SEARCH

SEARCH BY CITATION

Keywords:

  • heart failure;
  • australia;
  • hospital separations;
  • case fatality;
  • trends;
  • length of stay

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

Aim:

To assess trends in admissions of patients with heart failure (HF) to all hospitals in Australia between 1996–1997 and 2003–2004.

Methods and results:

We carried out a retrospective analysis of the official population-based National Hospital Morbidity Data in Australia. Although the absolute number of separations with a principal diagnosis of HF remained stable, the age- and sex-standardized separation rate for HF recorded as principal diagnosis decreased from 2.0 per 1000 population in 1996–1997 to 1.6 per 1000 population in 2003–2004. The corresponding values for HF recorded in any diagnostic position were 7.7 and 4.7 per 1000 population. Men had higher in-hospital mortality than women (8.9% versus 8.1%, p<0.001) and also a larger decrease in this measure over the study period (21.9% versus 14.4%). While the geometric mean length of stay for HF as principal diagnosis fell from 5.4days in 1996–1997 to 4.9days in 2003–2004, the proportion of bed-days related to such diagnoses relative to total bed days attributed to circulatory diseases increased from 12.8% to 13.7% (p<0.001).

Conclusion:

There were no increase in number of admissions involving HF and standardized rates of hospital separations with HF fell in Australia between 1996 and 2004. The explanation for the observed declines in in-hospital case fatality and the separation rates should be sought in whole-of-community studies.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

Heart failure (HF) is a chronic condition and makes a major contribution to the disease burden in most developed and developing countries. It affects 1–2% of the population, with a steep increase in the elderly [1]. Ageing of the population, improvement in survival of HF [2,3] and ischaemic heart disease (IHD) [4], and increases in the prevalence of diabetes mellitus (DM), obesity and overweight and sedentary life style are all factors that could contribute to an increase in the number of patients with HF. However, while this expansion of caseload has been described as an ‘epidemic’, there is actually little information on trends in the incidence rate of heart failure.

In addition to the high risk of death, HF is a debilitating condition associated with multiple admissions to hospital. Estimates of the expenditure on HF range between 1% and 2% of the total budget for health care [5] and the cost of inpatient treatment represents more than two thirds of the total cost attributed to HF [6]. Therefore, trends in hospital admissions for HF are of considerable interest as a measure of caseload and morbidity related to HF as well as the cost associated with this condition. In-hospital case fatality, total bed days, length of stay (LOS), numbers of diagnostic procedures and their changes over time are other important indices in the epidemiology of HF.

Although reports of increases in numbers and rates of admissions for HF as principal diagnosis or secondary diagnoses from different countries such as USA [7,8], Scotland [9], Spain [10], the Netherlands [11], Canada [12] and Singapore [13] support the existence of an ‘epidemic’ of HF, these studies investigated trends in the period before the late 1990s when ACE inhibitors became widely used. A greater range and availability of diagnostic techniques and better care and treatment for patients suffering from HF are factors that, together with changes in relevant risk factors, may alter the number and rate of admissions for HF.

In common with many other countries, information on the epidemiology of HF is scant in Australia. We recently reported a decline in HF mortality from 1997 to 2003 and showed that the reported epidemic of HF was not reflected in mortality data in Australia [14]. The present paper provides a similar analysis of National Hospital Morbidity Data (NHMD) for the corresponding years, to extend our understanding of the current epidemiology of heart failure in Australia.

2. Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

2.1. Data sources

The Australian Institute of Health and Welfare (AIHW) collates data on all hospital separations in Australia. A separation refers to the episode of care for an admitted patient, which can be the total hospital stay or part thereof when there is a change in type of care [15]. We obtained computerized records of all separations related to HF for the financial years 1996–1997 to 2003–2004 from AIHW (1st July to 30th June for each year). Since 1996–1997, each electronic record has included up to 30–50 codes reflecting various diagnoses, complications and procedures performed while the patient was in hospital. The diagnoses were classified according to ICD-9-CM (International Classification of Disease, 9th revision, Clinical Modification) [16] and ICD-10-AM (International Classification of Disease, 10th revision, Australian Modification) [17] for the years 1996–1997 to 1997–1998 and 1998–1999 to 2003–2004, respectively. We calculated rates using estimates of the population for each year of the study, disaggregated by sex and age, from the Australian Bureau of Statistics (ABS).

