By Adem Lewis / in , , , , , , , /


I’m Dr. Jeff Burns, Chief of Critical Care
at Boston Children’s Hospital and Harvard Medical School. Jeff, thank you very much. Hi everybody, my name is Peter Laussen, Chief
of the Department of Critical Care at the Hospital for Sick Children in Toronto and
I want to welcome you to a very special World Shared Practice Forum. This is actually a repeat of a lecture given
by Dr. Brian Kavanagh some years ago, and very sadly Brian passed away in June of this
year. Brian leaves a huge legacy as a clinician
as a friend as a scientist and as a teacher to so many of us across the world, and Jeff
thank you for reshowing this very pivotal video that he did around protocols and procedures
some years ago. It really exemplifies Brian, the way he thought,
and the contributions that he makes and I think this particular video will stand the
test of time. You and I were talking that it is classic,
it is timeless, and he challenges conventional wisdom in a way that
all of us need to really absorb. It’s very true, and that was Brian. Brian was a staff physician and held the Geoff
Barker Chair in Pediatric Critical Care Medicine at the Hospital for Sick Children. He was a senior scientist in translational
medicine in the Research Institute at the Hospital for Sick Children and professor of
anesthesia at the University of Toronto as some background, Jeff, he graduated from the
College of Dublin in 1985, completed his residency in internal medicine in Dublin, and then his
residency and fellowship training in Toronto in anesthesia, and then following that his
critical care training in Stanford. He came back to Toronto in the mid-1990s to
Toronto General Hospital and then in 1999 moved to Sick Kids where, as I said, he was
a very prominent clinician scientist. He has had a major impact on the career of
so many people. As a scientist, his laboratory investigated
mechanisms of lung injury, actions of carbon dioxide on the lung, and more recently on
better understanding of why patients with ARDS develop changes within their lungs in
association with that disease. And a new mode of–potential new mode of
mechanical ventilation. So Brian was very innovative, he had a very
active lab and I know is going to be missed among the scientific community because of
that. He completed two terms as Chair of the Department
of Anesthesia at the University of Toronto and that’s a very prominent position. During the 10-year tenure he really changed
the way in which research was conducted across the department of anesthesia. And now you start to see the product of that;
bringing the next generation through of anesthesiologists and scientists, and some of the work that
they’re now doing is just first-class, and a credit to Brian’s vision in bringing that
forward. Brian was a wonderful educator. Those of us that had the opportunity to see
him lead a team, present our backgrounds, we understood very clearly his attention to
process, his attention to detail, and it comes out in this video. He made sure that we understood what the problems
were, that we critically appraised what they were and were precise in our thinking. And they were always very good lessons and
as I said you’ll see that in the video. Finally, Brian was a dear friend. Not only to us at Sick Kids and to me personally,
but across the globe. He was a person of integrity, of such fine
character, and of wit that was so engaging, and so generous to people. No matter what they asked for, he would spend
the time to try and help you. And it wasn’t just within our own environment
at Sick Kids were we miss him dearly, but it was the impact he had across the world
as well. So thank you for giving me the opportunity
to introduce this video, to give testament to the career, and
the person who was Brian, and I hope you will enjoy this video, thank you. Peter, wonderful tribute to a wonderful colleague. Thank you. Welcome to World Shared Practice Forum. I’m Dr. Jeff Burns, Chief of Critical Care
at Boston Children’s Hospital and Harvard Medical School. We’re very happy to have with us today Dr.
