Surgical Never Event Statistics (2026): Incidence, Anesthesia, and Prevention

Never events are the medical errors that are not supposed to be possible: operating on the wrong site, leaving an object inside a patient, or a healthy patient dying under anesthesia. They are rare, clearly defined, and almost entirely preventable, which is exactly why they carry such weight in patient safety and litigation. For sedation providers, one never event category is directly on point: the death of a healthy patient under anesthesia.

What's in This Guide

1What Counts as a Never Event

The term has a precise definition, which is what makes never events measurable and accountable in a way that most adverse events are not. Understanding the definition is the starting point for understanding the data.

The term "never event" was introduced in 2001 by Dr. Ken Kizer, then head of the National Quality Forum, to describe errors so serious they should never happen. The NQF's list of serious reportable events now includes 29 adverse events across seven categories, and the Joint Commission classifies these same events as sentinel events. To qualify, an event must be unambiguous, meaning clearly identifiable and measurable; serious, meaning it results in death or significant disability; and usually preventable. The surgical category alone includes wrong-site surgery, wrong-patient surgery, wrong procedure, unintended retention of a foreign object, and intraoperative or immediately postoperative death in a healthy patient.

This precision matters because it separates never events from the broader universe of adverse outcomes. A poor result from a properly performed procedure is not a never event; operating on the wrong knee is. The clarity of the definition is what allows these events to be tracked, reported, and, since 2009, excluded from Medicare reimbursement. It also makes them powerful in litigation, because a never event is, almost by definition, evidence that something preventable went wrong.

Source: Never Events primer, AHRQ Patient Safety Network | NQF Serious Reportable Events list (National Quality Forum)

2How Often They Happen

Never events are, by design, rare, but rare is not the same as never, which is the entire point. The incidence data shows events that are individually uncommon but collectively persistent.

A 2013 study estimated that more than 4,000 surgical never events occur each year in the United States, even though any single hospital may experience a specific event like wrong-site surgery only once every 5 to 10 years. The Joint Commission has received roughly 800 sentinel event reports per year since 2007, with wrong-site surgery, foreign body retention, and falls consistently among the most common. Recent Joint Commission data has shown these events not declining but in some years rising: wrong surgery increased 26 percent and unintended retention 11 percent from 2022 to 2023, and retained objects rose for a fourth consecutive year through 2024.

The persistence is the story. Despite decades of concerted effort, checklists, protocols, and public reporting, never events have not been eliminated, and some categories have recently increased. This does not mean the safety systems have failed; it may partly reflect better reporting. But it confirms that these events remain a live risk in every setting where invasive procedures and anesthesia are performed, including office-based settings, and that vigilance rather than complacency is warranted.

Source: Never event incidence, AHRQ PSNet | 2023 Joint Commission sentinel event increases, American College of Surgeons

3The Malpractice Claim Data

Because never events are unambiguous and serious, they translate readily into malpractice claims. The largest analysis of never-event claims comes from the National Practitioner Data Bank and reveals which events dominate litigation.

A peer-reviewed analysis of 9,744 paid malpractice claims involving surgical never events reported to the National Practitioner Data Bank found that retained foreign body was the most common at 49.8 percent, followed by wrong procedure at 25.1 percent, wrong-site surgery at 24.8 percent, and wrong-patient surgery at 0.3 percent. These four events account for nearly all surgical never-event claims, and their relative frequency has been stable across studies. Retained objects lead in part because they can go undetected until symptoms appear, sometimes long after the procedure, extending both the harm and the liability window.

The claim data underscores how consequential these events are. A California study of never events found that roughly two-thirds of retained-object cases required additional surgery to correct, and regulators mandated an average of 13 corrective actions per incident, most focused on policy-adherence monitoring. Never events do not merely generate a claim; they trigger corrective mandates, regulatory scrutiny, and often a second procedure, compounding the harm to the patient and the exposure to the provider.

The dental and oral-surgery relevance is direct. Retained objects in dental and oral surgery have included dental retractor cords and instrument fragments, and wrong-site events include operating on or extracting the wrong tooth, one of the more commonly reported dental never events. Any invasive procedure carries never-event exposure, and office-based dental and oral surgery settings are not exempt from the verification and documentation disciplines developed in hospital surgery. The same protocols that prevent wrong-site surgery in an operating room, verification, marking, and a documented time-out, apply to the dental operatory.

