and the median age was 36 years (range 19 – 86 years). Of respondents, 71% identified as White, 11.0% Hispanic, 10.0% Black, 4.4% Asian, and 3.7% all other groups (Table 1).

Among respondents (Table 2), 53.1% used prefilled cartridge-based THC-containing EVPs, 37.9% used a tank system, 33.2% used disposable EVPs, and 32.7% used prefilled or refillable pods or pod cartridges in the past 3 months (Table 2). Less than 25% of respondents reported using dab rigs or vaporizers. The most common brand of THC-containing cartridge used was Dank Vapes (47.7%), followed by Golden Gorilla (38.7%), Smart Cart (21.8%), and Rove (20.1%). Nineteen percent of respondents reported using other brands. The most reported type of THC-based substance used in THC-containing EVPs was THC oils (69.6%), followed by marijuana herb (63.6%), THC concentrate (46.4%), butane hash oils (14.2%), and THC powder (11.6%).

Respondents reported obtaining their THC-containing EVPs from a variety of sources including, a recreational dispensary (41.1%), a friend or family member (38.6%), vape/smoke shop (38.4%), medical dispensary (25.5%), an illicit dealer (15.1%), grocery, drug or convenience store (9.0%), and online (7.4%).

Almost one-fourth (23.5%) of respondents used THC-containing EVPs daily, 23.0% weekly, 18.6% a few days per month, and 34.7% monthly or less. Most users (67.7%) reported using THC-containing EVPs 0–5 times per day, 13.7% reported 6–10 times per day, and 18.6% used over 10 times per day on the days they used THC-containing EVPs. Approximately 41% of respondents reported using THC-containing EVPs for more than a year, 19.2% for 7–12 months, 24.6% for 3–6 months, and 15.5% had been using for less than 3 months.

After adjustment for gender, race or ethnicity, and state nonmedical marijuana legalization status, respondents aged 45–86 years had lower adjusted odds of reporting daily use compared to those aged 25–34 years (aOR = 0.74; 95% CI = 0.61, 0.90) (Table 3). Compared to White respondents, Asian respondents had lower adjusted (for age, gender, state nonmedical marijuana legalization status) odds [aOR =0.55; 95% CI = 0.36, 0.84) of reporting daily THC-containing EVP use. Black or other respondents had higher adjusted odds of reporting THC-containing EVP use daily (Black respondents aOR = 1.48; 95% CI = 1.17, 1.86; other respondents, aOR = 1.51; 95% CI= 1.05, 2.17). Those living in legalized nonmedical marijuana states had higher adjusted (for age, gender, race/ethnicity) odds of using THC-containing EVPs daily than those living in non-legalized marijuana states (aOR = 1.18; 95% CI = 1.01, 1.37). No other significant differences were observed among the assessed groups.

After adjusting for age, race or ethnicity, and sex, adults reporting THC-containing EVP use via a disposable device, a dab rib, or a vaporizer had higher adjusted odds of living in a legalized nonmedical marijuana state (aOR for disposable device = 1.16; 95% CI = 1.01, 1.32; aOR for dab rig = 1.39; 95% CI = 1.18, 1.62; aOR for vaporizer = 1.21; 95% CI = 1.04, 1.42) compared to those not using those particular devices (Table 4). Further, those reporting using marijuana herb, butane hash oil, and THC concentrate had higher adjusted odds of living in a legalized nonmedical marijuana state (aOR for marijuana herb = 1.38; 95% CI = 1.20, 1.58; aOR for butane hash oil = 1.59; 95% CI = 1.32, 1.91; aOR for THC concentrate = 1.27; 95% CI 1.11, 1.43). There was limited variation in brands of THC products by residence in a legalized nonmedical marijuana state; however, users of the Golden Gorilla or ‘other’ brand had higher adjusted odds of living in a legalized nonmedical marijuana state (aOR for Golden Gorilla = 1.18; 95% CI = 1.03, 1.35; aOR for ‘other’ = 1.21; 95% CI = 1.02, 1.42). THC-containing EVP users who reported sourcing THC-containing EVPs through commercial means (e.g., recreational or medical dispensaries) had higher adjusted odds of living in a legalized nonmedical marijuana state (aOR for recreational dispensaries = 3.96; 95% CI = 3.45, 4.54, aOR for medical dispensary = 1.47; 95% CI = 1.27, 1.71; aOR for vape or smoke shop = 1.40; 95% CI = 1.23, 1.61). Correspondingly, accessing products via informal sources such as friends and family, the internet, or an illicit dealer was associated with lower adjusted odds of living in a legalized nonmedical marijuana state (aOR for illicit dealer = 0.47; 95% CI = 0.39, 0.58; aOR for family or friend = 0.56; 95% CI = 0.49, 0.64; and aOR for online source = 0.70; 95% CI = 0.54, 0.91). Those reporting first using THC–containing EVPs within the last year had lower odds of living in a legalized nonmedical marijuana state compared to adults who reported using THC-containing EVPs for more than a year (e.g., OR for < 3 months = 0.69; 95% CI = 0.57, 0.84). No other assessed groups had significant differences.

Almost half of survey respondents reported daily or weekly use of THC-EVP, and many accessed products through informal sources, even after EVALI began. These findings underscore that EVALI-like outcomes could occur again and highlight the importance of continued surveillance of THC-containing EVP use and expanded education and awareness about the potential risks associated with their use.

This work expands understanding of THC EVP use among adults in the U.S., which is particularly important in the midst of the evolving landscape of state-level marijuana policies and in the aftermath of the nation-wide outbreak of serious lung injuries associated with the use of THC-containing EVPs. Additional in-depth surveillance and research on EVP use behaviors, contents (including additives), product sources, associations with awareness of EVALI, and motivations for use, including the information provided from this report, can further inform prevention and education efforts and guide future outbreak response efforts.

CDC and FDA recommend that people not use THC-containing e-cigarette, or vaping, products, particularly from informal sources like friends, family, or in-person or online dealers. Because Vitamin E acetate is strongly linked to EVALI, it should not be added to any e-cigarette, or vaping, products. Additionally, people should not add any other substances not intended by the manufacturer to products, including products purchased through retail establishments. Evidence is not sufficient to rule out the contribution of other chemicals in either THC or non-THC products, in some of the reported EVALI patients. E-cigarette, or vaping, products (nicotine- or THC-containing) should never be used by youths, young adults, or women who are pregnant. THC use has been associated with a wide range of health effects, particularly with prolonged frequent use. The best way to avoid potentially harmful effects is to avoid the use of THC-containing e-cigarette, or vaping, products.