Brian Kavanagh. Dr. Kavanagh is Professor and Chair of Anesthesia
at the University of Toronto and the Jeffrey Barker Chair of Critical Care Medicine at
the Hospital for Sick Children in Toronto, Canada. Brian, welcome. Thank you very much indeed, Jeff and colleagues,
for this invitation. I’m going to speak on the topic of guidelines
and protocols and bundles. This is an important issue in contemporary
medicine, particularly critical care. And in fact, in many ways, it’s an emotional
issue. I have no financial conflicts of interest
with any of the subject matter of this talk. We’d like to start by turning to our colleagues
around the world to ask a question. In your answer, please first state your city
and country location. We are interested to find out what clinical
guidelines or protocols you are following in your pediatric intensive care unit. For example, do you have a protocol or guideline
in your PICU for the treatment of sepsis or the management of sedation and analgesia,
or any other protocols or guidelines that you’d like to share with us? One of the first things to get right out there
is, when you question guidelines, you upset people. People have careers in making guidelines,
and other people have careers in making people stick to them. So when you question that paradigm, you shouldn’t
be surprised that people will get somewhat upset. Let me start off with two very well-publicized
protocol guideline combinations. The first is a checklist. It’s a bundled aggregate of items that should
be checked before surgery. This is of worldwide importance, sponsored
by the WHO. This is the well-known surgical safety checklist. It involves items that should be checked off
before a patient undergoes surgery, and the idea is that if these elements are conformed
to, that the outcome will be better. Obviously, everybody would want outcomes to
be better. So Haynes and colleagues– and the senior
author was Atul Gawande– published in 2009 in The New England Journal the surgical safety
checklist to, as they say, “reduce morbidity and mortality in a global population. The first table I show you here is Table
5 from that paper. And what you can see is eight hospital sites
across the world. And in each of these, you can see that the
outcomes are improved. When you add all the outcomes together on
aggregate– surgical site infection, unplanned return to the operating room, death, any complication–
there’s about 1/3 less complications in those hospital centers after implementation of the
guideline checklist compared with before it was used. You would think, well, that’s pretty good. There are no medicines in modern medicine
that would give anything like a 30% reduction in complications. So you definitely want to understand this
approach. Many pragmatists in our world would say, look,
with outcomes as good as that, nothing more needs to be said. Just do it. That’s tricky, because anything that’s important
is worth understanding. Anything that’s to be understood needs and
is worth a closer look. On Table 6 of the same paper, what you can
see is how much implementation of the key processes was undertaken. On the left column, you see the eight hospital
sites. And then you see checking of the airway, checking
of the pulse oximetry, adequacy of intravenous anesthesia, intravenous access, airway assessment,
et cetera. And overall you can see that there’s lots
of implementation. That’s a good thing. What the authors did not do was to match up
implementation and outcome. If the elements of the checklist had a cause
and effect relationship with outcome, then you would anticipate concordance. Centers that had more compliance would be
the centers that had better outcome. Similarly, centers that didn’t have more compliance
would perhaps be the centers that didn’t have such good outcome. That doesn’t prove cause and effect, but it
is a necessary ingredient for there to be cause and effect. So when you cut and paste those two tables,
the outcome and the implementation, you can see for the eight sites several different
patterns. First are those centers that have significantly
less complications, that is to say, the desired outcome. And you can see there are three such centers. And then you can say, well, which are the
centers that had more implementation? And you can see that there are five such centers. And then you can see which are the centers
that had both more implementation and less complications, one center. So what you have from this surgical safety
checklist is no concordance. In fact, there’s almost perfect discordance. Don’t get me wrong. I strongly recommend every element of that
checklist. The idea that you would not check adequacy
of intravenous access or do an airway assessment is ludicrous. But you should not think that checking off
those elements on the checklist explain the better outcome, because quite clearly they
do not. I’ll move gears from perioperative anesthesia
care and outcome to sepsis, very relevant to emergency medicine and critical care. This is a study by Ferrer and colleagues published
