pharmacoepidemiologyanddrugsafety2014;23:1238–1246

Publishedonline11June2014inWileyOnlineLibrary(wileyonlinelibrary.com)DOI:10.1002/pds.3658

ORIGINALREPORT

Reductionsinreporteddeathsfollowingtheintroductionofextended-releaseoxycodone(OxyContin)withanabuse-deterrentformulation†

NelsonE.Sessler1,JerodM.Downing1,HrishikeshKale1,HowardD.Chilcoat1,3,ToddF.Baumgartner4andPaulM.Coplan1,2*

1DepartmentofRiskManagementandEpidemiology,PurduePharmaL.P.,Stamford,CT,USA

2Department ofClinicalBiostatisticsand Epidemiology UniversityofPennsylvania PerelmanSchool ofMedicine (Adjunct)Philadelphia,PA,USA

3DepartmentofMentalHealth,JohnsHopkinsBloombergSchoolofPublicHealth(Adjunct),Baltimore,MD,USA

4DepartmentofRegulatoryAffairs,PurduePharmaL.P.,Stamford,CT,USA

ABSTRACT

PurposeAbuseofopioidanalgesicsfortheirpsychoactiveeffectsisassociatedwithalargenumberoffatalities.Theeffectofmakingopi-oidtabletshardertocrush/dissolveonopioid-relatedfatalitieshasnotbeenassessed.Theobjectiveofthisstudywastoassesstheimpactofintroducingextended-releaseoxycodone(ERO[OxyContin®])tabletscontainingphysicochemicalbarrierstocrushing/dissolving(reformulatedERO)ondeathsreportedtothemanufacturer.

MethodsAllspontaneousadverseeventreportsofdeathintheUSreportedtothemanufacturerbetween3Q2009and3Q2013involvingEROwereused.Themeannumbersofdeaths/quarterinthe3yearsafterreformulatedEROintroductionwerecomparedwiththeyearbefore.Changesintheslopeoftrendsindeathswereassessedusingsplineregression.Comparisongroupsconsistedofnon-fatalreportsinvolving ERO andfatalityreportsinvolving ERmorphine.

ResultsReportsofdeathdecreased82%(95%CI:-89,-73)fromtheyearbeforetothethirdyearafter(131to23deathsperyear)reformulation;overdosedeathreportsdecreased87%(95%CI:-93,-78)andoverdosedeathswithmentionofabuse-relatedbehaviordecreased86%(95%CI:-92,-75).Incontrast,non-fatalEROreportsdidnotdecreasepost-reformulation,andreportedERmorphinefa-talitiesremainedunchanged.TheratioofEROfatalitiestoalloxycodonefatalitiesdecreasedfrom21%to8%intheyearpre-reformulationtothesecondyearpost-reformulation.

ConclusionsThesefindings,whenconsideredinthecontextofpreviouslypublishedstudiesusingothersurveillancesystems,suggestthattheabuse-deterrentcharacteristicsofreformulatedEROhavedecreasedthefatalitiesassociatedwithitsmisuse/abuse.©2014TheAuthors.PharmacoepidemiologyandDrugSafetypublishedbyJohnWileySons,Ltd.

keywords—OxyContin;extended-releaseoxycodone;abuse-deterrent;overdosedeath;pharmacovigilance;pharmacoepidemiology

Received11September2013;Revised1May2014;Accepted12May2014

INTRODUCTION

Opioidanalgesicsarerecommendedforthetreatmentofserious,persistentpainafternon-pharmacologictherapies and non-opioid medications have been

*Correspondenceto:P.Coplan,ExecutiveDirector,DepartmentofRiskManagementandEpidemiology,PurduePharmaL.P,OneStamfordForum,Stamford, CT06901,USA. E-mail:

†PriorPresentation:Aportion ofthesedata waspresented atPAINWeek2012 inLasVegas,NV,USA2012;thesecondInternationalConferenceonOpioidsBoston,MA,USA2012;the29thInternationalConferenceonPharma-coepidemiologyand Therapeutic RiskManagementinMontreal,Canada 2013.

used.1–5Extended-release(ER)andimmediate-release(IR)opioidanalgesicsaredispensedtoover4millionand56millionpatientsintheUSAperyear,respec-tively(IMShealth).However,overthelastdecade,prescriptionopioidabuse(forpsychoactiveeffects)hasincreasedgreatly,resultinginincreaseddeathsandburdentopublichealth.6–9Theaddictionpotentialandsequelaeofabuseincreaseexponentiallywhentabletsarecrushed/dissolvedfornon-oraladministra-tion(e.g.,snorting,injecting,andsmoking)toobtainrapidabsorptionoftheopioidexperiencedasa“high.”10–12 Initial oral abuse often progresses to

©2014TheAuthors.PharmacoepidemiologyandDrugSafetypublishedbyJohnWileySons,Ltd.

ThisisanopenaccessarticleunderthetermsoftheCreativeCommonsAttribution-NonCommercial-NoDerivsLicense,whichpermitsuseanddistributioninanymedium,providedtheoriginalworkisproperlycited,theuseisnon-commercialandnomodificationsoradaptationsaremade.

non-oralabusebythetimeofadmissiontoasubstanceabusetreatmentfacility.11,13–15

Scientificinnovationofabuse-deterrentformulationstopromotesafeprescriptionopioiduseisafocusofre-searchanddevelopmentamongpharmaceuticalcompa-nies.16,17Pharmacologicalapproachestoincorporatecharacteristicsdesignedtodeterabusehaveincluded:

(i)addinganopioidantagonist(e.g.,buprenorphineandnaloxone[Suboxone®],ERmorphineandseques-terednaltrexone[Embeda®],aswellaspentazocineandnaloxone[Talwin®NX]);(ii)addingagentsthatin-duceunpleasantsymptomswithexcessiveintake(e.g.,IRoxycodoneandaversiveagent[Oxecta®]);and(iii)incorporatingphysicochemicalbarriersintendedtocon-ferresistancetotablettampering(e.g.,ERoxycodone[OxyContin®],ERoxymorphone[Opana®ER],andERtapentadol[Nucynta®ER]).

OxyContin(ERO)isER-formulatedoxycodoneap-provedinthe USAin1995forthetreatmentofmoder-ate-to-severechronicpain,18whichhasbeenwidelyabused,19–21especiallybysnorting/injecting,requiringtabletcrushing/dissolving.22–25InApril2010,theFoodandDrugAdministration(FDA)approvedare-formulatedEROcontainingphysicochemicalbarrierstobreaking,crushing/dissolvingtodeterabuse,whichremainstheonlyavailableERformofoxycodoneavailableintheUSA.Pre-approvalstudiesdemon-stratedthatreformulatedEROisbioequivalentto,ismoredifficulttoextractoxycodonefrom,andislesslikedbyabusersthanthe original formulation.26–28AllshipmentsoforiginalEROtowholesalersstoppedon5August,andshipmentsofreformulatedEROstartedon9August2010.Thetransitiontothereformu-lationwasconductedwithoutnotificationofthegeneralpublic. Post-marketing studies have demonstrated a

reductioninreportedEROabuseindrugtreatmentcen-ter populationsand calls topoison centers.29–32InApril2013,EROreceivedFDA-approvedlabeling,indicat-ingthatitisexpected to beabuse-deterrentvia intrana-sal and intravenous routes of administration.33However,theeffectsonfatalityhavenotbeenreported.ThisreportfocusesontheimpactofreformulatedERO on reports of US fatalities submitted to themanufacturer’spharmacovigilancedatabase.Mortalitydatabases,suchastheNationalDeathIndexandstatemedicalexaminerdatabaseswerenotusedbecausetheydonotdifferentiatebetweenIRandERoxyco-done,andonly5%ofpatientsprescribedoxycodoneintheUSAreceivedERoxycodone(IMSHealth).However,theratioofdeathsassociatedwithEROre-ported to the manufacturer versus all oxycodonedeathsreportedtotheFDA’sAdverseEventReportingSystem(AERS)wasassessedtoprovideadditional

contexttothefindings.METHODS

Manufacturersreceive,archive,andsubmitspontane-

ousreportsofadverseeventsonmarketeddrugstonationaldrug-regulatoryauthorities,suchastheFDAintheUSA.34,35Searching themanufacturer’s adverseeventreportingdatabaseidentifiedallreportsoffataleventsoriginatingintheUSAinvolvingEROfrom3Q2009–3Q2013. Individualcasereportnarrativede-scriptions werereviewedand categorizedasmention-inganopioidoverdose-relatedeventand/ordrugabuse-relatedbehaviorusingcriteriadevelopeda priori(Table1).Thisreviewwasconductedbythetwoprimaryauthorswithanydisagreementsresolvedby consensus.

Table1. Criteria toidentify overdose-related eventand abuse-related behavior mentions

Overdose-relatedevent•Reporter describedevent usingverbatimterm“overdose”ora medically relatedterm (e.g.,drug poisoning,polydrugtoxicity,drugintoxication,andovermedicated);or

•Circumstancessurroundingdeathsuggestanoverdose-relatedevent(e.g.,ingestedmanypills,dosingmistake,tampering/snorting/injection ofdrug,anddrug obtained and ingestedataparty); or

•Coronerorphysiciandeemed fatality wasassociatedwithopioidoverdoseorpolydrug overdose(withorwithouttoxicologyevidenceofoxycodoneoropioidingestion).

Abuse-relatedbehavior•Subjectcurrentlyorpreviouslymanipulatedextended-releaseoxycodonewithintentionofabuse(e.g.,crushedandsnorted,dissolvedandinjected); or

•Extended-releaseoxycodonewasnotprescribedtothesubjectand/orsubjectwasobtainingdrugviaunlawfultransfer(e.g.,stolen,ataparty,fromparentssupply,andfromthestreet);or

•Subjectwasobtainingextended-releaseoxycodoneprescriptionsfromapillmill,multiplehealthcareproviders,and/ormultiplepharmacies;or

•Reporterstatessubjecthas historyofaddictiondisorderand/ordrug rehabilitationor indicatesthatsubjectiscurrentlyaddicted;or

•Reporterstatessubjecthadbeenusingillicitdrugsoralcoholincombinationwithextended-releaseoxycodone(e.g.,heroin,cocaine,marijuana,andamphetamines),or

•There was evidenceof subjectexposure toabenzodiazepine, anopioidother than oxycodone,and/or muscle relaxant/hypnoticinabsenceofmentionofprescription.Forthispurpose,anexposurewasdefinedas:theindividualwasobservedtakingdrug,orreportedtohavetakenthedrug,ordrugwasrevealedin toxicologicalresults.

Theanalysisfocusedonspontaneousfatalityreportsthatincludedmonth/yearofdeath,astimetrendsinmortalitycannotbeascertainedwherethisinformationisunknownandthedatewhenthereportwasreceivedbythemanufacturerdoesnotnecessarilycorrelatewiththedate of death.Reportsassociatedwithpost-market-ingstudies(includingthemanufacturer’sindividualpa-tientassistanceprogram),litigation(becausethesewerenotspontaneousreports),andthoselackingacorereportingelement(patient,reporter,suspectproduct,oradverseevent)wereexcluded.Allreportscontainingmonth/yearofdeathfrom3Q2009–2Q2013wereana-lyzedusingSASv9.2(SASInstitute,Inc,Cary,NC).

Fatalitiesweredividedintofourperiodscorrespondingto1yearpre-reformulation(3Q2009–2Q2010)andthefirst (3Q2010–2Q2011), second (3Q2011–2Q2012),

andthird(3Q2012–2Q2013)yearpost-reformulation.Themeanfatalitiesperquarterandchangesin theslopeoftrendsinfatalitieswerecalculatedbysplineregressionusingaPoissonmodelwiththeinflectionpointcorre-spondingtothetimeofEROreformulation.36,37

Severalsensitivityanalyseswereconductedtoas-sesstherobustnessoftheresults.Toassesstheimpactofprescriptionchangesonfatalities,countswereadjustedfor100,000EROprescriptionsdispensed(IMSNationalPrescriptionAuditdatabasesystem).38Becausereportingaccuracyvariesbysource,casesreportedbyhealthcareprofessionalswereanalyzedseparately.Toassesstheimpactofintentionalharm,fatalitiesexcludingsuicide/homicidewereanalyzedseparately.Theimpactofcaseswithoutdateofdeathwasassessedbycombiningallcasesandusingreportreceiptdateasaproxyfordateofdeath.Toassesstheimpactofcasescontainingmissingornonspecificformulationinformation,fatalitychangeswerecalcu-latedforcasesinwhichthereportermentionedbrandname“OxyContin.”Asensitivityanalysisassessedtheimpactofdelayedreportingbyremovingcasesthatwerereportedmorethan3months(or6months)aftereachquarterinthestudyperiod.

