Assessing the Evidence Base Series

Medication-Assisted Treatment WithBuprenorphine: Assessing the EvidenceCindy Parks Thomas, Ph.D.Catherine A. Fullerton, M.D., M.P.H.Meelee Kim, M.A.Leslie Montejano, M.A., C.C.R.P.D. Russell Lyman, Ph.D.

Richard H. Dougherty, Ph.D.Allen S. Daniels, Ed.D.Sushmita Shoma Ghose, Ph.D.Miriam E. Delphin-Rittmon, Ph.D.

Objective: Buprenorphine maintenance treatment (BMT) and meth-adone maintenance treatment (MMT) are pharmacological treatmentprograms for individuals with opioid use disorders. MMT is discussed ina companion article. This article describes BMT and reviews availableresearch on its efficacy. Methods: Two authors reviewed meta-analyses,systematic reviews, and individual studies of BMT from 1995 through2012. Databases surveyed were PubMed, PsycINFO, Applied SocialSciences Index and Abstracts, Sociological Abstracts, Social ServicesAbstracts, and Published International Literature on Traumatic Stress.They chose from three levels of evidence (high, moderate, and low) basedon benchmarks for the number of studies and quality of their method-ology. They also described the evidence of service effectiveness. Results:Sixteen adequately designed randomized controlled trials of BMT in-dicated a high level of evidence for its positive impact on treatment re-tention and illicit opioid use. Seven reviews or meta-analyses were alsoincluded. When the medication was dosed adequately, BMT and MMTshowed similar reduction in illicit opioid use, but BMT was associatedwith less risk of adverse events. Results suggested better treatment re-tention with MMT. BMT was associated with improved maternal andfetal outcomes in pregnancy, compared with no medication-assistedtreatment. Rates of neonatal abstinence syndrome were similar formothers treated with BMT and MMT during pregnancy, but symptomswere less severe for infants whose mothers were treated with BMT.Conclusions: BMT is associated with improved outcomes compared withplacebo for individuals and pregnant women with opioid use disorders.BMT should be considered for inclusion as a covered benefit. (PsychiatricServices in Advance, November 18, 2013; doi: 10.1176/appi.ps.201300256)

More than two million indi-viduals in the United Statesare addicted to opioids (1).

Two common options for pharmaco-logical maintenance treatment ofopioid dependence are the opioidagonistsmethadone and buprenorphine.Over 300,000 individuals receive meth-adone through outpatient treatmentprograms (2). Over half of these pro-grams and thousands of physicians nowoffer buprenorphine. Such pharmaco-logical treatment is typically provided incombination with psychosocial or othersupport services.

This article reports the results ofa literature review that was under-taken as part of the Assessing theEvidence Base Series (see box on nextpage). Methadone maintenance treat-ment (MMT) is reviewed in a com-panion article in this series (3). Asdiscussed in that review, research hasshown that MMT improves treatmentoutcomes for individuals with opioiddependence (4–7). However, MMT isassociated with serious adverse events,such as respiratory depression andcardiac arrhythmias (8–10). Becauseof concern about these adverse eventsand medication diversion, MMT isrestricted to dedicated opioid treat-ment programs that provide dailymedication dosing and offer psycho-social treatment services. In thisarticle, we review buprenorphinemaintenance treatment (BMT) asan alternative to MMT for the long-term management of opioid usedisorders.

For purposes of this initiative, theSubstance Abuse and Mental Health

Dr. Thomas and Ms. Kim are with the Heller School for Social Policy and Management,Brandeis University, Waltham, Massachusetts (e-mail: cthomas@brandeis.edu). Dr.Fullerton and Ms. Montejano are with Truven Health Analytics, Cambridge,Massachusetts. Dr. Lyman and Dr. Dougherty are with DMA Health Strategies,Lexington, Massachusetts. Dr. Daniels and Dr. Ghose are with Westat, Rockville,Maryland. Dr. Delphin-Rittmon is with the Office of Policy, Planning, and Innovation,Substance Abuse and Mental Health Services Administration (SAMHSA), Rockville. Thisarticle is part of a series of literature reviews that will be published in Psychiatric Servicesover the next several months. The reviews were commissioned by SAMHSA througha contract with Truven Health Analytics and were conducted by experts in each topicarea, who wrote the reviews along with authors from Truven Health Analytics, Westat,DMA Health Strategies, and SAMHSA. Each article in the series was peer reviewed bya special panel of Psychiatric Services reviewers.

PSYCHIATRIC SERVICES IN ADVANCE 1

Services Administration describesmedication-assisted treatment as adirect service that provides a personwho has a substance use or mentaldisorder with pharmacotherapy inconjunction with behavioral therapiesas treatment for associated symptomsor disabilities. BMT is a medication-assisted treatment that uses bupre-norphine or buprenorphine-naloxoneto treat individuals with an opioid usedisorder. A definition of medication-assisted treatment with buprenorphinefor opioid use disorders is presented inTable 1.

The objectives of this review wereto describe BMT and its primary andsecondary treatment goals, rate thelevel of evidence (methodologicalquality) of existing studies for thistreatment, describe the degree ofeffectiveness of this service on thebasis of the research literature, andcompare the relative advantages anddisadvantages of BMT and MMT.

Description of BMTBuprenorphine has been available asan injectable medication at low dosesto treat pain since the 1980s. In 2000,

Congress passed the Drug AbuseTreatment Act (DATA), whichallowed physicians to prescribe ap-proved medications for long-termopioid treatment in settings otherthan opioid treatment clinics, such asin office-based facilities (11). In 2002,the U.S. Food and Drug Administra-tion (FDA) approved high-dosesublingual formulations of buprenor-phine and buprenorphine-naloxonefor the treatment of opioid usedisorders (11,12). Naloxone induceswithdrawal symptoms if taken intra-venously but not if taken orally. Themanufacturer developed the combi-nation buprenorphine-naloxone med-ication to decrease the potential forabuse and diversion. Buprenorphineand buprenorphine-naloxone becamethe first medications to be approvedunder DATA and the first medica-tions available through DATA foroffice-based treatment of opioiddependence in the United States.Prescribing must be done within theguidelines of DATA, which requiresthat physicians receive specific train-ing and certification before prescrib-ing buprenorphine and that thenumber of patients they treat at onetime be limited to 100 (originally30 patients and amended in 2006)(13). In this review, we use buprenor-phine in reference to both buprenor-phine and buprenorphine-naloxonesublingual tablets. Although bupre-norphine can be used to managewithdrawal symptoms during acutedetoxification from opioids, BMTrefers to the maintenance use ofbuprenorphine to decrease illicit opi-oid use.

Because individuals remain depen-dent on buprenorphine, BMT is notconsidered an abstinence treatment.The goals of BMT are to reduce oreliminate illicit opioid use and, asa result, to decrease its associatednegative outcomes (Table 1). Thisassessment of the research will helpinform behavioral health policy lead-ers about the merits of BMT asdistinct from and in comparison toMMT. A summary of its value asa covered health benefit will also beof use to third-party payers, pro-viders, and people making personaldecisions about which medication touse.

Table 1

Description of medication-assisted treatment with buprenorphine

Feature Description

Service definition Medication-assisted treatment is a direct service that providesa person with a substance use or mental disorder withpharmacotherapy in conjunction with behavioral therapies astreatment for associated symptoms or disabilities. The natureof the services provided is determined by the person’s currentstatus or needs.

Buprenorphine maintenance therapy is a medication-assistedtreatment that uses buprenorphine or buprenorphine-nalox-one to help individuals with an opioid use disorder abstainfrom or decrease the use of illegal opioids (for example,intravenous heroin) or the use of opioids in a nonprescribedmanner (for example, abuse of prescription pain medications).

Service goals Retention in treatment; decrease in illegal opioid use; decrease inmortality; decrease in nonopioid drug use; decrease in criminalactivity; decrease in risk behaviors related to HIV and hepatitis C

Populations Adults with opioid use disorders; pregnant women with opioiduse disorders

Settings of servicedelivery

Office-based facilities; opioid treatment centers

About the AEB SeriesThe Assessing the Evidence Base (AEB) Series presents literature reviewsfor 14 commonly used, recovery-focused mental health and substance useservices. Authors evaluated research articles and reviews specific to eachservice that were published from 1995 through 2012 or 2013. Each AEBSeries article presents ratings of the strength of the evidence for the service,descriptions of service effectiveness, and recommendations for futureimplementation and research. The target audience includes state mentalhealth and substance use program directors and their senior staff, Medicaidstaff, other purchasers of health care services (for example, managed careorganizations and commercial insurance), leaders in community healthorganizations, providers, consumers and family members, and othersinterested in the empirical evidence base for these services. The researchwas sponsored by the Substance Abuse and Mental Health ServicesAdministration to help inform decisions about which services should becovered in public and commercially funded plans. Details about theresearch methodology and bases for the conclusions are included in theintroduction to the AEB Series (14).

2 PSYCHIATRIC SERVICES IN ADVANCE

MethodsSearch strategy

Two authors (CPT and CAF) withcomprehensive expertise in this topicconducted a literature search of majordatabases: PubMed (U.S. NationalLibrary of Medicine and National Insti-tutes of Health), PsycINFO (AmericanPsychological Association), Applied So-cial Sciences Index and Abstracts, So-ciological Abstracts, Social ServicesAbstracts, and Published InternationalLiterature on Traumatic Stress.They identified meta-analyses, re-

search reviews, clinical guidelines,and individual studies about BMTthat were published from 1995 through2012. They found additional literatureby examining the bibliographies ofmajor reviews and meta-analyses,major clinical texts, and professionalclinical society reviews. They relied onsystematic reviews and meta-analysesto summarize relevant findings fromearlier years. These review articleswere supplemented with individualrandomized controlled trails (RCTs)and quasi-experimental observationalstudies to provide additional informa-tion from recent years.The terms used to search the literature

were buprenorphine, buprenorphine/naloxone, opioid maintenance therapy,opioid treatment, addiction pharmaco-therapy, medication-assisted mainte-nance treatment, buprenorphinemaintenance therapy, and pregnancy.This review did not compare BMT tonaltrexone, another medication used inopioid maintenance treatment, becausethe literature review uncovered nostudies directly comparing the twomedications.

Inclusion and exclusion criteria

The two authors who conducted thesearch independently examined theabstracts of identified articles to de-termine compliance with the reviewinclusion and exclusion criteria. Theyaccepted articles on which they con-curred. They included the followingtypes of articles: RCTs, quasi-experimental studies, systematicreview articles, meta-analyses, andclinical guidelines; English-languagestudies conducted in the UnitedStates, including international studiesthat used U.S.-based sites and in-ternational reviews encompassing

U.S.-based studies; and studies thatfocused on BMT for individuals withopioid use disorders or the use ofBMT during pregnancy.

Excluded were case studies, cross-sectional studies, and those withsingle-subject designs. Also excludedwere studies that focused on bupre-norphine use for pain management orfor detoxification from opioids. Finally,reviews and meta-analyses that exam-ined only studies that did not meet theinclusion criteria were excluded.

Strength of the evidence

The methodology used to rate thestrength of the evidence is describedin detail in the introduction to thisseries (14). The authors who con-ducted the search independently ex-amined the research designs of thestudies identified during the literaturesearch. They chose from three levelsof evidence (high, moderate, and low)to indicate the overall research qualityof the collection of studies. Ratingswere based on predefined bench-marks that considered the number ofstudies and their methodological qual-ity. If the reviewers’ ratings were dis-similar (occurring for 13% of the studiesrated), the reviewers met to reach aconsensus opinion.

In general, high ratings indicateconfidence in the reported outcomesand are based on three or more RCTswith adequate designs or two RCTsplus two quasi-experimental studieswith adequate designs. Moderateratings indicate that there is someadequate research to judge the ser-vice, although it is possible that futureresearch could influence reportedresults. Moderate ratings are basedon the following three options: twoor more quasi-experimental studieswith adequate design; one quasi-experimental study plus one RCTwith adequate design; or at least twoRCTs with some methodologicalweaknesses or at least three quasi-experimental studies with some meth-odological weaknesses. Low ratingsindicate that research for this serviceis not adequate to draw evidence-based conclusions. Low ratings indicatethat studies have nonexperimentaldesigns, there are no RCTs, or thereis no more than one adequately de-signed quasi-experimental study.

The reviewers accounted for otherdesign factors that could increase ordecrease the evidence rating, such ashow the service, populations, and in-terventions were defined; use of statis-tical methods to account for baselinedifferences between experimental andcomparison groups; identification ofmoderating or confounding variableswith appropriate statistical controls;examination of attrition and follow-up;use of psychometrically sound mea-sures; and indications of potential re-search bias.

Effectiveness of the service

The reviewers described the effective-ness of the service—that is, how wellthe outcomes of the studies met theservice goals. They compiled the find-ings for separate outcomemeasures andstudy populations, summarized the re-sults, and noted differences across in-vestigations. They considered the qualityof the research design in their conclu-sions about the strength of the evidenceand the effectiveness of the service.

Results and discussionLevel of evidence

The literature search revealed 16 RCTs(15–30), a randomized cross-over study(31), a study using a self-administeredsurvey (32), and a retrospective de-scriptive study (33). Summaries ofthese studies are provided in Table 2.RCTs used either buprenorphinealone or buprenorphine-naloxone, asnoted in the table. The search alsofound seven reviews or meta-analyses(10,34–39), and summaries of theseare provided in Table 3.

Because of the large number oftrials, the overall evidence for BMTwas rated as high. Thus the level ofresearch evidence is similar for BMTand MMT (3). In addition, multiplemeta-analyses, reviews, and morethan three independent RCTs havecompared BMT with MMT on theprimary outcomes stated above, andthese results are also based on a highlevel of evidence in RCTs (19,20) orreviews (34,36). Secondary outcomes,such as use of other illicit drugs, cri-minal behaviors, and other measures ofaddiction severity or psychosocialfunctioning varied among studies; asa result, the evidence for these sec-ondary outcomes is not as strong.

PSYCHIATRIC SERVICES IN ADVANCE 3

Table 2

Individual studies of buprenorphine maintenance treatment (BMT) included in the reviewa

StudyDesign andobjectives

Population andconditions

Outcomesmeasured Summary of findings

Johnsonet al.,1995 (18)

RCT to assess earlyclinical effectivenessof buprenorphineversus placebo in anopioid-dependentpopulation

Patients randomlyassigned to placebo(N=60) or to 2 mg(N=60) or 8 mg (N=30) daily of sublingualbuprenorphine. Ondays 6–13, patientscould request a dosechange, knowing thatthe new dose would berandomly chosen fromthe 2 other alternatives.

Primary: percentage ofpatients in eachgroup requestinga dose change. Sec-ondary: positive urineopioid screens andpatient satisfactionwith treatment

Significant main effect of buprenorphineversus placebo. Patients takingbuprenorphine requested fewerdose changes (27% for 2 mg and32% for 8 mg versus 65% for placebo,p,.01). They also had fewer positiveurine drug screens (p,.05) and rateddose adequacy higher (p,.01). Effectswere significant for buprenorphineversus placebo but not for variousdoses.

Ling et al.,1996 (19)

RCT to evaluate safetyand efficacy of long-term, fixed-dose BMTversus low- and high-dose MMT

225 treatment-seekingpatients with opioiddependence ran-domly assigned toreceive 8 mg per dayof buprenorphine, 30mg per day of metha-done (low dose), or 80mg of MMT (highdose), all over a 1-year period

Primary: urine toxicology,retention, craving,and withdrawalsymptoms; safetydata

At 26 and 52 weeks, the high-doseMMT group had better retention(31% versus 20% at 52 weeks,p=.009) and less opioid use (p=.002)than the low-dose MMT or fixed-doseBMT groups. Results were compara-ble in the latter two groups. No seriousadverse health effects were noted for8 mg of buprenorphine.

Ling et al.,1998 (16)

RCT to evaluate safetyand efficacy of an 8mg per day sublingualdose of buprenorphineversus a 1 mg per daydose over a 16-weektreatment period in aheroin-dependentpopulation; second-ary analysis of 2 otherdose levels (4 mg and16 mg)

736 total patients in4 dose groups: 1 mg,N=185; 4 mg, N=182;8 mg, N=188; and16 mg, N=181. Totalof 375 completedthe full 16 treatmentweeks.

