I got a jolt reading Simon Hattenstone and Daniel Lavelle’s Empty Doorway article on the death of Gyula Remes, who was found by Westminster tube station last December. I realised I had autopsied his body at the request of the Southwark coroner.
I was not at the inquest. Pathologists are called to give evidence in person only if there is uncertainty about the cause of death or if the relatives wish it; but here, neither condition applied. My written cause of death (alcohol toxicity) was the conclusion, but my original full version did not appear, for I had added “homelessness – to be considered at inquest”.
It was considered, but it could not appear on the final death certificate because “homelessness” is not listed as a disease in the International Classification of Diseases (although it does feature in an annex entitled “Factors influencing health status or contact with health services”). I wrote it only to remind the coroner of what I thought was a fundamental reason why Remes died.
The Office for National Statistics (ONS) bulletin tells us that last year deaths among homeless people in England and Wales rose to 726, a fifth of them in London. Drug poisoning (for instance through illicit drugs such as opiates) rose as the cause of death in the homeless, while deaths from alcohol toxicity and suicide were proportionally similar to previous years. I have worked as a pathologist examining deaths for more than 40 years, and although I am part-time now I cannot help noticing that homeless rough sleeping features more often as the scenario of death than previously.
My personal observation is borne out by the statistics, which for the past year alone show a rise in homeless death of 22%. But shocking as the ONS numbers undoubtedly are, there is reason to believe that the true total could be higher. Accurate causes of death are essential for monitoring what goes on in society, and they advise governments where to put the money to improve matters in future by critical changes in policy and funding. Britain has a proud record in death certification, but for many reasons a substantial proportion overall, 20% or more, are known to be incorrect.
How the ONS arrives at these numbers – by no means a straightforward matter – is detailed in another document, the Quality and Methodology Information (QMI). These are based on death registration records (basic causes of death), linked with records of people sleeping rough or using emergency accommodation (homeless shelters, direct access hostels) at or around the time of death. A further proportion will have died not on the streets but later in hospitals.
The QMI acknowledges that the numbers are estimates, and that the real number may be higher. For one thing, the number of deaths registered in 2018 will not be the same as the number of deaths actually occurring in that year; the process of death investigation, including inquest, can take months and sometimes years to complete to registration.
The QMI states that “12% of deaths of homeless people are certified by a coroner”. Does this mean that the other 88% did not pass through the coronial system? The 12% must include, at the least, the rough sleepers. Anyone found dead on the streets is, by law, going to be reported to a coroner, who will be duty bound to investigate what happened by an autopsy. These unfortunate people will not have a note in their pocket indicating their GP, who would be able to produce a natural cause of death over the telephone to the coroner, thus avoiding an autopsy.
The ONS statistical bulletin also states that “looking at deaths identified as homeless in 2018 registrations, 90% were investigated by a coroner”. Lots of deaths are reported to coroners, but not all are accepted for full investigation, and only about 40% of reported deaths proceed to a coronial autopsy examination. Here, then, is another confusion: is it the majority of homeless deaths that get the full investigation by a coroner, or a small minority?
The next uncertainty is where the deaths are taking place. The ONS does not (perhaps cannot) give us the breakdown of homeless deaths into more precise categories of homelessness, including often/ever/never a rough sleeper. There are many different definitions of homelessness – some wide, others narrow – beyond the obvious and visible rough sleepers. The majority of homeless deaths probably happen not on the streets but in shelters and hostels for the homeless, or on NHS premises. Their death certificates are usually written by an attending doctor.
Only if the circumstances of death appear unnatural (a term with no comprehensive agreed definition), or if the final cause of death is not evident clinically, is the death referred to a coroner. From long experience, I am sceptical about the accuracy of such death certificates, compared with definite information from an autopsy.
From my experience, where the coroners do become involved there may be no consistency in how they depict these deaths in their conclusions. In non-rough-sleeper homeless deaths, they may not be informed that a contributory factor was homelessness. Going back to the death of Gyula Remes, the Record of Inquest said that in addition to the influence of alcohol he was “rough sleeping”: so this one was definitely counted – although not for 2018, the year of the death, since the inquest happened the following year.
Ultimately, registration delays, uncertainties over place of death and how thoroughly the deaths are being investigated should make us cautious about the accuracy of the statistics: the homeless mortality problem may be even more serious than we think.
Sebastian Lucas is honorary consultant pathologist at Guy’s & St Thomas’ NHS Foundation Trust
Follow Guardian Cities on Twitter, Facebook and Instagram to join the discussion, catch up on our best stories or sign up for our weekly newsletter