TheratioofdeathsassociatedwithEROreportedtothemanufacturerversusalloxycodonedeathsreportedtotheFDA’sAERSsystem(dataavailablethrough4Q2012)wascalculated.DateofdeathisnotincludedintheFDAAERSdatabecauseofprivacyregulations;therefore,reportreceiptdatebytheFDAwasusedintheanalysis.

RESULTS

Populationcharacteristics

Atotalof326uniquefatalitiesinvolvingERO,origi-natingintheUSA,werespontaneouslyreportedto

themanufacturerwithamonth/yearofdeathfrom3Q2009–2Q2013(Table2).Overdosewasmentionedin240reports,andabuse-relatedbehaviorwasmen-tionedin206reports.Reportsinvolvingfatalover-dosesweremostfrequentlyreceivedfromahealthcareprofessional,morefrequentlyinvolvinganadult(age18–64years)andofteninvolvingpolysubstanceuse.

Overall,therewerenolargedifferencesinreportcharacteristicsbetweenthepre-reformulationandpost-reformulationperiodsintermsofreportsource,gender,age,reportsource/type,andsourceoftoxicol-ogyinformation.However,forfataloverdoses,nota-bledecreasesintheproportionofreportsfromsouthernregions(40%to29%)andthosewithmentionsofbenzodiazepines(42%to33%)orotheropioids(37%to24%)wereobserved.

Decrease in reports offatalities

TherewasareductioninreportsoffatalitiesinvolvingEROinthepost-reformulationperiods,particularlyforthesubsetofcasesofoverdoseandoverdosewithmentionofabuse(Figure1).Thesereductionsbeganthefirstyearpost-reformulationandweremorepro-nouncedinsubsequentyears.Specifically,themeanofallreportsoffatalitiesintheyearpre-reformulationwas32.8perquarter,whichdecreasedby82%(95%CI:-89%to-73%)to5.8reportsperquarterinthethirdyearpost-reformulation;themeanoffatalityre-portsinvolvingoverdoseintheyearpre-reformulationwas26.0perquarter,whichdecreasedby87%(95%CI:-93%to-78%)to3.3reportsperquarterinthethirdyearpost-reformulation;andthemeannumberoffatalityreportsinvolvingbothoverdoseandabuse-relatedbehaviorintheyearpre-reformulationwas

23.3perquarter,decreasingby86%(95%CI:-92%to-75%)to3.3reportsperquarterinthethirdyearpost-reformulation(Table3).

Increasingtrendsinmeanquarterlyfatalityreportswereobservedinthepre-reformulationyear.Incon-trast,inthepost-reformulationyears,theslopeofthe3-yeartrendforallEROfatalreportsdecreasedanaverageof15.6%(95%CI:-18.7%,-12.3%)perquarter,representingachangeof-20.7%(95%CI:

-31.3%to-8.5%)frompre-reformulationtopost-re-formulation, which was statistically significant(p=0.0015).Similarstatisticallysignificantchangesinquarterly slopeswereobservedforreportsoffatali-ties involving overdose (-22.9%; CI: -34.7 % to

-8.9%,p=0.0022) and of fatalitiesinvolvingbothoverdoseand abuse (-22.2%; CI: -34.9% to -7%,p=0.0058).Changesfornon-overdosefatalities(e.g.,

Table2. Characteristics ofextended-releaseoxycodonefatalityreportsreceivedbymanufacturerwithdateofdeathduring1-yearperiodbeforeand3-yearperiodafterintroductionofreformulatedextended-releaseoxycodone

Allfatalcases(N=326)Subsetoffatalcasesofoverdose(N=240)

Pre-reformulation(3Q2009–2Q2010) / Post-reformulation(3Q2010–2Q2013) / Pre-reformulation(3Q2009–2Q2010) / Post-reformulation(3Q2010–2Q2013)
Fatalityreports
Total(N) / 131 / 195 / 104 / 136
Gender
Male / 63% / 66% / 65% / 68%
Female / 37% / 33% / 35% / 32%
Unknown / 0% / 1% / 0% / 1%
Agedistribution
13years / 2% / 6% / 3% / 7%
13to18years / 5% / 6% / 6% / 9%
18to65years / 69% / 68% / 77% / 71%
65years orolder / 6% / 3% / 3% / 1%
Unknown / 18% / 17% / 12% / 13%
Reportertype
Healthcareprofessional / 60% / 50% / 61% / 54%
Other / 40% / 50% / 39% / 46%
Reporterregion
Northeast / 17% / 20% / 15% / 18%
Midwest / 16% / 19% / 17% / 20%
South / 39% / 30% / 40% / 29%
West / 18% / 17% / 18% / 18%
Notreported(missing) / 10% / 13% / 9% / 14%
Oxycodonemention
OxyContin / 52% / 52% / 44% / 41%
oxycodoneNOS* / 48% / 48% / 54% / 57%
Otherproductmentions
Alcohol / 15% / 13% / 19% / 19%
Benzodiazepine / 34% / 23% / 42% / 33%
Opioid / 30% / 18% / 37% / 24%
Muscle relaxant/hypnotic / 7% / 8% / 9% / 12%
Illicit† / 18% / 16% / 22% / 21%
Anyofabove / 58% / 44% / 73% / 60%
AdditionalfindingsAutopsyreports / 37% / 35% / 43% / 46%
Toxicologyreports / 44% / 40% / 49% / 46%

*Oxycodonenototherwisespecified.Reportsinvolvingoxycodonetabletsthatdonotspecifyformulation(e.g.,immediate-releaseorextended-releaseformulation)areimpliedtohaveinvolvedextended-releaseoxycodone(OxyContin)becausethereporterhastakenthetimetospecificallytransmittheinformation tothemanufacturer. Duringthe evaluationperiodofthis study, only extended-release oxycodonewas soldbythe manufacturer,andno genericextended-release oxycodoneproduct was approved or sold.

†Illicitdrugsincludemarijuana,cocaine,amphetamines,andheroin.

deathnototherwisespecified,suicide/homicide,andcancer)werenotstatisticallysignificant (-15.8%;CI:-36.8%to12.3%,p=0.2421)buttrendeddown.

Trends in comparators

Nosubstantialchangeinadverseeventcasehandlingorpharmacovigilanceproceduresweremadebythemanufacturerduringthestudyperiod.Non-fatalre-portstothemanufacturer forEROwere384perquar-terintheyearpre-reformulationcomparedwith3129, 395, and294perquarterinthefirst,second,andthirdyearpost-reformulation,respectively.Thesecompara-tor results suggest that the reductions in fatalities

involvingEROpost-reformulationwerenotduetotemporalchangesinreportingpatterns.

Aspikeinadverseeventreportsappearedshortlyafterreformulation,mostofwhichoccurredwithin3monthsofthemarketplacetransition.Asurveyof1967subjectswhoreportedadverseeventsatthattimeindicatedthat93%werefromindividualswhohadusedEROforsometimeandwerereportingchangesfromwhattheywereaccustomedto.Thetransitiontothereformulationwasconductedwithoutnotificationofthegeneralpublic.

ReportsoffatalitiestothemanufacturerforERmor-phine(MSContin®)weretoofewtoprovideastatisti-calcomparatortrend(2.7,1.5,2.5,and2.0perquarter

Figure1. Numberof extended-release oxycodone(ERO)fatality reports per quarter. Categories entitled overdoseand overdose with mention of abuse-relatedbehavioraredefinedinmethods.DistributionofreformulatedEROtowholesalerswasinitiated9August2010(indicatedbythearrow).

intheyearpre-reformulation,andfirst,second,andthirdyearpost-reformulation,respectively),thoughtherewasnosubstantialdecrease.

Fatalityreportsforextended-releaseoxycodoneversusalloxycodone

TheratioofthenumbersoffatalitiesinvolvingEROreportedtothemanufacturerrelativetofatalitieswithanyoxycodonecategorizedassuspectdrugreportedtoFDAdecreasedsignificantly(p0.0001)from21%(131/637)intheyearpre-reformulationto22%(122/551),8%(50/616),and10%(12/120)inthefirst,second,andfirst6monthsofthirdyearpost-reformu-

lation,respectively.

Sensitivityanalyses

Toassesstherobustnessoftheprimaryresults,sensi-tivityanalyseswereconductedadjustingforthenum-berofdispensedEROprescriptions,missingdateofdeathinformation,reportertype,reportersource,for-mulation specificity, andreporting timelag(Table 3).

Relativetothepre-reformulationyear,thenumberofEROprescriptionsdispensedinretail,long-termcare,andmail-orderpharmaciesdecreasedby2%from1.72millionto1.69millionperquarterinthefirst,by9%to1.57millionperquarterinthesecond,

andby12%to1.51millionperquarterinthethirdpost-reformulationyear(Figure2).Decreasingtrendsinfatalitycountsweredetectably,butnotsubstan-tially, altered when adjusted for ERO prescriptionnumbers.Theprescription-adjustedrateofallfatalityreportsdecreasedby80%(95%CI:-87%to-69%)comparingtheyearpre-reformulationtothethirdyearpost-reformulation.Significantdecreasesinreportedfataloverdosecasesandfataloverdosecasesthatalsomentionedabuse-relatedbehaviorwerealsoobserved.Limiting the analysis to reports received fromhealthcareprofessionals,confiningtheanalysistofa-talitycaseswherebrandname“OxyContin”wasmen-tioned,removingfatalityreportswithtextualmentionofsuicideorhomicide(42cases),orinclusionofpost-marketingstudies(82cases)didnotchangetheresultssubstantially.Imposingaconsistentreportinglagperiodof3or6monthsforallquartersacrossthestudyperiodshowedsimilardecreasesinfatalityre-portspost-reformulation,suggestingthattheobserved

decreaseswerenotaffectedbyareportinglag.

Duringthestudyperiod,themanufacturerreceived376fatalitycasesmissingthedateofdeath.Incompar-isonwithreportsthatincludedthisinformation,thesereports,ingeneral,lackeddetailedinformationregard-ingautopsyfindings,toxicologyresults,patientage,and concomitant drugs. Analysis combining the

Table3. Changesinthenumberofextended-releaseoxycodone fatalityreportsperquarterreceivedbythemanufacturerfrom1yearbeforeto3yearsafterintroductionofreformulatedextended-releaseoxycodone

1-yearpre-reformulation(3Q2009–2Q2010)

Firstyearpost-reformulation(3Q2010–2Q2011)

Secondyearpost-reformulation(3Q2011–2Q2012)

Thirdyearpost-reformulation(3Q2012–2Q2013)

%change / %change / %change
Mean* / Mean / (95%CI) / Mean / (95%CI) / Mean / (95%CI)

Cases withdateofdeathreported(n=326)Allfatalreports

All / 32.8 / 30.5 / -7(-27,19) / 12.5 / -62(-72,-47) / 5.8 / -82(-89,-73)
Overdose / 26.0 / 21.0 / -19(-39,8) / 9.8 / -62(-74,-46) / 3.3 / -87(-93,-78)
Abuse-relatedbehavior / 23.3 / 17.5 / -25(-45,3) / 7.5 / -68(-79,-51) / 3.3 / -86(-92,-75)

Non-overdose6.89.541(-14,130)2.8-59 (-80, -18)2.5-63 (-82, -23)

Allfatal reports, per100000prescriptions of OxyContin†

All1.9031.802-5(-26,21)0.794-58(-70,-42)0.380-80(-87,-69)

Overdose1.5161.241-18(-38,10)0.619-59 (-72, -41)0.213-86 (-92, -75)

Abuse-relatedbehavior1.3591.033-23(-44,5)0.475-65 (-77, -47)0.213-84 (-91, -72)

