Prolonged opioid use among opioid-naive individuals after prescription for nonspecific low back pain in the emergency department

Long-term opioid therapy in older adults: Incidence and risk factors related to patient characteristics and initial opioid dispensed

BACKGROUND

The clinical benefits of long-term opioid therapy (LTOT) are not clearly known; however, LTOT is associated with various adverse outcomes.

OBJECTIVE

This study aimed to estimate the incidence and risk factors associated with LTOT in adults aged ≥ 65 years.

METHODS

This was a retrospective cohort study using Medicare claims data. Opioid-naïve older adults filling an opioid prescription between 2014 and 2016 were included. The outcome variable was LTOT, defined as an opioid use episode longer than 90 days and having > 60 cumulative days of supply. Predictor variables included patient characteristics, characteristics of initial opioid dispensed, and pain conditions. Multivariable logistic regression was performed to assess the association between predictors and LTOT.

RESULTS

Among 162,287 participants, 10,296 (6.3%) met the definition of LTOT. Key patient characteristics associated with LTOT were age > 85 years (adjusted odds ratio 1.13 [95% confidence interval 1.05-1.21]); > 5 comorbidities (1.55 [1.45-1.65]); and history of drug use disorder (1.53 [1.35-1.74]), alcohol use disorder (1.38 [1.23-1.54]), tobacco use disorder (1.31 [1.23-1.40]), and opioid use disorder (2.00 [1.69-2.37]). Characteristics of initial opioid associated with LTOT were dispensing long-acting opioids (1.72 [1.21-2.44]) and concomitant use of benzodiazepines (1.16 [1.08-1.25]), gabapentinoids (1.57 [1.47-1.67]), and prescription nonsteroidal anti-inflammatory drugs (1.24 [1.17-1.31]). Anxiety disorders were associated with 1.4-1.5 times increased odds of LTOT. Moreover, initial opioid supply of ≥ 30 days led to 11-16 times higher odds of LTOT than days' supply of 1-3 days.

CONCLUSIONS

Factors related to patient characteristics (age, number of comorbidities, substance use disorders, anxiety disorders) and initial opioid dispensation (duration of action, certain concomitant medications, days' supply) are associated with LTOT in older adults. Prescribers should consider these factors when prescribing opioids to senior patients.

Developing a Wearable Sensor-Based Digital Biomarker of Opioid Dependence

BACKGROUND

Repeated opioid exposure leads to a variety of physiologic adaptations that develop at different rates and may foreshadow risk of opioid-use disorder (OUD), including dependence and withdrawal. Digital pharmacovigilance strategies that use noninvasive sensors to identify physiologic adaptations to opioid use represent a novel strategy to facilitate safer opioid prescribing. This study aims to identify wearable sensor-derived features associated with opioid dependence by comparing opioid-naïve individuals to chronic opioid users with acute pain and developing a machine-learning model to distinguish between the 2 groups.

METHODS

Using a longitudinal observational study design, continuous physiologic data were collected on participants with acute pain receiving opioid analgesia. Monitoring continued throughout hospitalization and for up to 7 days posthospital discharge. Opioid administration data were obtained from electronic health record (EHR) and participant self-report. Participants were classified as belonging to 1 of 3 categories based on opioid use history: naïve, occasional, or chronic use. Thirty features were derived from sensor data, and an additional 9 features were derived from participant demographic and treatment characteristics. Physiologic feature behavior immediately postopioid use was compared among naïve and chronic participants, and subsequently features were used to generate machine learning models which were validated using cross-validation and holdout data.

RESULTS

Forty-one participants with a combined total of 169 opioid administrations were ultimately included in the final analysis. Four interpretable decision tree-based machine learning models with 14 sensor-based and 5 clinical features were developed to predict class membership on the level of a given observation (dose) and on the participant level. Ranges for model metrics on the participant level were as follows: accuracy 70% to 90%, sensitivity 67% to 100%, and specificity 67% to 100%.

CONCLUSIONS

Wearable sensor-derived digital biomarkers can be used to predict opioid use status (naïve versus chronic) and the differentiating features may be detecting opioid dependence. Future work should be aimed at further delineating the phenomenon identified in these models (including opioid dependence and/or withdrawal) and at identifying transition states where an individual changes from 1 profile to another with repetitive opioid exposure.

Initial and Long-Term Prescribing of Opioids and Non-steroidal Anti-inflammatory Drugs Following Total Hip and Knee Arthroplasty

Introduction

Limited evidence exists on health system characteristics associated with initial and long-term prescribing of opioids and nonsteroidal anti-inflammatory drugs (NSAIDs) following total hip and knee arthroplasty (THA/TKA), and if these characteristics differ among individuals based on preoperative NSAID exposure. We identified orthopedic surgeon opioid prescribing practices, hospital characteristics, and regional factors associated with initial and long-term prescribing of opioids and NSAIDs among older adults receiving THA/TKA.

Materials and Methods

This observational study included opioid-naïve Medicare beneficiaries aged ≥65 years receiving elective THA/TKA between January 1, 2014 and July 4, 2017. We examined initial (days 1-30 following THA/TKA) and long-term (days 90-180) opioid or NSAID prescribing, stratified by preoperative NSAID exposure. Risk ratios (RRs) for the associations between 10 health system characteristics and case-mix adjusted outcomes were estimated using multivariable Poisson regression models.

Results

The study population included 23,351 NSAID-naïve and 10,127 NSAID-prevalent individuals. Increases in standardized measures of orthopedic surgeon opioid prescribing generally decreased the risk of initial NSAID prescribing but increased the risk of long-term opioid prescribing. For example, among NSAID-naïve individuals, the RRs (95% confidence intervals [CIs]) for initial NSAID prescribing were 0.95 (0.93-0.97) for 1-2 orthopedic surgeon opioid prescriptions per THA/TKA procedure, 0.94 (0.92-0.97) for 3-4 prescriptions per procedure, and 0.91 (0.89-0.93) for 5+ opioid prescriptions per procedure (reference: <1 opioid prescription per procedure), while the RRs (95% CIs) for long-term opioid prescribing were 1.06 (1.04-1.08), 1.08 (1.06-1.11), and 1.13 (1.11-1.16), respectively. Variation in postoperative analgesic prescribing was observed across U.S. regions. For example, among NSAID-naïve individuals, the RR (95% CIs) for initial opioid prescribing were 0.98 (0.96-1.00) for Region 2 (New York), 1.09 (1.07-1.11) for Region 3 (Philadelphia), 1.07 (1.05-1.10) for Region 4 (Atlanta), 1.03 (1.01-1.05) for Region 5 (Chicago), 1.16 (1.13-1.18) for Region 6 (Dallas), 1.10 (1.08-1.12) for Region 7 (Kansas City), 1.09 (1.06-1.12) for Region 8 (Denver), 1.09 (1.07-1.12) for Region 9 (San Francisco), and 1.11 (1.08-1.13) for Region 10 (Seattle) (reference: Region 1 [Boston]). Hospital characteristics were not meaningfully associated with postoperative analgesic prescribing. The relationships between health system characteristics and postoperative analgesic prescribing were similar for NSAID-naïve and NSAID-prevalent participants.

Discussion

Future efforts aiming to improve the use of multimodal analgesia through increased NSAID prescribing and reduced long-term opioid prescribing following THA/TKA could consider targeting orthopedic surgeons with higher standardized opioid prescribing measures.

Conclusions

Orthopedic surgeon opioid prescribing measures and U.S. region were the greatest health system level predictors of initial, and long-term, prescribing of opioids and prescription NSAIDs among older Medicare beneficiaries following THA/TKA. These results can inform future studies that examine why variation in analgesic prescribing exists across geographic regions and levels of orthopedic surgeon opioid prescribing.

Oxycodone initiation in Australia (2014-2018): Sociodemographic factors and preceding health service use

AIMS

Oxycodone is the most commonly prescribed strong opioid in Australia. This study describes health service antecedents and sociodemographic factors associated with oxycodone initiation.

METHODS

Population-based new user cohort study linking medicine dispensings, hospitalizations, emergency department visits, medical services and cancer notifications from New South Wales (NSW) for 2014-2018. New users had no dispensings of any opioid in the preceding year. We analysed health service use in the 5 days preceding initiation and proportion of people on treatment over 1 year and fitted an area-based, multivariable initiation model with sociodemographic covariates.

