Hospitalisations trend by Aboriginality

Recording of Aboriginality in data collections

Estimating the size and composition of the Aboriginal population is difficult for a range of reasons, in particular the incomplete and differential recording of Aboriginality in administrative data collections. The Aboriginality is generally under-reported in administrative data collections for reasons such as staff reluctance to ask about Aboriginality and Aboriginal people’s reluctance to identify as Aboriginal in some circumstances. Recording and reporting is usually better in rural and remote regions than in major cities.

Recording of Aboriginality in the Admitted Patient Data Collection

The quality of recording of Aboriginality in health administrative datasets in NSW varies and may affect the reported hospitalisation rates. The estimated level of enumeration of Aboriginal people in hospital data (Admitted Patient Data Collection) in NSW was 88% in 2007 (Statistical Information Management Committee 2007).

Differences in population composition

The Aboriginal population is relatively young, with a median age of 21 years, compared with 36 years for the non-Aboriginal population. As age is closely related to health, care should be taken when comparing information for these two populations, except where rates have been age-standardised.

Standardisation by age allows comparing rates of disease in Aboriginal and non-Aboriginal populations. Comparisons are often limited, however, by very low numbers of Aboriginal persons in age groups of 55 years and older.

For example, in NSW in 2008 the proportion of Aboriginal young persons aged below 20 years was 47.6% while persons aged 55 years and older were 9.0% (estimated as 14,064 persons) of the total Aboriginal population. The corresponding proportions in non-Aboriginal population in NSW in 2008 were 25.3% and 34.3% (estimated as 1,733,605 persons). In NSW in 2008, there were estimated 1,613 Aboriginal persons and 458,987 non-Aboriginal persons aged 75 years and older.

These differences in age distribution result in a situation where chronic diseases of the old age may be relatively underrepresented in the Aboriginal older population and cases of rarer diseases may even be absent, leading to falsely favourable results in Aboriginal people when compared with non-Aboriginal people. 

To ensure a substantial number of older people in the Aboriginal population for comparison, the older age groups in both Aboriginal and non-Aboriginal populations are often amalgamated from age 55 years.  In this report, amalgamation from age 75 years is used for indicators that compare the Aboriginal and non-Aboriginal population data, while amalgamation from age 85 years is used for indicators that involve comparisons of the total population.  

Amalgamating older ages allows results in Aboriginal and non-Aboriginal populations to be compared, however it is not suitable for analysis of a disease distribution in the total population, because information concerning older ages from 55 to 85 years and over is lost.

References

Statistical Information Management Committee. Hospital under-identification for Aboriginal and Torres strait Islander peoples. Out of session agenda item No 4/2007. Canberra: AIHW, 2007.

Australian Bureau of Statistics. Experimental estimates and projections, Aboriginal and Torres Strait Islander Australians 1991-2021. Cat no 3238.0. Canberra: ABS, 2009. Available at http://www.abs.gov.au/ausstats/abs@.nsf/Products/946D4BC28DB92E1BCA25762A001CBF38?opendocument

1. Hospital statistics in NSW

All NSW public hospitals, public psychiatric hospitals, public multi-purpose services, private hospitals and private day procedure centres in NSW report data on patients admitted for care to the NSW Ministry of Health. Patient separations from developmental disability institutions and private nursing homes are not included. These reported data, from about 400 different facilities in NSW, are called the NSW Admitted Patient Data Collection (APDC).

The collection also includes data relating to NSW residents hospitalised interstate (see below in Imputation for more details) and in Commonwealth Department of Veterans’ Affairs facilities.

Each reporting facility has its own Patient Administration System (PAS), there are five types of these in NSW: HNMA Millenium (administered by Cerner), WinPAS, iPM (I-Soft), HOSPAS (legacy system being phased out and replaced with other systems) and ISCOS, which is the Inpatient Statistics Collection On-line System used by private hospitals.  Data from the PAS systems are loaded into the Health Information Exchange (HIE) in a standard format. The HIE is the data warehouse for the whole NSW health system maintained by the NSW Ministry of Health which is accessed by authorised staff across the health system and in the NSW Ministry of Health. Data from the HIE are extracted and loaded onto the Secure Analytics for Population Health Research and Intelligence (SAPHaRI) warehouse, administered by the Centre for Epidemiology and Evidence and used for this report.

2. Hospitalisation rate definition

A hospitalisation rate is an estimate of the proportion of a population that was hospitalised during a specified period. It is expressed in this report as the number of hospitalisations per 100,000 population per year (person-years). Age and sex standardisation (also called ‘adjustment’) adjusts for differences in the age and sex structure of populations and is performed to make rates comparable between different sub-populations in the same geographic area; in the same population over time; or between different geographic areas at the same time (for example between different States and Territories or different Local Health Districts). This assists in interpreting variations in patterns of hospitalisations between groups and over time after controlling for factors such as age, which are commonly associated with increased disease rates. 

