Self-Reported Cannabis Use and Symptom Burden Among Patients With Cancer Receiving Palliative Care

Abstract

Background

Patients are increasingly using cannabis products to manage a wide range of symptoms; however, few studies have examined the symptom burden of patients with cancer receiving palliative care who self-report using cannabis products.

Objectives

The purpose of this investigation was to compare the symptom burden of palliative care patients with cancer who self-reported using cannabis products vs those who did not.

Design

Retrospective study.

Setting/Participants

100 patients who self-reported using cannabis products and 300 patients who did not were randomly selected as the case and control groups, respectively. All participants were evaluated during their first consultation at a supportive clinic in a tertiary cancer center in the United States between January 1 and December 31, 2024.

Measurements

Data collected included self-reported cannabis use, demographics, primary cancer diagnosis, cancer treatment, symptoms, performance status, and medications.

Results

The proportion of patients who self-reported using cannabis products was 11.14%. Multivariate logistic regression analysis showed that female sex, older age, and Black or African American race were associated with lower odds of self-reported cannabis use. In contrast, living alone, higher levels of pain, anxiety, and appetite disturbance, were associated with increased odds of cannabis use.

Conclusions

Patients with higher symptom burden are more likely to report cannabinoid use. Universal screening accompanied by patient education is important in palliative care.

Keywords

cancer cannabis marijuana palliative care symptoms

Introduction

Cannabis, also known as marijuana, is a psychoactive drug derived from the cannabis plant (Cannabis sativa) which is primarily used for recreational and medical purposes.¹ In the United States, at the federal level, the Justice Department recently reclassified marijuana products into Schedule III.² As of June 26, 2025, 40 states, 3 territories, and the District of Columbia allow the use of cannabis products for medical use, while 24 states, 3 territories, and the District of Columbia allow or regulate their non-medical use.³

Cannabis products (such as tetrahydrocannabinol or cannabidiol) are used by patients for a wide range of cancer-related symptoms, including nausea, anorexia, pain, weakness, depressed mood, and anxiety.^4,5 Although the efficacy of cannabis products for the treatment of these symptoms remains limited,^6,7 its use among patients with cancer is widespread. A survey conducted across 12 National Cancer Institute Designated Cancer Centers in the United States found that 32.9% of the patients recently diagnosed with cancer reported using cannabis products; and the most common perceived benefits were for pain, sleep, stress, anxiety, and treatment side effects.⁸ Another survey reported that patients treated for cancer who used cannabis believed that it provided better pain control compared to opioids with fewer side effects, less addictive potential, and improved safety.⁹

These considerations are especially relevant in the palliative care setting because patients often experience high symptom burden, receive multiple drugs with potential interactions with cannabis, and quality of life is a central treatment goal. A systematic review of medicinal cannabis products in palliative care found some positive effects for pain, nausea/vomiting, appetite, sleep, and fatigue, but the quality of evidence was “low” or “very low”.¹⁰ The purpose of this investigation was to analyze the symptom burden among patients with cancer receiving palliative care who self-reported using cannabis products compared to those not who did not endorse such use.

Methods

This study consisted of a chart review of first consultations for patients evaluated at the Supportive Cancer Center at The University of Texas MD Anderson Cancer Center from January 1 to December 31, 2024. To be eligible, patients had to be 18 years old or older and had a diagnosis of cancer. The Institutional Review Board approved the study, and a waiver of informed consent was granted. Data collected included demographics, primary cancer diagnosis, cancer treatments, and medications.

During each visit to the Supportive Cancer Center, all patients are asked about the use of cannabis products (marijuana, cannabis concentrates, edibles, lotions, ointments, tinctures containing cannabis, cannabidiol-only products) and this information is documented in their chart. Additional routinely collected assessments include total Morphine Equivalent Daily Dose (MEDD), the Cut down, Annoyed, Guilty, and Eye opener - Adapted to Include Drugs score (CAGE-AID)^11,12; the Edmonton Symptom Assessment System (ESAS),^11,13 and the Screener and Opioid Assessment for Patients with Pain (SOAPP).^14,15

The CAGE-AID questionnaire is a screening tool for alcohol and illicit drug use, which is also utilized to identify maladaptive behaviors in patients who report exaggerated and erroneous need for opioids. It consists of four items; and scores ranging from 2 to 4 are considered positive for alcoholism and raise concerns about potential opioid misuse and chemical coping.¹⁵

The ESAS is a validated tool widely used in palliative care research to screen patients with cancer to assess for the presence and severity of symptoms. Symptoms are rated on a numerical scale from 0 (absence of symptoms) to 10 (worst possible severity). In this study, a 12-item version of the ESAS was used.^11,13

The SOAPP is used to assess risk of aberrant opioid behavior. It includes questions about antisocial behavior, history of substance abuse, the physician/patient relationship, medication-related behaviors, and the psychiatric and neurobiologic need for medicines.^14,15

Statistical Analysis

R version 4.4.1 was used for statistical analyses. Eight hundred ninety-eight patient records were screened to identify individuals with and without documented cannabis product use. Because cannabis use was relatively infrequent in the source cohort, a stratified sampling approach was used to ensure adequate representation of cannabis users for analytic purposes. All eligible patients who self-reported cannabis use and met inclusion criteria were eligible for selection up to the target sample size, and 100 were randomly selected. A comparison group of 300 non-users was then randomly selected from the same underlying cohort, yielding a 1:3 cannabis user to non-user ratio. This sampling strategy was designed to improve statistical efficiency for estimation of associations and is not intended to represent the underlying prevalence of cannabis use in the source population.

Associations between baseline demographic or clinical variables and the use of cannabis products were evaluated using logistic regression models. Univariable regression analyses were initially performed with the cannabis products outcome regressed for each variable. Variables whose P-value <0.20 were then retained in the multivariable model. For categorical variables, we kept those where one level was identified with a P-value <0.20, regardless of the significance of the remaining categories. Given the large number of possible multivariable model combinations, the Akaike Information Criterion (AIC) was used to identify the best fitting multivariable model.

Descriptive data were summarized as mean with the standard deviation (SD), median with the 25-75% interquartile range (IQR) or percentages. Fisher’s exact test or Wilcoxon rank-sum test were used for comparisons between groups. A P-value <0.05 was considered statistically significant.

Results

The proportion of patients who self-reported using cannabis products was 11.14% in the full screened cohort. Demographic and clinical characteristics, diagnosis and treatments are presented in Table 1. Compared to patients who did not report using cannabis products, those who self-reported using cannabis products were younger, more likely to be male, more likely to live alone, had higher SOAPP scores and a trend towards higher MEDD.

Table 1.

Demographic Information, Clinical Characteristics, and Treatments of the Patients

	Self-reported use of cannabis (N = 100)	No use of cannabis (N = 300)	P
Median age (IQR)	61.0 (46.0-67.0)	65.0 (54.0-74.0)	<0.001
Gender (%)
Female	44 (44)	171 (57)	0.028
Male	56 (56)	129 (43)	0.028
Race (%)
White or Caucasian	76 (76)	213 (71)	0.200
Black or African American	10 (10)	40 (13)
Asian, American Indian, Alaska Native, other	11 (11)	45 (15)
Declined to Answer	3 (3)	2 (1)
Hispanics (%)	18 (18)	52 (17)	>0.900
Marital status (%)
Married or significant other	60 (60)	216 (72)	0.064
Divorced	14 (14)	28 (9)
Widowed	7 (7)	25 (8)
Single	16 (16)	28 (9)
Other	3 (3)	3 (1)
Living alone (%)	40 (40)	84 (28)	0.033
Diagnosis (%)
Breast	13 (13)	48 (16)	0.082
Endocrine	6 (6)	11 (4)
GI	16 (16)	69 (23)
Gynecological	3 (3)	23 (8)
Head and neck	8 (8)	14 (5)
Hematological	17 (17)	46 (15)
Lung	14 (14)	26 (9)
Neuroendocrine	0 (0)	5 (2)
Neurological	2 (2)	10 (3)
Other	5 (5)	2 (1)
Sarcoma	4 (4)	11 (4)
Skin	2 (2)	6 (2)
Urological	10 (10)	29 (9.7%)
Treatment (%)
Chemotherapy	12 (12)	113 (38)	<0.001
Radiation	6 (6)	11 (4)
Hormone therapy	5 (5)	10 (3)
Surgery	2 (2)	6 (2)
Immunotherapy	28 (28)	56 (19)
Multiple	6 (6)	5 (2)
Other, pending, or none	41 (41)	99 (33)
Benzodiazepines (%)	24 (24)	68 (23)	0.800
SSRI (%)	11 (11)	49 (16)	0.300
SNRI (%)	13 (13)	22 (7)	0.100
Gabapentinoids (%)	28 (28)	68 (23)	0.300
Atypical antidepressants (%)	23 (23)	45 (15)	0.090
Opioids (%)	61 (61)	154 (51)	0.110
Antiemetics (%)	51 (51)	156 (52)	>0.900
Sleep aids (%)	35 (35)	94 (31)	0.500
Median MEDD (IQR)	5.0 (0.0-30.0)	0.0 (0.0-15.0)	0.070
Median CAGE score (IQR)	0.0 (0.0-0.0)	0.0 (0.0-0.0)	0.027
Median SOAPP score (IQR)	5.0 (3.0-8.0)	2.0 (1.0-4.0)	<0.001

CAGE: Cut down, Annoyed, Guilty, and Eye opener. ECOG: Eastern Cooperative Oncology Group performance status. ESAS: Edmonton Symptom Assessment System. IQR: 25-75% interquartile range. MEDD: Morphine Equivalent Daily Dose. SOAPP: Screener and Opioid Assessment for Patients with Pain. SNRI: Serotonin and Norepinephrine Reuptake Inhibitors. SSRI: Selective Serotonin Reuptake Inhibitors. Bold values indicate statistically significant results (p < 0.05).

Symptoms assessed by ESAS are presented in Table 2. In univariable analyses, patients who self-reported the use of cannabis products had higher overall ESAS scores, including higher scores for pain, nausea, depression, anxiety, appetite disturbance, financial distress, and spiritual pain.

Table 2.

The Edmonton Symptom Assessment System (ESAS), Total Score and Components

	Self-reported use of cannabis (N = 100)	No use of cannabis (N = 300)	P
ESAS Total score (IQR)	43.0 (28.0-56.0)	37.0 (21.0-48.0)	0.002
Pain	5.0 (3.0-7.0)	4.0 (1.0-6.5)	0.009
Fatigue	5.0 (3.0-7.0)	5.0 (3.0-7.0)	0.900
Nausea	1.0 (0.0-5.0)	0.0 (0.0-4.0)	0.026
Depression	2.0 (0.0-5.0)	1.0 (0.0-4.0)	0.001
Anxiety	4.0 (1.5-6.0)	2.0 (0.0-5.0)	<0.001
Drowsiness	2.0 (0.0-5.0)	2.0 (0.0-5.0)	0.700
Shortness of breath	1.0 (0.0-4.0)	1.0 (0.0-4.0)	0.500
Appetite	5.0 (2.0-7.0)	4.0 (1.0-6.0)	0.033
Wellbeing	5.0 (2.0-7.0)	5.0 (2.0-6.0)	0.200
Sleep	5.0 (3.0-7.0)	5.0 (2.0-6.0)	0.110
Financial	2.0 (0.0-5.0)	1.0 (0.0-4.0)	0.016
Spiritual pain	0.0 (0.0-2.0)	0.0 (0.0-1.0)	0.021

All values are expressed as median (25-75% interquartile range).

