Drug related mortality in NZ - What has been published so far?

Acute Mortality Related to Prescription and Illicit Drug Overdose in NZ

1998-2001

Research submitted for MSC thesis, 2003Elizabeth Morgan

Supervisor: Dr Nerida SmithSenior Lecturer in Clinical Toxicology Department of Pharmacology & Toxicology University of Otago

Drug related mortality in NZ - What has been published so far?

• Stream of literature lacks continuity – different regions over different time periods

Cairns et al (1983) – Auckland Dukes et al (1992) - Wellington

• A recent publications: • “NZ Drug Statistics” (MOH 2001)• “Surveillance of Chemical Induced Mortality in NZ”

(ESR for MOH, 2003)

Summary of the NZ Data

The information that is available suggests that:

Males are over-represented

Usually young – 20-30 yrs oldData not standardized/adjusted to population

↓ in barbiturate related deaths

TCAs common in late 1970s – early 1980s

CO deaths make up the largest proportion of deaths attributed to a single chemical/drug

The Present Study

Objectives

Examine deaths resulting acutely from prescription/illicit drugs in NZ, 1998 – 2001 using Coronial inquest data

Characteristics of the decedents & circumstances

Examine drugs involved

Identify preventable factors involved

Examine the quality & usefulness of the information available in the Coronial inquest files, for the purposes of population-based studies

Data Collection

Data collected during 2002

Case selection if death occurred between 1998 and 2001 AND if drug involvement was indicated

Deaths attributed solely to non-prescription drugs or substances not restricted by law were not included

Deaths did not have to be solely attributed to prescription/illicit drugs – additional circumstances such as disease or asphyxiation may have been named by Coroner as well

Exclusion Criteria

Death occurred as a result of long-term drug abuse (including disease as a result of drug use – HIV/AIDS)

Death as a result of withdrawal or abstinence syndromes

Drug/chemical implicated was available legally and without a prescription

Verdict – if verdict did not include any mention of drugs then case was excluded

Deaths among drug users where cause was not drug-related

Data Collected from Inquest files

File number

Date of death/date of inquest hearing

Verdict code assigned by Dept for Courts

Basic demographic data: age, gender etc. from Police Report for Coroner

Health status of the deceased

Post-mortem toxicological investigation

Cause of death – pathologist and Coroner

RESULTS

Two parts:

Describing the decedents – demographics

Post-mortem toxicology

Age & Gender

325 decedents 187 (58%) male/138 (42%) female

Aged 2-100 years – avg age 41 years Age-specific mortality data = males died younger

than females

0

5

10

15

20

25

30

35

40

45

50

0-4

5-9

10-1

4

15-1

9

20-2

4

25-2

9

30-3

4

35-3

9

40-4

4

45-4

9

50-5

4

55-5

9

60-6

4

65-6

9

70-7

4

75-7

9

80-8

485

+

age group

freq

uen

cy

Ethnicity

Ethnicity was recorded in 79% of cases Ethnicity data from PRC in 75% of these cases

0.00

1.00

2.00

3.00

4.00

male Maori male non-Maori female Maori female non-Maori

gender/ethnicity

mo

rtality

rate

per

100,0

00 p

op

ula

tio

n

Marital Status

Records available in 60% cases – Police report Of those 60% - almost three-quarters were single

(single, separated, divorced, widowed) Similar result for men & women

Employment Status

58% unemployed (S/B, unemployed, students, retired)

Proportion males > proportion of females On the whole, unemployed people were over-

represented compared to general population

Place of inquest

Auckland94 inquests

Wellington29 inquests

Christchurch36 inquests

Tauranga15 inquests

Dunedin10 inquests

New Plymouth 8 inquests

Palmerston North15 inquests

Place of Death

Own home 59%

Other 16%

Other’s residence 8%

Hospital 17%

Health Status

Very basic – decedents were defined by one of 4 categories

1. No record of mental or physical illness2. Medical history of mental illness (incl depression)3. Medical history of chronic physical illness4. Medical history of mental and chronic physical illness

Amount of information varied from file to file

People with no medical history in inquest file = included in group 1

→ undercounting of illness is likely

Health Status

No reported diagnoses of

mental or chronic physical

illness39%

32% females44% males

Reported clinically diagnosed mental

illness31%

36% females27% males

Reported chronic physical

illness18%

18% females18% males

Both mental and chronic physical illness reported

12%

14% females11% males

VerdictSuicide vs Non-Suicide

35% deaths included in this study were found to be suicide

Raw no. suicides over 4-years remained stableEven though the total no deaths each year dropped

Males outnumbered females in total… proportionally:

1998-2001 male Female

non-suicide 72% (135)54%(75)

suicide28%(52)

46%(63)

Gender, Health Status and Verdict

Health Status in cases found to be suicide:

Mental illness 46%

Physical 13%

Both 19%

Neither 22%

Health Status in cases that were not suicide

Mental illness 22%

Physical 21%

Both 9%

Neither 48%

Mental illness 22%

Physical 21%

Both 9%

Neither 48%

Gender Differences?

