|
|
|
|
 Home
Page
Search
Final
Report
Interim
Report
Evidence
- Transcripts
- Witness Statements
- Submissions
- CCNR Report and associated
documents
- Statistical Reports and associated
documents
- Background Papers
- Expert Papers
Inquiry
Seminars
About
the Inquiry
Help
|
 |
Daily transcripts
HEARING SUMMARY
15th September 1999
Hearings continued today with evidence from three expert witnesses: Dr Eric Silove,
Paediatric Cardiologist, Birmingham Children’s Hospital; and Professor Mark de Leval
and Mr Jaroslav Stark of Great Ormond Street Hospital. Dr Silove and Mr Stark are members
of the Inquiry’s Expert Group. Today’s theme was operative treatment and care
for babies born with congenital heart defects. The witnesses focussed on the development
of the arterial switch operation which was first undertaken in the mid 1960s and became
more prevalent during the 70s and 80s in the UK. They noted the reduction in the age of
babies undergoing this particular operation, discussed the risks and potential benefits
associated with it and how these were explained to parents. They highlighted the
difficulty in calculating accurate mortality percentages for a new operation. Then they
discussed the learning curve for surgeons introducing new operative procedures and the
benefits of learning from colleagues already proficient in the new techniques. Professor
de Leval commented on a study he has recently undertaken looking at the effect of human
factors on the outcome of paediatric cardiac surgery. Mr Stark showed the Inquiry a series
of slides to illustrate the developments in paediatric cardiac surgery, including advances
in surgical techniques, equipment, drugs and co-operation between members of the
multi-disciplinary team, which have led to reduced mortality figures. They went on to
discuss the supra-regional status of paediatric cardiac surgery and the optimum volume of
work carried out by individual supra-regional centres. The afternoon session concluded
with a look at Professor de Laval’s 1994 analysis of a cluster of surgical failures.
|
FULL TRANSCRIPT
1 Day 50, 15th September 1999
2 (9.40 am)
3 SEMINAR:
4 CONGENITAL HEART DEFECTS:
5 SURGICAL OPTIONS, INNOVATIONS and RISKS
6 THE CHAIRMAN: Good morning, everyone. Good morning
7 Mr Langstaff.
8 MR LANGSTAFF: Good morning, sir. Sir, today we have two
9 cardiac surgeons and Dr Silove to help us with the
10 particular problems of surgery, the factors that affect
11 surgery, and we are pleased to say, recent research that
12 Professor de Leval has been doing.
13 A word of explanation for the wider audience, who
14 will anticipate, I think, that Professor de Leval has
15 more than his surgical expertise to offer in general
16 terms to the Inquiry, because he of course was involved
17 in a report at an earlier stage into the actual surgery
18 conducted at Bristol.
19 We anticipate that we will hear from him at
20 a later stage on what one might call the "Bristol
21 specific" issues, and those matters of fact with which
22 he was involved.
23 He has been called today, in particular because he
24 has, as will become apparent when he gives his evidence,
25 done a considerable amount of work into risk factors in
0001
1 surgery. He has developed the use of the CUSUM analysis
2 in looking at individual surgeons' performance, and
3 considerable work in whether near misses may have
4 something to say in respect of improving surgical
5 outcome.
6 Although he is called as an expert today, again --
7 this is addressed to the wider public -- he is not one
8 of the panel of experts who is involved in the clinical
9 case review which is under way at the moment, and that
10 is from an abundance of caution, because, of course, of
11 his past involvement in looking at and expressing his
12 view upon a number of the Bristol cases. So he is not,
13 today, giving evidence about any specific Bristol case;
14 he is giving evidence generally as to cardiac surgery,
15 in that he has a uniquely valuable contribution to make.
16 Sir, with that introduction and explanation,
17 I hope, to those listening to this both in the chambers
18 outside and at a distance, may I invite both Professor
19 de Leval and Mr Stark to take the oath, and then I shall
20 invite each of them to tell us a bit about themselves.
21 Professor de Leval?
22 PROFESSOR de LEVAL (SWORN):
23 MR JAROSLAV STARK (SWORN):
24 MR LANGSTAFF: First, Professor de Leval. You are Professor
25 Marc de Leval?
0002
1 PROFESSOR de LEVAL: Correct.
2 MR LANGSTAFF: Would you tell us a bit about yourself and
3 your experience, and please do not be too modest.
4 PROFESSOR de LEVAL: I came to Great Ormond Street in 1972
5 and I was appointed there as a consultant surgeon in
6 1974. I have been there since, as a consultant and now
7 as a Professor of Cardiothoracic Surgery.
8 I have spent the last 25 years dealing with the
9 heart surgery of children, and also introducing and
10 developing a transplantation programme at Great Ormond
11 Street.
12 Besides that, I have been involved in a number of
13 research projects, basic research, clinical research,
14 and more recently I have been interested in the role and
15 impact of human factors on the outcomes of surgery,
16 which is probably the reason for my being here today.
17 MR LANGSTAFF: Mr Stark?
18 MR STARK: I graduated in 1958 at Charles University in
19 Prague. I came for further training to Great Ormond
20 Street in 1965. I had a brief spell doing research in
21 paediatric cardiology in Boston and Harvard, and in 1971
22 I was appointed a consultant in paediatric
23 cardiothoracic surgery at Great Ormond Street.
24 As my colleague and friend Mr de Leval said, in
25 1974 he joined me in our efforts to do what we did at
0003
1 that time. I was involved in trying to develop infant
2 cardiac surgery which, in the early 1970s, was really in
3 its infancy and the risk of those operations was high.
4 I retired in January this year.
5 In the last couple of years I became interested
6 and involved in data collection because I thought this
7 was really a prerequisite for us to develop possible
8 guidelines or standards. That has been one of my main
9 interests in the last couple of years.
10 MR LANGSTAFF: I want to focus in particular upon a number
11 of issues and questions, but perhaps first, if you would
12 begin from the surgeon's perspective in telling us about
13 the development of and the difficulties in development
14 of the arterial switch operation. You have a screen to
15 your right. Would it be helpful to have a diagram on
16 the screen?
17 PROFESSOR de LEVAL: If you have a diagram of the
18 transposition of the great arteries, it might be useful.
19 MR LANGSTAFF: We have two diagrams, so I will show you them
20 both first so you can see them. The first is ES 1/12.
21 PROFESSOR de LEVAL: I understand Dr Silove has gone through
22 the technicality of the procedures yesterday, so there
23 is no point in going through that again today.
24 MR LANGSTAFF: Yes.
25 THE CHAIRMAN: If I may just, in addition to welcoming our
0004
1 experts, make one plea, that of course we will be using
2 highly technical language during the day, and we have to
3 make sure that we make a transcription of what you say
4 for the benefit of all who will come to read it later,
5 so may I ask that one bears in mind the fact that there
6 is a transcriber and when one uses technical terms,
7 maybe go a little more slowly, without at the same time
8 interrupting your flow of thought, I hope.
9 DR SILOVE: May I just interrupt? I thought it might be
10 easier to do the operation than to demonstrate it on the
11 screen, and I did it in the most simple terms
12 yesterday. You might want to just do that again.
13 PROFESSOR de LEVAL: The treatment of transposition started
14 in the mid-1960s and there were two challenges. The
15 first one was to be able to do an open-heart operation
16 on a small infant; the second one was to design
17 a procedure, a surgical procedure, aiming at treating
18 the transposition.
19 The transposition, as Dr Silove said yesterday, is
20 a condition where the aorta arises from the right
21 ventricle and the pulmonary artery from the left, so the
22 procedures which were described in the early 1960s were
23 procedures which consisted in treating the transposition
24 by doing another transposition. They were transposing
25 the way the blood returns to the heart so as to correct
0005
1 the way the blood goes out of the heart. Do you
2 understand that? Those procedures were called the
3 Mustard and the Senning procedures.
4 So when I joined Mr Stark at Great Ormond Street
5 in 1972 and 1974, we were doing the Mustard procedure
6 and then the Senning procedure. The results had
7 improved quite significantly, to a point that the
8 mortality in the late 1970s/early 1980s was probably --
9 would you say about 5 per cent, Jaro?
10 MR STARK: 5 or under.
11 PROFESSOR de LEVAL: The arterial switch operation was
12 described. Can we get rid of this diagram if you want
13 me to do an operation on the screen?
14 MR LANGSTAFF: Can we provide a blank screen, please?
15 PROFESSOR de LEVAL: The switch operation had been described
16 a number of years before, but the first successful
17 switch was done, I think, by Jatene in the early 1970s.
18 This is the aorta with the coronary arteries, and
19 I would put the pulmonary artery next to it although it
20 is usually behind it.
21 The switch consists of dividing or transecting the
22 two great arteries, to switch them around. The key of
23 the operation is that the coronary arteries, which are
24 the feeding arteries to the heart, take off very close
25 to the aortic valve here. You cannot transect the aorta
0006
1 below the coronary arteries, so those two arteries have
2 to be taken separately and transferred to the neoaorta,
3 which is the preliminary artery.
4 That is the key of the operation in terms of
5 technical challenge. Very soon after the first few
6 switches, the surgeons and the cardiologists realised
7 that there was a variety of distributions of patterns of
8 those coronary arteries, and that some of them were
9 easier to deal with than others in terms of being able
10 to transfer them, technically.
11 MR LANGSTAFF: Can we for a moment have a look at INQ 6/3?
12 This is an illustration from a paper of yours which
13 I have shown because I think it may be helpful.
14 PROFESSOR de LEVAL: This shows the vessels in proper
15 relationship with the aorta being in front, the
16 pulmonary artery being behind. The two coronary
17 arteries have been removed with a cut of the arterial
18 wall from the aorta, to be transferred to the pulmonary
19 artery. This is the main technical point of the
20 operation.
21 The other aspect of it is that the operation is
22 done on new-born infants. You must do the switch
23 operation ideally below the age of 1 month because the
24 left ventricle, which is quite thick when you are born,
25 becomes thinner when the resistances in the pulmonary
0007
1 circulation fall, which is a normal phenomenon after
2 birth. The left ventricle, which is asked to perform
3 against the systemic resistance, the pump in the body,
4 needs to be thick enough to perform properly. This is
5 why the operation has to be done in small children.
6 So the challenge was the technical challenge of
7 transferring coronary arteries, which are vessels
8 measuring about 1 mm in diameter or less, and do a long
9 operation in new-born infants. So there is the problem
10 of cardiopulmonary bypass, long procedures, technically
11 challenging, and the post-operative care of those
12 patients.
13 THE CHAIRMAN: How long?
14 PROFESSOR de LEVAL: It depends on the technical complexity,
15 but the overall procedure can take up to 8 hours.
16 Taking into consideration the induction of anaesthesia
17 and insertion of catheters, I think it takes always at
18 least four hours, at a minimum.
19 The questions we had in the late 1970s and early
20 1980s was to decide to "switch from the switch", in
21 a way: it was to move from the Mustard which in those
22 days was giving a low mortality, low post-operative
23 morbidity and in those days at that time we had good
24 medium term results with the Mustard. So we had to
25 decide to go for an operation which was more logical,
0008
1 but this was purely intellectually, with totally unknown
2 long-term results, expected to be good, and obvious
3 higher mortality, and there was a problem of having to
4 go through a learning curve while you could treat those
5 patients with a low mortality.
6 MR LANGSTAFF: Do we yet know what the long-term results of
7 the arterial switch are?
8 PROFESSOR de LEVAL: It depends what we call "long-term
9 results". The longest we have is about 15 to 19 years.
