If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Requests for reprints should be addressed to Mukaila A. Raji, MD, MS(Pharm), Division of Geriatrics & Palliative Medicine, The University of Texas Medical Branch, 301 University Blvd. 6.600 Rebecca Sealy, Galveston, TX 77555-0177.
Division of Geriatrics and Palliative Medicine, Department of Internal MedicineSealy Center on AgingDepartment of Preventive Medicine and Population Health
Division of Geriatrics and Palliative Medicine, Department of Internal MedicineSealy Center on AgingDepartment of Preventive Medicine and Population HealthOffice of Biostatistics, University of Texas Medical Branch, Galveston
No data exist on comparative risk of cardiac arrhythmias among 3 Medication-Assisted Therapy (MAT) medications in patients with opioid use disorder. Understanding MAT medications with the least risk of arrhythmia can guide clinical decision-making.
Method
A multicenter retrospective cohort study was performed of patients 18 years or older diagnosed with opioid use disorder by the International Classification of Diseases, 10th revision, Clinical Modification without baseline arrhythmia in 2018-2019, using Clinformatics Data Mart Database (Optum, Eden Prairie, Minn). Everyone required 1 year of continuous enrollment prior to and after the diagnosis. Patients with MAT were propensity score-matched to those without MAT. Primary outcome was rate of arrhythmia across MAT (methadone, naltrexone, and buprenorphine). A multivariable logistic regression model was built to examine the outcome difference across 3 medications adjusted for patient's demographic and comorbidity.
Result
Only 14.1% of the 66,083 patients with opioid use disorder received MAT prescriptions in the 12 months after diagnosis. New-onset arrhythmia diagnoses occur more frequently among MAT vs non-MAT users (4.86% vs 3.92%), with 29% risk of incident arrhythmias among MAT users, even after adjusting relevant confounders (adjusted odds ratio [aOR] 1.29; 95% confidence interval [CI], 1.11-1.52). Incidence of arrhythmia varied by drugs: naltrexone (9.57%), methadone (5.71%), and buprenorphine (3.81%). Difference among the MAT drugs in incidence of arrhythmia remained significant even after adjusting covariates (aOR 2.44; 95% CI, 1.63-3.64 and buprenorphine aOR 0.77; 95% CI, 0.59-1.00, with methadone as reference).
Conclusion
MAT users had higher risk of cardiac arrhythmia than non-users. Naltrexone is associated with the highest risk of arrhythmia, suggesting caution with naltrexone use, especially in opioid use disorder patients with pre-existing heart conditions.
Patients with opioid use disorder who received medication-assisted therapy had a higher risk of cardiac arrhythmia than those who did not.
•
Rates of new-onset arrhythmia varied by medication type: naltrexone 9.57%, methadone 5.71%, and buprenorphine at 3.81%.
•
Our findings suggest that in opioid use disorder patients with significant cardiovascular conditions—especially conditions that increase the risk of arrhythmias—clinicians should consider buprenorphine as first-line therapy.
Introduction
Opioid overdose death—a growing public health problem in the United States—has substantially worsened during the ongoing COVID-19 pandemic.
A key evidence-based intervention to treat opioid use disorder and mitigate opioid overdose death is Medication-Assisted Therapy (MAT, also called medication for opioid use disorder or Medication-Assisted Treatment).
MAT medication use is associated with a significant reduction in opioid relapse after recovery and in opioid-related acute care use, overdose, and deaths, compared with no treatment.
Yet, MAT medications (buprenorphine, naltrexone, and methadone) are underused in opioid use disorder patients, reflecting multiple factors: access, cost, and insurance barriers, stigma, prescriber inertia, policies that restrict MAT prescribing, and concerns about cardiovascular and other side effects.
An understudied area of cardiovascular toxicity concerns related to the comparative risks of cardiac arrhythmias during treatment with any of the 3 MAT drugs: buprenorphine, naltrexone, and methadone. While the study shows the association of methadone use and increased odds of QT prolongation, few data exist on comparative cardiac toxicity among these 3 medications when used in the setting of opioid use disorder treatment.
We thus used nationally representative population-based data to examine comparative cardiovascular toxicity of MAT for opioid use disorder and their relative risks for QT prolongations and onset of new arrhythmias. Understanding MAT medications with the least risk of arrhythmia can guide clinical decision-making, especially in opioid use disorder patients with pre-existing cardiovascular conditions.
Methods
Data Source
A retrospective cohort study of patients diagnosed with opioid use disorder in 2018-2019 was performed using administrative claims data extracted from Optum's de-identified Clinformatics Data Mart (CDM) Database (Optum, Eden Prairie, Minn). CDM is one of the nation's largest commercial insurance databases containing patient demographics and clinical information such as prescription drugs dispensed and outpatient and inpatient claims. This study was reviewed and approved by the University of Texas Medical Branch Institutional Review Board.
Study Cohort
We identified patients aged 18 years and older in the analysis if they had a diagnosis of opioid use disorder in 2018-2019. Eligible individuals were identified using the International Classification of Diseases, 10th Revision, Clinical Modification (ICD‐10‐CM) based on the Centers for Medicare & Medicaid Services CMS Chronic Condition Data Warehouse (Supplementary Table 1, available online). The index date as defined by the first opioid use disorder diagnosis is 2018-2019. Patients were excluded if they were not continuously enrolled in the 12 months prior to and the 12 months following the index date. Another exclusion criterion was arrhythmia diagnoses in the 12 months prior to the index date, giving a final analytical sample size of 66,083 patients. The control group included those without any MAT in the 12 months after the index date. The Supplementary Figure (available online) details the cohort selection flowchart. For each patient, we identified whether they received MAT in the 12 months after the index date. We used National Drug Codes in pharmacy claims and Common Procedural Terminology codes in medical claims to identify MAT by receiving prescription or injection of methadone, naltrexone, and buprenorphine (Supplementary Table 2, available online). The control group included those without any MAT in the 12 months after the index date.
Supplementary FigureDiagram for selection of opioid use disorder (OUD) patients with and without Medication-Assisted Therapy (MAT).
We assessed whether both groups were diagnosed with arrhythmias, including long QT prolongation, within 1 year after the index date. Diagnosis of cardiac arrhythmias was based on the definition from the Elixhauser Comorbidity Index Supplementary Table 1. shows the ICD-10 codes of the study outcome. Age at the index date was obtained from the CDM database. We examined and adjusted for all conditions 12 months prior to the index date, which was included in the Elixhauser Comorbidity Index. Each condition was examined as a separate covariate. We also adjusted for sex and the geographic region (Midwest, Northeast, South, West) of each patient. To control for differences between the 2 groups, we performed propensity score matching. The propensity score of having MAT was generated using a logistic regression model including age, sex, region, and comorbidity related to MAT use and arrhythmias (congestive heart failure, chronic pulmonary disease, diabetes, fluid and electrolyte disorders, and hyperthyroidism). For each patient with MAT, we performed greedy nearest neighbor matching to select 2 patients without MAT within a caliper equal to 0.2 standard deviations (SD) of the logit of the propensity score. The date of first MAT use from the MAT user was assigned to the 2 matched non-MAT users. Then, we excluded the pairs in which MAT users had arrhythmias that occurred prior to MAT use, the pairs in which both non-MAT users had arrhythmias that occurred prior to their assigned MAT use, and the individual non-MAT user who had arrhythmias that appeared prior to their assigned MAT user. Our final propensity match cohort included 7511 pairs, with 5743 of them having 2 matched non-MAT users and 1768 of them having 1 matched non-MAT user. We regenerated the propensity score model for sensitivity analysis, including all Elixhauser comorbidity, hyperthyroidism, age, sex, and region. Then we repeated the matching process and had 7489 pairs of patients.
Statistics Analysis
Mean (SD) and frequency of patient characteristics and comorbidity were calculated for both groups prior to propensity match and compared by t test for continuous variables and chi-squared test for categorical variables. After the propensity score match, we used the standardized difference to assess the balance of covariates between 2 groups. A conditional logistic regression model was built to examine the association between MAT use and arrhythmia in the propensity match cohort. We further compared the rate of arrhythmia across 3 medications (methadone, naltrexone, and buprenorphine). In these analyses, we excluded 29 patients with more than one type of medication in the 12 months after the index date. A multivariable logistic regression model was built to examine the outcome difference across 3 medications adjusted for patient's demographic and comorbidity. All tests of statistical significance were 2-sided, and analyses were performed with SAS 9.4 (SAS Institute, Cary, NC).
