Measuring a Crisis: Questioning the Use of Naloxone Administrations as a Marker for Opioid Overdoses in a Large U.S. EMS System

Naloxone use by Aotearoa New Zealand emergency medical services, 2017-2021

OBJECTIVE

Emergency medical services (EMS) use of naloxone in the prehospital setting is indicated in patients who have significantly impaired breathing or level of consciousness when opioid intoxication is suspected. The present study characterised naloxone use in a nationwide sample of Aotearoa New Zealand road EMS patients to establish a baseline for surveillance of any changes in the future.

METHODS

A retrospective analysis of rates of patients with naloxone administrations was conducted using Hato Hone St John (2017-2021) and Wellington Free Ambulance (2018-2021) electronic patient report form datasets. Patient demographics, presenting complaints, naloxone dosing, and initial and last vital sign clinical observations were described.

RESULTS

There were 2018 patients with an equal proportion of males and females, and patient median age was 47 years. There were between 8.0 (in 2018) and 9.0 (in 2020) naloxone administrations per 100 000 population-years, or approximately one administration per day for the whole country of 5 million people. Poisoning by unknown agent(s) was the most common presenting complaint (61%). The median dose of naloxone per patient was 0.4 mg; 85% was administered intravenously. The median observed change in Glasgow Coma Scale score was +1, and respiratory rate increased by +2 breaths/min.

CONCLUSIONS

A national rate of EMS naloxone patients was established; measured clinical effects of naloxone were modest, suggesting many patients had reasons other than opioid toxicity contributing to their symptoms. Naloxone administration rates provide indirect surveillance information about suspected harmful opioid exposures but need to be interpreted with care.

An interpretable machine learning framework for opioid overdose surveillance from emergency medical services records

The goal of this study is to develop and validate a lightweight, interpretable machine learning (ML) classifier to identify opioid overdoses in emergency medical services (EMS) records. We conducted a comparative assessment of three feature engineering approaches designed for use with unstructured narrative data. Opioid overdose annotations were provided by two harm reduction paramedics and two supporting annotators trained to reliably match expert annotations. Candidate feature engineering techniques included term frequency-inverse document frequency (TF-IDF), a highly performant approach to concept vectorization, and a custom approach based on the count of empirically-identified keywords. Each feature set was trained using four model architectures: generalized linear model (GLM), Naïve Bayes, neural network, and Extreme Gradient Boost (XGBoost). Ensembles of trained models were also evaluated. The custom feature models were also assessed for variable importance to aid interpretation. Models trained using TF-IDF feature engineering ranged from AUROC = 0.59 (95% CI: 0.53-0.66) for the Naïve Bayes to AUROC = 0.76 (95% CI: 0.71-0.81) for the neural network. Models trained using concept vectorization features ranged from AUROC = 0.83 (95% 0.78-0.88)for the Naïve Bayes to AUROC = 0.89 (95% CI: 0.85-0.94) for the ensemble. Models trained using custom features were the most performant, with benchmarks ranging from AUROC = 0.92 (95% CI: 0.88-0.95) with the GLM to 0.93 (95% CI: 0.90-0.96) for the ensemble. The custom features model achieved positive predictive values (PPV) ranging for 80 to 100%, which represent substantial improvements over previously published EMS encounter opioid overdose classifiers. The application of this approach to county EMS data can productively inform local and targeted harm reduction initiatives.

Overdose and mortality risk following a non-fatal opioid overdose treated by Emergency Medical Services in King County, Washington

BACKGROUND

Emergency Medical Services (EMS) agencies respond to hundreds of thousands of acute overdose events each year. We conducted a retrospective cohort study of EMS patients who survived a prior opioid overdose in 2019-2021 in King County, Washington.

METHODS

A novel record linkage algorithm was applied to EMS electronic health records and the state vital statistics registry to identify repeat overdoses and deaths that occurred up to 3 years following the index opioid overdose. We measured overdose incidence rates and applied survival analysis techniques to assess all-cause and overdose-specific mortality risks.

