JOURNAL OF PALLIATIVE MEDICINE

Volume XX, Number XX, 2019 Brief Report

ª Mary Ann Liebert, Inc.
DOI: 10.1089/jpm.2019.0211

Perception of Benefits and Harms of Medical

Cannabis among Seriously Ill Patients
in an Outpatient Palliative Care Practice

Ali John Zarrabi, MD,1 Justine W. Welsh, MD,2 Roman Sniecinski, MD, MSc,3 Kimberly Curseen, MD,1
Theresa Gillespie, PhD, MA,4 Wendy Baer, MD,2 Anne Marie McKenzie-Brown, MD,3 and Vinita Singh, MD3
Downloaded by Linkoping University Library from www.liebertpub.com at 09/21/19. For personal use only.

Abstract
Introduction: Patients with serious illness often have pain, uncontrolled symptoms, and poor quality of life.
Evidence continues to evolve regarding the role of cannabis to treat chronic pain, nausea, and anorexia. Little is
known about how patients with serious illness perceive its benefits and harms. Given that an increasing number
of clinicians across the United States are treating patients with medical cannabis, it is important for providers to
understand patient beliefs about this modality. We assessed patient perceptions of benefits and harms of
cannabis who obtained a medical cannabis card within an ambulatory palliative care (APC) practice.
Methods: We recruited patients with a medical cannabis card, allowing for legal possession of cannabis oil,
from an APC practice in Georgia. All participants reported using cannabis products. Patients completed an
online survey that included questions about their cannabis use, concurrent opiate or controlled medication use,
and perceptions of benefits and harms of cannabis.
Results: All 101 patients invited to participate completed the survey. A majority had cancer (76%) and were
married (61%), disabled or retired (75%), older than 50 years of age (64%), and men (56%). Most patients
ingested (61%) or vaporized (49%) cannabis products. A majority of respondents perceived cannabis to be
important for their pain (96%) management. They reported that side effects were minimally bothersome, and
drowsiness was the most commonly reported bothersome harm (28%). A minority of patients reported cannabis
withdrawal symptoms (19%) and concerns for dependency (14%). The majority of patients were using con-
current prescription opioids (65%). Furthermore, a majority of cancer patients reported cannabis as being
important for cancer cure (59%).
Conclusion: Patients living with serious illnesses who use cannabis in the context of a multidisciplinary APC
practice use cannabis for curative intent and for pain and symptom control. Patients reported improved pain,
other symptoms, and a sense of well-being with few reported harms.

Keywords: cannabis; marijuana; palliative care; symptom management

Introduction ambulatory palliative care (APC) practices that care for

I n a majority of states, select patients are allowed to
use cannabis for medical use.1–4 Seriously ill patients
with a variety of illnesses may qualify for legal possession
of cannabinoids of various forms.5 Georgia law allows
qualified patients with a state-sanctioned card to possess
low 9-tetrahydrocannabinol (THC) oil (£5%) with equal or
greater amount of cannabidiol (CBD).6 Patients are in-
creasingly requesting authorization for legal cannabis in
patients with comorbid chronic pain, substance use disorder
(SUD), and serious illness.7–10 Many patients are referred
early in their disease trajectory when opioid therapy for
pain or other polypharmacy may confer risk of diversion,
dependence, or SUDs. Given the increase in cannabis re-
quests at our institution, we assessed perceptions of can-
nabis’ benefits and harms among patients who reported
using cannabis and had a state-sanctioned card in our APC
practice.

1
Department of Family and Preventive Medicine, 2Department of Psychiatry and Behaviorial Sciences, 3Department of Anesthesiology,
4
Department of Surgery, Emory University, Atlanta, Georgia.
Accepted August 22, 2019.

1
 2 ZARRABI ET AL.

