Doctors, death certificates and reporting to coroners — room for improvement

In this issue of the Journal, Neate and colleagues review 320 deaths due to external causes in Victoria in the 2010–11 financial year that should have been reported to the coroner by doctors, but were not.1 Based on the causes of death reported on the death certificates, the Registry of Births, Deaths and Marriages referred the cases to the coroner for investigation. These deaths represent 6.6% of the total 4857 deaths reported to the coroner that year.2 Of these unreported deaths, 307 (96%) were the result of injuries from falls, 80% of the deceased were aged 80 years or older, and 68% of them were in hospital at the time of death. In 309 cases (97%), the cause of death was changed based on forensic pathology advice. One case went to an inquest. The fact that the external cause of the death — overwhelmingly fracture of the hip or pelvis, or head injury — was evident on the death certificate would seem to indicate a misunderstanding about the reportability of the deaths, rather than a conscious attempt to subvert the system.

In the United Kingdom, two recent major inquiries have involved the death certification system: the Shipman Inquiry in 2005,3 and the Mid Staffordshire NHS Foundation Trust Inquiry this year.4 The report of the latter cited a review of death certification at the Trust, which found that in 22% of just over 200 cases, there was a significant difference between the cause of death recorded on the death certificate and the cause of death in the medical records. A further concern was that 27% of the reviewed cases were regarded as coroners’ cases, but had not been referred.5

The Shipman Inquiry exposed the weakness of the death certification (and coroners) system in detecting hidden homicide, even on a huge scale, at the hands of the certifier. Similarly, deficiencies in health care could also be hidden. As part of the UK Government’s response to the Shipman Inquiry, the Coroners and Justice Act 2009 created a new official in England and Wales: the independent medical examiner (IME).6

The IMEs will be doctors who are appointed by local authorities and are independent of coroners, hospitals and health authorities. They will review the majority of death certificates (dependent on allocation of resources) and may review related medical records and speak with treating doctors. Their function is to ensure that deaths that should be reported to the coroner are reported, and to ensure death certificates are completed accurately.5 The Trust Inquiry believed the role should also include seeking out any serious adverse health care events. IMEs will be authorised to ask the next-of-kin whether they had any concerns about the way the deceased died or about the care received.5 It is anticipated that IMEs will help to smooth the interface between medical practitioners and coroners. About 500 IMEs are expected to be appointed, mainly part-time, from April 2014.

Doctors are in a powerful position when signing death certificates. The oversight provided by the Registry of Births, Deaths and Marriages, as described by Neate and colleagues, is the only system-wide oversight of death certification in Victoria. For the 50% of deaths that are followed by cremation, medical authorisation is also required for the cremation. Doctors authorising cremation must certify that they have undertaken “careful and independent inquiry into the circumstances surrounding the death of the deceased” to properly assess the cause of death provided on the death certificate and to ensure that the death is not reportable to the coroner.7,8 Although Neate et al do not report the number of cremations among the 320 deaths included in their study, if this system is fully effective there would be none.

It is interesting that in New South Wales, virtually all of these 320 deaths could quite lawfully have had death certificates written. Section 38(2) of the Coroners Act 2009 (NSW) states, in summary, that if the deceased was aged 72 years or older and died from injuries sustained accidentally and attributable to the age of the person, and there was no act or omission by anyone else contributing to the accident, then the doctor can complete a death certificate. As a safeguard, section 38(3) provides that the practitioner must not issue the certificate if a relative of the deceased (who is not obliged to provide any reason) objects to the giving of the certificate.

This seems a reasonable approach, allowing families to avoid the stress of a coroner’s investigation into a death resulting from age-related frailty. The weakness of the NSW system is that prevention opportunities arising from collective evaluation of such deaths in a vulnerable group may be lost. If the failings reported by Neate et al were replicated in NSW, this weakness would be compounded (and the public health surveillance enabled by death certification compromised) by the fact that major changes were required to the cause of death on almost half of the reviewed death certificates.

While education of doctors, as advised by Neate et al, is required, correctly certifying the cause of death or deciding to report the death to the coroner should be an explicit responsibility of the senior, not the junior, doctor. Both the aforementioned UK inquiries highlighted difficulties for younger and less experienced doctors in correctly completing death certificates and in handling the interface with coroners. The Mid Staffordshire NHS Foundation Trust Inquiry concluded that:

This is self-evident, and some hospitals have their own audit systems in place to monitor the discharge of these responsibilities by their staff. The report by Neate and colleagues is an important pointer to the need for improvement.