2.2. Hospital separations related to HF and circulatory disease

For the purpose of this study a hospital separation was defined as involving HF if at least one of the recorded diagnostic codes was a relevant rubric in ICD-9-CM (428) or ICD-10-AM (I50). In addition, hospital separation was defined as related to circulatory disease if the principal diagnosis was one of the rubrics in chapter 7 of ICD-9-CM or in chapter 9 of ICD-10-AM. We excluded 19,417 (1.6%) separations as outliers because their LOS was longer than 50 days.

2.3. Statistical analysis

The data were analyzed using Stata version 8 [18]. We calculated annual age- and sex-specific separation rates for HF either as principal diagnosis or mentioned in any diagnostic position by dividing the number of separations attributed to HF by the relevant population during that year. The population for each reporting year was calculated by average of mid-year populations of included years. Separation rates were standardized using the ‘European’ population [19], treating people aged less than 60 and people aged 85 years or older as single groups and employing 5-year age-groups between ages 60 and 85. In addition, 95% confidence intervals (95% CI) were calculated for the age-standardized separation rates, age specific rates and proportion of bed-days involving HF. We used the method of Fay and Feuer [20] for calculating 95% CI for age-standardized separation rates. To calculate confidence intervals for the age-specific rates and the proportion of bed-days involving HF, we used the large-sample normal approximation. We also calculated the age- and sex-specific in-hospital case fatality for HF as principal diagnosis. We tested the difference in case-fatalities between men and women using the large-sample normal approximation at the 5% significance level. To deal with the very positively skewed distribution of LOS, we calculated the geometric mean. We also estimated and tested the temporal trends for in-hospital case fatality, hospital separation rates, proportion of bed-days involving HF and mean of LOS (geometric) using simple linear regression.

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

From a total of 48,562,285 separations from Australian hospitals, there were 1,161,526 (2.4%) separations involving HF either as principal or an additional diagnosis in the official record (Table 1). In 329,714 (0.7%) separations, HF was regarded as the principal diagnosis. Thus, separations involving HF were more than three times as likely to have conditions other than HF as the principal diagnosis. The median age for separations with a principal diagnosis of HF increased from 77 years in 1996–1997 to 79 in 2003–2004.

Table 1. Numbers and rates of hospital separations involving heart failure in Australia, 1996–1997 to 2003–2004
YearHeart failure as principal diagnosisHeart failure mentioned anywhere
 MaleFemaleMaleFemale
 NumberRateaCINumberRateaCINumberRateaCINumberRateaCI
  • a

    CI=95% confidence interval.

  • a

    Hospital separation rates per 1000 person-years standardized using the ‘European’ population as the external reference.

1996–199720,2192.392.36–2.4220,4771.631.61–1.6575,8549.008.93–9.0683,4826.636.59–6.68
1997–199820,9482.372.33–2.4020,5441.581.56–1.6180,6069.169.09–9.2288,2026.746.69–6.78
1998–199920,9542.302.27–2.3320,6831.541.52–1.5681,5958.988.92–9.0488,9646.596.54–6.64
1999–200020,6182.192.16–2.2220,8381.501.48–1.5278,7988.408.34–8.4685,6906.146.09–6.18
2000–200120,1022.072.04–2.0920,7311.431.42–1.4664,1906.606.55–6.6568,0674.734.70–4.77
2001–200220,5262.032.01–2.0621,0911.421.40–1.4460,0525.95590–6.0060,0524.292.25–2.32
2002–200320,6622.111.98–1.9520,1641.311.29–1.3359,6195.715.66–5.7661,3014.003.96–4.03
2003–200420,9711.951.92–1.9720,1881.281.26–1.3060,1205.585.53–5.6261,1313.893.86–3.93

