Heroin in Transition : US regional differences in the opioid pill and “heroin” overdose epidemics Jay Unick, PhD – University of Maryland Sarah Mars, PhD - UCSF Daniel Ciccarone, MD- UCSF #Rx Summit www.NationalRxDrugAbuseSummit.org Working Hypotheses  There are multiple but overlapping heroin and opioid related crises occurring in the US - Prescription opioids (PO) are widely available but the use in the population has leveled off - Heroin use and related harms are increasing  Problem: Current research confuses several different Licit issues. Supply Illicit Supply Licit Use Prescription Opiate use as prescribed Diverted Buprenorphine Use Illicit Use Doctor Shopping Heroin Use #Rx Summit www.NationalRxDrugAbuseSummit.org Working Hypotheses  There is evidence of a relationship between PO use and heroin use - PO and heroin are somewhat fungible i.e. attach to the same receptors and work for keeping addicted persons “well” - As PO use or potential for misuse has leveled off nationally; heroin use has increased suggesting a replacement - PO use frequently precedes heroin initiation, suggesting that PO use is a gateway to heroin. #Rx Summit www.NationalRxDrugAbuseSummit.org lane OPEN EACCESS Freely available online Intertwined Epidemics: National Demographic Trends in Hospitalizations for Heroin- and Opioid-Related Overdoses, 1993?2009 George Jay Uniclc". Daniel Rosenblumz. Sarah Mars]. Daniel Ciccarone] 1 School of Social Work, ?J'rversty oir ri'aryand. ?ialu?no a. Mary and. teo Enates Arre'ca. 2 Departme't of tconorrics. Da housie University. Halifax, Nova Scotia. Kansas. 3 Departn'e "1 of ram ly and Com 'nu" ty Medicine. Unive'sity of California San 5a" l-rancisco. California. United 3ta1es of America Abstract The historical pattems of opiate use show that sources and meth'?" is evidence that an enormous increase in the availability of pre drawing in new initiates to these drugs and changing the geogl reductions in prescription opiates may reduce harm. hut addictet this analysis. we test the hypothesis that changes in the Iates o' changes in the rate of heroin overdoses ICDQ codes from the Census were used to estimate overall and demogmphic spe 1993 and 1009. Flegression models were used to test for Iineartre used to model the interrelationship between PUD and HUD hos middleaged individuals had the largest increase in POD and HI increased in since 1001'. The lagged models show that increases years HUD admissions by a factor of 1.20 {p{0.00l] and that years by a factor of 1.51 Dur hypothesis of supported by these analyses. These ?ndings suggest that focusin in use to heroin rather minimizing the reduction in harm. Th resources on treatment and demand-side reduction is liltely to b: Shifting Patterns of Prescription Opioid and Heroin Abuse in the United States To THE EDITOR: From 1010 through 1013. there was a notable downturn in abuse of prescription opioids and a coincident increase in abuse of heroin in the United States.l Given that there is some evidence of a relationship between the two trends ie.g., some persons who abuse prescription opioids switch to heroin for a number of reasons and drug inteEhangeabiJity has been observed},M we sought to examine this relationship more closely, including the validity of reports suggest? ing regional differences in the balance between prescription opioid and heroin abuse.i Data on opioid abuse in the previous month were collected quarterly from January 1, 1003, through September 31. 