Professor Bertrand Guidet is Director, Medical Intensive Care, St. Antoine Hospital, Paris, France and
Medical Director, APHP Hospitals, East Paris.
You have a long-standing interest in ICU care and elderly patients. What prompted your interest?
We should not consider the ICU as a unique unit. It is part of the hospital and there is treatment before and after ICU. This is true for all ICU patients, but particularly for elderly patients. If you deny ICU admission for an elderly patient, what will happen to them? Maybe they will be admitted to an intermediate care unit, to a geriatric care unit or to internal medicine. If you admit a patient to the ICU what will the level of care be, what about: the ethical issues if the treatment is not successful, the discharge criteria, the relationship with the family and the relatives? This is a paradigm for the whole process and also deals with distributive justice. If you have constraints, because you do not have enough ICU beds, how do you cope? You have to make a choice. You will try to choose the patient that will benefit the most from ICU treatment. You might be wrong in both ways. You might admit a patient that should not be admitted, or refuse admission to a patient who could have benefited from ICU care. The issue of elderly patients is of paramount importance because of the elderly population, and because we are at the centre of the process in the ICU – with ethical and financial considerations and organisational issues.
When we talk about the ‘elderly’ patient with specialists the threshold is usually 80 years of age. The threshold should be at least 75, certainly not 65 (in some studies they talk about 65-75 years as ‘young old’). Age by itself does not say anything. The concept of frailty is more important than age, and you should probably collect information that describes precisely the patient you are talking about, including nutritional and functional status, quality of life, family and so on. This is very important information that we are not collecting on a daily basis. We use severity scores such as SAPS II, APACHE, but these do not cover any of the conditions specific to the elderly, such as cognitive impairments, cachexia or depression etc. This information is key for predicting mid- and long-term outcomes. If we want to assess whether an elderly patient should be admitted, we should not only consider immediate severity such as organ dysfunction, but also the underlying disease, functional status and quality of life. In our ICE-CUBI study (Boumendil et al. 2012) we found that if you ask for the information (e.g. “How many drugs do you take, did you fall in the last three months, have you been admitted to the hospital, what is your functional status according to the activities of daily living (ADL) score?”) you are able to collect it, even in an emergency situation. I convinced my colleagues that we need to collect this information for the decision-making process, whether via the patient, the family or the general practitioner. This is key for decision-making.
At the Durban World Congress Ethics Round Table most participants agreed that age cannot be the sole criterion on which healthcare decisions be made (Guidet 2014). They also agreed that it is important to provide data showing that outcome differences between elderly and nonelderly patients are partly related to decisions to forgo life-sustaining treatment. Do sufficient clinical trials include this data currently?
A big issue is related to the admission of elderly patients. Firstly, all the papers on triage are biased, because they only consider the patients that were proposed for ICU admission. They do not consider pre-selection by the patient him/herself, the family, general practitioner or emergency physician and so on. So when you say, “I refuse 30% of the patients”, what is the prioritisation process? Secondly, you will have two different policies: an open policy or a strict admission policy. If you have a liberal admission policy, so that you admit the elderly patient when there is a possibility of improvement, you will have to decide during the ICU stay whether you continue treatment without any limitation or if you introduce some sort of limitations. For example, I favour a liberal admission policy, but after 3-4 days we need to sit around the table with the team, the nurses, physicians and afterwards with the family and ask what we do from here. The opposite, if you have a very strict admission policy, e.g. admit only good candidates with no co-morbidities, single organ failure, who are easy to diagnose and treat, is that you will very rarely have to discuss end-of-life decisions during the ICU stay, but at the same time you will deny ICU admission to some patients that could in fact have benefited. My point is that there is a relationship between the admission policy and end-of-life decision policy while in the ICU.
You are leading the current clinical trial on Impact on Mid-term Mortality of Guidelines for ICU Admission of Elderly Patients Arriving in Emergency Departments (ICECUBII) (Assistance Publique-Hôpitaux de Paris) Please explain how you arrived at your hypothesis, and comment on any early findings and progress of the trial.
The ICECUBII study is funded by the French Ministry of Health. It is a randomised crossover study: half the hospitals will provide their usual standard care and half of the hospitals will admit all elderly patients from the emergency department, given that they have no active cancer, no cachexia and have preserved functional status, as defined by an Activities of Daily Living (ADL) score above 4. The primary endpoint is mortality at 6 months.
According to our previous ICECUBI study (Boumendil et al. 2012; 2011), we did not find any benefits of ICU admission for elderly patients, so there is no ethical problem with the study. We plan to include 3000 patients. We have around 500 left to include, so hopefully the last inclusion will be at the end of the winter in around March 2015. Then we have 6 months follow-up, and we need to clean the database, which will hopefully be at the end of 2015. I don’t have any data related to mortality in both arms. I have some data relating to the percentages of patients admitted to the ICU in both arms, and it’s higher in the intervention arm, i.e. the patients who had to be admitted to the ICU.
In the preceding study (ICECUBI), only around a quarter of eligible patients were admitted to intensive care. Did this finding surprise you?
In fact it’s higher than this. Our starting point was the emergency department. We showed that emergency physicians proposed only one out of four patients for admission. Among the proposed patients one of two was finally admitted, so the total selected is one out of eight, which is a huge triage. None of the papers about triage assess this pre-triage process. I was not expecting such a selection. I expected perhaps one out of four, but not one out of eight.
