How Failures Happen
The Anatomy of Harm

The Swiss cheese model
of organizational failure

The most influential model for understanding how healthcare failures occur was developed by British psychologist James Reason in the early 1990s. Known as the Swiss cheese model of organizational accidents, it offers a framework that has shaped patient safety thinking for three decades.

Reason's insight was this: organizations build multiple layers of defense against failure — policies, procedures, training, equipment checks, supervision, and technology. Each layer is intended to prevent harm. But no layer is perfect. Each has gaps — holes — through which a failure can potentially pass.

Harm occurs when the holes in multiple layers momentarily align — when the gaps line up — allowing a failure to travel all the way through every layer of defense and reach the patient. In isolation, a hole in one layer does not cause harm. The defenses around it catch the problem before it reaches the patient.

Visualize a stack of Swiss cheese slices. Each slice is a layer of defense. Each hole is a weakness in that layer. Most of the time, the slices are misaligned and the holes do not line up. Occasionally, they do. That is when harm occurs.

Active failures and
latent conditions

Reason also drew a critical distinction between two types of contributing factors in healthcare failures: active failures and latent conditions.

Active failures are the unsafe acts committed by people who are in direct contact with the patient or system at the point of care. They are the visible triggers — the wrong drug drawn, the missed vital sign, the misread label. Active failures are what investigations most readily find, and what person-focused responses most commonly target.

Latent conditions are the hidden vulnerabilities embedded in the system long before any individual error occurs. They arise from decisions made weeks, months, or years earlier — design choices, resource allocations, staffing decisions, equipment procurement, policy gaps, and cultural norms that create conditions in which errors are more likely to happen.

Latent conditions are far more dangerous than active failures, precisely because they are invisible until activated. They sit quietly in the system, waiting for the right combination of circumstances. A corridor where look-alike medications are stored together. A monitoring system with too many alerts, so staff learn to override them. A handover process that evolved through informal workarounds into something fundamentally different from the written protocol.

Human factors
and the limits of human performance

Human factors is the scientific discipline concerned with understanding the interaction between humans and the systems, environments, and tools they work with. In healthcare, human factors examines how cognitive, physical, and organizational factors influence the ability of clinical staff to perform safely and effectively.

The fundamental insight of human factors science is that humans have predictable limitations. Working memory is finite. Attention is selective. Performance degrades under fatigue, time pressure, noise, and interruption. These are not character flaws — they are universal features of human cognition. Safe system design accounts for these limitations rather than assuming they can be overcome through willpower or training.

Normalization of deviance — a concept introduced by sociologist Diane Vaughan — describes the gradual process by which unsafe practices become accepted as normal through repeated occurrence without immediate consequence. In healthcare, it appears as workarounds that become standard practice, protocols that are routinely skipped because they seem unnecessary, and warning signs that are habitually dismissed because they have never caused harm before.

Key concepts
from this lesson