Month: October 2012

An assault on trust: in defence of the patient-physician relationship at the end of life

The recent controversy about misuse of the Liverpool Care Pathway (LCP), arising from Professor Patrick Pullicino and channelled via the Daily Mail, has resulted in a sustained attack on the bond of trust between doctor and patient. It has been suggested that the LCP is commonly used to bring forward the death of patients for the purpose of vacating much-needed beds. The implicit suggestion is that very real economic and logistic problems faced by the NHS are influencing the therapeutic decisions being made by individual doctors. Critics of the LCP have been very emphatic and clear in their suspicions. Here are some quotes from Melanie Phillips in the Daily Mail:


In practice, however, the LCP has turned into something quite different. For while in some cases it has been used properly as intended, with numerous others it has become, instead, a backdoor form of euthanasia.


Horrifyingly, the LCP has become a self-fulfilling prophecy. When people are put on it, they are said to be dying. But they may not be dying at all — not, that is, until they are put on the ‘pathway’, whereupon they really do die as a result.

 In other words, they are killed. What’s more, they are killed in a most cruel and callous way through starvation or dehydration. And this in a health service that is supposed to be a national byword for compassion!

There are suspicions, based on much circumstantial evidence, that such patients are being dispatched via the LCP because — simply and crudely — the hospitals need their beds to meet overwhelming demand. This callous disregard for the most needy is of a piece with the all too frequent abuse of patients who are elderly, confused or difficult to manage and who may be treated in hospital with indifference, neglect or even cruelty. 

Doctors who have used the LCP are therefore suspected of having killed patients and of bowing to managerial pressure to clear their beds. A direct, destructive link has been created between bed pressure and the purity of therapeutic contract. This requires more examination. I will now describe how consultants balance the pressures on their service and the needs of their individual patients. This is my view and I cannot speak for others.

The performance of doctors is under continuous scrutiny. Markers of quality are recorded and fed back to us on a regular basis. For patients the most visible aspects of quality are the efficacy of treatment, its safety, and the preservation of their comfort and dignity. It is difficult to measure these things, but patient questionnaires, the monitoring of adverse events and, with the help of outside agencies, adjusted mortality statistics, do provide useful information. Another important measure is the ‘length of stay’ (LOS). This is the amount of time a patient stays in hospital after presenting with an acute problem. It is accepted that the shorter the LOS the more efficient the medical system. Of course a short LOS is acceptable only if the patient is diagnosed accurately, gets better and has a reasonably comfortable and safe experience. But if we can be sure that our medicine is working, the the next thing to concentrate on is its efficiency.

As more and more people reach their eighth and ninth decade of life, more emergency presentations occur. Having recovered from their illness many patients cannot go home because they have become weaker. In order to achieve discharge help from social services is required, for the provision of daily care at home, or placement in a residential or nursing home. This takes time. Beds are said, insensitively, to be ‘blocked’. LOS trends upwards. New patients keep coming to casualty, but there are fewer beds in which to place them. Pressure! What do we do about it? How do we shorten LOS?  Would I, would any doctor, really hasten a patient’s death to this end?

What do I think when I see a patient who may be dying? My attention is narrowed. I perceive a man or woman in the last days or weeks of their life and reflect on everything that has gone before. It is an unscientific, almost poetic moment. I am privileged, in a way, to meet an individual at this defining moment. Everything that they have done, all that they have seen, resides in that darkening mind, memories betrayed by the frailty of the tissue that has formed them.  I don’t know them well, of course. I find out what sort of family they have, how many sons or daughters or grandchildren or great-grandchildren. Those family members may have already been to the ward and I, or  members of my team, will have spoken to them. The details they provide form a rounded human being in the eyes of the doctors and nurses who barely three of four days ago had never heard of them. This rapid assimilation of impressions and emotions creates a three-dimensional image, albeit sometimes lifeless and uncommunicative, that is the centre of our decision-making.

As we consider whether to start the LCP we run through several mental processes. Is this really the end? Could a change in medication or management result in recovery and a successful discharge home? Would the patient wants us to work harder in an effort to avoid death or prolong life? Sometimes that the answer is obvious. Most experienced doctors will have seen over a hundred people reach the end of their life and the signs, although not scientifically verifiable or quantifiable, become recognisable. If it is obvious that death approaches despite our treatment we need to rationalise what we do to the patient. I have rehearsed some of these decisions in a previous blog post (‘What we talk about when we talk about death: a case’). What I don’t think about is money.

Part of my role as a responsible member of the NHS is to use resources appropriately and efficiently. Sitting at my desk I frequently review my use of particularly expensive treatments or procedures, but on the ward in front of frail patients, I am a clinician. I am not naive – I do need to be aware of resources and make sensible decisions on a day to day basis. But death is different. A good death, if there is such a thing, demands care and focus. It doesn’t cost much. Thoughts that do not enter my mind, as I look down on an emaciated, obtunded or cancer riddled patient, are bed management, targets, quality indicators or local health economics.

Yes, if the patient was to go home two days earlier, or to die two days earlier, the bed would be used by another patient, and a financial benefit could be calculated. My LOS data might look a little better.  But for the clinician such a calculation is irrelevant. When a doctor meets a dying patient all such thoughts are banished. The bond between doctor and patient is at its strongest at this unique moment, and will not be eroded by such intangible concerns.


What we talk about when we talk about death: a case

An elderly man, the precise age is immaterial, is admitted from home having had a fall. It is quickly determined that he has a chest infection, and antibiotics are started. We are optimistic that he will recover. On the second day he appears confused, and the amount of oxygen he requires goes up. Sometimes people get worse before they get better, and the antibiotics have barely got into his system…we carry on. We know now that despite his jovial character and bluff attitude, he was not in such good shape at home. He managed on his own, but the neighbours did all the shopping and he was rarely seen out of the house. He smoked two packs of cigarettes a day, always had done, as evidenced by his yellow hair, leathery skin and walnut coloured fingertips. His lungs are overlarge, expanded, most of the air-sacs within broken down and useless.


On the third day we begin to worry. A blood test that is used to track the progress of an infection (CRP, not infallible) has gone up, from 60 to 150. We ring the microbiologist for advice, essentially for permission to use a stronger antibiotic. These are commonly effective, but bring with them the risk of Clostridium difficile, a potentially lethal diarrheoa bug. We prescribe it.


We haven’t seen any relatives, but I ask my team to make sure that his next-of-kin, whoever it is, knows that things are not going so well. On the fourth day he looks a little better, on the fifth worse. He isn’t eating well, and we worry that poor nutrition will hinder his recovery. I ask a nurse to insert a feeding tube, but he pulls it out half an hour later. Another is inserted, carefully taped down, but he wriggles a finger underneath and yanks it out. Third time lucky…we secure the tube with a ‘bridle’, a small clip that attaches the tube to a piece of string that runs into one nostril and out of the other. It is now impossible for him to pull out the tube without traumatising his own nasal septum. The bridle is a standard technique, and not as unkind as it sounds. Thankfully, he stops pulling the tube and starts to receive proper sustenance.


On the sixth day we are told that his oxygen levels have fallen again. Another x-ray shows that the white shadow, the patch of pneumonia, has enlarged to occupy half the left lung. The new antibiotic has not worked. We make a decision – on the ‘ceiling of care’. We must decide, for the benefit of all who might be called to see him ‘out of hours’, how far things should go.


We have a high dependency ward, where he can be more closely monitored and receive assisted ventilation with a portable machine. The next step, should that fail, would be intensive care. That is a big step. It is run by another group of doctors the ‘intensivists’, who must agree that the patient is a ‘good candidate’, ie. has a reasonable chance of surviving. The treatments they can provide are often miraculous, but they are also invasive. In cases of severe pneumonia the only real option is to sedate the patient, insert a tube into the trachea, or windpipe, and do all the breathing for them. But to recover from this period of complete physical dependence one needs a reserve of bodily strength. Patients weaken on ICU and to admit someone when you know that the chance of recovery is slight is hard to justify. Not only do they need to be strong enough to survive the actual infection, they must have enough lung function left to allow the ventilator to be detached before it causes its own complications. Then they must be able to get back on their legs and back home to live a tolerable life. This is no easy judgement, and must be done without bringing your own set of values to their projected quality of life. It is safer to just concentrate on the physical recovery and whether they will survive or not.


We make a decision that going on a mechanical ventilator is a step too far. He is transferred to the high dependency ward and a tight mask is placed around the nose and mouth so that the portable ventilator can push oxygen enriched air into his lungs under pressure. We review all of the blood results and cultures and see if the culpable bacteria can be identified, but there are no clues. There are stronger antibiotics out there but we keep our faith with the second line choice.


Next day he is barely conscious. For some reason the chest infection has advanced through the lung despite everything we have done. I begin to wonder whether the ‘burden’ of treatment is justified if the truth is that he cannot survive this illness. And this is where we begin to discuss end of life management. If it is clear that he cannot survive based on the fragility of his lungs and the rapidity of the advancing infection perhaps it would be kinder to stop pretending that we can reverse things and start to think about his comfort.


We meet with his son has now come down from the north of the country. It is clear that the relationship is not a straightforward one. He has not seen his father for 18 months, and the interaction is not a particularly loving one, I can tell. Not all families function according to the ideal. The son is clearly uncomfortable during the discussion, as though we are asking him what to do. That is not the intention. We need him to understand the reasons for our decisions to make sure that he does not strongly object, and to allow him to tell us what his father might have wanted. He agrees with our plan.


So at the end of the 8th day we surrender, on the patient’s behalf, to the illness that was once called ‘old man’s friend’. A nurse detaches the mask and rests his head back against the pillow. We look at the drug chart and cross off the antibiotics that were intended to save him. He has a small tube one of his wrist arteries and I instruct the nurse to remove it. We request that he has no more blood tests and that the only injections he receives are those that make him more comfortable. I write up small doses of morphine and midazolam, a painkiller and sedative, to be used if he appears to be distressed. If his cannula comes out he will not have another one inserted, even if this means he does not get any extra fluids. His veins have become thready and bruised, it would take the junior doctor four or five attempts to get one in. The drugs that he may need can be injected under the skin.


We start the Liverpool Care Pathway. In this way any doctor or nurse asked to see him overnight will not have to guess at our intentions and conclusions, but will know that active or ‘heroic’ treatment is not appropriate. He is, of course, made DNAR…for after all, we have accepted that his heart will soon stop and any attempt to restart it would be indefensible. But even at this point I ask myself – are missing a chance to save him? He was clearly a tough chap. A week and half ago he was living alone, happy, doing the things that amused him. He hardly ever saw a doctor, as shown by the slim medical folder that has been retrieved from records. The cigarettes had never really bothered him. Perhaps if we continued to treat, or if we reached for the even stronger antibiotic, we might yet reverse the infection. Nothing is certain. Would the patient want us to continue with this aggressive policy in the face of almost certain failure? Would he want his sore nose and chin to be continually rubbed by the tight mask, to have blood tests twice a day, the feeding tube in his nostril, if the chance of success is only 5%…or less? I don’t know. His son doesn’t know. We make the decision because we have seen this situation hundreds, thousands of times.


Perhaps, by stopping the antibiotic, we will let the infection progress more rapidly still. Perhaps, by stopping the fluids, his blood pressure will go down and his heart stop a little earlier. Perhaps the tiny dose of morphine that he receives overnight will slow his breathing even more, bringing forward the final collapse. It is possible that he may survive another four days if we continue aggressive therapy, but only one day on the LCP. But those extra days will be full of discomfort, and that suffering will achieve nothing. Those three days will represent the prolongation of death, not life.


[This case is fictional]

Quality of life assessment and DNACPR decisions: unsafe, but unavoidable

There has been much discussion about ‘quality of life’ (QOL) judgments in DNACPR decision making recently. As the doctors censured for including ‘Down’s syndrome’ on a DNACPR form have learnt, it is never wise to refer to QOL. Such an assessment implies that the patient’s life, either now or following resuscitation, is insufficiently ‘good’ to warrant continued existence. That is clearly inappropriate. It is also perplexing, because I remember clearly that DNACPR forms used to include a box labelled ‘poor quality of life’ for us to tick!  The habit of considering QOL remains with us – and I would contend that it is important we do not ignore this aspect, for the assessment of quality remains highly relevant.


It is far safer to use ‘CPR likely to be ineffective’ as justification for a DNACPR decision. This is a far more concrete. The patient is debilitated, has a weak heart, terrible lungs…you can confidently predict that cardiac massage or electric shocks will not restart the heart. A binary outcome…alive or dead, no grey areas, no pejorative forecasts. Not so. The fact is that in making this ‘efficacy’ prediction we do in fact make a QOL calculation. ‘Ineffective’ merges with ‘poor QOL’, and they cannot be separated.


The overlap exists because resuscitation frequently does result in a restoration of the heartbeat. Life continues, but unless the patient wakes quickly, is able to breathe off a ventilator and has a blood pressure sufficient to perfuse their organs without infusions, its quality will indeed be poor in the short term. This is a worthwhile sacrifice if the end result, after days, weeks or months in intensive care, is discharge home to the care of loved ones, but statistics show that this is rare. So, although doctors may say that resuscitation will not work, they know that in a reasonable number of cases it may in fact cause the heart to restart. But they have made a judgment that the burden of treatment, be it the initial resuscitation or the inevitable admission to intensive care that must follow, is not outweighed by the potential benefit.


We can debate the fallibility of quality, and criticise those who fall into the trap of referring to it on paper, but we are dishonest if we deny making quality judgments when making DNACPR decisions.

Medicus Mendax: false final words

 I have been in this position a few times now.  To protect confidentiality, the following account is inspired by a number of experiences and does not describe a specific patient.


A young patient, well known to me, was admitted to the hospital once again.  She had become severely jaundiced and the cause, as usual, was heavy drinking.  Just a month ago she had spent three weeks in hospital being fed via a nasal tube and receiving steroid tablets to calm down the dangerous inflammation in her liver.  This time she was even more unwell.  Her arms and legs were thin, her abdomen was swollen and she was confused.  But, as before, after a few days her strength began to return and the jaundice improved.  Her family came in.  They tried to convince her to stay at home with them in order watch over her, but she preferred to live in a shared house, with her friends, friends who supplied her with alcohol.


As I left one night I saw her sister walk onto the ward.  ‘Pretty stable.’ I told her.  We knew that she was unlikely to survive more than six months given the gradual decline in liver function, but the time that she spent outside hospital was satisfactory to her, and not without pleasure, or freedom.  We had made every effort to help her recover from alcoholism, but to no avail.


As I drove in one morning I received a phone call to say that she had vomited two pints of blood on the ward.  I parked the car and rushed in, knowing that she would need to be admitted to the intensive care unit for resuscitation and an emergency endoscopy.  She lay on her side, a puddle of congealing blood extending from her head to her chest on the bed sheet.  The floor was sticky with it.  She was groaning, but still conscious, fully aware of what was going on.  The curtains had been pulled around all the other patients’ beds so that they did not have to witness this terrible scene.  


Within fifteen minutes she had been transferred to the intensive care unit and preparations were made to sedate and intubate her.  This would allow me to perform an endoscopy without her struggling and moving around.  I stood over her and she looked up.  I explained what we were planning to do.  She said, ‘Doctor, just tell me I’m going to wake up from this.’  I hesitated and said, ‘Yes, we should be able to stop the bleeding and stabilise you, you should be awake in a couple of days.’  An oxygen mask was put over her face and the first sedative was administered. 


I had lied to her. I knew her liver was probably not strong enough to sustain her through this massive haemorrhage.  Privately I gave her a 10 to 20% chance, at most, of pulling through.  A large enough chance to justify aggressive treatment, especially for such a young patient, but a long shot.  As the anaesthetist began to insert the endotracheal tube a dark fountain rose up from her mouth and poured onto the already sodden pillow.  Twin rivulets of blood fell from her nostrils.  Her hair was matted with it.


When I passed the endoscope into her oesophagus all I could see was red.  I then inserted a balloon into the stomach to squeeze the engorged vessels from below, but her blood had become too thin and she continued to bleed.  I tensed the balloon as much as possible, but over the next two hours she lost more blood and her heart began to fail.  The family arrived and I spoke with them.  Death was now certain.  She died half an hour later.  It was no surprise – not to me, not to the nurses, not to the other consultants who looked after her in the past.  But I could not forget that one of her last human interactions was with me, when I gave her reassurance that she would survive.  I wonder if, as the sedatives kicked in, and the sound and the meaning of my words drifted across her darkening mind, they offered any comfort at all.