Primary: retention intreatment, illicit opioiduse as indicated byurine drug screens,opioid craving, andglobal ratings

For retention, 40% in 1-mg groupcompleted treatment, 51% in 4-mggroup, 52% in 8-mg group, and 61%in 16-mg group. The 1-mg grouphad poorer retention than the 8-mg(p=.019) or 16-mg (p,.001) groups.The 8-mg group had significantlyfewer positive screens than the 1-mggroup, less craving, and higher globalratings (p,.05).

O’Connoret al.,1998 (25)

RCT to evaluate theeffect of thrice weeklyBMT in a primary caresetting versus a tra-ditional treatmentfacility

46 patients assigned toprimary care treat-ment (N=23) or tra-ditional treatmentsetting (N=23) for12 weeks

Primary: treatmentretention and urinedrug tests

A trend toward higher retention at 12weeks was noted in the primary caresetting (78% versus 52%, p=.06).Patients in that setting had significantlylower rates of illicit opioid use as mea-sured by urine drug tests (63% versus85%, p,.01) but no difference in ratesof cocaine use.

Johnsonet al.,2000 (20)

RCT to compare levo-methadyl acetate (75–115 mg), buprenor-phine (16–32 mg), andhigh-dose (60–100mg) and low-dose(20 mg) methadoneas treatments for opi-oid dependence

220 patients, with 55in each group; 51%completed the 17-week trial.

Primary: treatment reten-tion, opioid use (per-centage of positiveurine screens), de-gree of continuousabstinence from opi-oid use (at least 12consecutive opioid-free urine screens),and patients’ reportsof use. Secondary: per-centage of cocaine-positive urine screens,abstinence from co-caine use, breath al-cohol readings, sideeffects, and sex-relateddifferences

No difference was found between high-dose buprenorphine and high-dosemeth-adone in days in treatment (mean of 96and 105 days, respectively) or percentageof patients with 12 or more consecutivenegative screens (26% versus 28%, respec-tively). High-dose buprenorphine wassuperior to low-dose methadone forboth outcomes (mean days, 96 versus70, p,.001; consecutive negative screens,26% versus 8%, p=.005).

Continues on next page

4 PSYCHIATRIC SERVICES IN ADVANCE

Table 2

Continued from previous page

StudyDesign andobjectives

Population andconditions

Outcomesmeasured Summary of findings

Fudala et al.,2003 (17)

RCT to compare 4weeks of office-basedtreatment with dailysublingual tablets ofbuprenorphine (16mg) in combinationwith naloxone (4 mg),buprenorphine alone(16 mg), or placebofor patients addictedto opioids

323 patients receivingat least one dose ofstudy medication;109 randomly as-signed to the com-bination medication,105 to buprenorphinealone, and 109 toplacebo

Primary: percentage ofurine screens nega-tive for opiates andself-reported cravingfor opiates by patients

During each of the 4 weeks, meancraving scores in the combined andbuprenorphine groups were sig-nificantly lower than in the placebogroup (p,.001 for both). Both groupswith buprenorphine-based treatmentshad reduced opioid use. Opioid-negative screens: combined group,17.8%; buprenorphine group, 20.7%;and placebo group, 5.8% (p,.001for all)

Kakko et al.,2003 (15)

RCT to compare dailybuprenorphine (fixeddose) versus a 6-daytapered regimen ofbuprenorphine fol-lowed by placebo;12-month programcombined withpsychotherapy

40 patients randomlyassigned to fixed-dose buprenorphine(N=20) or the taperedregimen (N=20)

Primary: 1-year re-tention in treatmentand negative urinedrug screens

One-year retention was 75% in thebuprenorphine group and 0% in theplacebo group (p=.001). Roughly75% of the patients retained intreatment had negative urine screensfor illicit opiates, stimulants, canna-binoids, and benzodiazepines.

Jones et al.,2005 (28)

RCT to compare NASamong neonates ofMMT- and BMT-maintained pregnant,opioid-dependentwomen; provide pre-liminary safety andefficacy data

30 patients randomlyassigned to MMT(N=15) or to BMT(N=15); 11 and 9,respectively, com-pleted the study.

Primary: number ofneonates treated forNAS, amount of med-ication used to treatNAS, length of neonatalhospitalization, andpeak NAS score. Sec-ondary: treatmentretention and illicitopiate use

No significant difference in illicitopioid use between groups. Total of20.0% and 45.5% of BMT-exposedand MMT-exposed neonates, res-pectively, were treated for NAS(p=.23). Other primary outcomeswere also not significantly different,except that the BMT-exposedneonates had a shorter averagehospital stay (p=.021).

Fischeret al.,2006 (29)

RCT to evaluate theefficacy and safetyof MMT versusBMT for pregnant,opioid-dependentwomen

18 pregnant womenrandomly assignedto receive MMT(N=9) or BMT (N=9)during weeks 24–29of pregnancy. Afterdropout, data wereavailable from 14cases (6 for meth-adone and 8 forbuprenorphine.

Primary for mothers:treatment retention,urine drug screens,and nicotine use.Primary for neonates:routine birth dataand severity and dura-tion of NAS

For mothers, no significant differencein retention was found betweengroups. MMT group had sig-nificantly less use of additionalopioids (p=.029). For neonates,earlier onset of NAS was noted in theMMT group; 43% of neonates n bothgroups combined did not require NAStreatment. Duration of NAS treatmentwas short in both groups (mean 5 days).

Kakko et al.,2007 (24)

RCT to compareadaptive, BMTstepped care versusoptimal MMT

96 patients randomly as-signed to flexible-doseMMT group (N=48)or BMT stepped-caregroup (N=48). Instepped treatment,buprenorphine couldbe increased to32 mg. If participantsrequired additionalmedication, they wereswitched (stepped) tohigh-dose methadone.

Primary: 6-month treat-ment retention, neg-ative urine opioidscreens, and problemseverity

No differences between groups werefound for retention (76% for bothat 6 months) or the proportion ofnegative screens (80% for bothgroups). For the BMT stepped-caregroup, 17 completers did not switchto methadone and finished with amean buprenorphine dose of 29.6mg, and 20 completers switched tomethadone and completed with amean methadone dose of 111 mg.Methadone group ended with amean dose of 110 mg.

Comer et al.,2010 (31)

Randomized cross-overstudy to assess intra-venous abuse poten-tial of buprenorphine-naloxone comparedwith buprenorphineamong injection drugusers receiving BMT

12 intravenous drugusers living in a hos-pital for 8–9 weeksand receiving bupre-norphine-naloxoneunder 3 BMT doseconditions: 2 mg, 8mg, and 24 mg

Primary: reinforcingeffects of intravenousbuprenorphine-naloxone and bupre-norphine amongBMT-maintainedintravenous drugusers who were

Buprenorphine-naloxone intravenousabuse potential was lower thanbuprenorphine alone or heroin,particularly on higher maintenancedoses. Intravenous buprenorphine-naloxone was self-administered lessfrequently than buprenorphine orheroin (p,.001). Selective ratings for

Continues on next page

PSYCHIATRIC SERVICES IN ADVANCE 5

Table 2

Continued from previous page

StudyDesign andobjectives

Population andconditions

Outcomesmeasured Summary of findings

given a drug-versus-money choiceexercise

“drug liking” and “desire to take thedrug again” were lower for buprenor-phine-naloxone than for buprenorphinealone or heroin (p=.001).

Jones et al.,2010 (27)

RCT to examine neuro-behavioral effects forneonates exposed toMMT or BMT

175 pregnant womenwith opioid depen-dency assigned toMMT group (N=89)or BMT group(N=86)

Primary: reduction inopioid use, treatmentretention, percentageof neonates treatedfor NAS, NAS peakscore, length of hos-pital stay, morphinerequired to treatNAS

Treatment was discontinued by 18% ofwomen in the MMT group and 33%in the BMT group; 58 mothers ex-posed to buprenorphine and 73 ex-posed to methadone were followedto the end of pregnancy. Neonatesof the former group required lessmorphine (mean dose, 1.1 versus10.4 mg, p,.009), had a shorterhospital stay (10.0 versus 17.5 days,p,.009), and had a shorter durationof NAS treatment (4.1 versus 9.9days, p,.003).

Ling et al.,2010 (21)

RCT to determineefficacy of bupre-norphine implants(6 month) versusplacebo

163 patients receivedbuprenorphineimplants (N=108) orplacebo implants(N=55) after induc-tion with sublingualbuprenorphinetablets

Primary: treatmentretention and reduc-tion in illicit opioiduse as measured byurine drug screens.Secondary: drugcraving and with-drawal symptoms

Significantly more patients withbuprenorphine implants completedthe study (65.7% versus 30.9%,p,.001). The buprenorphine grouphad more negative screens (40.4%versus 28.3%, p=.04), reducedwithdrawal symptoms on the ClinicalOpiate Withdrawal Scale (p,.001),and the Subjective Opiate WithdrawalScale (p=.004), lower patient ratingsfor craving on the Visual Analog Scale–opioid craving (p,.001), fewersymptoms on the Clinical GlobalImpressions–Severity Scale (34.9%versus 19.1% with no symptoms,p,.001), and greater change on theClinical Global Impressions–Improvement Scale (56.0% versus23.4% reporting very much improve-ment at week 24, p,.001).

Lucas et al.,2010 (26)

RCT to compareclinic-based BMTwith case manage-ment and referraland an opioid treat-ment program withinan HIV clinic

93 HIV-positive,opioid-dependentpatients not receivingopioid agonist therapyand not dependent onalcohol or benzodi-azepines randomlyassigned to receiveBMT in an HIVclinic (N=46) or re-ferred to an opioidtreatment program,where they receivedeither buprenor-phine or methadone(N=47)

Primary: initiation andlong-term treatmentwith opioid agonisttherapy, urine screenresults, visit atten-dance with primaryHIV providers, useof antiretroviral ther-apy, and HIV treat-ment outcomes

A larger proportion of HIV clinic pa-tients were on agonist therapy at 12months (74% versus 41%; p,.001).Illicit opioid use was less in the clinic-based group (44% versus 65%;p=.015). HIV clinic patients hadsignificantly fewer cocaine-positivescreens and attended more HIV pri-mary care visits. No difference wasfound in use of antiretroviral therapyor in improvements in HIV-monitoring tests.

Bazazi et al.,2011 (32)

Self-administeredsurvey study toexamine use, pro-curement, andmotivations foruse of divertedbuprenorphine-naloxone

100 opioid users; 51injecting users and49 noninjectingusers

Primary: illicitpossession ofbuprenorphine-naloxone, use ofdiverted buprenor-phine-naloxone,reasons for use, anduse to “get high”

More noninjecting users reported everusing buprenorphine-naloxone to“get high” (69% versus 32%, p,.01).Most participants reporting past useof buprenorphine-naloxone statedthat use was to treat withdrawal symp-toms (74%) or to stop using other opi-oids (66%) or because they could notafford drug treatment (64%).

Continues on next page

6 PSYCHIATRIC SERVICES IN ADVANCE

Effectiveness of BMT

Buprenorphine versus placebo. Stud-ies since 1995have foundbuprenorphine

to be a safe and effective treatment foropioid dependence. Compared withplacebo, buprenorphine significantly

improved treatment retention at low(2–6 mg), medium (7–15 mg), andhigh ($16 mg) doses (15–17,34). In

Table 2

Continued from previous page

StudyDesign andobjectives

Population andconditions

Outcomesmeasured Summary of findings

Weiss et al.,2011 (22)

Multiphase RCT toevaluate efficacy ofbrief and extendedbuprenorphine-naloxone treatmentwith various coun-seling intensities

First phase (N=653):brief treatment withbuprenorphine-naloxone with a 2-week stabilization,2-week taper, and8-week postmed-ication follow-up.Patients entered thesecond phase if theyhad opioid-positiveurine samples dur-ing the first phase.Second phase(N=360): 12 weeksof buprenorphine-naloxone treatment,4-week taper, and8-week postmedi-cation follow-up. Inboth phases, patientswere randomly as-signed to receivestandard (15-minutemedical visits) orenhanced medicalmanagement (stan-dard medical man-agement plus opioiddependence counsel-ing during 45-minutevisits).

Primary: minimal orno opioid use asmeasured by urinesamples that confir-med self-reports

All urine samples were negative afterthe first phase for only 6.6% ofpatients. During extended treatmentwith buprenorphine-naloxone, 49.2%of patients had successful outcomes(opioid-negative urine samples); thisrate fell to 8.6% at 8-week follow-up.Addition of counseling had no effectin either phase.

Coyle et al.,2012 (30)

RCT to determine im-pact on infant neuro-behavior of in-uteroexposure to buprenor-phine or methadone

39 full-term infantsexposed to metha-done (N=21) orbuprenorphine(N=18)

Primary: neonatalneurobehavioral ef-fects, measured onthe neonatal inten-sive care unit’s Net-work NeurobehavioralScale

Infants exposed to buprenorphineexhibited fewer signs of stress absti-nence (p,.001) and were less ex-citable (p,.001), less overaroused(p,.01), less hypertonic (p,.007),and better self-regulated (p,.04).

Moore et al.,2012 (23)

RCT to investigate im-pact of directly ob-served therapy pluscognitive-behavioraltherapy versus usualtreatment amongpatients receivingBMT for 12 weeksin primary care

55 opioid-dependentpatients assigned tophysician managementwith weekly bupren-orphine dispensing(N=28) or with di-rectly observed,thrice-weekly bupren-orphine and cognitive-behavioral therapy(N=27)

Primary: treatmentretention and druguse as measured byself-reports or urinescreens

No difference was found betweengroups in treatment retention ordrug use.

Prithamet al.,2012 (33)

Retrospective descrip-tive study to examineopioid replacementtreatment in preg-nancy and effect onneonatal outcomes

152 opioid-dependentpregnant womenreceiving MMT(N=136) or BMT(N=16) during pre-gnancy and theirneonates

Primary: length ofhospital stay for NAS

Neonates with prenatal exposure toMMT spent more days in the hos-pital for NAS (21 versus 14 days) (p=.05).

a Studies are listed in chronological order. Abbreviations: MMT, methadone maintenance treatment; NAS, neonatal abstinence syndrome; RCT,randomized controlled trial

PSYCHIATRIC SERVICES IN ADVANCE 7

one meta-analysis, buprenorphineshowed an improvement in treatmentretention over placebo at low doses(relative risk [RR]=1.50, p,.05),medium doses (RR=1.74, p,.05), andhigh doses (RR=1.74, p,.05) (34).Higher dose ranges (16–32 mg) havebeen associated with better retentionin treatment, compared with thelower dose (69% versus 51%, p=.006)(35). At medium- and high-doseranges, buprenorphine significantlyreduced illicit opioid use comparedwith placebo or with buprenorphineat a very low dose, as measured byurine drug tests (15–18,34). Forexample, one RCT reported that forthe group receiving 16 mg of bupre-norphine, 38% of urine samples werenegative for opioids, compared with18% of samples for the group re-ceiving 1 mg (p,.001) (16); anotherstudy found 21% opioid-negativeurine samples with buprenorphinealone versus 6% with placebo (p,.001)(17). Studies have shown inconsistentresults regarding reductions in non-opioid illicit drug use (for example,cocaine). However, most studies ofbuprenorphine have shown no statis-tically significant impact on reducingnonopioid illicit drug use comparedwith placebo (15,17,18,34). Althoughthe addition of naloxone to buprenor-phine has been shown to decreaseabuse potential (31), naloxone has notbeen found to alter buprenorphine’sefficacy (40).Although buprenorphine implants

were not FDA-approved in theUnited States at the time of thisreview, Ling and colleagues (21)examined the effect of six-monthbuprenorphine implants comparedwith placebo in a phase III trial. Thestudy compared patients receivingbuprenorphine implants (N=108)and those receiving placebo implants(N=55) after induction with sublin-gual buprenorphine tablets. Bothgroups had the option of receivingsupplemental buprenorphine tabletsfor withdrawal symptoms or craving.Participants could also receive a sup-plemental dose upon request, if it wasdeemed suitable by the treating clini-cian. Results showed that a signifi-cantly higher percentage of thosereceiving buprenorphine implantscompleted the six-month study

(65.7% versus 30.9%, p,.001). Inaddition, patients in the buprenor-phine implant group had a signifi-cantly higher percentage of theirurine samples negative for illicitopioids (40.4% versus 28.3%, p=.04).In regard to secondary outcomes, thebuprenorphine implant group hadsignificantly reduced withdrawal symp-toms on the Clinical Opiate With-drawal Scale (p,.001), and theSubjective Opiate Withdrawal Scale(p=.004), lower patient ratings ofcraving on the Visual Analog Scale–opioid craving (p,.001), fewersymptoms on the Clinical GlobalImpressions–Severity Scale (34.9%versus 19.1% with no symptoms,p,.001), and greater change on theClinical Global Impressions–Im-provement Scale (56.0% versus23.4% reporting very much improve-ment at week 24, p,.001).

