Gk86gr2q 1458190630

No-one should get dud hospital care – it’s time to lift our game on quality and safety

Some hospitals have substantially higher costs. Others have higher rates of death. racorn/Shutterstock

No-one should get dud hospital care – it’s time to lift our game on quality and safety

Some hospitals have substantially higher costs. Others have higher rates of death. racorn/Shutterstock

In 2013-14, Australian governments spent A$105 billion on health; A$44 billion of that was on public hospitals.

The Commonwealth government is increasingly concerned with the size of the health budget and has acted to reduce the inappropriate use of Medicare benefits. But the Commonwealth government has less influence on public hospitals because the state and territory governments control their expenditure.

State governments are facing tighter budgets as demand for heath care increases due to an ageing population, greater rates of chronic disease and more service use generally.

The collection and analysis of data on the performance of our health-care system can be used to improve the quality of health services and maybe also reduce costs.

At a national level, the clinician-led Choosing Wisely campaign is developing lists of specific tests, treatments and procedures that may be unnecessary and sometimes harmful for individual patients. Recommendations include reducing use of CT scans in the emergency department and not ordering x-rays for patients with uncomplicated acute bronchitis.

But while improving the decisions made by individual doctors is important, there remain other causes of substantial variation in the safety and quality of care provided in Australian hospitals. This needs to be addressed.

Varied quality and safety

Efforts to improve the quality of care in hospitals have traditionally been left to individual hospitals and their managers. But we now have the data to compare different hospitals. We can identify the best and worst performers and, most importantly, determine how to boost the performance of the stragglers.

Identifying and intervening to improve low-quality care requires financial investment. But there are significant potential long-term savings, due to improved efficiency and better patient outcomes.

In New South Wales, the Bureau of Health Information has developed and tested methods for comparing the death rates within 30 days of treatment for heart attacks, strokes, pneumonia and hip fracture surgery.

For stroke patients, ten hospitals had noticeably higher-than-expected death rates for these conditions. An additional 16 deaths were observed in every 100 patients treated at a low-performing hospital compared to a high-performing hospital.

Clinical auditors and review panels should investigate differences in the care provided at the high- and low-performing hospitals and approaches to improve care quality.

Other data show the costs of treating similar conditions varies dramatically. A Grattan Institute analysis shows the average cost of performing a hip replacement at different hospitals ranges from under A$10,000 to more than A$30,000.

Further investigation may find the higher costs are due to the use of more expensive prostheses and to keeping patients in hospital for longer after surgery. Assessments can then be made about whether more expensive prostheses or extended lengths of stay produce better patient outcomes, which justify the additional costs.

The average cost of hip replacement varies from A$10,000-30,000. XiXinXing/Shutterstock

We have analysed hospital data to compare costs, outcomes and the care pathways of patients treated for similar conditions at the main public hospitals in South Australia.

After adjusting for differences in the types of patients presenting at emergency departments with chest pain, seven in every 100 patients presenting at a particular hospital were readmitted or died within 12 months. This compared to four to five patients at the other hospitals.

The same hospital spent up to A$669 more on each patient than the other hospitals. Over one year, these additional costs amount to almost A$1 million.

Analysis of the care pathways showed that the hospital with the highest rates of re-admission, premature death and costs, discharged more patients from the emergency department. This hospital also kept patients who were admitted to an inpatient bed in hospital for longer than the other hospitals.

This suggests some patients may have been inappropriately discharged home from the emergency department, while other patients could have been discharged earlier.

Further investigation might look more closely at how and why decisions are made to admit patients from the emergency department and at what might be causing admitted patients to stay longer in hospital.

Investing in improvement

State governments are increasingly interested in improving quality. The Queensland government has set up an Integrated Care Innovation Fund to invest in initiatives to improve efficiency and value. NSW set up a similar Translational Research Grants Scheme. In South Australia, the Transforming Health initiative aims to improve the quality and consistency of health care across all metropolitan public hospitals.

But while individual efforts to improve quality may have some effect, it is more likely that co-ordinated, systematic approaches will have a greater impact.

Data should be analysed across hospitals on an ongoing basis to identify areas of clinical activity with the greatest potential for improvement, such as the examples above. Findings that quality could be improved should be fed back directly to hospitals.

Specialist teams should be set up to work with hospitals to further investigate areas of concern and to develop and implement improvement strategies.

Rather than going back to the drawing board on health reform, governments need to improve what we’ve already got and bring the poor performing hospitals and departments in line with their better performing peers.


This article is part of our series Hospitals in Australia. Click on the links below to read the other instalments: