Enhancing data compatibility in an evolving landscape: Medical cannabis and polysubstance use protocols in the PhenX Toolkit

Abstract

Introduction:

With increasing legalization of medical cannabis and prevalence of polysubstance use in the United States comes the need for standard psychometrically validated measures to study these substances’ health effects in the population. The PhenX (consensus measures for Phenotypes and eXposures) Toolkit (www.phenxtoolkit.org) is a freely accessible catalog of recommended measurement protocols to promote data compatibility across studies, but this extensive catalog did not include measures in these important areas.

Methods:

In 2024, a PhenX Working Group of experts followed a well-established consensus process to identify and recommend measurement protocols suitable for inclusion in studies on medical cannabis and polysubstance use. The broader scientific community was invited to review and provide feedback that was considered in the process of finalizing recommendations.

Results:

In 2025, the PhenX Toolkit released 15 new medical cannabis and polysubstance use protocols, which assess medical and other cannabis use—including consumption levels, product types, sources, motives, expectancies, medical card status, provider–patient communication, and biochemical validation—as well as polysubstance use patterns and related overdose risk. These protocols complement existing substance-related content in PhenX Toolkit and facilitate future development of knowledge on health effects of cannabis and polysubstance use and clinical guidance on safety or dosing for medical cannabis.

Conclusion:

Researchers are encouraged to adopt these measurement protocols, so results across studies can be better compared and combined to efficiently and reliably evaluate the health effects of medical cannabis and polysubstance use. Measures in these domains will continue to be updated as new knowledge is gained.

1. Introduction

1.1. Need for standard measurement protocols of medical cannabis

The per capita rate of daily or near-daily cannabis use in the United States increased 15-fold from 1992 to 2022, with more daily and near-daily users of cannabis than alcohol recorded for the first time in 2022 (Caulkins, 2024). Despite such a rapid increase in prevalence, scientific knowledge about the health effects associated with diverse patterns of consuming cannabis products is limited. Most published studies only measured any (binary) or frequency of cannabis use without assessing product type (flower, edible, concentrates, etc.), quantity, or potency (Coelho et al., 2025). Yet, delta-9-tetrahydrocannabinol (THC) concentration, a measure of potency, varies significantly across products: from as low as 6% to 9% in some herbal cannabis products to 70% to 80% in cannabis concentrates (Craft et al., 2020). Quantifying cannabinoid exposure across the multitude of products available in the marketplace is challenging but urgently needed to improve the evidence for health effects of cannabis.

As of April 2025, medical and recreational use of cannabis was legal in 37 and 24 states, respectively (Bryan, 2024), with great heterogeneity in product availability and qualifying conditions to obtain a medical cannabis card. Use of medical cannabis for some qualifying conditions (e.g., chronic pain, childhood-onset seizures) are supported by research evidence (NASEM, 2017), and less robust evidence is available for other conditions, such as posttraumatic stress disorder (Rehman et al., 2021). Because cannabis remains federally classified as a Schedule I substance, defined as having no accepted medical use and high potential for abuse, clinicians who certify patients for medical cannabis are not required to provide dosing or safety recommendations. As a result, there is limited clinical guidance on safety or dosing for patients who seek to consume medical cannabis (Baral et al., 2025; Calcaterra et al., 2020; Ellison et al., 2024; Vidot et al., 2024). To study the effectiveness of cannabis for medical use, it is thus crucial to have standard assessments for tracking the symptoms that the patient uses cannabis to manage, the patient’s medical card status, communication with healthcare providers, and the product types and doses consumed.

1.2. Need for standard measurement protocols of polysubstance use

Reviews of vital statistics and emergency department data reveal that polysubstance use has contributed to increasing drug overdose morbidity and mortality in the evolving opioid epidemic, with fentanyl and stimulant co-involvement as the predominating pattern (Compton et al., 2021; Hébert & Hill, 2024; Kline et al., 2023; Schneider et al., 2021). Another emerging trend is the rising rates of co-use of cannabis, alcohol, and/or nicotine among young adults (Buu et al., 2020; Cohn & Chen, 2022; McCabe et al., 2021; Terry-McElrath & ME, 2018). Polysubstance use is typically measured in existing research as “concurrent co-use,” meaning using two or more substances within a prespecified time frame (e.g., 30 days or 12 months), which is often driven by the measures available in large-scale epidemiological surveys (Bunting et al., 2024). Yet, such a crude measure does not specify the relationship between substances and tends to mix heterogeneous co-use patterns that could carry different levels of addiction liability (Hindocha & McClure, 2021). In fact, two types of high-risk co-use patterns have been identified to be associated with greater consumption and symptomatology (Agrawal et al., 2012; Peters et al., 2021): simultaneous use (or co-administration)—multiple substances mixed together and used with one mode of administration (e.g., loading cocaine and heroin into a syringe for a speedball), and sequential co-use—multiple substances used in a single episode (e.g., “chasing” cannabis with a nicotine product). Fine-grained measures are needed to better characterize the health risk associated with different co-use patterns.

1.3. The PhenX Toolkit: Existing protocols and gaps

Funded by the National Human Genome Research Institute, with co-funding from other National Institutes of Health Institutes and Centers, the PhenX (consensus measures for Phenotypes and eXposures) Toolkit (www.phenxtoolkit.org) is a web-based catalog of recommended measurement protocols to promote data compatibility across clinical, epidemiological, and genomic research (Hamilton et al., 2011). Use of standard protocols from the PhenX Toolkit promotes cross-study analysis, enabling researchers to increase the impact of individual studies and to obtain greater statistical power and replication of results.

The scientific community drives the PhenX Toolkit’s content and features. A 15-member Steering Committee with broad expertise in biomedical research provides overarching guidance to the project. Working Groups (WGs) of domain experts use an established consensus process that relies on community feedback to select well-established protocols for inclusion in the Toolkit (Maiese et al., 2013).

