Table 3 The five-year vision for clinical analytics in NSW – a one-page narrative

In five years’ time …

Clinicians will use patient reported measures as a part of routine care. The measures will be used for diagnosis, prognosis and clinical decision making. Clinically validated algorithms will assess case histories, diagnoses and risk profiles; and will facilitate safe and effective clinical care. Targeted and well validated alerts will highlight risk and safety issues. Aggregated, time-series data will be collected unobtrusively through the electronic medical record (eMR) and routine clinical tasks.

Clinicians will have access to relevant and timely information that highlights any unwarranted clinical variation and supports reflective and current best practice. Information will be available at the point of care on concordance of clinicians’ care with evidence-based practice; risk adjusted patient outcomes; benchmarking and peer comparisons; time-series and patient trajectories. Advanced analytics or artificial intelligence (AI) approaches will be deployed to discern novel patterns in complex and large datasets and guide the development of algorithms. Analytics-driven clinical audit processes will draw on “virtual registries” to personalise learning.

Feedback will be informed by the evidence on clinical decision making – incorporating passive ‘automated’ predictive analytics as well as peer to peer and expert feedback. Data will be discussed within clinical teams so that clinicians can collectively assess the data and identify causes of variation and plan improvements. Clinical research will be informed by timely and efficient access to linked data, big data, “virtual registries” and analytics. Efforts will be underway to secure wider data linkage to incorporate non-health sources. Clinician training will incorporate the use of analytics and address issues such as managing risk and uncertainty.

Patients will be assured that their data are appropriately secure and used to support clinical care and quality improvement. They will be firmly established as key informants in healthcare – providing data about their health status, experience and outcomes. Patients who chose to, will be engaged in monitoring their health using technologies that can communicate with information systems. Patients will be enabled and supported to access their own data and to use it to manage their health. With their consent, patient self-management will be prompted by algorithm enabled alerts.

Managers will be confident that monitoring and measurement systems are reliably and sensitively assessing healthcare services. They will be able to test models of reconfiguration and structural changes using data analytics. Real time alerts regarding impending surges in demand in acute care areas such as emergency departments, operating theatres and critical care units will be used to manage workflows, staffing and bed management.

Service level and system managers will utilise data from clinical analytics alongside administrative and other data to guide policy development and improve performance. There will be a robust mechanism and framework to identify, prioritise and support the introduction of system wide clinical analytic initiatives.