3.1. Rates of separations associated with HF

The overall age-standardized separation rates for HF as principal diagnosis or mentioned anywhere were 1.8 and 6.2 per 1000 person-years, respectively. These rates rose sharply with age and in each age group the rate in men was higher than that in women, although the male excess decreased with increasing age (Table 2). Although the absolute numbers of separations with HF as the principal diagnosis remained stable, the rates fell by 20.0% (p for trend<0.001). The corresponding decline for HF mentioned anywhere was 39.0% (p for trend=0.001) (Table 1).

Table 2. Age-specifica and age-standardized hospital separation ratesb, Australia, 1996–1997 to 2003–2004
Age groups (years)HF as principal diagnosisHF mentioned anywhere
 MaleFemaleMale/Female ratioMaleFemaleMale/Female ratio
  • a

    95% confidence intervals for age-specific rates have a maximum width of less than 0.04 events per 1000 person-years.

  • b

    Other than the sex ratios, all figures in the table are rates per 1000 person-years.

  • c

    Standardized using the age strata indicated and the ‘European’ population as the external reference.

<600.30.11.80.80.51.6
60–643.41.81.911.06.61.7
65–696.53.61.820.913.21.6
70–7411.27.01.637.925.91.5
75–7919.112.91.565.547.21.4
80–8431.023.21.3107.684.51.3
≥8546.739.31.2170.3143.91.2
Age-standardizedc2.11.51.57.35.31.4

3.2. Length of stay for non-fatal separations

Over the period of study, LOS varied with year of admission, age, sex and diagnostic position of HF. The geometric mean LOS for women was higher than for men and was higher than for separations with any mention of HF than for separations with principal diagnosis of HF (Fig. 1). In 2003–2004, the geometric mean LOS for non-fatal admissions with HF as principal diagnosis increased with age, from 4.2 days in patients aged less than 60 to 6.9 in patients aged 85 and over. While the geometric means LOS for all ages for HF mentioned anywhere fluctuated over the period of study (p for trend=0.06), for separations with mention of HF as principal diagnosis it declined from 5.4 to 4.9 days (p for trend<0.001).

image

Figure Fig. 1. Trends in the geometric mean length of stay for non-fatal hospital episodes.

Download figure to PowerPoint

3.3. Bed days involving HF

Over the period of study, total days of hospitalisation attributed to HF as principal diagnosis and mentioned anywhere were 2,490,447 and 10,486,936, respectively (Table 3). The number of days of hospitalisation for HF as principal diagnosis and mentioned anywhere declined by 4.0% and 20.1%, respectively. Despite a higher number of separations in men with HF as principal diagnosis, total hospital days for HF were higher in women because of longer LOS among women (Table 3 and Fig. 1).

Table 3. Overall days of hospital admission and proportion of bed-days for separations involving heart failure as principal diagnosis in Australia, 1996–1997 to 2003–2004
YearHF as principal diagnosis (number)HF mentioned anywhere (number)Proportion of bed-days attributed to HF as principal diagnosis relative to total bed-days (%)Proportion of bed-days attributed to HF as principal diagnosis relative to bed-days for circulatory disease (%)
 MaleFemaleMaleFemaleMale (CI)Female (CI)Male (CI)Female (CI)
  1. a

    HF=heart failure.

  2. b

    CI=95% confidence interval.

