The Way We Die Now

by Seamus O'Mahony

  The HSMR statistical method was developed by the Dr Foster Intelligence Unit at the School of Public Health at Imperial College London, which founded a separate campus company to exploit its commercial potential. Many questioned the statistical methodology, including Paul Taylor, an expert in health informatics, who wrote an illuminating article, ‘Rigging the Death Rate’ for the London Review of Books. (The best information and analysis of medical issues is now found, not as one would expect, in the medical journals, but in literary magazines such as the LRB.) Taylor and several others demonstrated the crudity of this mortality ratio. It is subject to many biases and distortions, including in the accuracy of coding, the quality of local GP care, and access to hospice care. A commercial competitor of Dr Foster Intelligence, Caspe Healthcare Knowledge Systems (CHKS), advised another hospital – Medway – which also had a high HSMR. They advised the hospital trust that it had been ‘under-using’ the specific code for palliative care: by increasing the proportion of patients it coded as receiving palliative care, Medway lowered its HSMR dramatically.

  The first Francis Report gave a summary of an independent assessment of the HSMR statistical method by two epidemiologists from the University of Birmingham, Professor Richard Lilford and Dr M. A. Mohammed: ‘our most crucial finding is that the methodology used to derive the Dr Foster SMR is riddled with the constant risk-adjustment fallacy and so is not fit for purpose.’ Even Roger Taylor, director of research at the Dr Foster Unit, lamented that the statistics had been ‘woefully poorly misunderstood’. He went on: ‘there is no figure for the actual number of people who might have died avoidably. It is impossible to put an actual figure on it.’

  Although the public inquiries into Stafford Hospital were covered in some detail by the press, it was less widely reported that Dr Mike Laker of Newcastle University was asked by the Mid Staffordshire Trust to examine a number of cases where families felt that poor care had contributed to a relative’s death. He interviewed 120 families and examined 50 case-notes and concluded that poor care caused death in ‘perhaps one’.

  Even Robert Francis, in his voluminous reports, concluded that no firm conclusions could be drawn from the hospital mortality figures. He acknowledged that unkindness, rather than a high death rate, was the main concern of those who had given evidence: ‘It was striking how many accounts I received related to basic elements of care and the quality of the patient experience, as opposed to concerns about clinical errors leading to death or injury.’ Yet the public, the media and the politicians now assume that any deaths above the national average must be the consequence of poor care, and thus, avoidable. The Guardian, in a piece by Denis Campbell headlined ‘Mid Staffs hospital scandal: the essential guide’, opens with the following statement: ‘An estimated 400–1,200 patients died as a result of poor care over the 50 months between January 2005 and March 2009 at Stafford Hospital.’ If the deaths were the result of poor care, somebody must be to blame, and there were calls for criminal prosecutions at Stafford. Peter Dominiczak in the Daily Telegraph reported in June 2013:

  A review of deaths at Mid-Staffordshire NHS Foundation Trust by police and other officials has identified hundreds of cases between 2005 and 2009 where poor care could have led to deaths. Campaigners welcomed the announcement and warned that individuals ‘must be held to account’ to avoid another tragedy on a similar scale to Mid-Staffs... Politicians and campaigners have lamented the fact that no one involved in the scandal has so far faced any legal action.

  So the Stafford Hospital scandal has now fixed an idea in the public consciousness of hospital death as a failure of medical care, on a par with an industrial accident, which automatically triggers an investigation by the Health and Safety Executive. This, combined with the demise of the Liverpool Care Pathway, has, at least temporarily, put into reverse the programme of enlightened care of the dying in acute hospitals in the UK.

  The various Royal Colleges felt obliged to respond to the Francis Report. The response from the Royal College of Physicians, Putting Patients First: Realising Francis’ Vision, has the expected hand-wringing tone of such documents, with woolly aspirations about developing leadership among doctors, ‘helping to improve patient experience’, setting quality standards, improving training, and so on. There is a prevailing societal view that doctors and nurses must be more ‘accountable’ and (particularly in Britain) there are literally dozens of agencies, colleges and quangos charged with overseeing those who work in hospitals. The moral philosopher Onora O’Neill addressed this ‘accountability’ culture in her Reith Lectures in 2002, and observed that while in theory the new regulations make professionals more accountable, in practice they achieve little, except an increase in suspicion: ‘currently fashionable methods of accountability damage rather than repair trust’.

  WAREHOUSES OF THE DYING

  It is a curious statistic that, in Ireland, you are three times more likely to die in an Intensive Care Unit (ICU) than in a hospice. Death in an ICU is frequently held up as an example of the worst kind of ‘technological’ hospital death, but is this really true? The ICU, unlike the general ward, is above all else a controlled environment. Each patient has a designated nurse looking after them, and them alone. The consultant staff are on the ground. Any acute deterioration is detected quickly and acted upon. The squalor and chaos of the general ward is not in evidence: the ambience and ethos is different. The ICU in my hospital has a ‘quiet time’ between 12 noon and 2 p.m., when no ward rounds or procedures take place and the lights are dimmed. About one in five patients admitted to an ICU will die there, but in most such cases, ‘active’ treatment has been withdrawn in the days or hours leading up to death. ICU doctors are generally anaesthetists by training, and are thus particularly good at pain relief.

  So dying in an ICU may not be the worst type of hospital death. This is not the case in the US, where elderly patients with dementia and metastatic cancer are routinely admitted to an ICU. One American ICU doctor, interviewed by Atul Gawande, remarked bitterly that she was running a ‘warehouse for the dying’:

  Out of ten patients in her unit, she said, only two were likely to leave the hospital for any length of time. More typical was an almost eighty-year-old woman at the end of her life, with irreversible congestive heart failure, who was in the ICU for the second time in three weeks, drugged to oblivion and tubed in most natural orifices and a few artificial ones... Another woman, in her eighties, with end-stage respiratory and kidney failure, had been in the unit for two weeks. Her husband had died after a long illness, with a feeding tube and a tracheotomy, and she had mentioned that she didn’t want to die that way. But her children wouldn’t let her go, and asked to proceed with the placement of various devices: a permanent tracheotomy, a feeding tube, and a dialysis catheter. So now she just lay there tethered to her pumps, drifting in and out of consciousness.

  It has been said that in the US, only the very poor (and thus medically uninsured) manage to die with dignity. Intensive care is used much more sparingly in Britain and Ireland, but the inexorable trend is towards American-style ICU care. The average age of an ICU patient is sixty, but many patients in their eighties and even nineties are now admitted to ICUs, which would not have been the case when I trained in the 1980s. Doctors are also increasingly fearful of being labelled as ‘ageist’ by relatives and also by their geriatrician colleagues. Families, even those of the very old and frail, are increasingly demanding maximum intervention. This is exacerbated by media scare-stories about old people being denied ‘life-saving’ treatments.

  WILD DEATH

  Most hospital deaths, however, take place, not in the ICU, but in General Medical wards, which are sometimes chaotic, understaffed and overcrowded. These wards are occupied mainly by elderly, highly dependent patients, and often there are simply not enough nurses to look after them properly. Relatives complain bitterly of their loved ones not being fed, but would rarely entertain the idea of assisting the nurses in this task, as is common in other
countries. And it’s a vicious circle: an overstretched nurse, lacking leadership, faced with very dependent elderly patients and complaining relatives, may quickly burn out and get through the shift doing the bare minimum. Many of the very nurses the system really needs – those with decades of ward experience – have left to pursue nine-to-five jobs as specialist and out-patient nurses, leaving the heavy lifting to the young and the inexperienced.

  Patients in these general wards fall into the following three categories: first, the acutely ill; second, the ‘medically discharged’ (previously known, uncharitably, as ‘bed blockers’) – old folk who have recovered from their acute illness but who cannot go home and are awaiting ‘placement’ in a nursing home; and third, the dying. The acutely ill patients, understandably, are the priority for the nursing and medical staff. A single acutely unstable patient commonly distracts attention from the other patients: when staff and resources are limited, these other patients may be neglected.

  There are different kinds of hospital death: the rapid, acute death, caused by some sudden catastrophe such as a heart attack or a stomach haemorrhage; the semi-acute death, which occurs over weeks, from a chronic illness such as liver cirrhosis; and the slow, lingering death, such as that which eventually overtakes patients with dementia or a stroke. Ideally, patients who are acutely unstable should be treated in the Intensive Care Unit or a High Dependency Unit (a sort of halfway house between an ICU and a general ward), but beds are too few to accommodate the growing demand, so the ICU doctors can only take the very sickest.

  Without exception, the worst type of hospital death is the acute death on the general ward. The patient might not be recognized initially as being sick enough to require ICU admission, or there may simply be no ICU beds. The nurse may have several other sick patients to attend to, and the first port of call for medical help is often an inexperienced and terrified intern. There is inevitably, at the end, a gruesome scene at the bedside following the unsuccessful resuscitation attempt.

  Most of the deaths I witness are semi-acute or chronic. By far the commonest cause of this kind of death in my practice is from liver cirrhosis. Death in this particular group of patients is particularly difficult for people to comprehend: most of the patients are young (under fifty); the disease is ‘not cancer’; there is a known ‘cure’ (liver transplantation). Unfortunately, most patients admitted to hospital with liver failure caused by alcoholic cirrhosis do not qualify for a transplant; you have to have demonstrated abstinence from alcohol for a minimum of six months, and (paradoxically) you have to be well enough to undergo the operation – so, for example, you have to overcome infection or kidney failure first. Liver transplantation is therefore for a very select few: if you survive the liver failure and stop drinking, you generally don’t need one; if you go back to drinking alcohol, you don’t qualify for it; if you’re too sick, you can’t have it. Thus liver transplantation is reserved for that minority who, despite giving up alcohol, develop recurrent liver failure, but recover sufficiently from these bouts to undergo the transplant operation. Most of my patients with alcoholic cirrhosis continue drinking, and die, sooner or later, of liver failure.

  Over many years, I have become better at predicting those of my liver failure patients who will live, and those who will die. (I may be deluding myself in this regard, because several studies have shown that doctors are woefully inaccurate at predicting survival time in dying patients: they generally overestimate.) This art of prediction is a mixture of science (clinical risk scores), experience and intuition. It can be difficult, however, to persuade families that the overwhelmingly likely outcome is death. The family of the woman with liver failure I described in Chapter 1 never accepted my prognosis, a refusal which only added to the discomfort and distress of the dying woman. These patients rarely die in a hospice or at home. Most die on the acute ward, or in the ICU, after weeks, or even months, of uncomfortable and invasive treatment. Death from liver failure can be nasty and undignified, particularly if bleeding is the final insult. Luckily, most slip into a coma and are unconscious for several days before their death. I have witnessed the deaths of dozens of patients with liver cirrhosis; those who knew and understood what was happening, and who opted for a non-interventionist approach, were rare indeed.

  The family of one patient of mine did however take my advice, and gave their son a peaceful death. He came under my care many years ago, when I worked in Yorkshire. He was only thirty-five, and had several admissions with liver failure. He had a degree of intellectual disability, and could not truly comprehend what was happening to him. The origins of his alcohol dependence were sad, and slightly comic. Education and work had passed him by, and the only environment where he felt truly accepted was the local pub. Although he had the mind of a child, he learned certain manly phrases, and was adopted by the hard-drinking pub regulars as a sort of mascot. He had neither the insight nor the motivation to stop drinking. During his last admission, it became clear that he would not recover. I spoke to his devoted father, and advised him that intensive care had little or nothing to offer and that we should aim for comfort. To my relief, the father agreed, and the family spent the last night at the bedside, praying. Many patients and families, however, cannot embrace what they regard as ‘giving up’. Liver cirrhosis is a perfect example of how the modern practice of pushing treatment to the limit can go wrong.

  BRINKMANSHIP

  The American medical ethicist Daniel Callahan has defined the term ‘technological brinkmanship’ as:

  A powerful clinical drive to push technology as far as possible to save life while, at the same time, preserving a decent quality of life. It is well recognized by now that, if medical technology is pushed too far, a person can be harmed, that there is a line that should not be crossed. I define “brinkmanship” as the gambling effort to go as close to that line as possible before the cessation or abatement of treatment.

  This sounds good in theory, but Callahan identified the obvious limitation of this strategy:

  In particular, brinkmanship fails to reckon with two potent realities, each of which conspires to make it hard to locate the point at which the brinkmanship should stop, and just as hard to work up the will to stop once this point has been identified. The two realities are the vanishing line between life and death, which makes it difficult to determine when to stop the use of technology, and the continuing profound public and medical ambivalence about what is wanted and valued in coping with illness and dying.

  Callahan’s ‘vanishing line between life and death’ simply refers to the difficulty in being sure, in a patient with a chronic disease, when the process of dying has started. This vanishing line, this difficulty, contributed to the demise of the Liverpool Care Pathway in 2013. Medicine is, and always has been, messy, imprecise and uncertain. And nothing is as messy, imprecise and uncertain as predicting death.

  SOCIETY’S DUSTBIN?

  The GP and writer Iona Heath asked: ‘Why is it that so few of our patients die what would be recognized or described as a good death?’ She goes on to recount a patient’s story:

  Some years ago, an elderly patient on my list was admitted to hospital after she collapsed. She was in her late eighties, a widow and very frail. She was admitted to a coronary care unit and received the highest possible standard of care including fibrinolytic treatment [‘clot-busting’ drugs] delivered according to the latest evidence-based guidelines. She made a good recovery and was discharged home, apparently well, a week later. I went to see her and found her to be very grateful for the kindness she had been shown but profoundly shocked by a course of treatment that she perceived to be completely inappropriate. She explained to me that not only her husband but almost all her generation of friends and acquaintances were already dead, that her physical frailty prevented her from doing almost all the things that she had previously enjoyed and that she had no desire to live much longer. No one had asked her about any of this or attempted to discover whether the effective and therefor
e recommended treatment for her condition was appropriate in her particular case. She died three weeks later while asleep in bed.

  Let us for a moment examine Heath’s assertion that someone should have carried out what can only be described as an existential assessment of this old lady before giving fibrinolytic treatment. The patient was almost certainly treated by a busy, shift-working junior doctor, and the old lady was one of perhaps dozens of acute patients seen on that shift. It is rather fanciful to imagine that in such circumstances, an assessment of the patient’s sense of herself and her situation would be a priority. Furthermore, a decision to deny such a patient ‘clot-busting’ treatment could easily be construed as ageist – even the normally sensible Raymond Tallis (geriatrician and philosopher) has labelled doctors who do not treat the old as actively as they treat the young as ‘scoundrels’. It is an absurd suggestion that an on-call junior doctor should take such a professional risk. Even if the old lady – after the detailed existential assessment suggested by Iona Heath – was happy to reject such treatment herself, it is not unknown for relatives to emerge after the patient’s death, demanding to know why their loved one was denied potentially life-saving procedures. One can imagine the scene at the inquest:

  CORONER: Doctor, why did you not give this patient, with a known myocardial infarct, fibrinolytic therapy?’