in JAMA in 2008. It describes a process of care, implementation
associated with better outcome in Spain. And what you can see here is the survival
curve. On the y-axis is probability of survival. And what you can see is that after the intervention,
the survival was slightly but significantly greater than before the intervention. The hospital mortality that was the marker
of outcome was 44% before the bundle was implemented, and that was reduced to 39.7%. And you may say, well, that’s not much of
a reduction. Dead wrong. Sepsis is extremely common, and with a mortality
of 44%, any reduction is extremely important. If this were a cardiology study, this would
be far and away the most significant finding of the decade. So it’s a small effect, and the intervention
may have helped. But remember, like the surgical safety checklist,
this is a before versus after study. The process of care that they’re talking about
is a two-part care package. The first is a resuscitation bundle, the things
that are done initially in the early hours of care. This is followed by a management bundle, specific
medical interventions that occur after resuscitation. What you can see here on Table 2 is the compliance
beforehand and after, post the implementation of this care bundle. In resuscitation, what you can see is that
yes, the lactate was measured more often, blood cultures were drawn more commonly, fluids
and vasopressors were used more, as was use of a central venous pressure to a target level,
as was mixed central venous O2. In general, there’s more compliance and more
use of the components of the resuscitation bundle after the educational program was implemented. So this could explain the better outcome. Overall, you might say there is more resuscitation
after this bundle was instituted, and therefore maybe it explains or contributed to the better
outcome. Therefore, it gets a tick mark as a possible. It’s definitely a potential player in explaining
how outcome following sepsis was improved by this implementation. When you look, on the other hand, at the management
bundle, unfortunately the situation is not so clear. In fact, it is clear. It’s extremely clear, because all elements
of this bundle either don’t work– such as use of generic low-dose steroids, use of activated
protein C, use of tight glucose control, use of plateau-pressure control, because it was
the same before and after– are all management bundles. So none of the management elements can explain
the improved outcome, because either they’ve been subsequently shown not to work or they
weren’t used any more frequently afterwards than before. So therefore you would say, well, this is
a complex package, editorialized in JAMA, high profile, but only half of it could explain
the better outcome, and that’s the resuscitation half. So the authors here, I think maybe unwittingly,
explain why in fact, this is not the case, because they give us the long-term data. They give us data follow-up at one year. And what they show here, that the mortality
got better at time of intervention, and at one year stayed better. That’s good. You say, well, this is the kind of thing we
want. Institute something, and one year later it’s
still working. Excellent. And it is excellent. However, it’s not explained by the management
bundle, because we know the management bundle doesn’t work. None of the elements could possibly explain
the improved outcome. Therefore it must be the resuscitation bundle. Well, that doesn’t work either because the
resuscitation bundle was more implemented at the time of initial improvement, but by
one year had gone back to pre-implementation levels. So the resuscitation cannot actually explain
the sustained benefit. So the lower mortality in this high-profile
study does not relate to the studied interventions. This is not because of what was done in that
study. This is some other effect. But all mortality is very likely to be due
to an improvement in overall standards, so-called secular trend, or to something unrelated,
so-called Hawthorne effect, the effect of being watched or being studied. Don’t get me wrong. I strongly, strongly recommend expert presence
and expert resuscitation of any individual patient who might have an infection. I don’t disregard that for a second. I strongly emphasize it. But when you’re looking at before and after
studies, particularly those that are widely cited and highly influential, where there
is a clear and unambiguous association with benefit, they are equally clearly not representative
of cause and effect. And so you must be very clear on relying on
those interventions, because in fact, the benefit is due to something else. What is the Hawthorne effect? Elton Mayo, psychologist, social psychologist,
examined productivity in the Hawthorne Works plant in Western Electric, Illinois, a good
place to examine things, I guess, because he noted that when you dim the lights or when
you raise the lights, the productivity increased under either set of changed conditions. And the productivity increased if all of these
