Afteralcohol intoxication,opioidsare themostcommoncause ofpoisoninginpatientspresentingtoNorthAmericanemergency departments.1Mostopioidsmisusedbypatientsoriginate from prescriptionmedication.Most patientswhooverdoseonprescription opioids are taking their medications differently thanprescribedorareusingopioidsprescribedtosomeoneelse. These2main typesofnonmedical opioiduse representamajor causeofmorbidityandmortality.Someindividualswhomisuse opioidsareseekingeuphoriceffects,butothershavedeveloped dependence through chronic opioid use and are simply trying toavoidopioidwithdrawal.Opioid-relatedharmhasnowreached epidemic levels: emergency department visits for nonmedical useofprescriptionopioidsmorethandoubledfrom2004to2011, accountingforanestimated488 000visits in2011.1Deathshave more than tripled since 1999, with an estimated 16 235 deaths attributable to prescription opioids in 2013.2,3 In this issueof JAMA, the studybyHanet al4 examined the current scopeof theopioidepidemic in theUnitedStates.Using datafromnearlyhalfamillionrespondentstotheannualNational SurveyonDrugUseandHealth(NSDUH), theauthors foundthat overalltrendsinself-reportednonmedicaluseofprescriptionopioidsdecreasedfrom5.4%to4.9%overan11-yearperiod, includingadecline innewusersofopioids, from1%in2003to0.6%in 2013.Althoughthisoverall findingofa reduction innonmedical useofprescriptionopioidsisencouraging,thestudyalsoreported increases in theprevalenceofprescriptionopioidusedisorders (abuseandaddiction)and increases in theprevalenceofopioidassociatedmortality,usingdatafromtheNationalVitalStatistics System’sMultipleCauseofDeathFiles.Theauthorsalsoreported an increasedprevalenceof frequentopioiduse (>100days/year) and highly frequent use (>200 days/year), as well as a greater prevalenceofprescriptionopioidusedisorders inpatientswith majordepressiveepisodes(MDEs)thaninpatientswithoutthem. The findings ofHan et al suggest thatmore patients are experiencing an inexorable progression from initial opioid use to frequent use, highly frequent use, or an opioid use disorder. Another report in this issue of JAMA by Saloner and Karthikeyanaddressedarelatedquestion:Dopatientswithnonmedical use of opioids access treatment?5 Among individuals identifiedinNSDUHashavinganopioidusedisorder, theauthors examined utilization of substance abuse treatment during the sameperiodasHanetal (2004-2013).Adjustedratesof individualswith opioid use disorders receiving treatmentwere lowand essentiallyunchangedduringthereportingperiod(18.8%in20042008to19.7%in2009-2013).SalonerandKarthikeyanalsofound thatthenumberofsettingsvisitedfortreatment(ie, inpatientservices,outpatientclinics,physicians’offices) increased, from2.8 to3.3, includingreceiptoftreatmentinofficesettings(from25.1% to34.8%)whereuseofbuprenorphineismost likely.Theauthors cannot explain why so many patients apparently did not seek treatment, suchaswhether it isnot available,not affordable, or notof interest.Furthermore, theoutcomesof treatmentcannot bemeasuredwiththesedata,butotherstudieshavereportedthat the long-termeffectiveness ofmost such therapies ismodest.6 Saloner and Karthikeyan also found that over time, survey respondents were older and were less likely to have private health insurance. Despite the Mental Health Parity and Addiction Equity Act of 2008,7 which mandated that insurersoffermentalhealthandaddictionbenefits comparablewith their medical-surgical benefits, an increase in treatment opportunitieswasnot observed in the latter part of the studyperiod, after the act was implemented. This suggests that ability topaymost likelywasnot theprimary factor in thedecision to forgo treatment but does not clarify the roles of access to, or interest in, treatment for an opioid use disorder. Prescribingofopioidanalgesics,particularlyforchronicpain, appears to be amain factor in themajority of nonmedical use. Basedonotherdataavailable in theNSDUH,prescribersare,directlyor indirectly, the sourceofmostmisusedopioids.8Anestimated53%ofnonmedicalusersreportedobtainingprescription opioidsfromafriendorrelative,81%ofwhomreceivedtheirdrug fromaphysician. It isunclearwhether theseprescriptionswere issuedfortherapeuticpurposesororiginatedfromunscrupulous prescribers (ie, “pillmills”); regardless, the sourceofopioiduse andmisuse is often a seemingly legitimate prescription. There is little evidence for long-term benefit from opioid therapy formost typesof chronicpain.9 It remainsunclearwhy thispracticeofopioidprescribingcontinuesdespite recommendationstothecontrary.9,10Newopioidmedications,manyofthem withtamper-resistantformulations,continuetobemarketeddespite the lackofevidencethat thesepreparationsreducetherisk ofaddiction.11More than10%ofpatientswho initiate treatment withopioidswill likelyprogress tochronicuse,definedasongoingtreatmentformorethan3months.12Nearlyallpatientstreated withlong-termopioidtherapydeveloptoleranceanddependence tovaryingdegrees,about25%becomenonmedicalusers,and10% developfeaturessuggestiveofaddiction.13Thesearesoberingpercentagesinlightofthemillionsofpatientsprescribedthesedrugs everyyear.14Consequently,forthemanypatientswhoneedtreatment for addictionor complicationsof substancemisuse, there areoftensignificantbarriers toobtainingcare.6Dependence,addiction, anddose escalation resulting from tolerancemakediscontinuingopioidsdifficult.Manypatientsunderstandablyperceivetheyneedongoingopioidanalgesic therapybecause,when thedrugsarediscontinued,anunpleasantwithdrawalsyndrome withassociatedpainensues.Patientsquickly learnthat resumpRelated articles pages 1468 and 1515 Opinion