Non-overdose0.3870.56143(-12,135)0.176-55 (-78, -10)0.167-58(-80, -13)

Subsetofallfatalreportsfromhealthcareprofessionals

All19.815.8-20(-43,11)6.0-70 (-81, -52)2.8-86 (-93, -74)

Overdose15.811.329(-51,5)5.0-68 (-81, -42)2.0-87 (-94, -74)

Abuse-relatedbehavior14.811.0-25(-50,10)4.0-73 (-84, -53)2.0-86 (-94, -72)

Non-overdose4.04.512(-43,121)1.0-75 (-92, -25)0.8-81(-95, -36)

Subsetofallfatalreportsmentioningbrandname“OxyContin”

All17.015.8-7(-34,31)6.5-62(-76,-40)3.0-82(-90,-67)

Overdose12.09.0-25(-51,16)4.3-65 (-80, -38)1.3-90 (-96, -74)

Abuse-relatedbehavior10.57.8-26(-54,17)3.3-69 (-83, -42)1.3-88 (-95, -70)

Non-overdose56.835(-24,141)2.3-55 (-80, -1)1.8-65 (-85, -17)

Subsetofall fatal reportswith dataconfined tocasesreceivedduring3-monthperiodfollowingdate ofdeath

All11.518.057(8,127)4.0-65 (-80, -39)4.0-65 (-80, -39)

Overdose9.510.511(-29,71)3.3-66 (-82, -36)2.0-79 (-90, -55)

Abuse-relatedbehavior9.08.5-6(-41,51)2.8-69 (-84, -40)2.0-78 (-90, -52)

Non-overdose2.07.5275(72,718)0.8-62(-90,41)2.00(-62,166)

Subsetofall fatal reportswith dataconfined tocasesreceivedduring6-monthperiodfollowingdate ofdeath

All17.321.525(-9,71)7.3-58 (-73, -35)5.0-71 (-82, -52)

Overdose13.813.5-2(-33,43)5.3-62 (-77, -37)2.5-82 (-91, -64)

Abuse-relatedbehavior12.811.0-14(-42,29)4.8-63 (-78, -37)2.5-80(-90,61)

Non-overdose3.58.0129(22,328)2.0-43(-76,36)2.5-29 (-68, -61)

Caseswithdateofdeathreported(n=326)andnotreported(n=376)

Fatal reports with date of death+fatal reportswithout date of death (usingmanufacturerreceipt date as proxy)

All55.359.37(-11,29)31.8-43 (-54, -29)29.3-47 (-58, -34)

Overdose40.339.8-1(-21,23)22.8-43 (-56, -27)17.0-58 (-68, -44)

Abuse-relatedbehavior31.829.0-9(-29,17)16.3-49 (-62, -31)11.3-65 (-75, -50)

Non-overdose22.326.519(-10,58)11.0-51 (-66, -29)10.5-53 (-67, -32)

*Meannumberoffatalitycases perquarterwithvalues rounded uptoone decimal.

†IMSNationalPrescriptionAuditdatabase(includesretail,mailorderandlong-termcarepharmacyprescriptions).

fatalityreportswithoutdateofdeath,usingmanufac-turerreport-receiptdateasaproxyfordateofdeath,withthosereportscontainingdateofdeath,showeda47%(95%CI:-58%to-34%)decreaseinallfatalityreportsinthethirdyearpost-reformulation.Significantdecreasesinthenumberofoverdosefatalities,over-dosefatalitieswithmentionofabuse-relatedbehavior,andnon-overdosefatalitieswerealso observed.

DISCUSSION

ThenumberofspontaneousreportsofdeathinvolvingERO reported to the manufacturer decreased after

introductionofareformulatedEROthatwasdesignedtobeabuse-deterrent.Asvoluntaryspontaneousad-versereportsdonotcaptureallevents,temporaltrendsinthesereportsmaynotbeareliablesourceforcausalinterpretation.39–41However,thelargemagnitudeofthedecreaseinreportedfatalities,whilenon-fatalityadverseeventsremainedunchangedorincreased,sug-geststhedecreasewasnotduetochangingprocessesforreporting/collectingofEROadverseevents.Therelativestabilityofdeathreportsforanotheropioidproductduringthesameperiod,thoughmuchfewer(becauseoffewerprescriptions),suggeststhattherewasnosystematicprocesschangeforreportingor

Figure2.OxyContinprescriptionsperquarter.Retail,mailorder,andlong-termcarepharmacydispensingofOxyContinprescriptionsfor3Q2009–2Q2013extractedfromIMSNationalPrescriptionAuditdata-base. DistributionofreformulatedOxyContin towholesalerswasinitiated9August2010(indicatedbythearrow).

collectingreportsofdeathsbythemanufacturerduringthisperiod.Furthermore,analysesconductedtoassesswhethermethodologicalartifactsmightberesponsibleforthedeclineshowedlittleimpactofpotentialarti-facts,suchaschangesinprescriptionnumbers,re-porter type orsource, formulationspecificity,missingdateof death,andreportingtime lag. Therefore, theseresultssuggestadecreaseinthenumberoffatalitiesassociatedwithEROabuse/misuseasaresultofitsre-formulationwithphysicochemicalpropertiesthatdetercrushing/dissolving.

Otherstudieshavereportedsimilardecreasesinabuse/misuseofERO post-reformulation. Ciceroetal.29reportedthatinpatientswithopioiddepen-dencewho wereenteringtreatment,thechoiceofEROastheprimarydrugofabuseanduseofEROtogethighatleastonceinthelast30days,decreasedsignificantlypost-reformulation(SurveyofKeyInfor-mantsPatients’ProgramoftheRADARS®System).Butleretal.30reportedthatforindividualsassessedforsubstanceabusetreatment,oralandnon-oralabuseofreformulatedEROwere17%and66%lower,re-spectively,thanhistoricabuseoforiginalERO(NAVIPPROSurveillanceSystem),whereasabuse ofERoxymorphoneandERmorphineincreasedorremainedrelativelyunchanged,respectively.Severtsonetal.31reportedthatEROabuseexposurecallstopoisoncentersandreportsofdiversionofERO,onaper-catchmentareapopulationbasis,de-creased38%and53%,respectively,andthatthestreetpriceforreformulatedEROwassignificantlylowerthanoriginalERO(PoisonCenterStudyandDrugDiversionProgramof RADARS System). Coplanetal.32reportedthatpoisoncenterEROabuse,suicide,therapeuticerrorsamongpatients,accidentalexpo-suresamongchildren, and adversereaction exposures