women knew they were being watched. And that’s the same of all of us. When we feel we’re being watched or being
studied, we’re more productive. Or it could be an impact of an improving baseline
outcome. In fact, recent data suggesting a far lower
mortality associated with systemic sepsis than was previously thought either reflect
better ways of measurement or true improving baseline. Many people will say, look, this is all theorizing. You’re just being unnecessarily intellectual. Who cares how this works, if it does work? I would say, be very wary, be aware of this
so-called pragmatic approach. It has potential consequences. You may continue the harmful or useless elements. Examples in the sepsis would be activated
protein C, tight glucose control, or routine use of corticosteroids in septic shock. Thank goodness we now know that they don’t
work. In fact, they’re harmful. The second issue is that you may ignore or
discontinue the factor that does work. Why? Because you have not identified it. Third, it’s really hard to advance knowledge
if the issues are hidden. Knowledge requires things to be explicit. So you can form a hypothesis and test it. Doesn’t have to be a randomized trial, but
unless you can identify what your question is, you can never answer it. An important element in figuring out whether
a protocol is appropriate for your population is the concept of alignment. It’s possible the protocol is exactly what
your population requires. But if it isn’t, you run the risk of misalignment. This is a Gaussian curve showing normal distribution. What you can see on the x-axis is performance. Some centers are very high performers. I won’t say where those centers might be. Could be Boston, but I won’t say. Some centers are poor performers, and I certainly
won’t say where they might be. And it’s very obvious, it’s quite intuitive,
that if you’re a poor-performing center, poor performing because you have insufficient staff,
insufficient education or basic training, or do only very simple cases, that the kind
of training or protocols that will work well there would be to improve basics training,
to improve basic staffing, to use protocols such as BCLS, for example, and to focus on
simple cases. Clearly not very useful in those situations
to use a protocol for ECMO, or high-level care. In contrast, in the high-performing centers,
you may choose advanced training, you may choose advanced fellowships. You would have extensive experience, tertiary
or quaternary care. And in those situations, it would be important
that the protocols, where they’re used, reflect the ability to use high-end performance. BCLS may have a role, but BCLS would be a
sort of a ceiling effect in that context. So different problems, like much of life,
need different solutions. It’s possible that in translating a protocol
that works well in one context to a different context, you can have trouble. Now let me give you three examples where this
can be a problem. I’ll frame this in the way of three questions. The first is, could a bundle, group of elements
of care, that are developed for a sophisticated context, could they work well in a more rudimentary
context? Well, here’s the Kathryn Maitland study, the
study of mortality after fluid bolus in African children with severe infection, a study in
sub-Saharan Africa. Children with severe sepsis were treated according
to an algorithm. And you can see here, there were several thousand,
almost 8,000 children screened. There were 3,000 children enrolled. And if we ignore the patients in frank shock,
we’re left with 3,000, approximately 1,000 in each group, assigned to fluid bolus with
saline, fluid bolus with albumin, or no fluid bolus therapy. And the no fluid bolus therapy reflects usual
care in sub-Saharan Africa. Note the bolus therapy reflects usual care
in the developed countries, Boston, Toronto, Pittsburgh, CHOP, everywhere. That’s what the bolus care reflects. Well, it turns out, when you look at this
graph, you think that must be mislabeled because the mortality over 28 days is lower for the
usual care. In fact, implementing the Western world standards
of bolus therapy for sepsis, whether it be saline or albumin, was associated with a slightly
but significantly worse mortality. So instituting Western standards, in part,
to these children in sub-Saharan Africa, in fact worsened the outcome demonstrably. Why was that? Well, we don’t know. Was it the age of the kids? Well, no. We have young kids in all the world over. Was it the presence of malaria? It could have been. Was it the preponderance of anemia? Quite possibly. Was it the lack of access to mechanical ventilation? Quite possibly. Was it ventricular overload? Maybe. Was it the chloride load? Who knows? There’s clearly a reason, because this is
not a statistical blip. This is a misalignment of protocol intent
that seems to work well in the Western world but clearly causes harm in a more rudimentary
setting. Overall point, different context, and there’s
a different setting. So therefore, a bundle of care developed for
