Sedative and hypnotic drugs-Fatal and non-fatal reference blood concentrations

(1)

Sedative and hypnotic drugs-Fatal and

non-fatal reference blood concentrations

Anna K Jönsson, Carl Soderberg, Ketil Arne Espnes, Johan Ahlner, Anders Eriksson,

Margareta Reis and Henrik Druid

Linköping University Post Print

N.B.: When citing this work, cite the original article.

Original Publication:

Anna K Jönsson, Carl Soderberg, Ketil Arne Espnes, Johan Ahlner, Anders Eriksson,

Margareta Reis and Henrik Druid, Sedative and hypnotic drugs-Fatal and non-fatal reference

blood concentrations, 2014, Forensic Science International, (236), 138-145.

http://dx.doi.org/10.1016/j.forsciint.2014.01.005

Copyright: Elsevier

http://www.elsevier.com/

Postprint available at: Linköping University Electronic Press

(2)

AUTHOR QUERY FORM

Journal: FSI

Please e-mail or fax your responses and any corrections to:

E-mail:

corrections.eseo@elsevier.thomsondigital.com

Article Number: 7480

Fax: +353 6170 9272

Dear Author,

Please check your proof carefully and mark all corrections at the appropriate place in the proof (e.g., by using on-screen

annotation in the PDF file) or compile them in a separate list. Note: if you opt to annotate the file with software other than

Adobe Reader then please also highlight the appropriate place in the PDF file. To ensure fast publication of your paper please

return your corrections within 48 hours.

For correction or revision of any artwork, please consult

http://www.elsevier.com/artworkinstructions

.

Any queries or remarks that have arisen during the processing of your manuscript are listed below and highlighted by flags in

the proof. Click on the ‘

Q

’ link to go to the location in the proof.

Location in

Query / Remark:

click on the Q link to go

article

Please insert your reply or correction at the corresponding line in the proof

Q1

Please confirm that given names and surnames have been identified correctly.

Q2

The affiliation ‘a’ has been split into two different affiliations. Please check, and correct if necessary.

Q3

Please check the edit made in the sentence “To improve the interpretation....”, and correct if necessary.

Please check this box or indicate your approval if

you have no corrections to make to the PDF file

(3)

1

2

Sedative

and

hypnotic

drugs

—F

atal

and

non-fatal

reference

3

blood

concentrations

4 Q1

Anna

Kristina

Jo¨nsson

a,b

,

Carl

So¨derberg

c

,

Ketil

Arne

Espnes

d

,

Johan

Ahlner

e

,

5

Anders

Eriksson

f

,

Margareta

Reis

a

,

Henrik

Druid

c,

*

6 a

DepartmentofDrugResearch/ClinicalPharmacology,

Q2 FacultyofHealthSciences,Linko¨pingUniversity,Linko¨ping,Sweden

7 b

DepartmentofClinicalPharmacology,CountyCouncilofO¨stergo¨tland,Linko¨ping,Sweden

8 c

ForensicMedicineLaboratory,DepartmentofOncology-Pathology,KarolinskaInstitute,Stockholm,Sweden

9 d_Department_of_Clinical_{Pharmacology,}_St._Olav_University_Hospital,_Trondheim,_Norway

10 e

DepartmentofForensicGeneticsandForensicToxicology,NationalBoardofForensicMedicine,Linko¨ping,Sweden

11 f

SectionofForensicMedicine,DepartmentofCommunityMedicineandRehabilitation,Umea˚ University,Umea˚,Sweden

12

13 1. Introduction

14 Statisticsoffatalintoxicationsarehighlydependentonreliable 15 informationfrommedico-legaldeathinvestigations.Itishowever 16 a challengeto establish which drug/sis/are themain cause of 17 death.Referencedatabasedonpostmortemanalyticalresultsare

18 particularly valuable. Compilations of therapeutic and toxic

19 concentrations [1–3] in living subjects are often used when

20 interpreting the resultsof thetoxicological analyses.However,

21 suchconcentrationscannotdirectlybetranslatedtoapostmortem

22 setting [4–7]. Another problem is that compilations based on

23 postmortemdataareamixofliteraturereviewsandcasereports.

24 Thisisproblematicsincedruglevelsthatmaybeobservedincases

25 of death attributed to causes other than intoxication are not

26 commonly included. Furthermore, there are discrepancies in

27 samples chosen for analysis, sampling procedures, analytical

28 methods,selectionof cases,and methodologicalprocedures[8].

29 In Sweden, the blood sampling and handling procedures are

ForensicScienceInternationalxxx(2014)xxx–xxx

ARTICLE INFO

Articlehistory:

Received6September2013

Receivedinrevisedform30December2013 Accepted5January2014 Availableonlinexxx Keywords: Postmortem Intoxication Poisoning Forensictoxicology Toxicity Benzodiazepines ABSTRACT

Inpostmorteminvestigationsoffatalintoxicationsitisoftenchallengingtodeterminewhichdrug/s causedthedeath.Toimprovetheinterpretationofpostmortembloodconcentrationsofsedativeand hypnoticdrugsand/orclonazepam,allmedico-legalautopsiesinSweden–wherethesedrugshadbeen detectedinfemoralveinbloodduring1992–2006–wereidentifiedinthedatabasesoftheNationalQ3 BoardofForensicMedicine.Foreachdrug,concentrationsinpostmortemcontrolcases–wherethe causeofdeathwas notintoxicationand whereincapacitation bydrugscouldbe excluded–were compiledaswellasthelevelsfoundinlivingsubjects;druggeddrivingcasesandtherapeuticdrug monitoring cases. Subsequently, fatal intoxications were assessed with regards to the primary substances contributing to death, and blood levels were compiled for single and multiple drug intoxications.Thepostmortemfemoralbloodlevelsarereportedfor16sedativeandhypnoticdrugs, basedonfindingsin3560autopsycases.Thecaseswereclassifiedassinglesubstanceintoxications (N=498), multiple substance intoxications (N=1555) and postmortem controls (N=1507). Each autopsycasecouldberepresentedmorethanonceinthegroupofmultipleintoxicationsandamongthe postmortemcontrolsifmorethanoneoftheincludedsubstancesweredetected.Theconcentration rangesforallgroupsareprovided.Overlapinconcentrationsbetweenfatalintoxicationsandreference groups was seen formost substances.However, the concentrations found insingle and multiple intoxications were significantly higher than concentrations found in postmortem controls forall substancesexceptalprazolamandtriazolam.Concentrationsobservedamongdruggeddriverswere similartotheconcentrationsobservedamongthetherapeuticdrugmonitoringcases.Flunitrazepamwas thesubstancewiththehighestnumberofsingleintoxications,whenrelatedtosales.Insummary,this studyprovides referencedrugconcentrationsprimarily tobe used forimprovinginterpretationof postmortem druglevels inobscure cases, but which also may assist in drug safetywork and in pharmacovigilanceefforts.

ß2014PublishedbyElsevierIrelandLtd.

* Corresponding author at: Forensic Medicine Laboratory, Department of Oncology-Pathology, Karolinska Institute, Retzius v. 3, SE-171 77 Stockholm, Sweden.Tel.:+46703447544;fax:+4613364270.

E-mailaddress:henrik.druid@ki.se(H.Druid). GModel

FSI74801–8

Pleasecitethisarticleinpressas:A.K.Jo¨nsson,etal.,Sedativeandhypnoticdrugs—Fatalandnon-fatalreferencebloodconcentrations, ForensicSci.Int.(2014),http://dx.doi.org/10.1016/j.forsciint.2014.01.005

ContentslistsavailableatScienceDirect

Forensic

Science

International

j o urn a lhom e pa g e :ww w . e l se v i e r. c om / l oca t e / f ors ci i nt

0379-0738/$–seefrontmatterß2014PublishedbyElsevierIrelandLtd.

(4)

30 standardized and all analyses are performed at one national, 31 accreditedlaboratory.Fatal and non-fatal referenceblood con-32 centrations in Sweden have previously been assessed for 83 33 substances in 1997[9],for ﬂunitrazepamin 2001 [10] and for 34 antidepressantsin2007[11].

35 The aim of this study was to evaluate fatal and non-fatal 36 referencebloodconcentrationsinSweden,withafocusonsedative 37 andhypnoticdrugs,tocomparetheseconcentrationswithliving 38 controlsand torelatethenumberof fatalintoxications todrug 39 salesstatistics.

40 2. Methods

41 2.1. Studypopulations

42 Thiscompilationwasbasedon drugconcentrationsin blood 43 samples from medico-legal autopsies and individuals driving 44 undertheinﬂuence(DUI)inSwedenandfromtherapeuticdrug 45 monitoring(TDM)casesinNorway.

46 2.1.1. Postmortemcases

47 In Sweden,mostsuspectedandcertiﬁedunnaturaldeathsas 48 wellasunexplaineddeathsarereportedtothepolice.Thepolice 49 request a medico-legal autopsy in a majority of these cases. 50 Swedenhassixforensic medicinedepartments andall ofthem 51 follow a strictly standardized procedure for the sampling of 52 femoralvenousblood[9,12]toreducetheimpactofpostmortem 53 redistribution [4,6,9]. All drug analyses are performed at one 54 national,accreditedlaboratory[9],wherethebloodisscreenedfor 55 alcohols,pharmaceuticaldrugs and upon requestalsofor illicit 56 drugs.Alltoxicologicalresultsarerecordedinanationaldatabase 57 [12]. Femoral venous blood concentrations are consistently 58 expressedin

m

g/gsincetheyarebasedonweighedsamples.This 59 isalsotruefortheDUIcases.Theconcentrationscanbeconverted 60 to

m

g/mLbymultiplyingwith1.06(theaveragedensityofblood). 61 2.1.2. Inclusioncriteriaandassessmentprocedure

62 Allmedico-legalautopsycasesinwhichsedativeandhypnotic 63 drugs (ATC codes N05B and N05C)and clonazepam(ATC code 64 N03AE01)weredetectedinfemoralbloodduringthestudyperiod 65 fromJanuary1992throughDecember2006wereidentifiedinthe 66 combinedforensicmedicineandtoxicologynationaldatabase[12]. 67 Thisconstitutesareal-timedatabaseandalldataarecontinuously 68 generatedfromtheroutinecaseworkdatamanagementsystemfor 69 forensicmedicineandtoxicology.TheICD-9system,withsome 70 supplementaldiagnosestoimprovespecificity,isusedtotranslate 71 causesofdeathintocodes.Basedonthecauseofdeath,thecases 72 weredividedintothreemutuallyexclusivegroupsaccordingtoa 73 previously described procedure [9,11]. The first two groups 74 includedcaseswithintoxicationastheimmediatecauseofdeath; 75 inGroupAbyonedrug,andinGroupBbytwoormoredrugsand/ 76 orethanol.GroupC,thepostmortemcontrolcasegroup,included 77 violentsuicidesandselectaccidentaltraumadeaths(ICD9codes: 78 800–959,E953,E955,E956,orE958).Ifmorethanonesedativeor 79 hypnoticdrugweredetected,thiscasecouldberepresentedmore 80 thanonceinGroupBorGroupC.Caseswithacauseofdeaththat 81 mayimplyincapacitationwereexcluded(Table1)aswellascases 82 where death occurred in a hospital. Each case was reviewed 83 independentlybytwooftheauthors.Forcaseswithunexpectedly 84 highorlowconcentrations,andincaseswhereanyotherqueries 85 wereraised during theevaluation, the original data files were 86 examined in detail. Cases where the assessment between the 87 authorsdifferedwerediscusseduntilaconsensuswasreached. 88 Inordertodeterminewhichconcentrationofadrugthatcould 89 beregardedasinsignificantwithregardstotoxicity,a concentra-90 tionwindowwasdefinedforeach sedative/hypnoticsubstance,

91 usingadditionalinformationrelatingtoconcentrationsfoundin

92 theGroupCcases,theDUIcasesandtheTDMdata(seebelow).

93 Further, an extensive literature search on maximum serum

94 concentrations in clinical trials and in other TDM populations

95 wasperformedtoidentifyobviousnon-toxicserumlevelsforeach

96 drug. These arbitrary concentration windows were used as

97 guidelinelevelswhenreviewingtheGroupAandGroupBcases.

98 Whenevaluatingandlatercompilingthedruglevelsidentiﬁedin

99 GroupsA–Ctheconcentrationsoftheparentcompoundwereused.

100 However,forclonazepam,nitrazepam,ﬂunitrazepamand

propio-101 mazine, the 7-amino metabolites, and dihydropropiomazine,

102 respectively,wereaddedtotheparentcompounds,sinceallthese

103 metabolitesarealmostexclusivelyformedpostmortem.

104 2.1.3. Individualsdrivingundertheinﬂuence

105 The DUI material consisted of drivers apprehended while

106 driving under the inﬂuence of substances other than alcohol,

107 includingpharmaceuticaldrugsandillicitdrugs.AllDUIcasesin

108 Sweden where sedative and hypnotic drugs were detected in

109 blood,fromJanuary1992throughDecember2006,wereidentiﬁed

110 andformedGroupD.Caseswherethesubstanceofinterestwas

111 detected in lower concentrations than the LoQ of postmortem

112 analyses were excluded to enable group comparisons. Each

113 individualcouldofcoursebesuspectedofdruggeddrivingmore

114 than once during the study period. However, only the ﬁrst

115 analyticalresultforeachsubstancedetectedinasingleindividual

116 was used. If more than one sedative and hypnotic drug were

117 detectedintheanalyses,oneindividualcouldberepresentedmore

118 thanonceinthedataset.

119 2.1.4. Therapeuticdrugmonitoringcases

120 At the Department of Clinical Pharmacology in Trondheim,

121 Norway,serumsamplesfrompatientstreatedwithsedativeand

122 hypnoticdrugs wereanalyzedupon requestby theresponsible

123 psychiatrist orgeneral practitioner.TheTDMsamplescollected

124 during theyears 1999through 2007 wereassessed. To ensure

125 validityoftheTDMdata,onlyonesampleperpatientwasused.

