Data quality considerations when using county-level opioid overdose death rates to inform policy and practice

Objectives. To examine trends in opioid overdose deaths by race/ethnicity from 2018 to 2019 across 67 HEALing Communities Study (HCS) communities in Kentucky, New York, Massachusetts, and Ohio. Methods. We used state death certificate records to calculate opioid overdose death rates per 100 000 adult residents of the 67 HCS communities for 2018 and 2019. We used Poisson regression to calculate the ratio of 2019 to 2018 rates. We compared changes by race/ethnicity by calculating a ratio of rate ratios (RRR) for each racial/ethnic group compared with non-Hispanic White individuals. Results. Opioid overdose death rates were 38.3 and 39.5 per 100 000 for 2018 and 2019, respectively, without a significant change from 2018 to 2019 (rate ratio = 1.03; 95% confidence interval [CI] = 0.98, 1.08). We estimated a 40% increase in opioid overdose death rate for non-Hispanic Black individuals (RRR = 1.40; 95% CI = 1.22, 1.62) relative to non-Hispanic White individuals but no change among other race/ethnicities. Conclusions. Overall opioid overdose death rates have leveled off but have increased among non-Hispanic Black individuals. Public Health Implications. An antiracist public health approach is needed to address the crisis of opioid-related harms. (Am J Public Health. Published online ahead of print September 9, 2021:e1–e4. https://doi.org/10.2105/AJPH.2021.306431 )

Download Full-text

Opioid Overdose Education, Prevention, and Management (DRAFT)

10.1093/med/9780190265366.003.0013 ◽

2018 ◽

Cited By ~ 1

Author(s):

Bernd Wollschlaeger

Keyword(s):

The United States ◽

Opioid Use Disorder ◽

Opioid Overdose ◽

Opioid Use ◽

Death Rates ◽

Overdose Prevention ◽

Overdose Death ◽

Good Samaritan ◽

Opioid Deaths ◽

Good Samaritan Laws

In reviewing the elements of opioid overdose education, prevention, and management, this chapter focuses particularly on practical interventions that are available and deserve advocacy; e.g., provision of naloxone to those with opioid use disorder and to possible first responders. It moves from a discussion of the epidemiology of opioid deaths to the more individual topic of patient risk for overdose. Prophylactic interventions in the form of education of the patient’s family and friends, and agreements for treatment with informed consent are described. There follows a discussion of management of the opioid poisoning itself, including use/distribution of naloxone injection. Two figures are included: drug overdose death rates in the United States (2014); a map describing the current states with naloxone or “good Samaritan” laws impacting opioid overdose management. A text box with resources includes directions for initiation of community overdose prevention and intervention schemes.

Download Full-text

Impact of a community-based naloxone distribution program on opioid overdose death rates

Drug and Alcohol Dependence ◽

10.1016/j.drugalcdep.2019.06.038 ◽

2019 ◽

Vol 204 ◽

pp. 107536 ◽

Cited By ~ 6

Author(s):

Rebecca B. Naumann ◽

Christine Piette Durrance ◽

Shabbar I. Ranapurwala ◽

Anna E. Austin ◽

Scott Proescholdbell ◽

...

Keyword(s):

Opioid Overdose ◽

Community Based ◽

Death Rates ◽

Overdose Death

Download Full-text

Emerging and Continuing Trends in Opioid Overdose Decedent Characteristics during COVID-19

10.1101/2021.07.28.21261276 ◽

2021 ◽

Author(s):

Gian-Gabriel P. Garcia ◽

Erin Stringfellow ◽

Catherine DiGennaro ◽

Nicole Poellinger ◽

Jaden Wood ◽

...

Keyword(s):

North Carolina ◽

Rhode Island ◽

Race And Ethnicity ◽

The United States ◽

Vital Statistics ◽

Opioid Overdose ◽

Cross Sectional ◽

Death Rates ◽

Overdose Death ◽

Overdose Deaths

Background: Since COVID-19 erupted in the United States, little is known about how state-level opioid overdose trends and decedent characteristics have varied throughout the country. Objective: Investigate changes in annual overdose death rates, substances involved, and decedent demographics in opioid overdose deaths across nine states; assess whether 2019-2020 trends were emerging (i.e., change from 2019-2020 was non-existent from 2018-2019) or continuing (i.e., change from 2019-2020 existed from 2018-2019). Design: Cross-sectional study using vital statistics data to conduct a retrospective analysis comparing 2020 to 2019 and 2019 to 2018 across nine states. Setting: Alaska, Colorado, Connecticut, Indiana, Massachusetts, North Carolina, Rhode Island, Utah, and Wyoming. Participants: Opioid-related overdose deaths in 2018, 2019, and 2020. Measurements: Annual overdose death rate, proportion of overdose deaths involving specific substances, and decedent demographics (age, sex, race, and ethnicity). Results: We find emerging increases in annual opioid-related overdose death rates in Alaska (55.3% [P=0.020]), Colorado (80.2% [P<0.001]), Indiana (40.1% [P=0.038]), North Carolina (30.5% [P<0.001]), and Rhode Island (29.6% [P=0.011]). Decreased heroin-involved overdose deaths were emerging in Alaska (-49.5% [P=0.001]) and Indiana (-58.8% [P<0.001]), and continuing in Colorado (-33.3% [P<0.001]), Connecticut (-48.2% [P<0.001]), Massachusetts (39.9% [P<0.001]), and North Carolina (-34.8% [P<0.001]). Increases in synthetic opioid presence were emerging in Alaska (136.5% [P=0.019]) and Indiana (27.6% [P<0.001]), and continuing in Colorado (44.4% [P<0.001]), Connecticut (3.6% [P<0.05]), and North Carolina (14.6% [P<0.001]). We find emerging increases in the proportion of male decedents in Colorado (15.2% [P=0.008]) and Indiana (12.0% [P=0.013]). Limitations: Delays from state-specific death certification processes resulted in varying analysis periods across states. Conclusion: These findings highlight emerging changes in opioid overdose dynamics across different states, which can inform state-specific public health interventions.