decreasedsignificantlypost-reformulationbutin-creasedorremainedsteadyforothersingle-entityoxy-codoneproducts(NationalPoisonDataSystem).Havensetal.42reportedthatexperiencedopioidabusersinaruralKentuckycountyself-reportedalowfrequencyofabuseofreformulatedEROwhilemaintainingaconsistentlyhighfrequencyof abuseofIRoxycodone.TheFloridaMedicalExaminersCom-missionReportshowedadecreaseindeathsduetoallformsofoxycodone,bothERandIRformulations,fromapeakof1516in2010to735in2012.43How-ever,duringthisperiod,astatelawwasenactedtoim-posestricterrequirementsfordispensingofcontrolledsubstancestargetedspecificallyatpillmillsthatwereprimarilyprescribingordispensingIRoxycodone.50Nationally,theratiooffatalitiesinvolvingEROversusanyoxycodonemorethanhalved(from21%to8%)afterreformulation,furtherendorsingthespecificityoftheeffectof thereformulation.

Otherresearchershaveusedspontaneousadverseeventreportstoassesstheeffectsofopioidformula-tionsintendedtodeterabuse.Post-marketingadverseeventreportsreceivedforanERmorphinecapsulecombinedwithnaltrexone(anopioidantagonist)indi-catedalownumberofproducttamperingreportsandnocasesofconfirmedtamperingresultinginfatality.44Therearelimitationstocausalinferencefromvolun-taryspontaneousfatalityreports.However,whentheresultsofthisstudyareconsideredinthecontextofthefiveadditionalpublishedarticlesandtheonestatesurveillancereportthatdemonstrateanabuse-deterrentimpactofreformulatedERO,theBayesianpriorthattheobservedchangeinreportedfatalitiesiscaused

bythereformulationisincreased.45

Alternateexplanations forthedecline infatalityreportingwereconsidered.AreductioninEROdis-pensingoccurredduringthestudyperiod.However,combinedprescriptionreductionsoforiginalandreformulatedERO(2%,9%,and12%inthefirst,second,andthirdyearpost-reformulation,respectively)wereinsufficienttoaccountforthemuchlargerreduc-tioninfatalities.Dispensingoforiginal ERO graduallydecreasedinthe18monthsafterwholesaleshipmentsstopped.Ofnote,reportsreceivedbythemanufacturerpost-reformulationincludedexposurestobothoriginalandreformulatedEROandthereforeaportionofthefatalreportsinthepost-periodlikelyinvolvedoriginalERO,whichmayaccountforthegradualdecreaseinfatalitiespost-reformulation.

Severalinitiativestodeteropioidabuse/overdosecommencedorwereongoingduringthestudyperiod.TheFDA’sRiskEvaluationandMitigationStrategyforERandlong-actingopioidswasapprovedinJuly

%0.46However,itsprimarycomponentiscontinuingeducationofprescribers, whichbeganinMarch 2013,subsequenttothereductioninEROdeaths.Stateprescriptiondrugmonitoringprogramswereoperatingorinitiatedduringthestudyperiod.47Preliminaryevaluationoftheseprogramsindicateapositiveimpactonopioidabuse/misuse,48buttheireffectonfatalitiesisnotclear.49,50Community-basedopioidoverdosepreventionprogramshavereduceddeathsinthefewlocalregionswhereimplementedbutcannotaccountforlargenationalchanges.51,52DrugdisposalprogramshavebeenestablishedbuthavenotfocusedspecificallyonEROandhavenotbeenshowntoimpactdeathsassociatedwithopioids.53,54Theseinitiativesmayhavecontributedtodecreasesinopioid-relatedfatalities;however,theyareun-likelytoaccountforthelevelofdecreaseinEROfatalities.TheratiooffatalitiesinvolvingEROversusanyoxycodonemorethan halvedafterreformulation.

Someauthors attributedanincrease in heroinabusetothereformulationofEROwhenitwasfirstintro-duced,29andthishasledtowidespreadattributionofthecauseoftherapidlyescalatingheroinabusetoEROinmediareports.55However,currently,1.7%ofindividualsdispensedopioidanalgesicsintheUSAreceiveERO,and2.5%ofopioidprescriptionsdis-pensedareforERO.Theseproportionshaveremainedroughlyconstantoverthepast5years(IMSHealth,unpublisheddata).Furthermore,approximately4.3%ofdiagnosedoverdoseeventsininsuranceclaimsdata-basesin2011wereamongpeopleprescribedERO(MarketScan,unpublisheddata)andtheprevalenceofreformulatedEROabuseamongprescriptionopioidabusersindrugtreatmentcentersis12.1%.22There-fore,itisunlikelythatreducedabuseofERO,whichoccurredinasingle,abruptinterventionbeginninginAugust2010,couldaccountforspikesinheroinabuse3½yearslater.

Inconclusion,thenumberofspontaneousreportsoffatalitiesinvolvingEROhassignificantlyde-creasedafteritsreformulation,whereasnon-fatalreports involvingEROremainedunchangedorincreased.Thesefindings,whenconsideredinthecontextofpreviouslypublishedstudiesusingothersurveillancesystems,suggestthattheabuse-deterrentcharacteristicsofreformulatedEROhavedecreasedfatalitiesassociatedwithitsmisuse/abuse.Abuse-deterrentformulationsmaybeavaluableriskmanagementtool,suchthatinnovation,policing,regulation,carefulprescribing,andeducationcanbecombinedtomitigateseriousriskandimprovethebenefit-riskbalanceofopioidanalgesics.56

CONFLICTOFINTEREST

NelsonSessler,JerodDowning,HrishikeshKale,HowardChilcoat,ToddBaumgartner,andPaulCoplanarefull-timeemployeesofPurduePharmaL.P.

