Background and Methods

The project was borne of the observed increase in hospital demand in South Australian public hospitals over recent years. The following sections describe the four phases of the project, and the respective components.

Phase 1 | Phase 2 | Phase 3 | Phase 4

Phase 1

Aim: Establish rank ordering of DRGs with respect to increased numbers of separations and increased mean Length of Stay (LoS)

Tasks:

1. Map DRG codes over analysis time horizon (5 years, 10 years?)
2. Use DRG coding accuracy audit data to adjust DRG data
3. Establish numbers of same day and multi-day separations per month by DRG across public sector hospitals
4. Establish monthly estimates of the mean LoS by DRG across public sector hospitals
5. Establish monthly estimates of costs by DRG across public sector hospitals
6. Identify DRGs with evidence of increased demand and/or costs
7. Identify relevant time periods for comparison of pre- and post-increased demand and/or cost (separately for each DRG identified in Task 6)
8. Estimate increases in separations, LoS, and cost
9. Present rank ordered DRGs and associated increases in demand and costs to Steering Committee to inform the order of selection of case studies for Phase 2

Phase 2

Aims:

1. Identify specific patient groups with observed increases in demand
2. Quantify the proportion of increased demand attributable to technological advancement, and changes in thresholds for admission, quality of care, private to public transfers, population prevalence, and Specialty capacity protection.
3. Identify specific patient groups with the greatest potential to be managed more efficiently

Tasks:

1. For each selected DRG, re-do Phase 1 Tasks for each ICD-10 code
2. Select ICD-10 codes with greatest increased demand, and re-do Phase 1 Tasks for patient sub-groups, defined by patient characteristics collected by the Integrated South Australian Activity Collection (ISAAC)
3. Select patient groups with greatest increased demand, and for each group:
   1. Undertake analyses of local and national population prevalence/incidence data to inform potential magnitude of epidemiological causes of increased demand
   2. Survey clinical experts to identify changes in practice, with linked literature review to quantify expected effects of changes on threshold for admission and LoS
   3. Analyse readmission rates as a proxy for changes in quality of care, with potential for primary data linkage to link admissions across hospitals
   4. Re-do Phase 1 Tasks for private sector activity to inform the impact of private to public transfers (or does ISAAC inform whether a patient came from the private sector?)
4. Combine absolute measures of increased demand with attributable proportions to rank patient groups with respect to potential for cost-effective transfer of care to an out-of-hospital setting (e.g. those with largest absolute impact of ‘changing thresholds’ and ‘capacity protection’)

Phase 3

Aim: To identify alternative, more cost-effective, management pathways for ‘new demand’ patients (either in- or out-of-hospital)

Tasks:

1. For each selected patient group, analyse cross-hospital technical efficiency using the net benefit correspondence theorem (need to define methods for estimating cost per separation for included patient groups, and which measures of outcome to use)
2. Select high and low performing hospitals, and analyse management pathways to identify relevant differences that are causing alternative levels of efficiency. This will involve primary analysis of individual hospital processes in both the high and low performing hospitals, hopefully via survey/interview of relevant clinical staff.
3. Review published guidelines and health technology assessments (HTAs) for the selected patient group to identify evidence on the cost-effectiveness of alternative management pathways.
4. If evidence is lacking, undertake HTA using data from high performing hospitals (representing high technical efficiency) to inform current practice. Retrospective data will be sought from patient records. Data for the comparator(s) may be collected from areas in which other management pathways are implemented, or by synthesising primary and secondary data.

Phase 4

Aim and Task: To compare the impacts on costs, health, hospital beds, and equity of the alternative strategies tested in Phase 3 for all selected patient groups.