Results
Among 66,083 opioid use disorder patients who met selection criteria, 14.1% received MATs in the 12 months after diagnosis (Supplementary Table 3, available online). Patients who received MATs were younger, more likely to be male, residing in Northeast and Midwest regions, and have alcohol use disorder, depression, and liver disease, but less likely to have other chronic conditions. To account for these differences, we conduct propensity matching. After matching, patient demographics and comorbidity related to MAT and arrhythmia between MAT and non-MAT groups were very similar. The SD was <0.1, which shows these characteristics were well balanced between the 2 groups after matching (Table 1). However, the comorbidity not included in the propensity score model was different between the 2 groups (Table 2). Alcohol abuse was less common in the MAT group (9.77%) than in the non-MAT group (13.73%). Depression was less prevalent (39.66%) in the MAT group than in the non-MAT group (45.84%). On the contrary, 20.59% were obese in the MAT group, compared with 16.39% in the non-MAT group. The rate of arrhythmias was 4.86% and 3.92% among MAT and non-MAT groups, respectively. After adjusting for comorbidity, which was significantly different between the 2 groups, the risk of arrhythmias was 29% higher among MAT users (adjusted odds ratio [aOR] 1.29; 95% confidence interval [CI], 1.11-1.52) (Table 3). Other comorbidity associated with arrhythmias included valvular disease (aOR 2.55; 95% CI, 1.65-3.93), peripheral vascular disorders (aOR 1.51; 95% CI, 1.14-2.00), renal failure (aOR 1.40; 95% CI, 1.03-1.92), and liver disease (aOR 1.43; 95% CI, 1.06-1.92).
Table 1Descriptive Statistics of Patient Demographics and Comorbidity Related to MAT Use and Arrhythmias Between Patients with and Without MATs in the Matched Cohort (n = 20,765)
Among MAT users, we found that the rate of arrhythmia with the use of naltrexone was higher (9.57%) compared with methadone (5.71%), and the rate of arrhythmia was the least with the help of buprenorphine (3.81%). There were significant age differences among patients in different MAT groups. Patients in the methadone group were older (mean age 56.96 years) compared with buprenorphine (mean age 51.11) and naltrexone (mean age 37.93). Alcohol abuse was highest in naltrexone (50.76%) compared with buprenorphine and methadone group (Table 4). This finding can be explained because naltrexone is also used to treat patients with alcohol use disorder. After adjusting for demographics and comorbidity, the use of naltrexone was associated with a more than 2 times higher risk of arrhythmias than methadone (aOR 2.43; 95% CI, 1.61-3.65) (Table 5). In contrast, the use of buprenorphine was associated with a lower risk of arrhythmias than the use of methadone; however, this difference was marginally significant (aOR 0.78; 95% CI, 0.60-1.02).
Table 4Descriptive Statistics of Patient Demographics and Comorbidity Across Patients Received Methadone, Naltrexone, and Buprenorphine (n = 7482)
Variables
Methadone n = 1612 n (%)
Naltrexone n = 721 n (%)
Buprenorphine n = 5149 n (%)
Overall n = 7482*
P Value
Age, mean (SD)
56.96 (13.62)
37.93 (14.66)
51.11 (14.84)
51.18 (15.35)
.0001
Sex
Female
777 (48.2)
323 (44.8)
2410 (46.81)
3510
.3024
Male
835 (51.8)
398 (55.2)
2739 (53.19)
3972
Region
Midwest
246 (15.26)
164 (22.75)
752 (14.6)
1162
.0001
Northeast
171 (10.61)
97 (13.45)
628 (12.2)
896
South
647 (40.14)
284 (39.39)
2538 (49.29)
3469
West
548 (34)
176 (24.41)
1231 (23.91)
1955
Elixhauser comorbidities
Alcohol abuse
96 (5.96)
366 (50.76)
563 (10.93)
1025
.0001
Blood loss anemia
15 (0.93)
6 (0.83)
48 (0.93)
69
.9652
