It Worked While We Were Watching, Sustainable Improvement in Social Care.

I once spent about six months in a home with my sister (we worked together!) trying to turn things around. It was one of the hardest periods I can remember. The situations were complex, the pressure was constant, and there were weeks where we were running on a few hours’ sleep because something always needed sorting and none of it could wait.

We worked unbelievably hard. And by the time we left, the home was in a much better place. More consistency. Better oversight. Clearer expectations. Far less of the constant firefighting we’d walked into.

A few months later, we went back. A lot of what we’d put in place had slipped.

I can still remember how that felt. After six months of pouring everything into that service, it was honestly soul-destroying to see things starting to unravel again.

At the time, I thought it meant we’d failed. Looking back, I see it differently.

We had improved the service, but we were also part of what was holding that improvement in place.

This is something we don’t talk about much in turnaround work. When a service is under close scrutiny, everyone becomes laser-focused. Actions get chased. Expectations are clearer. Senior oversight increases and there’s far less tolerance for drift.

That is exactly what’s needed in those situations, but the problem is when those standards only hold because that level of attention is still there.

I’m always wary of improvement plans that look impressive because everything is marked “completed”. Training delivered, competencies checked, care plans updated — all good things, but none of them tell me whether the original issue has actually gone away.

I want to know whether practice changed afterwards. Whether the same issue popped up somewhere else. Whether the improvement was still there months later.

Otherwise, we end up being great at completing actions and not quite as good at checking whether they worked.

I’ve seen this play out in so many ways. A manager holds a service together because they’re constantly checking and chasing. Staff work to a higher standard because they know there’s extra scrutiny. Governance meetings become sharp and focused because everyone knows the service is under pressure. Then the attention reduces, normal life resumes, and you start to see which parts genuinely became part of the culture and which parts were being carried by the people around it.

That, for me, is where the real test starts. Not when everything looks better, but when the extra support steps back. Can the manager still spot the issue without someone pointing it out? Are staff still working to the same standard when nobody is checking as often? Does the service notice something slipping before it becomes a bigger problem?

That’s the hard bit. Sustainable improvement isn’t just about putting things right, it’s about making sure the service can keep them right once the pressure reduces.

And I think we sometimes pull support away too quickly because things look better and everyone is keen to move on.

I’d rather step support down gradually and watch what happens. If something was reviewed weekly, move it to fortnightly. If senior leaders have been chairing every improvement meeting, hand it back to the manager. If every incident has had central scrutiny, reduce it and sample instead.

You’re not trying to catch anyone out. You’re checking whether the improvement can stand on its own. If it can’t, that’s important to know. Maybe support came away too quickly. Maybe the original action didn’t address the real issue. Maybe the service still needs more leadership, more clinical oversight, or clearer accountability.

None of that is helpful if you only discover it six months later when the same issue reappears. You fix wound care, then nutrition starts slipping. Medication improves, but escalation is still weak. Care planning gets better, but follow-up is patchy.

That’s when I start wondering whether the original problem was ever really about wounds, medication or care plans. Sometimes the real issue is much more basic: weak oversight, poor challenge, lack of confidence, unclear responsibility, or people not following things through.

That’s why improvement work needs a proper look-back — thirty days, sixty days, ninety days, whatever makes sense. Not for the sake of another process, but because we need to know whether what we did actually lasted

That home taught me that the hard way.

We did make a difference while we were there. I don’t doubt that. But we hadn’t yet built enough around the service for that difference to survive without us.

That’s the bit I pay far more attention to now. For me, sustainable improvement in social care is not about how good things look while everyone is focused on the service. It is about whether those standards still hold once the extra support steps back.

Because that’s usually when you find out whether the improvement truly belongs to the service, or whether it still belongs to the people who drove it.