The multivariate logistic regression analysis (Table 3) showed that female sex, older age, and Black or African American race were associated with lower odds of self-reporting cannabis products use. In contrast, living alone and ESAS pain, anxiety, and appetite disturbance, were associated with higher odds of cannabis products use.

Table 3.

Multivariable Logistic Regression of Self-Reported Cannabis use by Akaike Information Criterion

	Odds ratio	P	95% Confidence interval
Female (reference: male)	0.435	0.002	0.255-0.731
Age	0.965	0.000	0.947-0.983
Black or African American^a	0.398	0.043	0.153-0.925
American Indian, Alaska Native, Asian, or Other^a	0.529	0.108	0.233-1.113
Declined to answer race^a	3.647	0.177	0.553-29.553
Living alone	1.884	0.023	1.088-3.259
ESAS pain score	1.102	0.034	1.008-1.208
ESAS anxiety score	1.156	0.002	1.055-1.270
ESAS appetite score	1.101	0.047	1.002-1.213
ESAS Wellbeing score	0.921	0.157	0.821-1.031

^aWhite or Caucasian was used as refence.

Discussion

In this study, female sex, older age, and Black or African American race were associated with lower likelihood of self-reported cannabis products use, whereas living alone, pain, anxiety, and appetite disturbance, were associated with higher likelihood of use.

A large study conducted at a primary care clinic in Los Angeles reported that 15.6% used cannabis for medical reasons only, of whom 75.7% used cannabis to manage symptoms such as pain, stress, and sleep.¹⁶ In 2022, the most common patient-reported qualifying conditions for medical marijuana use across 39 jurisdictions allowing medical cannabis use in the United States were chronic pain (48%), anxiety (14%) and posttraumatic stress disorder (13%).¹⁷ Similarly, Leung reported that the most common physical reasons were pain (53%), sleep (46%), headaches (35%), appetite (22%), and nausea/vomiting (21%), the mental health reasons were anxiety (52%), depression (40%), and PTSD/trauma (17%), and there were 11% who reported using cannabis for managing other drug or alcohol use and 4% for psychosis.¹⁸ In oncologic patients, the most common patient-reported reasons for cannabis products use (sleep, pain, nausea, appetite and anxiety) mirror those of the general population.^8,19,20

In palliative care patients with cancer, Donovan et al found that tiredness, lack of appetite, shortness of breath, anxiety, depression, and difficulty sleeping, younger age and male sex were predictive of cannabis use.²¹ Similar to this investigation, the retrospective design of our study precludes a definitive assessment of which symptoms motivated the self-use of cannabis products, whether symptom burden preceded cannabis use, or how effective patients perceived cannabis to be for symptom management. An important consideration, is that the symptom burden was higher among patients self-reporting using cannabis products, suggesting that these patients may have been using cannabis for symptom management rather than recreationally.

In patients with advanced cancer, it is clinically important for physicians to know whether cannabis products are being used. This population frequently receive centrally acting medications such as opioids, hypnotics or antidepressants, which may increase the risks of serious pharmacodynamic and pharmacokinetic interactions.²² Knowledge of cannabis use also facilitates open discussions about the patient expectations, potential benefits, limitations of current evidence, and financial considerations associated with cannabis. Additionally, patients would be more willing to report the use of cannabis products after a stronger relationship has been established during follow up.²³

