Dying to know why (Image: Team Static/Getty)
The best way of checking that doctors correctly diagnose causes of death is dying out, warns a study in the UK. The percentage of hospital deaths investigated by autopsy has plummeted from around 40 per cent in 1960 to less than 1 per cent today. Similar declines are occurring in the US and Australia.
The report鈥檚 authors say the consequences could include more people going to the grave with a misdiagnosed cause of death, misallocation of resources because death statistics are wrong and lost opportunities for medical students, doctors and researchers to investigate first-hand the causes of new diseases, such as SARS, and poorly understood but increasingly common conditions such as Alzheimer鈥檚 disease.
Advertisement
鈥淎t this rate, all autopsies will disappear within 20 years,鈥� says Angus Turnbull of Imperial College School of Medicine in London. 鈥淣ow, at medical school, you can go through six years of training without seeing one,鈥� he says.
Turnbull and his colleagues collected data on autopsy rates in 2013 from 184 of the 186 National 午夜福利1000集合 Service trusts. On average, the trusts autopsied just 0.69 per cent of deaths. Rates were lowest in Northern Ireland at just 0.46 per cent, and highest in Scotland at 2.13 per cent. Autopsies had already disappeared completely in a quarter of trusts.
Room for error
Reasons for the decline include doctors supposing that scanning techniques such as MRI and CT mean their diagnoses are more accurate than ever before, rendering autopsies unnecessary.
But this confidence can be misplaced, says Sebastian Lucas, professor of histopathology at St Thomas鈥� Hospital in London. 鈥淭he impression that all is known through pre-mortem imaging and diagnostic tests is pervasive, but wrong.鈥�
Misdiagnoses still occur. For example, one US study on autopsies of , half of which would have affected the choice of treatment. And it鈥檚 not just cancer cases: a review from 2012 of 31 studies involving intensive care units found that .
鈥淭he proportion of misdiagnosis is shockingly high for conditions that, had they been recognised while the patient was alive, very likely would have altered care and perhaps improved outcome,鈥� says Bradford Winters of Johns Hopkins University School of Medicine in Baltimore, Maryland, who led the 2012 review. 鈥淲ithout a reference measure such as autopsy to determine our misdiagnosis rate, efforts to prevent misdiagnosis will not be able to generate sufficiently valid data so that we can tell the public that we are making strides,鈥� he says. 鈥淲e will essentially be 鈥榞uessing鈥� whether the situation is improving or not.鈥�
Difficult conversation
Another possible explanation for the decline is that doctors may believe relatives of deceased patients will be sensitive about being asked for permission to conduct autopsies, especially in the wake of scandals such as the retention of children鈥檚 organs at Alder Hey hospital in Liverpool in the 1990s. 鈥淪ince Alder Hey, many doctors have assumed people won鈥檛 give consent for autopsies,鈥� says Turnbull, but this isn鈥檛 true.
A recent study at London鈥檚 St Bartholomew鈥檚 Hospital showed that 18 of 23 groups of . And Alder Hey itself has the highest autopsy rate in the country, at 21 per cent.
Turnbull thinks it will be a struggle to reverse the decline, but not impossible. If doctors had an obligation to refer at least one uncertain death per year, that could help, as would making consent forms simpler, he says.
Virtual autopsies 鈥� those carried out using high resolution MRI immediately after death 鈥� can help fill the gap, says Winters, but the false positive and false negative rates this gives for most conditions are still unknown. Although virtual autopsies have their strengths, the scans can miss important clues such as blood clots, says Turnbull. 鈥淎lso, who would do them, as there鈥檚 no training for it.鈥�
鈥淲hile autopsy is not perfect, it is the 鈥済old鈥� standard,鈥� says Winters.
Journal reference: Journal of Clinical Pathology, DOI:
