Use of diagnosis codes for detection of clinically significant opioid poisoning in the emergency department: A retrospective analysis of a surveillance case definition

Trends in maternal opioid use: Statewide differences by sociodemographic characteristics in Florida from 2000 to 2019

BACKGROUND

Maternal opioid use (MOU) remains a public health concern. Studies have demonstrated significant increases in MOU, but estimates using ICD-10-CM or stratified by sociodemographic variables are limited.

OBJECTIVES

Using a statewide, population-based dataset of Florida resident deliveries from 2000 to 2019, we examined the trend of MOU by age, race/ethnicity, education level, and insurance.

METHODS

Florida administrative data was used to conduct a retrospective cohort study. MOU was identified using opioid-related hospital discharge diagnoses documented prenatally or at delivery. Maternal sociodemographic variables were obtained from Florida vital statistics. Joinpoint regression was used to identify statistically significant changes in the trends overall and stratified by sociodemographic variables. Results are presented as annual percentage changes (APC) and 95% confidence intervals.

RESULTS

Our sample included over 3.6 million Florida resident mothers; of which, MOU was identified in 1% (n = 22,828) of the sample. From 2000 to 2019, MOU increased over ten-fold from 8.7 to 94.7 per 10,000 live birth deliveries. MOU increased significantly from 2000 to 2011 (APC: 32.8; 95% CI: 29.4, 36.2), remained stable from 2011 to 2016, and decreased significantly from 2016 to 2019 (APC: 3.9; 95% CI: -6.6, -1.0). However, from 2016 to 2019, MOU increased among non-Hispanic Black mothers (APC: 9.2; 95% CI: 7.5, 11.0), and those ages 30-34 (APC: 2.9; 95% CI: 1.2, 4.6) and 35-39 (APC: 6.4; 95% CI: 4.3, 8.4).

CONCLUSIONS

Accurate prevalence estimates of MOU by sociodemographic factors are necessary to fully understand prevalence trends, describe the burden among sub-populations, and develop targeted interventions.

Adverse outcomes associated with opioid prescription by dentists in the Veterans Health Administration: A national cross-sectional study from 2015 to 2018

OBJECTIVES

Opioids prescribed by dentists have been associated with serious adverse events, including opioid-related overdose and mortality. However, the downstream outcomes of opioids prescribed by dentists to Veterans who are at high risk for opioid misuse and overdose have yet to be determined.

METHODS

This was a national cross-sectional analysis of opioids associated with dental visits within the Veterans Health Administration from 2015 to 2018. Overprescribing was defined per guidelines as >120 morphine milligram equivalents (MME) or >3 days supply. The association of dental visit and patient characteristics was modeled separately for opioid-related poisoning and all-cause mortality using logistic regression.

RESULTS

Of 137,273 Veterans prescribed an opioid by a dentist, 0.1% and 1.1% were associated with opioid-related poisoning and mortality, respectively. There was no difference in opioid poisoning within 6 months for Veterans with opioid prescriptions >120 MME (aOR = 1.25 [CI: 0.89-1.78]), but poisoning decreased in Veterans prescribed opioids >3-days supply (aOR = 0.68 [CI: 0.49-0.96]). However, Veterans with opioids >120 MME were associated with higher odds of mortality within 6 months (aOR = 1.17 [95% CI: 1.05-1.32]) while there was no difference in prescriptions >3-days supply (aOR = 1.12 [CI: 0.99-1.25]).

CONCLUSION

Serious opioid-related adverse events were rare in Veterans and lower than other reports in the literature. Since nonopioid analgesics have superior efficacy for the treatment of acute dental pain, prescribing opioid alternatives may decrease opioid-related poisoning. Strategies for dentists to identify patients at high risk should be incorporated into the dental record.

Description of Opioid-involved Hospital Deaths that Do Not Have a Subsequent Autopsy

BACKGROUND

Surveillance systems rely on death records to monitor the most severe outcome of the opioid epidemic. However, few studies have linked data from hospital systems with death records to determine potential undercount of opioid-involved deaths occurring in hospitals. This study describes characteristics of decedents less likely to have an autopsy following an opioid-involved death in hospitals and estimates the resulting undercount.

METHODS

A probabilistic data linkage of hospital and medical examiner data involving 4,936 opioid-involved deaths among residents of Cook County, Illinois, US from 2016 to 2019. We included only hospital deaths that met a national case definition and presented with clinical signs of opioid overdose.

RESULTS

Decedents had higher odds of not having an autopsy if they were 50+ years, admitted to the hospital (aOR = 3.7: 2.1, 6.5), hospitalized for 4+ days (aOR = 2.2: 1.5, 3.1), and had a comorbid diagnosis of malignant cancer (aOR = 4.3: 1.8, 10.1). However, decedents exposed to heroin and synthetic opioids (aOR = 0.39: 0.28, 0.55), and concurrent exposure to stimulants (aOR = 0.44: 0.31, 0.64) were more likely to have an autopsy). Compared to estimates from the US Centers for Disease Control and Prevention (CDC), we observed undercounts of opioid overdose deaths ranging from 6% to 15%.

CONCLUSIONS

Surveillance systems may undercount decedents that do not meet the typical profile of those more likely to have an autopsy, particularly older patients with chronic health conditions. Our undercount estimate likely exists in addition to the estimated 20%-40% undercount reported elsewhere. See video abstract at, http://links.lww.com/EDE/B990 .

Sensitivity of health records for self-reported nonfatal drug and alcohol overdose

BACKGROUND AND OBJECTIVES

Public health surveillance for overdose sometimes depends on nonfatal drug overdoses recorded in health records. However, the proportion of total overdoses identified through health record systems is unclear. Comparison of overdoses from health records to those that are self-reported may provide insight on the proportion of nonfatal overdoses that are not identified.

METHODS

We conducted a cohort study linking survey data on overdose from a national survey of Veterans to United States Department of Veterans Affairs (VA) health records, including community care paid for by VA. Self-reported overdose in the prior 3 years was compared to diagnostic codes for overdoses and substance use disorders in the same time period.

RESULTS

The sensitivity of diagnostic codes for overdose, compared to self-report as a reference standard for this analysis, varied by substance: 28.1% for alcohol, 23.1% for sedatives, 12.0% for opioids, and 5.5% for cocaine. There was a notable concordance between substance use disorder diagnoses and self-reported overdose (sensitivity range 17.9%-90.6%).

DISCUSSION AND CONCLUSIONS

Diagnostic codes in health records may not identify a substantial proportion of drug overdoses. A health record diagnosis of substance use disorder may offer a stronger inference regarding the size of the population at risk. Alternatively, screening for self-reported overdose in routine clinical care could enhance overdose surveillance and targeted intervention.

SCIENTIFIC SIGNIFICANCE

This study suggests that diagnostic codes for overdose are insensitive. These findings support consideration of alternative approaches to overdose surveillance in public health.

Use of Diagnosis Codes to Find Blood Transfusion Adverse Events in Electronic Health Records

OBJECTIVE

Electronic health records (EHRs) and big data tools offer the opportunity for surveillance of adverse events (patient harm associated with medical care). We used International Classification of Diseases, Ninth Revision, codes in electronic records to identify known, and potentially novel, adverse reactions to blood transfusion.

METHODS

We used 49,331 adult admissions involving critical care at a major teaching hospital, 2001-2012, in the Medical Information Mart for Intensive Care III EHRs database. We formed a T (defined as packed red blood cells, platelets, or plasma) group of 21,443 admissions versus 25,468 comparison (C) admissions. The International Classification of Diseases, Ninth Revision, Clinical Modification , diagnosis codes were compared for T versus C, described, and tested with statistical tools.

RESULTS

Transfusion adverse events (TAEs) such as transfusion-associated circulatory overload (TACO; 12 T cases; rate ratio [RR], 15.61; 95% confidence interval [CI], 2.49-98) were found. There were also potential TAEs similar to TAEs, such as fluid overload disorder (361 T admissions; RR, 2.24; 95% CI, 1.88-2.65), similar to TACO. Some diagnoses could have been sequelae of TAEs, including nontraumatic compartment syndrome of abdomen (52 T cases; RR, 6.76; 95% CI, 3.40-14.9) possibly being a consequence of TACO.

CONCLUSIONS

Surveillance for diagnosis codes that could be TAE sequelae or unrecognized TAE might be useful supplements to existing medical product adverse event programs.

Pharmaceutical opioid poisonings in Victoria, Australia: Rates and characteristics of a decade of emergency department presentations among nine pharmaceutical opioids

BACKGROUND AND AIMS

Pharmaceutical opioids are a significant contributor to the global 'opioid crisis', yet few studies have comprehensively distinguished between opioid types. We measured whether a range of common pharmaceutical opioids varied in their contribution to the rates and characteristics of harm in a population-wide indicator of non-fatal overdose.

DESIGN

Retrospective observational study of emergency department (ED) patient care records in the Victorian Emergency Minimum Data set (VEMD), July 2009 to June 2019.

SETTING

Victoria, Australia.

CASES

ED presentations for non-fatal overdose related to pharmaceutical opioid use (n = 5403), where the specific pharmaceutical opioid was documented.

MEASUREMENTS

We compared harms across the nine individual pharmaceutical opioids most commonly sold, and considered where multiple opioids contributed to the overdose. We calculated supply-adjusted rates of ED presentations using Poisson regression and used multinomial logistic regression to compare demographic and clinical characteristics of presentations among nine distinct pharmaceutical opioids and a 10th category where multiple opioids were documented for the presentation.

FINDINGS

There were wide differences, up to 27-fold, between supply-adjusted rates of overdose. When considering presentations with sole opioids, the highest supply-adjusted overdose rates [per 100 000 oral morphine equivalents (OME); 95% confidence interval (CI)] were for codeine (OME = 0.078, 95% CI = 0.073-0.08) and oxycodone (OME =0.029, 95% CI = 0.027-0.030) and the lowest were for tapentadol (OME = 0.004, 95% CI = 0.003-0.006) and fentanyl (OME = 0.003, 95% CI = 0.002-0.004). These rates appeared related to availability rather than opioid potency. Most (62%) poisonings involved females. Codeine, oxycodone and tramadol were associated with younger presentations (respectively, 59.5%, 41.7% and 49.8% of presentations were 12-34 years old), and intentional self-harm (respectively 65.2%, 50.6%, and 52.8% of presentations). Relative to morphine, fentanyl [ 0.32 relative risk ratio (RRR)] and methadone ( 0.58 RRR) presentations were less likely to be coded as self-harm. Relative to morphine-buprenorphine, codeine, oxycodone and tramadol presentations were significantly more likely to be associated with the less urgent triage categories (respectively 2.18, 1.80, 1.52, 1.65 RRR).

CONCLUSIONS

In Victoria, Australia, rates and characteristics of emergency department presentations for pharmaceutical opioids show distinct variations by opioid type.

Emergency Department Drug Surveillance (EDDS) hospital's urinalysis results compared with expanded re-testing by an independent laboratory, a pilot study

INTRODUCTION

Most hospital urine toxicology screens detect a fixed, limited set of common substances. These tests are fast and accurate but may miss emerging trends in substance use in the community and clinical acumen alone is insufficient for identifying new substances.

METHODS

This prospective cohort study examined de-identified urine specimens obtained from patients visiting the Emergency Department (ED) at Prince George's Hospital Center (PGHC), between October 15, 2019 to November 6, 2019 and tested positive for one or more substances. The Emergency Department Drug Surveillance System (EDDS) collects quarterly exports from de-identified electronic health records (EHRs) containing urinalysis results for drug related ED visits. We performed a feasibility study of a new urine specimen submission by collecting a stratified sample of 151 urine specimens from PGHC ED patients. The specimens were tested for 240 drugs using liquid chromatography-tandem mass spectrometry (LC-MS/MS). This paper presents a comparison between the PGHC and expanded testing results.

RESULTS

The expanded urinalysis panel found more cocaine (37% vs. 20%; p < 0.01) and benzodiazepine positives (21% vs. 11%; p < 0.05) than would have been detected by the hospital screen. Additionally, the expanded toxicology panel identified fentanyl in 4-14% of the samples.

CONCLUSION

The EHR data submitted to EDDS from the hospital urine toxicology screen correctly identified hospital substance use patterns over the approximate 1 month study period. The expanded testing also uncovered drugs that the hospital might consider adding to their routine screen. EDDS is a feasible system for monitoring and confirming recent substance use trends among ED patients.

Defining indicators for drug overdose emergency department visits and hospitalisations in ICD-10-CM coded discharge data

Introduction

The drug overdose epidemic has worsened over the past decade; however, efforts have been made to better understand and track nonfatal overdoses using various data sources including emergency department and hospital admission data from billing and discharge files.

Methods and findings

The Centers for Disease Control and Prevention (CDC) has developed surveillance case definition guidance using standardised discharge diagnosis codes for public health practitioners and epidemiologists using lessons learnt from CDC’s funded recipients and the Council for State and Territorial Epidemiologists (CSTE) International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) Drug Poisoning Indicators Workgroup and General Injury ICD-10-CM Workgroup. CDC’s guidance was informed by health departments and CSTE’s workgroups and included several key aspects for assessing drug overdose in emergency department and hospitalisation discharge data. These include: (1) searching all diagnosis fields to identify drug overdose cases; (2) estimating drug overdose incidence using visits for initial encounter but excluding subsequent encounters and sequelae; (3) excluding underdosing and adverse effects from drug overdose incidence indicators; and (4) using codes T36–T50 for overdose surveillance. CDC’s guidance also suggests analysing intent separately for ICD-10-CM coding.

Conclusions

CDC’s guidance provides health departments a key tool to better monitor drug overdoses in their community. The implementation and validation of this standardised guidance across all CDC-funded health departments will be key to ensuring consistent and accurate reporting across all entities.

Using ICD-10-CM codes to detect illicit substance use: A comparison with retrospective self-report

BACKGROUND

Understanding whether International Classification of Disease, 10th Revision, Clinical Modification (ICD-10-CM) codes can be used to accurately detect substance use can inform their use in future surveillance and research efforts.

METHODS

Using 2015-2018 data from a retrospective cohort study of 602 safety-net patients prescribed opioids for chronic non-cancer pain, we calculated the sensitivity and specificity of using ICD-10-CM codes to detect illicit substance use compared to retrospective self-report by substance (methamphetamine, cocaine, opioids [heroin or non-prescribed opioid analgesics]), self-reported use frequency, and type of healthcare encounter.

RESULTS

Sensitivity of ICD-10-CM codes for detecting self-reported substance use was highest for methamphetamine (49.5 % [95 % confidence interval: 39.6-59.5 %]), followed by cocaine (44.4 % [35.8-53.2 %]) and opioids (36.3 % [28.8-44.2 %]); higher for participants who reported more frequent methamphetamine (intermittent use: 27.7 % [14.6-42.6 %]; ≥weekly use: 67.2 % [53.7-79.0 %]) and opioid use (intermittent use: 21.4 % [13.2-31.7 %]; ≥weekly use: 52.6 % [40.8-64.2 %]); highest for outpatient visits (methamphetamine: 43.8 % [34.1-53.8 %]; cocaine: 36.8 % [28.6-45.6 %]; opioids: 33.1 % [25.9-41.0 %]) and lowest for emergency department visits (methamphetamine: 8.6 % [4.0-15.6 %]; cocaine: 5.3 % [2.1-10.5 %]; opioids: 6.3 % [3.0-11.2 %]). Specificity was highest for methamphetamine (96.4 % [94.3-97.8 %]), followed by cocaine (94.0 % [91.5-96.0 %]) and opioids (85.0 % [81.3-88.2 %]).

CONCLUSIONS

ICD-10-CM codes had high specificity and low sensitivity for detecting self-reported substance use but were substantially more sensitive in detecting frequent use. ICD-10-CM codes to detect substance use, particularly those from emergency department visits, should be used with caution, but may be useful as a lower-bound population measure of substance use or for capturing frequent use among certain patient populations.

Health Care Utilization Due to Substance Abuse Among Homeless and Nonhomeless Children and Young Adults in New York

OBJECTIVE

Substance abuse, particularly among homeless youths, is a significant public health challenge in the United States. Detailed data about health care utilization resulting from this preventable behavior remain sparse. This study aimed to compare health care utilization rates related to substance abuse among homeless and nonhomeless youths.

METHODS

A secondary data analysis evaluated records of homeless and nonhomeless patients under age 25 with a primary diagnosis of substance abuse, identified in 2013 and 2014 New York Statewide Inpatient and Emergency Department (ED) Databases. Outcomes included ED visit rate, hospitalization rate, in-hospital mortality, cost, length of stay (LOS), intensive care unit (ICU) utilization, and revisit or readmission rate. Multivariable regression models with a year fixed effect and facility random effect were used to evaluate the association between homelessness and each outcome.

RESULTS

A total of 68,867 cases included hospitalization or an ED visit related to substance abuse (68,118 nonhomeless and 749 homeless cases). Rates of ED visits related to substance abuse were 9.38 and 4.96, while rates of hospitalizations related to substance abuse were 10.53 and 1.01 per 1,000 homeless and nonhomeless youths, respectively. Homeless patients were more likely to utilize and revisit the ICU, be hospitalized or readmitted, incur higher costs, and have longer LOS than nonhomeless youths (all p<0.01).

CONCLUSIONS

The hospitalization and ED visit rates related to substance abuse were 10 and two times higher among homeless youths compared with nonhomeless youths, respectively. Detailed observation is needed to clarify whether homeless youths receive high-quality care for substance abuse when necessary.

Interrupted time series analysis to evaluate the performance of drug overdose morbidity indicators shows discontinuities across the ICD-9-CM to ICD-10-CM transition

INTRODUCTION

On 1 October 2015, the USA transitioned from the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) to the International Classification of Diseases, 10th Revision (ICD-10-CM). Considering the major changes to drug overdose coding, we examined how using different approaches to define all-drug overdose and opioid overdose morbidity indicators in ICD-9-CM impacts longitudinal analyses that span the transition, using emergency department (ED) and hospitalisation data from six states' hospital discharge data systems.

METHODS

We calculated monthly all-drug and opioid overdose ED visit rates and hospitalisation rates (per 100 000 population) by state, starting in January 2010. We applied three ICD-9-CM indicator definitions that included identical all-drug or opioid-related codes but restricted the number of fields searched to varying degrees. Under ICD-10-CM, all fields were searched for relevant codes. Adjusting for seasonality and autocorrelation, we used interrupted time series models with level and slope change parameters in October 2015 to compare trend continuity when employing different ICD-9-CM definitions.

RESULTS

Most states observed consistent or increased capture of all-drug and opioid overdose cases in ICD-10-CM coded hospital discharge data compared with ICD-9-CM. More inclusive ICD-9-CM indicator definitions reduced the magnitude of significant level changes, but the effect of the transition was not eliminated.

DISCUSSION

The coding change appears to have introduced systematic differences in measurement of drug overdoses before and after 1 October 2015. When using hospital discharge data for drug overdose surveillance, researchers and decision makers should be aware that trends spanning the transition may not reflect actual changes in drug overdose rates.

Concurrent prescriptions for opioids and benzodiazepines and risk of opioid overdose: protocol for a retrospective cohort study using linked administrative data

INTRODUCTION

Opioid overdoses have increased substantially over the last 20 years, with over 400 000 deaths in North America. While opioid prescribing has been a target of research, benzodiazepine and opioid co-intoxication has emerged as a potential risk factor. Our aim was to assess the risk of opioid overdose associated with concurrent use of opioids and benzodiazepines relative to opioids alone.

METHODS AND ANALYSIS

A retrospective cohort study will be conducted using medical claims data from adult residents of Montréal, Canada. We will create a cohort of new users of opioids (ie, no opioid dispensations in prior year) in 2000-2014 from people with at least 2 years of continuous health insurance. Those with any diagnosis or hospitalisation for cancer or palliative care in the 2 years before their first opioid dispensation will be excluded. On each person-day of follow-up, exposure status will be classified into one of four mutually exclusive categories: (1) opioid-only, (2) benzodiazepine-only, (3) both opioid and benzodiazepine (concurrent use) or (4) neither. Opioid overdose will be measured using diagnostic codes documented in the hospital discharge abstract database, physician billing claims from emergency department visits and death records. Using a marginal structural Cox proportional hazards model, we will compare the hazard of overdose during intervals of concurrent opioid and benzodiazepine use to intervals of opioid use alone, adjusted for sociodemographics, medical and psychiatric comorbidities, and substance use disorders.

ETHICS AND DISSEMINATION

This study is approved by the McGill Faculty of Medicine Institutional Review Board and the Commission d'access à l'information (Québec privacy commission). Results will be relevant to clinicians, policymakers and other researchers interested in co-prescribing practices of opioids and benzodiazepines. Study findings will be disseminated at relevant conferences and published in biomedical and epidemiological peer-reviewed journals.

Using Natural Language Processing and Machine Learning to Identify Hospitalized Patients with Opioid Use Disorder

Opioid use disorder (OUD) represents a global public health crisis that challenges classic clinical decision making. As existing hospital screening methods are resource-intensive, patients with OUD are significantly under-detected. An automated and accurate approach is needed to improve OUD identification so that appropriate care can be provided to these patients in a timely fashion. In this study, we used a large-scale clinical database from Mass General Brigham (MGB; formerly Partners HealthCare) to develop an OUD patient identification algorithm, using multiple machine learning methods. Working closely with an addiction psychiatrist, we developed a set of hand-crafted rules for identifying information suggestive of OUD from free-text clinical notes. We implemented a natural language processing (NLP)-based classification algorithm within the Medical Text Extraction, Reasoning and Mapping System (MTERMS) tool suite to automatically label patients as positive or negative for OUD based on these rules. We further used the NLP output as features to build multiple machine learning and a neural classifier. Our methods yielded robust performance for classifying hospitalized patients as positive or negative for OUD, with the best performing feature set and model combination achieving an F1 score of 0.97. These results show promise for the future development of a real-time tool for quickly and accurately identifying patients with OUD in the hospital setting.

Trends in Opioid and Psychotropic Prescription in Pregnancy in the United States From 2001 to 2015 in a Privately Insured Population : A Cross-sectional Study

BACKGROUND

Opioid and psychotropic prescriptions are common during pregnancy. Little is known about coprescriptions of both medications in this setting.

OBJECTIVE

To describe opioid prescription among women who are prescribed psychotropics compared with women who are not.

DESIGN

Cross-sectional study.

SETTING

U.S. commercial insurance beneficiaries from MarketScan (2001 to 2015).

PARTICIPANTS

Pregnant women at 22 weeks' gestation or greater who were insured continuously for 3 months or more before pregnancy through delivery.

MEASUREMENTS

Opioid prescription, dosage thresholds (morphine milligram equivalents [MME] of ≥50/day and ≥90/day), number of opioid agents (≥2), and duration (≥30 days) among those with and without prescription of psychotropics, from 2011 to 2015.

RESULTS

Among 958 980 pregnant women, 10% received opioids only, 6% psychotropics only, and 2% opioids with coprescription of psychotropics. Opioid prescription was higher among women prescribed psychotropics versus those who were not (26.5% vs. 10.7%). From 2001 to 2015, psychotropic prescription overall increased from 4.4% to 7.6%, opioid prescription without coprescription of psychotropics decreased from 11.9% to 8.4%, and opioids with coprescription decreased from 28.1% to 22.0%. Morphine milligram equivalents of 50 or greater per day decreased for women with and without coprescription (29.6% to 17.3% and 22.8% to 18.5%, respectively); MME of 90 or greater per day also decreased in both groups (15.0% to 4.7% and 11.5% to 4.2%, respectively). Women prescribed opioids only were more likely to have an antepartum hospitalization compared with those with neither prescription, as were women with coprescription versus those prescribed psychotropics only. Compared with those prescribed opioids only, women with coprescriptions were more likely to exceed MME of 90 or greater per day and to be prescribed 2 or more opioid agents and for 30 days or longer. Number and duration of opioids increased with benzodiazepine and gabapentin coprescription.

LIMITATION

Inability to determine appropriateness of prescribing or overdose events.

CONCLUSION

Opioids are frequently coprescribed with psychotropic medication during pregnancy and are associated with antepartum hospitalization. A substantial proportion of pregnant women are prescribed opioids at doses that increase overdose risk and exceed daily recommendations.

PRIMARY FUNDING SOURCE

None.

Characteristics and outcomes of hospitalizations and readmissions for opioid dependence and overdose: nationally representative data

BACKGROUND

Despite the increasing rates of morbidity, mortality, and costs from the opioid addiction crisis, there is a paucity of literature on nationwide patterns of opioid abuse and dependence admissions and readmissions. We sought to investigate common comorbidities, readmission rates, and variables associated with readmission following index admission for opioid overdose or dependence. Methods: The 2013 Nationwide Readmission Database is a national database including data on more than 14 million US admissions. We used International Classification of Disease, Ninth Revision, Clinical Modification codes to identify index opioid abuse or dependence admissions, readmissions, and medical co-morbidities. We summarized all-cause readmission rates and reasons for readmission following index opioid dependence or overdose admission. We performed multivariable logistic regression, testing the association between characteristics of index admission and readmission. Results: 64,426 individuals were admitted for drug overdose or dependence during 2013. Of those, 30.1% were readmitted for all causes within one year and 8.7% were readmitted for opioid overdose or dependence within that year. The most common primary diagnoses on the readmission record were infection, kidney failure, drug related admission, and psychiatric admission. Predictors of readmission were smoking, male sex, younger age, alcohol, bipolar disorder, non-opioid drug use, admission to teaching hospitals in metropolitan areas, and discharge against medical advice. Conclusion: There is a high all-cause readmission rate following index admission for opioid overdose or dependence and a greater likelihood of readmission among young males with psychiatric comorbidities in metropolitan areas. Targeted interventions to address psychiatric comorbidities and transitions of care may be needed for the high-risk opioid dependence and overdose population.

Impact of reported NSAID "allergies" on opioid use disorder in back pain

BACKGROUND

It is crucial to identify patients at highest risk for opioid use disorder (OUD) and to address challenges in reducing opioid use. Reported nonsteroidal anti-inflammatory drug (NSAID) allergies may predispose to use of stronger pain medications and potentially to OUD.

OBJECTIVE

We sought to investigate the clinical impact of reported NSAID allergy on OUD in patients with chronic back pain.

METHODS

We conducted a retrospective study of adults receiving care at a tertiary health care system from January 1, 2013, to December 31, 2018. Back pain and OUD were identified using administrative data algorithms. We used propensity score matching and logistic regression to estimate the impact of self-reported NSAID adverse drug reactions (ADRs) on risk of OUD, adjusting for other relevant clinical information.

