The AIHW's latest quarterly Quality Indicator report, released on 17 July and covering January to March 2026 across 2,509 residential care homes, shows hospitalisation moving in the wrong direction. Across 19 quarters, both measures within the indicator, ED presentations alone and ED presentations or hospital admissions combined, have risen to a statistically significant degree. It is the only adverse-outcome indicator going backwards, and sector commentary has pointed to rising resident acuity, reduced GP access into homes and changing clinical escalation practices as possible drivers, though none of these is established in the data itself.
The hospitalisation trend deserves the attention it is getting. But sitting inside the same report is a pattern that says more about the everyday reality of quality management, and it is worth reading closely if you are responsible for indicators that refuse to move.
Five years of data, two classes of indicator
The AIHW's trend analysis now spans 19 quarters, close to the entire life of the expanded QI Program, across effectively the whole sector. At that scale the results are not noise. And they split cleanly.
One group of indicators is improving. Pressure injuries, both measures of unplanned weight loss, polypharmacy and antipsychotic use have all declined to a statistically significant degree since July 2021.
The other group is stuck. Restrictive practices show no significant change over five years, despite sustained regulatory attention. And the falls data contains a detail that most of this week's coverage has passed over: falls resulting in major injury have declined significantly, but total falls have not shifted at all. The sector has become measurably better at reducing the harm caused by falls without becoming any better at preventing them.
If your organisation's falls or restrictive practices numbers have been flat for two years while your pressure injury results improve, you are not an outlier. You are the national pattern. That distinction matters, because a flat number on a stuck indicator is routinely read inside organisations, and by boards, as a local performance problem. Five years of near-census data says it is a property of the indicator, not of your team.
Why some indicators move and others don't
Look at what the improving indicators have in common. Pressure injuries, unplanned weight loss and medication management respond to defined clinical protocols: wound care regimes, nutrition monitoring, prescribing reviews. They have clear intervention points, established playbooks and a short line between action and result.
The stuck indicators are a different class of problem. Reducing falls, rather than just their severity, and reducing restrictive practices requires sustained change across the whole operation: workforce stability and skill mix, environmental design, behaviour support capability, clinical governance and culture. There is no single protocol to deploy. Progress comes from many small adjustments compounding over time, which means it depends on seeing what is working, and what isn't, while there is still time to adjust course.
Hospitalisation is the extreme case of the same pattern: the most multi-factor indicator of all, shaped partly by forces outside the provider's walls, and the one that has been quietly deteriorating for four years.
Our own polling, conducted with representatives from more than 150 provider organisations ahead of our upcoming Quality Indicators Briefing, suggests providers already feel this distinction. Asked which Quality Indicator is hardest to improve, respondents put Falls and Major Injury (29%) and Restrictive Practices (26%) well clear of Medication Management (19%), Unplanned Weight Loss (17%) and Pressure Injuries (9%). The sector's instinct about where the hard problems sit tracks the national trend lines almost exactly.
The compound problems demand a different relationship with data
Here is the practical consequence. For protocol-driven indicators, quarterly retrospective data is survivable: the playbook is known, and a four-month-old picture still tells you whether it is being followed. For compound indicators, it is not. If progress depends on continuous adjustment, then by the time a deteriorating trend appears in published data, covering a quarter that ended four months earlier, the window to investigate and adjust has already closed.
And that is where our polling points to the sector's real gap. These were engaged providers, organisations motivated enough to register for a Quality Indicators briefing, and presumably more data-mature than average. Even among this group, while almost nine in ten were confident their reported QI data accurately reflects the care they deliver, only 29% said they could identify a Quality Indicator heading in the wrong direction in real time. More than half (56%) typically spot emerging issues during the reporting quarter, 13% only once results are published, and 1% when it affects their Star Rating. Benchmarking is thinner still: just 23% can compare themselves against the industry in real time, 42% can do it 'with effort', and the rest rely on published government data or cannot meaningfully compare at all. Across the wider sector, those numbers are unlikely to be better.
The hospitalisation trend shows what that gap costs at national scale. It has been building quarter on quarter since 2022. Providers with real-time visibility of their own ED presentations had the opportunity to see their contribution early and act on it. Most of the sector is reading about it this month.
The next phase of QI improvement is the hard part
The national data suggests the sector has largely banked the gains available from protocol-driven improvement. What remains, falls, restrictive practices, hospitalisation, is the harder class of problem, and it is also the class that most directly shapes residents' daily safety and freedom.
Quality Indicators have outgrown their origins as a compliance obligation. They inform governance, influence Star Ratings and increasingly function as operational performance measures. Accuracy, our polling suggests, is largely solved. The opportunity now is timeliness: identifying variation as it emerges, benchmarking in real time rather than with effort, and acting inside the quarter rather than explaining after it. For the indicators that are stuck, that is not a nice-to-have. It is the mechanism of improvement.
