Abstract
Background
A limited number of studies have analyzed prescribing among Medicare‐enrolled adults at risk for opioid overdoses. The objectives of this study were to evaluate prescribing for naloxone and central nervous system (CNS) active medications and to determine the relationships of patient characteristics with exposure to these medications.
Methods
This was a retrospective cross‐sectional analysis of a Medicare‐enrolled medication therapy management eligible cohort. Patients were stratified into two cohorts, individuals with a mean daily morphine milligram equivalent (MME) dose <50 and individuals with MME ≥50. Medications assessed included benzodiazepines, skeletal muscle relaxants (SMR), hypnotics, gabapentanoids, selective‐serotonin reuptake inhibitors (SSRI), serotonin–norepinephrine reuptake inhibitors (SNRI), tricyclic antidepressants (TCA), antipsychotics, barbiturates, other antiepileptics, hydroxyzine, and naloxone. Chi‐square with odds ratios and logistic regressions determined the relationships of medications and patient characteristics with mean daily MME ≥50. Relationship between medications and opioid dose was adjusted for age and sex.
Results
There were 3452 patients with a daily MME <50 and 1116 with a daily MME ≥50. After adjusting for age and sex, patients with a daily MME ≥50 were more likely to be prescribed hypnotics (OR: 1.41, 95% CI 1.17–1.70), SNRIs (OR: 1.39, 95% CI 1.17–1.64), and naloxone (OR: 3.21, 95% CI 2.49–4.12) (p < 0.001). Nine percent of eligible patients received naloxone. Age groups of persons <85 years of age had 1.58–4.04 (p ≤ 0.004) times the odds of being prescribed a mean daily MME ≥50.
Conclusion
Nearly one‐fourth of patients were prescribed a mean daily opioid therapy of MME ≥50. These patients were more likely to be prescribed hypnotics, SNRIs, and naloxone. Patients receiving chronic high‐dose opioid therapy were more likely to be in age groups of persons <85 years. Naloxone may be underprescribed among eligible adults. Targeted medication services may ensure optimal prescribing among Medicare patients with chronic opioid therapies.
Keywords: central nervous system active medication, chronic use, drug interactions, Medicare, medication therapy management, naloxone, opioid
Key points
After adjusting for age and sex, patients with high chronic opioid doses (mean daily MME ≥50) were more likely to receive naloxone and be prescribed hypnotics and serotonin and norepinephrine reuptake inhibitors compared to patients with lower chronic opioid doses (<50MME).
This study found only 9% of the eligible patients received naloxone over a two‐year period.
Why does this paper matter?
These findings highlight the need for comprehensive ongoing medication reviews among Medicare patients to identify opportunities for deprescribing of potentially harmful medications and prescribing potentially lifesaving naloxone.
INTRODUCTION
In 2016, the Centers for Disease Control and Prevention (CDC) published a guideline to address morbidity and mortality associated with opioid prescribing in the United States (US). 1 This guideline highlighted opioid doses and the co‐prescribing of medications such as hypnotics, benzodiazepines, and skeletal muscle relaxants (SMRs) associated with adverse events and increased healthcare utilization. 1 Additionally, the Food and Drug Administration (FDA) published a communication recommending to avoid prescribing of central nervous system (CNS) active medications among patients utilizing opioids due to increased risks for respiratory depression, overdose, and death. 1 , 2
Subsequently, the Centers for Medicare and Medicaid Services (CMS) 3 and the Department of Health and Human Services (HHS) 4 recommended naloxone co‐prescribing for patients at greater risk of an opioid overdose. These included individuals diagnosed with substance use disorder, utilizing ≥50 morphine milligram equivalents (MMEs) per day, having concurrent prescribing of benzodiazepines, or having already experienced an opioid overdose. 3 , 4
One means of improving the uptake of naloxone among patients at risk for opioid overdose is enrollment in Medication Therapy Management (MTM) services. These services are utilized to comprehensively assess medication regimens to identify medication‐related problems (MRPs). 5 These services provide recommendations that improve the uptake of evidence‐based medications and indicated medication doses and reduce the prevalence of supra‐therapeutic medication doses, drug–drug or drug–disease interactions, therapeutic duplications, and medication nonadherence. One study found an MTM intervention in a community pharmacy increased access to naloxone and reduced misuse of opioids. 6
A previous study found, older adults with chronic opioid prescribing were more likely to be prescribed certain CNS active medications when compared to patients without chronic opioid prescribing. 7 However, it remains unclear if patients with chronic opioid prescribing at higher doses are more likely to utilize CNS active medications despite their already increased risk for opioid overdose. It is also unknown whether patients prescribed higher doses of opioids have greater access to naloxone relative to those prescribed lower opioid doses. The objectives of this study were to (1) determine the prescribing trends of CNS active medications and naloxone among patients receiving chronic opioid therapy; and (2) determine relationships of patient characteristics with medication prescribing in a Medicare‐enrolled MTM eligible cohort.
METHODS
This was a retrospective cross‐sectional analysis of a cohort of MTM‐eligible patients from one Medicare Part D plan provider for the year 2019. Patients were excluded if no history of medication claims was available. Patients were included if they had ≥90 days of opioid medications in the 180 days prior to MTM qualification. Mean daily MME dose was calculated for each patient utilizing a fixed observation window of 180 days. 8 Patients were stratified into two cohorts, individuals with a mean daily MME ≥50 and individuals with a mean daily MME <50. A mean MME threshold of ≥50 was chosen because of the increased risk of opioid overdose at this threshold. 1 The Ohio State University Institutional Review Board approved this study (Study protocol: 2020H0103, 05/12/2020). Waiver of consent was obtained for human subjects research. Permission for research was obtained from the insurance plan provider.