RESULTS

Of 47,114 patients with chronic back pain, 3,620 (7.7%) had a reported NSAID ADR. In an adjusted propensity score-matched analysis, patients with NSAID ADRs had higher odds (odds ratio, 1.34; 95% CI, 1.07-1.67) of developing OUD as compared with those without NSAID ADRs. Additional risk factors for OUD included younger age, male sex, Medicaid insurance, Medicare insurance, higher number of inpatient and outpatient visits in the previous year, and comorbid anxiety and depression. Patients with listed NSAID ADRs also had higher odds of a documented opioid prescription during the study period (odds ratio, 1.22; 95% CI, 1.11-1.34).

CONCLUSIONS

Adults with chronic back pain and reported NSAID ADRs are at a higher risk of developing OUD and receiving opioid analgesics, even after accounting for comorbidities and health care utilization. Allergy evaluation is critical for potential delabeling of patients with reported NSAID allergies and chronic pain.

Incidence of Treatment for Opioid Use Disorder Following Nonfatal Overdose in Commercially Insured Patients

IMPORTANCE

Timely initiation and referral to treatment for patients with opioid use disorder seen in the emergency department is associated with reduced mortality. It is not known how often commercially insured adults obtain follow-up treatment after nonfatal opioid overdose.

OBJECTIVE

To investigate the incidence of follow-up treatment following emergency department discharge after nonfatal opioid overdose and patient characteristics associated with receipt of follow-up treatment.

DESIGN, SETTING, AND PARTICIPANTS

A retrospective cohort study was conducted using an administrative claims database for a large US commercial insurer, from October 1, 2011, to September 30, 2016. Data analysis was performed from May 1, 2019, to September 26, 2019. Adult patients discharged from the emergency department after an index opioid overdose (no overdose in the preceding 90 days) were included. Patients with cancer and without continuous insurance enrollment were excluded.

MAIN OUTCOMES AND MEASURES

The primary outcome was follow-up treatment in the 90 days following overdose, defined as a combined outcome of claims for treatment encounters or medications for opioid use disorder (buprenorphine and naltrexone). Analysis was stratified by whether patients received treatment for opioid use disorder in the 90 days before the overdose. Logistic regression models were used to identify patient characteristics associated with receipt of follow-up treatment. Marginal effects were used to report the average adjusted probability and absolute risk differences (ARDs) in follow-up for different patient characteristics.

RESULTS

A total of 6451 patients were identified with nonfatal opioid overdose; the mean (SD) age was 45.0 (19.3) years, 3267 were women (50.6%), and 4676 patients (72.5%) reported their race as non-Hispanic white. A total of 1069 patients (16.6%; 95% CI, 15.7%-17.5%) obtained follow-up treatment within 90 days after the overdose. In adjusted analysis of patients who did not receive treatment before the overdose, black patients were half as likely to obtain follow-up compared with non-Hispanic white patients (ARD, -5.9%; 95% CI, -8.6% to -3.6%). Women (ARD, -1.7%; 95% CI, -3.3% to -0.5%) and Hispanic patients (ARD, -3.5%; 95% CI, -6.1% to -0.9%) were also less likely to obtain follow-up. For each additional year of age, patients were 0.2% less likely to obtain follow-up (95% CI, -0.3% to -0.1%).

CONCLUSIONS AND RELEVANCE

Efforts to improve the low rate of timely follow-up treatment following opioid overdose may seek to address sex, race/ethnicity, and age disparities.

Eye-Related Emergency Department Visits and The Opioid Epidemic: a 10-Year Analysis

PURPOSE

To describe the epidemiology of Emergency Department (ED) visits related to opioid abuse with primary ophthalmic diagnoses in the United States (US).

METHODS

This retrospective cross-sectional study used National ED Sample (NEDS) (2006-2015), a representative sample of all US EDs, to analyze and compare the epidemiology of primary ophthalmic diagnoses in opioid abusers and a control group of non-opioid users. National incidence and descriptive statistics were calculated for demographics and prevalent diagnoses. Multivariable logistic regression was used to compare outcomes between primary ophthalmic diagnoses in opioid and non-opioid abusers.

RESULTS

An estimated 10,617 visits had a primary ophthalmic diagnosis and an accompanying opioid abuse diagnosis, and the incidence increased from 0.2 in 2006 to 0.6 per 100,000 US population in 2015. Opioid abuse group had more adults (6,747:63.5%) and middle-aged (3,361:31.7%) patients, while in controls adults (7,905,003:40.4%) and children (4,068,534:20.8%) were affected more. Leading etiologies were similar: traumatic and infectious etiologies were most common; however, opioid abuse patients had more severe ophthalmic diagnoses such as orbital fractures (8.4%), orbital cellulitis (7.4%), globe injury (3.4%) and endophthalmitis (3.2%) compared to controls. Patients in the opioid abuse group were also more likely to be admitted (adjusted Odds Ratio [aOR], 28.38 [95% CI, 24.50-32.87]).

CONCLUSIONS

In the era of opioid crisis, an increase in ED visits with ophthalmic complaints is seen, with increasing direct and indirect costs on the healthcare system. More research is needed to establish causality and devise strategies to lower this burden.

Receipt of Addiction Treatment After Opioid Overdose Among Medicaid-Enrolled Adolescents and Young Adults

IMPORTANCE

Nonfatal opioid overdose may be a critical touch point when youths who have never received a diagnosis of opioid use disorder can be engaged in treatment. However, the extent to which youths (adolescents and young adults) receive timely evidence-based treatment following opioid overdose is unknown.

OBJECTIVE

To identify characteristics of youths who experience nonfatal overdose with heroin or other opioids and to assess the percentage of youths receiving timely evidence-based treatment.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study used the 2009-2015 Truven-IBM Watson Health MarketScan Medicaid claims database from 16 deidentified states representing all US census regions. Data from 4 039 216 Medicaid-enrolled youths aged 13 to 22 years were included and were analyzed from April 20, 2018, to March 21, 2019.

EXPOSURES

Nonfatal incident and recurrent opioid overdoses involving heroin or other opioids.

MAIN OUTCOMES AND MEASURES

Receipt of timely addiction treatment (defined as a claim for behavioral health services, for buprenorphine, methadone, or naltrexone prescription or administration, or for both behavioral health services and pharmacotherapy within 30 days of incident overdose). Sociodemographic and clinical characteristics associated with receipt of timely treatment as well as with incident and recurrent overdoses were also identified.

RESULTS

Among 3791 youths with nonfatal opioid overdose, 2234 (58.9%) were female, and 2491 (65.7%) were non-Hispanic white. The median age was 18 years (interquartile range, 16-20 years). The crude incident opioid overdose rate was 44.1 per 100 000 person-years. Of these 3791 youths, 1001 (26.4%) experienced a heroin overdose; the 2790 (73.6%) remaining youths experienced an overdose involving other opioids. The risk of recurrent overdose among youths with incident heroin involvement was significantly higher than that among youths with other opioid overdose (adjusted hazard ratio, 2.62; 95% CI, 2.14-3.22), and youths with incident heroin overdose experienced recurrent overdose at a crude rate of 20 700 per 100 000 person-years. Of 3606 youths with opioid-related overdose and continuous enrollment for at least 30 days after overdose, 2483 (68.9%) received no addiction treatment within 30 days after incident opioid overdose, whereas only 1056 youths (29.3%) received behavioral health services alone, and 67 youths (1.9%) received pharmacotherapy. Youths with heroin overdose were significantly less likely than youths with other opioid overdose to receive any treatment after their overdose (adjusted odds ratio, 0.64; 95% CI, 0.49-0.83).

CONCLUSIONS AND RELEVANCE

After opioid overdose, less than one-third of youths received timely addiction treatment, and only 1 in 54 youths received recommended evidence-based pharmacotherapy. Interventions are urgently needed to link youths to treatment after overdose, with priority placed on improving access to pharmacotherapy.

ICD-10-CM-Based Definitions for Emergency Department Opioid Poisoning Surveillance: Electronic Health Record Case Confirmation Study

OBJECTIVES

Valid opioid poisoning morbidity definitions are essential to the accuracy of national surveillance. The goal of our study was to estimate the positive predictive value (PPV) of case definitions identifying emergency department (ED) visits for heroin or other opioid poisonings, using billing records with International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes.

METHODS

We examined billing records for ED visits from 4 health care networks (12 EDs) from October 2015 through December 2016. We conducted medical record reviews of representative samples to estimate the PPVs and 95% confidence intervals (CIs) of (1) first-listed heroin poisoning diagnoses (n = 398), (2) secondary heroin poisoning diagnoses (n = 102), (3) first-listed other opioid poisoning diagnoses (n = 452), and (4) secondary other opioid poisoning diagnoses (n = 103).

RESULTS

First-listed heroin poisoning diagnoses had an estimated PPV of 93.2% (95% CI, 90.0%-96.3%), higher than secondary heroin poisoning diagnoses (76.5%; 95% CI, 68.1%-84.8%). Among other opioid poisoning diagnoses, the estimated PPV was 79.4% (95% CI, 75.7%-83.1%) for first-listed diagnoses and 67.0% (95% CI, 57.8%-76.2%) for secondary diagnoses. Naloxone was administered in 867 of 1055 (82.2%) cases; 254 patients received multiple doses. One-third of all patients had a previous drug poisoning. Drug testing was ordered in only 354 cases.

