Table 3

Reports
Author(s)/year Population and Calendar period Methods Key findings pertinent to heart failure Validity and generalizability issues (including Indigenous identification)
1. Prevalence or incidence, either population-based or within clinical groups or clinical service settings
NATSIHS Survey reported in Penm (2008) [33] Whole of Australia Indigenous population (Residents in Very Remote areas not included in non-Indigenous NHS comparator group)
    Design
:
Standardised prevalence ratio of HF among Indigenous Australians 1.7 (males 1.9; females 1.6) • Ascertainment of HF based on self-report; conflated with self-report of oedema.
Cross-sectional survey
    Data source
: questionnaire of persons usually resident in private dwellings
• Comparator non-Indigenous data excluded subjects in Very Remote areas.
    Period
: August 2004 to July 2005
    Outcome
: Self-reported health problems
• Low precision of SPR estimate, especially for males
• Indigenous status according to self-identification in Census
2. Aetiology, risk factors, clinical presentation and pathophysiology
Nil
3. Co-morbidities
Nil
4. Mortality & survival
Field (2003) [34] SA, Qld, WA, NT population
    Design
: Descriptive study
Indigenous HF mortality rates almost threefold higher than non-Indigenous. Disproportionately high HF mortality among Indigenous males aged 55–64 years. • Rates calculated for population aged ≥45 years only
    Data source
: Administrative data (NMD)
• Inter-jurisdictional variation in Indigenous identification data quality
    Period
: 1995–96 to 1997–98 and 1998–99 to 2000-01
    Outcome
: Deaths
• Inherent shortcomings of HF identification on death certificates
Penm (2008) [33] SA, Qld, WA, NT population
    Design
: Descriptive study
Age-adjusted Indigenous HF mortality rates more than double non-Indigenous rates. • Inter-jurisdictional variation in Indigenous identification data quality
    Period
: 2002-05
    Data source
: Administrative data (NMD)
In 45–64 year age-group, mortality rate ratio 6.4.
• Inherent shortcomings of HF identification on death certificates
    Outcome
: Deaths
5. Quality of life
Nil
6. Therapeutic interventions
Nil
7. Health service utilisation (including medication adherence, outpatient attendances, hospitalisations, cardiac rehabilitation)
(a) Primary care attendances
BEACH Survey reported in AIHW (2008) [35] GP practices Australia-wide
    Design
: Cross-sectional survey
Crude proportion of HF encounters lower among Indigenous (1.0/100, CI 0.6-1.3) than non-Indigenous patients (0.7, CI 0.7-0.8) • Data difficult to interpret: not person-based (cannot identify recurrent attendances for the same person), estimates conflate differences in underlying morbidity with differences in service access and utilisation
    Period
: 2002–03 to 2006-07
    Data source
: Written questionnaires (100 consecutive encounters from ~1000 participating GPs nationwide)
Age-standardised proportion of HF encounters higher for Indigenous patients (ratio 2.6)
    Outcome
: Indications for GP encounters
• No formal basis for Indigenous identification; patients not providing Indigenous status conflated with ‘non-Indigenous’
• Imprecise estimates for Indigenous attendances
Beach Survey AIHW (2011) [36] GPs Australia-wide
    Design
: Cross-sectional survey
Crude proportion of HF encounters lower among Indigenous (0.9/100, CI 0.6-1.2) than non-Indigenous patients (0.7, CI 0.7-0.7) • Data difficult to interpret: not person-based (cannot identify recurrent attendances for the same person), estimates conflate differences in underlying morbidity with differences in service access and utilisation
    Period
: April 2004-March 2005 to April 2008-March 2009
    Data source
: Written questionnaires (100 consecutive encounters from ~1000 participating GPs nationwide)
Age-standardised proportion of HF encounters higher for Indigenous patients (ratio 2.6)
    Outcome
: Indications for GP encounters
• No formal basis for Indigenous identification; patients not providing Indigenous status conflated with ‘non-Indigenous
Imprecise estimates for Indigenous attendances
(b) Hospitalisations
Nichol (1999) [37] Patients admitted to Australian public and private hospitals
    Design
: Descriptive study
970 separations with principal diagnosis HF among indigenous; 39,305 Non-Indigenous • Separation rate-ratio not provided
    Data source
: Administrative data (NHMD)
Crude average length of hospital stay for ‘congestive heart failure) shorter for Indigenous than non-Indigenous patients (6.5 vs 9.4 days) • Data not person-based: cannot identify recurrent separations for the same person
    Period
:July 1995-June 1996
    Outcome
: Principal diagnosis reported for hospital separations
• Indigenous identification varies between jurisdictions, Indigenous identity likely under-identified at a single separation
• Caveats of HF-related code as principal diagnosis
Field (2003) [34] Patients admitted to SA and NT hospitals only
    Design
: Descriptive study
July 1998-June 2001 triennium: age-standardised separation rates (HF or hypertensive heart disease) higher among Indigenous than non-Indigenous patients (males: 1555/105 vs 743/105; females: 1579/105 vs 541/105) • Rates calculated for population aged ≥45 years only
    Data source
: Administrative data (NHMD)
• Data not person-based so cannot distinguish repeat recurrent separations for the same person.
    Period
: 1995–96 to 1997–98 and 1998–99 to 2000-01
    Outcome
: Principal diagnosis reported for hospital separations
HF hospitalisation rates fell among both sexes, in both Indigenous and non-Indigenous populations, between 1995–98 and 1998–2001 triennia.
• Not nationwide data: SA/NT only.
• Indigenous identity likely under-identified at a single separation.
AIHW (2008) [35] Patients admitted to private (excluding NT) and public hospitals in NSW, Vic, Qld, WA, SA and NT.
    Design
: Descriptive study
Age-standardised hospital separation ratio (Indigenous:non-Indigenous) for HF 3.4. • Data not person-based: cannot identify recurrent separations for the same person
    Data source
: Administrative data (NHMD)
Average bed days for congestive heart failure 5.7 (Indigenous patients); 7.7 (non-Indigenous)
• Report restricted to jurisdictions with better Indigenous identification, however this varies between included jurisdictions, Indigenous identity likely under-identified at a single separation
    Period
: July 2004 to June 2006
    Outcome
: Diagnoses reported for hospital separations
AIHW (2011) [36] Patients admitted to private (excluding NT) and public hospitals in NSW, Vic, Qld, WA, SA and NT.
    Design
: Descriptive study
Age-standardised hospital separation ratio (Indigenous:non-Indigenous) for HF 3.0. • Data not person-based: cannot identify recurrent separations for the same person
    Data source
: Administrative data (NHMD)
    Period
: July 2006 to June 2008
    Outcome
: Diagnoses reported for hospital separations
Average bed days for congestive heart failure 5.4 (Indigenous patients); 7.5 (non-Indigenous) • Report restricted to jurisdictions with better Indigenous identification, however this varies between included jurisdictions, Indigenous identity likely under-identified at a single separation
Steering Committee (2011) [38] Patients admitted to private (excluding NT) and public hospitals in NSW, Vic, Qld, WA, SA and NT.
    Design
: Descriptive study
Age-standardised hospital separation rates for congestive heart failure 6.1 (Indigenous) vs 2.0 (non-Indigenous) • Data not person-based: cannot identify recurrent separations for the same person
    Data source
: Administrative data (NHMD)
    Period
: 2008-09
    Outcome
: Diagnoses reported for hospital separations
• Indigenous identification varies between jurisdictions, Indigenous identity likely under-identified at a single separation
AIHW (2011) [39] Patients admitted to public and private hospitals in all states and territories.
    Design
: Descriptive study
Crude hospital separation rates for congestive heart failure: • Data from all states and territories.
    Data source
: Administrative data (NHMD)
Indigenous: 2.8/1000 Non-Indigenous: 2.1/1000 • Data not person-based: cannot identify recurrent separations for the same person
    Period
: 2008-2009
    Outcome
: Principal diagnosis reported for hospital separations
(Rate ratio: 1.33)
• Rates adjusted for Indigenous under-identification.
• Crude rates only.
Bureau of Health Information (NSW) (2011) [40] Patients >45 years admitted to public and private hospitals in NSW.
    Design
: Descriptive study
2% of ‘potentially avoidable’ HF admissions of patients occurred among patients identified as Aboriginal, with ‘2% of the NSW population’ considered to be Aboriginal. • Data not person-based: cannot identify recurrent separations for the same person
    Data source:
Administrative data (APDC)
    Period
: July 2009-June 2010
    Outcome
: ‘potentially avoidable’ admissions for specified conditions (including HF)
• Crude proportion only
No adjustment for Indigenous under-identification in hospitalisation data
Bureau of Health Information (NSW) (2012) [41] Patients >45 years with pre-existing record of HF hospitalisation admitted to public and private hospitals in NSW.
    Design
: Cohort study
Patients with pre-identified HF admitted on >1 occasion with HF during year of study were more likely to be Aboriginal (3%) than those with 0–1 HF admissions (2%) • Person-based data
    Data source
: Linked administrative data (APDC and mortality)
• Proportion of cohort identified as Aboriginal not stated
No adjustment for Indigenous under-identification
    Period
: July 2009-June 2010
    Outcome
: admissions and re-admissions
8. Health service delivery issues (including needs, access and barriers)
Nil
9. Costs related to HF diagnosis and care
AIHW (2011) [39] Patients admitted to public and private hospitals in all states and territories.
    Design
: Descriptive study
For congestive heart failure, patients identified as Indigenous accounted for 3.9% of total expenditure for this condition. Expenditure on CHF hospitalisation per person: • Data from all states and territories.
    Data source
: Administrative data (NHMD)
• Indigenous identification varies between jurisdictions, Indigenous identity likely under-identified at a single separation
    Period
: 2008-2009
    Outcome
: Expenditure on potentially preventable hospital separations
Indigenous $26.70
Non-Indigenous $16.90
(Indigenous:non-Indigenous expenditure ratio 1.58)

AIHW: Australian Institute of Health and Welfare.

APDC: Admitted Patient Data Collection (New South Wales).

BEACH: Bettering the Evaluation and Care of Health.

HF: heart failure.

NATSIHS: National Aboriginal and Torres Strait Islander Health Survey.

NHMD: National Hospital Morbidity Database.

NHS: National Health Survey.

NMD: National Mortality Database.

Woods et al.

Woods et al. BMC Cardiovascular Disorders 2012 12:99   doi:10.1186/1471-2261-12-99

Open Data