126 Based on the free text in the database, usually including the

127 requestingdoctors’notes,intentionalandunintentionaloverdose

128 cases were excluded [11]. All TDM concentration data were

129 originallygivenasnmol/Lserumwhereasthepostmortemandthe

Table1

ExclusioncriteriaforGroupCbasedonICD9codes(withsupplementarysufﬁces)a

duetopossibleincapacitationorsevereorganinjury. Causeofdeath(codes) Deﬁnition

852 Subduralhemorrhage

861K Injurytoheartandlung,gunshotwound

861M Injurytoheartandlung,laceration

864 Liverlaceration

869 Severeexternalandinternalinjuries

933 Foreignbodyinpharynxorlarynx

934 Foreignbodyintrachea,bronchiorlungs

940–949 Burns

991G Hypothermia

994B Drowningorothersubmersion Mannerofdeath(codes)

E850–E858 Accidentalpoisoningbydrugs

E880–E888 Accidentalfall

E910 Accidentaldrowningandothersubmersion

E954 Suicidebysubmersion

E958A Suicidebyjumping/lyingbeforeatrain

E958B Suicidebyﬁre

E958D Suicidebyhypothermia

E960–E969 Homicide/injuryinﬂictedbyotherperson

E980–E989 Injuryundetermined

a _Cer_tain_sufﬁces_are_supplements_to_the_ICD-9_codes_deﬁned_by_the_Swedish Medico-legalSocietyandusedbytheSwedishforensicpathologiststoprovidemore detaileddiagnoses.

A.K.Jo¨nssonetal./ForensicScienceInternationalxxx(2014)xxx–xxx 2

(5)

130 DUIdataweregivenas

m

g/gwhole blood.In ordertofacilitate 131 comparisons,theTDMdatawererecalculatedintothesameunit, 132 i.e.

m

g/g.Anassumptionwasmadethat1mLserumweighed1g.A 133 transformation factorwas calculated bythe equation10 6/Mw 134 (molecularweight)toconverttheconcentrationinnmol/Linto

m

g/ 135 g.Furthermore,thesamedruglimitsofquantiﬁcationasforthe 136 postmortemcaseswereappliedontheTDM-datatoenablegroup 137 comparisons (Table 2). Hence, by introducing arbitrary cut-off 138 detectionlevelsfortheTDMpopulationa‘‘true’’TDMcohortisnot 139 displayed in this paper but rather a restricted population for 140 comparisonpurposes.HenceforththeTDMdataarereferredtoas 141 GroupT.

142 2.2. Analyticalmethodsusedinmedico-legalcases(GroupsA–C) 143 FemoralbloodfromautopsycasesandwholebloodfromDUI 144 caseswereanalyzedatthesamelaboratoryusingthesamemethod 145 formost of the substances[9]. Inbrief, this methodimpliesan 146 alkaline and a neutral liquid–liquid extraction and gas chro-147 matographic(GC)analysis utilizingHP5890GCsequippedwith 148 HP7673autoinjectorsandNPdetectors.Alprazolam,clonazepam, 149 diazepam,ﬂunitrazepam,midazolam,nitrazepam,zaleplon, zolpi-150 dem, zopiclone, propiomazine, hydroxyzin and buspirone (until 151 2005)were analyzedinallmedico-legalcases.TheLoQs usedin 152 postmortem analyses are shown in Table 2. For clonazepam, 153 diazepam,ﬂunitrazepam,nitrazepamandpropiomazinethe meta-154 boliteswereanalyzedusingthesamemethodsastheirparentdrug. 155 Lorazepam,triazolam, clomethiazole, buspirone (from 2005) 156 and oxazepam were analyzed upon request by the forensic 157 pathologist.Clomethiazolewasextractedbybutyl-acetateatpH 158 7.0. The internal standard was allobarbital and prazepam. 159 Clomethiazole was separated using a DB-5 column and was 160 detectedusingNPD[13].Oxazepamwasextractedbyliquid–liquid 161 extractionatpH8.9andwasdetectedusingGC–UVwithdiode 162 arraydetection.Theorganicphasewasevaporatedanddissolved 163 in the mobilephase and injected on Agilent 1100 seriesHPLC 164 systemsafterseparationonC8columns.Buspironewasextracted 165 (from2005)bybutyl-acetateatpH11.0andanalyzedbyLC–MS– 166 MSonaPESciexAPI2000triplequadrupleinstrumentequipped

167 withturboion-sprayinterfaceafterseparationonaZorbaxStable

168 BlondCyanocolumn[14].

169 Externalaccuracywascheckedroutinelywithexternalcontrol

170 samples and precision was calculated from quality control

171 samples. As an example the precision over 12 months was

172 between4.8and14.0%forthebenzodiazepinesandtheZ-drugs

173 (N=220–450),10.0–14.4%forhydroxyzine,propiomazineandits

174 metabolite,and12.2% forbuspirone.Externalcontrolsgenerally

175 hadz-scoreslessthan1andalmostalwaysz-scoreslessthan2.

176 2.3. AnalyticalmethodsusedintheTDMcases

177 Alprazolam,diazepam,ﬂunitrazepam,clonazepam,nitrazepam,

178 oxazepam,meprobamate,zopicloneandzolpidemwereextracted

179 from serum with liquid–liquid extraction using a mixture of

180 dichloromethane/isopropanol (90:10) as extractionsolvent.

Pro-181 methazinewasextractedfromserumwithliquid–liquidextraction

182 usingamixtureofhexane/acetonitrile(98:2)asextractionsolvent.

183 All analytes were quantiﬁedby liquidchromatography–mass

184 spectrometry(LC–MS)onAgilentMSD1100LC–MSsystemsafter

185 separationonC18columnsandsubsequentdetectionas

pseudo-186 molecularions(M+1).Internalstandardswerealwaysused.Mobile

187 phasesconsistedofmethanol,acetonitrile,formicacid(25mmol/L)

188 orammoniumacetate(50mmol/L)indifferentratios.Togetherwith

189 the unknown samples, each analytical series contained seven

190 calibrators coveringtherapeutic, sub-therapeutic and toxic

con-191 centrations.Inaddition,threequalitycontrol (QC)samples with

192 representativetargetlevelswerealwaysincluded.

193 Analyticalaccuracywascheckedroutinelywithexternalcontrol

194 samples when available, and between series precision was

195 calculatedfromqualitycontrolsamples.Asanexampletheprecision

196 over12monthswasbetween5.5and12.0%forthebenzodiazepines

197 andtheZ-drugs.Externalcontrolsgenerallyhadz-scoresbetween

198 1and1andalmostalwaysz-scoresbetween 2and2.

199 2.4. Fataltoxicityindex

200 Fataltoxicityindex[11,15–17]wasestimatedbyrelatingthe

201 numberofcasesclassiﬁedasGroupsAandBtothesalesofthat

202 substanceinSwedenduring1992through2006.Salestatisticsfor

203 theincludedsubstances,deﬁnedasthenumberofDDD(deﬁned

204 dailydoses)soldinSweden,wasretrievedfromApotekensService

205 AB.Theanalysesofthefataltoxicityindexwererestrictedtothe

206 years 2000–2006for hydroxyzine,propiomazine, clomethiazole,

207 clonazepam, zopiclone,zolpidem, buspirone and zaleplon since

208 salesstatisticsfromtheearlieryearswerenotavailableforthese

209 substances. Wealsocompared thenumberofAandBcasesfor

210 eachsubstancewiththetotalnumberofdetectionsinpostmortem

211 casesasanadditionalmeansofestimatingtherelativetoxicity.

212 2.5. Statisticalanalyses

213 The Mann–Whitney U-test (unevenly distributed data) was

214 used for group comparisons. The p-values were adjusted for

215 multiplecomparisonsusingtheHolmprocedure[18].Statistical

216 signiﬁcancewasdeﬁnedasp<0.05.TheRsoftwarev2.15.0was

217 usedforstatisticalanalysis[19].

218 3. Results

219 3.1. Referenceconcentrations

220 Duringthestudyperiod,toxicologicalanalyseswereperformed

221 in73,321autopsies,andsedativesandhypnoticswereidentiﬁedin

222 12,608(17%)ofthese.Aftertheassessmentprocedure3560(29%)

223 individualswereincludedinthestudy;498individualsinGroupA,

Table2

Lower limitofquantiﬁcation for theincludedsubstances inthe postmortem analyses.

Substance Postmortemanalyses(mg/g)

Alprazolam 0.02 Buspirone 0.05a Clomethiazole 0.2 Clonazepam 0.05 7-Aminoclonazepam 0.05 Diazepam 0.05 Nordazepam 0.05 Flunitrazepam 0.02b 7-Aminoﬂunitrazepam 0.02b Hydroxyzine 0.05 Lorazepam 0.02 Midazolam 0.02 Nitrazepam 0.05 7-Aminonitrazepam 0.05 Oxazepam 0.1 Propiomazine 0.03 Dihydropropiomazine 0.03 Triazolam 0.002 Zaleplon 0.05 Zolpidem 0.05 Zopiclone 0.02

Abbreviations:DUI,Drivingundertheinﬂuence.

a

Until2005.From2006theLoQwas0.001mg/g.

b

TheLoQwas0.01mg/guntil1998and0.02mg/gthereafter.

A.K.Jo¨nssonetal./ForensicScienceInternationalxxx(2014)xxx–xxx 3 GModel

FSI74801–8

(6)

224 1555inGroupBand1507inGroupC.InGroupD7455individuals 225 wereincluded.Sedative and hypnotics wereidentiﬁed in 3593 226 TDManalyses,ofwhich1477wereincludedinGroupT. 227 Themedianagewas66yearsinGroupA;50yearsinGroupB;53 228 yearsinGroupC;32yearsinGroupDand;42yearsinGroupT.The 229 proportionofwomenwas54%(N=271)inGroupA,49%(N=766)in 230 GroupB,27%(N=403)inGroupC,15%(N=1099)inGroupD,and 231 57%(N=841)inGroupT.Amongthepostmortemcases,thetwo 232 maincausesofdeathinGroupCwerehanging(70%,N=1062)and 233 CNStraumasuchasgunshotwoundtothehead(19%,N=286).The 234 causeofdeathinGroupsAandBwasbydeﬁnitionintoxication,and 235 the manner of death was most often suicide; 76% in Group A 236 (N=376)and66%inGroupB(N=1124).

237 Table3 showsthefemoral bloodconcentrationsofsedatives

238 andhypnotics in intoxications and in controls. For nitrazepam, 239 ﬂunitrazepam,clonazepamandpropiomazine,theconcentrations 240 presentedinTable3arethesumoftheparentdrugandmetabolite 241 concentrations.

242 ThemedianconcentrationsinGroupAweresignificantlyhigher 243 than in Group B for oxazepam, propiomazine, zopiclone and 244 zolpidem,andthemedianconcentrationsinGroupsAandBwere 245 significantly higher than in Group C for all substances except 246 alprazolamandtriazolam.Althoughsignificantdifferencesinthe 247 median concentrations among groups were observed, the con-248 centrationsinsomegroups overlapped. Themedian concentra-249 tionsinGroupTweresimilartotheconcentrationsobservedin 250 GroupDformostsubstances (alprazolam,clonazepam, flunitra-251 zepam, nitrazepam, oxazepam, and zopiclone). However, for 252 diazepamthemedianconcentrationinGroupTwashigherthan 253 inGroupD.Formidazolamnocasesmettheinclusioncriteriafor 254 GroupC. Midazolam washowever detected in 46 postmortem 255 casesthatwereexcludedsincetheydiedinhospital.Inthosecases 256 the median concentration was 0.1

m

g/g femoral blood with 257 0.03

m

g/g and 0.5

m

g/g as the 10th and the 90th percentile, 258 respectively.Regardingtriazolam,onesubjectdiedinhospitaland 259 wasnotincludedinGroupC.Inthiscasetriazolamwasdetectedat 260 aconcentrationof0.01

m

g/gfemoralblood.

261 3.2. Fataltoxicityindex

262 Table4showsthenumberofintoxicationsforeachsubstance

263 relatedtotherespectivedrugsalesinSweden,aswellastothe 264 total number of cases where that substance was detected in 265 medico-legalautopsiesinSweden.Hydroxyzinewasthesubstance 266 with the highest number of fatal intoxications related to the 267 numberofDDDsold(2.02intoxicationsper106DDDsold).There 268 werehowevernotonesingleintoxicationduringthistimeperiod 269 wherehydroxyzinealonewasdeterminedtobetheprimarycause 270 ofdeath.Propiomazinehadthehighestnumberfatalintoxications 271 (GroupAandGroupBcombined)relatedtothetotalnumberof 272 detectionsinmedico-legalautopsies(24%),whereasﬂunitrazepam 273 had the highest number of fatal intoxications (Group A alone) 274 relatedtosales(0.43casesper106DDDsold)aswellasrelatedto 275 thetotalnumberofdetections(8%).Whencombined,thehypnotic 276 drugshadahighernumberoffatalintoxications(GroupsAandB 277 combined)related tosales compared withtheanxiolyticdrugs 278 (1.13and0.37casesper106DDDsold,respectively)aswellas 279 whenrelatedtothenumberofdetectionsinfemoralbloodamong 280 autopsycases(22%and5%,respectively).

281 4. Discussion 282 4.1. General

283 In this study we present reference values in postmortem 284 femoralbloodfor16sedative-hypnoticdrugsandclonazepamin

285 fatalintoxicationsandinpostmortemcontrols.Thesevalueswere

286 alsocomparedwithconcentrationsdetectedinDUIcasesandina

287 TDMpopulation.Referenceinformationaboutpostmortemblood

288 drugconcentrationscanbefoundinpublishedcompilations(e.g.

289

[3,20,21])butalmostalloftheseconsistofliteraturereviewsand

290 thequalityoftheoriginalpublicationsreferredtovaries.Insome

291 studies,thetypeofbloodsampleanalyzedisnotspeciﬁed.Inthis

292 study,likeinthreepreviouspublications[9–11]onthesameissue,

293 theresultsarebasedonastrategythatprovidesboth

concentra-294 tionsinfatalintoxicationsandinpostmortemcontrols.GroupCis

295 of particular importance, as it provides a better idea of drug

296 concentrationsthatmaybeobservedwithoutcausing

incapacita-297 tionimmediatelypriortodeath.

298 4.2. Methodologicalaspects

299 Itshouldbenotedthattheuseofreferenceinformationofthis

300 kindisintendedtoassistintheinvestigationofsuspectedacute

301 intoxicationsorobscuredeathsthatoccuroutsidehospitals.The

302 compilationexcludescaseswherethesubjectsdiedathospitalfor

303 twomainreasons;ﬁrstly,thecircumstancessurroundingdeathin

304 hospitalareoftenclearandmanymedicaldataarealreadyathand;

305 secondly, the treatment may include various pharmacological

306 interventions,artiﬁcialventilationand otherlife-savingmedical

307 support,distortingtheinterpretationofthedrugeffects.Wedo,

308 however,presentseparatelysomedatafromhospitalizedsubjects

309 formidazolamtoprovideanideaoftheconcentrationspectrumin

310 thissetting.Itshouldfurtherbepointedoutthatreferencelevels

311 donotreplaceameticulousanalysisoftheindividualcase.

312 RegardingDUIcases,itis obviousthatmany ofthesubjects

313 mighthavebeeninebriated,butstill,allofthemwerealiveand

314 wereabletodriveavehicle–althoughnotnecessarilysafely.The

315 concentrationscompriseallresultsobtained,withoutevaluationof

316 theindividualcases.Henceaconcentrationofasubstanceinone

317 casemightrepresentanincidentalﬁndingwhiletheconcentration

318 inanothercasemightbethereasonforpoordriving.

319 TheTDMlevelswerebasedonalargecollectionofanalysesof

320 serumsamplesfrompatientstreatedatvariousmedicalfacilities.

321 To make the concentrations identiﬁed in the TDM cohort

322 comparablewith theconcentrations identiﬁed in theDUI and

323 thepostmortemcases,thevolumeunit(nmol/L)wasconvertedto

324 amassunit(

m

g/g).Further,theforensiclimitsofquantiﬁcation

325 wereappliedontheTDMresults.Thisimpliesthatsubjectswith

326 lowdrugconcentrations(probablytherapeuticor

subtherapeu-327 tic) were excluded.The TDMconcentrations presented in this

328 paperisthereforenotrepresentativeofthewholeclinicalTDM

329 population.Thedifferencesobservedbetween GroupsC,Dand

330 GroupTmightpartlybeexplainedbydifferencesinmatrixused,

331 sinceitispossiblethatsomeofthedrugsincludedshowanuneven

332 distributionbetweenserumandtheerythrocytefraction[22–24].

333 This area is however not well explored.Accordingto Clarke’s

334 Analysis of Drugs and Poisons [24], there are known uneven

335 distribution between of the plasma: whole blood ratio for

336 alprazolam (1.35), clomethiazole (1.2), clonazepam (1.54),

337 diazepam (1.8) and oxazepam (0.9). In addition, there are

338 demographic differences between thepopulations includedin

339 thegroups.

340 Inourpresentandpreviouscompilationsofpostmortemblood

341 concentrations of drugs [9–11], we have reviewed the cases

342 without consideration of active metabolites and the results

343 presentedherearebasedonthelevelsoftheparentcompounds

344 only (withthe exceptionsofclonazepam,diazepam,

ﬂunitraze-345 pam, nitrazepam and propiomazine where the postmortem

346 metabolitewasaddedtotheparentdrug).Fortwoofthedrugs

347 included–buspironeandhydroxyzine–boththeparentdruganda

348 mainmetaboliteareactive[25–27].Fortheotherdrugsincluded,

(7)

Table3

FemoralbloodconcentrationsofdrugsinpostmortemcasesandbloodconcentrationsinDUIcasesandTDMcases(mg/g). Substance Group N 10thpercentilea

Median 90thpercentilea p-Value Alprazolamb A 4 0.13 0.30 0.40 vs.B0.23 B 67 0.10 0.16 0.60 vs.C0.08 C 51 0.02 0.05 0.12 vs.A<0.01 D 793 0.02 0.05 0.15 T 110 0.02 0.04 0.12 Buspironec A 0 B 11 0.05 0.20 1.10 C 0 D 1 0.06 0.06 0.06 T – Clomethiazolec A 5 4.50 7.00 28.0 vs.B0.4 B 8 1.10 5.40 11.0 vs.C0.05 C 5 0.30 0.40 3.00 vs.A0.04 D 21 0.40 1.20 3.60 T – Clonazepamd,e _A ₀ B 51 0.14 0.30 0.75 vs.C<0.01 C 32 0.06 0.14 0.29 D 214 0.05 0.07 0.15 T 94 0.05 0.07 0.15 Diazepamf A 3 1.20 1.40 6.00 vs.B0.85 B 20 1.09 1.45 2.60 vs.C<0.01 C 496 0.07 0.10 0.40 vs.A<0.01 D 4717 0.07 0.20 0.80 T 153 0.10 0.68 1.90 Flunitrazepamd,e A 175 0.11 0.30 0.74 vs.B0.23 B 418 0.07 0.15 0.41 vs.C<0.01 C 192 0.01 0.03 0.14 vs.A<0.01 D 555 0.02 0.03 0.06 T 28 0.02 0.04 0.07 Hydroxyzinec _A ₂ _0.70 _1.6₀ _2.50 _vs._B_0.34 B 104 0.30 0.60 3.30 vs.C<0.01 C 73 0.06 0.10 0.40 vs.A0.04 D 34 0.06 0.10 0.37 T – Lorazepamc A 0 B 2 0.29 0.55 0.80 C 0 D 35 0.04 0.1 0.37 T – Midazolamh A 0 B 1 0.6 0.6 0.6 C 0 D 30 0.02 0.06 0.12 T – Nitrazepamd,g _A ₆₄ _0.40 _1.10 _3.20 _vs._B_<0.0₁ B 258 0.30 0.60 1.70 vs.C<0.01 C 196 0.05 0.09 0.29 vs.A<0.01 D 446 0.05 0.08 0.28 T 92 0.05 0.11 0.44 Oxazepame A 4 2.39 3.85 5.59 vs.B0.02 B 79 1.08 1.80 3.44 vs.C<0.01 C 71 0.10 0.30 0.92 vs.A<0.01 D 476 0.20 0.40 1.70 T 726 0.07 0.23 1.50 Propiomazinec,d A 160 0.45 1.20 5.71 vs.B<0.01 B 660 0.23 0.60 2.70 vs.C<0.01 C 246 0.03 0.06 0.20 vs.A<0.01 D 26 0.03 0.04 0.10 T – Triazolamh _A ₁ _0.02 _0.02 _0.02 _vs._B_1.00 B 4 0.01 0.02 0.04 C 0 D 7 0.002 0.009 0.03 T –

GModel FSI74801–8

(8)

349 themetabolitesareeithernotactive,notdetectable,orpresentat 350 lowconcentrationsonly.

351 Forsomesubstances,thenumberofGroupAcases(singledrug 352 fatalintoxication)isloworevenlacking.Thequestionthenarises 353 whetherornot thesedrugsarepotentially lethal.Althoughthe 354 authorsdoconsiderthesedrugstobepotentiallylethalincertain 355 settings, caution is advised when referring to the drug levels 356 deﬁnedinthisstudy.Thereasonforpaucityorlackofsingledrug 357 intoxications with these substances might be either that they 358 typically are used in combination withother drugs with toxic

359 properties (resultingin classiﬁcation as a B case),or that they

360 actuallydohavealowtoxicity.Forthesedrugs,alargernumberof

361 cases needs to be reviewed to appreciate their toxicological

362 proﬁles.