You have observed that Advance Care Directives, when communicated, are of great assistance to the treating doctor. Should these be more widely used?
Firstly, Advance Care Directives are rare. Secondly, even if they exist they are not very often used, and if they are used we do not really know the impact in terms of outcome. But I think this is a nice piece of information to have. We should not use it as a law, but if it is clear and updated it certainly can contribute to the decisionmaking process. It’s a piece of paper, and it should not substitute for direct conversation and interaction that you need to have with the patient and the family. It’s just one piece of information you need to consider and which may change according to the patient’s condition.
In a recent editorial (Rusinova and Guidet 2014) you say that the awareness of risks/ rewards of treatment and a genuine dialogue between physicians and patients and families are becoming a priority and outweigh the impact of age in clinical decision-making. How can ICU services facilitate this kind of dialogue?
We miss a lot of opportunities for interaction with the patient and their families. We overlook opportunities to get information from the nurses, who have a lot of information through discussion with the patient and the family. It depends on the countries and the situation. Some physicians are very reluctant to talk directly to the patient and the family in order to get information about the willingness to receive intensive treatment. In the ICECUBI study ((Boumendil et al. 2012; 2011), in more than half of the cases the family of the patient was able to answer. Yet they were asked only in 10% of cases. When you consider their willingness to be admitted or not, it has a profound impact on the decision. If the patient says, “No, I don’t want to be admitted to the ICU”, he is not admitted. If a patient or the family says, “I want to be admitted, I need intensive care treatment”, then they are admitted. So we generally do not ask the patient or the family. However, if we do ask them, it has an impact on the decision, so we need to do it. We have a manuscript in preparation about this huge issue.
I would like to emphasise that ICU is a team. If you want to improve the outcome, I don’t feel that it will be a fancy new ventilator or a new drug that will do the job, but the key elements of culture and climate in the ICU, that is if the people working in the ICU have good communication, respect, ability to change and training. If staff are struggling, if they don’t communicate well, if there are, for example, problems with the head nurse, problems with the nurses and physicians, at the end of the process the patient will get suboptimal care. My point about looking at organisation is that it includes how the manager or leader should work in order to engage the whole team, including the cleaners. An ICU has to be clean and well organised, especially because we have to work sometimes in an urgent situation, e.g. cardiac arrest, three patients admitted at the same time. We need solidarity amongst the team and to share the same goals.
It means you need to share the decision-making process. Sometimes you need to sit down and talk about wrong decisions together, maybe in a debriefing session or with the help of a psychologist. For example, we do a common morbidity and mortality review across the hospital.
You organised a European Society of Intensive Care Medicine survey, which looked at the organisation and management of critical care globally. Can you tell us more?
Dr. Yên-Lan Nguyen and I did a survey about different ICU networks across the world, looking at the type of participating ICUs. We have shown that when you have bigger hospitals and bigger ICUs you have better organisation and maybe better outcomes. We have submitted a paper looking at the volume/ outcome relationship, and found that in most cases there was a relationship.
You commented in an editorial on research into night-time discharge (Guidet and Bion 2014), “If we had at our disposal a drug which could reduce mortality by 20-50%, we would all be using it. Avoiding nighttime discharge is that drug.” How can ICUs achieve this?
We wrote an editorial accompanying a paper from the ANZIC group (Gantner et al. 2014) and considering from the literature that roughly 15% of patients are discharged out of hours. This group was more often readmitted and had higher mortality. We need to reduce this percentage that did not decline over time, at least in Australia and New Zealand. In our editorial, we proposed some recommendations to reduce off-hours discharge or to improve the selection process for patients that should be discharged. In most cases, if you discharge the patient during the night, it’s to cope with the pressure, and you need to admit another patient. The response has to be a short-term response. Maybe during the morning rounds we need to select the patient that could have been discharged in the afternoon, or in the long term have more ICU beds, or intermediate care beds. In most cases when you say discharge, it’s discharge mostly from the ICU to the ward.
This interview will appear in ICU Management’s Winter issue, which has a cover story on early goal-directed therapy. What do you see as the main challenges in early goal-directed therapy?
There are two terms – early and goal-directed. Early, I fully agree, the sooner the better. Again it goes with the relationship of the ICU with the other parts of the hospital – the emergency department, the wards, the operating theatre. You need good communication, you need to reduce the delay, particularly for septic shock and anaphylaxis. Goal-directed is much more difficult e.g. for shock. What are the goals, what are the right algorithms? Several papers this year are raising concern about usefulness of the goals. What is optimum cardiac outputs, mean arterial pressure, for example – we don’t know. We need a kind of integrative approach towards the patients. We need first to look at tissue perfusion (mottled skin, cutaneous refilling time, lactate, urine output). Those are simple tools, simple clinical science that you can collect at the bedside. My approach would be much more patient-oriented instead of figures-oriented e.g. mean arterial pressure, it’s nonsense! If there are no signs of tissue hypoperfusion and MAP is 62 mmHg, do we really need to increase blood pressure? And when you talk about an elderly patient with stiff arteries or a patient with previous renal failure maybe they will need more.