Injury Statistics – Data Quality General Information

General information Data sources Data on claims for work-related injuries is from the Accident Compensation Corporation (ACC). All claims are included under the calendar year when the injury occurred. The most recent year’s figures are provisional at the time of release and may be updated in later releases. Claims for fatal work-related injuries can take up to two years to be finalised. Figures for the number of full-time equivalent (FTEs) are sourced from the Household Labour Force Survey (HLFS) and are used to calculate incidence rates. We also refer to figures from the 2023 Census and the HLFS within the release. Accuracy of the data Accredited Employers Programme (AEP) claims To increase the accuracy of work-related injury statistics, we include AEP claims in this release. Before 2015 we excluded them because of data quality issues. However, this resulted in undercounting, as AEP claims are a substantial proportion of work-related claims. There are known data quality issues with these types of claims and we are working with ACC to review this to improve future submissions from accredited employers. Age Age specifies an injured person's age at the date of injury. We calculate age using date of birth. This age may differ from the worker's age when the claim is lodged, the age when compensation is received, or the age at death (if the worker dies of the injury). We group data on work-related injuries for people aged 14 years and under into 'other and unspecified'. This is due to uncertainty about the ages recorded for people in this group. In previous years, claims by people aged 14 and under were less than 1 percent of total claims. Body site of injury reporting The body site of the injury for claims submitted through AEP is categorised as ‘Unobtainable’ for 2012–15 due to data quality issues. Claims involving treatment at hospital emergency departments Treatment lasting less than three hours, and provided at a hospital emergency department (ED), is bulk funded by ACC directly to the district health boards and is not recorded as individual claims. As a result, claims involving treatment at the ED for less than three hours are not included in this release. Ethnic group reporting Ethnic group reporting uses ‘total response’ for both the number of claims and the number of FTEs. Total response counts the injured person once in every ethnic group they identify with. For example, a person who identifies as being of Samoan, Tongan, and Māori ethnicities would be counted once in the ‘Pacific peoples’ category and once in the ‘Māori’ category. Counting individuals in more than one ethnic group means the sum of the ethnic groups is greater than the number of people. However, it also means that all ethnicities are counted and identified in a specific unambiguous ethnic group and that the relative size of the groups within the population is fairly represented. In this release, up to three separate ethnic groups can be reported for each claim. Any further ethnic groups reported are put in the 'other' ethnicity category. As multiple ethnicities are coded in the same manner by ACC and the HLFS, the numerator and denominator are consistent in the calculation of incidence rates. The ethnicity of people who submitted a claim through AEP is categorised as ‘Not specified’ due to data quality issues. Fatal claims Fatal accidents are referred to Coronial Services to determine the cause of death. The cause of death is official only when the coroner publishes their finding, which can take up to two years. The assessment of a fatal claim requires the coroner to determine the cause of death, because in some cases an underlying medical condition such as a heart attack may be the cause of the accident – therefore the claim is not covered under ACC legislation. The coroner can take up to two years to review a specific case before ACC can make a final decision. A fatal claim in 2019 may therefore only be finalised in 2021; it will only be included in the 2019 statistics if the claim was finalised in time for reporting to Stats NZ. Full-time equivalent employees (FTEs) Because we derive FTE numbers from a sample survey (the HLFS), they are subject to both sampling and non-sampling error, and should therefore be seen as indicative rather than definitive. 2023 Census of Population and Dwellings and 2023 post-enumeration survey and incidence rates On 30 June 2023, population estimates based on the 2023 Census of Population and Dwellings and associated post-enumeration information, were released. National population estimates were revised for the years ended 2018-23. All age-standardised rates for the whole population, Māori, and children during this period may have changed as a result of this revision. For more information, see Estimated resident population 2023: Data sources and methods. HLFS population rebase and incidence rates In 2016, the HLFS was redeveloped– the biggest substantial change to the survey since it started in 1985. The redevelopment improved the derivation of usual hours worked, which has contributed to an overall increase in FTE estimates. In 2016, we backdated incidence rates for 2002 to 2015 to capture improvements made to FTE estimates from the HLFS redevelopment. In 2017, the 2016 FTE number was affected by the redeveloped questionnaire being in the field for the first time (for the June 2016 quarter). In 2018, the 2017 FTE number was affected by the redeveloped questionnaire being in the field for the full year. The biggest effect this has on the 2017 release, is that the new questionnaire is now more likely to identify people as employed, particularly if they’re self-employed. We use FTE estimates to calculate the incidence rates in this release; we expect some rates will be lower because of the change. This needs to be considered when comparing rates between 2016, 2017 and any previous years’ rates, as a decrease in the rate may be due to the HLFS questionnaire changes rather than a real-life effect. We therefore recommend customers use caution when comparing 2016 and 2017 rates with previous years’ rates. 2016 HLFS redevelopment has more information. On 2 July 2025, the HLFS underwent a population rebase based on the 2023 Census of Population and Dwellings and 2023 post-enumeration survey. Estimates for the years ended 2018-2025 have been revised. The new FTE estimates from the HLFS resulted in revisions to incidence rates back to 2018 for the release in December 2025. Rates of work-related injuries during this period may have changed as a result of this rebase. See Household Labour Force Survey population rebase from 2023 estimated resident population for more information. Industry incidence rates Due to variation within the data, we do not currently recommend using HLFS industry estimates below Level 1 ANZSIC. This is why we do not currently produce rates for agriculture, forestry, and fishing as separate industries. Estimates of employment by industry are not benchmarked against any other real-world population estimates. Customers should take this into consideration when using industry incidence rates from this release. Type of injury reporting The type of injury for claims submitted through AEP has been categorised as ‘Other and undefined’ for 2012–14 due to data quality issues. Percentages All percentages are calculated from unrounded data with the exception of percentages for fatal claims, which are calculated from rounded data. Rounding To protect confidentiality, all fatal claim counts in the tables are randomly rounded using the standard Stats NZ random rounding to base 3 method. In this method, we randomly round all up or down to one of the adjoining multiples of three (eg, a count of five would be displayed as either 3 or 6, and a count of one would be displayed as either 0 or 3). Non-fatal claim counts are rounded to the nearest hundred. Rounding may result in totals disagreeing slightly with the total of individual items shown in the tables. Due to differences in random rounding, fatal claims which have been randomly rounded may have different values those presented in NZ.Stat tables. Final figures In this release, claims are included in the tables under the calendar year in which the injury occurred. For final figures, claims are only included if they were accepted by 20 June 2025. Claims for fatal work-related injuries can take up to two years to be finalised. Some ACC claims for a given year may not be represented in the statistics for that year because of the date of reporting. Claims could be reported at a later date due to late submission, or they may be more complex and require specialist assessments that take time to determine whether the accident is covered by ACC. Therefore, the claim is only included in the data if it was finalised in time for reporting to Stats NZ. Otherwise, such claims are included in the statistics at a later date and may cause a slight change in estimates for the given year. Consistency with other periods or datasets Error in previously released data ACC Accredited Employers Programme claim information was added into the injury data – work-related claims releases from 2015. In 2018, an error was found where claims made under the Accredited Employers Programme (AEP) incorrectly included non-accepted claims. The inclusion of these non-accepted claims caused the total number of claims made per year to be over counted by approximately 2.6% on average each year. This impacted the previous releases: Injury statistics – work-related claims: 2016, published 31 August 2017, Injury statistics – work-related claims: 2015, published 28 November 2016, and Injury statistics – work-related claims: 2014, published 15 October 2015. The data has been revised and corrected for the injury data – work related claims: 2017 release. Consequently, the trend data reported in these tables have been updated and are now different to data previously published. As a result, we do not recommend comparing the results from Injury statistics – work-related claims: 2017 release with any previous releases. Comparability with previous work-related claims releases Due to including AEP claims, introduced in 2015, and the HLFS rebase and HLFS restructure, we do not recommend comparing the results from this release with any work-related claims releases published before October 2015. Comparisons between this release and the 2016 release are meaningful. However, please consider that injury type and body site of injury data is now available in 2002–11 for accredited employers and may affect comparisons of these estimates. Caution also needs to be used when comparing 2016 rates backwards due to the HLFS development. See HLFS redevelopment effect on incidence rates above for more information. Comparability with the serious injury outcome indicators Data from ACC on work-related injury claims is also used for injury surveillance and monitoring through the serious injury outcome indicators (SIOI), which provide annual counts and rates for outcomes of serious injury in New Zealand. For work-related injury, the SIOI present measures for fatal injuries, and serious non-fatal injuries (these are injuries which meet a certain threshold of severity). They also provide a serious injury indicator, which combines the fatal and serious non-fatal data. While the data used in the indicators has some similarities with the data used in this information release, there are some key differences. The SIOI contain high-level figures, while this release contains a more detailed breakdown of the data. The work-related fatalities published in the SIOI release include work-related fatalities notified to