Researchers in Europe may have defined and given structure to the challenges associated with safety. Their white paper on Safety I vs. Safety II highlights where scientific methods, standardization, root cause analysis and reliability alone fall short. They assert that safety needs to progress beyond a bimodal model to employ multiple approaches for reducing error/failure and achieving fully resilient systems.

Safety I, as experienced today, is the push to eliminate all accidents, system failures and injury. The researchers define this as the "state where as few things as possible go wrong" with involved components — technology, procedures, humans and the organization in which they operate. This makes sense for intractable and well-engineered systems. However, Safety I is really just the starting point for healthcare; which by its very nature is intractable.

Safety II recognizes that medicine and healthcare processes are not an exact science and that there is sometimes an art to caregiving.

Going beyond Safety I involves a focus on what went right; usually the adjustments made in human performance to match the conditions of work. Below is a listing of key concepts from the whitepaper reinforcing the understanding that variability and flexibility and the need to ensure as many things as possible go right.

  1. It is false to assume good outcomes are simply due to systems functioning as they should, because people worked as imagined (an idealized view).
  2. The world and healthcare have become more complex; outcomes can no longer be viewed simply along linear models of cause-effect relations.
  3. Instead, task performance must adjust to constantly changing conditions of the work and environment. When this occurs, the work-as-done may be very different from the work-as-imagined and delineated in policy, protocol, guidelines, rules, checklists, etc.
  4. "In the normal course of clinical work, doctors, nurses, and all allied staff perform safely because they are able to adjust their work so that it matches the conditions."
  5. Healthcare leaders should acknowledge that things often go right, because clinicians adjusted their work to conditions, instead of working as imagined. In these cases, outcomes are emergent, not resultant.
  6. Eliminating variability and pushing the wrong highly reliable systems could increase the occurrence of poor outcomes.
  7. Safety management must become more proactive and promote interventions to prevent the unintended consequences of policy, protocol, guidelines, rules, checklists, etc.
  8. Safety II is based on the principle that performance adjustments are ubiquitous and that performance not only always is variable, but it must be so.
  9. Policy makers, regulators and healthcare leaders should recognize that they are removed in time and space from the actual operation of the systems and services. And, they have limited opportunities to observe or experience how work (caregiving) is actually done.

Thus, managers and leaders have an opportunity to recognize the important role of variability and flexibility in safety. This is accomplished when they:

  • Make the effort to truly consider how work is done,
  • provide the necessary resources and
  • prepare for the unexpected.