In 2011, the National Institute on Drug Abuse (NIDA) provided supplemental funding to PhenX to expand the depth and breadth of the protocols in the Toolkit’s Substance Abuse and Addiction (SAA) Collections. As a result of this project, 69 new protocols were added to the Toolkit and organized into a core collection and six specialty collections (Conway et al., 2014; National Institute on Drug Abuse, 2012; PhenX Toolkit, n.d.-f). In 2016, a PhenX Expert Review Panel (PhenX Toolkit, n.d.-d) reviewed the protocols in the SAA Collections and the Alcohol, Tobacco, and Other Substances Domain to ensure they remained scientifically relevant. Since their release, the SAA Collections have consistently been among the top 10 most accessed PhenX domains and collections. The PhenX Toolkit is a preferred protocol resource for NIDA-supported research investigators. In notices of funding opportunity announcements, NIDA strongly encourages investigators to use protocols in the Toolkit as a common set of assessments to support pooling of data across studies (National Institute on Drug Abuse, n.d.). Although the PhenX SAA Collections continue to offer useful protocols captured from rigorous research and surveillance programs, there is a need to update the collections with contemporary protocols to fill gaps in the Toolkit that have emerged with the changing substance use landscape.

In January 2024, NIDA provided administrative funding to PhenX to establish a WG to recommend protocols that enhance consistent measurement of medical cannabis use, polysubstance use, and recovery. Secondary areas of focus included identification of updates to language and imagery to address stigma and opportunities for linking SAA protocols to other collections and research domains in the PhenX Toolkit to support assessment of contextual factors (e.g., social determinants of health). This article delineates the procedure for proposing and revising protocols specific to medical cannabis and polysubstance use, the final protocols recommended by the Substance Use and Recovery (SUR) WG for inclusion in the Toolkit, and future measurement work that is needed to advance scientific knowledge in these important fields of research and clinical work. Recommended protocols, relevant issues, and future work for assessing recovery identified by the WG are reported elsewhere.

2. Methods

The PhenX SUR WG (PhenX Toolkit, n.d.-f) included nine experts with diverse research and clinical expertise in medical cannabis use, polysubstance use, and recovery. The WG was led by two co-chairs and supported by a scientific collaborator from NIDA, the PhenX National Human Genome Research Institute project scientist and project analyst, and members of the PhenX team. The NIDA scientific collaborator and NIDA extramural program officers identified WG candidates and defined the initial WG scope (Table 1).

Table 1.

Initial medical cannabis and polysubstance use scope elements for the Substance Use and Recovery Working Group.

Patterns of use

Route of administration and delivery method

Cannabis forms

Medical card status

Source of products

Co-use of medications or other substances (excluding alcohol and nicotine) and drug interactions

Biochemical validation of products and consumption

Assessment of drug types

Methods of combination

Timing, frequency, and progression of use

Motivation, expectancies, and context

The SUR WG recommended protocols for the Toolkit using the PhenX consensus process, which incorporates input from the broader research community, and is described in detail elsewhere (Maiese et al., 2013). The WG held an introductory meeting in February 2024 to review the project goals and initial scope. In March 2024, the WG refined the scope and assigned members to identify relevant measurement protocols for specific scope elements. WG members volunteered for measurement protocol search and selection based on area of expertise and research interests. The assigned expert presented a list of the existing well-established, validated measurement protocols and recommendations for WG consideration. The WG co-chairs facilitated the discussions, all WG members considered the recommendations, and protocols were selected to move forward based on consensus decisions to avoid potential conflicts of interest. WG members stated their involvement in the development of measurement protocols under WG consideration for transparency, when applicable. During this process, it was acknowledged that, although the WG scope was primarily focused on enhancing assessment of medical cannabis use, there was also opportunity to enhance Toolkit protocols to address cannabis use more broadly. In April 2024, the WG met twice to present and discuss recommended preliminary protocols for each scope element. The WG prioritized protocols based on criteria established by the PhenX Steering Committee (Table 2) (PhenX Toolkit, n.d.-b) and using PhenX terms and definitions (PhenX Toolkit, n.d.-c). These criteria required protocols to be well-established, existing instruments that impose a relatively low burden on investigators and participants; usable by experts and non-experts; and preferably from a publicly available source.

Table 2.

Criteria for selecting PhenX protocols.

The protocols should be

Clearly defined Well established Use in major reference study (e.g., NHANES) or consortium (e.g., EMERGE) Standard measurement protocols exist Broadly validated and demonstrated utility Reproducible Reliable Low burden to participants and investigators

Whenever possible

Cross-cutting relevance for population groups as well as diseases and conditions Prior use in omics research Open-source software and nonproprietary instruments preferred

Source: https://www.phenxtoolkit.org/about/criteria

During deliberations, the WG reviewed existing Toolkit protocols to (1) avoid duplication, (2) adopt existing Toolkit protocols where applicable, and (3) identify gaps in the Toolkit. The WG built on existing PhenX protocols, adding existing and new protocols not in the PhenX Toolkit to the Substance Use, Use Disorders, and Recovery Collections. In October 2024, the WG identified 17 preliminary protocols to assess medical cannabis and polysubstance use to share with the research community for feedback and comment. Additional questions were included in community outreach to gather feedback on recommended language to update stigmatizing language and to adjust terminology to be “person-first” whenever possible.

Community outreach was open from October 1 through October 31, 2024. The outreach was shared extensively with a variety of stakeholders in the research community, including PhenX registered users, NIDA funded investigators, the NIDA Clinical Trials Network, Helping to End Addiction Long-Term (HEAL) Initiative investigators, Advancing Addiction Research and Treatment through Engagement with Rural Marylanders Impacted by PolySubstance Use (ARTEMIS), All Of Us measurement experts, the Recovery Research Institute network, the Cannabis and Health Research Initiative network, the International Cannabinoid Research Society, Research Society on Marijuana network, the College on Problems of Drug Dependence network, and the Consortium for Medical Marijuana Clinical Outcomes Research network. Eighty-six people responded to the outreach survey.

