Which is the odd one out: a supermarket tin of baked beans, an airport suitcase or a hospital patient? The answer is the tin of beans, which is about 1,000 times less likely to be wrongly identified and mishandled than the suitcase or the patient, for each of which the risk is about 1 in 100. It is for this reason that the Hospital Authority is testing a supermarket style bar-code labelling system to help reduce mix-ups of patients' samples and specimens.
Sensational headlines in the international press, such as 'Embryo confusion: white couple have black twins' and 'Life-support blunder: hospital pulls plug on wrong patient' give the false impression that identification errors are spectacular and rare worldwide. In fact, misidentification of the cells, fluids, tissues, organs, limbs and bodies of patients is commonplace: the 'identity crisis' of modern medicine.
Reports in Hong Kong over the past year attest to mixed-up blood samples leading to transfusion error, mislabelled biopsies resulting in unnecessary mastectomy and wrong-patient prostate radiotherapy, mistaken implantation of prosthetics and confusion in the mortuary leading to cremation of the wrong body.
How reliably can we tell who's who and whose is who's in hospital? Let's start with the patient. The venerable hospital bracelet has been the mainstay of inpatient identification for decades. But, when scrutinised, wristband errors are detected in 7 per cent of cases: although in most cases the bands are simply missing, many others contain erroneous information and, astonishingly, 1 in 1000 patients is wearing a wristband belonging to someone else.
Wrong patient/site/operation errors have an incidence of about 1 per 25,000 procedures, and are the single commonest category reported by the Joint Commission on the Accreditation of Healthcare Organisations (which sets institutional health care standards in the US and internationally). Wrong-drug, wrong-patient mix-ups account for some 10 per cent of medication errors, another dominant cause of adverse events.
And then there are the tests. Mislabelling of laboratory blood samples occurs in about 1 in 500 instances. Studies show that up to 6 per cent of tissue pathology specimens are incorrectly labelled and that more than 4 per 1,000 surgical biopsy specimens are misidentified. Substantial patient harm results from at least 1 out of every 18 identification errors.
Why do these seemingly elementary mistakes occur so often? The answers lie in the pithy observation that 'every system is perfectly designed to achieve the results it gets'. Modern health care systems are predicated on faultless provider performance: in reality, medical care is delivered by skilled but error-prone risk-takers.
How can the health care system be modified to improve identification accuracy? Better design can reduce complexity, optimise information processing, automate and standardise. For example, bar-coding and radio-frequency tags have reading error rates of about 1 per million, and reduce medication errors by factors of two to five.
What can the health care consumer do to minimise their chance of misidentification? Two words: get involved. You check your luggage tags at the airport, right? But did you confirm the information on your wristband (let alone your biopsies) when you were last in hospital?
There is a good argument for a health care equivalent of the Universal Product Code, a unique identifier for individuals across all health care domains (the Smart ID card would be a convenient place to start). Medical misidentification events should be reported to a Patient Safety Agency in Hong Kong, and used as a quality-of-service indicator. Safety can then emerge as a product discriminator for hospitals in the public and private sectors, in much the same way as it has in the car industry (think Volvo).
Darren Mann is a clinical associate professor (honorary) at the Chinese University of Hong Kong and examiner in surgery of the Royal College of Surgeons of Edinburgh