10 Now there is a number of reports, many from France,
11 where they have looked at the coronary perfusions of
12 those patients, and there is a high incidence, or
13 a significant incidence -- I will not say high -- of
14 coronary problems in those patients who have been
15 studied. The incidence is difficult to assess because
16 a number of those patients were investigated because
17 they were symptomatic, so it is not a blind prospective
18 study, but there are some problems. And, you know, we
19 could question, wonder what will be the incidence of
20 coronary artery disease for example in those patients in
21 40 years from now. All this is still unknown, but we
22 know that the delayed results of the Mustard and the
23 Senning are not as good as expected, and there are
24 a number of patients who die suddenly, probably from
25 arrhythmias or from narrowing of the pathways being
0009
1 inside the atria. So I think in balance now, we are
2 quite confident that the switch operation is a better
3 operation.
4 In the early 1980s we were balancing the early and
5 the late risks, and one of the questions was, what kind
6 of lower risk can you afford, assuming that the
7 long-term results will be better? I do not think that
8 question has been answered. The acute problem was the
9 decision to deal with the learning curve. When we
10 started the switch operation, those who had started
11 before us started with a mortality of -- I mean, the
12 person I am quoting is Jan Quaegebeur, who has become
13 a master of the switch, who started with a mortality of
14 25 per cent.
15 The way he started is, he was more or less asked
16 by his mentor, Dr Brawn, to start the switch because he
17 was confident that it was the best operation. So it is
18 the association of a very wise, I think, and
19 forward-seeing mentor and a very good surgeon, who
20 started the switch, but they had to go through
21 a learning curve with a mortality of 25 per cent, which,
22 within a few years, came down to below 10 and now, in
23 his results, is probably about 2 or 3 per cent.
24 MR STARK: I would like to add if I may to the difficulties
25 of the decision-making in those days. Marc already has
0010
1 explained a few things. In children, unlike in adults,
2 you really are looking for the long-term results, which
3 means 50/60 years and this is why, when we introduce
4 a new operation on the presumption that the new
5 operation will be better than the old one, it is very
6 difficult. I think this, coupled with some of the
7 complications we were aware of -- and it has been
8 mentioned that the coronary arteries make late
9 problems -- it was also the right ventricle outflow
10 obstruction and the pulmonary obstruction, the flow from
11 the right ventricle to the lungs, which, with certain
12 techniques gave a very high incidence.
13 Then, at the time when, say, our results at Great
14 Ormond Street showed a mortality of the Senning
15 operation under 5 per cent, then some of the people who
16 were pioneers in arterial switch operation, like Jatene
17 in Brazil, his first survival was, I believe, after
18 three or four unsuccessful attempts. There was an
19 excellent French surgeon who made a lot of
20 contributions, and the first time he lectured to us at
21 Great Ormond Street, he had 12 or 14 consecutive cases
22 who had died.
23 So that was psychologically a very difficult time
24 for the cardiologists and for the surgeons to make this
25 change from the old-type operation that we thought the
0011
1 long-term results probably will not be so good and the
2 arterial switch operation where we thought the long-term
3 results probably will be much better. I think now,
4 20 years later, we are a little bit wiser.
5 MR LANGSTAFF: Just looking at the early operations, Jatene,
6 when he began, the French surgeon whom you quote, would
7 have been in a sense experimenting on the children who
8 were in their care?
9 MR STARK: Well, I think one can put it that way, although,
10 on the other hand, all the new operations you may in
11 summary call "experiments" because you cannot experiment
12 on animals because you do not have the animal model, and
13 even if you had the animal model, we are not, as you
14 know, allowed to try the operations on animals --
15 certainly in some countries this is possible but we
16 cannot do that. So to some extent "experimenting"
17 sounds a harsh word, but I think it was.
18 MR LANGSTAFF: In such a case, do you know to what extent
19 the parent of the child concerned was involved in taking
20 a decision which might involve the parent being told,
21 "The last two, the last three operations that I have
22 done", says the surgeon, "have all failed. On the other
23 hand, what I can offer you is a chance that your child
24 will live to a much longer and riper old age"?
25 Was the parent involved? Did the surgeon in those
0012
1 days take the decision for the parent? How did it
2 work?
3 PROFESSOR de LEVAL: When we were in the transition period
4 of moving from the Mustard or Sennings to the switch, we
5 were informing the parents of that change of attitude
6 and policy, and we were discussing the pros and cons of
7 those operations on a purely theoretical basis because
8 we had no hard data, but we had good reasons to believe
9 intellectually that a good switch was better than a good
10 Senning. We indicated to them that the risk, the early
11 risk, was higher, probably, for the switch than for the
12 Senning but we expected the long-term results to be
13 better and in balance, the switch to be better.
14 THE CHAIRMAN: Professor Jarman has a question, but you
15 finish this answer.
16 PROFESSOR de LEVAL: I have forgotten where I was.
17 MR LANGSTAFF: You said you expected the long-term results
18 to be better.
19 PROFESSOR de LEVAL: That is right. If I recall, in those
20 days we ended up the conversation with the parents
21 sometimes by doing a switch and sometimes by doing
22 a Senning. That was at the beginning, but as time went
23 by, we convinced ourselves that the mortality of the
24 switch was going down and therefore that, in balance, we
25 were making -- not a plea, but we were influencing the
0013
1 decision to have the switch.
2 MR LANGSTAFF: In a sense, that would have to be the case,
3 because if any new surgical technique is developed which
4 you as a surgeon would prefer to do, then there is
5 a disadvantage, is there, in doing a low volume of two
6 operations rather than a higher volume of one?
7 PROFESSOR de LEVAL: This is the discussion about volumes,
8 which is a very difficult one. It is volume of what?
9 Is it volume of operations you do which is important to
10 maintain your skill? Is it the number of bypass
11 operations? Is it the exact number of switches?
12 I think that the concept of volume is a general concept,
13 and it is important to indicate what we really analyse.
14 I think that in all fields there is a correlation
15 between volume and good performance, but what is the
16 most important? Is it that we do a switch or that we do
17 quite a number of operations on small babies? Or can
18 a surgeon do any operation to maintain his skills?
19 I think those points are important, because there
20 is a tendency to look at one diagnosis and if you look
21 nowadays at the performance of the very good surgeons,
22 you can just see that some years a surgeon would have
23 a higher mortality for one condition. So I think it is
24 difficult to correlate volume with one particular
25 condition.
0014
1 THE CHAIRMAN: May I interrupt just a second? I am sure we
2 are going to pursue the issue of volume, but before we
3 lose sight of the beginning of switch, Professor Jarman
4 had a question.
5 PROFESSOR JARMAN: I am sorry to interrupt. It is difficult
6 with this setup and it is a difficult question as well.
7 I realise how difficult it is to develop new operations,
8 particularly if you cannot operate on animals, but
9 I just wanted to ask you, had you had a young child of
10 that age at the time when the operative mortality of the
11 early operations, the Sennings, was 5 per cent, and
12 knowing that surgical mortalities were likely to reduce,
13 what would you have chosen: to go to the new operation
14 or to have stuck with the older one, for your own child,
15 personally?
16 PROFESSOR de LEVAL: The way I went about this myself is
17 that I could not cope with the learning curve,
18 basically, so what I have done -- maybe I am a bit
19 selfish, but I asked someone who had gone through it to
20 do the operations with me. This is what I did, so
21 Dr Quaegebeur, who happens to be a surgeon as well, came
22 to Great Ormond Street to help me starting the switch
23 operation. This is the way I did go through my learning
24 curve.
25 Posteriorly, I think it was probably a good thing
0015
1 to do, and if I would comment, to try to reduce the side
2 effects of learning curves is precisely to limit the
3 number of those who pioneer new procedures, and when
4 they master it, to have a system whereby the other
5 people can benefit and reduce their learning curve.
6 There are two types of learning curve. There is
7 the learning curve of established people to learn new
8 procedures, and there is the learning curve for training
9 for which conventional procedure has to be learned and
10 taught. I think that the way to go about it is to do it
11 with someone who has done it. We cannot get rid of
12 a learning curve completely, but we can certainly
13 minimise its effects. I felt unable to go through
14 a 25 per cent learning curve myself, so I have great
15 admirations for Jan Quaegebeur who did it, but
16 I personally would not have had the guts do it.
17 PROFESSOR JARMAN: I am not quite sure that quite answers my
18 question. Mr Stark?
19 MR STARK: It is obviously a very difficult question, but
20 I think it is a very pertinent one. I think precisely
21 because of this question, we probably started with the
22 arterial switch operation at Great Ormond Street
23 a little bit later than other people, and I think it is
24 very difficult -- it is a theoretical question but you
25 did not have much time for the decision. If you wanted
0016
1 to go for an arterial switch operation, as Professor
2 de Leval explained, it has to be done in the first four
3 weeks of life, so you cannot dither and wait. I think
4 probably in the early 1980s I would have opted for
5 Senning for my child. But then you would say I would be
6 preventing all the progress, and you would be right.
7 PROFESSOR JARMAN: Maybe if you were allowed to operate on
8 animals --
9 MR STARK: I think it would have helped, but it would still
10 not solve the problem completely, because I personally
11 believe, when we talk about learning curves, we
12 concentrate, perhaps rightly, very much on the surgeon.
13 But this particular operation, or any operation on an
14 infant, is such a team effort that a minor slip in
15 diagnosis, in perfusion, anaesthesia, post-operative
16 care, can completely negate the result. So although the
17 surgeon is a prerequisite and you have to have a good
18 operation, I think it is not everything. I am not sure
19 I have answered your question.
20 PROFESSOR JARMAN: You have, thank you.
21 PROFESSOR de LEVAL: Another way, not to answer that
22 question completely, but very early we identified that
23 some of those switches were higher risk than others
24 because they were coronary arteries, so we tried to know
25 pre-operatively and select the better cases. For the
0017
1 first few years we were doing the switches only on the
2 good cases and we were doing Sennings on the difficult
3 coronary arteries.
4 But I still cannot answer your question of what
5 I would have done as a parent, because it depends on the
6 information we receive, and I do not think that in the
7 1980s we were giving the parents the information you
8 expect today.
9 THE CHAIRMAN: Mrs Howard has a question.
10 MRS HOWARD: This may be evident to others, but could you
11 give me some feel of what you mean by "a good case" in
12 these early days?
13 PROFESSOR de LEVAL: I mean a low risk coronary artery, so
14 we identify some coronary arterial patterns which were
15 difficult to transfer and carried a higher risk, so for
16 those patients, we electively put them through a Senning
17 operation. So when we began on our own, we were doing
18 a switch on those cases who were expected to give good
19 results.
20 MR STARK: I would add, if I may, one other thing. We are
21 talking about the late 1970s or early 1980s, but I think
22 the arterial switch operation is both technically
23 demanding and demanding on the whole team, so it is
24 possible that this operation should only be performed in
25 a limited number of centres, because in some of the
0018
1 papers which I have submitted to you, even 1994/96 in
2 the State of New York, overall mortality was 10 per
3 cent. Not very long before the Congenital Cardiac
4 Surgeons Society in the United States, which encompasses
5 more or less the best units in the country, the
6 mortality for the so-called low risk operations was 6 to
7 13 per cent, but for the rest of the department in the study
23 to 80 per cent.
8 That, I think, would indicate that although the
9 good centres or the centres with low mortality managed
10 to get over the learning curves, it is still not an easy
11 operation.
12 MR LANGSTAFF: Can I explore for a moment the change that
13 there has been in parental expectations as to
14 information and the influence that that might have
15 had -- it is a theoretical question -- if the arterial
16 switch programme had not developed with Jatene in the
17 1970s but was proposed today?
18 Is there a conflict, do you think, between the
19 development of a new operation, the results of which are
20 speculative, when there is an operation the results of
21 which are relatively clear, although far from perfect,
22 such that if you had said to a parent, if surgeons
23 generally had said to parents, if Jatene had and the
24 French surgeon that you quote had said, "The risks of
25 operative death are very high here; we have not done any
0019
1 of these operations, or many of these operations
2 before. Do you want this for your child because I have
3 an idea that it will be better in the long-term?", the
4 parent might very well say, "No, I do not want that",
5 and you would not actually get the development of an
6 operation which is now thought to be of value generally.
7 PROFESSOR de LEVAL: This is the big issue of scientific
8 progress and comparing strategies which can come
9 subsequently or simultaneously. The only proper
10 scientific way would be to proceed with a prospective
11 double-blind study. That is the only scientific way to
12 compare two techniques. This should apply to the early
13 and long-term results.
14 We are talking here about new-born infants whose
15 life expectancy now, in our civilisation, in this part
16 of the world, is about 70 years, so normally we should
17 know what the results at 70 years are of the two
18 operations before we can decide which one is the best,
19 which is totally impossible.
20 The speed with which progress goes on in terms of
21 science is such that all the time new procedures, new
22 strategies, new operations are coming out, and this
23 question comes up all the time. I do not have an answer
24 for it, but we should realise that this is what happens
25 in the world, so there will be always patients who will
0020
1 receive for the first time a new procedure, whatever you
2 have done before to make sure it was the proper way to
3 do it, there will always be patients who will be at the
4 beginning of an experience, and it is unavoidable.
5 What we have to do is to introduce those new
6 strategies, new research, to make sure that we minimise
7 the casualties of this introduction, but I believe that
8 they are unavoidable.
9 MR LANGSTAFF: Mr Stark?
10 MR STARK: I would completely agree with what Marc said.
11 I think that one thing, when you say "when you discuss
12 it with the parents", you actually are not asking the
13 parents to make the decision, because I think, to some
14 extent, the way you see the benefits, you are willingly
15 or unwillingly influencing the parents. But the other
16 way round, the parents sometimes influence you.
17 I would like to give you an example. One of the
18 very difficult conditions is pulmonary atresia, with
19 major collaterals coming from the aorta.
20 For this condition, although the outlook has
21 improved, the usual scenario in the 1970s/80s was that
22 we would do two, three, four palliations in the first
23 3 years. Eventually there was nothing to offer. So on
24 those occasions when we told the parents this scenario
25 and suggested, because the outlook is so bad, that
0021
1 perhaps we should not treat the child, of course very
2 often the parents feel anything that could be done
3 should be tried, and we did, and then the scenario was
4 followed.
5 Then, a few years later, the parents would write
6 to me and say, "We are sorry we did not take your advice
7 because the misery we have suffered during those three
8 or four years was immense".
9 So I think that there is always both sides that
10 influence each other. I am not sure if I expressed
11 myself clearly.
12 MR LANGSTAFF: The problem that I was asking about is that
13 of scientific progress which inevitably may have a cost,
14 and the cost may be human.
15 MR STARK: Yes.
16 MR LANGSTAFF: It is one thing to see it in the general,
17 when the cost of progress for all may be reasonable,
18 possibly, arguably, even though there may initially be
19 harm to some. If you are looking at it in terms of the
20 individual patient who represents one of the "some", how
21 does one reconcile one's duty as a doctor to the
22 individual with the idea that there may be, but not
23 necessarily will be, progress for the many?
24 PROFESSOR de LEVAL: I think that there are ways to minimise
25 the risks of introducing new strategies, and nowadays,
0022
1 for example, in terms of education and training, there
2 are models, there are simulations; some of the
3 operations can be performed on animals, although the
4 switch, I think, would have been quite impossible to
5 do. I am involved personally in some research on
6 computation of flow dynamics, which is a highly
7 sophisticated technique of looking at circulation to try
8 to analyse flow dynamics before we do it on a patient.
9 There are ways to reduce as much as we can the
10 casualties, but I think it will be impossible to reach
11 a stage whereby the risk of an operation is the same
12 when you do it for the first time than when you do it
13 for the 100th time, I think.
14 THE CHAIRMAN: May I ask you a question following on? As
15 you rightly say, this is a very hard issue but it is one
16 that has to be confronted. You clearly, in your
17 professional lives, will have confronted it many times.
18 It was put to us quite early on in our evidence
19 that one way of responding to the challenge of
20 innovation, as you have described it, would be to insist
21 not only in the kind of measures you have described, the
22 use of models, whether computerised or whatever, but
23 also to insist that any proposal for innovation in the
24 form of surgery be passed before the relevant Ethics
25 Committee, which will be concerned with looking at
0023
1 whether the surgeon has had all this other experience,
2 whether all the opportunities have been exhausted to
3 acquire technical expertise and whatever else is needed,
4 as well as whether the balance has been properly weighed
5 to introduce a programme.
6 What observations would you have on that?
7 PROFESSOR de LEVAL: I think that to have a formal proposal
8 going to an appropriate Ethical Committee is certainly
9 a good idea, and actually, that exists for a number of
10 new innovations in medical treatments, new drugs,
11 et cetera.
12 THE CHAIRMAN: Certainly, but not traditionally in surgery?
13 PROFESSOR de LEVAL: No. There was an article in the JAMA
14 some years ago. The title was "There is no FDA approval
15 for the surgeons", which is true, and I think that that
16 could be of some benefit, although we must make sure
17 that it does not do more harm than good.
18 For example, the legislation nowadays in the
19 States is such that the they lag behind in a number of
20 fields, like orthopaedic surgery, fields where you have
21 to use prosthetic or artificial materials. It is very
22 difficult to have an improved prosthetic valve, for
23 example, accepted in the States and the clinical trials
24 have to be done outside the country because of those
25 regulations. So I think that the suggestion of having
0024
1 an approval for a new procedure is a good one, but we
2 must make sure it remains well balanced.
3 The other thing is that there are two types of
4 progress in science. One is the big new innovation,
5 very often totally unexpected. Then there is the daily
6 improvement in what you do. Personally, I have been
7 involved in the so-called Fontan operation, which is an
8 operation we do for a patient with a single ventricle.
9 At the end, I contributed to the design of a new
10 operation, but it did not start from scratch; it is the
11 evolution. I did it, and if I had had to put a proposal
12 for doing the operations I do today, I do not know when
13 I would have put it, because I was not aware of the end
14 point.
15 I think this distinction between the two types of
16 progress is important.
17 MR LANGSTAFF: We heard also at an earlier stage of our
18 evidence the view expressed that, so far as the arterial
19 switch was concerned, a number of surgeons were, for
20 reasons of demonstrating their prowess, keen to take up
21 what seemed to them to be a new operation: because
22 Jatene had done it, they wanted to do it, was the idea
23 expressed to us.
24 To what extent is that a reflection of a number of
25 surgeons' attitudes in the late 1970s, early 1980s?
0025
1 PROFESSOR de LEVAL: I have never seen a surgeon who did an
2 operation because of just wanting to do it, it was
3 good. I have never come across a surgeon with that
4 mentality, I must say. Have you, Jaro?
5 MR STARK: No, it is a new concept to me.
6 PROFESSOR de LEVAL: Have you, Eric?
7 DR SILOVE: I do not believe surgeons or cardiologists think
8 like that at all. I am surprised at the suggestion,
9 actually.
10 MR LANGSTAFF: It was one that was made to us.
11 If I can then ask, before we leave this difficult
12 area, to what extent would a parent, in the 1980s, in
13 the late 1980s, in the 1990s and now, looking at those
14 four different periods, understand that they were one of
15 the first and so, in a sense, their individual case
16 might be the sacrifice for the future of others to come?
17 To what extent might they have appreciated that
18 that would be the possibility?
19 PROFESSOR de LEVAL: I think that the parents were informed
20 that the procedure was new or that there was an
21 alternative, but we were implying or proposing a new
22 procedure, but I think all this was done in the context
23 of a relationship of confidence between the families,
24 the cardiologist first and the nursing staff, the
25 surgeon, and I do not think that parents ever considered
0026
1 that they were probably the "victims" or "guinea pigs"
2 or whatever you would call that. I think that they
3 were, you know, as fully informed as we thought they
4 should be, and we were totally open in what we were
5 doing. The fact that we were telling them that it was
6 a new procedure implied, without being necessarily
7 specific in spelling it, that there was probably
8 a higher risk in those days than now, because we have
9 done more.
10 But I think the relationship and the ambience
11 where all those things were taking place was such that
12 there was full confidence between parents and the
13 institution.
14 MR STARK: I think this is a very important point, because
15 I could give you an example of one operation which
16 I thought could be done and it has not been done before,
17 and when I talked to that family, I put it to them in
18 those simple terms: "It has not been done before.
19 I think it could work". The parents, and many other
20 parents, in those days usually did not comprehend fully
21 the anatomy of physiology, because, as I think was
22 probably explained already yesterday, it is quite
23 complex. Very often the answer was, "We know that you
24 will do your best and we trust you". So we went ahead
25 with the operation, that particular operation went well,
0027
1 but the feeling that the nurses, cardiologists, surgeons
2 had the full trust of the parents probably made the
3 explanation, even under such difficult conditions when
4 we started new operations or where we knew the risk was
5 still high, somewhat easier. I think it is much more
6 difficult today.
7 MR LANGSTAFF: The cardiologist would probably see the
8 parents before the surgeon.
9 DR SILOVE: Yes, the cardiologist would see the parents
10 first, but the cardiologists and the surgeons will have
11 discussed all of the ramifications in quite some detail
12 before the cardiologist ever puts something to the
13 parents.
14 I am sure it is the cardiologist who would be the
15 first to suggest to the parents, "We have discussed this
16 problem and we feel that the right way to go forward is
17 to go for the arterial switch operation". It then
18 becomes a question of how it is dealt with in an
19 individual institution, whether the parents see the
20 surgeon or whether the cardiologist and the surgeon see
21 the parents together, but the communication is terribly
22 important.
23 MR LANGSTAFF: One could take the trust point that Mr Stark
24 has made further, then, could one, and say that it is
25 a mistake to look at it simply in terms of
0028
1 a consultation between a parent or patient and
2 a surgeon, because in effect it is a team consulting
3 with the patient or parent?
4 DR SILOVE: That is exactly right. It is a team
5 consultation, it is a team decision and it is a team
6 responsibility. I do not think the surgeon should take
7 the flack for everything that goes wrong. I think the
8 whole team is responsible for things going right or
9 wrong.
10 MR LANGSTAFF: I see Mr Stark is nodding to that.
11 PROFESSOR de LEVAL: I fully agree with that. As Jaroslav
12 said, every single patient operated on has been
13 discussed once, twice or three times in great detail by
14 at least two consultants, a surgeon and a cardiologist,
15 but most of them by the junior staff, other consultants.
16 If it is a difficult problem, we would repeat
17 investigations. In the beginning we used to go back
18 with the echocardiograms, repeat an angiogram, to find
19 out where the coronary arteries were. Because the
20 coronary artery was so crucial, sometimes we went ahead
21 with a switch and found different coronary arteries than
22 expected and we had to back up to a Senning.
23 So all these decisions were team decisions in
24 which we all took responsibilities. I think this is
25 that type of attitude, corporate attitude, that was
0029
1 communicated to the parents, which I think allowed us to
2 work in an atmosphere of trust and confidence. I think
3 it is absolutely vital. I do not think the parents have
4 ever seen me, as a surgeon, as a single individual
5 within the hospital. I have been always part of a team,
6 and they knew when I talked to them that it was after
7 discussing with others, it was the decision which had
8 been taking place at the conference.
9 THE CHAIRMAN: May I ask one particular question in relation
10 to that? You said at one stage that to say that
11 something was a new procedure would be understood by
12 a parent to imply that there was a higher risk.