ETHICSSTATEMENT

Theauthorsstatethatnoethicalapprovalwasneeded.ACKNOWLEDGEMENTS

Casereview assistancewasprovided byBarbaraHar-ding.ManuscriptpreparationassistancewasprovidedbyMelindaPhilbrook.Writingassistancewaspro-videdbyLouisAlexander.ThisstudywassponsoredbyPurduePharmaL.P.

REFERENCES

1.AmericanAcademyofPainMedicine.UseofOpioidsfortheTreatmentofChronicPain.AmericanAcademyofPainMedicine:Chicago,IL, 2013.

2.AmericanGeriatricsSocietyPanelonPharmacologicalManagementofPersis-tentPaininOlderPersons.Pharmacologicalmanagementofpersistentpaininolderpersons.JAmGeriatrSoc2009;57:1331–1346.doi.10.1111/j.1532-5415.2009.02376.x.

3.ChouR,FanciulloGJ,FinePG,etal.Clinicalguidelinesfortheuseofchronicopioidtherapyinchronicnoncancerpain.JPain2009;10:113–130. doi:10.1016/j.jpain.2008.10.008.

4.ManagementofOpioidTherapyforChronicPainWorkingGroup.VA/DoDclinicalpracticeguidelineformanagementofopioidtherapyforchronicpain.DepartmentofVeteransAffairs,DepartmentofDefense:Washington,DC,2010.

5.NationalOpioidUseGuidelineGroup.Canadianguidelineforsafeandeffectiveuseofopioidsforchronicnon-cancerpain.PartB:recommendationsforpractice.NationalOpioidUse GuidelineGroup (NOUGG):Hamilton, Canada, 2010.

6.JonesCM,MackKA,PaulozziLJ.Pharmaceuticaloverdosedeaths,UnitedStates,2010.JAMA2013;309: 657–659. doi:10.1001/jama.2013.272.

7.OkieS.Afloodofopioids,arisingtideofdeaths.NEnglJMed2010;363:1981–1983.doi:10.1056/NEJMp1011512.

8.WhiteAG,BirnbaumHG,MarevaMN,etal.Directcostsofopioidabuseinaninsuredpopulationinthe United States.JManagCarePharm 2005; 11:469–479.

9.BirnbaumHG,WhiteAG,ReynoldsJL,etal.EstimatedcostsofprescriptionopioidanalgesicabuseintheUnitedStatesin2001:asocietalperspective.ClinJ Pain2006; 22: 667–676.

10.RaffaRB,PergolizziJV.Opioidformulationsdesignedtoresist/deterabuse.

Drugs2010;70:1657–1675.doi:10.2165/11537940-000000000-00000.

11.KatzN,DartRC,BaileyE,etal.Tamperingwithprescriptionopioids:natureandextentoftheproblem,healthconsequences,andsolutions.AmJDrugAlco-holAbuse 2011;37:205–217.doi:10.3109/00952990.2011.569623.

12.NationalDrug IntelligenceCenter. OxyContindiversionand abuse.January2001. 2012].

13.HaysL,KirshKL,PassikSD.SeekingdrugtreatmentforOxyContinabuse:achartreviewofconsecutiveadmissionstoasubstanceabusetreatmentfacilityinKentucky.JNatlComprCancNetw2003;1:423–428.

14.HaysLR.AprofileofOxyContinaddiction.JAddictDis2004;23:1–9.

15.ButlerSF,BlackRA,GrimesSerranoJM,etal.Characteristicsofprescriptionopioidabusersintreatment:prescriptionopioidusehistory,age,usepatternsandfunctionalseverity.JOpioidManag2010;6:239–252.

16.RomachMB,SchoedelKA,SellerEM.Updateontamper-resistantdrugformulations.DrugAlcoholDepend2013;130:13–23.doi:10.1016/j.drugalcdep.2012.12.028.

17.Moorman-LiR,MotyckaCA,IngeLD,etal.Areviewofabuse-deterrentopi-oidsforchronicnonmalignant pain.PT2012;37: 412–418.

18.OxyContin[fullprescribing information].Stamford,CT:PurduePharmaL.P.7/2013.

19.CentersforDiseaseControlandPrevention.CDCgrandrounds:prescriptiondrugoverdoses—aU.S.Epidemic. MorbMortal WklyRep2012;61:10–13.

20.ConeEJ,FantRV,RohayJM,etal.Oxycodoneinvolvementindrugabusedeaths:aDAWN-basedclassificationschemeappliedtoanoxycodonepostmor-temdatabasecontainingover1000cases.JAnalToxicol2003;27:57–67;discussion 67.

21.DhallaIA,MamdaniMM,SivilottiML,etal.Prescribingofopioidanalgesicsandrelatedmortalitybeforeandaftertheintroductionoflong-actingoxycodone.CMAJ2009;181:891–896.doi:10.1503/cmaj.090784.

22.ButlerSF,BlackRA,CassidyTA,etal.Abuserisksandroutesofadministrationofdifferentprescriptionopioidcompoundsandformulations.HarmReductJ2011;8:29–45.doi:10.1186/1477-7517-8-29.

23.ButlerSF,BudmanSH,LicariA,etal.Nationaladdictionsvigilanceinterven-tion andprevention program (NAVIPPRO):a real-time,product-specificpublichealthsurveillancesystemformonitoringprescriptiondrugabuse.PharmacoepidemiolDrugSaf2008;17:1142–1154.doi:10.1002/pds.1659.

24.KatzN, FernandezK,ChangA,etal.Internet-basedsurveyofnonmedicalpre-scriptionopioiduseintheUnitedStates.ClinJPain2008;24:528–535.doi:10.1097/AJP.0b013e318167a087.

25.PassikSD,HaysL,EisnerN,etal.Psychiatricandpaincharacteristicsofpre-scriptiondrugabusersenteringdrugrehabilitation.JPainPalliatCarePharmacother2006;20: 5–13.

26.ConeEJ,GiordanoJ,WeingartenB.Aniterativemodelforinvitrolaboratoryassessmentoftamperdeterrentformulations.DrugAlcoholDepend2013;131:100–105.doi:10.1016/j.drugalcdep.2012.12.006.

27.SellersEM,PerrinoPJ,ColucciSV,HarrisSC.AttractivenessofreformulatedOxyContin(R)tablets:assessingcomparativepreferencesandtamperingpotential.JPsychopharmacol2013;27(9):808–816.doi: 10.1177/0269881113493364.

