Medical Errors and Adverse Events

A medical error is a failure of a planned action to be completed as intended, or the use of a wrong plan to achieve an aim; an adverse event is an injury caused by medical care rather than by the patient's underlying condition. Distinguishing errors, adverse events, and preventable harm is foundational to measuring and reducing the burden of harm from health care.

Definition

A medical error is the failure of a planned action to be completed as intended or the use of a wrong plan to achieve an aim; an adverse event is an injury resulting from medical management rather than from the underlying disease, a subset of which is preventable.

Scope

This topic covers the definitions and taxonomy of error and harm (including the distinction between error, adverse event, and near miss), the major categories such as diagnostic, medication, and surgical error, methods for detecting harm, and the epidemiology of in-hospital adverse events. It is a reference entry on how harm is defined and studied and does not provide clinical management instructions.

Core questions

• How are medical errors, adverse events, and near misses distinguished?
• Which adverse events are preventable, and how is preventability judged?
• What are the major categories of error (diagnostic, medication, surgical)?
• How is harm detected and measured across institutions?
• How common are in-hospital adverse events, and are rates changing?

Key concepts

• Error, near miss, and adverse event
• Preventable versus non-preventable harm
• Active failures and latent conditions
• Diagnostic, medication, and surgical error
• Slips, lapses, and mistakes
• Trigger tools and chart review for detection
• Underreporting of incidents

Key theories

Error versus harm distinction

Not every error reaches or harms a patient, and not every adverse event results from error; clear separation of error, near miss, and (preventable) adverse event is necessary to measure safety meaningfully and to target interventions.

Mechanisms

Errors are commonly classified by cognitive type — skill-based slips and lapses versus knowledge- and rule-based mistakes — and by the stage of care in which they occur. Harm is detected through voluntary incident reporting, which substantially underestimates frequency, and through more sensitive methods such as structured record review and trigger tools. Reason's framework links individual active failures to latent organizational conditions, explaining why error counts alone do not capture the system origins of harm.

Clinical relevance

Understanding how errors and adverse events are defined and measured underpins safety reporting and improvement throughout health care. This entry is a reference on the concepts and epidemiology of harm; it does not offer guidance for diagnosing or treating any individual patient.

Epidemiology

A systematic review of in-hospital adverse events found a median incidence of around 9% of hospital admissions, of which a substantial share were judged preventable (de Vries 2008). A study using a trigger tool in North Carolina hospitals found high rates of harm with little change over time, suggesting safety gains were not yet widespread by the late 2000s (Landrigan 2010). Diagnostic errors are increasingly recognized as a major and under-measured category (Singh 2012).

Evidence & guidelines

To Err Is Human (2000) catalyzed measurement of harm; systematic reviews (de Vries 2008) and trigger-tool studies (Landrigan 2010) quantified it; and reviews of diagnostic error (Singh 2012) extended attention beyond medication and procedural harm. Detection methodology is itself an active area, since reporting and record-review methods yield very different estimates.

History

Although iatrogenic harm was studied earlier (notably the Harvard Medical Practice Study of the early 1990s), systematic attention followed To Err Is Human (2000). Subsequent systematic reviews and trigger-tool studies refined incidence estimates, and diagnostic error emerged as a recognized frontier in the 2010s.

Debates

How preventable are adverse events?

Estimates of the preventable fraction vary widely with the detection method and reviewers' judgments, making it difficult to set firm targets or compare institutions reliably.

Key figures

• Lucian Leape
• James Reason
• Hardeep Singh
• Mark Graber
• Christopher Landrigan

Seminal works

• kohn-2000
• devries-2008
• landrigan-2010

Frequently asked questions

Is every adverse event caused by an error?

No. An adverse event is an injury caused by medical care; some adverse events result from errors and are potentially preventable, while others occur despite appropriate care.

Why is incident reporting an incomplete measure of harm?

Voluntary reporting captures only a fraction of events because many go unrecognized or unreported, so methods such as structured record review and trigger tools usually detect far more harm.