Source: Mehtsun et al., Surgical never events in the United States, Surgery (NPDB analysis), PubMed

4The Anesthesia Never Event

Among the 29 never events, one is directly relevant to every sedation and anesthesia provider. It is defined narrowly and specifically, and it places anesthesia safety at the center of never-event prevention.

The National Quality Forum specifically defines "intraoperative or immediately postoperative/post-procedure death in an ASA Class I patient" as a never event. An ASA Class I patient is a normal, healthy person, and the NQF specification includes all such deaths where anesthesia was administered, whether or not the planned procedure was completed, with "immediately postoperative" meaning within 24 hours. In plain terms: when a healthy patient dies under or shortly after anesthesia, it is classified as a preventable event that should never have happened.

This is the never event most relevant to office-based sedation. A patient presenting for elective dental or oral surgery under sedation is frequently an ASA Class I or II patient, healthy apart from the dental issue. Their death under sedation falls squarely within the category the NQF considers a never event. This framing raises the stakes for sedation monitoring: it is not merely that a sedation death is a tragedy and a potential malpractice claim, but that it is formally classified as an event that proper care should prevent. The anesthesia never event is, in effect, the patient-safety system's judgment that a healthy patient should not die from sedation, and that when it happens, prevention failed. The underlying incidence of these deaths is covered in our review of dental sedation death statistics.

Source: NQF anesthesia never event definition, AHRQ PSNet | NQF event 1E, ASA Class I patient death specification

5The Reporting Gap

A crucial caveat runs through all never-event data: the numbers almost certainly undercount the true incidence, because most reporting is voluntary. Understanding this gap is essential to reading the statistics honestly.

Because reporting sentinel events to the Joint Commission is voluntary, the agency itself cautions that no conclusions should be drawn about the true relative frequency of events. The scale of the gap is illustrated by a striking comparison: when New York State used mandatory reporting, it received far more wrong-site surgery reports in two years than the Joint Commission received nationally over a period three times as long, a discrepancy suggesting voluntary reporting may underestimate true incidence by a factor of 20 or more. Historically, only about two-thirds of sentinel events were self-reported by the institutions involved, with the rest surfacing through patient complaints or media.

The reporting gap cuts in two directions for interpretation. It means the roughly 4,000 annual figure and the Joint Commission's counts are floors, not ceilings, the true numbers are higher. But it also means that year-to-year changes in reported events may reflect reporting behavior as much as actual incidence. The honest reading is that never events are more common than the official tallies show, and that the data is best used to understand patterns and drivers rather than precise rates. For a provider, the practical implication is that these events are a real and somewhat underappreciated risk, not a negligible one.

Source: Sentinel Event reporting and underreporting, StatPearls (NCBI)

6Prevention and the Record

Never events are, by their own definition, usually preventable, which means the prevention toolkit is well established. The tools share a common thread: structured verification and documentation at every critical step.

The cornerstone is the Joint Commission's Universal Protocol, which requires three steps before every procedure: pre-procedure verification of the patient, procedure, and site; marking of the operative site; and a time-out immediately before starting, during which the entire team confirms the patient's identity, the correct site, and the correct procedure. Structured surgical safety checklists extend this verification across the perioperative period. For anesthesia specifically, the same time-out applies, and the American Society of Anesthesiologists emphasizes that anesthesia providers actively participate in it, along with continuous physiologic monitoring and careful medication verification, since wrong-medication and wrong-dose errors are their own never-event risks. Each of these safeguards produces a documentable checkpoint. The ASA also recommends a separate anesthesia consent as part of this verification discipline, a point we cover in our review of informed consent litigation statistics in dentistry.

This is where documentation becomes inseparable from prevention. A time-out that is performed but not recorded, or monitoring that occurs but is not captured, leaves no evidence that the safeguard was in place, which matters both for continuous safety and for defensibility if an event occurs. The way missing monitoring documentation drives outcomes in litigation is examined in our review of standard of care violation statistics in sedation cases. This is where iSedate's SedationVault fits the never-event framework for sedation providers. By pulling vitals directly from compatible monitors such as Edan, MindRay, and Criticare, timestamping every entry, and producing an audit-ready PDF, SedationVault captures the continuous monitoring that guards against the anesthesia never event and documents that it was performed. It does not conduct the time-out, replace the verification steps, or substitute for clinical judgment and rescue readiness. What it does is ensure the monitoring safeguard against a healthy patient's death under sedation is both active and provable.

Source: Universal Protocol for preventing wrong-site surgery, AHRQ PSNet

Every Statistic in One Table

Frequently Asked Questions