CONCLUSIONS

The study findings suggest that heroin or other opioid poisoning surveillance definitions that include multiple diagnoses (first-listed and secondary) would identify a high percentage of true-positive cases.

Comparing the predictive capability of self-report and medically-verified non-fatal overdose in adults released from prison: A prospective data linkage study

BACKGROUND

Self-reported non-fatal overdose (NFOD) is a predictor of future overdose and is often used to target overdose prevention for people released from prison. However, the level of agreement between self-reported and medically-verified NFOD history remains unknown. This study aimed to determine the agreement between, and predictive value of, self-reported and medically-verified history of NFOD in people recently released from prison.

METHODS

Pre-release baseline survey data from 1307 adults in prison surveyed from 2008 to 2010 in Queensland, Australia were linked to ambulance, emergency department, and hospital records. We compared the agreement of self-reported NFOD history in the baseline survey and medically-verified NFOD ascertained through linked medical data. Unadjusted and adjusted regression models were used to determine the association between self-reported and medically verified NFOD history and medically-verified NFOD after release from prison.

RESULTS

224 (19 %) participants self-reported NFOD history only, 75 (5 %) had medically-verified NFOD history only, and 56 (4 %) both self-reported and had medically-verified NFOD history. Compared to those with no NFOD history, those who self-reported and had a medical history of NFOD (adjusted hazard ratio (AHR) 6.1, 95 %CI 3.1-11.9), those with a medical history only (AHR 3.4, 95 %CI 1.7-7.0), and those who self-reported only (AHR 1.8, 95 %CI 1.0-3.5) were at increased risk of medically-verified NFOD after release from prison.

CONCLUSIONS

Relying on self-report of NFOD is likely to miss people at increased risk of future NFOD, many of whom could be identified through medical records. Wherever possible, data related to NFOD should be triangulated from multiple sources.

Validation of an administrative claims coding algorithm for serious opioid overdose: A medical chart review

PURPOSE

A standardized definition for serious opioid overdose has not been clearly established for disease surveillance or assessing the impact of risk mitigation strategies. The purpose of this study was to use medical chart review to clinically validate a claims-based algorithm to identify serious opioid overdose events.

METHODS

The algorithm for serious opioid overdose required an opioid poisoning or external cause ICD-9-CM code occurring within 1 day of (a) an adverse effect code for serious central nervous system or respiratory depression or (b) a mechanical ventilation or critical care CPT code. The claims coding algorithm identified a sample of 145 individuals 18 years or older among patients that presented to the emergency department of two large hospitals in metropolitan Atlanta, Georgia from January 2014 to August 2015. Claims-defined cases were evaluated against rigorous clinical definitions for serious opioid overdose using (a) literature-based criteria for typical clinical manifestations of opioid overdose and/or (b) clinical response to the opioid-specific reversal agent naloxone. The positive predictive value (PPV) for a serious opioid overdose was calculated as the percentage of clinically confirmed cases (definite or probable).

RESULTS

Among 140 evaluable claims-defined cases, 107 fulfilled clinical criteria for a serious opioid overdose [95 definite and 12 probable; PPV of 76.4% (95% CI 69.4%, 83.5%)]. Among 30 nonconfirmed cases, 20 were polyintoxications involving one or more nonopioid psychoactive agents.

CONCLUSIONS

An administrative claims coding algorithm for serious opioid overdose had high clinical predictive performance in a medical chart review.

Do electronic health record prompts increase take-home naloxone administration for emergency department patients after an opioid overdose?

BACKGROUND AND AIMS

Distribution of take-home naloxone (THN) to emergency department (ED) patients who have survived an opioid overdose (OD) could reduce future opioid mortality, but is not commonly performed. We examined whether electronic health record (EHR) prompts provided to ED physicians when discharging a patient after an OD could improve THN distribution.

DESIGN

Interrupted time-series analysis to compare the percentage of OD patients who received THN during the 11 months before and after implementation of an EHR prompt on 18 June 2017.

SETTING AND PARTICIPANTS

A total of 3492 adult patients with diagnoses of OD discharged from nine EDs in a single health system in Western Pennsylvania from July 2016 to April 2018.

INTERVENTION AND COMPARATOR

The EHR prompt was triggered by the presence of specific terms in the nurse's initial assessment note. The EHR displayed a pop-up window during the ED physician discharge process asking the physician to consider prescribing or providing naloxone to the patient. The comparator was 'no EHR prompt'.

MEASUREMENTS

Measurements were based on standard criteria from ICD diagnostic codes and chief complaint keywords.

FINDINGS

In July 2016, 16.3% [95% confidence interval (CI) = 14.0, 18.5] of OD patients received THN, which decreased every month through June 2017 by 1.2% (P < 0.0001, 95% CI = 0.8,1.7). For each month post-EHR prompt there was an increase of 2.8% of OD patients receiving THN (P < 0.001, 95% CI = 2.0, 3.5). No increases occurred in the ED with the highest pre-EHR prompt THN distribution. Rates of THN distribution varied by patient age and race prior to, but not after, implementation of EHR prompts.

CONCLUSIONS

Electronic health record prompts are associated with increased take-home naloxone distribution for emergency department patients discharged after opioid overdoses.

Disparities in Access to Mental Health Services Among Patients Hospitalized for Deliberate Drug Overdose

OBJECTIVE

The authors examined patient and hospitalization characteristics associated with receiving a mental health assessment and disposition to an inpatient psychiatric facility among patients hospitalized for deliberate drug overdose.

METHODS

This retrospective analysis of 2012-2013 South Carolina all-payer data included adults ages 18-64 with at least one inpatient admission for a primary diagnosis of deliberate illicit or pharmaceutical drug overdose (N=2,686). Outcomes were receipt of a mental health assessment and disposition to an inpatient psychiatric facility. Multivariable logistic regression models were used to estimate the effects of patient and hospitalization characteristics on study outcomes.

RESULTS

Non-Hispanic blacks and people of other races-ethnicities were less likely than non-Hispanic whites to receive a mental health assessment (non-Hispanic blacks, adjusted odds ratio [AOR]=0.52, 95% CI=0.34-0.81; other races-ethnicities, AOR=0.24, 95% CI=0.12-0.49). Non-Hispanic blacks were also less likely than non-Hispanic whites to be discharged to an inpatient psychiatric facility than to home (AOR=0.60, 95% CI=0.47-0.77). Compared with persons without insurance, those with insurance, except those with Medicaid, were more likely to be discharged to an inpatient psychiatric facility than to home (Medicare, AOR=3.06, 95% CI=2.36-3.96; private, AOR=2.78, 95% CI=2.23-3.47; other, AOR=7.58, 95% CI=4.21-13.6).

CONCLUSIONS

Non-Hispanic white race-ethnicity and having insurance were predictive of receipt of a mental health assessment and disposition to an inpatient psychiatric facility among patients hospitalized for deliberate drug overdose. Study findings can inform clinical strategies and interventions aimed at reducing mental health care disparities among populations who are vulnerable to overdose or suicide.

Opioid Abuse or Dependence Increases 30-day Readmission Rates after Major Operating Room Procedures: A National Readmissions Database Study

BACKGROUND

Although opioids remain the standard therapy for the treatment of postoperative pain, the prevalence of opioid misuse is rising. The extent to which opioid abuse or dependence affects readmission rates and healthcare utilization is not fully understood. It was hypothesized that surgical patients with a history of opioid abuse or dependence would have higher readmission rates and healthcare utilization.

METHODS

A retrospective cohort analysis was performed of patients undergoing major operating room procedures in 2013 and 2014 using the National Readmission Database. Patients with opioid abuse or dependence were identified using International Classification of Diseases codes. The primary outcome was 30-day hospital readmission rate. Secondary outcomes included hospital length of stay and estimated hospital costs.

RESULTS

Among the 16,016,842 patients who had a major operating room procedure whose death status was known, 94,903 (0.6%) had diagnoses of opioid abuse or dependence. After adjustment for potential confounders, patients with opioid abuse or dependence had higher 30-day readmission rates (11.1% vs. 9.1%; odds ratio 1.26; 95% CI, 1.22 to 1.30), longer mean hospital length of stay at initial admission (6 vs. 4 days; P < 0.0001), and higher estimated hospital costs during initial admission ($18,528 vs. $16,617; P < 0.0001). Length of stay was also higher at readmission (6 days vs. 5 days; P < 0.0001). Readmissions for infection (27.0% vs. 18.9%; P < 0.0001), opioid overdose (1.0% vs. 0.1%; P < 0.0001), and acute pain (1.0% vs. 0.5%; P < 0.0001) were more common in patients with opioid abuse or dependence.

CONCLUSIONS

Opioid abuse and dependence are associated with increased readmission rates and healthcare utilization after surgery.

VISUAL ABSTRACT

An online visual overview is available for this article at http://links.lww.com/ALN/B704.

User-centred clinical decision support to implement emergency department-initiated buprenorphine for opioid use disorder: protocol for the pragmatic group randomised EMBED trial

INTRODUCTION

The goal of this trial is to determine whether implementation of a user-centred clinical decision support (CDS) system can increase adoption of initiation of buprenorphine (BUP) into the routine emergency care of individuals with opioid use disorder (OUD).