Illicit use of buprenorphine. Con-cerns regarding diversion or nonmedicaluse of buprenorphine have emerged,even with the buprenorphine-naloxonecombination (31,32,41). Comer and col-leagues (31) confirmed that buprenorphine-naloxone retains some potential forabuse intravenously, but the combi-nation has less abuse potential asmeasured by self-administration thanbuprenorphine alone or heroin. Sur-veys of individuals with opioid usedisorders suggest that up to half ofclients who use opioid drugs and seektreatment have used illicit buprenor-phine. The clients typically stated thatthey used opioids for management ofwithdrawal symptoms and in attemptsto decrease other opioid use (32,41,42).Individuals addicted to prescriptionopioids were more likely than thoseaddicted to intravenous heroin to usebuprenorphine to “get high” (32).

Prescription opioid dependence. Arecent study examined the use ofbuprenorphine to treat patients withprescription opioid dependence.Weissand colleagues (22) conducted thePrescription Opioid Addiction Treat-ment Study multiphase clinical trial incommunity treatment settings, report-ing outcomes compared with baseline.The first phase examined brief treat-ment with buprenorphine and pro-vided a two-week buprenorphinestabilization, two-week taper, andeight-week postmedication follow-up.

Patients entered the second phase ifthey had relapsed (opioid-positiveurine sample) during the initial phase.The second phase consisted of a12-week buprenorphine treatment,four-week taper, and eight-week post-medication follow-up. In both phases,patients were randomly assigned toreceive standard medical manage-ment (15-minute medical visits) orenhanced management (standard med-ical management plus opioid depen-dence counseling in 45-minute visits).Results showed that all urine sampleswere negative for only 6.6% of patientsafter the first phase (note that all pa-rticipants received buprenorphine).During extended treatment withbuprenorphine, 49.2% of patients hadsuccessful outcomes (all urine sampleswere opioid negative), but this per-centage fell to 8.6% at the eight-weekfollow-up after buprenorphine wasdiscontinued. Opioid dependencecounseling had no effect in eitherphase. The authors concluded thatpatients dependent on prescriptionopioids have good outcomes withimproved abstinence while takingbuprenorphine, but if they are taperedoff of this drug, the likelihood ofsuccessful outcomes in terms of noopioid use is low.

Psychosocial interventions

and support services

The addition of structured psycho-therapy to standard treatment—which may include peer supportservices, 12-step programs, and otherpsychosocial treatment provided atthe facility or office—has not beenshown to improve outcomes forpatients on opioid maintenance ther-apy. A meta-analysis examined theimpact of adding a more structuredpsychotherapy to standard treatmentthat included three types of opioidagonist therapy: levomethadyl acetate(LAAM; now off the U.S. market)(one study), methadone (28 studies),or buprenorphine (six studies) (37).The authors found no improvementsin treatment retention or abstinencefrom illicit opioids and no effect onother outcomes, compliance, or psy-chiatric symptoms. It is important tonote that in this meta-analysis, stan-dard treatment may have includedpeer support, psychosocial treatment

8 PSYCHIATRIC SERVICES IN ADVANCE

and counseling sessions, and referralsfor additional support, but the meta-analysis examined only the effects ofstructured treatment in addition tosupport services already provided. A

more recent study investigated theimpact of directly observed therapyplus cognitive-behavioral therapy com-pared with regular medical manage-ment of BMT (23). Results showed no

improvement in retention or drug use.It has been noted that the literature onpsychosocial treatments is heteroge-neous, and there is a lack of sufficient,high-quality studies to assess which

Table 3

Review articles about buprenorphine maintenance treatment (BMT) included in the reviewa

Study Focus of reviewPopulation andconditions

Outcomesmeasured Summary of findings

Barnettet al.,2001 (36)

Compare the effective-ness of buprenorphineand of methadone

Patients receivingmethadone atmedium-high (50–80 mg) and low(20–35 mg) dosesand buprenorphineat medium doses(6–12 mg) across5 RCTs

Primary: retention intreatment and urinedrug screens foropioids

Compared with patients on medium-highmethadone doses, those on mediumdoses of buprenorphine had 1.26 timesthe relative risk (RR) of discontinuingtreatment (p=.019), and the rate ofpositive drug screens was 8.3% higher(p=.002). Buprenorphine was moreeffective than low doses of methadonein treatment retention (RR of discon-tinuing treatment=.86; ns) and reduc-tion of positive drug screens (8.4%fewer, p,.05).

Matticket al.,2008 (34)

Compare the effects ofBMT with placebo andMMT on treatment re-tention and suppres-sion of illicit drug use

Evaluated 24 RCTsinvolving 4,497patients

Primary: retention intreatment and illicitdrug usesuppression

Treatment retention was higher with BMTcompared with placebo at low doses(RR=1.50, p,.05), medium doses(RR=1.74, p,.05), and high doses(RR=1.74, p,.05).

McCance-Katz et al.,2010 (38)

Examine literature onmethadone and bupre-norphine for druginteractions with con-current medications

Populations varied;extensive literaturereview with 93references

Primary: drug interac-tions with metha-done orbuprenorphine

Buprenorphine had fewer drug interac-tions than methadone, especially withHIV medications.

Amato et al.,2011 (37)

Evaluate the effectivenessof any psychosocialtreatment plus any ag-onist maintenancetreatment versus stan-dard agonist treatment

4,319 patients in35 studies

Primary: retention intreatment and opi-ate abstinence; sec-ondary: treatmentcompliance, psychi-atric symptoms, de-pression, and death

Adding any psychosocial support to stan-dard maintenance treatments did notappear to give additional benefits.

Martin et al.,2011 (10)

Examine literature, regu-latory actions, profes-sional guidance, andopioid treatment pro-gram experiencesregarding adverse car-diac events associatedwith methadone

Populations varied;extensive literaturereview with 108references and in-put from panel andfield experts

Primary: cardiacevents associatedwith methadone;impact on cardiacQT interval

The pharmacology of buprenorphineaffords it a better safety profile thanmethadone; buprenorphine (at standarddoses) did not affect cardiac electro-physiology by lengthening the cardiacQT interval.

Fareedet al.,2012 (35)

Meta-analysis to provideinformation aboutproper dosing in BMTto improve treatmentoutcomes

Compared higherdoses of buprenor-phine (16–32 mgper day) to lowerdose (,16 mg perday) across 21 RCTsinvolving 2,703patients

Primary: treatmentretention and re-duction in opioiduse

Higher doses of buprenorphine wereassociated with better treatment re-tention than the lower dose (69%versus 51%, p=.006).

Jones et al.,2012 (39)

Review literature on out-comes after maternaltreatment withbuprenorphine

Evaluated outcomes of3 RCTs and 44nonrandomizedstudies

Primary: fetal effects,neonatal effects,effects on breastmilk, and longer-term developmentaleffects

Maternal treatment with buprenorphinehad similar efficacy to methadone.Prenatal buprenorphine treatmentresulted in less severe neonatal absti-nence syndrome than methadonetreatment. No adverse effects on infantdevelopment of in-utero buprenorphineexposure were found. Dose increasesfor methadone and buprenorphine maybe needed during pregnancy.

a Studies are listed in chronological order. Abbreviations: MMT, methadone maintenance treatment; RCT, randomized controlled trial

PSYCHIATRIC SERVICES IN ADVANCE 9

psychosocial interventions have themost success in various populations(43).BMT versus MMT. Several studies

and meta-analyses have examined theuse of BMT compared with MMT.Dose levels have been shown to beimportant for efficacy of both drugs.In this discussion, we define metha-done dose ranges as high ($60 mg),medium (40–59mg), and low (,40mg).We define buprenorphine dose rangesas high (16–32mg), medium (7–15mg),and low (2–6 mg).Barnett and colleagues (36) per-

formed a meta-analysis of data fromfive RCTs conducted between 1992and 1997. The authors compared theefficacy of methadone at medium-high doses (50–80 mg) and low doses(20–35 mg) and buprenorphine atmedium doses (6–12 mg). Resultsshowed that patients on mediumdoses of buprenorphine had 1.26times the relative risk of discontinuingtreatment (p=.019), and the numberof positive urine samples was 8.3%higher than the number for patientson medium-high doses of methadone(p=.002). However, compared withlower doses of methadone (20–30 mgper day), buprenorphine was moreeffective in treatment retention (RRfor discontinuing treatment=.86, notsignificant) and in reduction of posi-tive urine drug tests (8.4% fewerpositive urine samples per patient,p,.05). Ling and colleagues (19)found similar results. High-dosemethadone (80 mg) was superior tomedium-dose buprenorphine (8 mg)and low-dose methadone (30 mg) fortreatment retention and opioid use.A more recent meta-analysis com-

paring BMT and MMT was based on25 RCTs and 4,497 participants (34).The authors found results that weresimilar to the study by Barnett andcolleagues (36). Specifically, this meta-analysis found mixed results formedium-dose buprenorphine versusmedium- and low-dose methadone inretaining patients. Three studies sug-gested that MMT was superior,whereas seven found no differencebetween the groups, although resultsdiffered by dose. Medium-dose bupre-norphine was less likely to suppressillicit opioid use than medium-dosemethadone (standard mean difference

[SMD]=.27, p,.05), but it was morelikely to suppress illicit opioid use thanlow-dose methadone (SMD=–.23,p,.05). Treatment retention wasworse for low-dose buprenorphine thanfor medium- and low-dose methadone(RR for both comparisons=.67, p,.05).Low-dose buprenorphine showed nodifference in illicit opioid use com-pared with low-dose methadone, butlow-dose buprenorphine was inferior tomedium-dose methadone in terms ofillicit opioid use (SMD=.88, p,.05). Inthe meta-analysis, flexible-dose bupre-norphine and methadone had similarresults for illicit opioid use, and meth-adone had a slight (but statisticallysignificant) edge for retention in treat-ment—despite the fact that moststudies found no difference. Of note,several of the studies used buprenor-phine in low- or medium-dose ranges,and the flexible-dose ranges were nothigher than 16 mg. No statisticallysignificant differences were found be-tween methadone and buprenorphineat any dose comparison for use of otherillicit drugs (primarily cocaine) orcriminal activity.

Johnson and colleagues (20) con-ducted a 17-week RCT (N=220) tocompare the effects of LAAM (75–115 mg), high-dose buprenorphine(16–32 mg), high-dose methadone(60–100 mg), and low-dose metha-done (20 mg). Although LAAM is nolonger marketed in the United States,the comparison of high-dose bupre-norphine, high-dose methadone, andlow-dose methadone is still important.The results supported the value ofhigh-dose buprenorphine; no differ-ence was found between high-dosebuprenorphine and high-dose metha-done in the mean number of days intreatment (96 and 105 days, respec-tively) or in the percentage of partic-ipants with 12 or more consecutiveurine samples that were negative forillicit opioids (26% and 28%). High-dose buprenorphine was superior tolow-dose methadone in terms of themean number of days in treatment(96 versus 70, respectively, p,.001)and percentage of participants withconsecutive negative urine samples(26% versus 8%, p=.005).

Kakko and colleagues (24) testedthe efficacy of a stepped-care strategythat used buprenorphine in increas-

ing doses. The researchers compareda flexible-dose MMT group (n=48)and a stepped-care BMT group(N=48). In the stepped-treatmentgroup that used a flexible-dose algo-rithm, buprenorphine could be in-creased up to 32 mg. If participantsrequired additional medication, theywere switched (stepped) to high-dosemethadone. The study found nodifferences between the stepped-care BMT and MMT groups in treat-ment retention (76% for both at sixmonths) or in the proportion of urinesamples that were free of illicit opioids(80% for both groups). In the bupre-norphine stepped-care group, 17 par-ticipants who completed treatment didnot switch to methadone and finishedwith a mean buprenorphine dose of29.6 mg, and 20 participants whocompleted treatment switched tometh-adone and finished with a mean meth-adone dose of 111.0 mg. Those in themethadone group ended with a meandose of 110.0 mg.

The pharmacology of buprenor-phine affords it a better safety profilethan methadone, which is importantconsidering that methadone is associ-ated with one-third of opioid-relatedoverdose deaths annually (44). Becauseit is a partial agonist at the mu opiatereceptor, it has a ceiling effect thatlimits its potential to cause respiratorydepression compared with methadone(45). However, this risk still exists,especially if buprenorphine is used incombinationwithother central nervoussystem depressants such as benzodia-zepines or alcohol (8) or is used inhigher doses. In addition, unlike meth-adone, buprenorphine at standarddoses does not affect cardiac electro-physiology by lengthening the cardiacQT interval—a mechanism that canlead to serious cardiac arrhythmias(10). Buprenorphine also has fewerdrug interactions than methadone,especially with HIV medications (38).

Taken together, the articles re-viewed suggest that the efficacy ofBMT is dose dependent, and dose isimportant to take into account whencomparing medications. For compar-isons at medium-dose ranges, evi-dence is mixed—some studies showsimilar effects of MMT and BMT andsome studies suggest that MMT im-proves treatment retention or reduces

10 PSYCHIATRIC SERVICES IN ADVANCE

illicit opioid use. Only one studyreviewed compared high doses ofbuprenorphine and methadone, and itshowed similar outcomes (20). Fi-nally, the stepped-care approach—inwhich individuals begin with bupre-norphine and switch to methadone ifbuprenorphine doses above 32 mg arerequired—suggests that MMT may beneeded for patients who require highdoses of opioid agonist treatment (24).Treatment setting. We reviewed

two studies examining the receipt ofBMT in an office-based setting com-pared with treatment in a traditionaldrug treatment program. In an earlyRCT (1998), O’Connor and col-leagues (25) compared patients ran-domly assigned to receive BMT ina primary care setting (N=23) ora traditional drug treatment program(N=23). During the 12-week study,retention showed a trend towardbeing higher in the primary caresetting, compared with the traditionalsetting (78% versus 52%, respectively,p=.06). Patients in the primary caresetting had significantly lower rates ofillicit opioid use on the basis of urinedrug tests (63% versus 85%, p,.01),but they showed no difference in ratesof cocaine use. Lucas and colleagues(26) compared outcomes of HIV-positive patients randomly assignedto receive BMT in an HIV clinic(N=46) or an opioid treatment pro-gram in which they received eitherbuprenorphine or methadone (N=47).A significantly higher proportion of thepatients in the HIV clinic were re-ceiving agonist therapy at 12 months(74% versus 41%, p,.001). Illicitopioid use, as measured by urine drugtests, was less in the clinic-based group(44% versus 65% of patients; p=.015).In addition, the study showed thatpatients treated in the HIV clinic hadsignificantly fewer cocaine-positiveurine drug tests and attended moreHIV primary care visits. The groupsdid not differ in use of antiretroviraltherapy or in improvements in testsused to monitor HIV. The authorsspeculated that streamlined accessto treatment in the clinic group was amajor reason for the improved results.None of the RCTs reviewed were

implemented in incarcerated popula-tions. A recent survey of criminal justiceagencies indicated that medication-

assisted treatment of incarcerated indi-viduals is generally limited to pregnantwomen and detoxification (46).

Buprenorphine use in pregnancy.MMT has been used to treat opioiddependence during pregnancy toimprove maternal and fetal outcomes(47,48). However, as discussed in thecompanion article (3), MMT putsnewborn infants at risk for neonatalabstinence syndrome (NAS). NASoften requires detoxification treat-ment in the hospital with a morphinetaper (49–53). As a result, cliniciansand researchers have studied BMT asan alternative to MMT during preg-nancy. RCTs were conducted withbuprenorphine alone, to avoid pre-natal exposure to naloxone.

Three RCTs and observational stud-ies (27–29,39) have compared use ofbuprenorphine with use of methadoneby pregnant women. Authors con-cluded that buprenorphine has similarefficacy to methadone in reducingillicit opioid use among pregnantwomen, and buprenorphine may leadto less severe NAS. With both MMTand BMT, dose increases may benecessary during pregnancy (39). Al-though the two smaller RCTs did notfind a difference in treatment re-tention between BMT and MMT(28,29), the largest RCT—the Mater-nal Opioid Treatment: Human Ex-perimental Research study (27)—found that a higher percentage ofpatients in the BMT group discon-tinued treatment before delivery(33% versus 18%, p=.02). Motherswere more likely to discontinue treat-ment in both groups if they had highercumulative lifetime months and re-cent days of heroin use (27). TwoRCTs showed no difference in illicitopioid use between the two medica-tions (27,28), whereas one RCTsuggested that methadone may besuperior in reducing illicit opioid use(29). Infants born to mothers main-

tained with buprenorphine versusmethadone had similar rates of NAS,but the manifestation of NAS was lesssevere. Infants whose mothers tookbuprenorphine required significantlylower doses of morphine to treat NASand needed fewer hospital days(27,30,33).

ConclusionsOverall, a high level of evidence wasfound for the effectiveness of BMT inimproving treatment retention anddecreasing illicit opioid use (see boxon this page). Research regarding theimpact of BMT on nonopioid illicitdrug use is less conclusive but sug-gests positive trends. The addition ofany type of psychosocial regimen toBMT or MMT has not been shown toimprove outcomes, but the hetero-geneity of interventions across trialslimits the ability to make strongconclusions. As with MMT, there isgrowing evidence that higher doses ofbuprenorphine (16–32 mg) are moreefficacious than lower doses; however,because of the pharmacology ofbuprenorphine, doses above 32 mgdo not provide additional efficacy.Research suggests that buprenor-phine may be as effective for patientswith prescription opioid dependenceas it is for patients with heroin de-pendence. When the medications aredosed similarly, BMT appears to be aseffective as MMT in reducing illicitopioid use. Results are mixed regard-ing treatment retention, but severalstudies suggest that MMT might con-fer some advantage. The advantagemay be due, in part, to the supportiveservices or social reinforcement inoutpatient MMT programs. However,buprenorphine has a better safetyprofile than methadone, and theability to prescribe buprenorphinein office facilities as opposed to only inopioid treatment programs improvesaccess to care and earlier initiation of

Evidence for the effectiveness of BMT: highEvidence clearly shows that BMT has a positive impact compared with placebo on:• Retention in treatment• Illicit opioid use

Evidence is mixed for its impact on:• Nonopioid illicit drug use

PSYCHIATRIC SERVICES IN ADVANCE 11

treatment. A key advantage of bupre-norphine is its availability. The numberof clinicians approved to prescribebuprenorphine is growing, althoughmany areas of the country do not haveaccess to methadone programs (2).Both BMT and MMT improve

pregnancy-related outcomes by re-ducing illicit drug use during preg-nancy. Infants of mothers treated withbuprenorphine during pregnancy maybe born with NAS, although NASappears to be less severe in infantsof mothers treated with buprenorphinethan of those treated with methadone.Potential areas for future research

include increased focus on the impactof BMT on secondary outcomes,additional investigation of appropriatedosing to enhance treatment out-comes, confirmation of the results ofthe stepped-care protocol, improvedinduction protocols to minimize initialproblems with treatment retention(and thus potentially enhance adoptionrates by providers), and examination ofthe differential effectiveness of BMTin specific subpopulations, such aspatients dependent on prescriptionopioids versus heroin. Differential ef-fects and access to BMT acrossracial and ethnic groups and geo-graphic areas should also be studied.Ongoing research needs do not

diminish the strong evidence for thistreatment approach. Given the poorsuccess rates of abstinence-basedtreatments for opioid use disordersand the limited access to and morerestrictive safety profile of MMT,BMT is an important treatment foropioid dependence. Policy makershave reason to promote access toBMT for patients in substance usetreatment who may wish to chooseBMT as a potentially safer alternativeto MMT. Administrators of substanceuse treatment programs, communityhealth centers, and managed careorganizations and other purchasers ofhealth care services, such as Medicare,Medicaid, and commercial insurancecarriers, should give careful consider-ation to BMT as a covered benefit.

Acknowledgments and disclosures

Development of the Assessing the EvidenceBase Series was supported by contractsHHSS283200700029I/HHSS28342002T,HHSS283200700006I/HHSS28342003T,

and HHSS2832007000171/HHSS28300001Tfrom 2010 through 2013 from the SubstanceAbuse and Mental Health Services Administra-tion (SAMHSA). The authors acknowledge thecontributions of Robert Lubran, M.S., M.P.A.,KevinMalone, B.A., and Suzanne Fields,M.S.W.,from SAMHSA; John O’Brien, M.A., from theCenters for Medicare & Medicaid Services;John Easterday, Ph.D., Linda Lee, Ph.D.,Rosanna Coffey, Ph.D., and Tami Mark, Ph.D.,from Truven Health Analytics; and SandrinePirard, M.D., Ph.D., from National Institute onDrug Abuse. The views expressed in this articleare those of the authors and do not necessarilyrepresent the views of SAMHSA.

The authors report no competing interests.

References

1. Topics in Brief: Medication-Assisted Treat-ment for Opioid Addiction. Bethesda,Md, National Institute on Drug Abuse,April 2012. Available at www.drugabuse.gov/publications/topics-in-brief/medication-assisted-treatment-opioid-addiction. AccessedSept 14, 2013

2. The N-SSATS Report: Trends in the Useof Methadone and Buprenorphine atSubstance Abuse Treatment Facilities:2003 to 2011. Rockville, Md, SubstanceAbuse and Mental Health Services Ad-ministration, Center for Behavioral HealthStatistics and Quality, 2013

3. Fullerton CA, Kim M, Thomas CP, et al:Medication-assisted treatment with metha-done: assessing the evidence. PsychiatricServices, 2013; doi 10.1176/appi.ps.201300235

4. Connock M, Juarez-Garcia A, Jowett S,et al: Methadone and buprenorphine forthe management of opioid dependence:a systematic review and economic evalua-tion. Health Technology Assessment 11(9):1–171, 2007

5. Fletcher BW, Battjes RJ: Introduction tothe special issue: treatment process inDATOS. Drug and Alcohol Dependence57:81–87, 1999

6. Hall W, Ward J, Mattick R: The Effec-tiveness of Methadone MaintenanceTreatment 1: Heroin Use and Crime.Utrecht, Netherlands, Harwood AcademicPublishers, 1998

7. Mattick RP, Breen C, Kimber J, et al:Methadone maintenance therapy versus noopioid replacement therapy for opioiddependence. Cochrane Database of Sys-tematic Reviews 3:CD002209, 2009

8. Webster LR, Cochella S, Dasgupta N,et al: An analysis of the root causes foropioid-related overdose deaths in theUnited States. Pain Medicine 12(suppl 2):S26–S35, 2011

9. Modesto-Lowe V, Brooks D, Petry N:Methadone deaths: risk factors in pain andaddicted populations. Journal of GeneralInternal Medicine 25:305–309, 2010

10. Martin JA, Campbell A, Killip T, et al: QTinterval screening in methadone mainte-nance treatment: report of a SAMHSAexpert panel. Journal of Addictive Diseases30:283–306, 2011

11. Fiellin DA, O’Connor PG: New federalinitiatives to enhance the medical treat-ment of opioid dependence. Annals ofInternal Medicine 137:688–692, 2002

12. Subutex and Suboxone Approved to TreatOpiate Dependence. T02-38. SilverSpring, Md, US Food and Drug Adminis-tration, 2002. Available at www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm191521.htm

13. Frequently Asked Questions AboutBuprenorphine and the Drug AddictionTreatment Act of 2000 (DATA 2000).Rockville, Md, Substance Abuse andMental Health Services Administration.Available at buprenorphine.samhsa.gov/faq.html#A11. Accessed Sept 19, 2013

14. Dougherty RH, Lyman DR, George P,et al: Assessing the evidence base for be-havioral health services: introduction to theseries. Psychiatric Services, 2013; doi10.1176/appi.ps.201300214

15. Kakko J, Svanborg KD, Kreek MJ, et al:1-year retention and social function afterbuprenorphine-assisted relapse preventiontreatment for heroin dependence inSweden: a randomised, placebo-controlledtrial. Lancet 361:662–668, 2003

