PATIENT SAFETY NOW

Incident reporting systems were originally intended to help organisations identify safety events and learn from them. Over time, they’ve become the default route for almost every organisational frustration. That’s not because staff are using them wrongly, it’s because, in many organisations, they’re the only mechanism that is available to them. People use incident reporting systems as a communication tool because the communication systems have failed.

Choosing the right route

Instead of reporting via your incident reporting system how about asking ‘what am I trying to achieve?’ Am I trying to:

• alert someone today
• record harm
• improve a process
• report abusive behaviour
• identify an emergent pattern
• count something

All of these should lead you to a different place.

• Learn about risk, potential harm and harm = incident reporting system
• Count activity or comply with data expectations = dashboards, audits, spreadsheets
• Escalate operational problems = line management, huddles, briefings and debriefings, escalation processes
• Raise concerns about behaviour = speak to the person, speak up to someone else, talk to HR
• Detecting early warning signs = safety huddles, observations, conversations with staff, video reflexive ethnography

First: we need to separate counting from learning

I have mentioned this many times. If the organisation wants to know how many falls, pressure ulcers, delayed medicines occur then put them in a spreadsheet. They are important but require surveillance. Spend your time talking to the patient, and managing the patient’s condition rather than filling out a form.

Second: Have a system for reducing the multiple reports

One of the issues is that the report might be a ‘one off’ for the individual staff member but for every one report there are likely to be many staff writing a report about the same thing. So the organisational response should be to detect that and send out a message that says ‘we are receiving multiple reports on a particular issue, you can stop reporting it now and we will get on and work on the issue’.

Third: Don’t add to the already overwhelmed reporting systems

The easy response when a staff member wants to talk about an issue is often, ‘got a problem, report it’. Near misses, report it. Good practice, report it. Pointless to add to a system that is already not working.

As I mentioned in a recent blog, we need to detect early warning signs. But we dont need to add them to the reporting system. There is a need to create a completely separate approach to early earning signs. A better way to capture the times when people are constantly working around something or catching something at the last minute. The last thing we want to do is use an incident reporting system for these – they will be lost. It would be much better to have structured debriefs, routine observation, conversations with staff.

Safety intelligence

The safest organisations are not necessarily the ones that collect the most reports. Instead of talking about incident reporting systems, let’s reframe what we are trying to achieve as safety intelligence. There are numerous existing ways to gather safety intelligence, for example:

• incidents (these should be the rare but important)
• huddles that can take just 5 minutes
• handovers, meetings, briefings and debriefings
• conversations with patients and patient stories
• conversations with staff and staff stories
• staff concerns
• operational data and other data requirements
• complaints
• compliments
• safety conversations
• morbidity and mortality meetings
• VRE findings
• safety walk rounds
• leadership walk rounds
• new staff interviews
• leaving interviews
• end of day handovers – with ‘did anything worry you today, did anything get in the way of you doing your job today?’ – 5 minutes, no forms, just a conversation

An incident report becomes one source of safety intelligence rather than the organising principle.

Critical design principle

I am absolutely not suggesting people report more but to choose the right way to communicate and learn. If any redesign requires people to do more it will fail before it begins.

We don’t need another reporting system it needs a more effective use of the mechanisms we already have. Some the design questions are:

• how do we make sense of all the above?
• how do we make it easier for people and organisations to learn?
• how do we maximise organisational attention while minimising the burden on staff?

At the moment people spend huge amounts of time entering information that ultimately does not seem to help. I am no digital expert but there will be people out there who are probably already working on this and there will be others who are starting to use AI as a tool to help. AI can be used to cluster similar reports, identify recurring themes, detect the weak signals, spot the emerging risks and summarise thousands of reports (in a second). The humans then make sense of the patterns and summaries provided.

A new foundation for the new way of thinking

We are not learning if our understanding of safety comes from a system that is overwhelmed with millions of incident reports that are barely glanced

We are not learning if our understanding of safety comes purely from failure

We are not learning if we fail to capture early warning signals

We are not learning if we rely on lots of retrospective investigations and inquiries

The science of safety has been slowly growing for the last 3 to 4 decades. We are reaching an important stage of understanding that there are different ways we should be looking at safety. Our safety intelligence has not kept pace with this emerging safety thinking.

Maybe we need a new foundation that underpins this new way of thinking.