Hospitalisation rates in this report are standardised to the 2001 Australian standard population. 

3. The Admitted Patient Data Collection and this report

The Admitted Patient Data Collection (APDC) is a census of all inpatients treated in NSW and includes data on NSW residents treated in other states.  The APDC was, up to 2002, known as the Inpatient Statistics Collection. The APDC contains approximately 70 variables and is based on a financial year cycle beginning on 1 July and ending on 30 June of the following year.

The APDC data used in this report are the SAPHaRI (Secure Analytics for Population Health Research and Intelligence) data warehouse, maintained by the Centre for Epidemiology and Evidence. The APDC data is extracted from the Health Information Exchange (HIE) and undergoes a quality assurance and standardisation process before being loaded onto SAPHaRI in a SAS dataset format. 

Further information on the APDC and SAPHaRI is available in this report in Methods tab under The Admitted Patient Data Collection title.

4. Use of selected variables of APDC in this report

4.1 Principal diagnosis and additional diagnoses coded to the ICD-9-CM and ICD-10-AM

Each hospital episode in the APDC is described by a principal diagnosis and additional diagnoses, which are coded using the International Statistical Classification of Diseases and Related Health Problems:  ICD-9-CM (up to June 1998) and ICD-10-AM (from July 1998 onwards). Please refer to a separate Methods section on Principal and additional diagnoses for a detailed discussion (this section accompanies selected indicators that include principal and additional diagnoses, for example diabetes hospitalisations).

In this report the count of hospitalisations for a condition is based mainly on principal diagnosis. If an indicator contains analysis of data in additional diagnosis fields as well ('comorbidities'), this is clearly stated in the title of an indicator and in the Notes. The Notes specify how many additional diagnosis fields were included in the analysis. Another exception is the topic of injury and poisoning where the count is based on ‘external cause of injury’. See Methods for injury and poisoning indicators.

The number of additional diagnoses that are included in the hospital records is restricted to conditions which fulfil several requirements, including direct relevance to the treatment and management of the principal diagnosis. This ensures that only the most resource intensive or clinically relevant conditions are listed for that hospitalisation rather than all comorbidities that a patient may have. Data on additional hospital diagnoses cannot therefore be used to estimate the prevalence of a condition in the community.

4.2 Procedures coded to Australian Classification of Health Interventions

Procedures performed in Australian hospitals are coded in medical records using the Australian Classification of Health Interventions (ACHI) published by the National Centre for Classification in Health.  This classification is based on the Commonwealth Medicare Benefit Schedule and relates to anatomy rather than surgical specialty.

Up to 30 June 1998, the ICD-9-CM Procedure Classification was used in Australia. This was based on WHO Surgical procedures (edition 5) and was not revised and carried beyond 1998 by the WHO due to the rapid advancements in the field of procedures  (National Coding Centre 1996).

On 1 July 1998, both the ACHI and the ICD-10-AM were introduced in NSW. They are revised every two years.  The ICD-10 is the WHO classification but the AM suffix stands for Australian Modification, which adds detail necessary to describe practice in Australian hospitals. The ICD-10-AM is fully compatible with ICD-10 (NCCH 2006).

Up to the fourth edition of ICD-10-AM in 2004, the ACHI was published as a part of the ICD manuals (volume 3 and 4), from the fifth edition in 2006 the titles of the ACHI publication emphasise that this is a classification independent from the ICD.

Both ICD-9-CM Procedure Classification (up to and including 1997-98) and Australian Classification of Health Interventions (from 1998-99 onwards) are used in this report.

4.3 Episode of care based count of hospitalisations

The count of hospitalisations in this report is based on an episode of care from 1 July 1998 (ISC EOC datasets on SAPHaRI). A patient can have several episodes of care during one hospital stay, that is, between the formal admission and the formal discharge (separation) from hospital.    

The episode of care is defined by a service category. An episode of care starts when the hospital stay starts or when the service category changes. There are ten service categories: acute care, rehabilitation care, palliative care, maintenance care, newborn care, other care, geriatric evaluation and management, psychogeriatric care, organ procurement-posthumous and hospital boarder. 

A new episode of care starts also when a patient is on leave from hospital: more than 4 days away in any hospital and more than 10 days away in psychiatric hospitals.

4.4 Overnight and day-only hospital stays are included

Both overnight and day-only hospitalisations are included in this report, unlike in the national reporting on hospital statistics (AIHW Health services series no. 40. Cat. no. HSE 107 2011), where day-only hospitalisations are not included.

The only exceptions are the count and rate of hospitalisation for falls in elderly (persons aged 65 years and over) in NSW, where day-only hospitalisations are excluded from the key indicator reports in NSW. NSW performance reporting of hospitalisations for falls in elderly in NSW includes overnight (and longer) hospitalisations. However, in this report, an analysis of day-only hospitalisations and the total of overnight and day-only hospitalisations are also included in the relevant indicator for comparison and completeness.