This study has several limitations. Data were collected retrospectively at a single tertiary medical center in a state that relatively recently approved the use of medical cannabis for patients with terminal cancer. Importantly, although the 1:3 sampling design was intentionally selected to improve statistical power, it may still introduce selection and/or classification bias which may influence the results and does not allow estimation of population prevalence. Also, routine screening relying on self-reported cannabis use may have resulted in underreporting,²³ particularly considering that our group previously found in the same setting that, between April 2017 and January 2018, that 37% (44/119) of the patients on long-term opioid treatment for cancer pain who underwent random urine drug screening, tested positive for cannabis.²⁴ It is important to acknowledge that cannabis use was based on self-report and we did not capture information about dose, frequency, route of use, product composition or intent of use which limits clinical interpretability. Finally, recent qualitative research suggests that stigma surrounding cannabis use may persist among cancer survivors and may influence patients’ willingness to disclose cannabis use to healthcare providers, potentially contributing to underreporting²⁵; in addition, observed differences by sex, age, and race/ethnicity may reflect unmeasured structural and contextual factors such as access, cultural attitudes, and legal concerns, which were not assessed in this study.

We conclude that female sex, older age, and Black or African American race were associated with lower likelihood of self-reported cannabis use in patients with cancer receiving palliative care, whereas living alone, pain, anxiety, and appetite disturbance were associated with higher likelihood of cannabis use. Prospective studies in diverse clinical settings are warranted to further characterize patterns of cannabis use and their relationship with symptom burden in this population.

Footnotes

ORCID iD

Patricia S. Bramati

Ethical Considerations

The Institutional Review Board of The University of Texas MD Anderson Cancer Center approved the study, and a waiver of informed consent was granted.

Authors’ Contribution

PSB and HA: conceptualization, methodology, acquisition of the data, formal analysis, original draft preparation. GLG and DG: methodology and acquisition of the data. CMM: methodology and formal analysis. EB: conceptualization, supervision, reviewing and editing.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Kramer

. Medical marijuana for cancer. CA Cancer J Clin. 2015;65(2):109-122. doi:10.3322/caac.21260.

U.S. Department of Justice . Approved Marijuana Products and Products Containing Marijuana Subject to a Qualifying state-issued License in Schedule III, Strengthening Medical Research While Maintaining Strict Federal Controls. U.S. Department of Justice. Accessed May 5, 2026.

Legislatures NCoS . State medical cannabis laws. https://www.ncsl.org/research/health/state-medical-marijuana-laws.aspx. (Accessed 02 20 2022).

Pergam

Woodfield

Lee

, et al. Cannabis use among patients at a comprehensive cancer center in a state with legalized medicinal and recreational use. Cancer. 2017;123(22):4488-4497. doi:10.1002/cncr.30879.

Waissengrin

Urban

Leshem

Garty

Wolf

. Patterns of use of medical cannabis among Israeli cancer patients: a single institution experience. J Pain Symptom Manag. 2015;49(2):223-230. doi:10.1016/j.jpainsymman.2014.05.018.

Mucke

Weier

Carter

, et al. Systematic review and meta-analysis of cannabinoids in palliative medicine. J Cachexia Sarcopenia Muscle. 2018;9(2):220-234. doi:10.1002/jcsm.12273.

Whiting

Wolff

Deshpande

, et al. Cannabinoids for medical use: a systematic review and meta-analysis. JAMA. 2015;313(24):2456-2473. doi:10.1001/jama.2015.6358.

Ellison

Helzlsouer

Rosenfield

, et al. Perceptions, prevalence, and patterns of cannabis use among cancer patients treated at 12 NCI-designated cancer centers. J Natl Cancer Inst Monogr. 2024;2024(66):202-217. doi:10.1093/jncimonographs/lgae029.

Worster

Ashare

Meghani

, et al. Prevalence and perceptions of cannabis use among racially diverse patients with cancer pain: results from a multi-site survey. J Clin Oncol. 2022;40(16_suppl):12125. doi:10.1200/JCO.2022.40.16_suppl.12125.

10.

Doppen

Kung

Maijers

, et al. Cannabis in palliative care: a systematic review of current evidence. J Pain Symptom Manag. 2022;64(5):e260-e284. doi:10.1016/j.jpainsymman.2022.06.002.

