• Thousands of patient misidentification incidents each year, including hundreds that cause harm
  • Most harm at a low level, but data also reveals two deaths in five years
  • Charity says “misidentifying a patient should be a basic error for which there is no place in a modern health care setting”

Hundreds of patients are being harmed each year because NHS organisations have incorrectly identified who they are, an HSJ investigation has found.

Responses to Freedom of Information requests from 166 trusts revealed 58,537 cases of patient misidentification logged in Datix or other patient safety systems between 2019 and 2023, including 4,713 causing some sort of harm. This is equivalent to 11,707 incidents a year, with 943 leading to harm.

It includes a wide range of errors, but harm typically happens when patients are given the wrong treatment or medicine, or miss out on the right treatment, as a result of errors in recording and/or miscommunication. A typical example is patients being given a wristband with the wrong name – or ID information resulting in patients not being treated, or the wrong treatment being given.

Some of the worst examples are where the wrong patient, or the wrong part of the body, is operated on.

Both the total number of incidents and incidents with harm were highest in 2023 – at 12,482 and 1,161 respectively – having been relatively high in 2019, before dipping between 2020 and 2022.

Although most of the incidents with harm were low harm, HSJ also found 250 cases of moderate harm, 10 cases of severe harm and two deaths across the five years. In 2023, there were 59 incidents involving moderate harm, three with severe harm and one death.

Acute trusts reported most of the incidents, both in total and those causing harm.

Earlier this year, a Health Services Safety Investigation Body report warned “patient misidentification is challenging to address and previous efforts to reduce the risk have not been as successful as hoped”.

Action against Medical Accidents chief executive Paul Whiteing said: “These details are a real cause for concern and are at a level and scale that suggests there is need for an urgent system-wide review as these errors are avoidable. And as the data shows, at their worst they can result in the death of patients, which is unacceptable.

“We understand that with high-risk medicine things can sometimes go wrong. But misidentifying a patient should be a basic error for which there is no place in a modern healthcare setting.”

Patients Association chief executive Rachel Power said: “It is concerning to see rising levels of patient misidentification, given what a serious risk to patient safety it is to give or not receive treatment intended for a particular patient.

“Like the prevention of never events, there should be systems in place in every hospital or clinic that proactively ensure patients are correctly identified.”

Patient Safety Learning CEO Helen Hughes called on NHS England to review cases nationally to identify root causes.

She added: “Where avoidable harm has occurred, it is vital these incidents are investigated, that causes and contributory factors are identified, and steps put in place to prevent their reoccurrence. In some cases, this may require standardisation of approaches to patient identification, while others may highlight contributory factors that are more difficult to address, such as staff fatigue.”

An NHSE spokesperson said: “Patient misidentification – including communications being sent to the wrong person – can represent clear healthcare risks and, while the vast majority of these incidents result in no harm, it is right that they are increasingly being recorded on local management systems so that steps can be taken by providers to reduce risks.

“The NHS offers expert advice and guidance to healthcare organisations on how to strengthen systems, including through implementation of barcode technology to match patients with their treatments, medications and surgical procedures .”

HSJ sent freedom of information requests to all English NHS trusts and received substantive responses from 166. The remainder did not respond, said they could not provide the information, or provided the information in a way that could not be easily compared to other trusts.

Some trusts indicated they had “fewer than five” events in certain years rather than providing an exact number. Where this has happened, and HSJ felt confident at least one incident had occurred, HSJ counted this as one instance.

Where trusts have provided figures but not indicated if there was harm, HSJ has assumed there wasn’t. Where trusts have indicated there were events with harm but not indicated what category, HSJ has assumed “low”.

Some trusts reported cases described as near misses, which HSJ has included in the total. 

Patient misidentification is not an officially defined term. Because of this, trusts may have interpreted HSJ’s question in different ways. Where trusts offered examples of type of incident they recorded, a common example given was incorrect information on a patient’s wristband. Another explanation which we provided to trusts when asked was “a patient being sent a communication meant for somebody else or a patient being given a procedure, medication or test results meant for a different patient”.