a sophisticated context might not work well in a more rudimentary setting. I’ll ask a second question. It’s sort of the first question in reverse. Could a rudimentary bundle impair performance
in a more expert setting? Well, how could that be? Here’s a study, and it’s far from sub-Saharan
Africa. This is a study from Germany. Extremely wealthy country, extremely high
standard of medical care, a study looking at individual optimization of hemodynamic
care in adult patients after cardiac surgery. They had 100 patients. It’s a small study. It’s randomized strength, but it’s non-blinded,
single center. So many flaws. There were two hemodynamic treatment protocols. But these protocols were implemented by the
same very sophisticated team of intensivists and surgeons. The simple protocol was an algorithm consisting
of mean arterial pressure, CVP, and heart rate. That’s all they had to go on. Of course, in most circumstances, you can
look after most patients. Maybe, maybe not can you look after the particularly
rare, very ill patients. With the comprehensive protocol, there were
more elements. There was mean arterial pressure, heart rate,
and CVP. In addition, there was end diastolic volume
index, there was the cardiac index, and there was extravascular lung water. So simple versus comprehensive. What could you see? The overall complication rate, in fact, was
higher if you were restricted to the simple protocol. Indeed, the ICU length of stay was also longer
if you were restricted to the simple protocol. There’s no difference in outcome, but there
never would be a difference in outcome in a small study like this. So there’s a possibility that yes, indeed,
a rudimentary protocol may dumb down expertise in a sophisticated setting. There’s a third way that bundled care or protocols
can be lost in translation, and it’s framed in the following question. Does care that works in one sophisticated
context always work well in a different sophisticated context? Here’s the Van den Berghe study of tight glucose control. You can see here 1/3 fewer deaths in a predominantly
cardiovascular surgical adult population, non-blinded study. This is an extraordinary finding that was followed
up by multiple RCTs, including the NICE-SUGAR study. The approach certainly works in Leuven. The data are true, the statistics are accurate. No doubt about it. The question is, when you extend this to a
more generalized setting, that is to say, Australasian and Canadian intensive care units,
what you in fact find is the exact opposite. You can see that the conventional glucose
control is associated with a higher probability of survival. That is to say, the tight glucose control
made matters worse. This is a problem. I’m not a health services researcher, but
if you imagine 6,000 patients, excess mortality in the NICE-SUGAR study of 2.6%. At least now we know the answer, that in adult
medical surgical ICU patients, we should not use tight glucose control as it was used in
the NICE-SUGAR study. But if you imagine in the US, in adult care
there’s about 5,000,000 patients in ICU. About 20% of those are supported with mechanical
ventilation, and thus would have been eligible for the NICE-SUGAR study. Think about it. If tight glucose control caused deaths in
2.6% of patients, that amounts to 2.6%, of 20% of 5,000,000 adult patients. That’s about 26,000 deaths a year in the US
alone. As I say, I’m not a health services researcher. Those data are very crude, but it’s not a
good signal. So care that does work in one sophisticated
context does not always translate to another sophisticated context. We’d like to turn now to our colleagues around
the world and ask a question. In your answer, please first state your city
and country location. The question is, when a study promoting a
protocol or checklist is published in the literature, is there a formal review process
in your PICU to determine whether or not to implement the protocol or checklist? If so, what does this process consist of? There are many, many situations when a protocol
is a very good idea, I would say when a protocol is mandatory. What situations are those? Well, one is where the concept is inherently
simple. If there are only three or four steps to take,
do them the same way every time. If the knowledge is explicit, that is to say,
everything contained that needs to be done can be specified in an explicit way, that
might work well for a protocol. And if there is a need– if there is a need
to minimize variability. Here is the result of a task forcethat I sat
on from the American Thoracic Society and Society of Critical Care Medicine and the
ACCP. And it outlines situations that favor the
use of the protocol much along the lines that I just said. In contrast, situations where protocols are
not such a good idea, it’s where you have competing goals for optimization, where you
have complex management, where you have processes that are uncertain or where knowledge is implicit. Simple idea. Titrating rocuronium neuromuscular blocker
could absolutely be protocolized. Setting up hemodialysis. The last thing you ever want is a resident
or a fellow in the middle of the night deciding a new way of setting up a hemodialysis circuit
or an ECMO circuit. That’s insane. You just have to had seen something go wrong
with one of these circuits, and you know absolutely there is one way to do it. Martin Tobin, in an editorial, talked about
the importance of rigidly adhering to protocols in research. He says the protocol in research must be followed
exactly. There can be quote, unquote, “no flexibility,
no weasel words.” It’s kind of harsh. He says the insight gained is the point of
the research, not the protocol. The protocol is there so that when you read
the paper, you know what was done. If they didn’t follow protocol, you simply
don’t know what was done in the research study. And therefore, the research, in a way, might
as well not have been done. He goes on to say that implementation of the
research findings is a whole different issue. You are using the insight gained from the
protocol and the study to manage a patient who may or may not have fitted into that study. When is a protocol not a good idea? Well, there are certain situationswhen it’s
a bad idea. One, if the idea is too complex to be protocolized. Second, if the knowledge inherent is tacit. And third, if you desire variability. Many people say, how on earth would you desire
variability? Let me explain. Tacit knowledge is knowledge that is difficult
to put into words. Anything that requires practice is likely to
have a high tacit knowledge component. Here’s a woodworker or metalworker working
with a lathe. Well, you can instruct someone how to use
a lathe, but if you’ve ever used one, you’ll need to practice for many months before becoming
competent, much less expert. This is the aperture in a flute. Many of you who play the flute know that getting
high end fingering is really not that hard if you have some talent. But the hallmark of a superb flautist is tone. And this is a picture of James Galway, one
of the better flautists in the last 50 years. And you can basically follow those instructions
till you’re blue in the face, but you will never, ever sound as good as James Galway. Why? Because he has innate talent, and he’s practiced
like crazy. And those issues reflect tacit knowledge. Here’s a BMX biker. Try showing a protocol that will reproduce
this sort of agility. Tacit knowledge, you may say, is all very
well for playing the flute and riding a BMX bike. But is it relevant to medicine? Well, William Osler said– and you see the
pictures here– use your five senses. “Learn to see,” he says. “Learn to hear, learn to feel, learn to smell. And then know that by practice alone can you
become expert. You must learn these techniques, and you must practice them. There are unintended consequences of protocols. Here is a case report by Polgreen and colleagues,
and they describe an outbreak of Clostridium difficile. And they relate that to inappropriate antimicrobial
therapy for community-acquired pneumonia. They point out that the hospital had a protocol
to treat better patients with community-acquired pneumonia, and that resulted in a massive
over-diagnosis of patients who didn’t have it, and a massive over-prescription of antibiotics
to those patients so that the individuals in the emergency room would be compliant with
the new protocol. Half of the patients who received treatment
for bacterial pneumonia may not have had pneumonia. And excessive use of the new pneumonia care
plan during the influenza season contributed, they think, to the intensity of the outbreak. That means a 10- or 11-fold increase in cephalosporin
use, and a fivefold increase in risk of death from C-difficile infection. That’s a very, very unfortunate, severe, unintended
consequence of what was a laudable goal, to treat better community-acquired pneumonia. And you say, well, guidelines work. They improve care. They standardize care. Well, not necessarily. This is a paper from Anesthesiology earlier
this year, Kappen and colleagues in the Netherlands. And they randomized patients to basically
care by usual standards versus care by the introduction of a protocolized approach. And what you can see here is an overlay here. The standards of care result in a good matching
of the need and the prescription of antiemetics. And on the right side you can see a mismatch. What you see here is an excess of prescription
of the antiemetics. So the antiemetic guideline here in fact resulted
in more antiemetics but the same amount of emesis. In critical care, Gordon Doig and colleagues
in this paper in JAMA, again in 2008, again editorialized with a laudatory editorial,
looked at the evidence, use of evidence-based feeding guidelines on mortality in critical
care adults. And what you can see, they had 30 plus ICUs
that were randomized. And 27 ended up being randomized in the cluster
randomized trial. And they had important endpoints. In those that followed the guidelines, there
was a higher proportion of patients receiving enteral or parenteral nutrition. So more patients were fed. Well, that may be a good thing. And you can see the energy delivery per patient