The WG met in February and March 2025 to review research community outreach results and finalize protocols for inclusion in the Toolkit. Overall, the feedback received from community outreach was positive. The respondents supported the WG’s recommendations and language updates; there was little disagreement over inclusion of recommended protocols to reconcile. The community provided suggestions for additional protocols for consideration by the WG.

The WG made several updates to their preliminary recommendations based on community feedback. The WG added the Cannabis Use Disorder Identification Test–Revised (CUDIT-R) (Adamson et al., 2010) as a screening tool for problematic cannabis use. The WG agreed to combine different versions of the Timeline Followback (Sobell & Sobell, 1992, 1995; Sobell et al., 1996) into a single protocol for Patterns of Substance Use – Alcohol and/or Drug Use to improve clarity. The WG also added two items from the International Cannabis Policy Study (ICPS) Survey Wave 6 (Hammond et al., 2023a, 2023b) to the Motives for Using Medical Cannabis protocol to capture opioid substitution as a rationale for medical cannabis use.

The prioritized protocols were loaded into the Toolkit database and processed through a bioinformatics pipeline to develop bioinformatics features supporting data collection and interoperability. The PhenX Steering Committee reviewed and approved the finalized SUR protocols and released them in the Toolkit in Spring 2025.

3. Results

The SUR WG recommended 15 protocols for medical cannabis and polysubstance use for inclusion in the Toolkit (Table 3). Twelve protocols were included in the Assessment of Substance Use and Substance Use Disorders Specialty Collection (PhenX Toolkit, n.d.-a), and three protocols were included in the Substance-specific Intermediate Phenotypes Specialty Collection (PhenX Toolkit, n.d.-e). Each protocol is briefly described as follows.

Table 3.

Final protocols for the PhenX Toolkit Substance Use, Use Disorders, and Recovery Collections.

