Director of Breast Cancer Services for Royal Melbourne & Royal Women's Hospital's, Professor of Surgery at the University of Melbourne, Director of Advanced Surgical Training, Royal Melbourne Hospital
Download powerpoint presentation by Professor Bruce Mann (PDF 311 KB)
Introduction by Norman Swan:
... who is a surgical oncologist, very strong interest in breast cancer surgery. He works at both the Royal Melbourne, Royal Women's and Peter Mac and he is still President of COSA? No? You've lifted that burden off your shoulders?
Please welcome Bruce.
Professor Bruce Mann:
Thank you very much for the opportunity to come here.
Feeling, as a surgeon, not quite out of my comfort zone talking to a group of radiation oncologists. I feel very comfortable talking to fellow loco regional therapists.
What I'm going to give today is my view, as a surgical oncologist and as someone who's been involved really in leading multidisciplinary teams in a number of different cancers, a view on what we need to seriously address in order to make the development of the regional cancer centres of excellence, to what do we need to do to make this a great benefit to our patients to really improve the overall care that we're delivering. And I think that there are some huge - there are obviously some huge opportunities but there are some risks associated with the developments.
(Small side discussion in getting slides to load)
I guess these are some personal views or considered personal views that in order to give best care to our patients, we require appropriate sub-specialisation. Now some of this sub-specialisation will occur naturally, some will need encouragement and in designing systems which is what this meeting is about, we need to design the systems and implement some changes to facilitate it.
I think the whole area of sub-specialisation is one where there is a lot of debate. In complex areas, I believe there's agreement that outcomes are better. That may be a knowledge issue, it may be an equipment issue, it may be a technical issue. In less complex areas, in less complex diseases, that link might not be so clear. However I think it is very dangerous to assume that the link isn't there and I'll just give you a couple of examples.
The examples I'll give you, the data will be from colleagues from the US, it seems much safer to give American data. The first is an area where there isn't much sort of theoretically - there is general agreement or there's philosophical agreement that a condition such as a Whipple pancreaticduodenectomy should be done in a specialised centre because outcomes should be better. This is an excellent example to prove that from the state of Maryland where Richard Schulick is the head of the Pancreas Program at Johns Hopkins.
They collected data over a 10 year period, '95 to 2004, and they looked at the volume distribution of Whipple procedures done in hospitals that did between one and 10 per year - and you can see there were 38 hospitals - 11 to 29 procedures per year and more than 30, that being Johns Hopkins. The striking number is the mortality rate and you can see here that the mortality rate in the state of Maryland - this is recent experience with modern anaesthesia, with modern antibiotics and supportive care - a huge difference between the low, intermediate and the high volume hospital.
The length of stay, something of a surrogate of care, it is a few days lower in the high volume hospital. The reason that the low volume hospital has shorter length of stay then the intermediate volume is firstly that the post operative mortality actually led to a lower length of stay. But in the intermediate risk hospitals, the intermediate volume hospitals, the quality of care was such that a number of patients had the morbidity that was rescued at the cost of long length of stay and a long and high cost.
Here is the hospital charges to the insurer, according to hospital, and Johns Hopkins is an expensive hospital, however the cost per patient for pancreas surgery significantly less in Johns Hopkins than it is even in the low volume hospital. Quite clearly a compelling case that pancreas surgery should be done in a centre of excellence. I wouldn't go too far to say any high volume hospital will be good. I think data as we've heard is necessary but this link is likely to continue.
The radiation here - I'm being very brave mentioning something about radiation - is the TROG trial that was mentioned this morning. I thought, and I'm sure that there are a number of issues but I'm also - so this is the trial that showed that in the centres that took part in the study where there was significant radiation compliance issues and there were significant radiation compliance in the low volume centres, that the outcome was poorer. There's the loco regional - the Kaplan-Meier curve looking at loco regional failure free survival and the low line is the one where there was a compliance issue that was predicted to have a therapeutic impact.
The other line is the one where there was a compliance issue not predicted to have impact. The top three lines are not statistically significant but the lower one is significant. I accept the argument that there will be a number of issues but I also have in my mind the knowledge that a patient being treated in a trial is likely to get, on average, a better level of care than a patient not in a trial. So for a patient in a trial to have this poorer outcome, makes me ask the question of what about everyone who's having this type of treatment off study?
So I suspect if it could be measured that the pancreas cancer, the striking difference probably does exist with other disciplines.
Top of page
What about medical oncology? The medical oncology area, there are some areas that quite clearly need a specialist centre; bone marrow transplant is the striking one. I'm not aware of data looking at some of the more toxic chemo programs whether there is a difference in morbidity, mortality according to centre. I suspect if the data was there - or it might be there, I'm not aware of it - that it might show a difference.
But certainly when looking at what can be delivered in medical oncology according to centre, clinical trials is one of the areas where the phase one studies or the studies that are applicable to a small niche of patients are only going to be available in a few centres, like to be a high volume centre. So if we're looking at a system to provide optimal care to the vast majority of the population, this issue needs to be addressed.
So it's easy when you're looking at this to say, well, there's the complex disease and we'd go the Whipples, we'd look at oesophageal cancer, recurrent rectal cancer, sarcoma, bone marrow transplant and we can think of a few. But is it only about complex diseases? The area that I practice in most is breast cancer and technically breast surgery, even though I would like to deny it. It's not the technically most difficult of surgery there is and well trained general surgeons can generally do a lumpectomy or a mastectomy just about as well as I can.
However in my years since I finished my training in breast cancer, the whole issue of sentinel node has come in. More recently the management of the positive sentinel node. The issue of the selection of a patient, radiation area - which patients with DCA should be radiated. Partial breast radiation. Systemic therapy for triple negatives. We still see referrals that this patient has a seven millimetre cancer, clearly she doesn't need any systemic therapy, yet it's a triple negative or a HER2-positive cancer, where during my training 10 years ago, these were not given systemic treatment. Clearly there is a strong argument that they should. The management of the gene carriers, the identification of the family history.
These are all things that have required significant effort on my behalf and others to remain current. How is it that or how do we make sure that the patient being looked after by a surgeon, a team who are responsible for the care of patients from a wide range of tumour are able to stay current? Thyroid cancer is another. Thyroid cancer, not many people die of it, it's maybe- it's not a lung cancer, it's not a pancreas cancer.
However the issue is there, it's another disease I know quite a bit about. But hemithyroidectemy versus total thyroidectemy, which should be done? Central neck dissection. Big debate. Difficult operation. Easy simply not to do it. Selection of patients for adjuvant radioiodine. How should these patients be surveilled? These are all things that a thyroid management multidisciplinary team will be on top of and we believe will lead to better long term outcomes. But how can this be delivered across a system?
So there is - I would put - a growing understanding that sub-specialisation is good. In all these areas, the sub-specialist is likely to have that edge of knowledge or experience or basically the time to remain current, so that when there is the case that requires a slight variation in normal, they are more likely to get it. But it can have major travel and logistic issues for patients. It also can have major impact - sub-specialisation can have major impact on practitioners and this is a big issue in surgery. The surgeon in the regional centre is required to provide a broad range of treatment. Now if various areas are sort of removed and said, "Well, that should be done down in town", there are a number of implications of such a suggestion.
The patients are voting with their feet. A JCO paper from last year - I hope you can read it - but what this shows is it looks at two time points, I believe at '97 and 2005. I can't read it from here. It looked at the proportion of cases in parts of north eastern US where the surgery was done in a very low, low up to a very high volume hospital. You can see for oesophagus and pancreas, that patients or doctors have worked it out. So that in the mid '90s the proportion being treated in the range of hospital, of cases being treated in the range of hospitals, was roughly the same. Yet the real move, fewer being treated in the low volume hospital, many more in the high volume hospital.
Somewhat similar but less marked has occurred in colon cancer - that's at the bottom left. In rectal cancer, there has been no apparent trend, which is interesting given that from a surgical point of view most general surgeons would recognise that rectal cancer is one of the areas where sub-specialisation has led to a great benefit.
The paper also looked at the impact - now this is north-east US, there's lots of hospitals there, travel distances aren't large. You could probably multiply it by a couple of orders of magnitude. But if you look there, the top line is for oesophagus, the distance travelled by the patient to get to the centre and you can see that there really is a linear association from the low volume hospitals very close but to travel to the high volume hospital, people are travelling, I think it's about 20 times as far on average.
Now if that were translated to the Australian setting, there are some long distances, clearly this is politically important and this is where the regional cancer centres are important, both by being potentially the medium to high volume centre as opposed to the low, and by reducing some travel. So coming back, we're looking at optimal care. What are some of the principles and I think it's clear that selected patients should be treated in major centres. There is little debate over that. Pancreas cancer should be treated in a centre that is very used to doing it, or I must say at least the surgery of it.