28.PerrinoPJ,ColucciSV,ApseloffG,etal.Pharmacokinetics,tolerability,andsafetyofintranasaladministrationofreformulatedOxyContin®tabletscomparedwithoriginalOxyContin®tabletsinhealthyadults.ClinDrugInvestig2013;33:441–449.doi:10.1007/s40261-013-0085-x.

29.CiceroTJ,EllisMS,SurrattHL.Effectofabuse-deterrentformulationofOxyContin.N EnglJ Med2012; 367: 187–189.doi:10.1056/NEJMc1204141.

30.ButlerSF,CassidyTA,ChilcoatH,etal.Abuseratesandroutesofadministra-tionofreformulatedextended-releaseoxycodone:initialfindingsfromasentinelsurveillancesampleofindividualsassessedforsubstanceabusetreatment.JPain2013;14:351–358.doi:10.1016/j.jpain.2012.08.008.

31.Severtson SG, Bucher BartelsonB, Davis JM, etal. Reduced abuse,therapeuticerrors,anddiversionfollowingreformulationofOxyContin®in2010.JPain2013;14:1122–1130.doi:10.1016/j.jpain.2013.04.011.

32.CoplanPM,KaleH,SandstromL,etal.Changesinoxycodoneandheroinexpo-suresinthenationalpoisondatasystemafterintroductionofextended-releaseoxycodonewithabusedeterrentcharacteristics.PharmacoepidemiolDrugSaf2013;22:1274–1282.doi:10.1002/pds.3522.

33.FoodandDrugandAdministration.2013b.FDAnewsrelease:FDAapprovesabuse-deterrentlabelingforreformulatedOxyContin,Availablefrom:

34.KesslerDA.IntroducingMEDWatch.Anewapproachtoreportingmedicationanddeviceadverseeffectsandproductproblems.JAMA1993;269:2765–2768.

35.U.S.FoodandDrugandAdministration(FDA).Adverseeventreportingsystem(AERS).Rockville,MD:FDA.Availablefrom: 2013].

36.GreenlandS.Doseresponseandtrendanalysisinepidemiology:alternativestocategoricalanalysis.Epidemiology1995;6:356–365.

37.RothmanKJ,GreenlandS,LashTL.ModernEpidemiology(3rdedn).Lippincott, Williams &Wilkins:Philadelphia, PA, 2008; 410.

38.IMShealthnationalprescriptionauditdatapurchasedandreceivedDecember2013

39.AhmadSR.AdversedrugeventmonitoringattheFoodandDrugAdministra-tion.JGenInternMed2003;18:57–60.

40.GibbonsRD,AmatyaAK,BrownCH,etal.Post-approvaldrugsafetysurveil-lance.AnnuRevPublicHealth2010;31:419–437.doi:10.1146/annurev.publhealth.012809.103649.

41.Rodriguez EM,StaffaA, GrahamD. The roleof databasesindrugpost-marketingsurveillance.PharmacoepidemiolDrugSaf2001;10:407–410.doi:10.1002/pds.615.

42.HavensJR,LeukefeldCG,DeVeaugh-GeissAM,CoplanPM,ChilcoatHD.Theimpactofareformulationofextended-releaseoxycodonedesignedtodeterabuseinasampleofprescriptionopioidabusers.DrugAlcoholDepend[availableahead ofprint27February2014]doi:10.1016/j.drugalcdep.2014.02.018.

43.Floridamedicalexaminerscommissionreportfor2012.Availableat:

44.BadalamentiVC,BuckleyJW,SmithET.SafetyofEMBEDA(morphinesulfateandnaltrexonehydrochloride)extended-releasecapsules:reviewofpost-marketingadverseeventsduringthefirstyear.JOpioidManag2012;8:115–125.

45.PearlJ.Causality:Models,ReasoningandInference.CambridgeUniversityPress:Cambridge, United Kingdom,2009; 5.

46.FDANewsRelease.FDAintroducesnewsafetymeasuresforextended-releaseandlong-actingopioidmedications.July9,2012.Availableat:

47.U.S.DepartmentofJusticeDrugEnforcementAdministration.Prescriptiondrugmonitoringprograms.Availableat:

48.ReiflerLM,DrozD,BaileyJE,etal.Doprescriptionmonitoringprogramsim-pactstatetrendsinopioidabuse/misuse?PainMed2012;13:434–442.doi:10.1111/j.1526-4637.2012.01327.x.

49.PaulozziLJ,KilbourneEM,DesaiHA.Prescriptiondrugmonitoringprogramsanddeathratesfromdrugoverdose.PainMed2011;12:747–754.doi:10.1111/j.1526-4637.2011.01062.x.

50.Sauber-SchatzEK,MackKA,DiekmanST,etal.Associationsbetweenpainclinicdensityanddistributionsofopioidpainrelievers,drug-relateddeaths,hos-pitalizations,emergencydepartmentvisits,andneonatalabstinencesyndromeinFlorida.DrugAlcoholDepend2013;133:161–166.doi:10.1016/j.drugalcdep.2013.05.017.

51.AlbertS,BrasonFW,SanfordCK,etal.ProjectLazaruscommunity-basedover-dosepreventioninruralNorthCarolina.PainMed2011;12(Suppl2):S77–S85.doi:10.1111/j.1526-4637.2011.01128.x.

52.Centersfor Disease ControlandPrevention.Community-basedopioidoverdosepreventionprogramsprovidingnaloxone—UnitedStates,2010.MMWRMorbMortal WklRep2012;61:101–105.

53.Nationaltake-backinitiative.USDepartmentofJustice,DrugEnforcementAd-ministration,OfficeofDiversionControlWebsite.

54.GrayJA.PrescriptiondrugabuseandDEA-sanctioneddrugtake-backevents:characteristicsandoutcomesinruralAppalachia.ArchInternMed2012;172:1186–1187.doi:10.1001/archinternmed.2012.2374.

55.SilvermanG.Heroinabuse:shootingup.FinancialTimes,February9,2014.Availableat

56.Office of NationalControlPolicy. Epidemic:responding to America’sprescrip-tiondrugabusecrisis,2011. March 2013].