Introduction
The NSW Ministry of Health has established Lumos; a pioneering initiative that links general practice data to a number of health system data collections, creating a unique data asset that spans health care services across NSW. Lumos sheds light on the patient journey through the NSW health system by “Linking Up and Mapping Of Systems”, providing an essential platform for the design, delivery and evaluation of integrated value-based care pathways. Over the next three years Lumos will expand state-wide, linking data from up to 500 general practices across all 10 PHNs and 17 LHDs in NSW.
Policy context and objective
With the growth and ageing of Australia’s population, and with the incidence of complex and chronic health conditions in the population, people’s health needs are changing. To meet these changing needs, the challenge for the health system is to integrate care across the continuum of care settings and providers, delivering person-centred, seamless, efficient and effective care.
Lumos aims to provide this “whole of system” view of the patient journey. It delivers a trusted source of information that can be used to guide service design and integration, the commissioning of services, improved patient experience and outcomes, and the strategic directions of the NSW health system.
Targeted population
The Lumos cohort comprises anyone who attended participating general practices as a patient since 2010. It builds on the NSW GP Data Linkage Pilot Project which has successfully linked the GP data of approximately 400,000 patients from 40 general practices in NSW. The number of patients will increase as the Lumos Program progresses with repeated and expanding data collections and linkages.
Highlights (innovation, impact and outcomes)
Lumos has provided participating GPs with a new picture of patients’ healthcare journeys. For instance, 20% of their patients attended an emergency department (ED) in the past year, and many of these were for semi-urgent and potentially avoidable reasons. GPs can view how their patients’ ED presentations relate to their characteristics and patterns of GP visits, and how this compares to other practices. This data is having a tangible impact on GPs, prompting one practice to extend its opening hours and another to follow up patients with chronic conditions to help reduce hospitalisations.
There are also new insights for acute care. People who smoke spend, on average, a full day longer in hospital than those who do not smoke. People with mental health diagnoses have higher use of all hospital services than people with other chronic conditions. Such statistics provide actionable insights into preventable hospital use.
Conclusions
Lumos has the potential to drive shared objectives across health services by providing an evidence-based platform to align strategy, policy and practice in patient care across NSW. The imperative to create the information needed to support health system redesign is now well established. Effort now concentrates on the processes needed to facilitate building secure, safe data assets that can deliver the much-needed insights for health care system reform and improved patient outcomes.

Healthcare is facing numerous challenges in an increasingly digital world. The availability of healthcare data is opening up opportunities to improve, or fundamentally change, methods of healthcare delivery. Artificial Intelligence, personalised medicine, big data are all offering outstanding potential (1). However all of these potentialities are underpinned by good data, and by comprehensive data. Both of these are currently lacking.

The path to good data relies in the efforts of clinicians to record data traditionally in only human readable form, into a form that is also machine readable (2). The next barrier is attempting to merge the various levels of data held in disparate systems, across primary, secondary care and across various institutions.

Outcome Health manages the POLAR program and its research arm, Aurora. Collecting data from over 600 practices across the eastern seaboard of Australia, it represents one of the largest GP data sources in the country. The program is designed around not just collecting data, but ensuring levels of data quality using a comprehensive strategy at all levels of data use (3).

But using primary care data alone is not sufficient to deliver care to patients. By using linked data we have developed a proof of concept of an emergency risk prediction tool using machine learning (4). By linking data between emergency departments and general practice, we were able to map the general practice journeys of those admitted to emergency over a 5 year period, and develop an electronic view of risk.

This has oped the door to an understanding that old concepts of data custodianship are outdated, and now to deliver care to patients also means having access to the relevant data about that patient, wherever it was created and is stored.

To do this we have embarked on a program of developing a framework for linked data between primary care, hospital settings and data held by the state. This raises significant challenges, both technical and social. Technical problems are the nature of ensuring adequate, verified linkage according to multiple needs, and quality measures of data. Social factors include the ethics and consent models, that are currently institution based, rather than patient based. Meeting these challenges is a priority and will be part of the presentation.


1. Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56.
2. de Lusignan S, Hague N, van Vlymen J, Kumarapeli P. Routinely-collected general practice data are complex, but with systematic processing can be used for quality improvement and research. Informatics in primary care. 2006;14(1):59-66.
3. Pearce C, Mcleod A, Rinehart N, Ferriggi J, Shearer M. What does a comprehensive, integrated data strategy look like: The Population Level Analysis and Reporting (POLAR) program. Stud Health Technol Inform. 2019;In Press.
4. Pearce C, McLeod A, Rinehart N, Patrick J, Fragkoudi A, Ferrigi J, et al. POLAR Diversion: Using General Practice Data to Calculate Risk of Emergency Department Presentation at the Time of Consultation. Appl Clin Inform. 2019;10(1):151-7.

Introduction & Aim:

In January 2019 the 2nd edition of the National Safety & Quality Health Service Standards was published, with a new standard 5 - Comprehensive Care. This standard required a more integrated approach to assessment of patient clinical risk. The aim of this project was to meet the requirements of the new standard, and also reduce unnecessary duplication in paperwork for hospital direct care staff.

Method/Change:

A new comprehensive clinical risk screening was designed as a standard tool for Healthscope hospitals, aimed at a general medical/surgical overnight patient population. The 4-page fold-out form comprised integrated risk screening in the areas of falls, pressure injury, cognitive impairment, malnutrition, venous thromboembolism, behavior, mental health and medication management. These areas were previously assessed across multiple forms, many of which contained duplicate documentation fields. Specialised versions of the form were modified for paediatric, obstetric and day procedure populations.

A formal evaluation was undertaken to ensure that removal of a numerical risk assessment scale did not result in an increase in falls. Throughout the process, the value of expert clinical judgement was recognized and emphasized, while still retaining checklists to ensure consistency of practice. The form contains both risk screening and prevention strategies, and is typically filed at the point of care, alongside the Nursing Care Plan, observation chart and medication chart. The comprehensive risk screening allows all clinical risks to be evaluated in a more integrated way, with strategies for prevention optimized. The paper form was also transferred into a format suitable for an Electronic Medical Record.

Outcomes & Lessons Learnt:

The new form was designed, implemented and evaluated with active involvement of direct care nurses, consumer representatives, allied health and nurse managers. The first draft of the form, with risk mitigation strategies within the care plan, was unsuccessful, and both forms were remodeled in line with user feedback, before a second multi-hospital trial was undertaken. Lessons learnt from the first trial were applied in the second trial form which has now been successfully implemented in 10 hospitals. From trial to implementation was approximately a 12 month period. The form will roll out to 43 Healthscope hospitals in all Australian states, showing broad applicability and transferability. The new tool will be sustainable, as multiple previous forms will be withdrawn as the project rolls out to all hospitals.

A highlight of the new tool has been the integration with both the Patient Health History (completed by the patient before admission) and the Nursing Care Plan. Updating these forms concurrently has allowed further removal of duplication of fields, which has increased staff satisfaction and acceptance of the form.

The new comprehensive risk screening form resulted in a reduction in paperwork from 10 pages to 4. This is a reduction in time spent on paperwork, as well as enhanced compliance with standards and best practice in risk screening and assessment. Clinical risks are identified in a timely way on admission, and risk mitigation strategies implemented, allowing comprehensive care to be the focus for all patients.

Introduction & Aim:

This paper describes how aligning two innovative core quality indicators with values and objectives drives improvements in patient safety and quality of care.