Data source
Data provided by the patient's insurance plan were obtained from a national MTM provider. Information provided to investigators included demographic information (age, sex, and zip code), prescription claims (medication names, doses, routes of administration, days' supply), diagnostic claims, and the number of unique medications and dispensing pharmacies used by the patient in the four months prior to MTM qualification. Race and ethnicity data was not provided to the MTM provider and as such could not be included in this analysis. The unique national provider identifier (NPI) of prescribers for medication claims was used to calculate the number of unique providers, and the NPI of the prescriber of each opioid claim was used to calculate the number of providers that prescribed opioids during the 180‐day look‐back period. To approximate socioeconomic status, data published by the Census Bureau describing percentages of households living below the federal poverty level (FPL) was cross‐referenced with patient zip codes. 9
Study population
This study evaluated patients who were enrolled in Medicare and eligible for MTM. Patients are eligible for MTM if they (1) were diagnosed with certain chronic conditions, (2) have prescription claims for a specific number of unique chronic medications, and (3) have incurred annual medication costs exceeding a predetermined threshold. 10 Enrollment criteria were determined by the insurance plan provider. In this study, patients were limited to one plan provider which offered insurance plans for patients subscribing to Medicare Advantage or Medicare Part D plans. Patients in this cohort were with ≥8 medications to treat chronic conditions, had at least 3 of the following conditions, diabetes, dyslipidemia, heart failure, hypertension, and osteoarthritis, and an estimated annual drug spend of at least $4000.
Medications assessed
Prescription claims data were collected for dates January 1 to December 31, 2019 and were limited to 180 days prior to MTM qualification. Naloxone claims were collected for dates January 1, 2018‐ December 31, 2019. Medications assessed included opioids, naloxone and CNS active medications.
The CNS active medications evaluated were benzodiazepines, SMRs, hypnotics, gabapentanoids, antidepressants including selective serotonin reuptake inhibitors (SSRIs), serotonin–norepinephrine reuptake inhibitors (SNRIs), tricyclic antidepressants (TCAs), maprotiline, bupropion, mirtazapine, serotonin reuptake inhibitor/antagonists, monoamine oxidase inhibitors, and vilazodone, first and second‐generation antipsychotics, hydroxyzine, barbiturates, and other antiepileptics. The full list can be found in Table S1.
Since the publication of reports by the CDC 1 and FDA, 2 research has linked prescribing of benzodiazepines, 11 SMRs, 12 hypnotics, 11 , 13 gabapentanoids, 14 , 15 and antipsychotics 16 with an increased risk of opioid overdose and mortality. For these reasons, these medications were included. Antidepressants were included as these were an important predictor of opioid‐induced respiratory depression among populations with chronic opioid use. 17 , 18 , 19 Hydroxyzine was included as it may be used as opioid‐sparing but has the potential to cause CNS depression. 20 Barbiturates and other antiepileptics were included given their classification as CNS active medications. 21
Criteria for naloxone
Based on recommendations by CMS 3 and HHS, 4 patients in this study were considered eligible for naloxone: having a mean daily MME ≥50 or being prescribed chronic opioid therapy with a diagnosis code for substance use disorder, a history of adverse events or accidental/intentional poisoning with an opioid, or with concurrent benzodiazepine prescribing. Diagnosis codes included are found in Table S2. 22 , 23
Statistical analyses
Data were coded and organized using Microsoft Excel (2016 MSO, Redmond, WA) and IBM SPSS software (v28.0, IBM Corp, Armonk, NY). The following data were transformed into ordinal variables, age (<65, 65–74, 75–84, ≥85 years), number of prescribers (1–3, 4–6, 7–9, 10–12, ≥13) number of prescribers for opioids (1, 2, 3, ≥4), number of unique medications (8–10, 11–13, 14–16, ≥17), number of unique dispensing pharmacies (1, 2, 3, ≥4), percentage of households below the FPL (0.00–9.99, 10.00–19.99, 20.00–29.99, 30.00–100.00), and number of CNS active medications (0, 1, 2, 3, 4, >5 and <4, ≥4). Descriptive statistics were calculated as appropriate. A multivariate logistic regression determined the relationship of patient characteristics with prescribing of mean daily MME ≥50. Variables included were age, sex, number of prescribers, number of prescribers for opioids, number of unique medications, percentage of households below the FPL, and number of dispensing pharmacies.
Chi‐square and odds ratios with 95% confidence intervals (CI) assessed if patients with a mean daily MME ≥50 were more likely to be dispensed the targeted medications when compared to individuals with a mean daily MME <50. This test was repeated for exposure to ≥4 CNS active medication classes. Figure 1 illustrated the differences in prescribing targeted medications between the two cohorts. Logistic regressions evaluated the medications associated with MME dose (<50 MME, MME ≥50) after controlling for age and sex.
FIGURE 1.
Percent of patients prescribed naloxone and central nervous system active (CNS) medications. Cohorts are grouped based on exposure to a mean opioid dose above or below 50 morphine milligram equivalents (MME). Green bars are patients with a mean daily opioid dose <50 MME and the prescribed medication. Blue bars are patients with a mean daily opioid dose ≥50 MME and the prescribed medication. aBonferroni corrected statistically significant difference after adjustment for age and sex (p < 0.004).
The count and percentage of patients for whom naloxone would have been appropriate to dispense were calculated. This analysis was performed among patients with and without diagnosis codes because diagnosis codes were not available for all patients. Chi‐square and odds ratios were used to assess the association between mean daily MME ≥50 and dispensing of naloxone. For the evaluation of associations of patient characteristics with being prescribed a mean daily MME ≥50, a Bonferroni corrected p‐value of 0.007 was set to establish significance. For associations between medication prescribing and being prescribed a mean daily MME ≥50, a Bonferroni corrected p‐value of 0.004 determined significance.
A post‐hoc analysis was performed to determine which provider specialties had prescribed naloxone in this cohort. Figure 2 illustrated the number of naloxone fills by specialty. Table S3 describes the number of naloxone fills by specialty utilizing a provider's Medicare provider description or taxonomy description. Provider NPI was utilized to record their primary taxonomy as listed by CMS. 24 The description of the specialty was then derived from the Medicare provider description found within a provider taxonomy crosswalk. 25 When descriptions were not available, specialty definitions were derived from a database managed by the National Uniform Claim Committee. 26
FIGURE 2.
Count of providers who prescribed naloxone by specialty
RESULTS
From an initial sample of 55,951, a total of 4568 patients were included, of which 3452 (76%) had a daily MME <50 and 1116 (24%) had a daily MME ≥50. The majority of patients were ≥65 years of age (75%), predominantly female (66%), and had ≥4 prescribers (74%) with 1 prescribing opioids (54%). Patients were also mostly with ≥11 unique medications (81%), used 2 pharmacies (58%), and predominantly (80%) resided in regions where more than 10% of households lived below the federal poverty level. Complete demographic information appears in Table 1.
TABLE 1.
Results from logistic regression assessing relationship between patient characteristics and prescribing of mean daily morphine milligram equivalent (MME) ≥50
| Characteristics | Total population n = 4568 | <50 MME n = 3452 | ≥50 MME n = 1116 | Adjusted odds ratio (95% confidence interval) | p‐value |
|---|---|---|---|---|---|
| N (%) | N (%) | ||||
| Age, years | <0.001 | ||||
| <65 | 1159 (25) | 727 (21) | 432 (39) | 4.04 (2.99–5.46) | <0.001 |
| 65–74 | 1641 (36) | 1229 (36) | 412 (37) | 2.35 (1.75–3.15) | <0.001 |
| 75–84 | 1179 (26) | 970 (28) | 209 (19) | 1.58 (1.16–2.15) | 0.004 |
| ≥85 (reference) | 589 (13) | 526 (15) | 63 (6) | ‐ | ‐ |
| Sex | 0.004 | ||||
| Female (reference) | 3009 (66) | 2332 (68) | 677 (61) | ‐ | ‐ |
| Male | 1559 (34) | 1120 (32) | 439 (39) | 1.24 (1.07–1.43) | 0.004 |
| Number of unique prescribers | 0.02 | ||||
| 1–3 (reference) | 1130 (25) | 937 (27) | 193 (17) | ‐ | ‐ |
| 4–6 | 1340 (29) | 998 (29) | 342 (31) | 1.40 (1.13–1.73) | 0.002 |
| 7–9 | 988 (22) | 718 (21) | 270 (24) | 1.37 (1.09–1.72) | 0.007 |
| 10–12 | 585 (13) | 422 (12) | 163 (15) | 1.22 (0.94–0.60) | 0.14 |
| ≥13 | 487 (11) | 349 (10) | 138 (12) | 1.15 (0.86–1.53) | 0.36 |
| Number of unique prescribers of opioid medications | 0.09 | ||||
| 1 (reference) | 2470 (54) | 1915 (56) | 555 (50) | ‐ | ‐ |
| 2 | 1330 (29) | 996 (29) | 334 (30) | 1.09 (0.93–1.28) | 0.19 |
| 3 | 523 (11) | 369 (11) | 154 (14) | 1.29 (1.03–1.61) | 0.01 |
| ≥4 | 245 (5) | 172 (5) | 73 (7) | 1.29 (0.95–1.76) | 0.11 |
| Number of unique medications | <0.001 | ||||
| 8–10 (reference) | 850 (19) | 688 (20) | 162 (15) | ‐ | ‐ |
| 11–13 | 1203 (26) | 934 (27) | 269 (24) | 1.18 (0.94–1.48) | 0.14 |
| 14–16 | 1080 (24) | 804 (23) | 276 (25) | 1.35 (1.08–1.69) | 0.01 |
| ≥17 | 1435 (31) | 1026 (30) | 409 (37) | 1.55 (1.25–1.93) | <0.001 |
| % of households below the federal poverty level in the patient's zip code | 0.07 | ||||
| 0.00–9.99 (reference) | 900 (20) | 674 (20) | 226 (20) | ‐ | ‐ |
| 10.00–19.99 | 1602 (35) | 1237 (36) | 365 (33) | 0.81 (0.66–0.99) | 0.04 |
| 20.00–29.99 | 1144 (25) | 884 (26) | 260 (23) | 0.77 (0.62–0.96) | 0.02 |
| 30.00–100.00 | 902 (20) | 642 (19) | 260 (23) | 0.90 (0.72–1.12) | 0.36 |
| Number of unique dispensing pharmacies | 0.08 | ||||
| 1 (reference) | 1907 (42) | 1528 (44) | 379 (34) | ‐ | ‐ |
| 2 | 1217 (27) | 898 (26) | 319 (29) | 1.22 (1.02–1.46) | 0.03 |
| 3 | 690 (15) | 489 (14) | 201 (18) | 1.28 (1.04–1.59) | 0.02 |
| ≥4 | 716 (16) | 509 (15) | 207 (19) | 1.17 (0.94–1.46) | 0.17 |
Note: Bonferroni corrected p‐value of 0.007. Percentages may not sum to 100% due to rounding.
A multivariate logistic regression found age (p < 0.001), sex (p = 0.003), and number of unique medications (p < 0.001) associated with having a daily MME ≥50. Patients <85 years of age compared to individuals ≥85 had 1.57–4.19 times the odds of being prescribed a daily MME ≥50. Males were more likely than females to be prescribed a daily MME ≥50, (OR: 1.25, 95% CI 1.08–1.45, p = 0.003). While the number of medications was significantly associated with being prescribed a daily MME ≥50, only patients with ≥17 unique medications compared to patients with 8–10 medications had a significant association with having a daily MME ≥50 (OR: 1.55, 95% CI 1.25–1.93, p < 0.001). Complete results from the regression are presented in Table 1.
Among patients identified with chronic opioid use, there were 44% patients prescribed oxycodone, 31% prescribed tramadol, 28% prescribed hydrocodone, 7% prescribed codeine, 7% prescribed morphine, and 6% prescribed fentanyl. Regarding CNS active medications, 27% had co‐prescribed benzodiazepines, 20% had co‐prescribed SMRs, 14% had co‐prescribed hypnotics, 54% had co‐prescribed gabapentanoids, 61% had co‐prescribed any antidepressant, 5% had co‐prescribed barbiturates, 10% had co‐prescribed other antiepileptics, 5% had co‐prescribed hydroxyzine, 2% had co‐prescribed first‐generation antipsychotics, and 14% had co‐prescribed second‐generation antipsychotics.
Patients prescribed a mean daily MME ≥50 were more likely to be co‐prescribed SMRs (OR: 1.43, 95% CI 1.22–1.68, p < 0.001), hypnotics (OR: 1.74, 95% CI 1.45–2.09, p < 0.001), SNRIs (OR 1.35, 95% CI 1.14–1.59, p < 0.001) TCAs (OR 1.61, 95% CI 1.25–2.07, p < 0.001), and gabapentanoids (OR 1.22, 95% CI 1.07–1.40, p = 0.004). After adjusting for age and sex only hypnotics (adjusted OR: 1.41, 95% CI 1.17–1.70, p < 0.001) and SNRIs (adjusted OR: 1.39, 95% CI 1.17–1.64, p < 0.001) remained significantly associated with having a mean daily MME ≥50. Complete results are presented in Table 2 and illustrated in Figure 1.
TABLE 2.
Results from Chi‐square and odds ratio (OR) analysis assessing relationship between prescribing of naloxone and central nervous system (CNS) active medications and a mean daily morphine milligram equivalent (MME) ≥50
| Medication | Total N = 4568 n (%) | <50 MME N = 3452 n (%) | ≥50 MME N = 1116 n (%) | OR (95% CI) | aOR (95% CI) a |
|---|---|---|---|---|---|
| Naloxone | 279 (6) | 134 (4) | 145 (13) | 3.70 (2.89–4.73) b | 3.21 (2.49–4.12) b |
| CNS active medications | |||||
| Benzodiazepine | 1231 (27) | 896 (26) | 335 (30) | 1.22 (1.05–1.42) b | 1.19 (1.02–1.39) |
| Skeletal muscle relaxant | 928 (20) | 650 (19) | 278 (25) | 1.43 (1.22–1.68) b | 1.07 (0.90–1.26) |
| Hypnotic | 628 (14) | 414 (12) | 214 (19) | 1.74 (1.45–2.09) b | 1.41 (1.17–1.70) b |
| Gabapentanoid | 2445 (54) | 1806 (52) | 639 (57) | 1.22 (1.07–1.40) b | 1.11 (0.96–1.28) |
| Any antidepressant | 2800 (61) | 2110 (61) | 690 (62) | 1.03 (0.90–1.18) | 1.08 (0.93–1.24) |
| SSRI | 1481 (32) | 1150 (33) | 331 (30) | 0.84 (0.73–0.98) | 0.88 (0.76–1.02) |
| SNRI | 855 (19) | 606 (18) | 249 (22) | 1.35 (1.14–1.59) b | 1.39 (1.17–1.64) b |
| TCA | 302 (7) | 201 (6) | 101 (9) | 1.61 (1.25–2.07) b | 1.41 (1.09–1.82) |
| First generation antipsychotic | 81 (2) | 55 (2) | 26 (2) | 1.47 (0.92–2.36) | 1.35 (0.83–2.18) |
| Second generation antipsychotic | 621 (14) | 469 (14) | 152 (14) | 1.00 (0.82–1.22) | 0.83 (0.68–1.02) |
| Hydroxyzine | 207 (5) | 148 (4) | 59 (5) | 1.25 (0.91–1.70) | 0.96 (0.72–1.36) |
| Barbiturates | 74 (2) | 59 (2) | 15 (1) | 0.78 (0.44–1.39) | 0.85 (0.47–1.51) |
| Other antiepileptics | 470 (10) | 353 (10) | 117 (11) | 1.03 (0.82–1.28) | 0.93 (0.74–1.17) |
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; SSRI, Selective serotonin reuptake inhibitor; SSRI, Serotonin norepinephrine reuptake inhibitor; TCA, Tricyclic antidepressant.
Adjusted for age and sex.
Statistically significant based on a Bonferroni corrected p‐value of 0.004.
In the overall cohort, 23% filled 1 CNS active medication, 24% filled 2 CNS active medications, 18% filled 3 CNS active medications, 12% filled 4 CNS active medications, 6% filled 5 CNS active medications and 4% filled ≥6 CNS active medications. Table 3 describes differences in the proportion of patients with CNS active medications by MME category. Patients with a mean MME ≥50 (306, 27%) were more likely than patients with an MME <50 (734, 21%) to be exposed to ≥4 CNS active medications (OR 1.40, 95% CI: 1.20–1.63, p < 0.001). However, this relationship was rendered nonsignificant after adjusting for age and sex.
