Ngā tapuae me ngā raraunga: Methods and data sources

Numerators

Data in this chart book were sourced from the Ministry of Health (including the New Zealand Health Survey), Statistics New Zealand, the Institute of Environmental Science and Research Ltd, the Action on Smoking and Health Year 10 Snapshot Survey and the Community Oral Health Service.

Where administrative data (ie, national collections and notifications) were used, the most recent 3 years of data were aggregated to provide stable rate estimates. For survey data, the most recent survey year was used.

Where the New Zealand Health Survey was used, the numerator was the sum of the weights for the respondents in the relevant subgroup (eg, Māori who had diabetes). Further information about survey weights for the New Zealand Health Survey can be found in Methodology Report 2013/14: New Zealand Health Survey.

Data sources and ICD-10-AM codes gives full details of the data sources and the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Australian Modification (ICD-10-AM) codes used for data from national collections.

Denominator

Statistics New Zealand’s mid-year (at 30 June) estimated resident population were used as denominator data in the calculation of population rates for deaths, hospitalisations and cancer registrations. Live births, sourced from Statistics New Zealand, were used as denominators for rates of infant death and low birthweight.

For New Zealand Health Survey data, the denominator was the sum of the weights for the respondents in the relevant population group (eg, Māori) (see Methodology Report 2013/14: New Zealand Health Survey).

Ethnicity classification

Most indicators compare Māori with non-Māori. The method of ethnicity classification is noted under each table or figure in this chart book. In general, prioritised ethnicity classification was used when people identified with more than one ethnic group. A person was classified as Māori if one of their recorded ethnicities as Māori; all other people were recorded as non-Māori, and represent a comparative or reference group. (For example, a person recorded as both Māori and New Zealand European was counted as Māori.) Unknown or missing ethnicity was counted as non-Māori.

Unless otherwise stated, all indicators used ethnicity as recorded on the relevant collection or survey. Recent research has shown that there is no net undercount of Māori deaths and while hospitalisations have been shown to undercount Māori in the past, an assessment of this by the authors of this chart book has shown that when compared with Māori deaths, Māori hospitalisations show no undercount. However, Māori cancer registration rates in this chart book may still be underestimated because the impact of the method used to assign ethnicity in the cancer registration is not well understood (see Ethnicity data in the cancer registration and public hospital event records for more detail).

Age-standardised and crude rates

Age-standardised rates account for differences in population structure, and can be used to compare groups with different age structures, such as Māori and non-Māori. This chart book uses direct age-standardisation; most rates (unless noted otherwise) are standardised to the 2001 Census Māori population. Most of the indicators are expressed as an age‑standardised rate per 100 (ie, a percentage), per 1000 or per 100,000. Age-standardisation is noted under the relevant tables or figures.

Standardising to the 2001 Census Māori population provides rates that more closely approximate the crude Māori rates (ie, the actual rates among the Māori population) than could be provided by other standard populations (eg, the World Health Organization (WHO) World Standard Population), while also allowing comparisons with the non-Māori population. Caution should be taken when comparing data in this chart book with data in reports that use a different population standard.

Where Census data, infectious disease notifications or data for a specific age were presented, crude rates were calculated. Crude rates are the number of events (eg, infant deaths) divided by the population of that age and multiplied by 1000. In this case, caution should be taken when comparing Māori with non-Māori results. Crude rates accurately portray a situation in each population, but make comparisons difficult, because they do not take into account different age distributions in each of the populations (eg, the Māori population is much younger than the non-Māori population).

Rates were not calculated for counts fewer than 5 in data from national collections, or where the population group being analysed (denominator) comprised fewer than 30 in data from surveys.

Confidence intervals

A confidence interval (CI) gives an indication of uncertainty around a single value (such as an age-standardised rate). CIs are calculated with a stated probability; in the case of this chart book, 95% (ie, each CI in this chart book has a 95% probability of enclosing the true value).

The CI is influenced by the sample size of the group. As the sample size becomes smaller, the CI becomes wider, and there is less certainty about the rate.

Rate ratios

Age-standardised rate ratios are used in this chart book to compare age-standardised rates between Māori and non-Māori. The rate ratio (RR) is equal to the age-standardised Māori rate divided by the age-standardised non-Māori rate. Thus the non-Māori population is used as the reference population. For example, an age-standardised RR of 1.5 means that the rate is 50% higher (or 1.5 times as high) in Māori than in non-Māori, after taking into account the different age structures of these two populations.

The main text of this chart book gives rate ratios and their 95% CIs. In this chart book, if the CI of the rate ratio does not include the number 1, the ratio is said to be statistically significant.

Unless otherwise stated, all differences presented in this chart book are statistically significant.

Back to top