1996–1997152,855167,472651,876789,6391.50 (1.49–1.51)1.37 (1.36–1.38)11.38 (11.33–11.43)14.42 (14.35–14.48)
1997–1998153,624164,000672,532812,4711.49 (1.48–1.50)1.34 (1.33–1.35)11.64 (11.59–11.69)14.38 (14.32–14.44)
1998–1999150,915162,023670,506806,2661.48 (1.47–1.49)1.33 (1.32–1.34)12.02 (11.96–12.8)14.85 (14.78–14.92)
1999–2000150,068165,430661,174785,1731.45 (1.44–1.46)1.35 (1.34–1.36)11.84 (11.78–11.89)15.20 (15.13–15.27)
2000–2001144,917161,339561,889655,9271.42 (1.41–1.43)1.32 (1.31–1.33)11.81 (11.75–11.87)15.07 (15.00–15.14)
2001–2002145,682163,088523,571611,5501.38 (1.37–1.39)1.29 (1.28–1.30)12.01 (11.95–12.07)15.39 (15.32–15.46)
2002–2003147,125154,299535,164598,2961.36 (1.35–1.37)1.21 (1.20–1.22)12.17 (12.11–12.22)14.86 (14.79–14.93)
2003–2004151,986155,624544,397606,5051.40 (1.39–1.41)1.22 (1.21–1.22)12.39 (12.33–12.45)15.34 (15.27–15.41)

Although the proportion of separations with HF mentioned anywhere that were non-fatal 1-day separations declined from 12.1% to 11.4%, this proportion when HF was recorded as the principal diagnosis increased from 10.0% and 8.5% in 1996–1997 to 14.6% and 13.2% in 2003–2004 in men and women, respectively (p for trend<0.001 in both sexes). In addition, the proportion of such separations with a record of at least one of medical, surgical or diagnostic procedure increased from 12.9% to 19.6% (Fig. 2).

image

Figure Fig. 2. Trends in crude proportion of non fatal 1-day separations for HF as principal diagnosis relative to all separations with HF as principal diagnosis.

Download figure to PowerPoint

3.4. HF as proportion of hospital activity

HF as a principal diagnosis accounted for 0.8% of all hospital separations and 9.7% of separations due to circulatory disease in 1996–1997. Both proportions declined statistically significantly over the period of study, to 0.6% (p for trend<0.001) and 9.5% (p for trend=0.002), respectively in 2003–2004. In terms of resources, HF as principal diagnosis accounted for 1.4% of total bed-days and 13.3% of hospital days attributed to circulatory disease. While the former proportion declined by 7.1% (p for trend=0.008), the latter increased from 12.8% to 13.7% (p for trend=0.002) (Table 3).

3.5. Case fatality

The overall age-standardized in-hospital case fatality for HF as principal diagnosis was 8.9% in men and 8.1% in women (p<0.001 for difference). Case fatality increased with age so that in patients aged 85 and over the risk of in-hospital mortality was more than four times the risk among patients aged less than 60 years. Case fatality declined over the period of study in both men and women, by 21.9% (p for trend=0.001) and 14.4% (p for trend<0.001), respectively (Fig. 3).

image

Figure Fig. 3. Trends in age-standardized in-hospital case fatality for HF as principal diagnosis.

Download figure to PowerPoint

3.6. More inclusive definitions of HF

Inclusion of other rubrics for HF such as alcoholic cardiomyopathy, dilated cardiomyopathy, ischaemic cardiomyopathy and hypertensive heart and/or renal disease which may lead to symptomatic HF did not change the observed trends in separation rates, the proportion of separations with principal diagnosis of HF relative to total hospital separations with diagnosis of circulatory disease, case fatality and 1-day separations. Even with a broader definition of HF, the rubrics 428 (in ICD-9CM) and I50 (in ICD-10AM) contribute to 96% of all separations with mention of HF in first position and in any position.

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

The present study shows that although the number of hospitalisations associated with HF recorded as principal diagnosis remained stable in Australia from 1996–1997 to 2003–2004, the rate of hospitalisation for HF, either as principal diagnosis or mentioned anywhere, declined. The decrease in rate was accompanied by a decline in separations involving HF as principal diagnosis as a proportion of total separations and of separations with diagnosis of circulatory disease. However, we observed an increase in the contribution of HF as a principal diagnosis to total hospital bed-days associated with circulatory disease and a change in clinical practice in terms of an increase in 1-day admissions.

