Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Standard view
Full view
of .
Look up keyword
Like this
0 of .
Results for:
No results containing your search query
P. 1
Massachusetts OD Study in BMJ

Massachusetts OD Study in BMJ

Ratings: (0)|Views: 179|Likes:

More info:

Published by: webmaster@drugpolicy.org on Mar 22, 2013
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





Opioidoverdoseratesandimplementationofoverdoseeducation and nasal naloxone distribution inMassachusetts: interrupted time series analysis
Alexander Y Walley
assistant professor of medicine, medical director of Massachusetts opioid overdose prevention pilot 
1 3
, Ziming Xuan
research assistant professor 
, H Holly Hackman
, Emily Quinn
statistical manager 
, Maya Doe-Simkins
public health researcher 
,Amy Sorensen-Alawad
program manager 
, Sarah Ruiz
assistant director of planning and development 
, Al Ozonoff
director, design and analysis core 
5 6
Clinical Addiction Research Education Unit, Section of General Internal Medicine, Boston University School of Medicine, Boston, MA, USA;
Department of Community Health Sciences, Boston University School of Public Health, USA;
Massachusetts Department of Public Health, USA;
Data Coordinating Center, Boston University School of Public Health, USA ;
Design and Analysis Core, Clinical Research Center, Children’sHospital Boston, USA ;
Department of Biostatistics, Boston University School of Public Health, USA
Toevaluatetheimpactofstatesupportedoverdoseeducationand nasal naloxone distribution (OEND) programs on rates of opioidrelateddeathfromoverdoseandacutecareutilizationinMassachusetts.
Interruptedtimeseriesanalysisofopioidrelatedoverdosedeathand acute care utilization rates from 2002 to 2009 comparingcommunity-yearstratawithhighandlowratesofOENDimplementationto those with no implementation.
19 Massachusetts communities (geographically distinct citiesand towns) with at least five fatal opioid overdoses in each of the years2004 to 2006.
OEND was implemented among opioid users at risk foroverdose,socialserviceagencystaff,family,andfriendsofopioidusers.
OENDprogramsequippedpeopleatriskforoverdoseandbystanders with nasal naloxone rescue kits and trained them how toprevent,recognize,andrespondtoanoverdosebyengagingemergencymedical services, providing rescue breathing, and delivering naloxone.
Adjustedrateratiosforannualdeathsrelatedto opioid overdose and utilization of acute care hospitals.
Among these communities, OEND programs trained 2912potential bystanders who reported 327 rescues. Both community-yearstratawith1-100enrollmentsper100000population(adjustedrateratio0.73, 95% confidence interval 0.57 to 0.91) and community-year stratawith greater than 100 enrollments per 100 000 population (0.54, 0.39 to0.76) had significantly reduced adjusted rate ratios compared withcommunities with no implementation. Differences in rates of acute carehospital utilization were not significant.
Opioidoverdosedeathrateswerereducedincommunitieswhere OEND was implemented. This study provides observationalevidencethatbytrainingpotentialbystanderstoprevent,recognize,andrespond to opioid overdoses, OEND is an effective intervention.
Poisoning, nine out of 10 of which are related to drugoverdoses,
hassurpassedmotorvehiclecrashestobetheleadingcause of death by injury in the United States.
Overdose is alsoa major cause of death in Canada,
5 6
IntheUnitedStates,increasesinfataloverdosesincethe mid-1990s have been driven by the growth in prescriptionsfor opioid analgesics
and their non-medical use.
9 10
Opioidrelatedemergencydepartmentvisitsandadmissionstohospitalhave increased over the same period.
In Massachusetts, since
Correspondence to: A Y Walley Boston Medical Center, Section of General Internal Medicine, 801 Massachusetts Avenue, 2nd Floor, Boston, MA02118, USA awalley@bu.eduExtra material supplied by the author (seehttp://www.bmj.com/content/346/bmj.f174?tab=related#webextra)Map of the 19 communities in Massachusettscoefficients for covariates in adjusted absolute and relative models of opioid overdose related unintentional deathscoefficients for covariates in adjusted absolute and relative models of opioid related acute care utilizationsControl models of OEND implementation and ratio of opioid related overdose deaths to cancer deathsControl models of OEND implementation and ratio of opioid related acute care utilizations to motor vehicle crash acute care utilizations
No commercial reuse: See rights and reprintshttp://www.bmj.com/permissionsSubscribe:http://www.bmj.com/subscribe
2013;346:f174 doi: 10.1136/bmj.f174 (Published 31 January 2013) Page 1 of 12
2005, annual opioid-related overdose deaths have exceededmotor vehicle deaths.