Specialty Collection	Protocol	Protocol Source
Assessment of Substance Use and Substance Use Disorders Collection	Alcohol and Cannabis Simultaneous Use	Cuttler, C., & Spradlin, A. (2017). Measuring cannabis consumption: Psychometric properties of the Daily Sessions, Frequency, Age of Onset, and Quantity of Cannabis Use Inventory (DFAQ-CU). PLoS ONE, 12(5), e0178194.
		Kolp, H., Horvath, S., Fite, P. J., Metrik, J., Stuart, G. L., Lisdahl, K. M., & Shorey, R. C. (2023). Development of the Alcohol and Cannabis Simultaneous Use Scale (ACSUS) in college students. Journal of Substance Use, 29(4), 509–516.
	Assessment of Cannabis Use Disorder	Adamson, S. J., Kay-Lambkin, F. J., Baker, A. L., Lewin, T. J., Thornton, L., Kelly, B. J., & Sellman, J. D. (2010). An improved brief measure of cannabis misuse: The Cannabis Use Disorders Identification Test-Revised (CUDIT-R). Drug and Alcohol Dependence, 110(1–2), 137–143.
	Biochemical Validation of Cannabis Consumption	Klawitter, J., Sempio, C., Morlein, S., De Bloois, E., Klepacki, J., Henthorn, T., Leehey, M., Hoffenberg, E., Knupp, K., Wang, G., Hopfer, C., Kinney, G., Bowler, R., Foreman, N., Galinkin, J., Christians, U., & Klawitter, J. (2017). An atmospheric pressure chemical ionization MS/MS assay using online extraction for the analysis of 11 cannabinoids and metabolites in human plasma and urine. Therapeutic Drug Monitoring, 39(5), 556–564.
	Cannabis Products, Types, and Modes of Use	Hammond, D., Goodman, S., Burkhalter, R., & Corsetti, D. (2023). International Cannabis Policy Study Survey 2023 (Wave 6): Cannabis products—Product types & modes of use section, items USE.MODES, MODE.FREQ.HERB, HERB.FREQ, IMAGE.CHOICE, HERB.JOINT2, HERB.JOINT3, HERB.JOINT1, HERB.JOINT.CONF, HERB.JOINT2.REDO, HERB.JOINT3.REDO, HERB.JOINT1.REDO, HERB.JOINT.TOBACCO, HERB.JOINT.TOBACCO2, HERB.AMOUNT1, HERB.AMOUNT2, HERB.AMOUNT.CONF, HERB.AMOUNT1.REDO, HERB.AMOUNT2.REDO, HERB.THCRATIO, HERB.THC1, HERB.THC2, MODE.FREQ.DROP, DROP.FREQ, LIQUID.DROP.AMOUNT, LIQUID.DROP.AMOUNT.CONF, LIQUID.DROP.AMOUNT.REDO, LIQUIDDROP.THCRATIO, LIQUIDDROP.THC1, LIQUIDDROP.THC.UNIT, LIQUIDDROP.THC2, MODE.FREQ.LIQUIDVAPE, LIQUIDVAPE.FREQ, LIQUIDVAPE.AMOUNT, LIQUIDVAPE.AMOUNT.CONF, LIQUIDVAPE.AMOUNT.REDO, LIQUIDVAPE.THCRATIO, LIQUIDVAPE.THC1, LIQUIDVAPE.THC.UNIT, LIQUIDVAPE.THC2, MODE.FREQ.EDIBLE, EDIBLE.FREQ, HOMEMADE.EDIBLES.TRIED, HOMEMADE.EDIBLESOURCE, HOMEMADE.EDIBLES.AMOUNT, EDIBLE.12.MONTH.TYPE, EDIBLE.TYPE, EDIBLE.AMOUNT, EDIBLE.AMOUNT2, EDIBLE.AMOUNT.CONF, EDIBLE.AMOUNT.REDO, EDIBLE.AMOUNT2.REDO, EDIBLE.SERVING, EDIBLE.THCRATIO, EDIBLE.THC1, EDIBLE.THC2, MODE.FREQ.DRINK, DRINK.FREQ, DRINK.TYPE, DRINKS.AMOUNT, DRINK.AMOUNT.CONF, DRINK.AMOUNT.REDO, DRINKS.THCRATIO, DRINKS.THC1, DRINKS.THC2, MODE.FREQ.CONCEN, CONCEN.FREQ, CONCEN.AMOUNT, CONCEN.AMOUNT.CONF, CONCEN.AMOUNT.REDO, CONCEN.TYPE, CONCEN.THCRATIO, CONCEN.THC1, CONCEN.THC.UNIT, CONCEN.THC2, MODE.FREQ.HASH, HASH.FREQ, HASH.THCRATIO, HASH.THC1, HASH.THC2, MODE.FREQ.TINCT, TINCT.FREQ, HASH.AMOUNT, HASH.AMOUNT.CONF, HASH.AMOUNT.REDO, TINCT.THCRATIO, TINCT.THC1, TINCT.THC.UNIT, TINCT.THC2, MODE.FREQ.TOPICAL, TOPICAL.FREQ
	Co-use of Opioids with Other Drugs and Overdose Risk	Elliott, L., Crasta, D., Khan, M., Roth, A., Green, T., Kolodny, A., & Bennett, A. S. (2021). Validation of the Opioid Overdose Risk Behavior Scale, Version 2 (ORBS-2). Drug and Alcohol Dependence, 223, 108721.
	General Cannabis Use	Hammond, D., Goodman, S., Burkhalter, R., & Corsetti, D. (2023). International Cannabis Policy Study Survey 2023 (Wave 6): Consumption section, items EVER.TRY, USE.RECENT, CURRENT.USE, USE.30DAYS, USE.CONSISTENT, TIME.TO.USE.
	Medical Cannabis Card Status	Hammond, D., Goodman, S., Burkhalter, R., & Corsetti, D. (2023). International Cannabis Policy Study Survey 2023 (Wave 6): Cannabis sources—Medical cannabis use section, items CANNABIS.MED1, USE.MENTALHEALTH, USE.MENTALHEALTH.CURRENT, USE.MENTALHEALTH.MAIN, EVER.MED, EVER.MED.CURRENT, EVER.MED.MAIN, OPIOID.SUB, OPIOID.SUB2
	Medical Cannabis Provider-Patient Communication	National Cancer Institute. (n.d.). NCI Cannabis Supplement: Core measures questionnaire. https://epi.grants.cancer.gov/clinical/nci-cannabis-supplement-core-measures-questionnaire.pdf
	Patterns of Substance Use – Alcohol and/or Drug Use	Sobell, L. C., & Sobell, M. B. (1992). Timeline Followback: A technique for assessing self-reported alcohol consumption. In R. Z. Litten & J. Allen (Eds.), Measuring alcohol consumption: Psychosocial and biological methods (pp. 41–72). Humana Press.
		Sobell, L. C., & Sobell, M. B. (1995). Alcohol consumption measures. In J. P. Allen & M. Columbus (Eds.), Assessing alcohol problems: A guide for clinicians and researchers (pp. 55–73). National Institute on Alcohol Abuse and Alcoholism.
		Sobell, L. C., Sobell, M. B., Buchan, G., Cleland, P. A., Fedoroff, I., & Leo, G. I. (1996). The reliability of the Timeline Followback method applied to drug, cigarette, and cannabis use. Presented at the Annual Meeting of the Association for Advancement of Behavior Therapy, New York, NY.
	Patterns of Substance Use – Calendar-based Method for Detailed Cannabis Use	Petrilli, K., Lawn, W., Lees, R., Mokrysz, C., Borissova, A., Ofori, S., Trinci, K., Dos Santos, R., Leitch, H., Soni, S., Hines, L., Lorenzetti, V., Curran, H., & Freeman, T. (2024). Enhanced cannabis timeline followback (EC-TLFB): Comprehensive assessment of cannabis use including standard THC units and validation through biological measures. Addiction, 119(4), 772–783.
	Polysubstance Use Frequency - Past Year	Gryczynski, J., McNeely, J., Wu, L., Subramaniam, G., Svikis, D., Cathers, L., Sharma, G., King, J., Jelstrom, E., Nordeck, C., Sharma, A., Mitchell, S., O’Grady, K., & Schwartz, R. (2017). Validation of the TAPS-1: A four-item screening tool to identify unhealthy substance use in primary care. Journal of General Internal Medicine, 32(9), 990–996.
	Sources of Cannabis	Hammond, D., Goodman, S., Burkhalter, R., & Corsetti, D. (2023). International Cannabis Policy Study Survey 2023 (Wave 6): Cannabis Sources—Purchasing & retail sources section, items CANNABIS.SOURCE1, CANNABIS.SOURCE.FRIEND, CANNABIS.SOURCE4, CANNABIS.SOURCE3
Substance-specific Intermediate Phenotypes Collection	Expectancies for Medical Cannabis Use	Weiss, J., Tervo-Clemmens, B., Potter, K., Evins, A., & Gilman, J. (2023). The Cannabis Effects Expectancy Questionnaire-Medical (CEEQ-M): Preliminary psychometric properties and longitudinal validation within a clinical trial. Psychological Assessment, 35(8), 659–673.
	Motives – Alcohol, Tobacco, and Other Substances – Cannabis	Dowd, A., Zamarripa, C., Sholler, D., Strickland, J., Goffi, E., Borodovsky, J., Weerts, E., Vandrey, R., & Spindle, T. (2023). A cross-sectional survey on cannabis: Characterizing motives, opinions, and subjective experiences associated with the use of various oral cannabis products. Drug and Alcohol Dependence, 245, Article 109826.
	Motives for Using Medical Cannabis	Hammond, D., Goodman, S., Burkhalter, R., & Corsetti, D. (2023). International Cannabis Policy Study Survey 2023 (Wave 6): Cannabis Sources—Medical cannabis use section, items CANNABIS.MED1, USE.MENTALHEALTH, USE.MENTALHEALTH.CURRENT, USE.MENTALHEALTH.MAIN, EVER.MED, EVER.MED.CURRENT, EVER.MED.MAIN, OPIOID.SUB, OPIOID.SUB2

Throughout the instrument review process, the WG frequently recommended replacing the term “marijuana” with “cannabis” or “medical cannabis,” as appropriate, to reflect contemporary terminology and align with scientific standards, particularly in medical or therapeutic settings. Accordingly, throughout this article, the term “marijuana” has been replaced with “cannabis” or “medical cannabis” when discussing these protocols. The WG also noted the use of stigmatizing language and triggering text or images in their review of protocols. Given the rapidly changing nature of the field and terminology, the WG recommended that updates to protocol language be appropriate based on the circumstance; for example, understanding the local drug use context to inform references to specific products. These recommendations from the WG are included in the Toolkit alongside the protocols in the specific instructions to investigators.

The WG also revisited potentially stigmatizing language, such as the term “abuse,” within the existing Toolkit collections and protocols. For example, the previous title, “Substance Abuse and Addiction (SAA) Collections,” was revised to “Substance Use, Use Disorders, and Recovery Collections.” Instances of the term “abuse” within the text of protocols were also revised to “use.”

3.1. Alcohol and Cannabis Simultaneous Use

The WG selected the Alcohol and Cannabis Simultaneous Use Scale (ACSUS) (Kolp et al., 2023) because this self-administered questionnaire is concise and validated for measuring simultaneous alcohol and cannabis use that is prevalent in the United States and linked to greater likelihood of impaired driving, social consequences, and harms to self, compared with alcohol use only (Subbaraman & Kerr, 2015). Respondents answer nine questions about simultaneous alcohol and cannabis use frequency, quantity, and impact. The instrument was developed using modifications of previously validated instruments (e.g., Alcohol Use Disorders Identification Test [AUDIT], Cannabis Use Disorder Test [CUDIT-R]; Adamson et al., 2010; Saunders et al., 1993) and validated using exploratory and subsequent confirmatory factor analysis as well as measuring the internal consistency of the ACSUS with AUDIT and CUDIT-R among a sample of Midwestern college students.

3.2. Assessment of Cannabis Use Disorder

The WG recommended the Cannabis Use Disorder Identification Test-Revised (CUDIT-R; Adamson et al., 2010), a brief, validated, and widely used screening tool to assess the severity of cannabis-related problems. The CUDIT-R consists of eight self-administered items designed to identify problematic cannabis use in individuals who have used cannabis in the past 6 months. Items assess consumption patterns, psychological dependence, functional impairment, and risky behaviors related to cannabis use. The CUDIT-R is suitable for general and medical cannabis–using populations, although the WG notes that high-frequency use for medical reasons may elevate scores even in the absence of problematic use. Researchers are encouraged to interpret CUDIT-R scores in context. The CUDIT-R has good psychometric properties (Adamson et al., 2010) and has been used with young adults (Dunbar et al., 2025) and college students (Schultz et al., 2019).

3.3. Biochemical Validation of Cannabis Consumption

Liquid chromatography–tandem mass spectrometry is the preferred method for determination of cannabis analytes in human plasma or urine. The protocol from Klawitter et al. (2017) was selected as an example of an established, validated protocol for the biochemical analysis of the concentration (nanograms per milliliter) of 11 cannabinoids and metabolites in plasma or urine, such as THC, cannabidiol (CBD), 11-nor-Δ9-tetrahydrocannabinol-9-carboxylic acid, and so on. The assay from Klawitter et al. (2017) provides details for purchasing analytes, internal standards, and reagents; performing quality control and calibration curves; extracting, processing, and storing samples; and analyzing samples using high-performance liquid chromatography and tandem mass spectrometry. The primary measure for cannabis exposure is THC-COOH concentration in whole blood, which has been demonstrated to be a reliable quantitative measure of cannabis use over the past few weeks (Huestis et al., 1992; Schwope et al., 2011).

3.4. Cannabis Products, Types, and Modes of Use

Up to 135 self-administered items from the ICPS Survey Wave 6 (Hammond et al., 2023a) were adopted to quantify consumption of a variety of cannabis products. The ICPS is a well-established international survey that has been administered since 2018 to monitor cannabis use in Canada, the United States, Australia, and New Zealand (Corsetti et al., 2023; Hammond et al., 2020). ICPS questions are based on questions from other well-established surveys (e.g., National Survey on Drug Use and Health). Respondents are first asked to identify which cannabis products they have used in the past 12 months (e.g., dried herb, edibles, concentrates). For each product endorsed (with skip rules implemented), follow-up questions capture frequency of use, usual amount used, and levels of THC and CBD. For each of the cannabis products, the questions enable calculation of standard milligrams (mg) of THC and CBD used by the respondent, which can then be used to estimate the number of standard THC units (i.e., 5 mg THC).

3.5. Co-use of Opioids with Other Drugs and Overdose Risk

The WG recommended the Opioid Overdose Risk Behavior Scale, Version 2 (ORBS-2) (Elliott et al., 2021), a 39-item interviewer-administered questionnaire with six subscales (Subscales A–F) that cover use and misuse of prescription opioids and combined use with other substances: (A) Prescription Opioid Misuse, (B) Risky Non-Injection Use, (C) Injection Drug Use & Speedballing, (D) Opioid & Alcohol Combinations, (E) Opioid & Benzodiazepine Combinations, and (F) Higher Order Polysubstance Combinations. The subscales can be used alone or in combination with any other subscales. To reduce participant burden, a checklist of substances is first provided to all participants to indicate which substances they have consumed in the past 30 days, and skip logic patterns can be used to adjust applicable subscale delivery. Participants provide the number of days they used each substance or combination of substances over the past 30 days. The ORBS-2 was validated using convergent, concurrent, and predictive validity with people who use drugs in New York City.