13 How do you know that?
14 PROFESSOR de LEVAL: I do not. Similarly, I do not
15 understand today whether the patients I speak to
16 understand, or the parents, what I say. I think that it
17 happens to me when I speak to people in other fields,
18 like the legal profession: they think I understand and
19 I do not, although they try to speak in simple terms.
20 I think this applies to the parents. Even if you make
21 a drawing of a switch operation, I cannot be sure that
22 what I have drawn, you will understand exactly what it
23 implies.
24 THE CHAIRMAN: Of course, but if there was, in your mind,
25 the notion that there were increased risks, why would
0030
1 that not be said, rather than merely say, a new
2 procedure would be understood by a parent to imply
3 greater risk?
4 PROFESSOR de LEVAL: No, I was specific about this as well.
5 Each time I see a family, for the last 25 years, I write
6 a note on the chart or the record of the patients with
7 the risk I quoted, and if it is 10 per cent or 15 per
8 cent, and in those areas, I mention the difference of
9 risks as well, and it is written down in the notes.
10 THE CHAIRMAN: Do bear in mind many of the questions I ask
11 you are so as to make sure we are clear on everything
12 you are telling us.
13 MR STARK: There is one thing. I think it is very important
14 to give as accurate information as possible, but I think
15 the more we learn about the statistics and statistical
16 evaluation, I personally realise that I have probably
17 not given very accurate information a number of times.
18 There has been recently a very nice paper in the BMJ on
19 the incidence of adverse effects that have not happened
20 yet.
21 So can the surgeon who operates 15 switches
22 without mortality tell the parents that the risk of that
23 particular operation is zero? No. Actually,
24 statistically, there is, within 95 per cent confidence
25 intervals, that he may lose 20 patients out of the next
0031
1 100 and it could be due to chance.
2 So that is a sort of new concept for me, because
3 if I have not lost 15 patients, I would probably quote
4 the risk of 5 to 10 per cent, but I would probably never
5 quote the risk of 20 per cent.
6 So I think that means that we will have to learn
7 more about these things and perhaps educate the parents
8 more, because otherwise if we say the risk is very low,
9 of course if the child dies, for the parents it is
10 100 per cent risk and it is a tragedy.
11 MR LANGSTAFF: What you have both indicated is that when you
12 were dealing with risks with patients, you would quote
13 a percentage. You have just been indicating, I think,
14 that the percentage is not worked out on a mathematical
15 basis because if you are reflecting your own
16 performance, taking your own series, you would, in the
17 instance you have quoted, have said 0 per cent. You
18 modify the mathematics in relation to other factors, so
19 it becomes a matter of judgment, does it, rather than
20 a question of mathematics?
21 MR STARK: It does.
22 MR LANGSTAFF: To what extent, then, does quoting
23 a percentage give a spurious impression of scientific
24 certainty to a patient?
25 PROFESSOR de LEVAL: I agree with that. I think what I have
0032
1 done in 25 years lacks of science completely, and
2 probably misled the patients. I agree with that.
3 MR STARK: I think very often we would quote the parents
4 actually very broad idea, like saying the risk is less
5 than 50:50. Only when the parents insisted, we put
6 together our own experience, we put together the data
7 from the literature, but it was not scientific;
8 I completely agree. Unfortunately, we did not have the
9 basis for that.
10 THE CHAIRMAN: Professor Jarman?
11 PROFESSOR JARMAN: I just wondered: would it be possible to
12 just give the confidence interval, the percentage, or
13 would they not understand?
14 MR STARK: It would, but I would like to know what
15 percentage of parents would understand "confidence
16 limits", because I did not understand until a few years
17 ago myself.
18 MR LANGSTAFF: Picking up on Professor Jarman's question, it
19 might be suggested that one way of putting it would be
20 to say, "I cannot say precisely what the risk is, and of
21 course no operation is ever successful to a percentage;
22 it is either successful or it is not, but it is within
23 the range, 15/25 per cent or 2/15 per cent, and using
24 the word "range" rather than "confidence interval".
25 Is that something which is ever done, or not?
0033
1 PROFESSOR de LEVAL: Nowadays, obviously, we are careful
2 what we say, what we write and we try to choose our
3 words, but I think that, frankly, when I was talking to
4 parents in 1985 about risks, I did not know exactly what
5 my results were and certainly not what the confidence
6 interval was, so it was a clinical impression of what
7 I had done; also a knowledge of what had been published
8 and what I had heard at meetings. Some of the
9 conditions, the number of cases I had done, we had done,
10 was very small, rare conditions, and the risk quoted was
11 the best I thought I could do in terms of assessing what
12 the risk was, plus taking into consideration my own
13 performance from previous cases, which is something you
14 have alluded to. You tell the parents that three of the
15 last four patients have died while all the others before
16 had survived? We did not, but I am sure that when I was
17 quoting a risk of an operation, having lost one or two
18 patients from the same condition, I was more
19 pessimistic. But this is no science. There was no
20 confidence interval quoted.
21 MR STARK: I think there may be perhaps another way in
22 future, if we master adequately data collection, we can
23 have on our desktop access to our previous data, and
24 instead of quoting overall numbers, we can say "For this
25 condition for the last three years I have done [so many]
0034
1 patients and those were the results". You can then
2 perhaps discuss around this figure, the percentages,
3 with the parents, but that means that we have to have
4 the systems and resources for that, et cetera.
5 MR LANGSTAFF: Because otherwise your practice of writing
6 the percentage that you quoted into the clinical notes
7 is there for the purposes of the record and presumably
8 it is there because, if anyone were to ask a question
9 about it, you could say, "That is the percentage
10 I quoted". But if you were challenged on it, if someone
11 said "How on earth do you quote 25 per cent when the
12 last 10 cases you have done have all been successful, or
13 all been unsuccessful?", you would have to answer. How
14 easy would that be in any given case?
15 PROFESSOR de LEVAL: I cannot answer it. I cannot
16 demonstrate why scientifically; it is just an
17 impression. I started to do this. I learned that from
18 my mentor in my clinic, Dwight McGoon who was doing it
19 in those days, 1973, and I thought it was a good way to
20 try to introduce the concept of risks and try to give an
21 idea to the parents, because this is what it is about:
22 that the parents should be informed that there is
23 a risk, the risk of death, the risk of morbidity, and to
24 give them an idea of the magnitude of the risk. It is
25 no science and we were trying to be as accurate as
0035
1 possible in terms of our own feeling at the time. But
2 I cannot demonstrate scientifically that the risks
3 quoted were the correct ones.
4 MR LANGSTAFF: There must come a point at which anyone
5 looking at a figure which has been quoted would be able
6 to say, "Well, that simply is not a proper professional
7 figure; we have careful surgeons whose figures may well
8 be right; we may have cavalier surgeons whose figures
9 are simply wrong. How does one judge the rightness and
10 the wrongness of the figures? What are the factors that
11 go, broadly speaking, to make up this sort of percentage
12 estimate?"
13 MR STARK: This is why I mentioned that for a long time
14 I tried to avoid percentages, because I thought for
15 a number of parents they were actually meaningless.
16 I tried to explain to the parents what were the
17 alternatives; if we did not operate, what sort of life
18 the child would lead or whether the child would die, and
19 also mention that even if the risk was 1 per cent, which
20 is very low risk, if that one child was their child,
21 therefore it was 100 per cent. I preferred to leave it
22 at that because I was aware that any percentage I give
23 them could be challenged. We did not have a better way
24 to do it.
25 PROFESSOR de LEVAL: I think we should also look at this in
0036
1 the context of what the mortality was 25 years ago for
2 open-heart surgery in children. 25 years ago, I think
3 the overall mortality for infant cardiac surgery was
4 poor, 25 per cent? Now we are --
5 MR STARK: I am sorry, I have these figures because I looked
6 it up in one of the early articles. In 1965 to 1970,
7 mortality for infants under 1 year done on bypass was
8 70 per cent.
9 PROFESSOR de LEVAL: Now we are at a mortality of probably
10 5 per cent overall, I suppose, so the concept of death
11 following surgery, in those days, was different from
12 what it is now. That has to be taken into consideration
13 in this discussion, because we are discussing events,
14 procedures and practices of 20 years ago, with our
15 today's mentality, which is a bit difficult, I think.
16 MR LANGSTAFF: Can I ask you to think perhaps over the
17 break -- because we have come, sir, I think to a time
18 when we perhaps ought to have a coffee break, but if
19 I can return after the break to the question of what, if
20 any, factors that would go to a percentage can be
21 identified, how one would set about, as it were, telling
22 the good and proper percentage figure from the
23 inadequate, wrong, cavalier figure?
24 That is what I would welcome your views on, as to
25 whether there is any touchstone or any factors which one
0037
1 can identify which would go to distinguish the one from
2 the other.
3 THE CHAIRMAN: That will start the discussion, but there
4 will be many other matters which will subsequently be
5 raised. We will now take a break for 15 minutes and
6 reconvene, therefore, at 11 o'clock.
7 (10.45 am)
8 (A short break)
9 (11.05 am)
10 MR LANGSTAFF: Returning to the question I left you puzzling
11 over at the start of the coffee break, what factors
12 would one consider in distinguishing a good percentage
13 judgment from a bad percentage judgment?
14 MR STARK: It is a professorial question.
15 PROFESSOR de LEVAL: I suspect that what you would like to
16 ask is how do I identify a surgeon who is
17 over-optimistic from a surgeon who is over-pessimistic?
18 I suspect you want to tell us that some surgeons are
19 over-confident and when they speak to parents they quote
20 risks which are lower than they actually are. I suspect
21 that is what you want to discuss now?
22 MR LANGSTAFF: Yes.
23 PROFESSOR de LEVAL: The only way to be sure that the risks
24 quoted are correct is the scientific way. I do not
25 think we can do it. I think that it is impossible, with
0038
1 small numbers, et cetera, to give you a risk for
2 conditions which you see once, twice, three or four
3 times a year. I think it is very difficult.
4 MR LANGSTAFF: What you were telling me, I think, before the
5 break was that if you ask the question, how do you tell
6 a surgeon who quotes risks which are lower or higher
7 than they actually are, is that that implies a starting
8 point that you are able, actually, to say what the risk
9 is. You were telling me you cannot, because --
10 PROFESSOR de LEVAL: That is right.
11 MR LANGSTAFF: -- the individual surgeon has individual risk
12 and may not do a sufficient number of operations to
13 identify his particular risk. Even if he does, his own
14 performance in that operation may vary from time to
15 time -- and this I think we will touch on in your CUSUM
16 analysis in a moment -- so that it is going to be
17 variable. Thirdly, there may be particular factors in
18 relation to recent history which may or may not indicate
19 a proper assessment of risk. Fourthly, one has to put
20 that in the context of a generally changing risk
21 elsewhere, because there are undoubtedly going to be
22 improvements in surgery, surgical techniques, the
23 abilities of the team to provide the result, and you are
24 also looking at not only your own surgical results, but
25 the results of the team around you who may vary in the
0039
1 same way.
2 Those are all the features that one might quickly
3 identify as being matters which would make it difficult
4 to quote a particular risk, as Mr Stark has indicated,
5 but it still leaves the question: can one, even despite
6 that, say of a particular estimate, "That is obviously
7 wrong" because it is too high or too low?
8 MR STARK: I have been discussing recently this aspect with
9 the statisticians, with Professor Gallivan from UCL (who
10 I think is actually on your panel) with regards to the
11 paper I mentioned when we analysed the results from the
12 five UK centres.
13 We find this question very difficult because in
14 that study we had 11 surgeons and the mortality, say,
15 for open-heart under 1 year varied widely between,
16 I think, 2 and 10 per cent, yet with the statistical
17 analysis, we found that all these results were within 95
18 performance confidence intervals; therefore they could
19 be due to chance; therefore it would be probably
20 inappropriate for the surgeon with the risk -- and again
21 we are talking about very small numbers -- based on the
22 results of say 10 per cent mortality, to say that his
23 mortality is any different from the surgeon who had
24 2 per cent mortality.
