44 Bowen Street
Private Bag 6995
Wellington 6141
New Zealand
T 64 4 894 5400
F 64 4 894 6100
www.nzta.govt.nz
3 July 2026
Nigel Gray
[FYI request #34772 email]
REF: OIA-22302
Dear Nigel
Request made under the Official Information Act 1982
Thank you for your email
of 23 May 2026 requesting the fol owing information under the Official
Information Act 1982 (the Act):
I am requesting the fol owing information regarding the classification of road crashes as
“drug-related”, “alcohol-related”, “medical-event-related”, or any similar causal category used in
official reporting, dashboards, or public statements.
This request arises from the contradiction between:
Agency statements that no single factor can be confidently identified as the primary cause of a
crash, and
The agency’s practice of classifying crashes into specific causal categories such as “drug-related”.
To resolve this contradiction, I request the following information:
1. The decision rules used to classify a crash as “drug-related”
Please provide:
The criteria, thresholds, decision rules, or tests used to determine that a crash is “drug-related”.
Whether toxicology detection alone is sufficient to classify a crash as “drug-related”.
Whether impairment must be established, and if so, how impairment is determined.
Whether causation must be established, and if so, how causation is determined.
2. The process for moving a crash from “pending toxicology” to “drug-related”
Please provide:
The workflow, internal guidance, or standard operating procedures governing this change.
The roles or units responsible for making the classification decision.
Any quality-assurance or review processes applied before the classification is finalised.
3. The evidential standard required to classify a crash as “drug-related”
Please provide:
The evidential threshold required (e.g., presence, impairment, contribution, causation).
Any scientific, legal, or policy basis for this threshold.
Any documents that define or describe the difference between:
presence of a substance
impairment
causation
contribution
4. Reconciliation of the contradiction
Please provide:
How the agency reconciles the statement that “no single factor can be confidently identified as the
primary cause of a crash” with the practice of assigning crashes to specific causal categories such
as “drug-related”.
Any internal documents, emails, memos, or guidance that address this issue.
5. Classification audit or accuracy checks Please provide:
Any audits, reviews, or assessments of the accuracy, consistency, or validity of crash-cause
classifications.
Any documents discussing limitations, uncertainties, or risks associated with assigning causal
labels based on toxicology results.
6. Historical or current definitions
Please provide:
Al definitions of “drug-related crash” used by the agency in the last 10 years.
Any changes to these definitions, including the dates and reasons for the changes.
Prior to responding to your request, I can advise that crash data is provided to the NZ Transport
Agency Waka Kotahi (NZTA) by NZ Police. NZTA staff within the Customer Operations Support Team
analyse the data for each crash and decide on which factors are likely to have caused or contributed
to the outcome of the crash. For each of these factors, a code is applied. You may find a list of codes
and their meanings on the NZTA website at
: https://www.nzta.govt.nz/assets/Safety/docs/cas-factor-
codes.pdf.
2
The al ocation of these codes is often a matter of judgement. The coder must interpret the narrative of
the attending police officer along with any other data provided.
A crash may, and often does have, more than one contributing factor, but a crash does not have a
primary cause attributed to it. When reporting, these contributing factors are sometimes grouped into
road safety categories. Because a crash can involve multiple factors, it can be counted in more than
one category.
1. The decision rules used to classify a crash as “drug-related”
Please provide:
The criteria, thresholds, decision rules, or tests used to determine that a crash is
“drug-related”.
Whether toxicology detection alone is sufficient to classify a crash as “drug-related”.
Whether impairment must be established, and if so, how impairment is determined.
Whether causation must be established, and if so, how causation is determined.
NZTA does not official y use the term “drug-related” when providing coding for crashes. The ‘Drugs
proven’ Code 109 is applied when:
a. a blood test result for drugs is positive, and the coder believes that where there is a positive
test for drugs, the positive results is not solely because pharmaceutical drugs have been
administered by an emergency responder after the crash, but before the administration of a
drug test, or
b. an oral fluid confirmatory test is positive.