TABLE 3.
Description of number of central nervous system active medication classes by morphine milligram equivalent (MME) category.
| Number of CNS active medications | Total n = 4568 | <50 MME n = 3452 | ≥50 MME n = 1116 |
|---|---|---|---|
| 0 | 535 (12) | 422 (12) | 113 (10) |
| 1 | 1063 (23) | 842 (24) | 221 (20) |
| 2 | 1090 (24) | 829 (24) | 261 (23) |
| 3 | 840 (18) | 625 (18) | 215 (19) |
| 4 | 561 (12) | 405 (12) | 156 (14) |
| 5 | 280 (6) | 192 (6) | 88 (8) |
| ≥6 | 199 (4) | 137 (4) | 62 (6) |
In the overall cohort, 2216(49%) were identified as being eligible for a naloxone prescription. A total of 1116 patients were with a daily MME ≥50 and 1231 patients were co‐prescribed benzodiazepines. There were 2165 patients with diagnostic claims. Among patients with diagnostic claims, 53% were eligible for naloxone. Among these patients, 474 were diagnosed with substance use disorder, and 20 diagnosed with an opioid adverse event or poisoning. Among all patients eligible for a naloxone prescription, 9% filled a naloxone prescription. Among the entire studied cohort, patients with a daily MME ≥50 were more likely to have filled naloxone (OR 3.70, 95% CI 2.89–4.73, p < 0.001) compared to individuals with a daily MME <50. This association remained significant after adjusting for age and sex (adjusted OR 3.21, 95% CI 2.49–4.12, p < 0.001).
A total of 219 providers prescribed naloxone filled by 202 patients. Among these providers, the three most predominant specialties included primary care (n = 74), nurse practitioners or nurse specialists (n = 31), and physicians who specialized in pain medicine and anesthesiology (n = 26). There were 17 pharmacists, physical medicine and rehabilitation specialized physicians, and physician assistants who also prescribed naloxone. Complete information is illustrated in Figure 2 and can be found in Table S3.
DISCUSSION
This study found nearly one‐fourth of MTM‐eligible Medicare patients with chronic opioid therapy were prescribed a mean MME dose ≥50. Individuals prescribed high‐dose opioid therapy were more likely to have certain CNS active medications compared to patients with lower chronic opioid doses. Furthermore, this study found nearly 1 in 5 patients were utilizing ≥4 CNS active medications. The data also revealed that patients <85 years of age were more likely to be prescribed a daily MME dose ≥50 when compared to adults ≥85 years. Only 9% of patients eligible for naloxone had been dispensed the drug over a two‐year period. These findings highlight the need for targeted services to identify opportunities to optimize medication prescribing among MTM‐eligible patients with chronic opioid use.
After adjusting for age and sex, hypnotics and SNRIs were more likely to be used among patients with a daily MME dose ≥50. It is also important to highlight that 27% of patients were prescribed benzodiazepines and 54% utilized gabapentanoids. Patients with chronic opioid use who utilize hypnotics, SMRs, gabapentanoids, and benzodiazepines are at greater risk of opioid overdose and mortality. 1 , 2 , 11 , 12 , 13 , 14 , 15 Additionally, approximately 15% of patients receiving opioids were found to be at risk of experiencing a drug–drug interaction resulting in increased dose requirements for adequate pain control. 27
Greater use of SNRIs among patients with chronic opioid use at higher doses was expected given SNRIs are used to treat chronic pain, neuropathic pain, and central pain sensitization. 1 , 28 However, psychotropic drugs and psychiatric diagnoses are individual risk factors for opioid misuse and opioid overdose. 1 , 17 , 18 , 19 Our study found that 62%, 2%, and 14% of the patients utilized any antidepressant, first‐generation antipsychotics, and second‐generation antipsychotics, respectively. No differences in the prescribing of antidepressants and antipsychotics between the two MME cohorts were found. These findings suggested the overall prescribing of mental health medications among patients with chronic opioid use did not differentiate between the analyzed cohorts. Importantly, these findings should not be used to deprescribe antidepressant medications, as a previous study found patients diagnosed with depression and prescribed antidepressants were less likely to experience an opioid overdose. 29
Naloxone was substantially underprescribed, and its filling was significantly higher among patients with a mean daily MME ≥50. Only 9% of patients who, according to guidance provided by CMS 2 and HHS 3 may benefit from a naloxone prescription, were with naloxone. However, due to the limitations of utilizing insurance claims, it is unclear if patients were provided access to naloxone through cash payment at pharmacies or from volunteer organizations. Notwithstanding, our findings are consistent with another study that utilized dispensing data, including cash payment methods, which found suboptimal rates of naloxone co‐prescribing among patients. 30 To address this problem, previous research highlighted important interventions in community, academic, pharmacy, and hospital settings 31 and through legislative mandates. 32 These findings may signal a greater need for awareness and specialized training among providers and pharmacists, de‐stigmatization of carrying naloxone, and a need for further reductions in copays for patients. 33
Given the recent decrease in naloxone co‐prescribing among Medicare and Medicaid patients during the COVID‐19 pandemic there is an urgent need for pharmacovigilance programs. 34 Insurance monitoring programs with complete prescribing data can be utilized to identify patients who would benefit from naloxone prescribing and alert prescribers to a patient's need. In addition to the implementation of targeted services, it is important for insurance plans to ensure patients can easily afford naloxone. Presently, 99% of Medicare plans are estimated to cover naloxone prescriptions with an estimated copay of 0–19$. 35 While encouraging, a 0$ copay would ensure patients who receive prescriptions for naloxone do not encounter financial barriers.
In this population, patients using higher doses of opioid medications were more likely to be <85 years of age, male, and have ≥17 concurrent medications. This is consistent with previous research that found younger patients, males, and individuals with greater disease burden were more likely to utilize high doses of opioids chronically. 36 It is important to highlight, patients <65 years of age were 4 times more likely to be prescribed higher doses of opioids compared to individuals ≥85 years of age. Medicare patients <65 years of age are likely to have qualified for Medicare due to disability 37 and have 10 times the risk of experiencing an opioid overdose compared to the national mean. 38
While the number of prescribers was not a significant predictor of having a mean MME ≥50 it is important to highlight patients were with 7 ± 4 unique prescribers. Patients within the US may receive prescriptions from various healthcare systems with little integration. A previous survey noted less than one‐half of primary care physicians (PCPs) were informed by specialists of changes to a patient's therapeutic regimen. 39 The risk of drug interactions is also increased when multiple providers are prescribing for the same patient, suggesting a breakdown in communication between providers regarding medications. 40 Furthermore, this study found almost 1 in 4 patients were utilizing ≥4 CNS active medications. Future research is needed to assess the effectiveness of interventions identifying deprescribing opportunities or the need for naloxone among patients on long term opioid therapy.
The findings of our study suggested that patients on chronic opioid therapy may greatly benefit from MTM services to identify and alert providers about potential MRPs. Communications targeting naloxone prescribing may yield the most benefit for patients if sent to their PCP given they prescribed the most naloxone in this cohort. Furthermore, they can be used to increase engagement, education and implementation of multimodal therapies to reduce supply side access to high doses of opioids. 1 , 40 Given the harm some insurance programs have caused through misapplication of prescribing guidelines, 41 MTM may be a more suitable means of engaging providers and patients to assess for deprescribing opportunities and reduce potentially avoidable healthcare utilization.
Limitations
This study was limited to one Medicare plan provider, which may limit the generalizability of results. Investigators were unable to obtain prescribing rationale. Prescriptions not purchased through the prescription benefit, for example, cash payment or provided to patients without a prescription (such as naloxone), could not be assessed. Medications investigated in this study may not comprehensively represent all medications associated with adverse events. It is important to note this study did not control for confounding chronic health conditions. The intent of this analysis was to determine if exposure to specific CNS active medications differed between high and low doses of chronic opioids. Patients requiring higher doses of opioids for pain compared to those with lower opioid doses are more likely to be diagnosed with specific mental health conditions and experience greater comorbidity burden. 42 Thus, differences in exposure to CNS active medications may have occurred, in part due to the differences in the prevalence of comorbid conditions for which these medications are indicated.
CONCLUSIONS
Approximately one‐fourth of Medicare MTM‐eligible patients with chronic opioid use were prescribed daily MME ≥50. Patients on higher opioid doses were more likely to utilize specific CNS active medications compared to patients with lower chronic opioid doses. Almost 1 in 4 patients with chronic opioid use utilized ≥4 CNS active medications. Furthermore, this study found potential underprescribing of naloxone over a two‐year period. Patients receiving chronic high‐dose opioid therapy were likely to be <85 years of age, male, and with ≥17 medications. Targeted medication reviews can identify patients at risk for adverse events and increase opportunities to improve access to naloxone among patients prescribed chronic opioid therapy.
AUTHOR CONTRIBUTIONS
Armando Silva Almodóvar and Milap C Nahata were involved conceptualization and design of the study and writing and revision of the manuscript. Armando Silva Almodóvar conducted the analysis and prepared the results. All authors reviewed the manuscript and approved it.
FUNDING INFORMATION
Dr. Armando Silva‐Almodóvar's is supported in part by the National Institute on Aging (R24AG064025). Dr. Milap Nahata was supported in part by the Avatar Foundation.
CONFLICT OF INTEREST
Dr. Armando Silva Almodovar is a research pharmacist at the Medication Management Program. Dr. Milap Nahata has no conflicts of interest to report.
SPONSOR'S ROLE
The National Institute on Aging or the Avatar Foundation had no role in the design or execution of this study, nor in how the data was collected, managed, analyzed or interpreted. Sponsors had no role in the preparation, writing, editing, or review of the manuscript; or in the decision to submit for publication.
Supporting information
Table S1. List of medications assessed in this study by medication class.
Table S2. List of diagnosis codes used to determine the potential need for naloxone prescription.
Table S3. Count of providers who prescribed naloxone by specialty and primary taxonomy.
ACKNOWLEDGMENTS
We appreciate the contributions of Jera Copley PharmD candidate and Nicole Glatz PharmD candidate in the organization of the research datasets.
Silva Almodóvar A, Nahata MC. High opioid doses, naloxone, and central nervous system active medications received by Medicare‐enrolled adults. J Am Geriatr Soc. 2023;71(1):98‐108. doi: 10.1111/jgs.18102
The results of this manuscript were presented in part at the Academy of Managed Care Pharmacy Annual Meeting between March 29–April 1, 2022, in Chicago Illinois.
Funding information Avatar Foundation; National Institute on Aging, Grant/Award Number: R24AG064025
REFERENCES
- 1. Dowell D, Haegerich TM, Chou R. CDC guideline for prescribing opioids for chronic pain – United States, 2016 [published correction appears in MMWR Recomm rep. 2016;65(11):295]. MMWR Recomm Rep. 2016;65(1):1‐49. doi: 10.15585/mmwr.Rr6501e1 [DOI] [PubMed] [Google Scholar]
- 2. United States Food and Drug Administration . FDA drug safety communication: FDA warns about serious risks and death when combining opioid pain or cough medicines with benzodiazepines; requires its strongest warning. United States Food and Drug Administration Web site. Updated 2016. Accessed July 13, 2022. https://www.fda.gov/drugs/drug-safety-and-availability/fda-drug-safety-communication-fda-warns-about-serious-risks-and-death-when-combining-opioid-pain-or
- 3. Announcement of calendar year (CY) 2020 Medicare advantage capitation rates and Medicare advantage and part D payment policies and final call letter. The Centers for Medicare and Medicaid Services. 2019.
- 4. Giroir B. HHS recommends prescribing or co‐prescribing naloxone to patients at high risk for an opioid overdose. HHS.gov U.S. Department of Health & Human Services Web site. Updated 2018. Accessed July 13, 2022. https://public3.pagefreezer.com/browse/HHS.gov/31-12-2020T08:51/https://www.hhs.gov/about/news/2018/12/19/hhs-recommends-prescribing-or-co-prescribing-naloxone-to-patients-at-high-risk-for-an-opioid-overdose.html
- 5. Viswanathan M, Kahwati LC, Golin CE, et al. Medication Therapy Management Interventions in Outpatient Settings [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US). 2014. (Comparative Effectiveness Reviews, No. 138.) Introduction. Available from: https://www.ncbi.nlm.nih.gov/books/NBK294478/ [PubMed]
- 6. Cochran G, Chen Q, Field C, et al. A community pharmacy‐led intervention for opioid medication misuse: a small‐scale randomized clinical trial. Drug Alcohol Depend. 2019;205:107570. doi: 10.1016/j.drugalcdep.2019.107570 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Silva Almodovar A, Nahata MC. Potentially unsafe chronic medication use among older adult chronic opioid users. Pharmacotherapy. 2019;39(2):140‐149. doi: 10.1002/phar.2218 [DOI] [PubMed] [Google Scholar]
- 8. Opioid oral morphine milligram equivalent (MME) conversion factors table for prescription drug coverage. US Department of Health and Human Services Web site. Updated 2020. Accessed August 17, 2021. https://www.hhs.gov/guidance/document/opioid-oral-morphine-milligram-equivalent-mme-conversion-factors-0
- 9. Explore Census Data United States Census Bureau Web site. Accessed August 17, 2021.
- 10. Chavez‐Valdez AL. Contract year 2022 part D medication therapy management program guidance and submission instructions. Centers for Medicare and Medicaid Services. 2021.
- 11. Cho J, Spence MM, Niu F, Hui RL, Gray P, Steinberg S. Risk of overdose with exposure to prescription opioids, benzodiazepines, and non‐benzodiazepine sedative‐hypnotics in adults: a retrospective cohort study. J Gen Intern Med. 2020;35(3):696‐703. doi: 10.1007/s11606-019-05545-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Li Y, Delcher C, Wei YJ, et al. Risk of opioid overdose associated with concomitant use of opioids and skeletal muscle relaxants: a population‐based cohort study. Clin Pharmacol Ther. 2020;108(1):81‐89. doi: 10.1002/cpt.1807 [DOI] [PubMed] [Google Scholar]
- 13. Ray WA, Chung CP, Murray KT, Malow BA, Daugherty JR, Stein CM. Mortality and concurrent use of opioids and hypnotics in older patients: a retrospective cohort study. PLoS Med. 2021;18(7):e1003709. doi: 10.1371/journal.pmed.1003709 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Gomes T, Juurlink DN, Antoniou T, Mamdani MM, Paterson JM, van den Brink W. Gabapentin, opioids, and the risk of opioid‐related death: a population‐based nested case‐control study. PLoS Med. 2017;14(10):e1002396. doi: 10.1371/journal.pmed.1002396 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Gomes T, Greaves S, van den Brink W, et al. Pregabalin and the risk for opioid‐related death: a nested case‐control study. Ann Intern Med. 2018;169(10):732‐734. doi: 10.7326/M18-1136 [DOI] [PubMed] [Google Scholar]
- 16. Szmulewicz AG, Bateman BT, Levin R, Huybrechts KF. Risk of overdose associated with Co‐prescription of antipsychotics and opioids: a population‐based cohort study. Schizophr Bull. 2022;48(2):405‐413. doi: 10.1093/schbul/sbab116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Ward R, Weeda E, Taber DJ, Axon RN, Gebregziabher M. Advanced models for improved prediction of opioid‐related overdose and suicide events among veterans using administrative healthcare data. Health Serv Outcome Res Methodol. 2022;22(2):275‐295. doi: 10.1007/s10742-021-00263-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Nadpara PA, Joyce AR, Murrelle EL, et al. Risk factors for serious prescription opioid‐induced respiratory depression or overdose: comparison of commercially insured and veterans health affairs populations. Pain Med. 2018;19(1):79‐96. doi: 10.1093/pm/pnx038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Webster LR. Risk factors for opioid‐use disorder and overdose. Anesth Analg. 2017;125(5):1741‐1748. doi: 10.1213/ANE.0000000000002496 [DOI] [PubMed] [Google Scholar]
- 20. Glazier HS. Potentiation of pain relief with hydroxyzine: a therapeutic myth? DICP. 1990;24(5):484‐488. doi: 10.1177/106002809002400509 [DOI] [PubMed] [Google Scholar]
- 21. Maust DT, Strominger J, Kim HM, et al. Prevalence of central nervous system‐active polypharmacy among older adults with dementia in the US. JAMA. 2021;325(10):952‐961. doi: 10.1001/jama.2021.1195 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Practice Research and Policy staff . Substance use disorders and ICD‐10‐CM coding. American Psychological Association Web Site. Updated 2015. Accessed August 17, 2021.