RESULTS

Oxycodone accounted for 30% of opioid initiations. Annually, 3% of the NSW population initiated oxycodone, and 5-6% were prevalent users; the new user cohort comprised 830 963 people. Discharge from hospital (39.3%), therapeutic procedures (21.4%) and emergency department visits (19.7%) were common; a hospital admission for injury (6.0%) or a past-year history of cancer (7.2%) were less common. At 1 year after initiation, 4.6% of people were using oxycodone. In the multivariable model, new use of oxycodone increased with age and was higher for people outside major cities, for example, an incidence rate ratio of 1.43 (95% confidence interval 1.36-1.51) for inner regional areas relative to major cities; there was no evidence of variation in rates of new use by social disadvantage.

CONCLUSION

About half of new oxycodone use in NSW was preceded by a recent episode of hospital care or a therapeutic procedure. Higher rates of oxycodone initiation in rural and regional areas were not explained by sociodemographic factors.

Nine recommendations for the emergency department for patients presenting with low back pain based on management and post-discharge outcomes in an Australian, tertiary emergency department

OBJECTIVE

To ascertain and develop recommendations for analgesic management, discharge planning and further healthcare utilisation of adults presenting to an Australian tertiary ED with radicular or low back pain (LBP).

METHODS

This prospective cohort study included adults presenting with non-specific LBP or radicular pain to an Australian tertiary ED. Participants with trauma/cancer-related pain, and those requiring hospital admission or surgical interventions were excluded. The primary outcome was pharmacological and non-pharmacological management delivered in ED, retrospectively collected via electronic medical records. The secondary outcomes include discharge management use, and changes made due to post-discharge healthcare utilisation, as observed by weekly telephone questionnaires over 4-weeks follow-up.

RESULTS

Of the 100 participants recruited, 94 completed follow-up. In ED, pharmacological management was received by 85%, including opioids (62%) and non-steroidal anti-inflammatory drugs (NSAIDS, 63%). Non-pharmacological management was received by 73%, including patient education (71%) and exercise prescription (37%). In the first-week post-discharge, changes to initial discharge plan occurred in 50%, mostly carried out by GPs (76%). Over the follow-up period, 51% received additional investigations/referrals. Pharmacological use decreased by 38% and non-pharmacological use decreased by 10%. 16% of opioid-naïve patients continued using opioids 4-weeks post-discharge.

CONCLUSION

ED presentations for LBP were more often treated pharmacologically than non-pharmacologically, with opioids commonly prescribed and NSAIDs potentially under-utilised. Post-discharge, additional investigations/referrals, discharge analgesia reductions and maintenance of non-pharmacological management were common. Opioid initiation as a result of LBP presentations, signifies a potential 'gateway' towards unintentional long-term use. Key study findings form our nine recommendations to inform ED LBP pain management.

Defining opioid naïve and implications for monitoring opioid use: A population-based study in Alberta, Canada

PURPOSE

Reducing initial exposure of "opioid naïve" patients to opioids is a public health priority. Identifying opioid naïve patients is difficult, as numerous definitions are used. The objective is to summarize current definitions and evaluate their impact on opioid naïve measures in Alberta.

METHODS

An exploratory data analysis of the literature was conducted over the last 10 years to identify definitions commonly used in the literature to define opioid naïve. Then, using these definitions as a guide, we descriptively report the proportion of patients in Alberta between 2017 and 2021 who would be considered as opioid naïve using these definitions and all opioid dispensing data.

RESULTS

Three categories of definitions were broadly identified: (1) no opioid use within the previous 30 days/6 months/1 year, based on dispensation date; (2) no opioid use based on dispensation date plus days of supply; and, (3) exclusion of codeine from Definitions 1 and 2. Applying these definitions to the Alberta population showed a very wide range in the proportion who would be considered as opioid naïve. Overall, 36.4% of Albertans (n = 1 551 075) had an opioid dispensation in 2017-2021. The average age was 46.6 ± 18.8 and 52.8% were female. The proportion of opioid naïve were most affected by the "opioid free" period, with 97.4%, 83.2%, and 65.6% being classified as opioid naïve using time windows from Definition 1 (30 days, 6 months, 1 year of no prior opioid use). Definitions 2 and 3 did not materially change the results. Further extending the "opioid free" period to 2 years showed only 35% were opioid naïve.

CONCLUSIONS

The most convenient definition for "opioid naïve" was the use of an "opioid free" period. The choice of window would depend on how the information may be used to assistant in clinical decisions with longer windows more likely to reflect true opioid naïve patients. Irrespective of definition used, a large proportion of opioid users would be considered opioid naïve in Alberta.

Continued Opioid Use and Adverse Events Following Provision of Opioids for Musculoskeletal Pain in the Emergency Department: A Systematic Review and Meta-Analysis

BACKGROUND

The prevalence of continued opioid use or serious adverse events (SAEs) following opioid therapy in the emergency department (ED) for musculoskeletal pain is unclear. The aim of this review was to examine the prevalence of continued opioid use and serious adverse events (SAEs) following the provision of opioids for musculoskeletal pain in the emergency department (ED) or at discharge.

METHODS

Records were searched from MEDLINE, EMBASE and CINAHL from inception to 7 October 2022. We included randomised controlled trials and observational studies enrolling adult patients with musculoskeletal pain who were administered and/or prescribed opioids in the ED. Continued opioid use and opioid misuse data after day 4 since ED discharge were extracted. Adverse events were coded using the Common Terminology Criteria for Adverse Events (CTCAE), and those rated as grades 3-4 (severe or life-threatening) and grade 5 (death) were considered SAEs. Risk of bias was assessed using the Quality in Prognosis Studies (QUIPS) tool.

RESULTS

Seventy-two studies were included. Among opioid-naïve patients who received an opioid prescription, 6.8-7.0% reported recent opioid use at 3-12 months after discharge, 4.4% filled ≥ 5 opioid prescriptions and 3.1% filled > 90-day supply of opioids within 6 months. The prevalence of SAEs was 0.02% [95% confidence interval (CI) 0, 0.2%] in the ED and 0.1% (95% CI 0, 1.5%) within 2 days. One study observed 42.9% of patients misused opioids within 30 days after discharge.

CONCLUSIONS

Around 7% of opioid-naïve patients with musculoskeletal pain receiving opioid therapy continue opioid use at 3-12 months after ED discharge. SAEs following ED administration of an opioid were uncommon; however, studies only monitored patients for 2 days.

PROTOCOL REGISTRATION

10.31219/osf.io/w4z3u.

Examining buprenorphine diversion through a harm reduction lens: an agent-based modeling study

BACKGROUND

Recent policies have lessened restrictions around prescribing buprenorphine-naloxone (buprenorphine) for the treatment of opioid use disorder (OUD). The primary concern expressed by critics of these policies is the potential for buprenorphine diversion. However, the population-level effects of increased buprenorphine diversion are unclear. If replacing the use of heroin or fentanyl, use of diverted buprenorphine could be protective.

METHODS

Our study aim was to estimate the impact of buprenorphine diversion on opioid overdose using an agent-based model calibrated to North Carolina. We simulated the progression of opioid misuse and opioid-related outcomes over a 5-year period. Our status quo scenario assumed that 50% of those prescribed buprenorphine diverted at least one dose per week to other individuals with OUD and 10% of individuals with OUD used diverted buprenorphine at least once per week. A controlled prescription only scenario assumed that no buprenorphine would be diverted, while an increased diversion scenario assumed that 95% of those prescribed buprenorphine diverted and 50% of individuals with OUD used diverted buprenorphine. We assumed that use of diverted buprenorphine replaced the use of other opioids for that day. Sensitivity analyses increased the risk of overdose when using diverted buprenorphine, increased the frequency of diverted buprenorphine use, and simulated use of diverted buprenorphine by opioid-naïve individuals. Scenarios were compared on opioid overdose-related outcomes over the 5-year period.