3.6. General Cannabis Use

The WG recommended adopting six items from the ICPS Survey Wave 6 (see detailed description in 3.4; Hammond et al., 2023a). Respondents are asked if they ever tried cannabis, when they last used cannabis, how often they use cannabis, how often they used cannabis in the last 30 days, length of cannabis use, and how long after waking they use cannabis (daily users only).

3.7. Medical Cannabis Card Status

This protocol includes two items from the ICPS Survey Wave 6 (see detailed description in 3.4; Hammond et al., 2023a). Respondents are first asked whether they have ever received a recommendation for medical cannabis, then are asked whether they have a medical cannabis card.

3.8. Medical Cannabis Provider-Patient Communication

The WG recommended the National Cancer Institute (NCI) Supplement Cannabis Core Measures Questionnaire because it is well established and publicly available (National Cancer Institute, n.d.). It is used by NCI-designated cancer centers in regions with varying state cannabis laws to estimate cannabis use and perceived harms and benefits among cancer patients (Ashare et al., 2023; Baral et al., 2024; Ellison et al., 2024). Psychometric properties of the benefit/harm questions have been established by an item response theory analysis (Jones et al., 2024). This protocol includes four questions (listed as questions 7–10) from the original 37-item questionnaire, asking whether the participant has discussed using cannabis with their healthcare provider for their cancer symptoms or diagnosis and with what type of healthcare provider they communicated.

3.9. Patterns of Substance Use – Alcohol and/or Drug Use

The Timeline Followback (TLFB) (Sobell & Sobell, 1992) is a well-established, valid, and reliable instrument that has been used to retrospectively monitor use of a variety of substances (e.g., alcohol, marijuana, cocaine). It can be completed over a variety of time frames by self-, interviewer, or computer administration. The TLFB provides quantitative data of substance use, and relevant to the current WG, same-day polysubstance use behaviors. The psychometric properties of the TLFB have been extensively studied across populations and setting type (Hjorthøj et al., 2012). Adapted versions of the TLFB exist for specific substance use patterns, including cannabis (Petrilli et al., 2024), and cocaine, alcohol, and cannabis co-use (Fitzgerald et al., 2022).

3.10. Patterns of Substance Use – Calendar-Based Method for Detailed Cannabis Use

The WG recommended the Enhanced Cannabis Timeline Followback (EC-TLFB) calendar-based method because of the need for specificity of detailed medical and nonmedical cannabis use that enables calculation of standard THC units in alignment with NIDA recommendations. The EC-TLFB uses a series of five pictures to help respondents identify the type and amount of cannabis used, then captures frequency of use via a calendar (Petrilli et al., 2024). Feasibility and validity through biological measures have been established (Lees et al., 2025; Petrilli et al., 2024).

3.11. Polysubstance Use Frequency - Past Year

The WG recommended the Tobacco, Alcohol, Prescription medication, and other Substance use Part I (TAPS-I) (Gryczynski et al., 2017) because it is concise, well-established, and validated. It is highly specific to low-risk use of alcohol and other substances, including cannabis, and can be adapted easily for a specific substance or time frame. The TAPS-I is a self- or interviewer-administered questionnaire with five items asking about past 12-month use for four substance categories: tobacco, alcohol, prescription drugs, and illicit substances. A skip pattern asks men and women a different question about the quantity of alcohol used in the past 12 months. The tool’s robustness and factorial invariance have been demonstrated across multiple study populations and practice contexts (Gryczynski et al, 2017; McNeely et al, 2016; Schwartz et al, 2017; Wu et al, 2016).

3.12. Sources of Cannabis

The WG recommended adopting four self-administered items from the ICPS Survey Wave 6 (see detailed description in 3.4; Hammond et al., 2023a). Respondents are asked about cannabis obtained from family or friends, online, or from a store or dispensary in the past 12 months.

3.13. Expectancies for Medical Cannabis Use

The Cannabis Effects Expectancy Questionnaire–Medical (CEEQ-M) (Weiss et al., 2023) is self-administered, with 21 items that assess what expectations people have for the use of cannabis for medical symptoms (e.g., “cannabis relieves nausea,” “cannabis improves appetite”). The CEEQ-M was evaluated in a series of studies with US adults and demonstrated good psychometric properties, including a two factor structure and scalar invariance (Weiss et al., 2023).

3.14. Motives – Alcohol, Tobacco and Other Substances – Cannabis

The WG recommended the Comprehensive Marijuana Motives Questionnaire (CMMQ) Short Form (Dowd et al., 2023) because it is easy to score and is based on a validated instrument that has demonstrated satisfactory reliability and validity (Lee et al., 2009) in measuring use-motives within samples that have a medical cannabis certification or use cannabis recreationally. The CMMQ includes 15 self-administered items covering a variety of reasons for using cannabis (e.g., “to help you sleep,” “to forget your problems”).

3.15. Motives for Using Medical Cannabis

This recommended protocol includes nine self-administered items from the ICPS Survey Wave 6 (see detailed description in 3.4; Hammond et al., 2023a). Respondents are asked about motives for using medical cannabis (i.e., managing symptoms for anxiety, pain, etc.) and use of cannabis as a complement to or substitution for prescription and other drugs (e.g., opioids).

4. Discussion

This article presents a set of 15 standard protocols to address critical gaps in the PhenX Toolkit for medical cannabis and polysubstance use. Through an expert consensus process, the PhenX SUR WG identified and recommended instruments that reflect contemporary patterns of use, evolving cannabis products, and the complex nature of co-use behaviors. These protocols are designed to enhance data consistency, enable data integration, improve clinical and epidemiologic research, and support public health surveillance in a rapidly shifting substance use landscape. Use of standard protocols for medical cannabis and polysubstance use from the PhenX Toolkit facilitates comparing and combining data across studies, increasing the impact of individual studies (Hamilton et al., 2011).