Females: proportions of suicide/non-suicide were similar when “health status” categories were examined separately

Males: proportions of suicide/non-suicide equal where history of mental illness was indicated

BUT in contrast to females

Only about 18% of deaths among males with chronic physical illness were suicide

Employment Status vs Suicide/Non-suicide

Proportions of suicide/non-suicide appeared to be similar for unemployed and employed decedents when “unemployment” was viewed as a whole

~ 60-65% non-suicide

Subcategories of “unemployment”:

sickness beneficiary 47% suicide

retired 42% suicide

student 38% suicideunemployed 25% suicide

Post-Mortem Toxicology Examinations

PM toxicology – ESR reports

92% cases in this study subject to tox exam 3% of these cases – report unavailable

No tox exam in remaining 8% casesprevented by decomposition, time delay etc

Of those cases subjected to testing…

3 sample sites48%

2 sample sites31%

4+ sample sites9%

1 sample sites12%

A closer look at the sample sites

• Femoral blood samples were tested in 64% of these cases

• Most of these cases a biological sample from at least 1 other site was tested

• 14% examined cases, blood was from “unknown” site. In most cases this was the only blood sample tested

• Heart blood utilised in 5% of cases – usually blood was tested from other sites too

How Many Drugs Detected PM?

3 drugs21%

2 drugs32%

1 drug21%

No testing/none detected10%

6+ drugs4%

5 drugs5%

4 drugs7%

Drugs detected most frequently

Alcohol 45%Morphine/heroin 16%Diazepam, methadone 14%Zopiclone 10%

Drugs Named by Coroner as Agents Resulting in Death

Taken from Coroners Statements

1 drug57%

2 drugs14%

3 drugs5%

4 drugs1%

5+ drugs1%

Not specified22%

Drugs most frequently involved

In cases where death was attributed to one drug:Morphine/heroin 12%Methadone 9%Dothiepin 6%Doxepin 5%Zopicolne 4%

Where death was attributed to multiple drugs:Alcohol 12%Diazepam 5%Methadone 4%Zopiclone 3%Amitriptyline 3%

Gender vs Drugs

Drugs most frequently detected PM:

Males (specific drugs named by Coroner in 145 cases) methadone>morphine>diazepam>cannabis>zopiclone

Coroner’s statements: opioids dominated deaths among males

Females (specific drugs named by Coroner in 107 cases) dothiepin>morphine>zopiclone>diazepam>amitriptyline

Coroner’s statements: TCAs dominated deaths among females

Antidepressants in general were a more prominent feature of deaths among females

Age vs Drugs

3 age groups:0 – 29 years (about 29% of studied population)

30 – 49 years (about 44% of studied population)

50+ years (about 26% of studied population)

Drugs detected post-mortem were different for each age group

0 -29 years: opioids>chemical>benzo’s & TCA’s30 – 49 years: opioids>chemical>benzo’s50+ years: chemical>TCA’s>benzo’s>opioids

Source of Drugs

Of those deaths which underwent PM toxicological examination:

Records detailing the source of the drugs detected were identified in ~41% of cases

Of the most commonly detected drugs:

OPIOIDS source identified in 33% cases; 62% illicit

morphine – 90% illicitmethadone – 60% illicit

TCAs source identified in 53% cases; 95% prescribed

BENZOssource identified in 55% cases; 75% prescribed

How complete was the data set in this study?

What proportion of ALL drug related deaths occurring in this period did I gather?

Deaths in 1998 = the most complete data set

Looked at how many inquests were processed each year versus year of death, for example of the 107 deaths occurring in 1998:

64% of inquests were processed in 1998 27% inquests were processed in 1999 9% inquests were processed in 2000 and 2001

These results are similar to ESR’s estimates (2003)

Data Completeness…

For the majority of deaths examined in this study, the inquest was completed within 2 years

Data sets for 1998 and 1999 = reasonably complete 2000 and 2001 less so

BUT…different factors may be affecting different sub-sets of the studied population, for example suicides

→ perhaps suicide investigations are completed sooner?Perhaps the number of suicides involving drugs is actually increasing?

Drug Related Mortality in the Present study…OBSERVATIONS…

SHARING MEDICINES & “SELF-MEDICATING”Noted by several Coroners

STOCKPILING OF MEDICATIONSOpioids – cancer treatment patientsBarbiturates - elderly peoplePeople being treated for illnesses known to be associated with

increased suicide risk; often had access to large amounts

INADEQUATE STORAGEMethadone – naive users, not necessarily seeking a highColchicine – teenagers, lack of knowledge about medicines

Limitations of the Present Study

1. Data collection methodsOne person collecting data – no validation

2. Comparability limitationsmany definitions of “drug-related death”

3. Completeness and quality of the data…for example – the “ethnicity” results

Completeness and quality of the data…

Study was retrospective = inherent difficulties

Disparity between the objectives of the inquest and the research objectives

inquest = focused on the individual case-by-case pop based study = requires uniform data

How can this be addressed?

How can this be addressed?

minimum dataset requirementsthese could cover the basic data needs of population based studies: demographics, circumstances of death, aspects of the inquest etc. in a way that does not impose on the Coroner

This would ensure consistency in data source etc. research would be of higher quality mortality data would be more meaningful

Where to from here?