25 I know that you desperately want the answer, but
0040
1 I think that I personally find it very difficult to
2 answer it in any scientific way.
3 DR SILOVE: I think there are very few operations actually
4 which a surgeon does sufficiently frequently to be able
5 to build up adequate statistics. If you are talking
6 about, say, ventricular septal defect, he might have
7 done enough VSDs to be able to give you reasonable
8 statistics, but if you are talking about one of the less
9 frequently seen conditions, such as truncus arteriosus
10 or total anomalous pulmonary venous drainage, no surgeon
11 will have done enough to be able to quote accurate
12 statistics, and certainly not any confidence limits that
13 can be relied upon.
14 MR LANGSTAFF: Can I take an example and ask what you might
15 think it appropriate to say in a case such as this?
16 Suppose you have a surgeon who, in terms of repairing
17 VSDs, AVSDs, Fallot's, has near perfect results, but of
18 the last three operations for truncus arteriosus which
19 he has done -- he has just done three -- two have died;
20 one, let us suppose, dies not even reaching the surgeon
21 on the table because he dies in the anaesthetic room.
22 In such a case, a fourth child suffering from
23 truncus arteriosus comes to him. What risks does he
24 quote?
25 MR STARK: I think it would be in my view inappropriate to
0041
1 quote any figures because he would not be in a position
2 to quote figures of, say, other surgeons in the UK
3 because he would not know them; he would know perhaps
4 figures from literature and, as I mentioned during the
5 coffee break, results in peer review journals are only
6 the best results; the editor would not accept anything
7 else.
8 I think in the example you mentioned, that surgeon
9 obviously is an outstanding surgeon if, for the five
10 conditions he has not lost a patient, and here he lost
11 two out of three, to tell the parents that in his view
12 the risk would be 60 per cent would be inappropriate.
13 But I would not know. I think I would advise such
14 a surgeon not to quote any figures and just say the
15 results of the operation, this child's chances of
16 surviving without operation is nil, and it is a high
17 risk procedure, and I think I would leave it at that.
18 I do not know if you have something else on that,
19 Marc?
20 PROFESSOR de LEVAL: No. I think prospectively, the use of
21 CUSUM is another simple way to try to assess your own
22 performance, or the institution's performance, is that
23 you may have an alert, an alarm, indicating that the
24 performance is deteriorating. But if you do that, I do
25 not think that to base your analysis purely on mortality
0042
1 is sufficient. I think that, again, you are talking
2 about a period of cardiac surgery where mortalities were
3 high in all centres, and we extrapolate this to
4 a discussion where the mortality currently is much
5 lower.
6 I think that if your performance, for example, in
7 repairing ventricle septal defects, and look at the
8 mortality nowadays, it is an exercise which will be
9 meaningless, to some extent. You have to look at other
10 indicators of performance which are more subtle, such as
11 the length of intubations post-operatively, et cetera.
12 But I think that the difficulty I have here is that you
13 are asking current questions for past history. It is
14 very difficult to know if you want us to answer as being
15 in the past or as being today. I think it is different.
16 THE CHAIRMAN: Before you go on Mr Langstaff, you said,
17 Mr de Leval, that you must look at other indications and
18 you gave one example. Could you spell that out a little
19 bit more? What other relevant indications?
20 PROFESSOR de LEVAL: I think that, again, if you look at
21 analysis of results, what we normally do is to go to
22 a statistician. You have two extremes: one who is going
23 to look at 500 variables and parameters, and someone
24 else who will look at 2.
25 If you were asking me to have one indicator of
0043
1 performance of cardiac surgery outside the operating
2 room, or even including the operating room, I would tell
3 you the times of intubation post-operatively because
4 that reflects more or less the haemodynamic conditions
5 coming out from the operating theatre, the condition of
6 the body, therefore the way the body was perfused, the
7 lungs were perfused; the patient is able to breathe,
8 therefore, neurologically intact. You can remove all
9 the interlinked catheters in the lungs, the neck, the
10 arteries; you reduce the risk of infection.
11 So, as a simple surgeon, I would tell you that
12 that would be my indicator because it means a lot of
13 things. But I am sure statisticians would be totally
14 disappointed with that and disagree and say "You have
15 much much more to know", but this is what I would say.
16 DR SILOVE: Could I come in on the question you posed about
17 the surgeon with very good results in one regard but who
18 has lost some patients with truncus?
19 Another approach that that surgeon could use in
20 discussing it with the parents, I think, is he could
21 say, "Well, the actual technical approach to correcting
22 truncus arteriosus has many similarities and uses many
23 of the same strategies and techniques that we use for
24 other conditions which we see more commonly". I mean
25 the Rastelli operation, for example, is done in a number
0044
1 of situations. "The difference with your child, because
2 he has truncus arteriosus, is that he is very, very ill
3 at this moment and there are other complicating
4 factors." I just wonder whether that approach could not
5 be used. I do not know what the surgeons feel about
6 that.
7 MR STARK: I think it certainly could be used, but it still
8 does not get us away from your question.
9 MR LANGSTAFF: The question is complicated by what Professor
10 de Leval said about looking at it through present day
11 eyes. If one steps back ten years, how would it have
12 been approached then? Would people then have seen the
13 shortcomings of the percentage approach, or not?
14 PROFESSOR de LEVAL: If you go back another 10/20 years,
15 where 25 per cent of the patients were dying, it was
16 very difficult in those days first of all to understand
17 exactly what the mortality was. The distinction between
18 human performance and medical problems was very
19 difficult; it still is now. Now we reach a point where
20 the technicality of the operation, the medical aspect of
21 it, is resolved for many conditions, and this emphasises
22 the importance of performance and human factors, which
23 is what we like to discuss here. But we should realise
24 that in those days the medical component of the outcome
25 was much greater than now and we are quoting risks
0045
1 giving outcomes having not dissociated the two
2 components which lead to a good or bad outcome. Again,
3 there was lack of science, lack of knowledge, lack of
4 understanding, but this is the way it was. I think it
5 is difficult to be more scientific.
6 MR LANGSTAFF: If you take a 10-year snapshot, how was it
7 then?
8 PROFESSOR de LEVAL: This is, as usual, a progression.
9 I think that the results were better and therefore there
10 was more room to assess performance, and then we
11 probably started to see differences between individuals
12 or institutions where human factors play proportionately
13 a more important role in the outcomes.
14 MR LANGSTAFF (to Mr Stark): So I get it on the transcript,
15 you were nodding at that?
16 MR STARK: Yes, I agree with that.
17 MR LANGSTAFF: Can I come back to where we began this
18 discussion? We will come back to human factors in
19 a moment or two. We still have on the screen a simple
20 diagram showing the coronary artery transfer which is
21 necessary for the arterial switch to take effect.
22 You were saying that the biggest variable in
23 making an operation successful or not was the coronary
24 arteries.
25 PROFESSOR de LEVAL: It is not exactly what I said, or meant
0046
1 to say.
2 MR LANGSTAFF: It is my inadequate reflection of it, I am
3 quite sure.
4 PROFESSOR de LEVAL: I think that in a statistical analysis,
5 the most important risk factor of success or failure is
6 the pattern of the coronary arteries. In other words,
7 we had identified a number of distributions of coronary
8 arteries which were carrying a much higher risk than
9 others. This is what I said.
10 MR LANGSTAFF: Can you tell us a bit more about that?
11 I think this may relate to recent work which you have
12 completed and is about to be published?
13 PROFESSOR de LEVAL: I think the work we have recently done
14 confirmed that the distribution of the coronary arteries
15 was a key issue in terms of risks of a switch operation,
16 but that was known. We confirmed that. We confirmed
17 that the type of coronary arteries which carries the
18 higher risk is the so-called intramural coronary artery,
19 where one of the arteries is actually inside the wall of
20 the aorta before it comes out of it to be distributed
21 over the surface of the heart. We have confirmed that.
22 What this paper said is that -- does it explain
23 all? We try to say that it does not; we try to say that
24 human factors play a major role in the outcome of those
25 patients; and more importantly, in those with the higher
0047
1 risk, I think.
2 MR LANGSTAFF: We can have a look at your paper; it has been
3 scanned in if you want to look at any table or any part
4 of it. Looking at it first of all generally and then
5 focusing in, what human factors do you identify as being
6 of importance?
7 PROFESSOR de LEVAL: The word "human factors" has been
8 brought from the high-technology industry, aviation,
9 nuclear power plants, where human factors include
10 institutional factors, organisational factors; it is not
11 only just what the human performance is.
12 What the study has confirmed is that even if you
13 allow for the risk for coronary arteries, for example,
14 success or failure can be related to the presence of
15 human factors.
16 MR LANGSTAFF: Those human factors, as you have indicated,
17 are not simply a question of focusing on the surgeon's
18 hands.
19 PROFESSOR de LEVAL: Absolutely not, no.
20 MR LANGSTAFF: But they are looking at, if you like, the
21 team aspects, the organisation, finance, equipment,
22 those sort of things, are they?
23 PROFESSOR de LEVAL: That is right, yes.
24 MR LANGSTAFF: In your study (which is about to be
25 published) you looked at a series, I think, of 230
0048
1 arterial switches, but for statistical reasons, I think
2 you considered 165 of those?
3 PROFESSOR de LEVAL: First of all, I would like to confirm
4 that this paper is not published yet. I spoke to the
5 editorial office last night and they have allowed me to
6 discuss it today.
7 The intention was to look at all the arterial
8 switches done in the UK for an 18 months period. All
9 the cardiac surgeons doing the switch came to Great
10 Ormond Street for one day with cardiologists,
11 anaesthetists, and they all agreed to take part in the
12 study.
13 I had a grant from the British Heart Foundation
14 for the salary of two human factor researchers. The
15 most senior one is a person who has a PhD in a nuclear
16 power plant. We started the work by looking at a death
17 after a switch as being, let us say, a crash in
18 aviation. So we used the same type of technology that
19 had been used in the so-called organisational accident
20 to look at this.
21 We had three sets of data. We looked at the
22 classical procedural and patient variability data,
23 so we did a conventional analysis of the coronary
24 distribution, the age of the patient, et cetera.
25 This is listed in Appendix A -- can we show that? It is
0049
1 page 22.
2 MR LANGSTAFF: If we can have it on the screen, is INQ 6/58.
3 PROFESSOR de LEVAL: So this would be what a detailed
4 scientific paper on the switch would include in terms of
5 procedures, patient specific, and the post-operative
6 variables. This was filled in by the surgeons at the
7 end of each operation.
8 Then we used a questionnaire, again that was based
9 on --
10 MR LANGSTAFF: Let me stop you for a moment and let us
11 scroll down the screen so we have those.
12 PROFESSOR de LEVAL: In the patients you have the coronary
13 arterial pattern which, as has been indicated, is the
14 most important factor related to patients.
15 The procedures: we have all the technology, the
16 procedural aspect, the technical aspect of the
17 operation, the way the heart was protected with
18 cardioplegia, the type of cardioplegia, the technique
19 used to implant the coronary arteries, the presence or
20 absence of an additional ventricular septal defect, the
21 drugs used after the operation, whether the patients
22 required assistance or not, and then the post-operative
23 variables: the bleeding, major renal failure,
24 neurological complications and time of discharge from
25 the intensive care unit. Then you can see I have also
0050
1 the time of extubation, which I am so keen to include.
2 Then we had a questionnaire, which is again
3 a methodology used by human factor experts --
4 MR LANGSTAFF: Page 59(INQ 06/59), please.
5 PROFESSOR de LEVAL: We were guided in this by Professor Jim
6 Reason, Professor of Psychology in Manchester, who, as
7 you probably know, has written a book on human error.