prescriptiondrugtakeback days, safe opioid prescribing guidelines, and educationprograms seek to reduce opioid misuse and/or diversion topeople who do not have prescriptions. While these strategiesare promising, none has been demonstrated in clinical trials orcontrolled observational studies to reduce overdose rates.Methadone maintenance treatment
14 15
and supervised injectionfacilities
are strategies associated with decreased fatalitiesfrom overdose in controlled studies.Naloxone is an opioid antagonist that reverses the effects of opioidoverdose.Overdoseeducationandnaloxonedistribution(OEND) programs tackle overdose by educating people at risk for overdose and bystanders in how to prevent, recognize, andrespond to an overdose. Participants in the program are trainedto recognize signs of overdose, seek help, rescue breathe, usenaloxone, and stay with the person who is overdosing. From1996 through 2010, over 50 000 potential bystanders weretrained by OEND programs in the United States, resulting inover 10 000 opioid overdose rescues with naloxone.
In March2012, the United Nations Commission on Narcotic DrugsrecognizedoverdoseasaglobalpublichealthissuethatwarrantsfocusbytheWorldHealthOrganizationandmembercountries,including the use of naloxone for the prevention of opioidoverdose.
Studies of OEND programs have demonstratedfeasibility,
increased knowledge and skills,
and aconcomitant reduction in fatal overdoses after initiation of OEND.
27 28
AcontrolledstudyofOENDandoverdoserateshasnotbeencompleted.ImplementationofOENDinMassachusettsin communities with a high burden of opioid overdose createdthe opportunity to study the impact of OEND on opioid relatedfatal overdose and acute care hospital utilization rates, usinghighburdencommunitieswithlowornoOENDimplementationas concurrent controls.
We conducted an interrupted time series analysis of annualopioidrelatedratesofoverdosefatalitiesandutilizationofacutecare hospitals comparing communities and years where OENDwasimplementedwiththosewhereitwasnot.Theanalysiswasconducted at the city/town level. Massachusetts consists of 351geographically distinct cities and towns (referred to ascommunities). We included the 19 communities with five orgreateropioidrelatedunintentionalorundeterminedintentionalfatal poisonings in each year from 2004 to 2006, which weretheyearsimmediatelyprecedingtheimplementationofOEND.
The Massachusetts OEND program
In 2006-07, two community public health agencies beganproviding OEND.
The Massachusetts Department of PublicHealth expanded the program to four more organizations in2007 and two more in 2009. These agencies, which providedHIV education and prevention services to substance users,providedOENDtopotentialoverdosebystandersthroughtrainednon-medical public health workers under a standing order fromtheOENDmedicaldirector.Potentialoverdosebystanderswereopioidusersatriskforoverdose,aswellassocialserviceagencystaff,family,andfriendsofopioidusers.Trainingsitesincludedsyringe access programs, HIV education drop-in centres,addictiontreatmentprograms,emergencyandprimaryhealthcaresettings, and community meetings, such as support groups forfamily members of opioid users.Training curriculums were initially developed by the HarmReduction Coalition and the Chicago Recovery Alliance,
19 21 27
and adapted for nasal naloxone. OEND trainers completed afourhourcourse,knowledgetest,andtwotrainingsofpotentialbystanders supervised by a master trainer. The training of program participants by OEND trainers were conducted ingroups or individually, took as little as 10 minutes for enrolleeswith substantial pre-existing knowledge and as much as 60minutes for groups that generated discussion or had enrolleeswithout prior knowledge of overdose, and were tailored to thetraining setting. Key elements included minimizing the risk of overdose by reducing polysubstance misuse (for example,concomitant alcohol, benzodiazepine, or cocaine), accountingfor reduced tolerance after abstinence, and not using alone;recognizing overdose by assessing for unresponsiveness anddecreasedrespirations;andrespondingtoanoverdosebyseekinghelp,providingrescuebreathing,administeringnasalnaloxone,and staying with the person until medical personnel arrived orthe person recovered. Trainings concluded with enrolleesdemonstratingproperassemblyofthenaloxonedeviceandhownaloxone should be administered. Naloxone rescue kitscontained instructions, two prefilled syringes with 2 mg/2 mLnaloxonehydrochloride,andtwomucosalatomizationdevices.Two doses were included in case one dose was not sufficientorifoverdosesymptomsreturned,becausethehalf-lifeofmanyopioids is longer than that of naloxone.
Data collection and measures
Fatal opioid overdose rates 
For the fatal opioid overdose outcome, we calculated rates of unintentional and undetermined intentional opioid related drugpoisonings by community of residence using in-state occurrentdeaths from the electronic database maintained by theMassachusetts Registry of Vital Records and Statistics,Massachusetts Department of Public Health. Death certificateson fatal poisonings in Massachusetts are completed through asingle centralized, statewide office of the chief medicalexaminer,wheretheyarerequiredbylawtobereported.Opioidrelated deaths were defined by ICD-10 (internationalclassification of diseases, 10th revision) codes indicatingunintentionalorundeterminedintentionalpoisoning(X40-X44,Y10-Y14) in the underlying cause of death field and an opioidspecificTcodeofT40.0-T40.4and/orthenarcoticTcodeT40.6in any of the multiple cause of death fields. The use of T40.6toidentifyopioidrelateddeathsisrecommendedinjurisdictionswhere a high proportion of deaths with this code is opioidspecific.
An unpublished review of 2007 Massachusetts deathcertificate literals indicated that T40.6 had a positive predictivevalue of 98% for an opioid related death. Furthermore, 96.7%ofunintentionalorundeterminedintentionaldeathsbypoisoningin Massachusetts in 2007 received at least one ICD-10 code inthe range (T36-T50.8), indicating that specific information onagent or class of agent was present on death certificates fornearly all drug related deaths.
Opioid overdose related acute care hospital utilization rates 
We used the Massachusetts inpatient hospital and outpatientemergencydepartmentdischargedatabasesadministeredbytheMassachusetts Division of Health Care Finance and Policy toquantifyacutecarehospitalinpatientandemergencydepartmentdischarges associated with opioid poisoning by city or town of residence. Submission of external cause of injury codes (Ecodes) are required by state regulation on all cases with a
No commercial reuse: See rights and reprintshttp://www.bmj.com/permissionsSubscribe:http://www.bmj.com/subscribe
2013;346:f174 doi: 10.1136/bmj.f174 (Published 31 January 2013) Page 2 of 12
principlediagnosisofinjuryorpoisoning,ensuringhighqualitydata for state injury surveillance. Cases were defined asdischargeshavinganICD-9-CM(internationalclassificationof diseases, ninth revision, clinical modification) code of one ormore of the following opioid related discharge diagnosis or Ecodes: 965 (.00, .01, .02, .09), E850 (.0, .1, .2). We excludedthose cases receiving an E code indicating that the poisoningwastheresultofintentionalselfharm,assault,anadverseeffectof a drug in therapeutic use, or legal intervention. To avoidduplicate counts with the fatality measure we excluded deathsoccurring during the hospital event from this outcome.
Descriptive variables from enrollment andnaloxone rescue attempt questionnaires
TheMassachusettsDepartmentofPublicHealthOENDprogramdatabase included information from program questionnairescollectedatbothenrollmentandwheneveranenrolleerequestedan additional naloxone kit. The completed questionnaires werescannedbyformreadingsoftwareandenteredintotheprogramdatabase.Atenrollment,zipcodeofresidence,drugusehistory,and overdose history were collected. We defined users asparticipants who reported active use or being in treatment or inrecovery.Non-userswereallotherparticipants,typicallysocialservice agency staff, family, and friends of opioid users. Aquestionnaire was completed when a participant requested analoxone refill because naloxone had been used during anoverdose rescue. Staff were trained to define an overdose whenadministeringthequestionnaireasanepisodewhenanindividualwas unresponsive and had signs of respiratory depression afterusing substances. We only counted events where participantsreported their own overdose rescue attempts if another personadministered the naloxone. Self administered naloxone wasrarelyreportedandwasnotcountedasarescueattemptbecausea person able to self administer the drug was not considered tobe unresponsive. We considered naloxone to be successfullyadministered if the person’s unresponsiveness and respiratorydepression improved. Other descriptive variables included thezip code of the place in which the overdose occurred,relationship to the person who overdosed, setting (public orprivate),numberofnaloxonedosesused,whethernaloxonewassuccessful, emergency medical system involvement, rescuebreathing, and staying with the person who overdosed.
