Introduction
Opioids are the cornerstone of managing moderate–to-severe cancer-related pain (Mercadante Reference Mercadante2023). Similar to the general population, cancer patients are also at risk of opioid-related overdoses (Merlin et al. Reference Merlin, Black and Midboe2023). It is recommended that naloxone, a fast-acting opioid antagonist, be co-prescribed along with opioids for patients with a high risk of an accidental overdose (Dowell et al. Reference Dowell, Haegerich and Chou2016). Some characteristics of high-risk individuals include those who are on a high dose of opioids (MEDD > 90 mg), methadone use, concurrent use of opioids and benzodiazepines and/or other sedating drugs, history of overdose, pulmonary (Chronic Obstructive Pulmonary Disease COPD, Obstructive Sleep Apnea OSA, and oxygen use), renal, and liver disease (WHO 2025). Oftentimes, the lifesaving responsibility of administering naloxone falls on a caregiver (Chimbar and Moleta Reference Chimbar and Moleta2018). This highlights the importance of caregivers’ role and their understanding of naloxone and preparedness in case of an accidental overdose among patients with cancer.
Perceptions of naloxone among clinicians significantly influence prescribing practices (Binswanger et al. Reference Binswanger, Koester and Mueller2015; Haug et al. Reference Haug, Bielenberg and Linder2016; Wilson et al. Reference Wilson, Spicyn and Matson2016). Clinicians often cite barriers such as limited confidence in identifying at-risk individuals and uncertainty about how to educate patients on naloxone use (Binswanger et al. Reference Binswanger, Koester and Mueller2015; Haug et al. Reference Haug, Bielenberg and Linder2016; Wilson et al. Reference Wilson, Spicyn and Matson2016). While most patients view naloxone positively, some express concerns about stigma or being perceived as misusing opioids (Binswanger et al. Reference Binswanger, Koester and Mueller2015; Haug et al. Reference Haug, Bielenberg and Linder2016; Wilson et al. Reference Wilson, Spicyn and Matson2016). In a cross-sectional survey among patients with cancer (n = 150) in an outpatient tertiary cancer center, 70% of participants identified their caregiver’s discomfort with the administration of naloxone as a concern (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). A qualitative study of 40 caregivers who lived with individuals at high risk of overdose and used rural pharmacies found that caregivers tend to prefer non-stigmatizing language, such as “bad reaction” instead of “overdose,” and respond more favorably to analogies like “EpiPen” to understand naloxone’s role (Hosea et al. Reference Hosea, Mashburn and Kennelty2023).
Despite their vital role in reversing accidental overdose events, caregiver perspectives on naloxone remain underexplored. Addressing this gap is essential for improving the safety and acceptability of naloxone prescription in high-risk cancer patients receiving opioids. Our team previously published a cross-sectional survey study, which showed that cancer patients with increased risk of overdose perceived naloxone as beneficial (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). In this study, our aim was to capture the perception toward naloxone prescription amongst caregivers of high-risk cancer patients receiving opioids (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). Due to the limited research, we wanted to explore perceived risk vs benefits of naloxone prescription, assess the cancer patient’s caregiver beliefs, knowledge, and confidence in administering naloxone, and examine the association between patients’ and caregivers’ perceptions.
Methods
This prospective cross-sectional survey study, approved by MD Anderson’s Institutional Review Board, was conducted among established patients seen in the outpatient Supportive Care Center (SCC) for cancer pain at MD Anderson Cancer Center. The patient perceptions have been previously reported (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). This study focused on the caregiver characteristics and their perceptions regarding naloxone prescription.
Participant enrollment
In our SCC, patients who carry risk factors for opioid overdose and caregivers (when present) are provided with a prescription for intranasal naloxone along with accompanying education regarding opioid safety, recognizing signs of an overdose, and proper naloxone administration techniques.