RESULTS

In the year following the index opioid overdose, the overdose (fatal or non-fatal) incidence rate was 23.3 per 100 person-year, overdose mortality rate was 2.7 per 100 person-year, and all-cause mortality rate was 5.2 per 100 person-year in this cohort of overdose survivors (n=4234). Overdose incidence was highest in the first 30 days following the index overdose (43 opioid overdoses and 4 fatal overdoses per 1000 person-months), declined precipitously, and then plateaued from the third month onwards (10-15 opioid overdoses and 1-2 fatal overdoses per 1000 person-months). Overdose incidence rates, measured at 30 days, were highest among overdose survivors who were young, male, and experienced a low severity index opioid overdose, but these differences diminished when measured at 12 months.

CONCLUSIONS

Among EMS patients who survived an opioid overdose, the risk of subsequent overdose is high, especially in the weeks following the index opioid overdose. Non-fatal overdose may represent a pivotal time to connect patients with harm-reduction, treatment, and other support services.

Redefining and categorizing emergency medical service opioid-related incidents in Massachusetts

AIMS

To create a novel emergency medical service (EMS) opioid-related incident (ORI) tiering framework to describe more accurately the opioid epidemic in Massachusetts. By classifying the data, we could more accurately detail differing trends among the new categories.

DESIGN

Free-text fields of Massachusetts EMS reports, from 2013 through 2020, were analyzed to identify ORIs and then categorized into a five-tier severity cascade based on symptom presentation: 'dead on arrival,' 'acute overdose,' 'intoxication,' 'withdrawal' and 'other ORI.' As a validation of the new classification, an emergency medical technician, paramedic and emergency medical physician reviewed clinical reports and assigned a severity category to 100 randomly selected cases. The algorithm then assessed the same 100 cases to determine if it could accurately identify the severity category for each case.

FINDINGS

Validation of the algorithm by clinical review indicated a substantial level of agreement between the algorithm and the reviewers. Over half of all ORIs were acute overdose (55%), 21% were intoxication, 20% were other ORI, 3% were withdrawal, and 1% were dead on arrival. Overall ORIs decreased in 2020, but the number of 'dead on arrival' increased 32% from 2019. Administration of naloxone also differed between the categories, with 95% of acute overdose and 29% of intoxication receiving naloxone.

CONCLUSIONS

This novel categorization of emergency medical service opioid-related incidents in Massachusetts, United States, reveals new trend details and strains on the emergency medical service system. Using these categories also improves dataset linkage within the state and interstate rate comparisons.

The changing epidemiology of opioid overdose in Baltimore, Maryland, 2012-2017: insights from emergency medical services

INTRODUCTION

An estimated 100,306 people died from an overdose from May 2020 to April 2021. Emergency Medical Services (EMS) are often the first responder to opioid overdose, and EMS encounter records can provide granular epidemiologic data on opioid overdose. This study describes the demographic, temporal, and geographic epidemiology of suspected opioid overdose in Baltimore City using data from Baltimore City Fire Department EMS encounters with the administration of the opioid antagonist naloxone.

METHOD

The present analyses used patient encounter data from 2012 to 2017 from the Baltimore City Fire Department, the city's primary provider of EMS services. The analytic sample included patient encounters within the city that involved naloxone administration to patients 15 years of age or older (n = 20,592). Negative binomial regression was used to calculate the incidence rates based on demographic characteristics, year, and census tract. Choropleth maps were used to show the geographic distribution of overdose incidence across census tracts in 2013, 2015, and 2017.

RESULTS

From 2012 to 2017, the annual number of EMS encounters with naloxone administrations approximately doubled every 2 years, and the temporal pattern of naloxone administration was similar to the pattern of fatal opioid-related overdoses. For most census tracts, incidence rates significantly increased over time. Population-based incidence of naloxone administration varied significantly by socio-demographic characteristics. Males, non-whites, and those 25-69 years of age had the highest incidence rates.

CONCLUSION

The incidence of naloxone administration increased dramatically over the study period. Despite significant cross-sectional variation in incidence across demographically and geographically defined groups, there were significant proportional increases in incidence rates, consistent with fatal overdose rates over the period. This study demonstrated the value of EMS data for understanding the local epidemiology of opioid-related overdose. Key MessagesPatterns of EMS encounters with naloxone administration appear to be an excellent proxy for patterns of opioid-related overdoses based on the consistency of fatal overdose rates over time.EMS plays a central role in preventing fatal opioid-related overdoses through the administration of naloxone, provision of other emergency services, and transportation to medical facilities.EMS encounters with naloxone administration could also be used to evaluate the impact of overdose prevention interventions and public health services.

Identification of Non-Fatal Opioid Overdose Cases Using 9-1-1 Computer Assisted Dispatch and Prehospital Patient Clinical Record Variables

Background: The current epidemic of opioid overdoses in the United States necessitates a robust public health and clinical response. We described patterns of non-fatal opioid overdoses (NFOODs) in a small western region using data from the 9-1-1 Computer Assisted Dispatch (CAD) record and electronic Patient Clinical Records (ePCR) completed by EMS responders. We determined whether CAD and ePCR variables could identify NFOOD cases in 9-1-1 data for intervention and surveillance efforts. Methods: We conducted a retrospective analysis of 1 year of 9-1-1 emergency medical CAD and ePCR (including naloxone administration) data from the sole EMS provider in the response area. Cases were identified based on clinician review of the ePCR, and categorized as definitive NFOOD, probable NFOOD, or non-OOD. Sensitivity, specificity, positive and negative predictive values (PPV and NPV) of the most prevalent CAD and ePCR variables were calculated. We used a machine learning technique-Random-Forests (RF) modeling-to optimize our ability to accurately predict NFOOD cases within census blocks. Results: Of 37,960 9-1-1 calls, clinical review identified 158 NFOOD cases (0.4%), of which 123 (77.8%) were definitive and 35 (22.2%) were probable cases. Overall, 106 (67.1%) received naloxone from the EMS responder at the scene. As a predictor of NFOOD, naloxone administration by paramedics had 67.1% sensitivity, 99.6% specificity, 44% PPV, and 99.9% NPV. Using CAD variables alone achieved a sensitivity of 36.7% and specificity of 99.7%. Combining ePCR variables with CAD variables increased the diagnostic accuracy with the best RF model yielding 75.9% sensitivity, 99.9% specificity, 71.4% PPV, and 99.9% NPV. Conclusion: CAD problem type variables and naloxone administration, used alone or in combination, had sub-optimal predictive accuracy. However, a Random Forests modeling approach improved accuracy of identification, which could foster improved surveillance and intervention efforts. We identified the set of NFOODs that EMS encountered in a year and may be useful for future surveillance efforts.

Ineffectiveness of paramedic naloxone administration as a standalone metric for community opioid overdoses and the increasing use of naloxone by community members

Introduction:With Canada's growing opioid crisis, many communities are attempting to monitor cases in real-time. Paramedic Naloxone Administration (PNA) has become a common metric for monitoring overdoses. We evaluate whether the use of naloxone administration counts represents an effective monitoring tool for community opioid overdoses.Methods:The electronic ambulance call report database of Peterborough Paramedics (Ontario, Canada) was examined. De-identified records from 2016-2019 with problem codes of "Opioid Overdose", along with all patients documented as receiving naloxone were extracted. Chi-square and Bonferroni-adjusted post hoc proportion tests were used for comparison of counts.Results:558 opioid overdoses were identified, 124 (22%) of which had PNA documented, 181(32%) had naloxone prior to arrival documented and 264 (47%) received no naloxone. Over the three years, the annual number of overdose cases increased, while the proportion of patients receiving PNA decreased significantly each year. PNA was also associated with calls in a residence. Naloxone was administered by a non-paramedic in 262 cases, with 181 of these identified as opioid overdoses and was more common in later years and in cases occurring in public places.Conclusion:PNA calls did not account for a significant percentage of opioid overdoses attended to by paramedics. The strong association between PNA and call location being a residence, along with increasing use of community naloxone kits, may cause certain populations to be under-represent if PNA is used as a standalone metric. The decreasing association with time may also lead to a falsely improving metric further reducing its effectiveness. Thus, PNA when used alone may no longer be a suitable metric for opioid overdose tracking.