8 He is the father of the theory of organisational
9 accident. He had designed a questionnaire; we call it
10 the STAR Questionnaire, where we had investigated
11 organisational factors.
12 Can you put up page 23 now?
13 MR LANGSTAFF: It is on the screen.
14 PROFESSOR de LEVAL: Thank you. We looked at
15 organisational, situational, team and personnel factors,
16 what I call human factors (not only the individuals) and
17 you can see that organisational factor, the personnel
18 availability, equipment, bed space in the ITU,
19 scheduling of the operation, situational factors: free
20 to concentrate, distractions and interruptions during
21 the operation, physical conditions in the theatres,
22 atmosphere, equipment design or reliability, monitoring
23 of displays or equipment, team factors, pre-operative
24 briefing of the team, confidence in the team members,
25 team's ability.
0051
1 On the confidence of team members, it is
2 important, for example when you have a new first
3 assistant, a new scrub nurse or an anaesthetist you are
4 not familiar at working with, personnel factors related
5 to those who fill the forms. Those forms are filled in
6 after the operation by all the members of the team, so
7 the surgeon, the two assistants, if there were two, the
8 scrub nurse, the perfusionists, the anaesthetists. So
9 we were actually very hopeful that this type of
10 questionnaire would give us a large amount of useful
11 information for our analysis.
12 MR LANGSTAFF: The research you did was, I think, unusual in
13 that it actually had a researcher in the operating
14 theatre for virtually all the operations.
15 PROFESSOR de LEVAL: I am coming to that, yes. The third
16 component of the material we had --
17 MR LANGSTAFF: The next page, please.
18 PROFESSOR de LEVAL: -- was page 24. The human factor
19 researcher was observing the operation and those human
20 factor researchers depict themselves as a fly who is on
21 the wall. The fly can see everything and nobody can see
22 the fly. It is not really true. The switch operation
23 in the UK was better in those 18 months because she was
24 on site; but no way they should not be seen.
25 So they wrote a detailed report of the operation
0052
1 as the operation took place, having some information of
2 what happened pre-operatively and following the patient
3 until the handover in the intensive care unit.
4 When we analysed the data we had collected,
5 obviously we did the usual statistical analysis --
6 I must say, I am not a statistician so I am not able to
7 tell you the details of this analysis, they have been
8 done by two professional statisticians.
9 MR LANGSTAFF: Can I ask you to pause there, because we
10 ought to let those who are looking at this see the full
11 list of major events.
12 PROFESSOR de LEVAL: We will come back to this, if you do
13 not mind.
14 MR LANGSTAFF: Certainly.
15 PROFESSOR de LEVAL: The procedural factors, again, the
16 coronary arteries came up as the most important risk
17 factor, and this is on Table I, I think.
18 MR LANGSTAFF: Table I is INQ 6/63.
19 PROFESSOR de LEVAL: What they did was to look at the odds
20 ratios, which means that you compare two types of
21 coronary arteries and you can see here that if you
22 compare -- I am sorry, Table I shows you the mortality
23 related to the pattern of the coronary arteries, so
24 before I go further, I should say that we divided the
25 outcomes into four categories: outcome 1 is survival
0053
1 with extubation within 72 hours; outcome 2 is survival
2 with no sequelae, but intubation for more than 72 hours
3 and minor reversible problems. Outcome 3, which is near
4 miss, is the need for mechanical support after the
5 operation, such as ECMO or major complications such as
6 neurological problems, deep-seated infections,
7 et cetera. Type 4 is death.
8 The mortality was 6.5 per cent and the incidence
9 of what I have called "near misses" was about 18.5 per
10 cent. If you put the near misses and deaths together,
11 you have a failure rate, if I consider type 3 a failure,
12 of 25 per cent. So we did this in order to have more
13 negative events, to be able to analyse them. All the
14 analyses have been done for deaths or near misses as
15 a negative event.
16 You can see here that the pattern of the coronary
17 arteries, number 6, Table I, which is an internal
18 coronary artery, had a mortality of 25 per cent and
19 a risk of death and/or near miss of 69 per cent.
20 MR LANGSTAFF: There would only be one in the series of 16
21 which was either a category 1 or category 2? One actual
22 case? No, I am sorry, I am wrong on that; there would
23 be 5.
24 PROFESSOR de LEVAL: No, there were 16.
25 MR LANGSTAFF: You are absolutely right.
0054
1 PROFESSOR de LEVAL: If you take Table II now --
2 MR LANGSTAFF: That is page 64(INQ 6/64).
3 PROFESSOR de LEVAL: -- here we have the odds ratios. If
4 you take, for example, the pattern of coronary arteries
5 4, 5 or 6 compared to 1, which is here, you have a risk
6 of death which is 7.5 higher for 4, 5 and 6 than it is
7 for 1.
8 THE CHAIRMAN: It may help to have a blank screen again and
9 do that line again -- it is not a comment on you.
10 Dr Silove found that the pen --
11 PROFESSOR de LEVAL: I hope my surgery is better than this!
12 THE CHAIRMAN: I think we all do!
13 PROFESSOR de LEVAL: The next one, please, the following
14 one.
15 MR LANGSTAFF: The fourth line down. Can we highlight
16 that?
17 PROFESSOR de LEVAL: The odds ratio is 7.5, and I think that
18 if you look at the risk, if you could take the next
19 table, Table III now, the next page(INQ 06/65), it is the same type
20 of table which shows you the odds ratios for deaths or
21 deaths and near miss. Similarly, you can see that 4, 5
22 and 6, which is here, versus the type 1, have an odds
23 ratio of negative event of 5.1.
24 MR LANGSTAFF: Can we highlight the fourth line, please?
25 PROFESSOR de LEVAL: Maybe I could come back later to the
0055
1 timings, because it is a difficult issue.
2 What we did then is to add on to this the human
3 factors, to take into consideration the risk and see if
4 the human factors would influence the risk negatively or
5 positively.
6 Again, we were hoping that the STAR Questionnaire
7 would give us a great deal of information.
8 Unfortunately, we essentially did not find any of those
9 factors as influencing, in a strong statistical way, the
10 outcomes.
11 I do not think there is much point in discussing
12 the reason for that. I think that one of the
13 deficiencies was that those questionnaires were filled
14 in after the event. I think that with hindsight, be it
15 biassed, if you ask a surgeon how he felt before, if
16 everything went fine he felt great; if things did not go
17 very well, he said he felt terrible. So I think that it
18 is not very good to do it that way.
19 Also, you could say that we did not target the
20 proper factors, so those questionnaires may have to be
21 revised. If I had to revise them, I certainly would ask
22 a number of questions to each member of the team before
23 the operation rather than afterwards.
24 But what becomes much more useful is to go through
25 this very detailed report from the human factor
0056
1 researchers. We have only 165 reports because, for 40
2 cases, there were switches taking place in two or three
3 units at the same time in the UK. Also, we dropped
4 5 switches from each researcher as being a learning
5 curve. And one of the youngest of the three researchers
6 was a student who had seen 10 switches and when we read
7 his observations, we really felt that they were not as
8 professional as the other ones. So we have only 165
9 reports.
10 When we looked at those reports, we extracted what
11 we have called a number of minor and major negative
12 events. Those events were listed. We also looked at
13 whether they were compensated or not. The compensation
14 is a mechanism or actions which either rectify the
15 problem or allow people to cope with it but to prevent
16 the effects of a major event.
17 Those events were found to be closely related with
18 outcomes. Before I go to the outcomes, maybe I could
19 now go to the list of those events --
20 MR LANGSTAFF: Page 60(INQ 06/60), Appendix C.
21 PROFESSOR de LEVAL: I think it is probably important to go
22 into some details here. We looked at a number of
23 pre-operative events. The pre-operative events started
24 at the time of induction of anaesthesia -- no, even if
25 there had been a problem before that, they would be
0057
1 included, like a cardiac perforation doing balloon
2 atrial septostomy, or a serious error prior to the
3 induction of anaesthesia. An error in the diagnosis of
4 coronary arteries would not be a major event because it
5 is a minor event, but a cardiac arrest or major sepsis
6 would be a major event before surgery.
7 Anaesthesia, failure to gain sufficient vascular
8 access: one of the key parts of a switch operation is
9 the induction of anaesthesia. Those patients can be
10 very unstable. The circulation of a transposition is
11 very critical because in transposition -- could I have
12 the diagram of the transposition again?
13 MR LANGSTAFF: Certainly. It is ES 1/11 -- I think it is
14 11; it may be 12.
15 PROFESSOR de LEVAL: In the transposition, the blood from
16 the cavae, the blue blood, goes to the right ventricle
17 and then the aorta, and comes back to the right
18 ventricle. The pink blood from the pulmonary veins goes
19 to the left ventricle, to the pulmonary artery.
20 The two circulations are in parallel so the blue
21 blood goes around the body; the pink blood goes round
22 the pulmonary circulations. This is incompatible with
23 life, so those patients should die immediately as soon
24 as they are born, unless there is, some way in the
25 circulation, a communication between the blue blood and
0058
1 the red blood. This takes place usually here at the
2 atrial level. The cardiologist, before the operation,
3 usually does what we call an atrial balloon septostomy,
4 so they make that hole bigger. But yet, the only part
5 of the blood which circulates in the body which is
6 efficient in those patients is the amount of blood which
7 goes through that hole. They live on that amount of
8 blood which goes in both directions at atrial level,
9 so they have a very critical situation.
10 Before the operation they come to the anaesthetic
11 room. They become cold. If you are cold you have
12 a vasal constriction so the resistance in your systemic
13 circulation increases. That can change the way the
14 blood mixes at atrial level.
15 More importantly, we need some monitoring lines,
16 we need an artery, at least two veins. This can take
17 quite a long time and you need expert anaesthetists.
18 So the quality of the patient coming to the
19 operating theatre depends greatly on the expertise and
20 the quality of the induction of anaesthesia.
21 If I can come back now to the major events, which
22 is page 24 --
23 MR LANGSTAFF: That is 60(INQ 6/60), yes.
24 PROFESSOR de LEVAL: Anaesthesia: you have failure to gain
25 sufficient vascular access. Some of those patients went
0059
1 to the operating theatre actually with a peripheral line
2 instead of having proper lines inserted. I quite like
3 the word my researcher has used here: "pincushioning
4 during lines insertion leading to serious cardiac
5 events", which is an attempt to put in needles to try to
6 get access to the arteries.
7 "Delayed diagnosis of major deterioration in the
8 patient's condition": those patients may or may not be
9 monitored, because traditionally the monitors are in the
10 operating theatre and the monitoring of those patients
11 is not as close during the induction of anaesthesia.
12 Sometimes it is absent. Nowadays we would have
13 a saturation monitor and a percutaneous oximeter which
14 will indicate if the patient becomes desaturated, or, as
15 soon as you have vascular access, you should send blood
16 to make sure that the pH is correct, et cetera, but if
17 you spend an hour in the induction of anaesthesia, you
18 may have a patient whose haemodynamic condition has
19 deteriorated greatly between the time of leaving the
20 bed in the ward and the time of being in the theatre.
21 So a lack of monitoring, delaying diagnosis of
22 a serious event: you can see that it happened on
23 8 occasions amongst those 165 patients.
24 The pre-bypass: there were 27 major events.
25 Haemodynamic instability of the patient due to
0060
1 rough surgical manipulations: this is very common if you
2 have -- we see that at Great Ormond Street usually the
3 first two weeks of July because it is when the new
4 residents come. They come from a programme of other
5 cardiac surgery and they manipulate the tissues in a way
6 which is not appropriate for small babies.
