Skip to main content

Advertisement

SpringerLink
Log in

The risk-adjusted vision beyond casemix (DRG) funding in Australia

International lessons in high complexity and capitation

  • Original Papers
  • Published:
The European Journal of Health Economics, formerly: HEPAC Aims and scope Submit manuscript

An Erratum to this article was published on 01 May 2004

Abstract

Hospitals throughout the world using funding based on diagnosis-related groups (DRG) have incurred substantial budgetary deficits, despite high efficiency. We identify the limitations of DRG funding that lack risk (severity) adjustment for State-wide referral services. Methods to risk adjust DRGs are instructive. The average price in casemix funding in the Australian State of Victoria is policy based, not benchmarked. Average cost weights are too low for high-complexity DRGs relating to State-wide referral services such as heart and lung transplantation and trauma. Risk-adjusted specified grants (RASG) are required for five high-complexity respiratory, cardiology and stroke DRGs incurring annual deficits of $3.6 million due to high casemix complexity and government under-funding despite high efficiency. Five stepwise linear regressions for each DRG excluded non-significant variables and assessed heteroskedasticity and multicollinearlity. Cost per patient was the dependent variable. Significant independent variables were age, length-of-stay outliers, number of disease types, diagnoses, procedures and emergency status. Diagnosis and procedure severity markers were identified. The methodology and the work of the State-wide Risk Adjustment Working Group can facilitate risk adjustment of DRGs State-wide and for Treasury negotiations for expenditure growth. The Alfred Hospital previously negotiated RASG of $14 million over 5 years for three trauma and chronic DRGs. Some chronic diseases require risk-adjusted capitation funding models for Australian Health Maintenance Organizations as an alternative to casemix funding. The use of Diagnostic Cost Groups can facilitate State and Federal government reform via new population-based risk adjusted funding models that measure health need.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

Notes

  1. Consideration of the relative disadvantage of The Alfred vis a vis other teaching hospitals and the size of any other Risk Adjusted Specified Grants for other teaching hospitals can be further explored using the following formulae in the case of COPD, where severity markers are included into the equation, along with teaching hospital dummy variables for each teaching hospital and all other variables. This specification can be used both to predict why certain hospitals are more expensive than others, as well as to understand whether some factors systematically vary or are the same across all hospitals.

    Y:

    βo+β11*D1BR+β12*D1LPA+β13*D1HLT+β14*D1LT+β15*D1BIBAP+β16 1AGE+ß18*D1SEX+ß19*D1PROC+ß110*D1DIAG+ß111*D1DISEASE *D1LVF+β17*D1AGE+β18*D1SEX+β19*D1PROC+β110*D1DIAG+β111*D1DISEASE TYPES+β112*D1COMPLEX+β113*D1OUTLIER+β114*D1EMERG+β21*D2BR+β22*D2LPA+β23SEX+ß29*D2PROC+ß210*D2DIAG+ß211*D2DISEASE *D2HLT+β24*D2LT+β25*D2BIPAP+β26*D2LVF+β27*D2AGE+β28*D2SEX+β29 N5*DNBIPAP+ßN6*DNLVF+ßN7*DNAGE+ßN8*DNSEX+ßN9*DNPROC+ßN10*DNDIAG+ßN11*DNDISEASE *D2PROC+β210*D2DIAG+β211*D2DISEASE TYPES+β212*D2COMPLEX+β213*D2OUTLIER+β214*D2EMERG+...βN1*DNBR+βN2*DNLPA+βN3*D*** *DNHLT+βN4*DNLT+βN5*DNBIPAP+βN6*DNLVF+βN7*DNAGE+βN8*DNSEX+β N9*DNPROC+βN10*DNDIAG+βN11 *D NDISEASE TYPES+βN12*DNCOMPLEX+βN13*DNOUTLIER+βN14*DNEMERG+E

    Where

    Y:

    =Per patient costs

    βo :

    =Y intercept

    βij :

    =array of coefficients, one set for each of j hospitals

    D1BR:

    =Dummy variable bronchiectasis teaching hospital D1=1, other=0.

    D1LPA:

    =Dummy variable lung part absence teaching hospital D1=1, other=0

    D1HLT:

    =Dummy variable heart and lung transplantation teaching hospital D1, =1, other=0

    D1LT:

    =Dummy variable lung transplantation teaching hospital D1=1, other=0.

    D1BIPAP:

    =Dummy variable Bilevel Positive Airway pressure (BIPAP) teaching hospital D1=1, other=0.

    D1LVF:

    =Dummy variable Left Ventricular Failure teaching hospital D1=1, other=0.

    D1AGE:

    =Patient age, teaching hospital D1=1.

    D1SEX:

    =Dummy variable 1 if male, other=0 (gender of patient), teaching hospital D1

    D1PROC:

    =Number of procedures at teaching hospital D1

    D1DIAG:

    =Number of diagnoses at teaching hospital D1

    D1DISEASE TYPES:

    =Number of body systems at teaching hospital D1

    D1COMPLEX:

    =Dummy variable at teaching hospital D1, 1 of patient classified as high complexity case (PCCL) level 1 if 4. 0 if 3.

    D1OUTLIER:

    =Dummy variable at teaching hospital D1, 1 if patient an outlier on length of stay, otherwise 0

    D1EMERG:

    =Dummy variable at teaching hospital D1, 1 if patient admitted through emergency department, otherwise 0

References

  1. Antioch KM (2002) Book review on the social economics of health care. Aust N Z J Public Health 26(3):288–289

    Google Scholar 

  2. Antioch KM, Walsh MK (2000) Funding issues for Victorian hospitals: the risk-adjusted vision beyond casemix funding. Aust Health Rev 23(3):145–53

    CAS  Google Scholar 

  3. Antioch KM, Walsh MK (2002) Risk-adjusted capitation funding models for chronic disease in Australia: alternatives to casemix funding. Eur J Health Econ 3:83–93

    Article  Google Scholar 

  4. Antioch KM, Zhang X (2000) Organ transplant AN-DRGs: modifying the exceptions hierarchy in casemix classification. Aust Health Rev 23(1):137–152

    CAS  Google Scholar 

  5. Antioch KM, Walsh MK, Anderson D et al (1998) A new ambulatory classification and funding model for radiation oncology: non-admitted patients in Victorian hospitals. Aust Health Rev 21(1):62–76

    CAS  Google Scholar 

  6. Antioch KM, Zhang X, Raw J (1998) Using endoscopic procedures for AN-DRG assignment: Australia leads the way. Aust Health Rev 21(4):80–95

    CAS  Google Scholar 

  7. Antioch KM, Walsh MK, Anderson D et al (1999) Forecasting hospital expenditure in Victoria: lessons from Europe and Canada. Aust Health Rev 22(1):133–155

    CAS  Google Scholar 

  8. Antioch KM, Chapman R, Elliot B et al (2001) Cost-effective clinical pathways at the Alfred Hospital: international lessons from Bayside Health. Aust Health Rev 24(4):21–29

    CAS  Google Scholar 

  9. Antioch KM, Jennings G, Botti M et al (2002) Integrating cost-effectiveness evidence into clinical practice guidelines in Australia for acute myocardial infarction. Eur J Health Econ 3(1):26–39

    Article  Google Scholar 

  10. Ash AS, Ellis RP, Pope GC et al (2000) Using diagnoses to describe populations and predict costs. Health Care Financ Rev 21(3):7–28

    CAS  Google Scholar 

  11. Australian States and Territories (2002) Facing the future: opportunities for better health services.

    Google Scholar 

  12. Carpenter CE, Rosko MD, Louis DZ et al (1999) Severity of illness and profitability: a patient level analysis. Health Services Manage Res 12(4):217–226

    CAS  Google Scholar 

  13. Duan N et al (1983) Smearing estimate: a non-parametric retransformation method. J Am Stat Assoc 78:605–690

    Google Scholar 

  14. Duckett SJ (1995) Hospital payment arrangements to encourage efficiency: the case of Victoria, Australia. Health Policy 34:113–134

    Article  CAS  PubMed  Google Scholar 

  15. Ducket SJ, Agius PA (2002) Performance of diagnosis-based risk adjustment measures in a population of sick Australians. Aust N Z J Public Health 26(6):500–507

    PubMed  Google Scholar 

  16. Duckett SJ, Stow S (2001) Payment policy for chemotherapy in public hospitals: a case for splitting DRGs 780/R63Z. Aust N Z J Public Health 25(6):552–525

    CAS  PubMed  Google Scholar 

  17. Dwyer JM (2002) The next Australian Health Care Agreements: what clinicians want. Aust Health Rev 25(6):17–23

    PubMed  Google Scholar 

  18. Ellis R (1999) Diagnostic cost group (DCG) risk adjustment models. DxCG, Waltham

  19. Ellis RP, Ash A (1995) Refinements to the diagnostic cost group model. Inquiry 32(4):1–12

    Article  Google Scholar 

  20. Ellis RP, McGuire TG (1986) Providers behavior under prospective reimbursement: cost sharing and supply. J Health Econ 5:129–51

    Article  CAS  PubMed  Google Scholar 

  21. Ellis RP, Pope GC, Iezzoni LI, Ayanian JZ, Bates DW, Burstin H, Ash AS (1996) Diagnosis-based risk adjustment for Medicare capitation payments. Health Care Financ Rev 12(3):101–28

    Google Scholar 

  22. Ellis RP, Pope GC, Iezzoni LI et al (1996) Diagnostic cost group (DCG) and hierarchical coexisting condition and procedures (HCCP) models for Medicare risk adjustment. Final report. Health Care Financing Administration: Baltimore

    Google Scholar 

  23. Etzioni A (1988) The moral dimension: towards a new economics. Free Press: New York

    Google Scholar 

  24. Federal Department of Health and Aged Care (1999) The Australian coordinated care trials. interim technical national evaluation report. Federal Department of Health and Aged Care: Australia

    Google Scholar 

  25. Kennedy P (1992) A guide to econometrics. Basil Blackwell, Oxford

  26. Kugler C, Freytag S, Stillger R, Bauer P, Ferbert A (2000) Australian refined diagnosis related groups. Formal and inherent problems of grouping with the example of stroke care. Dtsch Med Wochenschr 125(51–52):1554–1559

    Google Scholar 

  27. Leary RS, Johantgen, ME, Farley D et al (1997) All-payer severity—adjusted diagnosis-related groups: a uniform method to severity adjust discharge data. Top Health Inf Manage 17(3):60–71

    CAS  PubMed  Google Scholar 

  28. Lilien DM, Startz R, Ellsworth S, Noh J, Engle R (1995) E views users guide econometric view: micro TSP for Windows and the Macintosh. Quantitative Micro Software: Irvine

    Google Scholar 

  29. Manning WG (1998) The logged dependent variable, heteroskedasticity and the retransformation problem. J Health Econ 17(3):283–295

    Article  CAS  PubMed  Google Scholar 

  30. Manning WG et al (1987) Health Insurance and the demand for medical care: evidence from a randomised experiment. Am Econ Rev 77:251–277

    CAS  PubMed  Google Scholar 

  31. McNair P, Duckett S (2002) Funding Victoria’s public hospitals: the casemix policy of 2000–2001. Aust Health Rev 25(1):72–98

    Google Scholar 

  32. Mullahy J (1998) Much ado about two: reconsidering retransformation and the two-part model in health econometrics. J Health Econ 17(3):247–282

    Article  CAS  PubMed  Google Scholar 

  33. Peacock S, Segal L (2000) Capitation funding in Australia: imperatives and impediments. Health Care Manage Sci 3:77–88

    Article  CAS  Google Scholar 

  34. Podger A, Hagan P (1999) Reforming the Australian health care system: the role of government. Occasional papers. New series no 1. Commonwealth Department of Health and Aged Care

    Google Scholar 

  35. Pope CG, Ellis RP, Liu CF et al (1998) Revised diagnostic cost group (DCG)/hierarchical coexisting conditions (HCC) models for Medicare risk adjustment. Final report. Health Care Financing Administration. Health Econ Res: February

    Google Scholar 

  36. Pope, GC, Ellis, RP, Ash, AS et al (2000) Principal inpatient diagnostic cost group model for Medicare risk adjustment. Health Care Financ Rev 21(3):93–118

    CAS  PubMed  Google Scholar 

  37. Productivity Commission (2002) Managed Competition in Health Care Workshop proceedings: Canberra 23 August, 2002. Productivity Commission: Australia

    Google Scholar 

  38. Reinhardt UE (1992) Reflections on the meaning of efficiency: can efficiency be separated from equity? Yale Law Policy Rev 10:302–315

    Google Scholar 

  39. Rice T (1998) The economics of health reconsidered. Health Administration: Chicago

    Google Scholar 

  40. Rice N, Smith R (1999) Approaches to capitation and risk adjustment in health care: an international survey. Centre for Health Economics, University of York: York

  41. Rochell B, Roeder N (2000) Vergleichende Betrachtung und Auswahl eines Patienten-klassifikationssystems auf der Grundlage von Diagnosis Related Groups. Krankenhaus nos 4, 5

  42. Roeder N, Rochell B, Juhra C et al (2001) Empirical comparison of DRG variants using cardiovascular surgery data: initial results of a project at 18 German hospitals. Aust Health Rev 24(4):57–80