To determine the cumulative enrollment rates for the 19communities with high overdose burdens we used thecommunity of residence based on the zip code of residence onthe enrollment questionnaire. We modeled OENDimplementation in two ways. Firstly, we categorized OENDimplementationintothreegroupswithineachyearbasedonthemedian cumulative enrollment rate (relative model). Groupsincluded community-year strata with no implementation, thosebelow the median (low implementers), and community-yearstrata with enrollment rates above the median (highimplementers).Secondly,todetermineifanabsolutepopulationdensity of enrollment was associated with overdose rates, wecategorizedcommunitiesineachyearintothreecategoriesbasedon cumulative enrollment rate levels of no implementation,1-100per100000populationand>100per100000population(absolute model). In the models we used enrollment cut pointsof0,1-75,>75andinsensitivityanalysescutpointsof0,1-150,and >150.
To account for geographic differences in overdose risk, weadjustedouranalysesfordemographics.Welinearlyinterpolatedcommunity specific data (age, sex, race or ethnicity, poverty)for each year from the community specific 2000
and 2010
US Census Bureau data.We used data from the Massachusetts prescription drugmonitoringprogramtoadjustforopioidprescriptionsto“doctorshoppers,” defined as individuals who had schedule II opioidprescriptions from four or more prescribers and filledprescriptionsatfourormorepharmaciesina12monthperiod.
WecalculatedtheproportionofscheduleIIopioidprescriptionsdispensed to doctor shoppers per total opioid prescriptions foreach community-year stratum.Inpatient medically supervised withdrawal (detox) results in aperiodofabstinencethatcanincreaseoverdoserates,
33 34
whereasengagement in methadone treatment results in decreased ratesof overdose.
Office based buprenorphine treatment expandedduring the study period. To adjust for these three treatmentservices, we calculated population rates of methadonemaintenance,buprenorphinemaintenance,anddetoxeventsforeachcommunity-yearstratumusingdatafromtheMassachusettsDepartment of Public Health Bureau of Substance AbuseServices treatment database. Any treatment program licensedby and contracted with the Substance Abuse Services wasrequired to report admission and discharge information.We accounted for linear trends over the study period by usinga time variable T, expressed as 1 for the index year 2002 andincreasing in integer increments for each year of the studyperiod.
Statistical analysis
For the interrupted time series we used the annual rates of fatalopioid related overdose and acute care hospital utilizationassociated with non-fatal opioid overdose by community of residence for the units of analysis. The denominators were thecommunity population based on US Census estimates. Basedon the independent variable definitions, we coded individualcommunity-year combinations with an indicator variabledenoting“implementation.Asinotherstudiesofinjurytrends
and program implementation,
we used Poisson regressionmodels to test our hypotheses that those community-year stratawith higher implementation would have lower rates. Wemodeled rates directly with a log-linear statistical model byincluding counts as the dependent outcome and population atrisk as an offset term. We controlled for community levelcovariates by including them in the model. All hypothesis testsused a significance level (α) of 0.05. We performed regressiondiagnostics,includingquasilikelihoodinformationcriteria,
toassess goodness of fit. Based on these, we chose first orderautoregressive covariance structure to account for theinterdependence of repeated measures.To determine whether our findings were specific to overdoseoutcomes or due to an unmeasured health system effect, suchas healthcare reform, we conducted additional sensitivityanalyses. We refit the adjusted Poisson fatal overdose modelssubstituting fatal opioid overdose rates with overdose death tocancer death rate ratios for each community-year stratum usingdata from the Massachusetts Registry of Vital Records andStatistics.Wealsorefitmodelssubstitutingacutecareutilizationrates associated with non-fatal opioid related poisoning withnon-fatal opioid related poisoning or non-fatal motor vehicletraffic related injury rate ratio for each community-yearstratum.
To define cancer deaths we used ICD-10 codes
No commercial reuse: See rights and reprintshttp://www.bmj.com/permissionsSubscribe:http://www.bmj.com/subscribe
2013;346:f174 doi: 10.1136/bmj.f174 (Published 31 January 2013) Page 3 of 12

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->