Methods were detailed in our original published study (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). In brief, a total of 150 high-risk cancer patients and 29 caregivers were included in the study. The study was conducted amongst our outpatient SCC patients who were followed for ongoing cancer-related pain & other symptoms between 18 July 2020 and 30 April 2022. Eligible patients were >18 years old, had active cancer, could read and consent in English, were on opioids and received a co-prescription of take-home naloxone.
Survey administration
Due to COVID-19 pandemic-related restrictions, the survey was administered to caregivers who were accessible by telephone after a telemedicine visit or during an in-person visit with enrolled patients (Reddy et al. Reference Reddy, Arthur and Dalal2021). A 28-item survey was administered to caregivers assessing their overall understanding of opioids and naloxone, the risk of overdose for their loved ones and naloxone administration. As described in the original study, survey questions were formulated based on recommendations from the Centers for Disease Control and Prevention, the National Comprehensive Cancer Network and a literature review (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024).
Details on patient survey administration were described in the original study (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). In brief, patients were enrolled based on their availability for in-person versus telephone interviews, and a survey comprising 29 statements was administered to understand patients’ knowledge of opioids, risk for overdoses, and naloxone administration (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024).
COVID-19-related restrictions limited the establishment of caregiver-patient dyads, as caregivers were often not present during visits and not reliably reachable by phone.
Statistical analysis
The primary objective of this study was to determine the caregivers’ perception of a prescription of naloxone. Variables of interest included caregivers’ demographics, clinical information, and perceptions regarding the prescription of naloxone in a variety of aspects. Univariate analysis was performed using a T-test, chi-squared test, or Fisher’s exact test. Results with p-values that were less than 0.05 were considered statistically significant, as all testing was two-sided. Cohen’s κ was run to determine if there was agreement between patients’ and caregivers’ judgment whether naloxone was beneficial. Interpretation of the results was based on the original paper by Landis et al (Landis and Koch Reference Landis and Koch1977). We used SAS 9.4 software for data analysis.
Results
A total of 150 cancer patients and 29 corresponding caregivers were included in this analysis. Overall, 100/150 (67%) patients and 23/29 (79%) caregivers perceived naloxone prescription as beneficial. The majority of patients who participated in the survey who found naloxone prescription beneficial were retired, white, and male with a bachelor’s degree who did not have a history of smoking, drug, or marijuana use (Amaram-Davila et al. Reference Amaram-Davila, Vega and Kim2024). Table 1 shows the caregiver characteristics. There were no significant differences in the caregiver characteristics between those who agreed that a prescription of naloxone was beneficial for their loved one and those who disagreed.
Characteristics of caregivers surveyed
Table 1 Long description
Caregiver characteristics are reported for Overall, Agree That Prescription Was Beneficial, and Disagree or Neutral That Prescription Was Beneficial, with p values shown for each variable. Age had N 28 overall, with mean and standard deviation 55.79 and 12.00 overall, 54.45 and 12.63 for agree, and 60.67 and 8.41 for disagree or neutral; minimum to maximum ages were 26.00 to 74.00 overall, 26.00 to 73.00 for agree, and 49.00 to 74.00 for disagree or neutral, with p value 0.2688. Gender counts were 8 male and 21 female overall; 6 male and 17 female for agree; and 2 male and 4 female for disagree or neutral, with p value 1.0000 marked with an asterisk. Ethnicity counts were 18 White, 5 Hispanic, and 6 Others overall; 13 White, 4 Hispanic, and 6 Others for agree; and 5 White, 1 Hispanic, and 0 Others for disagree or neutral, with p value 0.3700 marked with an asterisk. Employment status counts were 19 employed and 10 unemployed or retired overall; 14 employed and 9 unemployed or retired for agree; and 5 employed and 1 unemployed or retired for disagree or neutral, with p value 0.6328 marked with an asterisk. Education counts were 13 some college high school or others and 16 bachelor advanced degree overall; 11 and 12 for agree; and 2 and 4 for disagree or neutral, with p value 0.6628 marked with an asterisk. History of smoking counts were 25 never smoker and 4 previous smoker overall; 21 and 2 for agree; and 4 and 2 for disagree or neutral, with p value 0.1798 marked with an asterisk. History of drug use other than Marijuana counts were 28 never used and 1 previous user overall; 23 never used and 0 previous user for agree; and 5 never used and 1 previous user for disagree or neutral, with p value 0.2069 marked with an asterisk. History of Marijuana use counts were 24 never used, 4 previous user, and 1 unknown or missing overall; 21 never used, 2 previous user, and 0 unknown or missing for agree; and 3 never used, 2 previous user, and 1 unknown or missing for disagree or neutral, with p value 0.1350 marked with an asterisk.