    CAS  PubMed  Google Scholar 

  43. Segal L, Donato, R Richardon J, Peacock S (2002) Strengths and limitations of competitive versus non-competitive models of integrated capitated fundholding. J Health Serv Res Policy 17 [Suppl 1]:56–64

    Google Scholar 

  44. Shen Y, Ellis RP (2002) How profitable is risk selection: a comparison of four risk adjustment models. Health Econ 11(2):165–174

    Article  PubMed  Google Scholar 

  45. Van de Ven WPMM, Ellis R (2000) Risk adjustment in competitive health plan markets. In: Newhouse JP, Culyer AJ (eds) Handbook of health economics. Elsevier: Amsterdam, pp 755–845

  46. Victorian Department of Human Services (2000) Victorian public hospitals policy and funding guidelines 2000–2001. Acute Health Division, Victorian Government: Melbourne

  47. Victorian Department of Human Services (2002) Victorian public hospitals policy and funding guidelines 2002–2003. Metropolitan Health and Aged Care Division Victorian Government: Melbourne

    Google Scholar 

  48. Willcox S (2001) Promoting private health insurance in Australia: a policy in search of evidence? Health Aff (Millwood) 20(3):152–161

    Google Scholar 

  49. Zhao Y Ellis RP, Ash AS et al (2001) Measuring population health risks using inpatient diagnoses and outpatient pharmacy data. Health Services Res 26(6):180–193

    Google Scholar 

Download references

Acknowledgements

K.A., Chair of the State-wide Risk Adjustment Working Group (RAWG), would like to thank Profs. Randall P. Ellis and Wynand PMM Van de Ven for excellent tuition in risk-adjustment methodologies during the 1999 International Health Economists Association Conference, The Netherlands, and for subsequent assistance. We also thank the reviewers of the European Journal of Health Economics for excellent comments. Earlier versions of this contribution were presented at the Fourth European Conference on Health Economics, Paris, 2002, the Financial Review Conferences: the Fifth Health Congress: Australia’s Leading Health Industry Event. Sydney 2003, the International Health Economists Association Conference, 2003, United States and the Australian and New Zealand Health Services Research Conference, 2003, Melbourne.

Author information

Authors and Affiliations

  1. Bayside Health, The Alfred Hospital, Melbourne, Australia

    Kathryn M. Antioch & Michael K. Walsh

  2. Department of Epidemiology and Preventive Medicine, Faculty of Medicine, Monash University, Melbourne, Australia

    Kathryn M. Antioch

  3. Bayside Health, The Alfred Hospital, Commercial Rd Prahran, 3181, Melbourne, Victoria, Australia

    Kathryn M. Antioch

Authors
  1. Kathryn M. Antioch
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Michael K. Walsh
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Kathryn M. Antioch.

Additional information

An erratum to this aticle can be found at http://dx.doi.org/10.1007/s10198-004-0239-0 finden.

Glossary

Glossary

Bed-day gap is a measure used by the HRT. It is a measure of the inlier days beyond the 75th percentile plus the outlier days beyond the outlier trim point. It therefore measures the potential efficiencies that could be obtained if all stays were reduced to the 75th percentile LOS.

Relative stay index is the casemix adjusted LOS. Casemix adjustment is achieved by comparing each hospital’s number of bed-days with an expected number of bed-days using the Health Round Table in the prior year as the benchmark.

White heteroskedasticity test examines whether the error variance is affected by any of the regressors, their squares or their cross-products. It tests whether any heteroskedasticity that is present causes the variance-covariance matrix of the ordinary least squares estimator to differ from its usual formula [25]. The output from the test is an F statistic and a statistic that has an asymptotic χ2 distribution with a number of degrees of freedom equal to the number of regressors and squared regressors in the test regression. Each statistic provides a test of the hypotheses that the coefficients of the variables in the augmented regression are all zero. This is a general test for model misspecification since the null hypothesis underlying the test assumes that the errors are both homoskedastic and independent of the regression, and that the linear specification of the model is correct. Failure on any of these conditions could lead to a significant test statistic. Conversely, a non-significant test statistic is very reassuring since it implies that none of the three conditions is violated [28].

Rights and permissions

Reprints and permissions

About this article

Cite this article

Antioch, K.M., Walsh, M.K. The risk-adjusted vision beyond casemix (DRG) funding in Australia. HEPAC 5, 95–109 (2004). https://doi.org/10.1007/s10198-003-0208-z

Download citation

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10198-003-0208-z

Keywords

Search

Navigation