Diabetes complications in Victoria 2001/02 - Introduction

Page contents: Background | Purpose | Classifications of diabetes complications | Data sources and methods | Download document

Background

Based on findings in the 2000-01 report from the Victorian ambulatory care sensitive conditions study 1, diabetes complications was the leading cause of admission for ambulatory care sensitive conditions. Diabetes complications, along with asthma, vaccine-preventable influenza and pneumococcal pneumonias, have previously been identified as key conditions where interventions could be realistically set up with a demonstrable and significant short-term impact on demand for emergency hospital admission.

Purpose

The purpose of this report is to provide a detailed analysis of specific diabetes complications, with regional and community profiles for long and short-term complications to support community planning and evaluation.

Classifications of diabetes complications

For the purpose of this analysis, diabetes complications were classified into two main groups:

• short-term complications, comprising:
• ketoacidosis (without mention of coma)
• diabetic coma (including coma with ketoacidosis)
  
• long-term complications (including all other diabetes complications) comprising four main groups:
• renal complications
• ophthalmic complications
• neurological complications
• peripheral circulatory complications
• complications that did not fit these categories formed a fifth group, 'other complications'
• a sixth category, 'unspecified complications', was also included.

Detailed listings of codes and conditions included within these categories are provided in the appendices.

top of page

Data sources and methods

Hospital admissions data

Hospital separation data were obtained from the Victorian Admitted Episodes Dataset (VAED). The VAED is a minimum dataset containing data on all admitted patient activity submitted by all public and private acute hospitals, including acute facilities in rehabilitation and extended care institutions and day procedure centres 7.

Clinical data were stored as twelve ICD-9-CM diagnosis and procedure codes until 2000-01, and then as twenty-five ICD-10-AM diagnosis and procedure codes thereafter 8-9. The VAED records were selected on the basis of diagnosis codes. The appendices contain diabetes complications identified using ICD-9-CM and ICD-10-AM codes in the diagnosis fields of the VAED.

New standards for coding of diabetes and its complications were effective from July 2000 9-10. The higher number of admissions for diabetes complications in 2000-01 and 2001-02 compared with earlier years reflects this change.

Population data

Population data by gender and five-year age groups were obtained using the estimated resident population figures produced by the Australian Bureau of Statistics. The population data were used for calculating admission rates and 95 per cent confidence intervals (CI in the tables).

Population data for 1996 were used as the denominator for calculating admission rates relating to the years up to and including 1996-97. Thereafter, population data relating to 1997 through 2001 formed the basis of admission rate calculations.

Boundaries of catchment areas

Analyses of individual diabetes complication admissions to hospital require the calculation of admission rates for defined geographic areas. In Victoria, the boundaries that make up local government areas under the Australian Standard Geographic Classification have changed significantly over the past decade. Currently, there are 200 statistical local areas, which make up 78 local government areas. These boundaries have been collapsed into 32 PCP catchment areas. The boundaries were introduced in 1996, but were not available for hospital admissions data until 1997-98. For this reason, comparisons across the PCPs are made between 1997-98 and 2001-02, while comparisons across the entire nine years of the study are made at the departmental region level. Victoria is divided into nine health regions, of which four encompass metropolitan Melbourne and five comprise the non-metropolitan areas in Victoria.

Calculation of standardised admission rates

Estimates at the local government area level were used to calculate admission rates and 95 per cent confidence intervals for the 32 PCP catchments in Victoria. Admission rates were age and sex-standardised (direct method) using the 1996 Victorian population as the reference. The 95 per cent confidence intervals for the standardised rates were based on the Poisson distribution.

Trend analysis

Data from 1993-94 to 30 June 2002 were used in this analysis. Before 1993, not all hospitals contributed to the database. In addition, casemix funding was introduced for hospitals in that year.

Rural/metropolitan differences

Standardised admission rates and rate ratios are used throughout this report, providing important information on the relative impact of diabetes complications in rural and metropolitan settings, based on the departmental classification of rural and metropolitan regions. It is valuable, however, to review the crude numbers of admissions in PCP catchments to identify opportunities for high-yield targeted interventions.

Download document

Diabetes complications in Victoria 2001/02 - Introduction (23kb, pdf)

Download entire Diabetes complications in Victoria 2001/02 report

top of page