- 23. Weiss AJ, McDermott KW, Heslin KC. Opioid‐Related Hospital Stays Among Women in the United States, 2016: Statistical Brief #247. 2019 Jan 22. In: Healthcare Cost and Utilization Project (HCUP) Statistical Briefs [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US). 2006. Table 1, ICD‐10‐CM diagnosis codes defining different opioid‐related conditions. Available from: https://www.ncbi.nlm.nih.gov/books/NBK538344/table/sb247.tab1/ [PubMed]
- 24. Centers for Medicare and Medicaid Services. Data dissemination. Centers for Medicare and Medicaid Services Web site. Updated 2021. Accessed Feb 2, 2022. https://www.cms.gov/Regulations-and-Guidance/Administrative-Simplification/NationalProvIdentStand/DataDissemination
- 25. Centers for Medicare and Medicaid Services. Medicare provider and supplier taxonomy crosswalk. Centers for Medicare and Medicaid Services Web site. Updated 2021. Accessed Feb 8, 2022. https://data.cms.gov/provider-characteristics/medicare-provider-supplier-enrollment/medicare-provider-and-supplier-taxonomy-crosswalk
- 26. National Uniform Claim Committee . Health care provider taxonomy code set. National Uniform Claim Committee Web site. Updated 2022. Accessed Feb 8, 2022. https://taxonomy.nucc.org/
- 27. Michaud V, Bikmetov R, Smith MK, et al. Use of drug claims data and a medication risk score to assess the impact of CYP2D6 drug interactions among opioid users on healthcare costs. J Pers Med. 2021;11(11):1174. doi: 10.3390/jpm11111174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Finnerup NB, Attal N, Haroutounian S, et al. Pharmacotherapy for neuropathic pain in adults: a systematic review and meta‐analysis. Lancet Neurol. 2015;14(2):162‐173. doi: 10.1016/S1474-4422(14)70251-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Sherry TB, Nguyen T, Dick AW, Bradford AC, Simon K, Stein BD. A national retrospective study of antidepressants' effects on overdose and self‐harm among adults treated with opioid analgesics [published online ahead of print, 2022 Jun 30]. Psychiatr Serv. 2022;appi.ps.20220070. doi: 10.1176/appi.ps.20220070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Stein BD, Smart R, Jones CM, Sheng F, Powell D, Sorbero M. Individual and community factors associated with naloxone Co‐prescribing among long‐term opioid patients: a retrospective analysis. J Gen Intern Med. 2021;36(10):2952‐2957. doi: 10.1007/s11606-020-06577-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Dunne RB. Prescribing naloxone for opioid overdose intervention. Pain Manage. 2018;8(3):197‐208. doi: 10.2217/pmt-2017-0065 [DOI] [PubMed] [Google Scholar]
- 32. Green TC, Davis C, Xuan Z, Walley AY, Bratberg J. Laws mandating Coprescription of naloxone and their impact on naloxone prescription in five US states, 2014‐2018. Am J Public Health. 2020;110(6):881‐887. doi: 10.2105/AJPH.2020.305620 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Guy GP Jr, Khushalani JS, Jackson H, Sims RSC, Arifkhanova A. Trends in state‐level pharmacy‐based naloxone dispensing rates, 2012‐2019. Am J Prev Med. 2021;61(6):e289‐e295. doi: 10.1016/j.amepre.2021.05.042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. O'Donoghue AL, Biswas N, Dechen T, et al. Trends in filled naloxone prescriptions before and during the COVID‐19 pandemic in the United States. JAMA Health Forum. 2021;2(5):e210393. Accessed August 30, 2021. doi: 10.1001/jamahealthforum.2021.0393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. GoodRx. Naloxone. GoodRx Web site. Accessed February 25, 2022. https://www.goodrx.com/naloxone/medicare‐coverage?dosage=2ml‐of‐2mg‐2ml&form=syringe&label_override=naloxone&quantity=2&sort_type=popularity
- 36. Musich S, Wang SS, Slindee L, Kraemer S, Yeh CS. Prevalence and characteristics associated with high dose opioid users among older adults. Geriatr Nurs. 2019;40(1):31‐36. doi: 10.1016/j.gerinurse.2018.06.001 [DOI] [PubMed] [Google Scholar]
- 37. Who is eligible for medicare? U.S. Department of Health & Human Services Web site. Updated 2014. Accessed August 17, 2021.
- 38. Meara E, Horwitz JR, Powell W, et al. State legal restrictions and prescription‐opioid use among disabled adults. N Engl J Med. 2016;375(1):44‐53. doi: 10.1056/NEJMsa1514387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Doty MM, Tikkanen R, Shah A, Schneider EC. Primary care Physicians' role in coordinating medical and health‐related social needs in eleven countries. Health Aff (Millwood). 2020;39(1):115‐123. doi: 10.1377/hlthaff.2019.01088 [DOI] [PubMed] [Google Scholar]
- 40. Ong MS, Olson KL, Chadwick L, Liu C, Mandl KD. The impact of provider networks on the Co‐prescriptions of interacting drugs: a claims‐based analysis. Drug Saf. 2017;40(3):263‐272. doi: 10.1007/s40264-016-0490-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Dowell D, Haegerich T, Chou R. No shortcuts to safer opioid prescribing. N Engl J Med. 2019;380(24):2285‐2287. doi: 10.1056/NEJMp1904190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Kobus AM, Smith DH, Morasco BJ, et al. Correlates of higher‐dose opioid medication use for low back pain in primary care. J Pain. 2012;13(11):1131‐1138. doi: 10.1016/j.jpain.2012.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. List of medications assessed in this study by medication class.
Table S2. List of diagnosis codes used to determine the potential need for naloxone prescription.
Table S3. Count of providers who prescribed naloxone by specialty and primary taxonomy.