* Fisher’s exact test.
Table 2 shows the perceptions and beliefs regarding the co-prescription of naloxone nasal spray through their survey responses. Results show that caregivers felt confident that they can administer naloxone nasal spray appropriately and call for help in case of accidental overdose by their loved ones. Agreement between patient and caregiver perceptions was fair (Cohen’s κ = 0.34, p = 0.034).
Patient and caregiver survey answers
Table 2 Long description
Survey questions compare Patients, n 150, and Caregivers, n 29, reporting count and percent for the response, Agree, plus p-values3. Narcan prescription was seen as beneficial by 100 patients, 66.7 percent, and 23 caregivers, 79.3 percent, p-values3 0.4531. Risk awareness items show Agree for taking pain meds with other sedating substances at 138 patients, 92.6 percent, and 28 caregivers, 96.6 percent, p-values3 1.0000, and for taking more than prescribed at 144 patients, 96.0 percent, and 28 caregivers, 96.6 percent, p-values3 1.0000. Awareness of dangers to others not prescribed the medication shows Agree at 148 patients, 98.7 percent, and 28 caregivers, 96.6 percent, p-values3 1.0000, and one current pain killer could be fatal to someone else shows Agree at 129 patients, 86.6 percent, and 24 caregivers, 82.8 percent, p-values3 1.0000. Feeling safe having Narcan at home shows Agree at 124 patients, 83.2 percent, and 28 caregivers, 96.6 percent, p-values3 1.0000; education at prescribing shows Agree at 127 patients, 84.7 percent, with no caregiver value shown. Caregiver presence at initial prescribing shows Yes at 77 patients, 51.3 percent, and 22 caregivers, 75.9 percent, p-values3 0.5000; among those present, education for the caregiver shows Agree at 70 patients, 92.1 percent, and 20 caregivers, 87.0 percent, with p-values3 listed as NA2, and Unknown or Missing at 74 patients and 6 caregivers. Confidence to respond appropriately shows Agree at 129 patients, 86.6 percent, and 26 caregivers, 89.7 percent, p-values3 1.0000; additional distress item shows Agree at 44 patients, 29.7 percent, and 17 caregivers, 58.6 percent, with no p-values3 shown. Feeling at risk for overdose shows Agree at 4 patients, 2.7 percent, and 2 caregivers, 6.9 percent, p-values3 0.5000; alcohol plus pain medication risk shows Agree at 144 patients, 96.6 percent, and 29 caregivers, 100.0 percent, with p-values3 listed as NA2. Feeling suspected of inappropriate use shows Agree at 20 patients, 13.5 percent, and 0 caregivers, with p-values3 listed as NA2; being less careful shows Agree at 1 patient, 0.7 percent, and 1 caregiver, 3.4 percent, p-values3 1.0000. Worry about needing pain medications due to overdose risk shows Agree at 11 patients, 7.4 percent, and 13 caregivers, 44.8 percent, p-values3 1.0000; belief that a loved one needs Narcan shows a caregiver Agree value of 13, 44.8 percent, with no patient value shown. Several rows list Unknown or Missing counts, including 1 patient and 0 caregivers for sedating substances risk, 1 and 0 for fatal to someone else, 1 and 0 for safe at home, 1 and 0 for confidence, 2 and 0 for additional distress, 1 and 0 for alcohol risk, 2 and 0 for suspected inappropriate use, 4 and 0 for less careful, and 1 and 0 for worry about needing pain medications.
Discussion
Patients and caregivers understand the importance of a naloxone prescription. Yet over the years, we have seen that there may be a disconnect in the number of prescriptions offered to patients on opioid pain medications. Reasons have included limited physician understanding, concerns for stigmatizing patients, and worries that providing naloxone might encourage individuals to use opioids less cautiously (Bennett and Elliott Reference Bennett and Elliott2021). Amaram-Davila et al showed that efforts made for educational intervention to clinicians unfortunately did not increase naloxone prescription rates (Amaram-Davila et al. Reference Amaram-Davila, Reddy and Arthur2023). Our study supported that both patients and caregivers in this study perceived naloxone prescription as beneficial and unlikely to cause harm, supporting the acceptability of naloxone co-prescription in oncology settings. Patients who were surveyed were followed by our SCC for ongoing cancer-related pain and their caregivers who were involved in multiple levels of their loved one’s care. Only fair agreement between patient and caregiver perceptions highlights the need for improved communication around naloxone’s purpose and use.
Our survey results suggest that patients and caregivers understand the importance of appropriate opioid use and are aware of the potential adverse effects of taking pain medications other than as prescribed. These findings help avoid the misconception that prescribing Naloxone encourages inappropriate opioid use. A total of 100% of caregivers understood that mixing alcohol and pain medications could increase the risk of overdose resulting in harm or death. Also, 100% of caregivers did not feel that a prescription of naloxone labeled their loved one as someone who would inappropriately use pain medications. Overall, the survey responses help us remove assumptions that a naloxone prescription could stigmatize, mislabel, or encourage further opioid use without caution.
One major limitation of this study is its small sample size. Caregiver enrollment was significantly affected by COVID-19 pandemic-related visitation policy restrictions. However, our study adds value to the much less explored area of perception toward naloxone among caregivers of those with cancer. These findings are reassuring in that caregivers generally value naloxone prescriptions. However, some caregivers may continue to worry about the discomfort, distress, or stigma associated with a naloxone prescription and its administration. Since caregivers play a critical role in promptly administering naloxone during an emergency, shared understanding and agreement between patients and caregivers regarding its importance are essential. Naloxone prescriptions are less likely to be filled when patients do not perceive the medication as beneficial, and caregivers are less likely to seek education or maintain ready access to it when they do not recognize its value. Future work should identify which caregiver subgroups experience these concerns and evaluate strategies to address them – such as individualized education, hands-on demonstration of naloxone administration, and structured follow-up discussions. Additional research is also warranted to guide safer opioid prescribing practices for patients without caregiver support, ensuring equitable access to overdose prevention resources in cancer care.
Conclusion
Our study provides helpful insights into caregivers of cancer patients’ perceptions of naloxone prescription. Despite the small sample size, our findings encourage us to continue prescribing naloxone to our high-risk cancer patients receiving opioids and providing comprehensive education to both patients and caregivers.
Acknowledgments
We want to acknowledge the late Ms. Diana Urbauer (Department of Biostatistics (D.U.), The University of Texas MD Anderson Cancer, Houston, Texas, USA), who passed away on 4 June 2024, for her invaluable contributions to the statistical analysis. The authors did not use any AI and AI-assisted technologies to write this manuscript.
Author contributions
V.P. and J.A.-D. have joint first authorship. A.R. has senior authorship.
Funding
A.R. has received research funding for MD Anderson Cancer Center from Teikoku Pharma USA. This research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Competing interests
V.P., J.A.-D., A.R.d.M., L.T., and E.B. declare no conflict of interest.
Open access