RESULTS

Our status quo scenario predicted 10,658 (credible interval [CI]: 9699-11,679) fatal opioid overdoses. A scenario simulating controlled prescription only of buprenorphine (i.e., no diversion) resulted in 10,741 (9895-11,650) fatal opioid overdoses versus 10,301 (9439-11,244) within a scenario simulating increased diversion. Compared to the status quo, the controlled prescription only scenario resulted in a similar number of fatal overdoses, while the scenario with increased diversion of buprenorphine resulted in 357 (3.35%) fewer fatal overdoses. Even when increasing overdose risk while using diverted buprenorphine and incorporating use by opioid naïve individuals, increased diversion did not increase overdoses compared to a scenario with no buprenorphine diversion.

CONCLUSIONS

A similar number of opioid overdoses occurred under modeling conditions with increased rates of buprenorphine diversion among persons with OUD, with non-statistical trends toward lower opioid overdoses. These results support existing calls for low- to no-barrier access to buprenorphine for persons with OUD.

Patient motivators to use opioids for acute pain after emergency care

Introduction

Patients are stakeholders in their own pain management. Factors motivating individuals to seek or use opioids therapeutically for treatment of acute pain are not well characterized but could be targeted to reduce incident iatrogenic opioid use disorder (OUD). Emergency departments (EDs) commonly encounter patients in acute pain for whom decisions regarding opioid therapy are required. Decision-making is necessarily challenged in episodic, unscheduled care settings given time pressure, limited information, and lack of pre-existing patient provider relationship. Patients may decline to take prescribed opioids or conversely seek opioids from other providers or non-medical sources.

Methods

Using a framework analysis approach, we qualitatively analyzed transcripts from 29 patients after discharge from an ED visit for acute pain at a large, urban, academic hospital in the midwestern United States to describe motivating factors influencing patient decisions regarding opioid use for acute pain. A semi-structured interview guide framed participant discussion in either a focus group or interview transcribed and analyzed with conventional content analysis.

Results

Four major themes emerged from our analysis including a) pain management literacy, b) control preferences, c) risk tolerance, and d) cues to action.

Discussion

Our findings suggest targets for future intervention development and a framework to guide the engagement of patients as stakeholders in their own acute pain management.

Effectiveness of Opioid Analgesic Medicines Prescribed in or at Discharge From Emergency Departments for Musculoskeletal Pain : A Systematic Review and Meta-analysis

BACKGROUND

The comparative benefits and harms of opioids for musculoskeletal pain in the emergency department (ED) are uncertain.

PURPOSE

To evaluate the comparative effectiveness and harms of opioids for musculoskeletal pain in the ED setting.

DATA SOURCES

Electronic databases and registries from inception to 7 February 2022.

STUDY SELECTION

Randomized controlled trials of any opioid analgesic compared with placebo or a nonopioid analgesic administered or prescribed to adults in or on discharge from the ED.

DATA EXTRACTION

Pain and disability were rated on a scale of 0 to 100 and pooled using a random-effects model. Certainty of evidence was assessed using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework.

DATA SYNTHESIS

Forty-two articles were included (n = 6128). In the ED, opioids were statistically but not clinically more effective in reducing pain in the short term (about 2 hours) than placebo and paracetamol (acetaminophen) but were not clinically or statistically more effective than nonsteroidal anti-inflammatory drugs (NSAIDs) or local or systemic anesthetics. Opioids may carry higher risk for harms than placebo, paracetamol, or NSAIDs, although evidence is very uncertain. There was no evidence of difference in harms associated with local or systemic anesthetics.

LIMITATIONS

Low or very low GRADE ratings for some outcomes, unexplained heterogeneity, and little information on long-term outcomes.

CONCLUSION

The risk-benefit balance of opioids versus placebo, paracetamol, NSAIDs, and local or systemic anesthetics is uncertain. Opioids may have equivalent pain outcomes compared with NSAIDs, but evidence on comparisons of harms is very uncertain and heterogeneous. Although factors such as route of administration or dosage may explain some heterogeneity, more work is needed to identify which subgroups will have a more favorable benefit-risk balance for one analgesic over another. Longer-term pain management once dose thresholds are reached is also uncertain.

PRIMARY FUNDING SOURCE

None. (PROSPERO: CRD42021275293).

Transcutaneous electrical nerve stimulation effects on pain-intensity and endogenous opioids levels among chronic low-back pain patients: A randomised controlled trial

BACKGROUND

Transcutaneous electrical nerve stimulation (TENS) is a promising non-pharmacological modality for the management of chronic low back pain (CLBP), but its efficacy and mode of action have not been clearly established.

OBJECTIVE

To evaluate the responses of plasma beta-endorphin (βE), met-enkephalin (ME), and pain intensity (PI) among patients with CLBP exposed to TENS or sham-TENS.

METHODS

This double-blind trial involved 62 participants (aged 53.29 ± 5.07 years) randomised into TENS group (frequency 100 Hz, burst-rate 2 Hz, burst-width 150 μs, intensity 40 mA, duration 30 min), and sham-TENS group. The PI and plasma concentrations of βE and ME were measured at baseline, immediately (0 hr), 1 hr, 24 hrs, and 48 hrs post-intervention. Data were analysed using general linear model repeated measures, ordinal regression, one-way analysis of variance, Kruskal-Wallis test, independent and paired samples t-tests, Mann-Whitney U test, Wilcoxon signed-rank test, and Kendall's tau coefficient.

RESULTS

There was a significant temporal difference in PI between groups, F (1, 58) = 18.83, p< 0.001; the TENS group had better pain relief. The relative analgesic effect of TENS started immediately after the intervention (median difference [M⁢D] =-3, p< 0.001), peaked at 1 hr (M⁢D=-4, p< 0.001), and worn out by 24 hrs (M⁢D=-1, p= 0.029). However, there was no significant difference in βE and ME between the groups from 0 hr to 24 hrs post interventions, and no significant correlation between the PI, and βE, or ME.

CONCLUSION

TENS significantly reduced PI up to 24 hrs after treatment.

Impact of individual and treatment characteristics on wearable sensor-based digital biomarkers of opioid use

Opioid use disorder is one of the most pressing public health problems of our time. Mobile health tools, including wearable sensors, have great potential in this space, but have been underutilized. Of specific interest are digital biomarkers, or end-user generated physiologic or behavioral measurements that correlate with health or pathology. The current manuscript describes a longitudinal, observational study of adult patients receiving opioid analgesics for acute painful conditions. Participants in the study are monitored with a wrist-worn E4 sensor, during which time physiologic parameters (heart rate/variability, electrodermal activity, skin temperature, and accelerometry) are collected continuously. Opioid use events are recorded via electronic medical record and self-report. Three-hundred thirty-nine discreet dose opioid events from 36 participant are analyzed among 2070 h of sensor data. Fifty-one features are extracted from the data and initially compared pre- and post-opioid administration, and subsequently are used to generate machine learning models. Model performance is compared based on individual and treatment characteristics. The best performing machine learning model to detect opioid administration is a Channel-Temporal Attention-Temporal Convolutional Network (CTA-TCN) model using raw data from the wearable sensor. History of intravenous drug use is associated with better model performance, while middle age, and co-administration of non-narcotic analgesia or sedative drugs are associated with worse model performance. These characteristics may be candidate input features for future opioid detection model iterations. Once mature, this technology could provide clinicians with actionable data on opioid use patterns in real-world settings, and predictive analytics for early identification of opioid use disorder risk.

Trends in potentially inappropriate opioid prescribing and associated risk factors among Korean noncancer patients prescribed non-injectable opioid analgesics

Introduction

The aim of this study was to investigate trends in the prevalence of potentially inappropriate opioid prescribing (PIOP) and identify potential risk factors among Korean noncancer patients.

Methods

We conducted a cross-sectional study of annual national patient sample data from the Korean Health Insurance Review and Assessment Service (HIRA-NPS) for the period 2012-2018. Noncancer patients who were prescribed non-injectable opioid analgesics (NIOAs) at least once were included. The proportion of patients with at least one PIOP in terms of concurrent use of benzodiazepines or gabapentinoids, substance use disorder, treatment duration, and dosage was evaluated. Multivariable logistic regression was performed to identify the risk factors associated with PIOP.