Method/Change:

In 2017, Healthscope selected two new measures for Executive and Board quality Key Performance Indicator reporting for each of their 43 hospitals:
• Quality Clinical Outcomes – measured by Hospital Acquired Complication (HAC) rate
• Patient Experience – measured by overall rating of the quality of treatment and care using the Australian Hospital Patient Experience Question Set (AHPEQS)

The focus on two core measures was a new innovation aimed at simplifying and aligning core Board and Executive strategies, values and company purpose. Selection of the measures was a collaborative top down and bottom up approach, with direct care staff, managers, executive and board involvement in selecting the measures to align with the core values of the company – Quality Clinical Outcomes and Exceptional Patient Care. Over 100 clinical quality indicators are measured and reported by Healthscope, however streamlining the measures and focusing on priorities was a key.

Neither of these measures had previously been used in the industry, however HAC rate was foreshadowed by the Independent Hospital Pricing Authority as a penalty measure in their commonwealth public hospital funding model.

Outcomes & Lessons Learnt:

As both measures were new in the industry, implementation was a major change management project. Frequency of monitoring was increased to monthly, with real-time dashboards at ward level.

Focus and simplification of core quality indicators, with alignment to objectives was a major catalyst for change and continuous improvement in quality and patient safety. Both core measures showed significant improvement in year 1, with a continued positive trend in year 2.

Hospital acquired complication rate in year 1 was initially a pure rate, with hospital targets focused on improvement over time, as well as a focus on data quality. In year 2, risk adjusted methodology allowed hospitals to compare specific complication types with peer hospitals, public and private, with monitoring of HACs in real time.

The focus for the AHPEQS patient experience measure in year 1 was at hospital level, with ward-level real-time reporting and a person-centred care strategy in year 2 driving further improvements.

Conclusion:

Aligning innovative core quality measures with organisation purpose and goals, coupled with monthly reporting to Board and Executive, has driven significant change and improvement.

Aim:

This paper explains an innovative method to analyse Hospital Acquired Complications (HACs) and compare performance with peer hospitals to improve patient safety and quality of care.

Design/Methodology:

In 2017, Healthscope was an early adopter of the Hospital Acquired Complications (HAC) methodology developed by the Australian Commission on Safety & Quality in Healthcare, using HAC rate as a core Quality KPI for Board and Executive reporting. The HAC measure was ideal for integrated care monitoring of complications, being a combination of 16 different potentially preventable hospital acquired complications, such as falls, infection, haemorrhage, embolism, falls and medication errors.

After using a pure HAC ratio for 12 months, two methods for risk adjustment were developed and trialed. These were based on the Independent Hospital Pricing Authority (IHPA) risk adjustment methodology utilised by the public hospital sector.

In methodology (i) Healthscope and IHPA risk adjusted coefficients were calculated for each risk factor and then used to calculate expected HAC cases and the risk adjusted HAC ratio (expected HAC cases/actual HAC cases) for each HAC group for each of Healthscope’s hospitals. Calculated ratios were displayed in a separate funnel plots.

In methodology (ii) IHPA risk adjusted coefficients were applied to every patient episode, with complexity score calculated for each HAC for the target hospital. Overall complexity score was then used to identify peer hospitals with the same patient complexity, from a dataset of 400 public and private hospitals. After identifying peers, HAC ratios were compared and the target hospital ranked as Best Practice, Better than Peers, Equal to Peers, Worse than Peers or Worst Practice.

Outcomes/Findings:

Methodology (i) facilitated internal benchmarking and demonstrated that Healthscope hospitals were well within the confidence interval range of the funnel plot.
Methodology (ii) benchmarked performance against external private and public hospitals with the same complexity/casemix. A web based benchmarking portal was designed, allowing hospitals to view results in real time, and drill down to patient level details. Benchmarking each complication type allowed hospitals to identify which complication was causing the most harm, after risk adjustment.

Conclusion:

Methodology (ii) was identified as the most suitable model, as it had the ability to compare ratios with peer hospitals, identify best practice and focus on opportunities for improvement. Use of a risk adjustment methodology facilitates focus on complications that have the greatest potential for change and improvement in patient safety and quality of care.