Emergency Medical Services and Syndromic Surveillance: A Comparison With Traditional Surveillance and Effects on Timeliness

OBJECTIVE

Traditional public health surveillance of nonfatal opioid overdose relies on emergency department (ED) billing data, which can be delayed substantially. We compared the timeliness of 2 new data sources for rapid drug overdose surveillance-emergency medical services (EMS) and syndromic surveillance-with ED billing data.

METHODS

We used data on nonfatal opioid overdoses in Kentucky captured in EMS, syndromic surveillance, and ED billing systems during 2018-2019. We evaluated the time-series relationships between EMS and ED billing data and syndromic surveillance and ED billing data by calculating cross-correlation functions, controlling for influences of autocorrelations. A case example demonstrates the usefulness of EMS and syndromic surveillance data to monitor rapid changes in opioid overdose encounters in Kentucky during the COVID-19 epidemic.

RESULTS

EMS and syndromic surveillance data showed moderate-to-strong correlation with ED billing data on a lag of 0 (r = 0.694; 95% CI, 0.579-0.782; t = 9.73; df = 101; P < .001; and r = 0.656; 95% CI, 0.530-0.754; t = 8.73; df = 101; P < .001; respectively) at the week-aggregated level. After the COVID-19 emergency declaration, EMS and syndromic surveillance time series had steep increases in April and May 2020, followed by declines from June through September 2020. The ED billing data were available for analysis 3 months after the end of a calendar quarter but closely followed the trends identified by the EMS and syndromic surveillance data.

CONCLUSION

Data from EMS and syndromic surveillance systems can be reliably used to monitor nonfatal opioid overdose trends in Kentucky in near-real time to inform timely public health response.

Linking Emergency Medical Services and Emergency Department Data to Improve Overdose Surveillance in North Carolina

Introduction

Linking emergency medical services (EMS) data to emergency department (ED) data enables assessing the continuum of care and evaluating patient outcomes. We developed novel methods to enhance linkage performance and analysis of EMS and ED data for opioid overdose surveillance in North Carolina.

Methods

We identified data on all EMS encounters in North Carolina during January 1–November 30, 2017, with documented naloxone administration and transportation to the ED. We linked these data with ED visit data in the North Carolina Disease Event Tracking and Epidemiologic Collection Tool. We manually reviewed a subset of data from 12 counties to create a gold standard that informed developing iterative linkage methods using demographic, time, and destination variables. We calculated the proportion of suspected opioid overdose EMS cases that received International Classification of Diseases, Tenth Revision, Clinical Modification diagnosis codes for opioid overdose in the ED.

Results

We identified 12 088 EMS encounters of patients treated with naloxone and transported to the ED. The 12-county subset included 1781 linkage-eligible EMS encounters, with historical linkage of 65.4% (1165 of 1781) and 1.6% false linkages. Through iterative linkage methods, performance improved to 91.0% (1620 of 1781) with 0.1% false linkages. Among statewide EMS encounters with naloxone administration, the linkage improved from 47.1% to 91.1%. We found diagnosis codes for opioid overdose in the ED among 27.2% of statewide linked records.

Practice Implications

Through an iterative linkage approach, EMS–ED data linkage performance improved greatly while reducing the number of false linkages. Improved EMS–ED data linkage quality can enhance surveillance activities, inform emergency response practices, and improve quality of care through evaluating initial patient presentations, field interventions, and ultimate diagnoses.