was also slightly higher. But once you look at the outcomes,you see
there’s absolutely no difference. In fact, if anything, maybe a tendency towards
a worse outcome in terms of mortality or mean hospital length of stay. So what you have with feeding guidelines is
more feeding and the same outcome. Can we trust guidelines? Well, I think the integrity and the intent
of most people with clinical practice guidelines is positive. But there are situations where you have to
wonder. I have no problem with pharma, and pharma’s
need to survive and grow. It’s a very important part of modern medicine. When you click on Guideline Central and you
click on Pharma and Device sections, you see here society-endorsed guidelines recommending
your product. Quote, unquote, “What more could you ask for?” This is taken directly from the website. So that’s very desirable if you’re selling
a product. May not be desirable if you’re a patient. I do have confidence in the future in this
regard. Physicians choose critical care after a primary
specialty, pediatrics, anesthesia, medicine, or surgery. Why? Because among these specialties, there is
a rich, dense heritage of knowledge that they’ve immersed themselves in. And this is very important. Patients and their families will expect that
physicians have received an in-depth education, and that they will understand the patient or that
they will seek to understand the patient. That doesn’t jibe well with checking off a
guideline list. Why textbooks, and why in depth, and why not
an immediate Google- or Yahoo-type search of all of the issues that crop up on ward
rounds? Well, it’s very important. I refer you to the book Proust and the Squid
by Maryanne Wolf. And she decries the instant approach to modern
medical learning. She makes the point, “We are not only what
we read, but we are also how we read.” That is to say, knowledge acquired through
judicious use of texts, going backwards and forwards in an in-depth manner is very different
to the sort of knowledge you acquire when you look it up on a Google-type search and
move on to the next issue. The process, she says, of in-depth reading,
indeed shapes the mind. When critically ill– this is a pictureof
a patient from our ICU who indeed was critically ill. This child has meningococcal septicemia. When our patients are critically ill,they
are absolutely vulnerable. They need all of the tools that we have available. They need our training, they need our knowledge,
they need our experience. They need our wisdom. They need our procedural skills. They need our presence. They need our integrity. And yes, certainly, they do need guidelines
and protocols. Remember, questioning guidelines upsets people. You should expect people to be upset. It’s upsetting. Protocols, for use when procedures are mechanical,
first class. Guidelines, you should definitely use guidelines
if you’re unfamiliar or unsure. But if unfamiliarity or uncertainty are a frequent
occurrence, then you should probably consider textbooks, expert clinicians, and more time
learning. Thank you. Dr. Kavanagh, you have many decades of experience
as a clinician and as a scientist. And one of the questions I’m sure our colleagues
around the world are wondering is, how do you stay current in the field? It’s a challenge for all of us. It’s a challenge for people who are learning
the field, and it’s a challenge for those of us who are practicing in the field. And that is, how do you stay current with
the literature? Could you tell us first, what is your reading
plan? How do you stay current? Do you have a rigid plan that you stick to,
or is it ad hoc? And secondly, could you share with us, if
there were three, four journals that you would recommend to colleagues around the world that they read
frequently, what would they be? It’s very unkind of you to say that I’ve several
decades of experience. We both do, Brian. But I’ll take the compliment. I think the key issue of a reading plan– I
think that’s difficult. I have a very varied professional life, and
so it’s quite hard for me to have particular time during a day or particular days during
a week. But I kind of make sure, on aggregate, that
I at least skim the general medical journals, the five main general medical journals. And I skim also, frankly, all of the critical
care journals. And you might say, well, that’s a very large
reading load. Yes, it is. I don’t– maybe it just reflects my bias–
I don’t skim table of contents for pediatric journals. I probably should. But I am drawn, of course, through secondary
reading to much in the pediatric literature, particularly pediatric critical care. So I read the New England, Lancet, British
Medical Journal, Canadian Medical Association Journal, JAMA. I almost always read the table of contents,
and anything of remote interest, I try to make time to actually read it. I think in critical care, I read the blue
journal, American Journal of Respiratory and Critical Care Medicine, I read Intensive Care