WorkSafe New Zealand or those for which an ACC claim was made. The data is presented differently. This release separates the data into all claims, entitlement claims, and fatal claims. The SIOI release presents counts and rates based on some injury outcomes (ie fatal or not fatal) for serious and fatal claims only. The SIOI present data for fatal injuries as three-year moving averages, due to the small numbers. This release publishes fatal data for single years, but uses random rounding on all numbers. See Serious injury outcome indicator reports for more information. Interpreting the data Counting claims vs counting injuries The data in this release is for claims for work-related injuries, and is not a definitive count of all work-related injuries. This is because not all work-related injuries result in a claim to ACC. As for fatal injuries, the statistics in this release are not a definitive count of work-related fatalities. Firstly, not all fatal work-related injuries are the subject of claims to ACC. Secondly, in this release fatal work-related injuries are counted in the year that the injury took place and are not included in the count of work-related fatalities for the year of death of the injured person. More information Statistics in this release are produced in accordance with the Official Statistics System principles and protocols for producers of Tier 1 statistics for quality. They conform to the Stats NZ Methodological Standard for Reporting of Data Quality. Period-specific information This section has information about data that has not changed between releases. Reference period This release contains provisional statistics for work-related claims for injuries in the 2024 calendar year. It also includes final statistics for injuries in the 2023 calendar year. Both of these are as reported by 20 June 2025. Changes since the last release Error in previous release’s data In 2018, an error was found in which previously published data incorrectly included non-accepted claims, for claims made under the Accredited Employers Programme. The inclusion of these non-accepted claims caused the total number of claims made per year to be over counted by approximately 2.6% on average each year. This data has been revised and corrected for the injury data – work related claims: 2017 release. Consequently, the trend data reported in these tables have been updated and are now different to data previously published. Household Labour Force Survey (HLFS) redevelopment In 2016, the HLFS was redeveloped – the biggest substantial change to the survey since it started in 1985. The redevelopment improved the derivation of usual hours worked, which has contributed to an overall increase in full-time equivalent (FTE) estimates. We use FTE estimates to calculate the incidence rates in this release; we expect some rates will be lower because of the change. The redeveloped HLFS survey was in place for three-quarters of 2016 and the whole of 2017. This needs to be considered when comparing 2017 rates with previous years’ rates, as a decrease in the rate may be due to the HLFS questionnaire changes rather than a real-life effect. See HLFS redevelopment effect on incidence rates under General information. Improved data quality for type of injury and body site of injury In 2018, there is improved data quality for Accredited Employers Programme (AEP) claims in two measures: type of injury and body site of the injury. Previously, all AEP data was recorded as ‘Other and undefined’ for type of injury and ‘Unobtainable’ for body site of the injury, due to data quality issues. Now AEP data is included in the breakdowns for type of injury and body site of injury for 2002–11 and 2015-2017. For 2012–14, this data is still not available and claims from AEP are recorded as ‘Other and undefined’ and ‘Unobtainable’, respectively. AEP claims use an additional category in body site of injury, stated as ‘Back except Head Vertebrae’. These claims are assigned to a category for body site of injury using the Read Code variable as reported by the treatment provider. Change in age category Due to growing interest in an ageing population the age group categories have been expanded this year. The category of ‘65 years and older’ has been split into two categories: ‘65-74 years old’ and ‘75 years and older’. Experimental industry tables Previous releases have only reported industry data in trend tables from 2009 onwards. This is because in 2008 ACC changed its industry classification codes from Australian and New Zealand Standard Industrial Classification (ANZSIC) 1996 to ANZSIC 2006. As a result, industry data was not comparable between these time periods. In 2018, an experimental series for work-related claims by industry for the Injury statistics – work-related claims: 2017 – trends tables for 2002–17 has been introduced. Industry Classification Unit (CU) is used by ACC for insurance purposes. It is a detailed description of business activities grouped by risk to determine the ACC levy rates employers pay and share the cost of injuries fairly. A concordance was used to approximate ANZSIC06 industry codes from industry classification units for years 2002-2008. This concordance was developed by ACC in 2017 to map classification units, levy risk groups, and level 1 industry codes by year to their current (2017 calendar year) equivalents for 2002–present. Using this approach industry could be mapped to ANZSIC 2006 V1. This was deemed to be the most reliable approach, as it was the most consistent with current industry codes: one industry varied by 9%, five by +/- 1%, and figures were the same for thirteen of the nineteen industries. While this table provides a more complete time-series of work-related claims by industry, Stats NZ advise not to compare this experimental table to the original industry data presented. en-NZ