Using Surveillance With Near-Real-Time Alerts During a Cluster of Overdoses From Fentanyl-Contaminated Crack Cocaine, Connecticut, June 2019

In 2019, Connecticut launched an opioid overdose-monitoring program to provide rapid intervention and limit opioid overdose-related harms. The Connecticut Statewide Opioid Response Directive (SWORD)-a collaboration among the Connecticut State Department of Public Health, Connecticut Poison Control Center (CPCC), emergency medical services (EMS), New England High Intensity Drug Trafficking Area (HIDTA), and local harm reduction groups-required EMS providers to call in all suspected opioid overdoses to the CPCC. A centralized data collection system and the HIDTA overdose mapping tool were used to identify outbreaks and direct interventions. We describe the successful identification of a cluster of fentanyl-contaminated crack cocaine overdoses leading to a rapid public health response. On June 1, 2019, paramedics called in to the CPCC 2 people with suspected opioid overdose who reported exclusive use of crack cocaine after being resuscitated with naloxone. When CPCC specialists in poison information followed up on the patients' status with the emergency department, they learned of 2 similar cases, raising suspicion that a batch of crack cocaine was mixed with an opioid, possibly fentanyl. The overdose mapping tool pinpointed the overdose nexus to a neighborhood in Hartford, Connecticut; the CPCC supervisor alerted the Connecticut State Department of Public Health, which in turn notified local health departments, public safety officials, and harm reduction groups. Harm reduction groups distributed fentanyl test strips and naloxone to crack cocaine users and warned them of the dangers of using alone. The outbreak lasted 5 days and tallied at least 22 overdoses, including 6 deaths. SWORD's near-real-time EMS reporting combined with the overdose mapping tool enabled rapid recognition of this overdose cluster, and the public health response likely prevented additional overdoses and loss of life.

Characterizing Opioid Overdoses Using Emergency Medical Services Data : A Case Definition Algorithm Enhanced by Machine Learning

OBJECTIVES

Tracking nonfatal overdoses in the escalating opioid overdose epidemic is important but challenging. The objective of this study was to create an innovative case definition of opioid overdose in North Carolina emergency medical services (EMS) data, with flexible methodology for application to other states' data.

METHODS

This study used de-identified North Carolina EMS encounter data from 2010-2015 for patients aged >12 years to develop a case definition of opioid overdose using an expert knowledge, rule-based algorithm reflecting whether key variables identified drug use/poisoning or overdose or whether the patient received naloxone. We text mined EMS narratives and applied a machine-learning classification tree model to the text to predict cases of opioid overdose. We trained models on the basis of whether the chief concern identified opioid overdose.

RESULTS

Using a random sample from the data, we found the positive predictive value of this case definition to be 90.0%, as compared with 82.7% using a previously published case definition. Using our case definition, the number of unresponsive opioid overdoses increased from 3412 in 2010 to 7194 in 2015. The corresponding monthly rate increased by a factor of 1.7 from January 2010 (3.0 per 1000 encounters; n = 261 encounters) to December 2015 (5.1 per 1000 encounters; n = 622 encounters). Among EMS responses for unresponsive opioid overdose, the prevalence of naloxone use was 83%.

CONCLUSIONS

This study demonstrates the potential for using machine learning in combination with a more traditional substantive knowledge algorithm-based approach to create a case definition for opioid overdose in EMS data.

Michigan system for opioid overdose surveillance

Community rapid response may reduce opioid overdose harms, but is hindered by the lack of timely data. To address this need, we created and evaluated the Michigan system for opioid overdose surveillance (SOS). SOS integrates suspected fatal overdose data from Medical Examiners (MEs), and suspected non-fatal overdoses (proxied by naloxone administration) from the Michigan Emergency Medical Services (EMS) into a web-based dashboard that was developed with stakeholder feedback. Authorised stakeholders can view approximate incident locations and automated spatiotemporal data summaries, while the general public can view county-level summaries. Following Centers for Disease Control and Prevention (CDC) surveillance system evaluation guidelines, we assessed simplicity, flexibility, data quality, acceptability, sensitivity, positive value positive (PVP), representativeness, timeliness and stability of SOS. Data are usually integrated into SOS 1-day postincident, and the interface is updated weekly for debugging and new feature addition, suggesting high timeliness, stability and flexibility. Regarding representativeness, SOS data cover 100% of EMS-based naloxone adminstrations in Michigan, and receives suspected fatal overdoses from MEs covering 79.1% of Michigan's population, but misses those receiving naloxone from non-EMS. PVP of the suspected fatal overdose indicator is nearly 80% across MEs. Because SOS uses pre-existing data, added burden on MEs/EMS is minimal, leading to high acceptability; there are over 300 authorised SOS stakeholders (~6 new registrations/week) as of this writing, suggesting high user acceptability. Using a collaborative, cross-sector approach we created a timely opioid overdose surveillance system that is flexible, acceptable, and is reasonably accurate and complete. Lessons learnt can aid other jurisdictions in creating analogous systems.