The decline in the rate of hospitalisation for HF contradicts older reports [9][10][9–11], but recent studies suggest that the reported epidemic of HF is no longer reflected in hospital data. In fact, the decline in the rate of hospitalisation from HF in Australia is consistent with recent reports from Scotland [21], the Netherlands [22] and Canada [23]. In these countries, the rate of hospitalisation mostly remained stable or declined during the mid-1990s. By contrast, in USA, a recent report based on the National Hospital Discharge Survey (NHDS) showed that the rate of hospital admission for HF as principal diagnosis increased by 19% in women, from 1990 to 1999 [24]. In Australia, the rate of hospitalisation for HF as principal diagnosis declined gradually over the latter part of this period. Recent studies have shown declines elsewhere in both first-ever and recurrent admissions with HF [25].

An increase in prevalence of some risk factors for HF (such as diabetes, obesity and physical inactivity), an increase in awareness among medical practitioners of this condition (perhaps through the advent of new drugs for managing it) and use of new diagnostic technologies are three factors that potentially lead to an increase in the caseload attributed to HF. Yet, despite these influences and an ageing population, separation rates for HF as a principal diagnosis have been falling in Australia over the last decade. The reasons for the observed decline are not clear, but could include a fall in the incidence of HF [3], advances in treatment of HF particularly through increased use of ACE inhibitors and angiotensin receptor blockers, and other improvements in the quality of care (in-hospital and out-patient). For example Michalsen et al. assert that proper management of HF can decrease the rate of hospital admission by 40% [26].

However, care needs to be taken in interpreting trends in separations from hospital for HF. Ambiguity in diagnosis of HF (as an ill-defined condition) may be less likely to distort the recording of HF as a principal diagnosis in hospital data (as opposed, for example, to identification of HF as the underlying cause of death in death certificates [14]), but changes in policies affecting admission to hospital and evolution in medical practices regarding diagnosis and treatment of HF are important factors that may affect the trends in hospital statistics for HF.

Such factors make valid comparisons of the rates of hospital admission in different countries very difficult. For example, while the rate of hospitalisation for HF as a principal diagnosis in 1996–1997 was between 1.5 and 2.2 per 1000 in Canada [23], Australia and Scotland [21], the corresponding rate in the US was about 5 per 1000 in 1995 [8]. This difference may partly stem from a lower threshold for admission of patients with HF and differences in the approach to diagnosis of HF in the US.

The decline in the rate and number of mentions of HF in any diagnostic position in Australia is at odds with reports from Scotland [21] and the US [8] showing increases in the rate of HF mentioned anywhere during the 1990s. Part of the explanation for this divergence may lie in changes to rules governing the recording of prevalent conditions that did not necessarily affect the patient's clinical course but could potentially ‘inflate’ the Diagnosis Related Group and hence payment to the hospital. From 2000 in Australia, recording of a condition as an additional diagnosis was limited to those conditions which demonstrably affected in-hospital management in terms of diagnostic procedures, nursing care or therapeutic strategy [17]. Strict application of this new rule in 2000–2001 has resulted, by 2003–2004, in a reduction of 34.9% in admissions with a secondary diagnosis of HF recorded and had a similar effect on recording of other conditions such as diabetes [27].

Despite an increase in median of age from 77 to 79, the geometric mean LOS for non-fatal separations with HF recorded as principal diagnosis declined from 5.4 to 4.9 days over the period of our study. However, removing non-fatal 1-day stays from these separations revealed a much smaller decrease in mean LOS from 6.6 to 6.4 days. Indeed, the decline in LOS is mostly due to an increase in the number of non-fatal 1-day admissions.