4.1. Current issues and future directions for measuring medical cannabis use

Delineating “medical” cannabis use from “nonmedical” cannabis use is a current challenge and opportunity for clarity in future standard measurement tools. Consumption of cannabis has mechanistic potential to impact the endocannabinoid system, so assessments of “medical” use are needed. However, if measuring behavior, there may be differences in health outcomes for those who consume for medical reasons compared with those who self-report use for recreation, which is allowed legally in 24 of the United States. This specification is important context when interpreting results from measures such as the CUDIT-R. Furthermore, given that the demarcation between medicinal and recreational use is not clear-cut and that people can use cannabis for both medicinal and recreational reasons, it is important to parse the degree to which an individual uses cannabis for both reasons concurrently. The SUR WG has identified an existing measure, the Daily Sessions, Frequency, Age of Onset, and Quantity of Cannabis Use Inventory (DFAQ-CU; Cuttler & Spradlin, 2017), that has an item asking about medicinal only or both purposes; and another item about what percentage of the time the patient uses cannabis for recreational (rather than medicinal) purposes.

Quantifying the primary psychoactive compound in cannabis products, THC, that are consumed by individuals has been identified as a fundamental and yet challenging step to studying the relationship between cannabis consumption and health effects, as THC concentration varies widely across products (Craft et al., 2020). In recent years, researchers, including leadership at NIDA, have been advocating for 5 mg as the standard THC unit, similar to the concept of a standard drink (Freeman & Lorenzetti, 2020; Volkow & Weiss, 2020). To operationalize this concept, the SUR WG recommended the Cannabis Products, Types, and Modes of Use protocol (Hammond et al., 2023a) that employs pictures of different product amounts in reference to a bottle cap (for solid products) or a graduated transfer pipet (for liquid products) to facilitate participants’ estimation of the quantity consumed within a day in the unit of gram, milliliter (ml), or ounce (oz). Participants could also choose to report in terms of the number of joints, drops, or hits. The protocol provides example calculations for average mg of THC used per day for tinctures based on the quantity (with the input in ml, oz, or drop) and the THC level (with the input in mg/ml or percentage). Although the visual aids were designed to facilitate participants’ estimation of the quantity, community outreach respondents raised concerns about their efficacy that warrant further studies collecting quantitative data of response time and qualitative data of participants’ thought process. Furthermore, this protocol does not provide concrete calculations for the consumption quantity reported in hits and does not account for loss of THC due to the method of administration (e.g., side smoke, heating method, filtering). A recent study (Borodovsky et al., 2024) proposed to deal with the former issue by calculating grams-per-hit ratios and the latter issue by assigning method of administration efficiency constants based on product types (including smoking flower, vaping flower, vaping concentrate, and dabbing concentrate). This approach, based on responses to survey questions of quantity and potency with limited response options, has been validated against cannabis use disorder criteria from the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 2022) using a large online convenience sample of adult daily cannabis consumers in the United States. Recent pharmacokinetic findings highlight that innovative formulations such as sublingual tablets, rectal suppositories, and vaporizers differ markedly in absorption rate, bioavailability, and variability (Tarlovski et al., 2025). Taken together, these data underscore the importance of incorporating route-specific measures into cannabis exposure assessments, and more research in this area is needed.

Survey-based measures (e.g., the Cannabis Products, Types, and Modes of Use protocol) are designed to capture average or typical consumption of cannabis. Yet, recent studies found that young adults who vape tend to underestimate their actual consumption levels, based on a comparison of their retrospective reports on a conventional survey question with the corresponding real-time reports on ecological momentary assessment (EMA) via smartphones (Yang et al., 2023). The trend persisted in qualitative analysis of young adults’ open-ended responses concerning what they learned about their vaping after completing the 7-day EMA (Nam et al., 2025). In addition to the consumption level, EMA can provide insight into the use contexts and subjective effects of cannabis use that have intervention implications (Buu et al., 2023). In fact, a recent study (Coelho et al., 2025) tested the feasibility of using EMA to quantify THC and CBD consumption at the event level and examine acute consequences as a function of the THC intake (controlling for CBD intake). The feasibility was built upon the training that participants received to report quantities of different products and the availability of standard product labels describing THC and CBD content in Canada. Another recent study (Smith et al., 2024) supplemented EMA with a smart vaporizer device (to use cannabis oil) and an application to assess objective vaping behavior. Although these recent studies had small sample sizes and aimed to establish feasibility, they demonstrated the potential of assessing cannabis use behavior in real time and real life using technology. Given the automatic, and possibly unconscious, cognitive process underlying substance use (Baker et al., 2004), such capacity holds promise for improving cannabis measurement and estimates of exposure.

The need for rigorous research on minor cannabinoids is increasingly evident as the landscape of medical cannabis use expands. Although THC and CBD have been the primary focus of research (Murray et al., 2024; Senator et al., 2024), a growing body of evidence suggests that minor cannabinoids, including cannabigerol (CBG), cannabinol (CBN), and other compounds, may also exert distinct therapeutic effects (Cammà et al., 2025; Walsh et al., 2021). However, current cannabis assessment tools, including the DFAQ-CU (Cuttler & Spradlin, 2017); the Cannabis Engagement Assessment (CEA; Schluter & Hodgins, 2022); and the ICPS Survey (Hammond et al., 2023a), tend to focus on THC/CBD content and do not adequately capture the use of minor cannabinoids. Given the increasing commercial availability of products containing substantial doses of minor cannabinoids (Anthony, 2024), refining assessment methodologies to include detailed characterization of these cannabinoid profiles is crucial. The PhenX protocol Biochemical Validation of Cannabis Consumption, recommended by the SUR WG, can characterize a subset of minor cannabinoids such as CBG and CBN in urine or plasma (Klawitter et al., 2017). Future research should incorporate more comprehensive measures of various cannabinoids, including minor cannabinoid use, enabling a better understanding of their therapeutic potential, safety profiles, and real-world health outcomes (Thrul & Vandrey, 2024).