7 So the problem here is one of supervision. The
8 staff surgeon should be present at the beginning of the
9 operation. Again there was a great variety of this
10 practice within the UK during the 18 months here.
11 You can have an error, of course, in the
12 laceration that the doctors have to use in performing
13 bypass. A careful surgeon will do it on bypass or if it
14 is difficult, will not pursue a dissection before having
15 established a bypass.
16 Recognising a deterioration in the patient's
17 condition, low blood pressure, acidosis before the
18 bypass: again, this is the jargon of the human factor
19 researcher, which is "the cognitive tunnel vision",
20 which is very very common, is what we do. You
21 concentrate on a small problem disregarding what goes on
22 around yourself. This is a fairly common problem.
23 Delayed administration of heparin: heparin is the
24 drug you use before connecting the patient to the
25 heart/lung machine to make sure the blood does not clot
0061
1 into the extracorporeal circuit. I think it was never
2 forgotten, but there was some delay at least in one
3 case.
4 During bypass obviously there are a number of
5 errors. Quite a few were related to the implantation of
6 the coronary arteries, or technical errors. Obviously
7 if you make a hole in the aorta it is not a very good
8 start and it is a major error. But we had perforation
9 of the aorta when the cardioplegia was given. The
10 cardioplegia is a solution of potassium which is
11 arresting the heart before the operation.
12 Major perfusion problems, such as air in the
13 arterial circuit would be a major problem.
14 Anaesthetic problem: failure to cool the
15 cardioplegia. The cardioplegia should be cold to have
16 a proper effect.
17 Failure to flush the cardioplagia.
18 Post bypass, there were quite a number of events.
19 Again, it is very often due to a lack of supervision,
20 that the surgeon would leave the operating theatre
21 instead of doing the haemostasis himself or herself, or
22 surgical omissions, lack of a pacing wire, serious
23 deterioration of a patient when you close the sternum.
24 Quite frequently the heart is oedematous and has
25 increased in volume at the end of the procedure. If you
0062
1 approximate the sternum, the blood pressure goes down.
2 Quite frequently we leave the sternum open and again,
3 this is often not seen or not observed properly by
4 surgeons and more often, when the closure is done by
5 a junior doctor, he does not have sufficient experience.
6 So this is a list --
7 THE CHAIRMAN: Mrs Howard has one question, interrupting you
8 for a moment.
9 MRS HOWARD: I am sorry to interrupt your flow. In the
10 section headed "During Bypass" you refer to the LeCompte
11 manoeuvre. I have read other details about the LeCompte
12 manoeuvre. Could you explain a little about why that
13 clearly is seen as a technical error in terms of
14 twisting the artery? Why is that particular manoeuvre
15 seen as a problem?
16 PROFESSOR de LEVAL: Can I have the drawing of the switch?
17 MR LANGSTAFF: ES 1/12, please.
18 PROFESSOR de LEVAL: No, the drawing from my paper on the
19 cluster, the switch operation.
20 MR LANGSTAFF: That is INQ 6/3.
21 PROFESSOR de LEVAL: In transposition, the pulmonary artery
22 is behind and the aorta is in the front. You transect
23 the pulmonary artery just above the pulmonary valve.
24 Then you bring the distal pulmonary arteries forward in
25 front of the aorta to be anastomosed to the neoaorta
0063
1 here. That is called the LeCompte manoeuvre, or French
2 manoeuvre because LeCompte is French. When you do that,
3 you can obviously twist the pulmonary artery. I think
4 on one occasion the surgeon managed to rotate it
5 180 degrees and bring it to the neoaorta. Obviously the
6 blood has some difficulty to find its way out and that
7 had to be rectified.
8 MRS HOWARD: So the issue is how the arteries are
9 manipulated, not the use of the manoeuvre?
10 PROFESSOR de LEVAL: That is right. This is an error in
11 doing the manoeuvre, but the manoeuvre is a good
12 manoeuvre. Can we go back?
13 MR LANGSTAFF: Can we go back, please, to 6/60?(INQ6/60)
14 PROFESSOR de LEVAL: If I take those major errors and see
15 their impact, again, we have looked at those errors,
16 taking into consideration the risks of the coronary
17 arteries. So we try to analyse this taking into
18 consideration the type of coronary arteries we have.
19 THE CHAIRMAN: May I ask a question? You stop at the point
20 of transfer to the intensive care unit. Does the
21 analysis follow while the patient remains in intensive
22 care?
23 PROFESSOR de LEVAL: No. Those reports ended after the
24 handover. What follows the handover is part of an
25 on-going study, but I can tell you already now that one
0064
1 of the most crucial parts of the treatment of those
2 patients is particularly that handover, when the patient
3 has to be transferred from one team to another team.
4 To give you an example, I would do a difficult
5 switch operation with an anaesthetist with whom I have
6 been working for 25 years. A switch operation like this
7 may last 6/7 hours, and both of us, having together 50
8 years of experience of paediatric surgery, would watch
9 that heart on an electrocardiogram beat by beat for
10 5 hours.
11 The patient is transferred to the intensive care
12 unit. The patient has lots of lines, electrocardiograms
13 and so forth. There are four or five nurses trying to
14 connect the patient to a new monitor, so the patient is
15 not monitored for that period of time because we have to
16 reconnect it. Then you may or may not find a system
17 where your level of expertise will be matched in the
18 intensive care unit.
19 This is the big problem of a multiple team dealing
20 with a patient. Ideally, what you have to do is to make
21 sure that the knowledge of that particular patient at
22 the time of leaving one team must be acquired by the
23 other team to make sure you keep the same level of
24 expertise. The only way to achieve that is to have
25 either people who overlap that part of the treatment, or
0065
1 having the two teams overlapping at one period of
2 treatment, which is either at the end of the operation
3 or when you come to the intensive care unit. It is not
4 uncommon that the person who takes that patient in
5 charge would be a junior doctor who is in the intensive
6 care unit for the first week, who hardly knows what
7 a switch is, and then you have the whole expertise of
8 two people, 50 years which goes down to nothing. It is
9 the ups and downs of the expertise which is causing
10 major problems in the handover. But again, this is
11 being investigated now and I do not have any hard data
12 to discuss this here.
13 THE CHAIRMAN: When do you expect that study to be
14 completed?
15 PROFESSOR de LEVAL: It depends on funding.
16 THE CHAIRMAN: Mrs Howard has a question.
17 MRS HOWARD: Directly for the point about the ups and downs
18 of expertise, I note in your lecture that you very
19 graciously passed on to us that you talked about the
20 difference between the junior doctor and the very
21 experienced intensive care nurse. I wonder whether
22 there is any comment you would want to make about
23 handover at that point and where perhaps the skilled
24 nurse could come into it.
25 PROFESSOR de LEVAL: This is a slightly different issue.
0066
1 I think that in that lecture I summarised my views on
2 how high-technology medicine should be practised.
3 I think the care should be provided by specialists, and
4 it is a mistake to believe that the specialists are the
5 doctors, basically.
6 If you look at the analysis of those major events,
7 this is Table VII.
8 MR LANGSTAFF: Page 69(INQ 6/69).
9 PROFESSOR de LEVAL: This table shows you that the number of
10 major events per case has a very, very high risk of
11 deaths or near misses with a p-value of less than 0.001
12 for both negative outcomes.
13 Obviously this is not surprising. If you transect
14 the aorta, the patient will not survive unless you do
15 something, so it is a bit of common sense. I do not
16 want to say that it is a highly scientific discovery
17 here. Obviously if you have a major problem, it can
18 lead to a bad outcome.
19 What is more important is number 2, which tries to
20 look at the effect of compensation. If you have
21 a compensation, you can reduce very significantly your
22 risk of death or near misses. A compensated major event
23 will not lead to death if it is compensated.
24 This is, I think, an important message of this
25 study which goes along very well with what is done now
0067
1 in high-technology areas such as the aviation industry.
2 People realise that errors are unavoidable and there are
3 always failures. Rather than concentrating on
4 abolishing errors and failures, what the human factor
5 experts do today is to educate people to compensate.
6 The way I see what I do today, or the switch
7 operation, we should look at this as an eventful
8 procedure. It would be impossible to remove the fact
9 that it is eventful. What you have to train people to
10 do is to compensate for those negative events. It will
11 be impossible to avoid a major problem, but you have to
12 tell people to compensate to make sure that you can
13 reach a good outcome.
14 The study shows that a good compensation will
15 provide a good outcome, even with difficult coronary
16 arteries.
17 We had a patient who had an intramural coronary
18 artery. The coronary was implanted 7 times, so you have
19 the resilience of a surgeon who never gives up, who
20 carries on, who is meticulous from the first minute to
21 the last minute, even after 10 hours, and this can still
22 lead to a good outcome. So that was an important
23 message of this today.
24 The minor events: if you now go back to the
25 Appendix --
0068
1 MR LANGSTAFF: It is page 61(INQ 6/61).
2 PROFESSOR de LEVAL: Again, I think it is interesting, but
3 I do not think we should give it too much value. It is
4 the first time we do this study; it is the first time we
5 do an analysis like this. I am sure next time we
6 probably would do it more prospectively, but I think
7 that the results of this analysis gives some indication
8 of what human factors can play in terms of the role they
9 can play in terms of outcomes.
10 If you look at the pre-operative events,
11 a co-ordination problem with the bloodbank, problems
12 with the availability of a bed ITU -- I hate to say it,
13 but you would not believe the number of times -- it
14 happens at Great Ormond Street -- that a patient is fed
15 one hour before going to the operating room. It is just
16 unbelievable; it happens all the time. I do not know
17 how to prevent that. We put big signs in red, in blue,
18 in green, but despite that it happens all the time.
19 I do not know what to do.
20 Problems with the scheduling of operations of
21 course happens also very frequently. You can try to
22 have any protocol, the patient will be booked in the
23 wrong theatre at the wrong time.
24 These are all small things but which at the end of
25 the day count for the result.
0069
1 Anaesthesia, cognitive tunnel vision on lines of
2 insertion, to neglect monitoring of the patient on the
3 electrocardiogram.
4 Inappropriate delegation of tasks is one of the
5 features we have noticed on many occasions as well.
6 Equipment problems: the alarm systems are
7 a nightmare as they do not alarm when they should.
8 That happens all the time. Incorrect ECG readings,
9 human resource problems, no general anaesthetic
10 assistant is available, no technical assistance.
11 Pre-bypass and bypass, positioning and tension errors by
12 the surgical assistants. Very often the surgeons try to
13 have an assistant who is familiar with the operation,
14 but it is sometimes impossible in some institutions,
15 doing for example adult and paediatric cardiac surgery,
16 you have junior doctors who rotate between different
17 services and you may end up one morning by having
18 a junior surgeon to help you do your switch operation,
19 which is important. It is a minor event, but as you
20 will see, it probably counts at the end of the day.
21 Inappropriate task delegation at every level,
22 surgical, anaesthetic, perfusion, an inexperienced
23 assistant, problems with equipment, communication
24 errors, for example, the indication that the heparin has
25 been given or that some drugs are started by the
0070
1 anaesthetist without having checked with the surgeon
2 that they were happy to go ahead with the introduction
3 of a new drug.
4 Absence of a senior team member in the theatre:
5 this is also very interesting. If you look at the
6 seniority level in the theatre and the atmosphere in the
7 theatre, you can see that you have a high level. If the
8 surgeon leaves the theatre, a number of things can
9 happen. Usually it continues to be all right, but
10 someone turns the music on, for example, or the senior
11 anaesthetist will also leave to have a break. You end
12 up, again, by a very high level of experience and
13 expertise down to a very low one in a very short period
14 of time, which can also be important.
15 Perfusion problems, airlocks, poor visibility
16 because of unsatisfactory blood aspiration.