For ‘Drugs suspected’ Code 108, this code is applied against a driver in the fol owing circumstances:
a. the driver refused a test for drugs, or
b. the police officer attending the crash had good cause to suspect that the driver has been
taking drugs, or
c. the driver was arrested pursuant to section 120 of the Land Transport Act 1998, or
d. there were two positive oral fluid screening tests, and the result of a blood test is not available.
There is a Code 221 for ‘Impaired non-driver (pedestrian / cyclist / passenger, etc)’. It is used when
drugs were present in non-driving parties such as passengers, cyclists or pedestrians and may have
contributed to the crash.
In rare cases where drugs may have contributed to the cause or outcome of a crash but where none of
the above-mentioned factors are applicable, another factor – Code 220 ‘Other drugs’ may be applied
against a person involved in the crash.
2. The process for moving a crash from “pending toxicology” to “drug-related”
Please provide:
The workflow, internal guidance, or standard operating procedures governing this change.
The roles or units responsible for making the classification decision.
Any quality-assurance or review processes applied before the classification is finalised.
3
We do not use “pending toxicology” or “drug-related” as codes, or as a way of specifying states that
transition from one to the other. Therefore, there is no process to move a crash between these two
states.
If a driver in a crash has had a blood test or an oral confirmatory test for drugs, but the result of the
test is not available when the traffic crash report is sent to NZTA (the screen wil say the test result is
“Blood pending”), NZ Police may send the result of the test later and the data for the crash wil be
updated, including codes applied to the driver, if necessary.
As the information does not exist, this part of your request has been refused under section 18(e) of the
Act.
3. The evidential standard required to classify a crash as “drug-related”
Please provide:
The evidential threshold required (e.g., presence, impairment, contribution, causation).
Any scientific, legal, or policy basis for this threshold.
Any documents that define or describe the difference between:
presence of a substance
impairment
causation
contribution
As noted in the response to question 1 above, “drug-related” is not a term used in NZTA for
contributing factors in a crash.
The fol owing document is an excerpt which contains information about the codes we use regarding
alcohol and drugs as factors in cashes:
• Attachment 11 - Crash Analysis System - Movement and Factor Codes (excerpt).
4. Reconciliation of the contradiction
Please provide:
How the agency reconciles the statement that “no single factor can be confidently
identified as the primary cause of a crash” with the practice of assigning crashes to
specific causal categories such as “drug-related”.
Any internal documents, emails, memos, or guidance that address this issue.
The agency’s position that no single factor can be confidently identified as the primary cause of a
crash reflects the reality that crashes typically involve multiple contributing factors. These can include
driver behaviour, vehicle condition, road environment, and external conditions, which often interact in
complex ways.
For reporting and analysis purposes, crashes are categorised based on identified contributing factors.
This does not mean that a single factor is treated as the sole cause. Instead, it indicates that the factor
was present and contributed to the crash. A single crash may therefore be included in more than one
category where multiple contributing factors are identified.
4
As a result, there is no inconsistency to reconcile. The statement reflects how crashes occur in
practice, while the use of categories is simply a way of analysing and reporting on contributing factors
to support road safety interventions and monitoring of trends in high-risk behaviours. Therefore, your
request for internal documents, emails, memos, or guidance that address this issue is refused under
section 18(e) of the Act as the information doesn’t exist.
There is however reference to “Drug-related deaths and serious injuries (SAFE7) in NZTA’s Statement
of Performance Expectations which we publish on our website, with the 2025/26 report available
online as an example (page 29): www.nzta.govt.nz/assets/resources/statement-of-performance-
expectations/2025-2026/spe-2025-2026.pdf.
We have confirmed internally at NZTA that this measure includes both alcohol and drugs as
contributing factors. The definition of SAFE7 from the NZTA statement of performance expectations
2025/26 explanatory notes states that it counts the number of road deaths and serious injuries where
a driver has tested positive for alcohol and/or drugs with data from the Crash Analysis System -
www.nzta.govt.nz/assets/resources/statement-of-performance-expectations/2025-2026/spe-2025-
2026-performance-measure-explanatory-notes.pdf.
We also report the separate components in page 11 of the quarterly report for the breakdown.
5. Classification audit or accuracy checks Please provide:
Any audits, reviews, or assessments of the accuracy, consistency, or validity of
crash-cause classifications.
Any documents discussing limitations, uncertainties, or risks associated with assigning
causal labels based on toxicology results.
The last review of codes for contributing factors in crashes was undertaken at the end of 2015. Since
that time, there have been no review of the codes. We have tried to locate the 2015 review, including
doing a digital sweep of our internal document storage system but, due to the passage of time, was
unable to find a copy of it. I am therefore refusing this part of your request under section 18(e) of the
Act, as the document al eged to contain the information does not exist or, despite reasonable efforts to
locate it, cannot be found.
However, the following documents were related to the 2015 review and are enclosed:
• Attachment 1 - Rationale for news codes 29032016
• Attachment 2 - Code appendices
Please note that NZTA plans to introduce new codes for zero alcohol, refused blood test, and
impairment due to substances other than drugs.
The coding of fatal crashes is peer reviewed and arbitrated (if the original coder and peer reviewer
differ on what codes they think apply). Also, coding of a sample of crashes by staff members new to
coding is peer reviewed. This is done via the Coding Support Application and there is no audit record
of these reviews.
For fatal crashes, NZTA receives serious crash unit reports and coroner’s reports. Coders read these
reports and revise the coding of these crashes when they think it necessary.
5
Since late 2025, the accuracy of the coding of alcohol above the limit or blood refused in Code 103
has been measured every month.
The remaining documents in the appended table fall within the scope of this part of your request are
enclosed.
Certain information in Attachments 3 and 4 are being withheld under section 9(2)(a) of the Act to
protect the privacy of natural persons.
6. Historical or current definitions
Please provide:
Al definitions of “drug-related crash” used by the agency in the last 10 years.
Any changes to these definitions, including the dates and reasons for the changes.
There are no definitions for “drug-related crash” used in NZTA in the last 10 years. I am refusing this
part of your request under section 18(g) of the Act as the information requested is not held by the
department or venture or Minister of the Crown or organisation and the person dealing with the
request has no grounds for believing that the information is either—
(i)
held by another department (for itself and for a departmental agency hosted by it or an
interdepartmental executive board serviced by it) or interdepartmental venture or Minister
of the Crown or organisation, or by a local authority; or
(ii)
connected more closely with the functions of another department (for itself and for a
departmental agency hosted by it or an interdepartmental executive board serviced by it)
or interdepartmental venture or Minister of the Crown or organisation or of a local
authority.
Under section 28 of the Act, you have the right to ask the Ombudsman to review my decision to
partially refuse this request. The contact details for the Ombudsman can be located at
www.ombudsman.parliament.nz.
If you would like to discuss this reply with NZTA, please contact the Ministerial Services team, by
email to [NZTA request email].
Yours sincerely
Jackie Smith
Manager – Ministerial Services
6
OIA-22302 Document Schedule
Ref
Document
Date
1.
Attachment 1 - Rationale for news codes 29032016
29 March 2016
2.
Attachment 2 - Code appendices
Unknown
3.
Attachment 3 - MINUTES Crash Data Working Group - 22 Sep 2025 22 September
2025
4.
Attachment 4 - MINUTES Crash Data Working Group - 13 April
13 April 2026
2026
5.
Attachment 5 - Crash Data Quality Report Nov 2025
November 2025
6.
Attachment 6 - Crash Data Quality Report Dec 2025
December 2025
7.
Attachment 7 - Crash Data Quality Report Jan 2026
January 2026
8.
Attachment 8 - Crash Data Quality Report Feb 2026
February 2026
9.
Attachment 9 - Crash Data Quality Report Mar 2026
March 2026
10. Attachment 10 - Crash Data Quality Report Apr 2026
April 2026
11. Attachment 11 - Crash Analysis System - Movement and Factor
30 June 2026
Codes (excerpt)
7