METHODS

A pragmatic cluster randomised trial is planned to be carried out in 20 emergency departments (EDs) across five healthcare systems over 18 months. The intervention consists of a user-centred CDS integrated into ED clinician electronic workflow and available for guidance to: (1) determine whether patients presenting to the ED meet criteria for OUD, (2) assess withdrawal symptoms and (3) ascertain and motivate patient willingness to initiate treatment. The CDS guides the ED clinician to initiate BUP and facilitate follow-up. The primary outcome is the rate of BUP initiated in the ED. Secondary outcomes are: (1) rates of receiving a referral, (2) fidelity with the CDS and (3) rates of clinicians providing any ED-initiated BUP, referral for ongoing treatment and receiving Drug Addiction Act of 2000 training. Primary and secondary outcomes will be analysed using generalised linear mixed models, with fixed effects for intervention status (CDS vs usual care), prespecified site and patient characteristics, and random effects for study site.

ETHICS AND DISSEMINATION

The protocol has been approved by the Western Institutional Review Board. No identifiable private information will be collected from patients. A waiver of informed consent was obtained for the collection of data for clinician prescribing and other activities. As a minimal risk implementation study of established best practices, an Independent Study Monitor will be utilised in place of a Data Safety Monitoring Board. Results will be reported in ClinicalTrials.gov and published in open-access, peer-reviewed journals, presented at national meetings and shared with the clinicians at participating sites via a broadcast email notification of publications.

TRIAL REGISTRATION NUMBER

NCT03658642; Pre-results.

Development and evaluation of a standardized research definition for opioid overdose outcomes

Background: Increasing epidemiologic and intervention research is being conducted on opioid overdose, a serious and potentially fatal outcome. However, there is little consensus on how to verify opioid overdose outcomes for research purposes. To ensure reproducibility, minimize misclassification, and permit data harmonization across studies, standardized and consistent overdose definitions are needed. The aims were to develop a case criteria classification scheme based on information commonly available in medical records and to compare it with reviewing physician clinical impression and simple encounter documentation. Methods: In 2 large health systems, we developed a case criteria classification scheme for opioid overdose based on prior literature, expert opinion, and pilot testing with sample medical records. We then identified emergency department and hospital encounters (n = 259) with at least 1 International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code representing a broad range of opioid and non-opioid related poisonings. Physicians conducted structured medical record reviews to identify the proposed case criteria and generate a clinical impression, and trained abstractors verified documentation. We then compared the case criteria classification scheme with clinical impression and encounter documentation. Results: We developed a quantitative opioid overdose case criteria classification scheme that included 3 sets of major criteria and 9 minor criteria (supporting documentation). For the encounters identified using poisoning codes, the proportion verified as opioid overdoses using the case criteria classification scheme, clinical impression, and encounter documentation ranged from 50.4% to 52.7% at one site and 55.5% to 67.2% at the second site. Discrepancies across approaches and sites related to differences in available records and documentation of clinical signs of overdose. Conclusions: We propose a novel case criteria classification scheme for opioid overdose that could be used to rigorously and consistently define overdose across multiple research settings. However, prior to widespread use, further refinement and validation are needed.

National Trends in Hospitalizations for Stroke Associated With Infective Endocarditis and Opioid Use Between 1993 and 2015

Background and Purpose- There has been a recent sharp rise in opioid-related deaths in the United States. Intravenous opioid use can lead to infective endocarditis (IE) which can result in stroke. There are scant data on recent trends in this neurological complication of opioid abuse. We hypothesized that increasing opioid abuse has led to a higher incidence of stroke associated with IE and opioid use. Methods- We used the 1993 to 2015 releases of the National Inpatient Sample and validated International Classification of Diseases, Ninth Revision, Clinical Modification codes ( ICD-9-CM) to identify hospitalizations with the combination of opioid abuse, IE, and stroke (defined as ischemic stroke, intracerebral hemorrhage, or subarachnoid hemorrhage). Survey weights provided by the National Inpatient Sample were used to calculate nationally representative estimates and population estimates from the United States. Census data were used to calculate annual hospitalization rates per 10 million person-years. Joinpoint regression was used to assess trends. Results- From 1993 through 2015, there were 5283 hospitalizations with stroke associated with IE and opioid use. Across this period, the rate of such hospitalizations increased from 2.4 (95% CI, 0.5-4.3) to 18.8 (95% CI, 14.4-23.3) per 10 million US residents. Joinpoint regression detected 2 segments: no significant change in the hospitalization rate was apparent from 1993 to 2008 (annual percentage change, 1.9%; 95% CI, -2.2% to 6.1%), and then rates significantly increased from 2008 to 2015 (annual percentage change, 20.3%; 95% CI, 10.5%-30.9%), most dramatically in non-Hispanic white patients in the Northeastern and Southern United States. Conclusions- US hospitalization rates for stroke associated with IE and opioid use were stable for ≈2 decades but then sharply increased starting in 2008, coinciding with the emergence of the opioid epidemic.

Urban-rural variation in the socioeconomic determinants of opioid overdose

BACKGROUND

Prescription opioid overdose (POD) and heroin overdose (HOD) rates have quadrupled since 1999. Community-level socioeconomic characteristics are associated with opioid overdoses, but whether this varies by urbanicity is unknown.

METHODS

In this serial cross-sectional study of zip codes in 17 states, 2002-2014 (n = 145,241 space-time units), we used hierarchical Bayesian Poisson space-time models to analyze the association between zip code-level socioeconomic features (poverty, unemployment, educational attainment, and income) and counts of POD or HOD hospital discharges. We tested multiplicative interactions between each socioeconomic feature and zip code urbanicity measured with Rural-Urban Commuting Area codes.

RESULTS

Percent in poverty and of adults with ≤ high school education were associated with higher POD rates (Rate Ratio [RR], 5% poverty: 1.07 [95% credible interval: 1.06-1.07]; 5% low education: 1.02 [1.02-1.03]), while median household income was associated with lower rates (RR, $10,000: 0.88 [0.87-0.89]). Urbanicity modified the association between socioeconomic features and HOD. Poverty and unemployment were associated with increased HOD in metropolitan areas (RR, 5% poverty: 1.12 [1.11-1.13]; 5% unemployment: 1.04 [1.02-1.05]), and median household income was associated with decreased HOD (RR, $10,000: 0.88 [0.87-0.90]). In rural areas, low educational attainment alone was associated with HOD (RR, 5%: 1.09 [1.02-1.16]).

CONCLUSIONS

Regardless of urbanicity, elevated rates of POD were found in more economically disadvantaged zip codes. Economic disadvantage played a larger role in HOD in urban than rural areas, suggesting rural HOD rates may have alternative drivers. Identifying social determinants of opioid overdoses is particularly important for creating effective population-level interventions.

Drug overdose in the ED: a record linkage study examining emergency department ICD-10 coding practices in a cohort of people who inject drugs

Background

Drug overdose is a leading cause of mortality and morbidity amongst people who inject drugs (PWID). Drug overdose surveillance typically relies on the International Classification of Diseases (ICD-10) coding system, however its real world utilisation and the implications for surveillance have not been well characterised. This study examines the patterns of ICD-10 coding pertaining to drug overdoses within emergency departments for a cohort of known PWID.

Methods

Cohort data from 688 PWID was linked to statewide emergency department administrative data between January 2008 and June 2013. ICD-10 diagnostic codes pertaining to poisonings by drugs, medicaments and biological substances (T-codes T36-T50) as well as mental and behavioural disorders due to psychoactive substance use (F-codes F10-F19) were examined.

Results

There were 449 unique ED presentations with T or F code mentions contributed by 168 individuals. Nearly half of the T and F codes used were non-specific and did not identify either a drug class (n = 160, 36%) or clinical reaction (n = 46, 10%) and 8% represented withdrawal states. T and F codes could therefore be used to reasonably infer an illicit drug overdose in only 42% (n = 188) of cases. Majority of presentations with T or F overdose codes recorded only one diagnostic code per encounter (83%) and representing multiple-drug overdose (F19.- = 18%) or unidentified substances (T50.9 = 17%) using a single, broad diagnostic code was common.

Conclusions

Reliance on diagnoses alone when examining ED data will likely significantly underestimate incidence of specific drug overdose due to frequent use of non-specific ICD-10 codes and the use of single diagnostic codes to represent polysubstance overdose. Measures to improve coding specificity should be considered and further work is needed to determine the best way to use ED data in overdose surveillance.

When DAWN went dark: Can the Nationwide Emergency Department Sample (NEDS) fill the surveillance gap left by the discontinued Drug Abuse Warning Network (DAWN)?

BACKGROUND

This study assessed whether the Nationwide Emergency Department Sample (NEDS) could reliably fill the national drug surveillance gap caused by the discontinuation of the Drug Abuse Warning Network (DAWN).

METHODS

Estimates of the drug-related emergency department (ED) visits derived from DAWN (2004-2011) and NEDS (2006-2013). Estimates of the underlying reason for the drug-related ED visit, patient characteristics, and the specific drugs involved were compared for 2011, the most recent overlapping data year in DAWN and NEDS. Trends in ED visits for major drugs of abuse were then compared over the period 2004-2013.

