Persons experiencing fatal overdoses in 2015-2016 in Maryland who had opioid prescriptions, hospital records, and/or criminal justice records prior to overdose, separated by type of opioid death. More information on opioid prescriptions, hospital records, and justice records can be found in the Methods section.
Fatal opioid overdose death rates per 100 000 presented for any opioid (prescription opioid, fentanyl, or heroin). To calculate this rate, the denominator (shown in parentheses) represents the number of individuals with at least 1 of the labeled records in 2015-2016. The numerator is the number of fatal overdoses in 2015-2016 for individuals who also had at least 1 of the labeled records in 2015-2016. Maryland mean was calculated by taking the number of fatal opioid overdoses in 2015-2016 and dividing by the US Census estimate for the population of Maryland. More information on opioid prescriptions, hospital records, and justice records can be found in the Methods section.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Eisenberg MD, Saloner B, Krawczyk N, et al. Use of Opioid Overdose Deaths Reported in One State’s Criminal Justice, Hospital, and Prescription Databases to Identify Risk of Opioid Fatalities. JAMA Intern Med. 2019;179(7):980–982. doi:10.1001/jamainternmed.2018.8757
The United States is in the midst of an opioid overdose epidemic, with 45 000 opioid overdose deaths in 2017, most involving fentanyl and heroin.1 The President’s Commission on Combating Drug Addiction and the Opioid Crisis has recommended data integration between state-based prescription drug monitoring programs and other systems to identify individuals who are at an elevated risk of overdose.2 Linking prescription drug monitoring program data with other large databases can provide insight into how different service systems could have reached many individuals who fatally overdosed and how risk rates for each subgroup compare with statewide means.
We identified Maryland residents with at least 1 record in 2015-2016 in any of 3 state-level data sets: opioid prescriptions in the prescription drug monitoring program data (n = 1 740 332), inpatient hospitalization or emergency department visits in the Health Services Cost Review Commission data (n = 2 047 397), or at least 1 record for an adjudicated arrest, incarceration, or community supervision record (parole or probation) related to a property or drug offense in the Department of Public Safety and Correctional Services data (n = 42 925). These data were linked with opioid overdose death records (intentional and unintentional) from the Office of the Chief Medical Examiner (n = 2902), which could be separated into deaths involving heroin (n = 1938), fentanyl (n = 1452), and/or prescription opioids (n = 765) (numbers do not sum to 2902, as multiple types of opioids could be involved in a death). Data were linked and deidentified through a health information exchange that maintains a sharing agreement with the Maryland Department of Health by using a validated algorithm.3 The study was approved by the institutional review boards at theJohns Hopkins School of Public Health and the Maryland Department of Health.
We described the proportion of individuals with a fatal opioid overdose who previously appeared in 1 or more of the above-described data sets. In addition, we compared the overdose death rate across each combination of data sources. All analyses were performed using STATA/MP, version 15 (StataCorp).
Most individuals with fatal opioid overdose events appeared in at least 1 of the 3 data sets between 2015 and 2016 (Figure 1): 27.7% had opioid prescriptions and hospital records, 19.7% had hospital records only, 7.1% had opioid prescriptions only, and 5.9% had criminal justice records (either alone or in combination with clinical records). A total of 39.6% of individuals with fatal overdoses could not be linked with records in any data set. Individuals whose death involved prescription opioids were more likely to have had opioid prescriptions than individuals whose death involved heroin or fentanyl.
The statewide opioid overdose rate for Maryland residents during the 2-year period (Figure 2) was 49 of 100 000 individuals compared with 63 of 100 000 individuals who were prescribed opioids, 99 of 100 000 individuals with inpatient hospital records, and 413 of 100 000 individuals with adjudicated arrest records. Risk of overdose was considerably higher for individuals appearing in multiple data sets, particularly in criminal justice systems in combination with other data (eg, for community supervision and opioid prescription records: 765 of 100 000).
This study presents data showing how state-based data linkage across prescription drug, hospital, and criminal justice records can be used to identify populations at an elevated risk of fatal opioid overdose. All 3 types of records are associated with elevated risk. Consistent with prior research,4-6 approximately one-third of individuals with overdose deaths had histories of prescribed opioids, almost half of all individuals with overdose deaths had hospital records, and individuals within the criminal justice population had significantly high risk for overdose deaths, especially those also appearing in data from clinical settings. Linked databases are a promising tool for population overdose surveillance. Policymakers can use study results to focus resources toward high-risk groups or more broadly distributed risk. This study did not examine long-term associations or important differences within categories, such as people who concurrently use benzodiazepines and opioids or who are in treatment for opioid use disorder. Data linkage efforts such as these can support future research to understand risk across subgroups.
Accepted for Publication: December 22, 2018.
Corresponding Author: Matthew D. Eisenberg, PhD, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD (firstname.lastname@example.org).
Published Online: April 15, 2019. doi:10.1001/jamainternmed.2018.8757
Author Contributions: Dr Eisenberg had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Eisenberg, Saloner, Weiner.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Eisenberg, Saloner.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Eisenberg, Saloner, Weiner.
Obtained funding: Saloner, Weiner.
Administrative, technical, or material support: Krawcyzk, Ferris, Schneider, Lyons, Weiner.
Supervision: Saloner, Weiner.
Conflict of Interest Disclosures: Ms Ferris reported receiving grants from the Bureau of Justice Assistance Harold Rogers Prescription Drug Monitoring Program Category 2 during the conduct of the study and being employed by Audacious Inquiry and contracted full time with the Chesapeake Regional Information System for our Patients, the Maryland state-designated health information exchange. Ms Lyons reported receiving grants from the Bureau of Justice Assistance, US Department of Justice, during the conduct of the study. Dr Weiner reported receiving grants from the US Department of Justice to the Maryland Department of Health during the conduct of the study. No other disclosures were reported.
Funding/Support: This study was funded by a Harold Rogers Prescription Drug Monitoring Program grant awarded by the US Department of Justice, Office of Justice Programs, Bureau of Justice Assistance, aimed at improving prescription drug monitoring programs to prevent and reduce misuse and abuse of prescription drugs. The Maryland Department of Health, Behavioral Health Administration, was awarded the grant in partnership with the Center for Population Health Information Technology at the Johns Hopkins Bloomberg School of Public Health for data analyses and subject matter expertise and the Chesapeake Regional Information Systems for our Patients to link the data sets. Mss Krawczyk and Schneider were supported by T32 grant 5T32DA007292-25 from the National Institute on Drug Abuse.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Kate Jackson, MPH, Maryland Department of Health’s Behavioral Health Administration and the Chesapeake Regional Information System for our Patients, assisted with assembling the data. She was not compensated for her contributions. Tom Richards, MS, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, provided programming support. He was compensated for his contributions. We also thank the Office of the Chief Medical Examiner, the Health Services Cost Review Commission, and the Department of Public Safety and Correctional Services for providing data.