Congestive heart failure
79 (4.9)
9 (1.25)
198 (3.85)
286
.0001
Chronic pulmonary disease
445 (27.61)
116 (16.09)
1253 (24.33)
1814
.0001
Coagulopathy
78 (4.84)
13 (1.8)
124 (2.41)
215
.0001
Deficiency anemia
116 (7.2)
30 (4.16)
297 (5.77)
443
.0115
Depression
666 (41.32)
429 (59.5)
2331 (45.27)
3426
.0001
Diabetes
407 (25.25)
52 (7.21)
851 (16.53)
1310
.0001
Fluid and electrolyte disorders
192 (11.91)
87 (12.07)
502 (9.75)
781
.015
AIDS/HIV
12 (0.74)
7 (0.97)
28 (0.54)
47
.3178
Hypertension complicated
145 (9)
18 (2.5)
287 (5.57)
450
.0001
Hypertension uncomplicated
847 (52.54)
178 (24.69)
2261 (43.91)
3286
.0001
Hypothyroidism
289 (17.93)
72 (9.99)
736 (14.29)
1097
.0001
Hyperthyroidism
12 (0.74)
5 (0.69)
60 (1.17)
77
.2212
Liver disease
170 (10.55)
60 (8.32)
509 (9.89)
739
.2503
Cancer
115 (7.13)
15 (2.08)
173 (3.36)
303
.0001
Obesity
347 (21.53)
94 (13.04)
786 (15.27)
1227
.0001
Other neurological disorders
150 (9.31)
51 (7.07)
391 (7.59)
592
.0577
Pulmonary circulation disorders
48 (2.98)
7 (0.97)
88 (1.71)
143
.0008
Peptic ulcer disease excluding bleeding
23 (1.43)
NA
90 (1.75)
116
.0229
Peripheral vascular disorders
232 (14.39)
19 (2.64)
386 (7.5)
637
.0001
Paralysis
30 (1.86)
4 (0.55)
46 (0.89)
80
.0016
Psychoses
35 (2.17)
45 (6.24)
193 (3.75)
273
.0001
Renal failure
183 (11.35)
14 (1.94)
288 (5.59)
485
.0001
Rheumatoid arthritis/collagen
269 (16.69)
32 (4.44)
643 (12.49)
944
.0001
Valvular disease
71 (4.4)
11 (1.53)
161 (3.13)
243
.001
Weight loss
78 (4.84)
25 (3.47)
221 (4.29)
324
.3135
AIDS = acquired immunodeficiency syndrome; HIV = human immunodeficiency virus.
Supplementary Table 1International Classification of Diseases, 10th Revision, Clinical Modification (ICD‐10‐CM) Based on the CMS Chronic Condition Data Warehouse for OUD, Arrhythmia, and Long QT Syndrome
Our sensitivity analyses with the matching propensity score including all comorbidity were well balanced between MAT and non-MAT groups (Supplementary Table 4, available online). The association between MAT use and arrhythmias was robust (aOR 1.30; 95% CI, 1.13-1.50). Also, the association between type of medication and arrhythmias remained similar for naltrexone (aOR 2.44; 95% CI, 1.63-3.64) and buprenorphine (aOR 0.77; 95% CI, 0.59-1.00) in comparison with methadone.
Supplementary Table 4Descriptive Statistics of Patient Demographics and Comorbidity Between Patients with and Without MATs in the Matched Cohort from Sensitivity Analyses
Our findings can be summarized as follows. Only 14.1% of the 66,083 patients with opioid use disorder received MAT prescriptions in the 12 months after diagnosis. New-onset arrhythmia diagnoses occur more frequently among MAT vs non-MAT users (4.86% vs 3.92%), with a 29% risk of incident arrhythmias among MAT users, even after adjusting for relevant confounders. The incidence of arrhythmia varied by drugs: naltrexone (9.57%), methadone (5.71%), and buprenorphine (3.81%). The difference among the MAT drugs in arrhythmia risks remained significant even after adjustment for covariates: naltrexone is about 2.44 times, and buprenorphine is 0.77 times as likely as methadone to be linked to new-onset arrhythmia.
The overall low rate of MAT prescribing for opioid use disorder patients is consistent with prior research that showed a low rate of MAT use, ranging from 10% to 40% of opioid use disorder patients receiving MAT.
found that <10% of the 6932 Medicare enrollees who died from opioid overdose in 2012-2016 had any prescription for MAT. The persistently low adoption of MAT intervention for the opioid use disorder population is concerning, considering the proven effectiveness of MAT in saving lives and treating opioid use disorder.