RESULTS

In 2011, DAWN and NEDS produced statistically similar estimates of the overall number of drug-related ED visits (5.1 vs. 4.9 million) and those involving drug misuse or abuse (2.65 vs. 2.77 million). Among the latter, estimates by gender, age group, and patient disposition were generally consistent across data systems, suggesting that NEDS and DAWN samples draw from a similar population. Main analyses reveal statistically similar estimates across data systems in both levels and trends for cocaine, amphetamines, and narcotic pain relievers. In contrast, the number of ED visits for sedatives and heroin was significantly undercounted in NEDS, whereas marijuana-related ED visits were undercounted in DAWN.

CONCLUSIONS

This study demonstrates the utility of NEDS for conducting post-DAWN drug surveillance. Because NEDS cannot provide targeted surveillance of certain established (e.g., heroin) and emerging (e.g., fentanyl) drugs, however, it is critical that a data system that employs medical record-based reviews be implemented to augment the known weaknesses of NEDS.

Assessing the effectiveness of New York's 911 Good Samaritan Law-Evidence from a natural experiment

BACKGROUND

Drug overdose is the leading cause of accidental death in the United States. Nationally, opioids are the primary drugs associated with accidental overdoses. In response to increasing overdose deaths, 40 states and the District of Columbia have enacted Good Samaritan Laws (GSLs). Generally, these policies attempt to encourage witnesses or those experiencing an overdose to call 911 by providing limited immunity from arrest, charge and/or prosecution of possession of narcotics. The aim of the current study is to evaluate the effectiveness of New York State's 911 GSL.

METHODS

We exploit a difference in state law between New York State, where the policy was adopted in 2011, and New Jersey, where the policy was not adopted until 2013, to provide a reasonable comparison condition. We examine variation in accidental opioid overdose emergency department visits and inpatient admissions from 2010 to 2012 across 270 hospitals in New York and New Jersey at the quarterly level controlling for hospital fixed effects and time trends using State Emergency Department Databases (SEDD) and State Inpatient Databases (SID).

RESULTS

Accidental opioid overdose emergency department visits and inpatient hospital admissions were increasing in both New York and New Jersey. After the enactment of New York's 911 GSL, emergency department visits and inpatient hospital admissions for accidental heroin overdoses increased differently in New York and New Jersey, with an incident rate ratio (IRR) of 1.34 (95% CI = 1.00, 1.86). The results were inconclusive for accidental non-heroin opioid overdoses (IRR = 0.98, 95% CI = 0.86, 1.13).

CONCLUSIONS

Accidental heroin overdose emergency department visits and inpatient hospital admissions increased in New York State after the enactment of the 911 GSL, consistent with the intended effect of the GSL. Preliminary evidence suggests that either persons who use heroin and/or those around them were impacted by the policy change.

Mental Health-related Emergency Department Visits Associated With Cannabis in Colorado

BACKGROUND

Cannabis legalization in Colorado resulted in increased cannabis-associated health care utilization. Our objective was to examine cooccurrence of cannabis and mental health diagnostic coding in Colorado emergency department (ED) discharges and replicate the study in a subpopulation of ED visits where cannabis involvement and psychiatric diagnosis were confirmed through medical review.

METHODS

We collected statewide ED International Classification of Diseases, 9th Revision, Clinical Modification diagnoses from the Colorado Hospital Association and a subpopulation of ED visits from a large, academic hospital from 2012 to 2014. Diagnosis codes identified visits associated with mental health and cannabis. Codes for mental health conditions and cannabis were confirmed by manual records review in the academic hospital subpopulation. Prevalence ratios (PRs) of mental health ED discharges were calculated to compare cannabis-associated visits to those without cannabis. Rates of mental health and cannabis-associated ED discharges were examined over time.

RESULTS

Statewide data demonstrated a fivefold higher prevalence of mental health diagnoses in cannabis-associated ED visits (PR = 5.35, 95% confidence interval [CI], 5.27-5.43) compared to visits without cannabis. The hospital subpopulation supported this finding with a fourfold higher prevalence of psychiatric complaints in cannabis attributable ED visits (PR = 4.87, 95% CI = 4.36-5.44) compared to visits not attributable to cannabis. Statewide rates of ED visits associated with both cannabis and mental health significantly increased from 2012 to 2014 from 224.5 to 268.4 per 100,000 (p < 0.0001).

CONCLUSIONS

In Colorado, the prevalence of mental health conditions in ED visits with cannabis-associated diagnostic codes is higher than in those without cannabis. There is a need for further research determining if these findings are truly attributed to cannabis or merely coincident with concurrent increased use and availability.

Features of prescription drug monitoring programs associated with reduced rates of prescription opioid-related poisonings

BACKGROUND

The United States is in the midst of an opioid epidemic. In addition to other system-level interventions, all states have responded during the crisis by implementing prescription drug monitoring programs (PDMPs). This study examines associations between specific administrative features of PDMPs and changes in the risk of prescription opioid-related poisoning (RxORP) over time.

METHODS

This longitudinal, observational study utilized a 'natural experiment' design to assess associations between PDMP features and risk of RxORP in a nationally-representative population of privately-insured adults from 2004 to 2014. Administrative health claims data were used to identify inpatient hospital admissions and emergency department visits related to RxORP. Generalized estimating equation Poisson regression models were used to examine associations between specific PDMP features and changes in relative risk (RR) of RxORP over time.

RESULTS

In adjusted analyses, states without PDMPs experienced an average annual increase in the rate of RxORP of 9.51% over the study period, while states with operational PDMPs experienced an average annual increase of 3.17%. The increase in RR of RxORP over time in states with operational PDMPs was significantly less than increases in states without PDMPs. States with specific features, including those that monitored more schedules or required more frequent data reporting, experienced stronger protective effects on the RR of RxORP over time.

CONCLUSION

This study examined associations between specific PDMP features and RxORP rates in a nationally-representative population of privately-insured adults. Results of this study may be used as empirical evidence to guide PDMP best practices.

A statewide evaluation of seven strategies to reduce opioid overdose in North Carolina

BACKGROUND

In response to increasing opioid overdoses, US prevention efforts have focused on prescriber education and supply, demand and harm reduction strategies. Limited evidence informs which interventions are effective. We evaluated Project Lazarus, a centralised statewide intervention designed to prevent opioid overdose.

METHODS

Observational intervention study of seven strategies. 74 of 100 North Carolina counties implemented the intervention. Dichotomous variables were constructed for each strategy by county-month. Exposure data were: process logs, surveys, addiction treatment interviews, prescription drug monitoring data. Outcomes were: unintentional and undetermined opioid overdose deaths, overdose-related emergency department (ED) visits. Interrupted time-series Poisson regression was used to estimate rates during preintervention (2009-2012) and intervention periods (2013-2014). Adjusted IRR controlled for prescriptions, county health status and time trends. Time-lagged regression models considered delayed impact (0-6 months).

RESULTS

In adjusted immediate-impact models, provider education was associated with lower overdose mortality (IRR 0.91; 95% CI 0.81 to 1.02) but little change in overdose-related ED visits. Policies to limit ED opioid dispensing were associated with lower mortality (IRR 0.97; 95% CI 0.87 to 1.07), but higher ED visits (IRR 1.06; 95% CI 1.01 to 1.12). Expansions of medication-assisted treatment (MAT) were associated with increased mortality (IRR 1.22; 95% CI 1.08 to 1.37) but lower ED visits in time-lagged models.

CONCLUSIONS

Provider education related to pain management and addiction treatment, and ED policies limiting opioid dispensing showed modest immediate reductions in mortality. MAT expansions showed beneficial effects in reducing ED-related overdose visits in time-lagged models, despite an unexpected adverse association with mortality.

Mortality Quadrupled Among Opioid-Driven Hospitalizations, Notably Within Lower-Income And Disabled White Populations

Hospitals play an important role in caring for patients in the current opioid crisis, but data on the outcomes and composition of opioid-driven hospitalizations in the United States have been lacking. Nationally representative all-payer data for the period 1993-2014 from the National Inpatient Sample were used to compare the mortality rates and composition of hospitalizations with opioid-related primary diagnoses and those of hospitalizations for other drugs and for all other causes. Mortality among opioid-driven hospitalizations increased from 0.43 percent before 2000 to 2.02 percent in 2014, an average increase of 0.12 percentage points per year relative to the mortality of hospitalizations due to other drugs-which was unchanged. While the total volume of opioid-driven hospitalizations remained relatively stable, it shifted from diagnoses mostly involving opioid dependence or abuse to those centered on opioid or heroin poisoning (the latter have higher case fatality rates). After 2000, hospitalizations for opioid/heroin poisoning grew by 0.01 per 1,000 people per year, while hospitalizations for opioid dependence or abuse declined by 0.01 per 1,000 people per year. Patients admitted for opioid/heroin poisoning were more likely to be white, ages 50-64, Medicare beneficiaries with disabilities, and residents of lower-income areas. As the United States combats the opioid epidemic, efforts to help hospitals respond to the increasing severity of opioid intoxication are needed, especially in vulnerable populations.