363 4.3. Comparisonwithpreviouswork

364 The referencevalues presentedin this paper are difﬁcultto

365 compare with other reported concentrations due to major

366 differences in study design. However, for substances where

Table3(Continued)

Substance Group N 10thpercentilea

Median 90thpercentilea p-Value Zaleplonc A 0 B 0 C 2 0.06 0.1 0.40 D 5 0.07 0.10 0.14 T – Zolpidem A 25 0.6 1.50 3.32 vs.B<0.01 B 148 0.30 0.90 3.13 vs.C<0.01 C 52 0.06 0.16 0.59 vs.A<0.01 D 558 0.08 0.23 0.63 T – Zopiclonec _A ₅₄ _0.40 _0.8 _2.84 _vs._B_0.02 B 344 0.30 0.70 1.90 vs.C<0.01 C 353 0.02 0.05 0.21 vs.A<0.01 D 502 0.03 0.08 0.27 T 274 0.02 0.05 0.56

Abbreviations:DUI,drivingundertheinfluence;TDM,therapeuticdrugmonitoring.Group A:Certifieddeathbyintoxicationwithonesingledrug.Theinfluenceofalcohol (ethanol<0.1%)orothersubstances,aswellasothercontributoryfactorscouldclearlyberuledout.GroupB:Certifieddeathbyintoxicationwithmorethanonedrugand/or withdrug/sincombinationwithasignificantconcentrationofethanol.GroupC:Certifiedothercauseofdeath,inwhichthecircumstancesexcludedthepossibilityof incapacitation.GroupD:Individualsdrivingundertheinfluence.GroupT:TherapeuticDrugMonitoringdata.

a_The_10th_and_90th_percentiles_are_shown_if_the_number_of_cases_exceeds_10,_for_lower_N_the_minimum_and_maximum_values_are_shown. b

TDMsamplesduring1999–2005.

c

Lorazepam,propiomazine,zaleplonarenotregisteredinNorway.Buspirone,clomethiazineandhydroxyzinearenotincludedintheTDManalysis.

d

TheconcentrationsofparentsubstanceandmetabolitewereaddedforGroupsA–C.

e

f

g _TDM_samples_during_1999–2001.

h_There_were_no_TDM_analyses_for_midazolam_and_triazolam_during_the_study_period_.

Table4

ThenumberofGroupAandGroupBcasesrelatedtothetotalnumberofpositiveidentiﬁcationforeachdruginthedatabasesoftheNationalBoardofForensicMedicine (femoralbloodsamples),andtoSwedishsalesstatistics.ThesubstancesregroupedassedativesorhypnoticsandlistedaccordingtothenumberofGroupAandGroupBcases relatedtosales.

Substance Totalnumberof

detectionsamong Swedish medico-legal autopsies(N) GroupA cases(N) GroupAand GroupB cases(N) Proportionof GroupAcases amongtotal detections(%) Proportionof GroupAand GroupBcases amongtotal detections(%) Totalsales inSweden (DDD,10 6 ) Numberof GroupAcases relatedtosales Numberof GroupAand GroupBcases relatedtosales Hypnotics Propiomazinea ₁₅₀₇ ₆₈ ₃₆₆ _4.50 _24.30 ₂₄₅ _0.28 _1.49 Flunitrazepamb 2233 176 594 7.88 26.60 414 0.42 1.43 Nitrazepamb 1709 64 322 3.74 18.80 250 0.26 1.29 Clomethiazolea 31 2 6 6.50 19.40 7 0.30 0.89 Zopiclonea 1306 30 229 2.30 17.50 289 0.10 0.79 Zolpidema 525 20 112 3.80 21.30 214 0.09 0.52 Midazolamb ₂₂₇ _– ₁ _– _0.44 ₅ _– _0.22 Triazolamb ₁₂ ₁ ₅ _8.30 _41.70 ₂₄ _0.04 _0.21 Zaleplona 9 – – – – 5 – – Sedatives Hydroxyzinea 223 – 69 – 20.90 34 – 2.02 Clonazepama 177 – 4 – 2.30 5 – 0.81 Alprazolamb 480 4 71 0.83 14.8 14 0.03 0.52 Buspironea ₂₅ _– ₅ _– _20.0 ₁₀ _– _0.49 Oxazepamb ₅₉₆ ₄ ₈₃ _0.67 _13.9 ₁₉₂ _0.02 _0.43 Diazepamb 3873 3 23 0.08 0.59 280 0.01 0.08 Lorazepamb 10 – 2 – 20 33 – 0.06

Abbreviation:DDD,deﬁneddailydoses.

a

Duringtheyears2000–2006.

b

Duringtheyears1992–2006.

(9)

367 previously published case reports and case compilations

368 [3,20,21,28–41]areavailable,theconcentrationsinfatal

intoxica-369 tioncasesgenerallycomparewellwiththerangespresentedinthis 370 study.Mostofthecasereportsandcasecompilations[20,21,29–

371 32,34,36–40] do not provide control cases, which impedes

372 evaluation of their proposed toxic levels. The deviations from 373 theresultsinthisstudyarelikelytobemainlyattributedtothe 374 largersample size and thestandardized procedures; thelatter 375 reducingtheincidenceofextremeresults.Minordifferencesmight 376 berelated tothecharacteristicsofthestudypopulations.Apart 377 fromourpreviouspaper[9],nopublishedpostmortemreference 378 valuesregardingpropiomazinecouldbefound.Regarding diaze-379 pam,thelevelsinmultidrugintoxications(GroupB)showagood 380 correlationwithpreviouslypublishedmaterial[31].Theextremely 381 high number of detections and low number of single drug 382 intoxicationswithdiazepamstronglysupportthegeneral appre-383 ciationofdiazepamasasubstancewithverylowtoxicity.Actually, 384 amongtheCcases,therewere14caseswheretheconcentration 385 exceeded1

m

g/g.Thenitrazepam levelsinsingledrug intoxica-386 tions were substantially lower than previously reported [20], 387 whichmightbeexplainedbyourlargerpopulationsize. 388 Severalofthesubstancesanalyzedinthisstudywerepreviously 389 reviewed using the same strategy in 1997 [9]; alprazolam, 390 clonazepam,diazepam,nordazepam,ﬂunitrazepam,hydroxyzine, 391 nitrazepam,oxazepam,propiomazin,zolpidemandzopiclone.In 392 general, the extended study period for substances in this 393 communication compared to the previous publication has 394 provideda largerstudysample inallgroups (GroupsA–C). For 395 these drugs, evaluated twice by different reviewers and with 396 differentsamplesizes,themedianconcentrationsdidnotdiffer, 397 lending support to the robustness of the evaluation strategy. 398 Havingsaidthat,therangesinallgroupsshowedatendencytobe 399 lowerthaninthepreviousstudy[9];inparticular,the concentra-400 tions in Group B for alprazolam, hydroxyzin and oxazepam 401 (0.16

m

g/gvs.0.30

m

g/g,0.60

m

g/gvs.1.30

m

g/gand1.80

m

g/gvs. 402 3.60

m

g/g,respectively).RegardingoxazepamtheGroupAmedian 403 concentration was lower than in previously published data 404 (3.85

m

g/gvs.5.30

m

g/g).This canmosteasily beexplainedby 405 thelargeincreaseinthenumberofcaseswithinthesegroupsas 406 comparedwithourpreviousstudy(morethan10-foldincreasein 407 GroupB,andadoublingofoxazepamcasesinGroupA). 408 4.4. Fataltoxicityindex

409 Toassessdifferencesintheriskoffatalintoxicationsbetween 410 substances,fataltoxicityindexesarecommonlyused[15,16].The 411 fatal toxicityindex hasbeenshown toberelated to thelethal 412 toxicityinanimals,tophysiochemicalfactorsknowntoberelated 413 to toxicity in humans and measures of cardiac effects (for 414 antidepressants) [15,16]. In this study, flunitrazepam was the 415 mostfrequentdrugin GroupAwhenrelatedtosales.Thefatal 416 toxicityindexshould,however,beinterpretedwithcautiondueto 417 the limited number of intoxications for some substances. 418 Moreover, some of the substances included in this study are 419 commonlyusedfor‘‘recreationalpurposes’’andhencea propor-420 tionofthesubjectsmighthavetakenthesewithoutaprescription. 421 In a Swedish study [42] of DUI subjects, only 24% of the 422 flunitrazepampositive cases,and 26%of thediazepampositive 423 cases, had a prescription within the last 12 months. The 424 correspondingproportionforzolpidemwas79%andforzopiclone 425 70%.Whenrelatingthenumberoffatalintoxicationsinthisstudy 426 toalldetections,flunitrazepamwasstillintop.

427 Inthisstudy,hypnoticswereassociatedwithahighernumber 428 ofintoxicationsinrelationtosalescomparedwithanxiolytics.This 429 hasalsobeenobservedinaBritishstudy[43]basedoncauseof 430 deathstatisticsandintwopreviousSwedishstudies[44,45]based

431 on medico-legal data. The number of fatal intoxications with

432 anxiolyticsinrelationtosaleshaspreviouslybeenshowntobe

433 small in comparison with the number of fatal intoxications

434 attributedtobarbiturates[17].During1961–1974,118deathsdue

435 tobarbituratesand6deathsduetobenzodiazepinespermillion

436 prescriptionswereidentiﬁedinEnglandandWales[43].

437 We have previouslyperformed similaranalyseson

intoxica-438 tions related to antidepressants [11]. The overall number of

439 intoxicationswassomewhatlowerinthis studycomparedwith

440 theantidepressantswhenrelatedtosalesandtothetotalnumber

441 ofdetections.Thetopsubstanceinthatstudywastrimipramine

442 with1.91GroupAintoxicationspermillionDDDsold,compared

443 with0.41GroupAintoxicationsforﬂunitrazepampermillionDDD

444 soldinthepresentstudy.

445 4.5. Strengthsandlimitations

446 The majorstrengths withthis study are thestrict inclusion

447 criteriathatallpostmortemcasesincludedwereassessedbytwo

448 independentreviewers,andthatincaseswithunexpectedhighor

449 lowconcentrations,theoriginalﬁleswerescrutinized.Moreover,

450 onlyfemoralvenousbloodconcentrationswereconsidered, and

451 thesampleswerecollectedandhandledaccordingtoa

standard-452 izedprocedure andanalyzedat asinglelaboratory.Further, for

453 most substances a large number of cases were evaluated and

454 comparisonswithpreviousevaluationsusingthesamestrategy

455 showedthattheresultswererobust.

456 However,forsomesubstancesthenumberofcasesinoneor

457 moregroups waslow,andtheir mediansandrangesshouldbe

458 used with caution. The low occurrence of these drugs in our

459 postmortem data might be multifactorial. Since many more

460 sedativesandhypnoticsaremarketedinotherpartsoftheworld,

461 weencouragemorestudiesusingthesamestrategy.

462 4.6. Conclusions

463 Inconclusion,thisstudyprovidesbloodreference

concentra-464 tionsthatmayassistforensicpathologistsandtoxicologistsinthe

465 interpretationofpostmortemdruglevels.

466 Conﬂictofinterest

467 Theauthorsdeclarenoconﬂictofinterest.

468 Acknowledgements

469 WewouldliketothankAnitaHolmgrenandLudvig

Johannes-470 sen for assistance in database processing, Andreas Tillmar for

471 performing the statistical analyses and Gunilla Thelander for

472 assistancewithdescribingtheanalyticalmethods.Thisstudywas

473 supported by The Swedish Research Council and the Swedish

474 NationalBoardofForensicMedicine.

475 References

476

[1]C.Hiemke,P.Baumann,N.Bergemann,A.Conca,O.Dietmaier,K.Egberts,etal.,

477

AGNPconsensusguidelinesfortherapeuticdrugmonitoringinpsychiatry:update

478

2011,Pharmacopsychiatry44(2011)195–235.

479

[2]A.H.Stead,A.C.Moffat,Quantiﬁcationoftheinteractionbetweenbarbituratesand

480

alcoholandinterpretationoffatalbloodconcentrations,Hum.Toxicol.2(1983)

481

5–14.

482

[3]InternationalAssociationofForensicToxicology(TIAFT).ReferenceBloodLevel

483

ListofTherapeuticandToxic Concentrations,http://www.tiaft.org/(accessed

484

08.06.13).

485

[4]O.H.Drummer,Postmortemtoxicologyofdrugsofabuse,ForensicSci.Int.142

486

(2004)101–113.

487

[5]O.H.Drummer,B.Kennedy,L.Bugeja,J.E.Ibrahim,J.Ozanne-Smith,Interpretation

488

ofpostmortemforensictoxicologyresultsforinjurypreventionresearch,Inj.Prev.

489

19(2013)284–289.

A.K.Jo¨nssonetal./ForensicScienceInternationalxxx(2014)xxx–xxx 7 GModel

FSI74801–8

(10)

490 [6]D.J.Pounder,A.K.Hartley,P.J.Watmough,Postmortemredistributionand

degra-491 dationofdothiepin.Humancasestudiesandananimalmodel,Am.J.Forensic

492 Med.Pathol.15(1994)231–235.

493 [7]T.Hilberg,S.Rogde,J.Mørland,Postmortemdrugredistribution–humancases

494 relatedtoresultsinexperimentalanimals,J.ForensicSci.44(1999)3–9.

495 [8]M.R.Repetto,M.Repetto,Concentrationsinhumanﬂuids:101drugsaffectingthe

496 digestivesystemandmetabolism,J.Toxicol.Clin.Toxicol.37(1999)1–9.

497 [9]H.Druid,P.Holmgren,Acompilationoffatalandcontrolconcentrationsofdrugs

498 inpostmortemfemoralblood,J.ForensicSci.42(1997)79–87.

499 [10]H.Druid,P.Holmgren,J.Ahlner,Flunitrazepam:anevaluationofuse,abuseand

500 toxicity,ForensicSci.Int.122(2001)136–141.

501 [11]M.Reis,T.Aamo,J.Ahlner,H.Druid,Referenceconcentrationsofantidepressants.

502 Acompilationofpostmortemandtherapeuticlevels,J.Anal.Toxicol.31(2007)

503 254–264.

504 [12]H.Druid,P.Holmgren,P.Lo¨wenhielm,Computer-assistedsystemsforforensic

505 pathologyandforensictoxicology,J.ForensicSci.41(1996)830–836.

506 [13]A.W.Jones,Drivingundertheinﬂuenceofchlormethiazole,ForensicSci.Int.153

507 (2005)213–217.

508 [14]M.Josefsson,R.Kronstrand,J.Andersson,M.Roman,Evaluationofelectrospray

509 ionisationliquidchromatography–tandemmassspectrometryforrational

determi-510 nationofanumberofneurolepticsandtheirmajormetabolitesinhumanbodyﬂuids

511 andtissues,J.Chromatogr.BAnalyt.Technol.Biomed.LifeSci.789(2003)151–167.

512 [15]L.A.King,A.C.Moffat,Apossibleindexoffataldrugtoxicityinhumans,Med.Sci.

513 Law.23(1983)193–198.

514 [16]N.A.Buckley,P.R.McManus,Canthefataltoxicityofantidepressantdrugsbe

515 predicted with pharmacologicalandtoxicological data?DrugSaf. 18(1998)

516 369–381.

517 [17]N.A.Buckley,P.R.McManus,Changesinfatalitiesduetooverdoseofanxiolytic

518 andsedativedrugsintheUK(1983–1999),DrugSaf.27(2004)135–141.

519 [18]S.Holm,Asimplesequentiallyrejectivemultipletestprocedure,Scand.J.Stat.6

520 (1979)65.

521 [19]R developmentCore Team.,R:A Languageand Environment for Statistical

522 Computing, Rfoundationforstatisticalcomputing,Vienna,Austria,2012.

523 [20]A.H.Stead,A.C.Moffat,Acollectionoftherapeutic,toxicandfatalblooddrug

524 concentrationsinman,Hum.Toxicol.2(1983)437–464.

525 [21]R.C.Baselt,J.A.Wright,R.H.Cravey,Therapeuticandtoxicconcentrationsofmore

526 than100toxicologicallysigniﬁcantdrugsinblood,plasma,orserum:atabulation,

527 Clin.Chem.21(1975)44–62.

528 [22]A.W.Jones,H.Larsson,Distributionofdiazepamandnordiazepambetween

529 plasmaandwholebloodandtheinﬂuenceofhematocrit,Ther.DrugMonit.26

530 (2004)380–385.

531 [23]M.D. Osselton,M.D. Hammond,A.C.Moffat, Distributionofdrugsandtoxic

532 chemicalsinblood,J.ForensicSci.Soc.20(1980)187–193.

533 [24]Clarke’sAnalysisofDrugsandPoisons,4thedition,in:A.CMoffat,M.D.Osselton,

534 B.Widdop,WattsJ.(Eds.).www.medicinescomplete.com(accessed12.11.13).

535 [25]R.C.Dockens,D.E.Salazar,I.E.Fulmor,M.Wehling,M.E.Arnold,R.Croop,

Phar-536 macokineticsofanewlyidentiﬁedactivemetaboliteofbuspironeafter

adminis-537 trationofbuspironeoveritstherapeuticdoserange,J.Clin.Pharmacol.46(2006)

538 1308–1312.

539

[26]F.M.Gengo,J.Dabronzo,A.Yurchak,S.Love,J.K.Miller,Therelativeantihistaminic

540

andpsychomotoreffectsofhydroxyzineandcetirizine,Clin.Pharmacol.Ther.42

541

(1987)265–272.

542

[27]M.W.Jann,Buspirone:anupdateonauniqueanxiolyticagent,Pharmacotherapy

543

8(1988)100–116.

544

[28]B.C.Wolf,W.A.Lavezzi,L.M.Sullivan,R.A.Middleberg,L.M.Flannagan,

Alprazo-545

lam-relateddeathsinPalmBeachCounty,Am.J.ForensicMed.Pathol.26(2005)

546

24–27.

547

[29]K.Michaud,M.Augsburger,N.Romain,C.Giroud,P.Mangin,Fataloverdoseof

548

tramadolandalprazolam,ForensicSci.Int.105(1999)185–189.

549

[30]A.Steentoft,K.Linnet,Bloodconcentrationsofclonazepamand

7-aminoclona-550

zepaminforensiccasesinDenmarkfortheperiod2002–2007,ForensicSci.Int.

551

184(2009)74–79.

552

[31]B.S.Finkle,K.L.McCloskey,L.S.Goodman,Diazepamanddrug-associateddeaths.

553

AsurveyintheUnitedStatesandCanada,JAMA242(1979)429–434.

554

[32]O.H.Drummer,M.L.Syrjanen,S.M.Cordner,Deathsinvolvingthebenzodiazepine

555

ﬂunitrazepam,Am.J.ForensicMed.Pathol.14(1993)238–243.

556

[33]S.Darke,M.Deady,J.Duﬂou,Toxicologyandcharacteristicsofdeathsinvolving

557

zolpideminNewSouthWales,Australia2001–2010,J.ForensicSci.57(2012)

558

1259–1262.

559

[34]B.P.Joynt,TriazolambloodconcentrationsinforensiccasesinCanada,J.Anal.

560

Toxicol.17(1993)171–177.

561

[35]A.W.Jones,A.Holmgren,Concentrationdistributionsofthedrugsmostfrequently

562

identiﬁedinpost-mortemfemoralbloodrepresentingallcausesofdeath,Med.

563

Sci.Law49(2009)257–273.

564

[36]I.Brødsgaard,A.C.Hansen,A.Vesterby,Twocasesoflethalnitrazepampoisoning,

565

Am.J.ForensicMed.Pathol.16(1995)151–153.

566

[37]A.Tracqui,P.Kintz,P.Mangin,A.A.Lugnier,A.J.Chaumont,Afatalityinvolving

567

secobarbital,nitrazepam,andcodeine,Am.J.ForensicMed.Pathol.10(1989)

568

130–133.

569

[38]V.R.Spiehler,R.I.Fukumoto,Anotherfatalcaseinvolvinghydroxyzine,J.Anal.

570

Toxicol.8(1984)242–243.

571

[39]A.E.Robinson,R.D.McDowall,Toxicologicalinvestigationsofsix

chlormethia-572

zole-relateddeaths,ForensicSci.Int.14(1979)49–55.

573

[40]P.Kintz,B.Godelar,P.Mangin,Gaschromatographicidentiﬁcationand

quantiﬁ-574

cationofhydroxyzine:applicationinafatalself-poisoning,ForensicSci.Int.48

575

(1990)139–143.

576

[41]P.J.Boniface,S.G.Russell,Twocasesoffatalzopicloneoverdose,J.Anal.Toxicol.20

577

(1996)131–133.

578

[42]M.Tja¨derborn,Usingforensictoxicologyregistersincombinationwithdrug

579

dispensingdatabasestoidentifynon-therapeuticuseofprescriptiondrugs

(ab-580

stract),Presented at51 Annual Meetingofthe InternationalAssociation of

581

ForensicToxicology,Funchal,Madeira(2013).

582

[43]M.Serfaty,G.Masterton,FatalpoisoningsattributedtobenzodiazepinesinBritain

583

duringthe1980,Br.J.Psychiatry163(1993)386–393.

584

[44]A.Jo¨nsson,P.Holmgren,J.Ahlner,FatalintoxicationsinaSwedishforensic

585

autopsymaterialduring1992-2002,ForensicSci.Int.143(2004)53–59.

586

[45]B.Jonasson,T.Saldeen,Citalopraminfatalpoisoningcases,ForensicSci.Int.126

587

(2002)1–6.