Public Health Surveillance of Behavioral Health Emergencies through Emergency Medical Services Data

Objective: To identify the demographic, clinical and EMS characteristics of events documented as behavioral health emergencies (BHE) by EMS. Methods: This was a cross-sectional study using the 2018 National Emergency Medical Services Information System (NEMSIS) Version 3 dataset. All events that had patient care provided with a documented impression (field diagnosis) of ICD-10 codes F01-F99 (i.e., mental, behavioral, and neurodevelopmental disorders) were labeled a BHE and included. Descriptive statistics were calculated. Results: A total of 1,594,821 (7.3%) EMS calls had a BHE impression. The most common was mental and behavioral disorders due to psychoactive substance use (42.3%). More males than females had BHEs (54.6% vs. 45.4%), and most patients were ages 18-34 (31.5%). Most BHE occurred in urban settings (89.6%). Almost half (47.9%) were dispatched with a complaint unrelated to behavioral health. Conclusion: BHEs were noted in 7.3% of NEMSIS events, and the majority were associated with substance use disorders. EMS professionals need comprehensive training on best practices for BHE. Stakeholders should have information on prevalence of BHEs to ensure proper educational standards, training practices, and resource allocation.

Non-transport after Prehospital Naloxone Administration Is Associated with Higher Risk of Subsequent Non-fatal Overdose

Objective: U.S. opioid overdoses increased nearly sixfold from 1999 to 2018, and greater than 1% of all emergency medical services (EMS) encounters now involve naloxone administration. While "treat and release" protocols may have low short-term mortality, the risk of subsequent non-fatal overdoses is not known. This study compares the risk of repeat overdose encounters between patients transported to an emergency department (ED) and those who refused transport after prehospital naloxone administration. Methods: All EMS charts within a large single-tier fire-based urban EMS system between January 1 and August 31, 2018 were reviewed if either naloxone administration or a clinical impression related to opioid overdose was documented. Charts were excluded if there was no documented evidence of an opioid toxidrome (respiratory depression or altered mental status), if there was another clear explanation for the symptoms (e.g., hypoglycemia), or if naloxone was not administered. Ten percent of charts were reviewed by a second author to assess reliability. Cox regression (survival analysis) was used to estimate the risk of a subsequent EMS encounter with naloxone administration following an index encounter with naloxone administration. Results: Of the 2143 charts reviewed, 1311 unique patients with 1600 overdose encounters involving naloxone administration were identified. Inter-rater reliability for chart inclusion was strong [κ = 0.83 (95% CI: 0.72-0.90)]. Police/bystanders administered naloxone in 208/1600 (13.0%) encounters. A substantial proportion of encounters resulted in transport refusal (674/1600, 42.1%). The final Cox model included only refusal vs. acceptance of transport to an ED during the index EMS encounter. Patient age, gender, and naloxone administration prior to EMS arrival were not statistically significant in univariate or multivariable analyses, nor were they significant confounders. Refusal of transport was associated with a hazard ratio of 1.66 (95% CI: 1.23-2.23) for subsequent EMS encounters with naloxone administration. Conclusions: Non-transport after prehospital naloxone administration is associated with an increased risk of subsequent non-fatal overdose requiring EMS intervention. Limitations include the use of a single EMS agency as patients may have had uncaptured overdose encounters in neighboring municipalities.

Refusals After Prehospital Administration of Naloxone during the COVID-19 Pandemic

OBJECTIVE

To determine if COVID-19 was associated with a change in patient refusals after Emergency Medical Services (EMS) administration of naloxone.

METHODS

This is a retrospective cohort study in which the incidence of refusals after naloxone administration in a single EMS system was evaluated. The number of refusals after naloxone administration was compared across the before-pandemic interval (01/01/20 to 02/15/20) and the during-pandemic interval (03/16/20 to 04/30/20). For comparison the incidence of all other patient refusals before and during COVID-19 as well as the incidences of naloxone administration before and during COVID-19 were also reported.

RESULTS

Prior to the widespread knowledge of the COVID-19 pandemic, 24 of 164 (14.6%) patients who received naloxone via EMS refused transport. During the pandemic, 55 of 153 (35.9%) patients who received naloxone via EMS refused transport. Subjects receiving naloxone during the COVID-19 pandemic were at greater risk of refusal of transport than those receiving naloxone prior to the pandemic (RR = 2.45; 95% CI 1.6-3.76). Among those who did not receive naloxone, 2067 of 6956 (29.7%) patients were not transported prior to the COVID-19 pandemic and 2483 of 6016 (41.3%) were not transported during the pandemic. Subjects who did not receive naloxone with EMS were at greater risk of refusal of transport during the COVID-19 pandemic than prior to it (RR = 1.39; 95% CI 1.32-1.46).

CONCLUSION

In this single EMS system, more than a two-fold increase in the rate of refusal after non-fatal opioid overdose was observed following the COVID-19 outbreak.

Endogenous Opiates and Behavior: 2018

This paper is the forty-first consecutive installment of the annual anthological review of research concerning the endogenous opioid system, summarizing articles published during 2018 that studied the behavioral effects of molecular, pharmacological and genetic manipulation of opioid peptides and receptors as well as effects of opioid/opiate agonists and antagonists. The review is subdivided into the following specific topics: molecular-biochemical effects and neurochemical localization studies of endogenous opioids and their receptors (2), the roles of these opioid peptides and receptors in pain and analgesia in animals (3) and humans (4), opioid-sensitive and opioid-insensitive effects of nonopioid analgesics (5), opioid peptide and receptor involvement in tolerance and dependence (6), stress and social status (7), learning and memory (8), eating and drinking (9), drug abuse and alcohol (10), sexual activity and hormones, pregnancy, development and endocrinology (11), mental illness and mood (12), seizures and neurologic disorders (13), electrical-related activity and neurophysiology (14), general activity and locomotion (15), gastrointestinal, renal and hepatic functions (16), cardiovascular responses (17), respiration and thermoregulation (18), and immunological responses (19).

The Detection of Opioid Misuse and Heroin Use From Paramedic Response Documentation: Machine Learning for Improved Surveillance

Background

Timely, precise, and localized surveillance of nonfatal events is needed to improve response and prevention of opioid-related problems in an evolving opioid crisis in the United States. Records of naloxone administration found in prehospital emergency medical services (EMS) data have helped estimate opioid overdose incidence, including nonhospital, field-treated cases. However, as naloxone is often used by EMS personnel in unconsciousness of unknown cause, attributing naloxone administration to opioid misuse and heroin use (OM) may misclassify events. Better methods are needed to identify OM.

Objective

This study aimed to develop and test a natural language processing method that would improve identification of potential OM from paramedic documentation.

Methods

First, we searched Denver Health paramedic trip reports from August 2017 to April 2018 for keywords naloxone, heroin, and both combined, and we reviewed narratives of identified reports to determine whether they constituted true cases of OM. Then, we used this human classification as reference standard and trained 4 machine learning models (random forest, k-nearest neighbors, support vector machines, and L1-regularized logistic regression). We selected the algorithm that produced the highest area under the receiver operating curve (AUC) for model assessment. Finally, we compared positive predictive value (PPV) of the highest performing machine learning algorithm with PPV of searches of keywords naloxone, heroin, and combination of both in the binary classification of OM in unseen September 2018 data.

