Safety by design: why wait for the harm to happen?

There is a growing interest in proactive approaches to safety management – putting systems in place to ensure safety, rather than responding when things go wrong. We asked Simon Dodds, a consultant surgeon who also runs SAASoft, a small company that designs hospital processes and systems, to explain the changing approach.

How can proactive safety management improve patient safety?

In healthcare, we tend to wait for the harm to happen. We record the number of times that something bad occurs – the ‘never events’ (things that should never happen) and other serious incidents, such as radiation, drug events and avoidable harm. These reactive approaches to safety and harm tend to lead to a a ‘check and correct’ design. They do reduce harm, but they create complex, slow and inefficient processes.

Proactive safety management, on the other hand, involves preventing the harm from happening in the first place. If we design and create systems to ensure that they are fundamentally safe, it’s much more efficient.

For example, when a patient goes in for an operation, they’ll have their name and address checked about eight times between arriving and the operation itself. At every handover, there’s a risk of an error creeping in.

With a safe design, we would check the patient’s name just once before surgery and know that we can trust that. It’s been estimated that something like 50% of NHS costs relate to people checking, correcting or ‘covering their backs’ – documenting what they do in case an error happens, just to show that they did what they were told to do.

How can healthcare process and system design help with that?

Hugely. When you do systems design you are always focusing on safety, flow and cost at the same time. It’s about creating a system where there aren’t any trade-offs between those three dimensions. I recently did an operational redesign for our trust pharmacy. I looked at it from a system design perspective, made recommendations for how to improve throughput, and we reduced the time patients waited for their take-home medication to be dispensed by 50% in one day. It isn’t always that quick, though – if we’re looking at 18-week performance, then the effect of a design change could take weeks to wash through.

We already have some reasonably good safety mitigation processes in place in the NHS. For example, in surgery we have the World Health Organization Safe Surgery checklist. But most of the 'niggles' in healthcare are operational design flaws, such as poor scheduling, operational policies that generate variation, excessive complexity, and multiple handoffs. This is why it is important to look at the way the whole system functions – including not only the parts, but the relationships between the parts.

What are the technical challenges in switching to proactive safety management?

Proactive safety management involves eliminating risks and measuring improvements as we go. So, if the harm is ‘avoidable death’, the precursor might be cardiac arrest. And the precursor to that might be a medication prescribing error, and so on. The further back we go along that causal chain, the more minor the factor will seem, but this is where we will identify the apparently minor issues – the ‘niggles’ that may play a key role in causing harm.

To measure the safety, we need to look not just at the factors, but at the time sequence in which they arise (‘time-series data’). With this type of data it’s not valid to use standard comparative statistical methods such as pie charts or averages, as they ignore the time sequence. The time-series statistics that are needed to interpret them are actually easier to use once we learn them.

Part of safe system design involves putting in a measuring system that alerts us if anything is a hazard or has a risk of causing harm. For example, software operating a ventilator will alert us before it fails, and will include a back-up battery and alarms, which are all part of safe system design.

The benefits are so clear, but I’ve yet to meet many people in healthcare who are trained in safe system design. I think the lack of investment in this training is a threat to the NHS. The opportunity is there to adapt the well-known approaches to healthcare in order to make it safer, quicker and more affordable. We have an ageing population and a fixed budget. If we don’t do something, we will come to a sticky end – and I believe it’s avoidable.