Swiss Cheese Model and the Hot Cheese Model

A yellow colored slice of Emmental Swiss cheese

The Swiss Cheese Model and the Hot Cheese Model revolutionize the way we traditionally think of human error. In short, both models force us to move away from “blaming someone” for human error; instead, it pushes us to look at human error as a consequence, not as a cause. In fact, the Hot Cheese Model is an improvement to the classic Swiss Cheese Model.

About Swiss cheese

Following is a special type of Swiss cheese slice. It has holes in it. Among many, this is only one variety of Swiss cheese. This variety is from the Emmental area of Switzerland and also known as Alpine cheese. It is yellow, medium-hard, and therefore not flexible.

A yellow colored slice of Emmental Swiss cheese

A slice of Emmental Swiss cheese

Source: Emmental Swiss Cheese Slice; National Cancer Institute; Renee Comet (photographer): Wikimedia Commons on the public domain

About the Swiss Cheese Model

In 2000, James Reason, a professor in Psychology at the University of Manchester introduced this model. He published it in British Medical Journal under the title, “human error: models and management”. This is a “mental model” that helps us to visualize human errors in healthcare through a “system lens” instead of a “moral lens”. As a result, it forces us to move away from the emotional trap of the “blame game”.

James Reason’s Swiss Cheese model: Source: BMJ 2000 March 18.

This is one of my favorite conceptual models that helps to discuss behavior change options without prejudice. I have used it before as a community physician at a private hospital’s family and corporate health program in Sri Lanka in 2014 after my retirement from the government service.

Swiss Cheese Model in patient safety

Swiss Cheese Model in patient safety newsletter

Following are the James Reason’s thoughts about it;

Do not see human error as a moral problem

Think of deaths due to COVID-19. We, particularly those who hold authority, can easily find faults with the “general public” for not adhering to the recommendations; on the other hand, others can find faults with health officials, decision-makers, and politicians. We always fall into this trap of making it personal because blaming someone is emotionally satisfying.

Blaming individuals is emotionally more satisfying than targeting institutions.

James reason; “hUman error: models and management”, 2000, British medical journal

Instead, look at human error as a consequence, not as a cause.

James Reason conceptualizes that “humans are fallible and errors are to be expected even in the best organization system”. As a result, we need to see human errors as consequences, not as causes. In this paradigm shift, we do not have any room for the blame – game. In other words, this is an “upstream” journey; those who hold power cannot pass the ball down. Everyone has to bear equal responsibility and accountability.

This is the basic premise of this system approach. In fact, the systems approach forces us to think of humans as only a cog of a larger wheel of systems. So, here a human error is not seen as a failure but a variability. The focus is on the conditions or environments under which they operate or behave.

Swiss cheese model’s systems approach

James visualizes this system approach in layers: Some are engineered such as alarms, physical barriers, and automatic shutdowns, etc; others are human barriers such as physicians, pilots, control room operators in the aviation industry. Procedures and admin controls are other layers.

Ideally, any layer should not have holes; but in reality, there are many. And, the holes are dynamic; it constantly changes in size, shape, and shifts locations. Having holes in one or a few layers does not necessarily result in bad outcomes or disasters. Instead holes in each layer need to be aligned together for a disaster to happen. When a disaster occurs, according to the model, one cannot look at only one layer; we need to examine all the layers to prevent it from happening again.

Active failures and Latent failures

He grouped the holes into two:

  • Active failures
  • Latent

The active “holes” consist of slips, lapses, fumbles, mistakes, and procedural violations. For example, administering the wrong drug is an active failure.

Sara Garfield and Bryony Dean Franklin in the Pharmaceutical Journal in 2016 describes those active holes drawing example situations from the pharmacy field;

SlipIntended to dispense one drug but dispense another similar to the prescribed one
Lapseintended to dispense the second drug but forgot.
Violation dispensed a controlled drug that did not meet the handwriting requirement
Mistakedispensed the wrong form of the prescribed drug due to lack of knowledge

The Latent ones are the “holes” in the management decisions that lead to understaffing, time pressure, lack of facilities, unworkable procedures, etc. As its name suggests the latent conditions lie dormant years in the system. He used another analogy here; active holes are like mosquitoes; although we can remove them one by one they keep coming. The Latent ones refer to the breeding places we can remove.

However, this model contain several limitations.

Hot Cheese Model: An improved Swiss Cheese Model

Li Y and Thimbleby H. in 2014 introduced several improvements to the James’ Swiss Cheese Model and named it has the Hot Cheese Model. This model addresses the limitations of the classic model. As the name implies, this Hot Cheese Model is hot, fragile, flexible, and new holes can form with time. in contrast to the classic model, this model emphasizes that adding more layers does not necessarily bring more protection. It might add new holes to the system.

A comparison between the classic Swiss Cheese Model and the Hot Cheese Model

Following are the key differences between the classic Swiss Cheese Model and the Hot Cheese Model.

  • The original model sees managerial, policy, and procedural actions as latent ones. The Hot Cheese Model argues these layers are also active, not latent.
  • The original model implies more layers minimize the chances of error. The Hot Cheese Model argues more layers can also introduce more errors (drips) in addition to the errors that pass through from previous layers.
  • The original model does not accommodate the shape-changing nature of a loophole. The Hot Cheese Model recognizes the hole’s shape can change with time. For example, those involved may find method short-cuts with time.

Further reading


Author: Prasantha De Silva

A specialist in Community Medicine board-certified in Sri Lanka and a research analyst in Canada

Leave a Reply

Your email address will not be published.