Methods Table 1. Patient Demographics

Recruitment of participants % (standard
Patients were recruited from an academic APC practice Frequency error of %)
serving primarily advanced cancer patients and other serious Gender (n = 97)
illnesses. Patients who met inclusion criteria for invitation to Male 54 55.7 (5.1)
complete the survey were (1) at least 18 years of age, (2) an Female 43 44.3 (5.1)
active APC patient, (3) functionally and cognitively capable Age, years (n = 97)
of completing an online survey, (4) met a qualifying condi- <21 0 0.0 (0.0)
tion for a state-sanctioned cannabis card, and (5) actively 21–29 7 7.2 (2.6)
procured and carried the card. All patients who reported using 30–39 13 13.4 (3.5)
cannabis products who met the inclusion criteria were asked 40–49 15 15.5 (3.7)
to complete a survey at the end of their regularly scheduled 50–59 25 25.8 (4.5)
clinic appointment. The study was approved by the Institu- 60–69 25 25.8 (4.5)
tional Review Board. 70 and older 12 12.4 (3.4)
Level of education (n = 98)
Survey Did not complete high 4 4.1 (2.0)
school
Downloaded by Linkoping University Library from www.liebertpub.com at 09/21/19. For personal use only.

We used experiential data and patient feedback from >1 year High school or GED 23 23.5 (4.3)
of cannabis counseling within the APC clinic to draft the sur- Trade or vocational school 10 10.2 (3.1)
vey. The survey was voluntarily administered from December Some college 26 26.5 (4.5)
2017 to July 2018 during scheduled clinic appointments. No Bachelor’s degree 24 24.5 (4.3)
monetary incentives were provided for participation, and pa- Master’s degree or above 11 11.2 (3.2)
tients provided online informed consent for the study. Survey Marital status (n = 98)
items assessed concurrent opioid and controlled medication Married 60 61.2 (4.9)
use, means of cannabis consumption, and perceptions of Divorced/separated 20 20.4 (4.1)
benefits and harms. Perceived benefits and harms of cannabis Never married 18 18.4 (18.4)
were assessed using a Likert scale from ‘‘extremely impor- Employment (n = 97)
tant’’ to ‘‘definitely not important,’’ and ‘‘extremely both- Currently on disability, 55 56.7 (5.1)
ersome’’ to ‘‘noticeable, but not bothersome,’’ respectively. and unable to work
Full or part-time 15 15.5 (3.4)
All survey items were optional to complete. employment
Because of the small sample size and variety of cannabis Retired 18 18.6 (4.0)
products that patients used, this study is limited to descriptive Homemaker 7 7.2 (2.6)
analysis. Descriptive statistics, including frequency tables Student 2 2.1 (1.5)
with proportions and bar graphs, were used to summarize the Qualifying condition (n = 94)
data. Wald confidence intervals for proportions were set at Cancer 73 75.5 (4.3)
95%. All calculations were performed using SAS software Peripheral neuropathy 16 17.0 (3.9)
version 9.4 (SAS Institute, Inc., Cary, NC). Othera 5 5.3 (2.3)
Frequency of cannabis use (n = 94)
Results Multiple times daily 55 58.5 (5.1)
Demographics and qualifying conditions
Average of once daily 16 17.0 (3.9)
Average of one to five 19 20.2 (4.2)
for cannabis times weekly
Of the 101 patients who were invited to participate in the Less than once per week 4 4.2 (2.1)
survey, 100% responded. Fifty-four participants (55.7%) were a
Includes multiple sclerosis/amyotrophic lateral sclerosis (n = 1),
men. The majority were 50 years of age or older (64.0%) graft-versus-host disease (n = 1), HIV (n = 1), seizures/epilepsy
(Table 1). Most patients had at least a high school degree or (n = 1), Crohn’s disease (n = 1).
equivalent (82.6%) and were married (61.2%). Most patients GED, general education diploma.
were on disability and reported to be unable to work (56.7%).
None were actively seeking employment. Only 15 patients
able to select more than one product (Fig. 1). A minority of
were employed.
patients used non-oil forms of cannabis (20.0%). The re-
Most patients reported cancer as their qualifying diag-
mainder used pure CBD oil (14.9%), >5% THC oil (12.9%),
nosis for a state-issued card (75.5%) followed by severe
tetrahydrocannabinolic acid (THCa) oil (2.0%), or were
peripheral neuropathy (17.0%). The remaining diagnoses
unsure of the contents of their oil (8.9%). Patients were
were for amyotrophic lateral sclerosis/multiple sclerosis,
allowed to free-text responses, and responses were re-
graft-versus-host disease, seizures/epilepsy, and Crohn’s
classified to their appropriate category for analysis. Among
disease.
surveyed patients who used cannabis in the oil form, the
majority consumed it orally (61%) or inhaled the oil as a
Forms and frequency of use
vapor (49%), known popularly as ‘‘vaping.’’ Patients also
The most commonly reported products used by patients smoked plant-based cannabis (23%), ingested edibles
were state-sanctioned low-dose THC (£5%) oil (49.5%) (16%), or used a topical product (2%). Patients commonly
and mixed THC/CBD oil (33.7%). Of note, patients were used cannabis multiple times daily (59%) (Table 1).
 CANNABIS AND PALLIATIVE CARE 3
Downloaded by Linkoping University Library from www.liebertpub.com at 09/21/19. For personal use only.

FIG. 1. (A) Forms of cannabis used, (B) mode of cannabis consumption, and (C) other substances used, represented as
percentage of respondents selecting each choice in the ‘‘select all that apply’’ questions in survey items. Error bars represent
95% Wald confidence intervals of the percentage.

Cannabis and other medication use
Most patients registered for a state-issued card reported
using concurrent opioid therapy (65.3%), and a minority of
patients used prescription anxiolytics/sedatives (29.7%), al-
cohol (12.9%), and tobacco (10.8%) (Fig. 1). No surveyed
patients reported taking other schedule I drugs such as heroin,
cocaine, or psychedelics.
Perceived benefits and harms
A majority of patients found cannabis to be important or
extremely important in reducing pain (95.5%), decreasing
anxiety (79.0%), relieving depression/improving coping
(77.3%), providing a sense of well-being (73.9%), facilitat-
ing sleep (73.0%), increasing energy (69.0%), control-
ling nausea (66.7%), and stimulation of appetite (62.5%)
(Table 2). Patients on concurrent opioid therapy also over-
whelmingly reported cannabis as being important or ex-
tremely important for pain (95.5%). Among patients with
cancer, a majority reported using cannabis products for cancer
cure (62.2%).
Patients did not report significant harms from cannabis
(Table 3). A minority of patients reported drowsiness from
cannabis (27.6%), although a few found it bothersome or
 4 ZARRABI ET AL.

Table 2. Patient-Reported Benefits of Cannabis Products

Perception of importance
Extremely important Neither important/unimportant Definitely
or important or not applicable not important
Reported benefit F % SEP F % SEP F % SEP
Pain 84 95.5 2.2 3 3.4 1.9 1 1.1 1.1
Sleep 65 73.0 4.7 24 27.0 4.7 0 0 0
Nausea 58 66.7 5.1 28 32.2 5.1 1 1.1 1.1
Anxiety 68 79.0 4.4 17 19.8 4.3 1 1.2 1.2
Depression/coping 68 77.3 4.5 19 21.6 4.4 1 1.1 1.1
Appetite 55 62.5 5.2 27 30.1 4.9 6 6.8 2.7
Energy 60 69.0 5.0 23 26.4 4.8 4 4.6 2.3
Well-being 65 73.9 4.7 20 22.7 4.5 3 3.4 1.9
Cancer cure 52 59.1 5.3 32 36.4 5.2 4 4.5 2.2
F, frequency; SEP, standard error of percent.
Downloaded by Linkoping University Library from www.liebertpub.com at 09/21/19. For personal use only.

extremely bothersome (8.1%). Over 10% of patients reported
mild to extremely bothersome dizziness, problems with fo-
cusing, and confusion. Less than 5% of patients reported
bothersome or extremely bothersome (not including mildly
bothersome as a category) nausea/vomiting, constipation, hy-
potension, tachycardia, paranoia, anxiety, depression, head-
ache, thoughts of self-harm, coughing, imbalance, problems
with focusing, confusion, mood changes, or hallucinations.
Patients were able to free text brief narrative responses
explaining perceived benefits and harms of cannabis. Selec-
tions representative of the collective responses include the
following: (1) ‘‘makes life with advanced cancer much more
bearable.allows me not to focus on my own health issues
and to enjoy life as much as possible,’’ (2) ‘‘I no longer wake
every 20 to 30 minutes of dread, anxiety, and an utter sense of
overwhelming loss. My sleep is sound, deep, and fulfilling,’’
(3) ‘‘reduce dependence on opioids,’’ and (4) ‘‘peace.’’
Among noted harms, the most common free-texted response
was ‘‘none’’ (five respondents) followed by ‘‘dry mouth’’
(two respondents) (Supplementary Tables S1 and S2).
Concerns for dependency and SUDs
A minority of patients reported unpleasant physical or
emotional symptoms suggestive of withdrawal including ir-
ritable mood, changes in sleep, and pain when stopping
cannabis products (19.1%). Few patients reported a percep-
tion of mental or physical dependence to cannabis (13.5%).
Only three patients were concerned about becoming ‘‘ad-
dicted’’ to cannabis. Given the increased risk of developing a
SUD among patients with a personal or family history, we
asked patients about their personal and family histories. Pa-
tients reported a personal (2.2%) or family history (11.1%) of
drug or alcohol dependency resulting in rehabilitation pro-
gram enrollment, and only one patient reported both a per-
sonal and family history (Supplementary Fig. S1).

Table 3. Patient-Reported Harms of Cannabis Products

Perception of importance
Extremely bothersome, bothersome, Noticeable but Not
or mildly bothersome not bothersome applicable
Reported harms F % SEP F % SEP F % SEP
Nausea/vomiting 5 5.8 2.5 7 8.1 3.0 74 86.0 3.8
Constipation 4 4.7 2.3 8 9.3 3.2 74 86.0 3.8
Hypotension 3 3.4 2.0 6 6.9 2.7 78 90.0 3.3
Palpitations 4 4.5 2.2 10 11.3 3.4 74 84.1 3.9
Paranoia 5 5.7 2.5 12 13.8 3.7 70 80.5 4.3
Anxiety 8 9.1 3.1 11 12.5 3.5 69 78.4 4.4
Depression 4 4.6 2.3 8 9.2 3.1 75 86.2 3.7
Headache 3 3.4 2.0 8 9.2 3.1 76 87.4 3.6
Suicidality 0 0 0 4 4.5 2.2 84 95.5 2.3
Cough 5 5.8 2.5 12 14.0 3.8 69 80.2 4.3
Drowsiness 24 27.6 4.8 29 33.3 5.1 34 39.1 5.3
Dizziness 11 12.5 3.5 10 11.4 3.4 67 76.1 4.6
Focus 13 14.9 3.8 26 29.9 4.9 48 55.2 5.4
Confusion 15 17.2 4.1 11 12.6 3.6 61 70.1 4.9
Mood 6 6.9 2.7 22 25.3 4.7 59 67.8 5.0
Hallucinations 4 4.7 2.3 4 4.7 2.3 78 90.7 3.2
 CANNABIS AND PALLIATIVE CARE
Discussion
Patients with serious illness who use medical cannabis
under the supervision of an APC practice report improved
pain, symptoms, and a sense of well-being with few reported
harms, although a minority of patients endorsed symptoms
suggestive of cannabis withdrawal. This study is significant
because it explores the benefits and harms of cannabis use in
patients with serious illness being seen in an APC that reg-
isters patients for medical cannabis cards, and counsels and
monitors patients’ cannabis use. This differs from studies in
palliative care in which only providers’ perceptions are ex-
plored11 or cannabis is framed as an illicit drug.12
The scope of these data extends beyond the state of Georgia
to all states with medical cannabis laws. Of note, Georgia and
a number of states predominantly in the Southeastern United
States have medical cannabis laws but no dispensaries, which
may explain why surveyed patients who had low THC oil
Downloaded by Linkoping University Library from www.liebertpub.com at 09/21/19. For personal use only.
cards were also using other cannabis products.13 Further re-
search is needed to clarify the impact of state cannabis laws on
patterns of cannabis use.
One startling finding in our study is the high percentage of
cancer patients who felt that cannabis has curative properties.
This perception may result from messaging in popular me-
dia,14 and necessitates more research to understand how and
why certain patients perceive cannabis to be curative.
Our study has several key limitations. We did not verify
actual use of cannabis products by drug screening and relied
on self-report of cannabis use alone. Our cohort was a self-
selected patient population and likely had a greater baseline
interest in cannabis as a medical therapy. We limited our
study to patients who requested to have a state-sanctioned
card and to those who reported use of cannabis. Given the
lack of regulation over cannabis products and patients’ di-
verse use of multiple formulations, we were not able to cor-
relate symptom response to specific formulations.
There are significant barriers limiting cannabis-related
research including its Schedule I designation.1,15,16 Future
investigations should prospectively study the impact of stan-
dardized, controlled cannabis formulations on pain, symp-
toms, and quality of life in populations with serious illness.
Based on our clinical experience observing patients’ cannabis