Medicine, Critical Care Medicine, and Critical Care. Do I read every paper? No. I look at the title, and many times I look
at the abstract. But I sort of make a big effort that if the
paper is remotely of interest, either in my area or of interest even not in my area, but
shows an important mechanism or important perspective, that I will often print that
paper out. May be a waste of paper, but I think if it’s
important, you should probably print it out. And that’s a controversial thing in a way,
because you have a lot of paper. I accumulate a lot of paper. But yeah, that’s important. And I’m uncertain as to whether you should
get the journals online or journals arriving in the mail. It’s a huge paper waste for journals arriving
in the mail, but you actually have a far higher chance of reading them. It’s a little bit like the last point I made
in the lecture. If you have a paper journal, you’re much more
apt to flick backwards and forwards, and to put it down and pick it up again. That process doesn’t work so well with electronic
media. We’d like to turn now to our colleagues around
the world and ask you a question. In your answer, please first state your city
and country location. The question is, how do your PICU position
faculty stay up to date with the medical literature? First, in particular, is reading a journal
online or reading a journal in print preferred? Second, on average, approximately what percentage
of your efforts to stay up to date is through reading a journal online, and what percentage
of your effort is from reading a journal in print? How do you stay current, or how would you
strike the balance between your academic focus versus reading outside your academic focus? So you’re an expert, and your research has
been in the area largely of lung injury. But how do you find time to keep up with the
latest in renal replacement therapy, or the latest in bowel ischemia, translocation, et
cetera? I would say, if you talk to my research colleagues,
they say I don’t keep up enough with research. If you talk to my clinical colleagues,they
say I don’t keep up enough with clinical. I keep up pretty well, actually. I think that the only way to do this is to
put in the time, whether you do it every Monday, whether you do it two hours a day, an hour
a day. But you have to put in the time, and it has
to be substantial. And you simply cannot– particularly if you
have a research career, the level of reading there has to be in-depth. In the clinical career, you might want to get
away with just the sort of bottom line take home message. It’s not great, because if it’s central to
your practice, you should really understand it well. But if you’re going to research in the area,
the papers will have to be read several times, put down, taken up again, discussed with your
colleagues, read again, because after all, you’re basing your next hypothesis and your
next body of work, in part, on those papers. So there’s a danger that the answer could
sound too sanctimonious, to say, oh, well, you know, in the good old days or the bad old
days, you just have to read for many, many hours a week. I’m afraid you have to read for many hours
a week. Dr. Kavanagh, another question that arises,
I’m sure, among our colleagues is, what conferences do you make an effort to attend regularly,
and how often? Now for many of our colleagues around the
world, it’s just not practical to be able to get to many of the conferences that you
might mention. But nevertheless, we’d be curious to know,
what do you consider to be the essential conferences to stay current in our field in critical care? Two of them. I said I had no conflict of interest, and
now I do. So I’ll go to the most important conference
first. I think the most important critical care conference
in general is the American Thoracic Society, the ATS. And we’re speaking here as pediatric intensivists,
and one may say, well, there’s not a lot of pediatric ICU at the ATS. And yes, that’s true, that’s true. But it is the conference of which the highest level of science in our field is presented. And that’s very, very important, because whilst
that does not tell you what to do the next day and next week in the intensive care unit,
it absolutely shapes the future of our specialty. So the best thing to do is to go to a conference
that has the highest level of clinical science, clinical science. It’s not reasonable to expect a clinician
practitioner to go to a sort of a very molecular-type conference. That doesn’t make sense. But the American Thoracic Society, I think,
covers it all, but at very high level. There are other conferences that are more
concerned with, I would say, teamwork, and and that’s important. But it’s never really going to advance the
ability to diagnose or develop new treatments. And the second recommendation I have for Canadians
is the conference I chair the organizing committee. That’s the Critical Care Canada Forum. And all I can say is that other people say
it’s very good conference. So that’s a plug. Terrific.


Leave a Reply

Your email address will not be published. Required fields are marked *