Health Data Analytics

Rajiv Jayasena
Rajiv Jayasena
Team Leader
Health Data Analytics

Our work in this area supports acute hospitals by applying evidence-driven strategies to support improved health outcomes. For example, hospital overcrowding and timing of discharge are commonly linked to sub-optimal patient flow, poor quality of care, and unnecessary mortality. Consequently hospital services subscribe to theoretical targets for occupancy levels and discharge times. A better understanding of how occupancy levels and discharge times influence patient flow parameters, and more precise targets based on these, derived through modelling and simulation, would improve capacity management strategies and care outcomes.

1. Linking   ambulance, ED and admissions data Developing a data integration tool to produce linked datasets of clinical and administrative data, without   identifying information such as names and addresses
2. Disease   surveillance Using a variety of models (eg. adaptive   cumulative sum plans, internet search data, Twitter) to identify disease outbreaks as early as possible.
3. ED Length   of Stay performance Reviewing historical LOS performance; Identifying the relationship between ED LOS and in-hospital mortality; Understanding how individual waypoints in the patient journey impact the patient’s total time in the ED, and identify which waypoints act as critical flow bottlenecks.
4. Better   bed demand prediction Developing, validating and implementing web-based applications (Patient Admission Prediction Tool) to predict ED   presentations and subsequent hospital admissions and discharges across time of day, and day of the year.
5. Patient   flow visualisation Creating a user-friendly support tool   for bed administrators, to support their decision-making through viewing and analysing routinely collected hospital data.
6. Patient   flow and Hospital Occupancy Identifying stages of decline in patient flow system performance (‘choke points’) as a function of hospital   occupancy.
7. Bed   configuration Developing simulation models for patients admitted to inpatient beds from ED to assess how changing the numbers of beds in different specialities affects the waiting times for inpatient beds.
8. Adverse   event analysis Examining the relationship between daily hospital occupancy rates and the occurrence of reported adverse events
9. Early   discharge strategies Investigating the effects of varying inpatient discharge timing on ED length of stay and hospital occupancy, to determine the ‘whole of hospital’ response to discharge timing.
10. Readmission prediction Developing models that use patient data to understand the characteristics of ‘frequent flyers’ and how they utilise   the health system. This model can be used to identify inpatients pending discharge who have high risk of readmission to hospital.

Case Studies

Cutting Waiting Times

Software and Solutions

Patient Admission Prediction Tool