4.2. Current issues and future directions for measuring polysubstance use

Existing measures for polysubstance use tend to focus on specific substances (mostly a combination of two substances) and the frequency of simultaneous use or sequential co-use, including the PhenX protocols recommended by the SUR WG, Alcohol and Cannabis Simultaneous Use (Kolp et al., 2023) and Co-use of Opioids with Other Drugs and Overdose Risk (Elliott et al., 2021). Yet, more dynamic measurement is needed, including measurement of motivation for polysubstance use given that motivation might provide critical information for intervention (Bunting et al., 2024). For example, a second substance might be taken to relieve the withdrawal symptoms, enhance the effects of a primary substance, or for instrumental purposes to enhance alertness, as has been observed in qualitative studies on opioid-stimulant polysubstance use (Ivsins et al., 2022; McNeil et al., 2021). Although the former case may be resolved by treating the disorder of the primary substance, the latter case may warrant co-treatment of both substances and closer consideration of motives and instrumental uses. Another construct that is usually omitted in existing measures is the route of administration, which may carry important implications for health risk, intervention, and policy. Particularly, this construct is relevant to those substances with rising novel products, such as cannabis and nicotine. For example, co-vaping nicotine and cannabis via e-cigarettes, which is prevalent among young people (Ouellette et al., 2023), may carry high risk for addiction, because it usually involves high-potency cannabis concentrates (Chadi et al., 2020). As highlighted by the PhenX selection process, it is difficult to measure the multiple components of polysubstance use and to do so using tools with low burden for participants. Although the SUR WG recommended the use of the TLFB, it is recognized that this method can have significant participant burden. Overall, the measures selected for inclusion for polysubstance use were recognized as imperfect, owing to the dynamicity and complexity of the phenomenon.

To overcome limitations in current tools, researchers should continue to advance brief, quantitative assessments of polysubstance use. One such example, which was not available at the time of the PhenX selection process, is the Polysubstance Assessment Tool (PAT). This tool was recently developed after four phases: qualitative interviews, expert feedback, cognitive interviews, and psychometric validation (Bunting, 2024). PAT provides a comprehensive assessment of diverse substances with the flexibility to create different combinations tailored to participants’ unique co-use patterns. It measures same-occasion and same-day polysubstance use. It covers frequency of use, usual route of administration, motivation for use, and numbers of lifetime and past 30-day overdose events. Two versions were created: (1) the interviewer-administered version, programmed via REDCap; and (2) the self-administered version, built as an iOS application for delivery on a tablet. Both versions are freely available for individual research through New York University’s Technology Opportunities catalog (NYU Langone Health, n.d.). The self-administered version allows for the capture of any polysubstance use patterns, and the interviewer-administered version relies on skip patterns predicated on substances known to contribute to overdose. Future research to advance measures of polysubstance use is warranted.

4.3. Limitations

While the expert consensus and outreach process was strong, there are some limitations. Protocols that meet the selection criteria for the PhenX Toolkit (listed in Table 2) may not be available for each WG scope (listed in Table 1). As outlined in Section 2, the PhenX Steering Committee criteria for selecting PhenX protocols requires protocols to be validated, well established, and low-burden (PhenX Toolkit, n.d.-b). Available protocols may not include the most up-to-date terminology, such as definitions and lists of substances. To address these limitations, the SUR WG provided additional instructions to address language in the protocols that was not optimal for prospective study participants. The SUR WG could only propose a limited number of protocols for community outreach and for the Toolkit; thus, the WG had to identify the highest priority set of measures for inclusion.

Outreach was conducted via online survey distribution through community listservs; feedback was limited to short written responses. Respondents included a wide range of PhenX-registered users and NIDA-funded investigators who are likely to be most frequent users of such measures. The number of respondents was high in comparison to other PhenX outreach. However, outreach may have excluded researchers in other scientific fields that may incorporate substance use-related outcomes in their research or investigators funded by organizations other than NIH. Additionally, as participation was voluntary, outreach respondents were those with both the time and the interest to contribute.

The SUR WG identified terminology and language that investigators should be cognizant of when using these protocols. Although there are limitations to adapting protocols that have been validated for a particular context, investigators should consider their relevance and sensitivity to their study population.

5. Conclusion

The PhenX Toolkit provides well-validated measurement protocols for medical cannabis and polysubstance use through reaching consensus among the experts in the SUR WG and incorporating input from the research community. Researchers are encouraged to adopt these protocols, so results across studies can be better compared and combined to more efficiently and reliably evaluate the health effects of medical cannabis and polysubstance use.

Highlights.

• Fifteen medical cannabis & polysubstance use protocols are added to PhenX Toolkit.

• New protocols facilitate studies on health effects of cannabis & polysubstance use.

• Standard measures enhance data compatibility and integration across studies.

Glossary

Core Collection

Includes measurement protocols that are deemed relevant for all studies in a specific topic or field of research to ensure collection of comparable data across studies.

Domain

A PhenX (consensus measures for Phenotypes and eXposures) domain is a field of research or topic with a unifying theme, which may be a research area, disease, or condition, and with easily enumerated quantitative or qualitative measures. Examples include Demographics, Anthropometrics, Environmental Exposures, Pregnancy, Cancer, and Genomic Medicine Implementation.

Measure

A standard way of capturing data on a certain characteristic of, or relating to, a study subject.

Protocol

A standard data collection procedure, recommended by a PhenX Working Group. Also referred to as a “measurement protocol.”

Scope Element

A topic critical to the domain or collection proposed by the funding agency and its Working Group.

Specialty Collection

Provides in-depth assessments in a specific topic or field of research.

Source

PhenX (consensus measures for Phenotypes and eXposures), https://www.phenxtoolkit.org/help/glossary