use in our APC, the role of cannabis in treating comorbid
chronic nonmalignant pain, malignant pain, opiate depen-
dence/SUDs, and caregiver quality of life is worthy of inquiry.
Acknowledgments
We thank David Murphy, MD, PhD, and Tammie E.
Quest, MD for their contributions to this article.
Funding Information
Ali J. Zarrabi–Fostering the Academic Mission in the
Emory Department of Medicine (FAME) Grant, Emory
University. Vinita Singh is a KLZ Scholar at the Georgia
Clinical and Translational Science Alliance.
Supplementary Material
Supplementary Figure S1
Supplementary Table S1
Supplementary Table S2
5
References
1. Haffajee RL, MacCoun RJ, Mello MM: Behind schedule—
Reconciling federal and state marijuana policy. N Engl J
Med 2018;379:501–504.
2. Pergam SA, Woodfield MC, Lee CM, et al.: Cannabis use
among patients at a comprehensive cancer center in a state
with legalized medicinal and recreational use. Cancer 2017;
123:4488–4497.
3. Cyr C, Arboleda MF, Aggarwal SK, et al.: Cannabis in
palliative care: Current challenges and practical recom-
mendations. Ann Palliat Med 2018;7:463–477.
4. National Cancer Opinion Survey, Harris Poll on behalf
of ASCO, 2018. https://www.asco.org/research-progress/
reports-studies/national-cancer-opinion-survey (last accessed
September 11, 2019).
5. Choo EK, Feldstein Ewing SW, Lovejoy TI: Opioids out,
cannabis in: Negotiating the unknowns in patient care for
chronic pain. JAMA 2016;316:1763–1764.
6. Haleigh’s Hope Act; GA Code x 16-12-190 (2015).
7. Arthur J, Edwards T, Reddy S, et al.: Outcomes of a spe-
cialized interdisciplinary approach for patients with cancer
with aberrant opioid-related behavior. Oncologist 2018;23:
263–270.
8. Arthur JA, Haider A, Edwards T, et al.: Aberrant opioid use
and urine drug testing in outpatient palliative care. J Palliat
Med 2016;19:778–782.
9. Schenker Y, Merlin JS, Quill TE: Use of palliative care
earlier in the disease course in the context of the opioid
epidemic: Educational, research, and policy issues. JAMA
2018;320:871–872.
10. Zarrabi A, Broglio K, Drew Rosielle, et al.: Medical Cannabis
in Outpatient Palliative Care: Evidence and Approaches for
Integrating Cannabis into Clinical Practice. Orlando, FL:
American Academy of Hospice and Palliative Care Annual
Assembly, 2019.
11. Luba R, Earleywine M, Farmer S, Slavin M.: Cannabis in end-
of-life care: Examining attitudes and practices of palliative
care providers. J Psychoactive Drugs 2018;50:348–354.
12. Donovan KA, Chang YD, Oberoi-Jassal R, et al.: Rela-
tionship of cannabis use to patient-reported symptoms in
cancer patients seeking supportive/palliative care. J Palliat
Med 2019. [Epub ahead of print]; DOI: 10.1089/jpm.2018
.0533.
13. Lopez G: Marijuana is legal for medical purposes in 33
states, Vox, Editor. 2019. https://www.vox.com/identities/
2018/8/20/17938366/medical-marijuana-legalization-states-
map (last accessed May 11, 2019).
14. Shi S, Brant AR, Sabolch A, Pollom E.: False news of a
cannabis cancer cure. Cureus 2019;11:e3918.
15. Hazekamp A: The trouble with CBD oil. Med Cannabis
Cannabinoids 2018;1:65–72.
16. Kilmer B: Recreational cannabis—Minimizing the health
risks from legalization. N Engl J Med 2017;376:705–707.
Address correspondence to:
Ali John Zarrabi, MD
Emory University
1365 Clifton Road, Clinic A, 4th Floor
Atlanta, GA 30322
E-mail: ali.zarrabi@emoryhealthcare.org