The Expert:

Dr Carl Horsley

Dr Carl Horsley is a dual trained intensivist who is the current clinical head of the Critical Care Complex at Middlemore Hospital. He is a member of the Resilient Healthcare Network and recently co-authored a chapter on a project he introduced to enhance the ability of staff to adapt to changing situations within the Complex. This approach sees people as the key resource required to deal with the complexity of modern healthcare and has led to improved staff engagement, improved quality care and a more patient centred view.

Their View:

Despite the promises following the 1999 Institute of Medicine report To Err Is Human, the rates of patient harm as a result of their care continue to remain high. Recent reports have even postulated that healthcare related harm is third only to heart disease and cancer as the leading causes of death in the developed world.

It’s not only patients who are suffering; healthcare workers involved in bad patient outcomes frequently experience second victim effects with a profound challenge to their sense of professional identity.

The current aim of safety is that as few things as possible go wrong, and our current model of safety clearly isn’t working.

Safety-I: The current view of safety
The current view of safety is that good outcomes result from normal functioning and compliance, and bad outcomes result from malfunctioning or noncompliance.[1] Therefore we only need to focus on when something goes wrong; to find the root cause of the bad outcome and prevent it from happening again.

We then design our healthcare systems to focus on fixing the problems, ensuring compliance and erecting more and more safeguards to prevent things from going wrong. As a result, healthcare systems become progressively more regulated, restricted and potentially complex. We treat healthcare like a linear and predictable factory and the people in it, both staff and patients, as sources of variability that prevent safe care.

The problem with this approach to safety is that it’s retrospective, reactive and biased. It fails to helps us learn what happens in our system when things go right despite difficulties or how to deal with new safety problems that emerge in the future. It creates brittleness because we don’t see the sources of adaptability and innovation. It also creates a conflict between safety and productivity, because conformity to the procedures and checklists designed to ensure safety is time-consuming. Finally, it exacerbates burnout by asking staff to be passive ‘robots’ and to focus on compliance rather than care.

The reality is that healthcare is a complex adaptive system which is constantly changing and thus requires constant adjustment and adaptation. In the real work environment, staff constantly have to make trade-offs between efficiency and thoroughness. This reality requires a different view of safety.

Safety-II: The new view
The ability to juggle complex dynamic situations is how we achieve normal success; it is also how things sometimes go wrong. Safety-II recognises that both good and bad outcomes stem from the same performance adjustments people make every day to get their jobs done.[1] No action is intrinsically good or bad – the outcome depends on whether the action suits the context.

In this model of safety, the system only succeeds because people can adjust to meet the conditions of work. Complexity is the problem and people are the solution.

The new aim of safety is for as many things as possible to go right. It is proactive and seeks to understand how things usually go right to identify the difference in context when they occasionally go wrong.

Lessons for the way we work
1. Make it easier to do the right thing. Success and failure come from the same source, so aim to make success more likely, not to make failure less likely.

2. Learn from all events. Think about why things went right as well as why things went wrong. In the CM Health Critical Care Complex, we ask four questions after every intervention:

  • What went as planned?
  • What were the surprises?
  • What did we have to work around and make-do?
  • What did we miss?

This brief reflection on why things worked despite being difficult ensures we learn the value of things that helped interventions go well and highlight anything with potential to cause a problem in the future.

When you look at safety incidents, change the question from ‘Why didn’t you follow the rules?’ to ‘Why did that seem the right thing to do at the time?’ If you don’t understand the context which made a choice seem the right one, someone else could make the same choice again.

3. Understand work-as-done. Design safety interventions to fit the reality of the work environment and how it operates.

4. Build resilient teams and systems. Resilience is the ability of a team or system to monitor and adjust performance to achieve its goals, even when the unexpected happens. Building resilience involves four steps:

  • Anticipation – knowing what to expect
  • Monitoring – knowing what to look for
  • Response – knowing what to do
  • Learning – knowing what has happened

This means moving from the idea of a single all-knowing expert to a team-based approach that recognises the value of bringing many different perspectives to bear on a problem.

It also means ensuring staff feel safe to speak up and ask questions. Psychological safety is the key determinant of performance for teams working in high levels of uncertainty and interdependence.

A change of perspective
Safety-II is a change of perspective. It changes the way we design our systems, build our teams and look at ourselves. Let’s create a safety system that works in the real world to keep our staff and patients safe.

1. Hollnagel E. Safety-I and Safety-II: the past and future of safety management. CRC Press; 2014.


This article is adapted from Dr Horsley’s Grand Round presentation for Patient Safety Week.

Last modified: