ABOUT six months ago, June Henley was admitted to the emergency ward at the hospital where I work in intensive care. By the time I was asked to see her she was suffering from all the features of a very serious infection, in this case an infected bladder. She was barely conscious, her blood pressure was low, she had a high fever and her kidney and liver functions were impaired. Like many older people – she is seventy-six – June’s first reaction to the illness, weeks earlier, had been stoic. She was feeling out of sorts but hadn’t wanted to bother the staff at the nursing home where she lives. By the time her GP saw her, a major counter-offensive was under way in her body and she was rushed to hospital.
As an intensive care specialist, much of my business comes from retirement villages and nursing homes, often as a result of the type of infection June was experiencing. Older people don’t have the same level of immune response as they had when they were younger. They can also have problems with incontinence, which sometimes means they must use pads. Certain bacteria thrive in these conditions, rapidly dividing and sometimes spreading further up the urinary tract into the kidneys. The body increases its temperature – bacteria do less well in higher temperatures and the body’s defences, including the killing power of the white blood cells, do better – and the battle is on. It can be a life-or-death struggle, especially if the immune system overreacts by producing chemicals of such potency that its own organs become damaged.
June had many of the “co-morbidities” – high blood pressure, high cholesterol, coronary artery disease, previous strokes, diabetes, heart failure, osteoarthritis and heart palpitations – that afflict older patients. In fact, these are so common that I wonder if we clinicians should simply give them a single name – maybe “ageing related conditions” – to counter our tendency to medicalise normal processes. The stroke had left her confined to a wheelchair, and had also affected her ability to swallow. She had a percutaneous endoscopic gastrostomy tube inserted directly into her stomach, enabling liquid food to be given without the danger of its being aspirated into the lungs as a result of her damaged swallowing mechanism.
Not surprisingly, June was on many medications: blood pressure pills, pills to lower cholesterol, pills to improve her heart function, and Warfarin, a blood-thinning pill to prevent further strokes. She was also in shock; in medical terms, this means a poor blood flow to the cells in the body – poor enough to result in serious cell dysfunction, which can eventually cause organs to fail and ultimately result in death.
WE HAD decisions to make. In view of June’s poor state of health and the serious nature of her condition, should we subject her to invasive treatment? If so, how far should we go? As far as she was able to concentrate, I explained the situation to June. The treatment for urinary tract infections goes something like this, in increasing order of invasiveness. First come antibiotics, which are usually very effective if they are given early enough. But June, not wanting to be a bother, had left it too late, and the infection had spread into her circulation. Intravenous fluids usually correct the shock that occurs in these cases; they’re easy to give, requiring only that a small tube, a cannula, be inserted into the vein. Antibiotics can be given at home; to take intravenous fluids you need to be in hospital – the first step up in the treatment hierarchy. Sometimes the low blood pressure doesn’t respond to intravenous fluids alone. Inotropes and vasopressors are next – powerful drugs that get the heart and circulation going, counteracting the shock. By now we are in the intensive care unit. If things get worse, we get the big guns out: the breathing might be affected, in which case it’s artificial ventilation; if the kidneys fail, then we use dialysis, usually as a temporary measure until the patient recovers.
At some point on this continuum, a decision often needs to be made about whether further treatment is the best option. In some cases, patients have thought of circumstances like these long before we see them, and have formulated an “advance care directive,” or living will. These forms are simple and easy to complete; unfortunately, dying and its precise diagnosis is often not quite so simple. The exact data about who will or will not survive a given condition and treatment isn’t precise. It depends on the severity of the illness and your body’s ability to fight it. With certain conditions – terminal cancer, for instance – death in the short term is obviously inevitable, but even then we are never sure exactly when the end will be. An advance care directive is useful in cases where we can prolong life, or at least prolong the dying process. Some people are clear in their views, and might not want to be on life support for the last few days of life. Or they might want to squeeze every last hour out of their life, even if they are unconscious and on a ventilator. But in my experience members of the latter group are rare and their wishes can change as they are subjected to more operations and chemotherapy.
How do you express the general feeling that you don’t want to end up on a life-support machine when there is no hope of recovery? What words reflect all the possibilities, all the ifs and buts? We have at least ten elderly patients admitted to intensive care each year after falling off a ladder. They would have died if they hadn’t been put on artificial ventilation for a few days. Instead, they go home as good as before the fall. In practice, the statement, “I do not want to be on a life-support machine,” may not reflect their wishes. To make an informed directive they would have needed more information – and, as they age, that information can become more complex, can change over time and can depend on which stage they’ve reached in the disease process. Like the treatment of urinary tract infections, much of healthcare is about a hierarchy of treatments that are available at a particular time. We usually commence a certain range of treatments and see what the response is. Things may get a whole lot better, in which case it might be worth continuing. Or the patient may deteriorate despite the therapy, in which case escalating the treatment might not be a good idea. The general expression that you do not want heroic measures when they are likely to be futile is a good basis for further discussions when the time comes; it’s hard to be more precise.
The hierarchy of treatment options was straightforward but were they all “appropriate” for June, and what role should she play in the decision? Antibiotics, intravenous fluids, intravenous drugs to support the circulation, a ventilator to support the lungs and dialysis for the kidneys have all been shown to work in certain circumstances. Whether they would be effective in June’s case was a different matter. The disease process was advanced and June’s ability to fight the disease was severely compromised. We have computer models that can roughly tell us a patient’s chances of survival, but the opinion of an experienced physician has been shown to be just as accurate. The model even takes into account the changing odds according to how the body responds to the treatment – but, again, no more accurately than an experienced physician.
But there are also other, non-medical factors to consider in the decision-making process. To make sure that we cover all these in a sometimes rushed and emotion-charged atmosphere, we use an ethical framework that sets out four main principles. First, it calls for “beneficence” – in other words, we should administer all the possible treatments that might benefit June. This might seem a simple criterion, but in some cases one treatment will work just as well on its own as some or all of the available treatments would in combination. In this case, we would be acting in a patient’s best interests by limiting the treatments.
“Non-maleficence,” the second principle, reminds us that treatments can do harm, just as not administering them can do harm, so we need to take account of – and avoid – the danger of administering treatments regardless of their impact on the individual patient.
Autonomy is the third principle. If I had interpreted this principle in its most literal sense, I would have given June a list of all the options and let her make the decisions. Autonomy drives much of medical practice in the United States. In this literal sense, patients’ choices largely depend on their knowledge and, importantly, on the way the options are presented. A doctor might say, “Here’s what we can do. Which do you want?” and thereby appear to be implying that all of the options can be successful so they should all be given.
This runs into the fourth principle – distributive justice. If we gave June all the therapies, even if we knew the effort would almost certainly be futile, it would come at a great cost to the health system. (If June lived in the United States she would have had to bear much of the cost herself.) I don’t know of any doctor who withholds care in Australia because of concerns about the cost to society. Maybe this is reassuring, perhaps not. But if the doctor is paid for each service he or she delivers, there is a perverse incentive to encourage the patient to take more options. Many Americans become bankrupt as a result of a slow death in intensive care units, the therapy escalating in the face of increasingly certain futility. And many doctors become very wealthy. That’s part of the reason why the United States has the most expensive health system in the world but ranks twenty-seventh in terms of health outcomes, marginally ahead of Cuba.
These conversations can be even more difficult when the patient is no longer able to communicate his or her wishes. The nearest relatives might want active treatment to cease when there is still a high chance of the patient’s recovering, or they might want active treatment continued when there is no hope. It’s not clear who has rights here: the relatives who are acting in what they see as the patient’s interests or the doctors who have a right not to be forced into delivering care that they think would be inappropriate or futile. And what rights does the community have in the challenge of funding increasingly expensive terminal care?
An important part of the decision-making process for doctors is taking into account the frailty of the patients – or what we sometimes call their “reserves.” Frailty is difficult to define. We all know that ageing compromises the body’s ability to combat disease. June obviously didn’t have the reserves of a twenty-year-old, but her degree of frailty was almost impossible to estimate. Frailty has something to do with age, the number and type of co-morbidities and the patient’s pre-existing health; perhaps there are other factors that we cannot measure – factors that relatives are getting at when they say things like, “She’s a real fighter and not ready to go just yet.” Ultimately, frailty can only be measured in retrospect. Some patients deteriorate and die rapidly; others recover just as quickly.
I explained to June that I would take her into the intensive care unit and give her antibiotics and intravenous fluids and then we would take it from there, depending on how well she responded. I said no more. I didn’t mention ventilation and dialysis, as I believed that they would probably be ineffective if she didn’t respond to the antibiotics and fluids – just as I wouldn’t have discussed a heart transplant or snap freezing her body when she died and unfreezing her when medicine had a cure for her condition. Mentioning other possibilities might have implied that they could help but were being withheld.
June responded well to the antibiotics and fluids. She didn’t need the powerful drugs to support her circulation and she didn’t need artificial ventilation or dialysis. She was discharged back to the village after three days in the intensive care unit and a week in hospital. Of course, her difficulties didn’t entirely end at that point.
We doctors who work in acute hospitals often forget about convalescence. June was too debilitated to walk the fifty metres from her room to the dining room for three weeks. During that time she was either confined to bed or needed assistance to sit in her chair. She was anxious and depressed, had problems sleeping and, when sleep came, was subject to nightmares. After three weeks, though, her physical strength and confidence gradually returned. After three months she was back to the same state of health she had experienced before the infection. Like frailty, the need for convalescence is hard to measure. We are just learning that when patients have been in intensive care for more than two days they will almost invariably leave with debilitating symptoms; longer visits to the unit can lead to anxiety, depression, relationship problems and in some cases post-traumatic stress disorder. Some intensive care units run special follow-up clinics to reassure patients and, where possible, to treat the symptoms.
The urinary tract infection had severely knocked June about; the slow recovery was another measure of her frailty, of her reserves, of her tenuous hold on life. And yet she survived the setback.
MERLE HAWKINS had the same disease. She was eighty-four years old and also lived in a nursing home. She presented at the same late state of the disease: in shock, with evidence of kidney and liver dysfunction. There was an important difference: she had severe dementia. She was in a special section of the village, under the care of psychogeriatricians. She presented late because she could not communicate her symptoms. It was not until she changed from being occasionally violent and always confused to being lethargic and drowsy that the infection was picked up.
Merle had no formal advance care directive but had made it clear to her children that she would never have wanted to be demented and totally dependent on others. For her children, Merle’s dementia was a traumatic experience; when they visited her in our unit they seemed almost as disoriented as Merle, barely recognising her for who she was or who she used to be.
The ambulance was called by the staff at the nursing home. Having heard the details, the paramedics felt uncomfortable about urgently transporting this drowsy, peaceful-looking woman to the local hospital. The nursing home staff also felt uncomfortable. They would have been happy to leave her in her bed and tend to her dying over the next day or two. Merle’s children would have been happy for that to happen and, according to them, so would Merle. But the system, as it stands, doesn’t allow discretion. Merle was on the conveyor belt.
The staff from the emergency department contacted me about admitting Merle to the intensive care unit. Her disease was at the same stage as June’s had been. She had similar, though perhaps fewer, co-morbidities than June. Merle was older but perhaps slightly healthier. I declined her admission to the unit and suggested we speak to the children, keep her comfortable and not even give her anti-biotics. But the conveyor belt was still operating. Because I was not going to admit her to the unit, the emergency physicians contacted the on-duty general physician, who accepted care. It is rare at this point to refer the patient back to the nursing home or to an alternative institution where her needs could have been met in a more appropriate way, concentrating on making sure she wasn’t suffering and treating her with dignity. Acute hospitals are not good places to die. Their business is about actively treating conditions – they specialise in starting and continuing therapy, not in deciding when it’s time to stop and how to manage that process.
Merle was admitted to the general wards and given antibiotics and intravenous fluids. Overworked staff in the general wards of an acute hospital cannot administer fluids with the same precision and in the same amounts that we do in intensive care. It’s our bread and butter. Consequently, Merle’s blood pressure remained low and within twenty-four hours she had quietly died. The dying process was not acknowledged in any active sense but the result was much the same as it would have been had she not been given intravenous fluids. The decision to give the fluids was made because staff trained to treat patients actively find it difficult simply to watch and make sure a patient isn’t suffering. Giving intravenous fluids gave the staff the satisfaction that they had “done everything.”
I felt uncomfortable about my complicity in this sequence of events. I was not consoled by the knowledge that I’d advised my colleagues that we should adopt comfort care rather than a more active intervention. Perhaps I should have advised the relatives that they shouldn’t consent to further treatment, including antibiotics, but I would have been in conflict with my colleagues and she was their patient. Even if I had advised the family not to continue further active treatment, various family factions might have disagreed and accused the relatives or me of not caring, given that all sorts of family dynamics are uncovered during these difficult times. The fact is that I am part of a system in which active treatment in the face of futility will continue to happen, often against the wishes of the patients and their carers and certainly not in the interests of broader society. Almost a third of the health budget is spent in the last few months of life for very little gain – and in many cases a diminution – in quality of life.
A few months ago, I was discussing these issues on a radio program when a woman aged around eighty from a rural area rang in and related her story. Her husband had a terminal disease, and the couple had discussed the consequences at length. If he deteriorated, he was not to be admitted to hospital; he wanted to die at home. They had found out all you need to know about community nursing support, and their family doctor was aware of their wishes and agreed with them. While driving to the shopping centre one day, the husband suddenly slumped next to his wife, who was driving. She stopped the car, confirmed that he was dead, restarted the car and began to think through the issue rationally.
She knew that calling the ambulance or a quick drive to the nearby hospital would probably mean he would have unwanted CPR. She had heard that after five minutes or so, the chances of a meaningful recovery were almost nil. So she slowly drove around the town. Hearing the bell at the railway crossing she headed over there and waited until the train had passed. Then she drove to their family doctor, careful not to exceed the speed limit and taking her time to find a parking space. The doctor rushed out, called the ambulance, pulled the man out of the car and commenced CPR. It was all too late, of course, and he was declared dead at the hospital. It’s often hard to die in peace. •
Ken Hillman is professor of intensive care at the University of New South Wales. Pseudonyms and composite characteristics have been used in this article to protect the privacy of patients.
Also in Inside Story: The myth of CPR
How did such a poorly proven intervention become a routine end to many people’s lives, asks Ken Hillman