The low rate of MAT use underscores the urgent need for the development and implementation of practical policy and practice guidelines by state and federal agencies, insurance payers, and health systems to lessen barriers to MAT prescribing for the growing population of opioid use disorder patients.
National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Health Sciences Policy; Committee on Medication-Assisted Treatment for Opioid Use Disorder
Mancher M Leshner AI Medications for Opioid Use Disorder Save Lives. National Academies Press (US),
Washington, DC2019
This is particularly critical at this moment of the COVID-19 pandemic, with its associated social isolation and limited access to medical care. For example, state and federal policies should consider allowing all clinicians who already prescribe opioid analgesics to have automatic approval to prescribe MAT drugs for individuals living with opioid use disorder. Government policies should also consider the elimination of the limit on opioid use disorder patient census in the clinicians' patient panels. Finally, decision-makers in insurance companies and health systems should consider removal of the onerous preauthorization process for MAT prescribing.
National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Health Sciences Policy; Committee on Medication-Assisted Treatment for Opioid Use Disorder
Mancher M Leshner AI Medications for Opioid Use Disorder Save Lives. National Academies Press (US),
Washington, DC2019
Our finding of the highest rate of incident arrhythmia in naltrexone users is unexpected, given the large body of literature on the arrhythmogenic risk of methadone. The reason for this finding is unclear. One possibility for the higher arrhythmia rate in naltrexone than methadone is that MAT prescribers might avoid methadone in opioid use disorder patients with any history of cardiac conditions or abnormal electrocardiography (ECG) findings. This possibility is plausible given that prescribing guidelines recommend ECG to rule out arrhythmias and measure QTc prior to initiating methadone. In this scenario,
high arrhythmia risk patients are screened out when prescribers are considering methadone for opioid use disorder, thus giving the higher rate for naltrexone. Although we excluded participants with prior diagnoses of cardiac arrhythmias, we found that the naltrexone group had a baseline lower rate of heart failure, hypertension, and valve disease than the methadone group. We do not, however, have detailed information on ECG to explore the possibility of differences in the rate of ECG abnormalities in methadone users vs users of other MATs.
Conclusion
Our study showed that the rate of arrhythmia with the use of naltrexone was higher (9.57%) compared with methadone (5.71%), and the rate of arrhythmia was the least with the use of buprenorphine (3.81%). We also found that the rate of arrhythmia was significantly higher (27.59%) when 2 or more MAT medications were used in combination.
Limitations
Our findings must be interpreted in view of several limitations. First, information on outcomes and comorbidities is based on diagnosis codes included in charges for outpatient and hospitalization services. Such diagnoses are not always accurate or complete. Second, given the retrospective nature of this study, undetected selection bias may have affected the findings. For example, patients who received MATs may have been more likely than their counterparts to have had subsequent diagnoses. However, we attempted to address this potential bias by propensity matching for a broad range of demographics and comorbidity. Third, our database lacked information on several important factors, such as race/ethnicity and socioeconomic status. Fourth, prescription claims data do not capture data on drugs obtained outside the plan. Given the various stigmas and restrictions associated with MATs, some patients may have accessed these drugs outside their health care setting.
Clinical Implications
Our findings suggest that in opioid use disorder patients with significant cardiovascular conditions—especially conditions that increase the risk of arrhythmias—clinicians should consider buprenorphine as first-line therapy. Of course, the final clinical decision for MAT prescribing is made on a case-by-case basis, with consideration by the prescriber of the overall clinical profile of the individual patient with opioid use disorder and with the health care preferences of the patient.
National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Health Sciences Policy; Committee on Medication-Assisted Treatment for Opioid Use Disorder
Mancher M Leshner AI Medications for Opioid Use Disorder Save Lives. National Academies Press (US),
Washington, DC2019
Funding: YFK and MAR are supported by the National Institute on Drug Abuse (R01‐DA039192; Kuo, Raji, MPI). The content is solely the authors' responsibility and does not necessarily represent the official views of the National Institutes of Health. XY is supported by a research career development award (K12HD052023: Building Interdisciplinary Research Careers in Women's Health Program-BIRCWH; Berenson, PI) from the National Institutes of Health/Office of the Director (OD)/National Institute of Allergy and Infectious Diseases (NIAID), and Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD).
Conflicts of Interest: None.
Authorship: All authors had access to the data and all the authors had a key role in writing the manuscript.
00095-X&id=gr1.jpg){kind=link}