1014, with the use of self-administered surveys that were completed anonymously by independent cohorts of 15,111r patients with opioid dependenceI as defined by the Diagnostic and Statistical Moment ofi'slmtoi Dis- orders, 4th edition, who were entering nonmeth? adone?maintenance treatment programs through- out the United States. Of these patients, 101 agreed to online intmviews to gather quali? tative information in order to amplify and in? terpret findings from the structured national survey. Figure 1 shows the unadjusted rates of abuse of prescription opioids only, abuse of prescription opioids and heroin, or abuse of heroin only among respondents who reported such abuse in the pre? vious month from 1000 through 1014. Rates of exclusive prescription opioid abuse remained sta- ble from 1000 through 1010, at 10%, but then decreased smadily, with an average annual reduc- tion of to less than 50% in 1014. Con? versely, concurrent abuse of both heroin and pre- scription opioids in the previous month increasedI with an average annual increase of 10.3%, from 13.0% in 1000 to 41.0% in 1014. Although the exclusive use of heroin was low in this popula- tion, it more than doubled from 1000 through 1014 [from 4.3% to The national data obscure important re? gional differences [Fig. 51 in the Supplementary Appendix, available with the full teat of this let- tet at The Northeast showed the most striking shifts in patterns of abuse. The West followed closely behind, with concurrent heroin via-ct} ocroata &Drug Abuse a Heroin swan 173353 Data and Methods  Analyses of hospitalizations using the AHRQ, Nationwide Inpatient and Nationwide Emergency Department Samples and deaths using the CDC Multiple Causes of Death File will examine national and regional trends in prescription opiate overdoses (OPOD) and heroin overdoses (HOD) rates.  US government data from the DEA and National Forensics Laboratory (NFLIS) will illustrate changes in heroin supply and form as well as reported evidence for contamination. These data will be correlated with HOD hospitalization rates by region. #Rx Summit www.NationalRxDrugAbuseSummit.org Problematizing the relationship  There are reasons to see the overlap between PO and heroin related health crises but there are also reasons to see them as different.  The populations at risk are not the same  OPOD are most prevalent in for individuals in their 50s  HOD are most prevalent for individuals in their 20s #Rx Summit www.NationalRxDrugAbuseSummit.org Opiate ODHate {If Dpiate age adjusted I 1549 20-24 25-29 30-34 35-39 4?-44 45-4' Elli-54 55-59 50?54 21112 2:112 212114 NATIONAL a &DrU9Abt?e #Rx Summit Heroin sum? Herein DD: US 1 5 1D 5 Hate {If Heroin DDSH - age adjust?d I 15-15 2121-24 25-25 5131-24 55-55 45-44 45-45 55-54 55-55 1513-54 2512 21:112. 217114 NATIONAL a B1DrU9Abuse #Rx Summit a Heroin 51111111117 Historically, risk of HOD was more evenly spread across the life span. But something changed around 2007. #Rx Summit www.NationalRxDrugAbuseSummit.org PDMPs Prescribing Standards Provider Education Historically, risk of OPOD was more evenly spread across the life span. But something changed around 2003. #Rx Summit www.NationalRxDrugAbuseSummit.org Problematizing the relationship  The National data may suggest that there is a population wide increase in demand for opiates but different sources for older and younger folks.  But…There is a difference in the risk of heroin use in different locations across the country.  This suggest more local – structural explanations rather than a simple pills = heroin relationship. #Rx Summit www.NationalRxDrugAbuseSummit.org Census Regions and Divisions of the United States MIDWEST LEGEND 1% REGIDN I Reg ion Map sI NATIONAL B?T?gr?i?f?n #Rx Summit m or f i Un r ac 4 all s os s ion g re PDMPs Prescribing Standards Provider Education Prescription Opioid Overdose Hospitalizations by Region #Rx Summit www.NationalRxDrugAbuseSummit.org Midwest Northeast separate Bunched up Heroin Overdose Hospitalizations by Regions #Rx Summit www.NationalRxDrugAbuseSummit.org Notice Decline in the West Prescription Opioid Overdose ED Admissions by Region #Rx Summit www.NationalRxDrugAbuseSummit.org Notice Flat in the West Heroin Overdose ED Admissions by Region #Rx Summit www.NationalRxDrugAbuseSummit.org Differences in heroin overdose by location in the US is Dramatic  There are places where the heroin overdose was bad and has gotten worse  Northeast  Places where it has gone from not so bad to very bad  Midwest/Appalachia  Places where it is starting to pick up but is not yet as bad as other locations  South, Mountain and West. #Rx Summit www.NationalRxDrugAbuseSummit.org PDMPs Prescribing Standards Provider Education Division Map #Rx Summit www.NationalRxDrugAbuseSummit.org Relatively High Rate Relatively Low Rate Prescription Opioid Hospitalization ODs #Rx Summit www.NationalRxDrugAbuseSummit.org Relatively Low Rate Relatively High Rate Heroin Hospitalization ODs #Rx Summit www.NationalRxDrugAbuseSummit.org Decrease in Prescription Opiate OD Hospitalizations between 2012 and 2014 in most age groups #Rx Summit www.NationalRxDrugAbuseSummit.org Mixed changes in heroin OD hospitalizaitons Subtitle #Rx Summit www.NationalRxDrugAbuseSummit.org Prescription ODs mostly unchanged between 2012 and 2014 #Rx Summit www.NationalRxDrugAbuseSummit.org Large increases in heroin OD hospitalizations for 25-35 yo. #Rx Summit www.NationalRxDrugAbuseSummit.org Differences in heroin overdose by location in the US is Dramatic  Demographic patterns of risk are different across different divisions  Race groups are differently at risk #Rx Summit www.NationalRxDrugAbuseSummit.org NATIONAL &Drug Abuse Heroin sum? - race adjusted Rate of Heroin 0Dsl100,000 2 I I White Heroin OD: New_England Black 2012 2013 2014 Large increase in Hispanic heroin OD admissions #Rx Summit Increase in white heroin OD admissions, decrease in black heroin OD admissions. Subtitle #Rx Summit www.NationalRxDrugAbuseSummit.org Differences in heroin deaths across state lines  Neighboring states are experiencing different changes to death rates associated with heroin  Further complicating the pills to heroin story #Rx Summit www.NationalRxDrugAbuseSummit.org The heroin epidemic overlaps and diverges from the supply of prescription opiates #Rx Summit www.NationalRxDrugAbuseSummit.org Oregon has a high rate of prescriptions and high rate of heroin deaths. Washington lower rate of prescriptions but escalating rate of heroin deaths. #Rx Summit www.NationalRxDrugAbuseSummit.org Michigan, Indiana, and Ohio have high rates of prescriptions but Ohio has a unique heroin overdose death problem. #Rx Summit www.NationalRxDrugAbuseSummit.org ?2005 2010 2015 Kentucky Tennessee West Virginia Same with West Virginia #Rx Summit Massachusetts has a low rate of prescriptions, Connecticut a moderate rate and Rhode Island the highest rate but these patterns do not correspond to heroin overdose deaths. #Rx Summit www.NationalRxDrugAbuseSummit.org Fentanyl does not look like the best explanation  Differences in deaths and OD from “heroin” could be to adulteration with fentanyl  But Fentanyl can not be the only cause - Show up to late in seized samples - Is present in only a small number of samples. (National Forensic Laboratory Information System (NFLIS) data) #Rx Summit www.NationalRxDrugAbuseSummit.org 0 Massachusetts NFLIS 2009 2010 2011 2012 2013 2014 - Hydrocodone - Oxycodone - heroin Fentanyl NATIONAL i &DrU9Abuse #Rx Summit a Heroin sum? Ohio NFLIS I I I I I 2009 2010 2011 2012 2013 2014 - Hydrocodone - Oxycodone - heroin Fentanyl 20,000 30,000 I 10,000 NATIONAL a &DrU9Abuse #Rx Summit a Heroin sum? West Virginia NFLIS I I I I 2009 2010 2011 2012 2013 2014 - Hydrocodone - Oxycodone - heroin Fentanyl 2,000 3,000 I 1 ,000 NATIONAL a 3(DrU9Abuse #Rx Summit Heroin sum? Discussion  No pattern emerges that tells a single story.  Heroin overdose deaths are increasing in many     states but not all states. Even states that had large prescription opiate overdose rates, have not seen the shift to heroin. Differences occur by race, age and region Fentanyl is at best a partial explanation Economic conditions do not seem to explain the difference. For example why is West Virginia experiencing twice the rate of increase compared to Kentucky? #Rx Summit www.NationalRxDrugAbuseSummit.org Conclusion  There is evidence of a relationship between PO use and heroin use - They are intertwined  There is also strong evidence that we are experiencing a set of local epidemics rather than a national epidemic.  Solutions to the current opiate/heroin crises require local knowledge - Need to understand … • what is driving use • changing risk of use #Rx Summit www.NationalRxDrugAbuseSummit.org