17 Post-bypass, the same error of positioning,
18 tension and instrument handling errors, communications,
19 anaesthetic problems, communication problems during the
20 handover in the ITU. I have already mentioned that
21 before. Lack of seniority during transfer in the team
22 admitting the patient to the ITU.
23 This is very, very important. I think that this
24 part of the operation has to be supervised by the
25 presence of the most senior people. It is one of the
0071
1 most difficult parts of the operation.
2 Absence of monitoring.
3 When we look at those minor events, there may be
4 many more, we missed some or some should not be counted,
5 but if we assume that those events have some values, if
6 you put them in a statistical analysis.
7 MR LANGSTAFF: Back to page INQ 6/69, Table VII.
8 PROFESSOR de LEVAL: You have here the number of minor
9 events per case is 3; p-value for both deaths and near
10 misses of less than 0.001, which to me is the most
11 important message of this analysis: that those minor
12 events which actually are often not even noticed by
13 those who perform the operation, they have a cumulative
14 effect. You add them on. When in isolation there is no
15 influence, but at the end of the day they have a major
16 impact on outcome, having again taken into consideration
17 the stratification of the patient in terms of risk for
18 coronary arteries.
19 MR LANGSTAFF: Can we just highlight them?
20 PROFESSOR de LEVAL: It is number 3 there (indicating).
21 I think this is probably the most important message of
22 this research, besides the importance of compensation.
23 To try to go a bit further in the understanding of
24 those minor events, we looked at those having accounted
25 for the major event and the major compensated event to
0072
1 see if it is still important. This is Table VII.
2 MR LANGSTAFF: Page 70(INQ 6/70).
3 PROFESSOR de LEVAL: I forgot to say, there is no
4 compensation for those events, very little effective
5 compensation. Precisely because they are not noticed,
6 there is no attempt to compensate them.
7 If you have Table VIII, you a number of minor
8 events are still of significance, the last line of this
9 table here, for the deaths; certainly for the near
10 miss. In the near miss, you still have a p-value of
11 0.001, having taken into consideration both the major
12 event and the number of uncompensated major events.
13 My interest now is to try to understand the
14 dynamics of those events. We understand a lot of this
15 is common sense, although we think we have demonstrated
16 that. The last figure which I will show here is, as we
17 began this research by saying that we could look at the
18 deaths after the switch as an organisational accident,
19 I have taken Jim Reason's diagram of an organisational
20 accident, which is figure 3, I believe.
21 MR LANGSTAFF: It is 78(INQ 6/78).
22 PROFESSOR de LEVAL: Jim Reason's theory of accident is the
23 'Swiss cheese' model, where in theory it says that an
24 accident is the result of the concatenation between
25 active and latent failures. The active failures are the
0073
1 errors or failures made at the sharp end of a system:
2 the surgeon, the pilot, the nurse, the engineer. Then
3 you have latent failures.
4 According to this theory, those casts are the
5 different mechanisms. The perforations are the failures
6 or the errors. He said that for an accident to happen,
7 you need all those errors to be aligned so that an arrow
8 can penetrate them to produce the accident.
9 I think this is probably applicable to the minor
10 events. If you align them, they have a cumulative
11 effect and they can lead to death. I wanted to show
12 there you have a high hazard procedure, let us say an
13 intramural coronary artery. For this we know it is
14 a high risk procedure, but for this to lead to death you
15 have the human component. Many of the minor events,
16 although they take place at the sharp end of the system
17 and the operation, have their origin much further away
18 from the operating theatre: there is the teaching
19 system, you have an institution which is dealing with
20 other than cardiac surgery, an intensive care unit
21 dealing with meningitis and switches, et cetera. But
22 all these minor errors, or many of them, actually have
23 their origin not on site but further remote from them.
24 I think this is the essence of this paper.
25 MR LANGSTAFF: We have come to a time when we would
0074
1 naturally, sir, have a break. I am conscious that we
2 still have a number of questions to ask Professor
3 de Leval, and much that he can tell us not only by
4 further exploration of his recent paper but by his work
5 on the CUSUM analysis, looking at the individual surgeon
6 and his own performance.
7 But we have the advantage that Mr Stark brought
8 with him some slides which will be shown by making
9 arrangements over the luncheon break. For that reason,
10 we will need something of an hour rather than our more
11 usual 45 minutes. I shall invite him to present his
12 slides to us immediately after the lunch break. It may
13 be that we may then need to dismantle that before we
14 begin in the format that we have at the moment.
15 THE CHAIRMAN: What you are saying is that the time needed
16 to continue with Professor de Leval will probably be too
17 long from now for us to carry on?
18 MR LANGSTAFF: I think that is right.
19 THE CHAIRMAN: I take your advice, and thank you.
20 Therefore, yes, we will adjourn for an hour because
21 there is this need to set up some equipment. I express
22 my thanks to all three of our experts. We will continue
23 what is a very important dialogue after a break of about
24 an hour, therefore reconvening at almost 1.15.
25 (12.15 pm)
0075
1 (Adjourned until 1.15 pm)
2 (1.15 pm)
3 MR LANGSTAFF: I am going to invite Mr Jaroslav Stark to
4 talk us through the slides which he has brought with
5 him. May I say he will be happy after each slide to
6 take any questions that there are -- he will not
7 necessarily invite them -- but if there are any if the
8 Panel would wish to ask them in addition to any that
9 I might ask or either Professor de Leval or Dr Silove
10 might ask.
11 MR STARK: I was asked if I can briefly illustrate the
12 complexity of the work we do. I think the first slide,
13 as you see here, is exactly how the slide should not
14 look. It is far too much, but I just am putting it up
15 to show you the spectrum of various procedures that
16 we have to do, and I think one has to realise that in
17 some patients we do several operations at the same time.
18 In essence, when you look at the work of
19 paediatric cardiac surgeons (Slide), you can say that we are
20 closing defects but also creating defects. We are
21 opening valves, we are repairing or replacing valves,
22 removing or bypassing obstructions, and you have all the
23 combinations between them. So, unless the surgeon has
24 full understanding of the diagnosis and all the
25 procedures, it is quite clear that at any stage one can
0076
1 fail.
2 Now, the procedures that we do vary (Slide) from a very
3 simple one where you just ligate this structure which is
4 patent ductus arteriosus connection between aorta and
5 pulmonary artery, and, as probably Dr Silove explained
6 yesterday, nowadays that can be done during cardiac
7 catheterisation by placing a little umbrella or coil, or
8 what is called a Blalock-Taussig shunt, when there is an
9 obstruction to flow from the right ventricle to the
10 pulmonary artery, you bypass that obstruction by either
11 turning down the artery or interposing a small segment
12 of prosthetic material.(Slide)
13 But the majority nowadays, the majority of
14 operations are done on heart/lung machines (Slide) , on so-called
15 open heart procedures. Again just to familiarise you
16 briefly with what it involves, the blood from the
17 superior and inferior vena cava, that is the blue blood
18 which normally returns to heart, is syphoned off from
19 the body, from the heart to the oxygenator, which
20 provides the exchange of gases, to the pump, to the heat
21 exchanger, because we may have to cool and then re-warm
22 the patient, and then it is pumped back to the patient.
23 Now, the technology of this has improved really
24 tremendously. This is early 1970s, (Slide) when we were using
25 this monstrous machine which was used even for small
0077
1 children and infants. The volume of fluid which we had
2 to put in was something like 5 litres. Now, if you can
3 imagine that the volume of a 3 kilo baby is only 240
4 mls, you could imagine what biochemical and fluid
5 changes it can cause, so then we come to the further
6 development, and nowadays you can see a tiny machine
7 that squeezes under the table next to the surgeon.(Slide)
8 It is much kinder to the organs of the child but also
9 it has some dangers because the time that elapses
10 between the alarm sounding that the level of the blood
11 is too low, and that means that you can start sucking in
12 some air and pumping to the patient, may be as short as
13 five or six seconds. So you realise that that means the
14 perfusionist has to be on alert all the time.
15 It also shows you that the operation, what we
16 already heard so elegantly from Marc, where you have
17 possibility of errors, it is errors of all members of
18 the team. You have here usually two anaesthetist
19 surgeons, two assistants, one or two scrub nurses, two
20 perfusionists, and that has to be co-ordinated and all
21 in their individual field have to be very professional.
22 Then I just thought I would illustrate one
23 operation to you which probably will show many of these
24 points. I also brought a video of the AV septal defect
25 repair, but unfortunately we do not have the facilities
0078
1 to show that.
2 This is the repair of truncus arteriosus which is
3 a condition which usually presents in the first few days
4 or weeks of life and it consists of, instead of two
5 arteries, aorta and pulmonary artery coming out of the
6 heart, there is just one common trunk, and, from that,
7 the aorta and the pulmonary artery arises.
8 So here you can see the arterial cannular for
9 bypass; aorta is cross clamped and the pulmonary artery
10 has to be detached first (slide). Then the defect in the aorta
11 has to be closed. The right ventricle is opened and
12 here you can visualise the hole between the two pumping
13 chambers, right and left ventricle, and that is closed
14 with a piece of prosthetic material. (Slide) In the next step,
15 the hole is closed but then you have to --
16 MR LANGSTAFF: Can I just stop you there? Is the prosthetic
17 material Goretex or is it likely to actually be tissue?
18 MR STARK: That can be more or less anything, depending on
19 your preference and the time when it was done.
20 Originally this was a piece of Dacron, then it was
21 Goretex. Some surgeons would nowadays use bovine
22 pericardium, so it may be a different material but the
23 important thing is it has to close the hole. It will
24 not grow and it will be covered by the heart's own
25 tissue probably in three to six weeks, so it will just
0079
1 form the substrate of the closure and on both sides this
2 will be covered by the heart's own tissue.
3 Then, the last step of the operation is to bridge
4 the gap between the pulmonary artery and the right
5 ventricle. On this slide, this is a piece of Dacron
6 tube with the porcine valve inside, but nowadays most
7 probably this will be done with a human valve, either
8 aortic or pulmonary, which is usually harvested from
9 cadavers.
10 MR LANGSTAFF: So this slide would illustrate the procedure
11 at what period of time?
12 MR STARK: This is almost at -- you mean period of time --
13 MR LANGSTAFF: Chronologically.
14 MR STARK: Chronologically. It would be probably up to
15 mid-1970s. After the mid-1970s most of us use the
16 homograft.
17 Here, you can see (Slide) more or less the completed
18 operation when the blood from the left ventricle will go
19 to the aorta and from the right ventricle through this
20 conduit to the pulmonary arteries. I think it
21 illustrates one more point, that if you do this
22 operation on a small infant, as the infant grows,
23 whatever you put here will not grow, and one or two
24 replacements of this tube will be required after
25 10 years, 15 years or so.
0080
1 THE CHAIRMAN: Mrs Maclean has a question.
2 MRS MACLEAN: Just a small point. I was interested in the
3 use of valves harvested from cadavers. Presumably these
4 must be the proper size. Is the supply of such valves
5 problematic?
6 MR STARK: Supply is always a problem but the size of the
7 valve -- we usually try to put a size bigger than would
8 be appropriate for the child so that the child will, so
9 to speak, grow into it, but for some very small babies
10 we need very small valves and supply of those small
11 valves is really a major problem.
12 Here, the operation is completed (Slide). As I say, it is
13 only one of the operations which we do but I thought
14 perhaps it would illustrate some of the most technical
15 difficulties, the problems that the child will grow, so
16 that we have to keep that in mind, and, although we
17 cannot in this instance avoid re-operation for
18 replacement of the conduit, we should at least try to
19 make the number of replacements as few as possible.
20 MR LANGSTAFF: Roughly how long would an operation such as
21 that shown on the screen take today?
22 MR STARK: I would think that it would not take less than
23 five and hopefully not mo