Results

In total, 54,359 trip reports were filed from August 2017 to April 2018. Approximately 1.09% (594/54,359) indicated naloxone administration. Among trip reports with reviewer agreement regarding OM in the narrative, 57.6% (292/516) were considered to include information revealing OM. Approximately 1.63% (884/54,359) of all trip reports mentioned heroin in the narrative. Among trip reports with reviewer agreement, 95.5% (784/821) were considered to include information revealing OM. Combined results accounted for 2.39% (1298/54,359) of trip reports. Among trip reports with reviewer agreement, 77.79% (907/1166) were considered to include information consistent with OM. The reference standard used to train and test machine learning models included details of 1166 trip reports. L1-regularized logistic regression was the highest performing algorithm (AUC=0.94; 95% CI 0.91-0.97) in identifying OM. Tested on 5983 unseen reports from September 2018, the keyword naloxone inaccurately identified and underestimated probable OM trip report cases (63 cases; PPV=0.68). The keyword heroin yielded more cases with improved performance (129 cases; PPV=0.99). Combined keyword and L1-regularized logistic regression classifier further improved performance (146 cases; PPV=0.99).

Conclusions

A machine learning application enhanced the effectiveness of finding OM among documented paramedic field responses. This approach to refining OM surveillance may lead to improved first-responder and public health responses toward prevention of overdoses and other opioid-related problems in US communities.

Identifying high-risk areas for nonfatal opioid overdose: a spatial case-control study using EMS run data

PURPOSE

The objective of our study was to incorporate stricter probable nonfatal opioid overdose case criteria, and advanced epidemiologic approaches to more reliably detect local clustering in nonfatal opioid overdose activity in EMS runs data.

METHODS

Data were obtained using emsCharts for our study area in southwestern Pennsylvania from 2007 to 2018. Cases were identified as emergency medical service (EMS) responses where naloxone was administered, and improvement was noted in patient records between initial and final Glasgow Coma Score. A subsample of all-cause EMS responses sites were used as controls and exact matched to cases on sex and 10-year-age category. Clustering was assessed using difference in Ripley's K function for cases and controls and Kulldorff scan statistics.

RESULTS

Difference in K functions indicated no significant difference in probable nonfatal overdose EMS runs across the study area compared to all-cause EMS runs. However, scan statistics did identify significant local clustering of probable nonfatal overdose EMS runs (maximum likelihood = 16.40, P = 0.0003).

CONCLUSIONS

Results highlight relevance of EMS data to detect community-level overdose activity and promote reliable use through stricter case definition criteria and advanced methodological approaches. Techniques examined have the potential to improve targeted delivery of neighborhood-level public health response activities using a near real-time data source.

EMS naloxone administration as non-fatal opioid overdose surveillance: 6-year outcomes in Marion County, Indiana

BACKGROUND AND AIMS

Despite rising rates of opioid overdose in the United States, few studies have examined the frequency of non-fatal overdose events or mortality outcomes following resuscitation. Given the widespread use of naloxone to respond to overdose-related deaths, naloxone administration may provide a useful marker of overdose events to identify high-risk users at heightened risk of mortality. We used naloxone administration by emergency medical services as a proxy measure of non-fatal overdose to examine repeat events and mortality outcomes during a 6-year period.

METHODS

We conducted a retrospective investigation of all cases in Marion County, Indiana between January 2011 and December 2016 where emergency medical services used naloxone to resuscitate a patient. Cases were linked to vital records to assess mortality and cause of death during the same time-period. We used Cox regression survival analysis to assess whether repeat non-fatal overdose events during the study period were associated with the hazard of mortality, both overall and by cause of death.

RESULTS

Of 4726 patients administered naloxone, 9.4% (n = 444) died an average of 354 days [standard deviation (SD) = 412.09, range = 1-1980] following resuscitation. Decedents who died of drug-related causes (34.7%, n = 154) were younger and more likely to have had repeat non-fatal overdose events. Patients with repeat non-fatal overdose events (13.4%, n = 632) had a ×2.07 [95% confidence interval (CI) = 1.59, 2.71] higher hazard of all-cause mortality and a ×3.06 (95% CI = 2.13, 4.40) higher hazard of drug-related mortality.

CONCLUSIONS

Among US emergency medical service patients administered naloxone for opioid overdose, those with repeat non-fatal opioid overdose events are at a much higher risk of mortality, particularly drug-related mortality, than those without repeat events.