Barriers to effective surveillance

The following barriers to effective surveillance have been identified with regard to the current methods and systems used for measuring and monitoring occupational disease and injury in New Zealand:

Expertise required to manage and administer surveillance systems

Maintaining a database of occupational disease and/or injury data requires specialist knowledge and skills, in areas such as epidemiology, statistics, and health informatics. It also requires skills in relationship management, as external relationships with providers and users of data, and internal relationships with those who enter and those who use data, can have an important bearing on data quality. The range of skills required suggests that a multi-disciplinary team is needed, in most instances, to manage and administer an occupational disease or occupational injury surveillance system. This requires a high level of resource.

The lack of expertise available to manage and administer the systems is compounded by the fact that the various systems are disparate, with little coordination between them. Crucially, nobody has been appointed as the responsible person for doing the coordination at DoL. At Statistics New Zealand, the Injury Information Manager has started to address this issue for occupational injury. However, quite different skills, systems, and methods are required for occupational disease.

Inadequate hazard/exposure assessment

The reviewed databases, on the whole, collect minimal information on exposures to hazards and risk factors, such as environmental, technological, organisational, human, and other factors that contribute to occupational injury and disease. This limits the current potential for conducting robust analysis of incidence, prevalence, distribution, and trends of occupational disease and injury.

Data ownership and access issues

Ethical and privacy issues present potential barriers for access to identified micro-data for linkage purposes. There are potential cost issues if data requests are complex or time-consuming for data managers to meet. Fundamental conflicts in purpose can create a barrier in certain circumstances. For example, from a DoL perspective, it is desirable to have access to ACC data as soon as possible to identify potential serious harm occurrences for investigation and enforcement. However, from an ACC perspective, such an arrangement would be undesirable if it created a disincentive for individuals to claim.

Knowledge gaps

Key knowledge gaps include:

significant gaps in the coverage and accuracy of the existing data sets
the inability to identify new associations between occupational disease/injury and factors such as occupation, industry, and exposures, due to the lack of accurate and appropriate numerator data and a shortage of appropriate denominator data
a lack of knowledge on the part of some GPs, employers, and employees regarding matters such as occupational disease risk factors, the existence of NODS and Workbench (formerly HASARD), and statutory reporting requirements.

Reporting

Collecting, analysing, and disseminating data can involve significant costs, and these can present a barrier to effective surveillance. These problems are exacerbated by a lack of electronic data (requiring time-consuming searches of paper-based records), a lack of coding (requiring extra data preparation and cleaning), and gaps in the available data (requiring expensive one-off studies, such as surveys, to address data requirements).

Presently, no agency is providing regular, comprehensive reporting on occupational disease and injury incidence, prevalence, trends, and distribution, due to limitations inherent in the surveillance systems.

Primary purpose of data collection

Currently, there are no systems that exist specifically to collect occupational disease and occupational injury data. Furthermore, the primary purpose of each agency conflicts to some extent with surveillance objectives. For example:

NODS and Workbench (formerly HASARD) support occupational health and safety investigation and enforcement. Fear of investigation and/or prosecution is a disincentive for employers to report cases to DoL. Efficient use of investigation resources by DoL involves filtering and prioritising of cases for investigation, which impacts on data collection.
The ACC claims database primarily supports claims administration and case management. Full registration of all claims is unnecessary for these purposes and is therefore an unnecessary cost that can be avoided by entering limited information for minor claims. Also, the claims database only records cases that fall within ACC eligibility criteria and for which a claim is submitted, which results in incomplete coverage of some circumstances of injury and disease. For work-related chronic diseases, these problems of incomplete coverage are significant, with only a tiny fraction of cases being identified through ACC claims.
The purpose of the NMDS is to provide comprehensive hospital discharge information on inpatients and day patients. Thus, it offers good capture of acute occupational disease cases but potentially poor capture of chronic occupational disease.

Standardisation, accuracy, and integrity and work-related fields

The following table summarises key data elements for the major occupational disease and injury surveillance systems. Much of the data collected for the surveillance of occupational disease and injury is of a moderate standard, with moderate use for surveillance purposes.

The table indicates what data elements are captured within each system, how they are captured, and, where possible, how well they are captured.

TABLE 2	Standardisation of data elements – occupational disease and injury
DATA ELEMENT		ACC	WORKBENCH (HASARD)	NMDS	NODS	MORTALITY	NZCR
Unique identifiers		ACC45, NHI [H]	Case ID [H]	NHI, ACC45 [H]	Case ID [H]	NHI, BDM registration number [H]	NHI, NMDS event ID, record ID. tumour ID [H]
Other identifying and demographic details		Name, age, sex, address [H]	[M]	Name, age, sex, address [H]	Name, age, address	Name, age, sex, address	Name, age, sex, address
Ethnicity		Non-standard coding system [L–M]	[L–M]	Stats NZ Level 2 [L–M]	Non-standard coding system [M]	Stats NZ Level 2 [M]	Stats NZ Level 2 [M]
Occupation		Stats NZ [M–H for work-related claims]	[L–M]	Free text or occ code [L]	Non-standard coding system [L–M]	Free text [L]	Free text [L]
Occupational history		May give partial longitudinal history	Not recorded	Not recorded	Not recorded	Not recorded	Not recorded
Industry		Stats NZ [M–H]	[L]	Not recorded	Non-standard coding system [L–M]	Not recorded	Not recorded
Work-relatedness indicator		Yes [H – for occ disease]	[H]	E code (since 1 July 2004)	All cases should be work-related	Injury [M]	No
Work-relatedness scope		Employee at work (incl in work car), inter alia	Not recorded	Injury while working for income	Not applicable	Accident while working at place of paid employment	Not applicable
Diagnosis		ICD-10-AM and Read codes [M–H]	Not recorded	ICD-10-AM [H]	Non-standard, non-specific coding system [M]	ICD-10-AM [H]	ICD-10 [H]
Injury type		[H]	[M]	[H]	N/A	Injury [H]	Not applicable
Death		[L]	[M–H]	[H]	Not recorded	ICD-10-AM [H]	Not recorded
Severity		ICD-10-AM, Read codes [L–M]; Time off work [H]	Not recorded	ICD-10-AM [L]	Not recorded	E codes [L for occ disease]	ICD 10 [H]
Site		Non-standard coding system [M–H]	[H]	ICD-10-AM [H]	Non-standard coding system [M]	Not recorded	ICD 10 [H]
Activity		[H]	Not recorded	[H]	N/A	Injury [M]	Not recorded
Mechanism		[L–M]	[M]	Not recorded	N/A	Injury [M]	Not recorded
Agent/exposure		Non-standard coding system [L–M]	[M]	E codes [L for occ disease]	Non-standard coding system [M]	Not recorded	Not recorded
Cost		ACC and claimants costs recorded. Hosp costs imported from NMDS [H]	Not recorded	Relevant (hospital admission) cost weights recorded [H]	Not recorded	Not recorded	Not recorded
Key: H Collected to a high standard of specificity, completeness, and accuracy and in a format suitable for data linkage and surveillance M Collected to a moderate standard or in a format with moderate utility for surveillance (e.g. non-specific coding; partially recorded) L Collected to a low standard or in a format with low utility for surveillance (e.g. not coded; not recorded electronically) NA Not applicable or irrelevant