4.5 Full census of hospitalisations from 1993-94 and sampling factor prior to 1993-94

In 1993-94 the APDC (then known as the Inpatient Statistics Collection) was a fully enumerated survey (that is a census) for the first time. Datasets containing financial years earlier than 1993-94 are based on estimates from a sample of all inpatient data in some hospitals, partiularly in rural areas, and their use in analysis requires the inclusion of a sampling factor weighting variable. 

4.6 Separation date determines the year of hospitalisation

The main record file of the hospital stay for a patient is created or completed on separation from hospital when all relevant documentation is made available to hospital medical record departments. Consequently, the main reason for hospitalisation (principal diagnosis) completed at separation, may be different from the admitting diagnosis. The hospitalisation is counted in the year when the separation took place even if the hospitalisation period occurred predominantly in the previous financial year. For example a patient discharged on 1 July after a 4 week hospital stay would be counted as hospitalised in the new financial year.

4.7 Imputation of interstate hospitalisations in the last year of data

At the time when the NSW Ministry of Health completes the Admitted Patient Data Collection for the previous financial year, data on hospitalisations of NSW residents in other states are usually not available and so interstate hospitalisations for the latest year is always ‘imputed’ from the previous three years in this report. This imputation process estimates the number of hospitalisations occurring in other states for each diagnosis in previous years and adds these numbers to the latest year of data. When the actual data become available, the number and rates for the affected year are re-calculated. 

4.8 Summary of major changes in the hospital data on SAPHaRI

1988-89:           Hospital data included from the Inpatients Statistics Collection (ISC) counted on the basis of a "period of stay" in hospital (ie multiple episodes of care).  ICD-9-CM used and sampling from some hospitals (requiring a sampling factor for weighting numbers in these hospitals). 

1993-94:             Admitted Patient Data Collection (APDC) is a fully enumerated census of hospitalisations.

1998-99 onwards:             APDC starts to be counted as episodes of care. ICD-10-AM replaces ICD-9-CM; and Australian Classification of Health Interventions replaces ICD-9-CM Procedure Classification. 

Imputation for interstate hospitalisations: The latest year of data (number and rates) are ‘imputed’ for missing interstate hospitalisations for NSW residents. The actual number of interstate hospitalisations is included when available. 

5. Differences between figures published by the AIHW and the NSW Ministry of Health and other institutions

5.1 Differences in publication schedules

The APDC dataset is continuously updated because it is not uncommon to receive additional records or additional information on records already supplied well after the close of a financial year. Consequently data on hospitalisations in NSW from different sources, such as this report and the AIHW report, will always differ slightly due to different publication schedules.

5.2 Imputation of interstate hospitalisations in the latest year of data

In this report imputation methods are used to estimate the number of hospitalisations of NSW residents occurring elsewhere in Australia in the latest year due to a lag in receiving this information (see above). Over some periods of time two last years of data need to be imputed. Refer to Notes in the indicator for confirmation how many years are imputed at the time. Estimates may vary slightly from the estimates in other releases of this report or other reports because of the imputation. The actual number of interstate hospitalisations is included when relevant. 

5.3 Definition of hospitalisation

Both overnight and day-only hospitalisations are included in this report, unlike in the national reporting on hospital statistics, where only overnight hospitalisations are included (AIHW Health services series no. 40. Cat. no. HSE 107 2011).

5.4 Different projected populations

For the calculation of rates, the NSW Ministry of Health uses population projections based on the NSW Health Population Projection Series, prepared by the Statewide Services Development Branch of the NSW Ministry of Health in collaboration with the NSW Department of Planning. Refer to Methods in Population-Demography topic for further information on projected populations and other issues mentioned here. The rates in this report are expressed as a number per 100,000 population.

Rates published by the Australian Institute of Health and Welfare may be expressed differently (per 1,000 or 10,000 population) and use different projected population estimates for NSW. The population estimates, which are not projected, are likely to be the same, as these are based on the estimated residential populations published by the ABS.

6. Other datasets holding data on activity in hospitals

The APDC datasets contain information on inpatient hospital activity. Emergency department data and Outpatient activity data are also available to report on hospital activity.

7. References

Australian Institute of Health and Welfare. Australian hospital statistics 2009–10. Health services series no. 40. Cat. no. HSE 107. Canberra: AIHW, 2011. Available at http://www.aihw.gov.au/publication-detail/?id=10737418863 (Cited on April 2011).

National Centre for Classification in Health. The International statistical classification of diseases and related health problems, 10th Revision, Australian Modification (ICD-10-AM). Australian Coding Standards. Sydney: NCCH, 2006.

National Coding Centre. The Australian version of the International Classification of Diseases, 9th revision, clinical modification (ICD-9-CM). Sydney: University of Sydney, 1996.