16. Ling W, Charuvastra C, Collins JF, et al:Buprenorphine maintenance treatment ofopiate dependence: a multicenter, ran-domized clinical trial. Addiction 93:475–486,1998

17. Fudala PJ, Bridge TP, Herbert S, et al:Office-based treatment of opiate addictionwith a sublingual-tablet formulation ofbuprenorphine and naloxone.NewEnglandJournal of Medicine 349:949–958, 2003

18. Johnson RE, Eissenberg T, Stitzer ML,et al: A placebo controlled clinical trial ofbuprenorphine as a treatment for opioiddependence. Drug and Alcohol De-pendence 40:17–25, 1995

19. Ling W, Wesson DR, Charuvastra C, et al:A controlled trial comparing buprenor-phine and methadone maintenance inopioid dependence. Archives of GeneralPsychiatry 53:401–407, 1996

20. Johnson RE, Chutuape MA, Strain EC,et al: A comparison of levomethadyl ace-tate, buprenorphine, and methadone foropioid dependence. New England Journalof Medicine 343:1290–1297, 2000

21. Ling W, Casadonte P, Bigelow G, et al:Buprenorphine implants for treatmentof opioid dependence: a randomized con-trolled trial. JAMA 304:1576–1583, 2010

22. Weiss RD, Potter JS, Fiellin DA, et al:Adjunctive counseling during brief andextended buprenorphine-naloxone treat-ment for prescription opioid dependence:a 2-phase randomized controlled trial.Archives of General Psychiatry 68:1238–1246, 2011

23. Moore BA, Barry DT, Sullivan LE, et al:Counseling and directly observed medica-tion for primary care buprenorphinemaintenance: a pilot study. Journal of Addic-tion Medicine 6:205–211, 2012

12 PSYCHIATRIC SERVICES IN ADVANCE

24. Kakko J, Grönbladh L, Svanborg KD, et al:A stepped care strategy using buprenor-phine and methadone versus conventionalmethadone maintenance in heroin de-pendence: a randomized controlled trial.American Journal of Psychiatry 164:797–803, 2007

25. O’Connor PG, Oliveto AH, Shi JM, et al: Arandomized trial of buprenorphine main-tenance for heroin dependence in a pri-mary care clinic for substance users versusa methadone clinic. American Journal ofMedicine 105:100–105, 1998

26. Lucas GM, Chaudhry A, Hsu J, et al:Clinic-based treatment of opioid-dependentHIV-infected patients versus referral to anopioid treatment program: a randomizedtrial. Annals of Internal Medicine 152:704–711, 2010

27. Jones HE, Kaltenbach K, Heil SH, et al:Neonatal abstinence syndrome after metha-done or buprenorphine exposure. New Eng-land Journal of Medicine 363:2320–2331, 2010

28. Jones HE, Johnson RE, Jasinski DR, et al:Buprenorphine versus methadone in thetreatment of pregnant opioid-dependentpatients: effects on the neonatal abstinencesyndrome. Drug and Alcohol Dependence79:1–10, 2005

29. Fischer G, Ortner R, Rohrmeister K, et al:Methadone versus buprenorphine inpregnant addicts: a double-blind, double-dummy comparison study. Addiction 101:275–281, 2006

30. Coyle MG, Salisbury AL, Lester BM, et al:Neonatal neurobehavior effects followingbuprenorphine versus methadone expo-sure. Addiction 107(suppl 1):63–73, 2012

31. Comer SD, Sullivan MA, Vosburg SK,et al: Abuse liability of intravenousbuprenorphine/naloxone and buprenorphinealone in buprenorphine-maintained intra-venous heroin abusers. Addiction 105:709–718, 2010

32. Bazazi AR, Yokell M, Fu JJ, et al: Illicit useof buprenorphine/naloxone among inject-ing and noninjecting opioid users. Journalof Addiction Medicine 5:175–180, 2011

33. Pritham UA, Paul JA, Hayes MJ: Opioiddependency in pregnancy and length of

stay for neonatal abstinence syndrome.Journal of Obstetric, Gynecologic, andNeonatal Nursing 41:180–190, 2012

34. Mattick RP, Kimber J, Breen C, et al:Buprenorphine maintenance versus pla-cebo or methadone maintenance for opioiddependence. Cochrane Database of Sys-tematic Reviews 2:CD002207, 2008

35. Fareed A, Vayalapalli S, Casarella J, et al: Effectof buprenorphine dose on treatment outcome.Journal of Addictive Diseases 31:8–18, 2012

36. Barnett PG, Rodgers JH, Bloch DA: Ameta-analysis comparing buprenorphine tomethadone for treatment of opiate de-pendence. Addiction 96:683–690, 2001

37. Amato L, Minozzi S, Davoli M, et al: Psy-chosocial combined with agonist mainte-nance treatments versus agonist maintenancetreatments alone for treatment of opioid de-pendence. Cochrane Database of SystematicReviews 10:CD004147, 2011

38. McCance-Katz EF, Sullivan LE, Nallani S:Drug interactions of clinical importanceamong the opioids, methadone and bupre-norphine, and other frequently prescribedmedications: a review. American Journal onAddictions 19:4–16, 2010

39. Jones HE, Heil SH, Baewert A, et al: Bupre-norphine treatment of opioid-dependentpregnant women: a comprehensive re-view. Addiction 107(suppl 1):5–27, 2012

40. Chiang CN, Hawks RL: Pharmacokineticsof the combination tablet of buprenor-phine and naloxone. Drug and AlcoholDependence 70(suppl):S39–S47, 2003

41. Monte AA, Mandell T, Wilford BB, et al:Diversion of buprenorphine/naloxonecoformulated tablets in a region with highprescribing prevalence. Journal of Addic-tive Diseases 28:226–231, 2009

42. Schuman-Olivier Z, Albanese M, NelsonSE, et al: Self-treatment: illicit buprenor-phine use by opioid-dependent treatmentseekers. Journal of Substance AbuseTreatment 39:41–50, 2010

43. Drummond DC, Perryman K: Psychoso-cial Interventions in Pharmacotherapy ofOpioid Dependence: A Literature Review.London, St George’s University of London,

Division of Mental Health, Section of Ad-dictive Behaviour, 2007

44. Vital signs: risk for overdose from metha-done used for pain relief - United States,1999–2010. Morbidity and MortalityWeekly Report 61:493–497, 2012

45. Fareed A, Vayalapalli S, Byrd-Sellers J,et al: Safety and efficacy of long-termbuprenorphine maintenance treatment.Addictive Disorders and Their Treatment10:123–130, 2011

46. Friedmann PD, Hoskinson R, Gordon M,et al: Medication-assisted treatment incriminal justice agencies affiliated withthe criminal justice-drug abuse treatmentstudies (CJ-DATS): availability, barriers,and intentions. Substance Abuse 33:9–18,2012

47. Kandall SR, Doberczak TM, Jantunen M,et al: The methadone-maintained pregnancy.Clinics in Perinatology 26:173–183, 1999

48. Hulse GK, Milne E, English DR, et al: Therelationship between maternal use of her-oin and methadone and infant birthweight. Addiction 92:1571–1579, 1997

49. Kaltenbach K, Berghella V, Finnegan L:Opioid dependence during pregnancy:effects and management. Obstetrics andGynecological Clinics of North America25:139–151, 1998

50. Finnegan L, Kaltenbach K: Neonatal ab-stinence syndrome; in Primary PediatricCare. Edited by Hoekelman RA, FriedmanSB, Nelson NM, et al. St Louis, Mo,Mosby–Year Book, 1992

51. Ebner N, Rohrmeister K, Winklbaur B,et al: Management of neonatal abstinencesyndrome in neonates born to opioidmaintained women. Drug and AlcoholDependence 87:131–138, 2007

52. Dashe JS, Sheffield JS, Olscher DA, et al:Relationship between maternal methadonedosage and neonatal withdrawal. Obstetricsand Gynecology 100:1244–1249, 2002

53. McCarthy JJ, Leamon MH, Parr MS, et al:High-dose methadone maintenance inpregnancy: maternal and neonatal out-comes. American Journal of Obstetrics andGynecology 193:606–610, 2005

PSYCHIATRIC SERVICES IN ADVANCE 13