Download Full-text

Predicting County-Level Overdose Death Rates Amid the Escalating Overdose Crisis in the United States: A Statistical Modeling Approach Predicting Deaths Through 2022

10.31219/osf.io/83zkh ◽

2022 ◽

Author(s):

Charles Marks ◽

Daniela Abramowitz ◽

Christl A. Donnelly ◽

Daniel Ciccarone ◽

Natasha Martin ◽

...

Keyword(s):

Negative Binomial ◽

The United States ◽

Similar Efficacy ◽

Original Model ◽

Mortality Data ◽

County Level ◽

Online Application ◽

Death Rates ◽

Overdose Death ◽

The Us

Aims. U.S. overdose (OD) deaths continue to escalate but are characterized by geographic and temporal heterogeneity. We previously validated a predictive statistical model to predict county-level OD mortality nationally from 2013 to 2018. Herein, we aimed to: 1) validate our model’s performance at predicting county-level OD mortality in 2019 and 2020; 2) modify and validate our model to predict OD mortality in 2022.Methods. We evaluated our mixed effects negative binomial model’s performance at predicting county-level OD mortality in 2019 and 2020. Further, we modified our model which originally used data from the year X to predict OD deaths in the year X+1 to instead predict deaths in year X+3. We validated this modification for the years 2017 through 2019 and generated future-oriented predictions for 2022. Finally, to leverage available, albeit incomplete, 2020 OD mortality data, we also modified and validated our model to predict OD deaths in year X+2 and generated an alternative set of predictions for 2022.Results. Our original model continued to perform with similar efficacy in 2019 and 2020, remaining superior to a benchmark approach. Our modified X+3 model performed with similar efficacy as our original model, and we present predictions for 2022, including identification of counties most likely to experience highest OD mortality rates. There was a high correlation (Spearman’s ρ = 0.93) between the rank ordering of counties for our 2022 predictions using our X+3 and X+2 models. However, the X+3 model (which did not account for OD escalation during COVID) predicted only 62,000 deaths nationwide for 2022, whereas the X+2 model predicted over 87,000.Conclusion. We have predicted county-level overdose death rates for 2022 across the US. These predictions, made publicly available in our online application, can be used to identify counties at highest risk of high OD mortality and support evidence-based OD prevention planning.

Download Full-text

Using Geographic Information System Analysis to Understand West Virginia’s Growing Opioid-Overdose Epidemic - What Are We Missing?

Prehospital and Disaster Medicine ◽

10.1017/s1049023x19001249 ◽

2019 ◽

Vol 34 (s1) ◽

pp. s53-s53

Author(s):

Sasha Rihter ◽

Nathan Menke

Keyword(s):

Information System ◽

Geographic Information System ◽

The United States ◽

Geographic Information ◽

The State ◽

Opioid Overdose ◽

Death Rates ◽

Overdose Death ◽

The Impact ◽

Overdose Deaths

Introduction:The opioid epidemic is overwhelming communities across the United States. West Virginia (WV) has been devastated, heralding a 86% increase in deaths from 2012-2016, and over 1,000 deaths last year as per WV Health Statistics Center. Treatment centers and providers have emerged throughout the state to provide medication-assisted treatment (MAT). The impact of these clinics on the opioid abusing population is not yet fully understood.Aim:Utilizing Geographic Information System (GIS), a comparison of MAT provider locations versus regions of historical overdoses can indicate areas of deficiency. If no providers emerge in underserved counties, overdose deaths in those areas will continue to rise.Methods:Maps were created using current DEA-X licenses in WV registered through Substance Abuse and Mental Health Services Administration (SAHMSA). Overdose death rates were taken from WV Public Health Records from 2010-2017. Two maps and corresponding data were compared for overlap or lack thereof.Results:Of the 338 locations of DEA-X licenses registered, 17.5% are in Cabell County, which led the state in overdose deaths in 2017. Only 2.5% of the total providers are currently in Wayne County, which had the second highest overdose death rate. Berkeley County, which was 3rd highest, has a mere 6.5% of total providers. Comparatively, Kanawah County, home to the state’s capital, has over twice this number of providers despite consistently having at or below the state average of overdose rates. Resources are pulled towards population-dense areas or university centers, where the epidemic is present but misses counties with higher overdose rates.Discussion:Results show a lack of MAT providers in many of WV’s devastated counties. Treatment centers exist throughout the state but are concentrated in regions with large cities or academic centers. This distribution limits accessibility to a marginalized patient population, making improvements unlikely in WV’s future opioid-overdose death rates.

Download Full-text