US regional and demographic differences in prescription opioid and heroin-related overdose hospitalizations

BACKGROUND

US opioid overdose death rates have increased between 2000 and 2014. While, the increase in prescription opioid use has been linked to the increase in heroin use, there are reasons to view this relationship as a partial explanation for the recent increase in heroin-related harms. This study documents the differences in trends in prescription opioid overdose-related (POD) and heroin overdose-related (HOD) hospitalizations.

METHODS

Data come from the National Inpatient Sample (NIS) for the years 2000 through 2014. POD and HOD hospitalizations were abstracted from ICD-9 codes. Rates of POD and HOD by census region and census division were constructed along with separate rates for age and race. Regression analysis analyzing trends across region were estimated along with graphs for documenting differences in POD and HOD rates.

RESULTS

POD hospitalization rates were highest in the South and lowest in the Northeast. HOD hospitalization rates were highest in the Northeast region and grew the fastest in the Midwest. There was statistically significant heterogeneity in HOD trends but not POD trends across the four regions between 2000 and 2014. Between 2012 and 2014 POD rates decreased in eight of the nine census divisions, with only New England showing an increase. HOD hospitalization rates increased in all nine census divisions between 2012 and 2014. Both POD and HOD rates show different demographic patterns across the nine census divisions.

CONCLUSION

Comparing POD and HOD hospitalization trends reveals significant disparities in geographic as well as demographic distributions. These epidemics are evolving and the simple opioid-to-heroin transition story is both supported and challenged by this paper. The opioid pill, heroin and fentanyl crises are intertwined yet increasingly have drivers and outcomes that support examining them as distinct. Addressing these complex and interrelated epidemics will require innovative public health research and interventions which need to consider local and regional contexts.

Rates and Risk Factors for Opioid Dependence and Overdose after Urological Surgery

PURPOSE

Effective pain management is a critical component of the perioperative process with opioids representing a mainstay of therapy. The opioid epidemic is a growing concern in the United States. The goal of this study was to quantify the risk of opioid dependence or overdose among patients undergoing urological surgery and to identify risk factors of opioid dependence or overdose.

MATERIALS AND METHODS

We retrospectively reviewed data on urological surgery from 2007 to 2011. Data sources included the HCUP (Healthcare Cost and Utilization Project) inpatient, ambulatory surgery and emergency department data sets. Outcomes of postoperative opioid dependence and overdose were identified by previously validated ICD-9 codes. Multivariable logistic regression adjusted for surgical procedure was performed to identify predictors of opioid dependence or overdose following urological surgery.

RESULTS

Overall 675,527 patients underwent urological surgery, of whom 0.09% were diagnosed with opioid dependence or overdose. Patients in whom opioid dependence or overdose developed were younger (median age 51 vs 62 years), carried nonprivate insurance (69.6% vs 66%), underwent an inpatient procedure (81.0% vs 42.4%) and had a longer length of stay (median 3 vs 0 days) and a history of depression (14.4% vs 3.4%) or chronic obstructive pulmonary disease (20.3% vs 8.9%, all p <0.001). On adjusted multivariable analysis these factors remained independent risk factors for opioid dependence or overdose.

CONCLUSIONS

Postoperative opioid dependence or overdose affects 1 of 1,111 urological surgery patients. Risk factors for opioid dependence or overdose included younger age, inpatient surgery and increasing hospitalization duration, baseline depression, tobacco use and chronic obstructive pulmonary disease as well as insurance provider, including Medicaid, Medicare (age less than 65 years) and noninsured status.

Performance Measures of Diagnostic Codes for Detecting Opioid Overdose in the Emergency Department

OBJECTIVES

Opioid overdose mortality has tripled in the United States since 2000 and opioids are responsible for more than half of all drug overdose deaths, which reached an all-time high in 2014. Opioid overdoses resulting in death, however, represent only a small fraction of all opioid overdose events and efforts to improve surveillance of this public health problem should include tracking nonfatal overdose events. International Classification of Disease (ICD) diagnosis codes, increasingly used for the surveillance of nonfatal drug overdose events, have not been rigorously assessed for validity in capturing overdose events. The present study aimed to validate the use of ICD, 9th revision, Clinical Modification (ICD-9-CM) codes in identifying opioid overdose events in the emergency department (ED) by examining multiple performance measures, including sensitivity and specificity.

METHODS

Data on ED visits from January 1, 2012, to December 31, 2014, including clinical determination of whether the visit constituted an opioid overdose event, were abstracted from electronic medical records for patients prescribed long-term opioids for pain from any of six safety net primary care clinics in San Francisco, California. Combinations of ICD-9-CM codes were validated in the detection of overdose events as determined by medical chart review. Both sensitivity and specificity of different combinations of ICD-9-CM codes were calculated. Unadjusted logistic regression models with robust standard errors and accounting for clustering by patient were used to explore whether overdose ED visits with certain characteristics were more or less likely to be assigned an opioid poisoning ICD-9-CM code by the documenting physician.

RESULTS

Forty-four (1.4%) of 3,203 ED visits among 804 patients were determined to be opioid overdose events. Opioid-poisoning ICD-9-CM codes (E850.2-E850.2, 965.00-965.09) identified overdose ED visits with a sensitivity of 25.0% (95% confidence interval [CI] = 13.6% to 37.8%) and specificity of 99.9% (95% CI = 99.8% to 100.0%). Expanding the ICD-9-CM codes to include both nonspecified and general (i.e., without a decimal modifier) drug poisoning and drug abuse codes identified overdose ED visits with a sensitivity of 56.8% (95% CI = 43.6%-72.7%) and specificity of 96.2% (95% CI = 94.8%-97.2%). Additional ICD-9-CM codes not explicitly relevant to opioid overdose were necessary to further enhance sensitivity. Among the 44 overdose ED visits, neither naloxone administration during the visit, whether the patient responded to the naloxone, nor the specific opioids involved were associated with the assignment of an opioid poisoning ICD-9-CM code (p ≥ 0.05).

CONCLUSIONS

Tracking opioid overdose ED visits by diagnostic coding is fairly specific but insensitive, and coding was not influenced by administration of naloxone or the specific opioids involved. The reason for the high rate of missed cases is uncertain, although these results suggest that a more clearly defined case definition for overdose may be necessary to ensure effective opioid overdose surveillance. Changes in coding practices under ICD-10 might help to address these deficiencies.

Assessing the accuracy of opioid overdose and poisoning codes in diagnostic information from electronic health records, claims data, and death records

PURPOSE

The purpose of this study is to assess positive predictive value (PPV), relative to medical chart review, of International Classification of Diseases (ICD)-9/10 diagnostic codes for identifying opioid overdoses and poisonings.

METHODS

Data were obtained from Kaiser Permanente Northwest and Northern California. Diagnostic data from electronic health records, submitted claims, and state death records from Oregon, Washington, and California were linked. Individual opioid-related poisoning codes (e.g., 965.xx and X42), and adverse effects of opioids codes (e.g., E935.xx) combined with diagnoses possibly indicative of overdoses (e.g., respiratory depression), were evaluated by comparison with chart audits.

RESULTS

Opioid adverse effects codes had low PPV to detect overdoses (13.4%) as assessed in 127 charts and were not pursued. Instead, opioid poisoning codes were assessed in 2100 individuals who had those codes present in electronic health records in the period between the years 2008 and 2012. Of these, 10/2100 had no available information and 241/2100 were excluded potentially as anesthesia-related. Among the 1849 remaining individuals with opioid poisoning codes, 1495 events were accurately identified as opioid overdoses; 69 were miscodes or misidentified, and 285 were opioid adverse effects, not overdoses. Thus, PPV was 81%. Opioid adverse effects or overdoses were accurately identified in 1780 of 1849 events (96.3%).

CONCLUSIONS

Opioid poisoning codes have a predictive value of 81% to identify opioid overdoses, suggesting ICD opioid poisoning codes can be used to monitor overdose rates and evaluate interventions to reduce overdose. Further research to assess sensitivity, specificity, and negative predictive value are ongoing. Copyright © 2017 John Wiley & Sons, Ltd.

Identifying Patients for Overdose Prevention With ICD-9 Classification in the Emergency Department, Massachusetts, 2013-2014

The national rise in opioid overdose deaths signifies a need to integrate overdose prevention within healthcare delivery settings. The emergency department (ED) is an opportune location for such interventions. To effectively integrate prevention services, the target population must be clearly defined. We used ICD-9 discharge codes to establish and apply overdose risk categories to ED patients seen from January 1, 2013 to December 31, 2014 at an urban safety-net hospital in Massachusetts with the goal of informing ED-based naloxone rescue kit distribution programs. Of 96,419 patients, 4,468 (4.6%) were at increased risk of opioid overdose, defined by prior opioid overdose, misuse, or polysubstance misuse. A small proportion of those at risk were prescribed opioids on a separate occasion. Use of risk categories defined by ICD-9 codes identified a notable proportion of ED patients at risk for overdose, and provides a systematic means to prioritize and direct clinical overdose prevention efforts.