Results

Of the 9,772,503 noncancer patients, 1,583,444 (16.2%) were prescribed NIOAs at least once. Among them, 15.7% were exposed to PIOP, and the prevalence was much higher (31.6%) in the elderly group (age: ⩾65 years). The prevalence of PIOP increased 1.1-fold over 7 years (14.8-16.8%) among the total NIOA users and was more pronounced in non-tramadol NIOA users (a 1.5-fold increase, from 13.2% to 19.4%). Multivariable logistic regression indicated that older age, beneficiaries of medical aid or national meritorious service, exposure to polypharmacy, psychological disorder, chronic pain indication, and concomitant sedative use were independently associated with higher odds of PIOP.

Discussion and Conclusion

We found that the prevalence of PIOP was 15.7% among Korean noncancer patients, and it increased over the 7-year study period. This increasing trend is alarming because it was more drastic with non-tramadol NIOAs compared with that with tramadol. Several patient-level risk factors associated with PIOP would be useful in targeted management strategies for the safe use of opioids.

Plain Language Summary

Potentially inappropriate opioid prescribing and related risk factors among noncancer patients prescribed non-injectable opioids in Korea In Korea, the prevalence of non-injectable opioid analgesic (NIOA) use in noncancer patients steadily increased from 15.3% in 2012 to 17.1% in 2018.Also, the prevalence of potentially inappropriate opioid prescribing (PIOP) increased from 14.8% in 2012 to 16.8% in 2018.The following factors were associated with a markedly increased risk of PIOP: age, beneficiaries of medical aid or national meritorious service, polypharmacy, psychological disorder, chronic pain, and concomitant medications.

Trends and risk factors for opioid administration for non-emergent lower back pain

BACKGROUND

Non-emergent low-back pain (LBP) is one of the most prevalent presenting complaints to the emergency department (ED) and has been shown to contribute to overcrowding in the ED as well as diverting attention away from more serious complaints. There has been an increasing focus in current literature regarding ED admission and opioid prescriptions for general complaints of pain, however, there is limited data concerning the trends over the last decade in ED admissions for non-emergent LBP as well as any subsequent opioid prescriptions by the ED for this complaint.

AIM

To determine trends in non-emergent ED visits for back pain; annual trends in opioid administration for patients presenting to the ED for back pain; and factors associated with receiving an opioid-based medication for non-emergent LBP in the ED

METHODS

Patients presenting to the ED for non-emergent LBP from 2010 to 2017 were retrospectively identified from the National Hospital Ambulatory Medical Care Survey database. The “year” variable was transformed to two-year intervals, and a weighted survey analysis was conducted utilizing the weighted variables to generate incidence estimates. Bivariate statistics were used to assess differences in count data, and logistic regression was performed to identify factors associated with patients being discharged from the ED with narcotics. Statistical significance was set to a P value of 0.05.

RESULTS

Out of a total of 41658475 total ED visits, 3.8% (7726) met our inclusion and exclusion criteria. There was a decrease in the rates of non-emergent back pain to the ED from 4.05% of all cases during 2010 and 2011 to 3.56% during 2016 and 2017. The most common opioids prescribed over the period included hydrocodone-based medications (49.1%) and tramadol-based medications (16.9), with the combination of all other opioid types contributing to 35.7% of total opioids prescribed. Factors significantly associated with being prescribed narcotics included age over 43.84-years-old, higher income, private insurance, the obtainment of radiographic imaging in the ED, and region of the United States (all, P < 0.05). Emergency departments located in the Midwest [odds ratio (OR): 2.42, P < 0.001], South (OR: 2.35, < 0.001), and West (OR: 2.57, P < 0.001) were more likely to prescribe opioid-based medications for non-emergent LBP compared to EDs in the Northeast.

CONCLUSION

From 2010 to 2017, there was a significant decrease in the number of non-emergent LBP ED visits, as well as a decrease in opioids prescribed at these visits. These findings may be attributed to the increased focus and regulatory guidelines on opioid prescription practices at both the federal and state levels. Since non-emergent LBP is still a highly common ED presentation, conclusions drawn from opioid prescription practices within this cohort is necessary for limiting unnecessary ED opioid prescriptions.