Although HF is becoming less prominent in the overall activity of hospitals, it is simultaneously becoming more prominent in hospital cardiological practice. Improvement in the survival of patients with HF [2,3], recognition of HF in patients admitted with other chronic and debilitating conditions and a decline in average LOS for circulatory disease (from 9.8 days in 1993–1994 to 8.0 days in 2001–2002) are important factors contributing to this increase in prominence.

The increase in the proportion of non-fatal, 1-day admissions for HF as a principal diagnosis is one of the interesting findings in this study and probably reflects efforts to maximize efficiency in use of hospital resources. The corresponding increase in number of such separations with at least one procedure indicates that clinical practice in HF is changing; the percentage of separations with at least one surgical, medical or diagnostic procedure increased from 12.9% in 1996–1997 to 19.6% in 2003–2004. None of the available international studies [8,10,12,13,21,23,25] has reported on trends in 1-day admissions.

For those separations with HF as a principal diagnosis, inpatient case fatality declined in both sexes. However, the relative decline in men (21.9%) was higher than that in women (14.6%) such that, by 2003–2004, the case fatalities in both sexes were around 8.0%. Despite this improvement, the absolute number of fatalities at age 85 and over increased from 1,223 to 1,424. In 1996–1997 specifically, the case fatality in Australia (9.6%) was lower than that reported in Scotland (15.2% and 15.6% in men and women, respectively) [21] but higher than that in the USA (8.0%) in 1995 [8].

5. Study limitations

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

This study was based on the analysis of official NHMD. The collection covers all separations for the whole of a country whose population is now around twenty million. As hospital data in Australia are based on separations and do not identify individual patients, comparison between the rates in this study with studies that used linked data is difficult [21,25]. In addition, different policies regarding thresholds for hospital admission and variation in the definition and diagnosis of HF between doctors mean we should be cautious about generalizing our findings to other populations. However, these factors should have no effect on the observed trends for HF as principal diagnosis and these can be legitimately compared with those from other countries.

6. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

This study uses very recent data to show that the reported epidemic of HF is not reflected in hospital admissions in Australia. However, the contribution of HF to total bed-days attributed to circulatory diseases is still increasing. Increases in the proportion of non-fatal 1-day admissions and in separations with at least one procedure are indications of change in clinical practice regarding HF. Explanation for the observed declines in separation rates and in-hospital case fatality should be sought in a whole-of-community study that includes HF seen in outpatients and primary care settings.

Acknowledgement

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References

We are grateful to the Australian Institute of Health and Welfare for providing us with an anonymous copy of the national hospital morbidity records. The Government of Iran has provided a scholarship to support Dr. Najafi's studies in Australia.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Results
  6. 4. Discussion
  7. 5. Study limitations
  8. 6. Conclusion
  9. Acknowledgement
  10. References
  • 1
    Remme W.J. Swedberg K. Task Force for the Diagnosis and Treatment of Chronic Heart Failure: Guidelines for the diagnosis and treatment of chronic heart failure Eur Heart J 2001 22 1527 1560
  • 2
    Roger V.L. Weston S.A. Redfield M.M. et al. Trends in heart failure incidence and survival in a community-based population JAMA 2004 292 344 350
  • 3
    Levy D. Kenchaiah S. Larson M.G. et al. Long-term trends in the incidence of and survival with heart failure N Engl J Med 2002 347 1397 1402
  • 4
    McGovern P.G. Jacobs D.R. Jr. Shahar E. et al. Trends in acute coronary heart disease mortality, morbidity, and medical care from 1985 through 1997: the Minnesota Heart Survey Circulation 2001 104 19 24
  • 5
    Berry C. Murdoch D.R. McMurray J.J. Economics of chronic heart failure Eur J Heart Fail 2001 3 283 291
  • 6
    Stewart S. Jenkins A. Buchan S. et al. The current cost of heart failure to the National Health Service in the UK Eur J Heart Fail 2002 4 361 371
  • 7
    Croft J.B. Giles W.H. Pollard R.A. et al. National trends in the initial hospitalization for heart failure J Am Geriatr Soc 1997 45 270 275
  • 8
    Haldeman G.A. Croft J.B. Giles W.H. et al. Hospitalization of patients with heart failure: National Hospital Discharge Survey, 1985 to 1995 Am Heart J 1999 137 352 360
  • 9
    McMurray J. McDonagh T. Morrison C.E. et al. Trends in hospitalization for heart failure in Scotland 1980–1990 Eur Heart J 1993 14 1158 1162
  • 10
    Rodriguez Artalejo F. Guallar Castillon P. Banegas Banegas J.R. et al. Trends in hospitalization and mortality for heart failure in Spain, 1980–1993 Eur Heart J 1997 18 1771 1779
  • 11
    Reitsma J.B. Mosterd A. de Craen A.J. et al. Increase in hospital admission rates for heart failure in The Netherlands, 1980–1993 Heart 1996 76 388 392
  • 12
    Feldman D.E. Thivierge C. Guerard L. et al. Changing trends in mortality and admissions to hospital for elderly patients with congestive heart failure in Montreal CMAJ 2001 165 1033 1036
  • 13
    Ng T.P. Niti M. Trends and ethnic differences in hospital admissions and mortality for congestive heart failure in the elderly in Singapore, 1991 to 1998 Heart 2003 89 865 870
  • 14
    Najafi F. Dobson A. Jamrozik K. Is mortality from heart failure increasing in Australia? An analysis of official data on mortality for 1997–2003 Bull WHO 2006 84 722 728
  • 15
    Australian Institute of Health and Welfare (AIHW) Australian hospital statistics 2003–2004 Health Service Series 2005 vol. 23 Canberra AIHW
  • 16
    American Medical Association International Classification of Disease, 9th revision, Clinical Modification: ICD-9-CM 1999 1999 Chicago American Medical Association
  • 17
    National Centre for Classification in Health International Classification of Disease and related health problems, 10th revision, Australian Modification (ICD-10-AM) 2000 2nd ed. Sydney NCCH
  • 18
    Stata Corporation Stata 8 Statistics/Data analysis Special edition College station 2003 8 ed. Texas Stata Corporation
  • 19
    Breslow N. Day N. Statistical methods in cancer research Volume II - The design and analysis of cohort studies 1987 Oxford Oxford University Press
  • 20
    Fay M. Feuer E. Confidence intervals for directly adjusted rates: a method based on the gamma distribution Stat Med 1997 16 791 801
  • 21
    Stewart S. MacIntyre K. MacLeod M.M. et al. Trends in hospitalization for heart failure in Scotland, 1990–1996. An epidemic that has reached its peak? Eur Heart J 2001 22 209 217
  • 22
    Mosterd A. Reitsma J.B. Grobbee D.E. Angiotensin converting enzyme inhibition and hospitalization rates for heart failure in the Netherlands, 1980 to 1999: the end of an epidemic Heart 2002 87 75 76
  • 23
    Cujec B. Jin Y. Quan H. et al. The province of Alberta, Canada avoids the hospitalization epidemic for congestive heart failure patients Int J Cardiol 2004 96 203 210
  • 24
    Koelling T. Chen R. Lubwama R. et al. The expanding national burden of heart failure in the United States: the influence of heart failure in women Am Heart J 2004 147 74 78
  • 25
    Schaufelberger M. Swedberg K. Koster M. et al. Decreasing one-year mortality and hospitalization rates for heart failure in Sweden Eur Heart J 2004 25 300 307
  • 26
    Michalsen A. Konig G. Thimme W. et al. Preventable causative factors leading to hospital admission with decompensated heart failure Heart 1998 80 437 441
  • 27
    Phillips G. The impact of ICD coding standard changes for diabetes hospital morbidity data 2003 Canberra AIHW