11.

Arthur

Tanco

Park

, et al. Personalized pain goal as an outcome measure in routine cancer pain assessment. J Pain Symptom Manag. 2018;56(1):80-87. doi:10.1016/j.jpainsymman.2018.03.004.

12.

Dev

Parsons

Palla

Palmer

Del Fabbro

Bruera

. Undocumented alcoholism and its correlation with tobacco and illegal drug use in advanced cancer patients. Cancer. 2011;117(19):4551-4556. doi:10.1002/cncr.26082.

13.

Bruera

Kuehn

Miller

Selmser

Macmillan

. The edmonton symptom assessment system (ESAS): a simple method for the assessment of palliative care patients. J palliat care. 1991;7(2):6-9.

14.

Butler

Fernandez

Benoit

Budman

Jamison

. Validation of the revised screener and opioid assessment for patients with pain (SOAPP-R). J Pain. 2008;9(4):360-372. doi:10.1016/j.jpain.2007.11.014.

15.

Yennurajalingam

Edwards

Arthur

, et al. Predicting the risk for aberrant opioid use behavior in patients receiving outpatient supportive care consultation at a comprehensive cancer center. Cancer. 2018;124(19):3942-3949. doi:10.1002/cncr.31670.

16.

Gelberg

Beck

Koerber

, et al. Cannabis use reported by patients receiving primary care in a large health system. JAMA Netw Open. 2024;7(6):e2414809. doi:10.1001/jamanetworkopen.2024.14809.

17.

Boehnke

Sinclair

Gordon

, et al. Trends in U.S. medical cannabis registrations, authorizing Clinicians, and reasons for use from 2020 to 2022. Ann Intern Med. 2024;177(4):458-466. doi:10.7326/M23-2811.

18.

Leung

Chan

Stjepanovic

Chung

JYC

Hall

Hammond

. Prevalence and self-reported reasons of cannabis use for medical purposes in USA and Canada. Psychopharmacology (Berl). 2022;239(5):1509-1519. doi:10.1007/s00213-021-06047-8.

19.

Behl

D'Andre

Hankins

Tiet

Portugal

Parise

. Demographics, methods of use, and perceived benefits among patients with cancer who use cannabis. Support Care Cancer. 2026;34(3):197. doi:10.1007/s00520-026-10452-0.

20.

Vinette

Cote

El-Akhras

Mrad

Chicoine

Bilodeau

. Routes of administration, reasons for use, and approved indications of medical cannabis in oncology: a scoping review. BMC Cancer. 2022;22(1):319. doi:10.1186/s12885-022-09378-7.

21.

Donovan

Chang

Oberoi-Jassal

, et al. Relationship of cannabis use to patient-reported symptoms in cancer patients seeking Supportive/Palliative care. J Palliat Med. 2019;22(10):1191-1195. doi:10.1089/jpm.2018.0533.

22.

Madden

Tanco

Bruera

. Clinically significant drug-drug interaction between methadone and cannabidiol. Pediatrics. 2020;145(6):e20193256. doi:10.1542/peds.2019-3256.

23.

Bramati

Cadavid

Bansal

Fellman

Tanco

Bruera

. Letter to the editor: prevalence of self-reported cannabis use among patients at a supportive care clinic in a tertiary cancer center. J Palliat Med. 2022;25(8):1176-1178. doi:10.1089/jpm.2022.0199.

24.

Arthur

Nguyen

, et al. Random vs targeted urine drug testing among patients undergoing long-term opioid treatment for cancer pain. JAMA Oncol. 2020;6(4):580-581. doi:10.1001/jamaoncol.2019.6756.

25.

Levy

Meghani

Worster

, et al. Cannabis stigma and symptom management considerations in cancer survivors: a mixed-methods exploration of patient perspectives. Support Care Cancer. 2026;34(4):313. doi:10.1007/s00520-026-10523-2.