Selected aspects of treatment of many cancers should be done in a major centre. At the same time, all patients will require aspects of their care delivered locally. It is important that we think of cancer care, there's surgery, there's medical oncology, there's radiation oncology. There also is supportive care to the patient and all of those around and that can only be delivered locally. But that can be difficult to deliver that local care if it's completely divorced from the other treatment.
I would say, and I think I've given some examples, that it may be hard to identify the patient or the aspect of care that needs to be delivered centrally. What are some drivers for sub-optimal care - I think sub-optimal care would be patients travelling long distance for a treatment that can and should be delivered locally. The patient with a very routine breast cancer who comes from 400 kilometres away to see me - and I do an operation that I know well trained colleagues locally could have done just as well - it's a waste of time. It's a waste of time and particularly if there is a complication, I'm sure their treatment is sub-optimal.
Then the other side of the sub-optimal care is the patient having treatment locally that is best delivered somewhere else. What are the drivers? We have to accept when everything is done that is sub-optimal, money is likely to be an aspect and I know we would like to say that everyone in the profession and associated with it has the best motivation and I think that, in general, we do have a very good motivation. But I think it would be somewhat na´ve to say that this is the only group of humans for whom money does not drive some decisions.
The other aspect is pride. There is a lot of ego. I reflect when I'm having, in a particularly feisty department of surgery meeting, that people didn't get here if they didn't have an ego and when you've got a lot of egos together, it doesn't always lead to an optimal outcome. So how do you move to optimal care? The first line, I'm just going to put as a line. I'm happy to talk about it. But role delineation is essential. It is not appropriate - I was talking to Rhonda Coleman before about the importance of choice, should patients be able to choose to have treatment locally. And you could say, well, a well informed patient should be able to choose to make an error. It's an interesting philosophy.
But I must say, I think we've gone beyond the point where a clinician, a practitioner and I'll say a surgeon who has done a single thyroid a year, has a patient with a thyroid cancer and feels like doing a thyroidectemy and a central neck dissection and just does it, when there is a colleague at the same hospital who does this operation every week. I think that shouldn't happen, if it can be done by peer pressure that would be good. If it's going to be done by peer pressure, why not back it up with some regulation and I think it's a discussion that needs to be had in the colleges. But I think it's a discussion that needs to be encouraged and, dare I say, enforced by our system.
But having said that, there are things that can be done in our system to facilitate, the CanNET is a local example of allowing the interaction between institutions that can improve things. I've got an example that I heard from a colleague in UC Davis, Sacramento, California, Richard Bold. They have a world leading program of what they're calling virtual tumour boards and I'll just run through it because I really think that this is a model that could be very helpful for us.
So UC Davis, it's an NCI accredited cancer centre. There are four community based programs. Some distance, up to 185 kilometres away from the central site. Some of them in fairly impoverished areas in the central valley of California. Truckies up near Lake Tahoe, which is very much a rural centre. They set up with very good IT support, these virtual tumour boards where all four sites are online at the same time and they have the facilities to present high definition radiology, pathology and have a free discussion.
Over the last few years, they've set up four virtual tumour boards, which meet weekly. The areas of genitourinary urinary, breast, thoracic and GI. The system that they have introduced, they have introduced a series of rules where all patients, patients are to be presented prior to initiation of therapy. It's not every patient but the time of presentation is before therapy. They develop standardised templates, the requirement for the presentation to be prepared and submitted in advance, particularly to allow radiology and pathology to prepare. Their wording was servility despite passion was demanded. In particular, they were aware of the potential sensitivities associated with real or perceived criticism.
An attendance was encouraged but not demanded and effort was gone to encourage attendance of the community centres even when cases weren't being presented. The findings were that, in general, UC Davis clinicians were sub-specialist, they went to their tumour board. The community participants were mainly generalist and didn't attend all each week but would attend a number of different ones.
I've asked Richard - he sent me a manuscript but I don't believe it's published and there isn't the evidence that this has led to improved care. But their perceptions of the results were that the community sites benefited because they had access to the sub-specialist's opinion without having to send the patient down for management. So there wasn't the fear of losing the patient.
It certainly facilitated appropriate referral that once presented, the referral becomes easy. Access to information, and these are some of the things I was discussing in breast, this is ideal for it, so that the differing case, triple negative, no, it didn't use to be bad, we've now realised that it is and then access to various services.
The central site, it benefited because when the patients did come down, they were well prepared and they knew what they were going to get. The personal links developed and they reported access of patients for their clinical trials. But most importantly, the patients benefited with this access to the academic oncologist without the need to travel and also, although not reported, the ability to divide the patient journey aspects delivered close to home but often the adjuvant chemotherapy in the major centre, but the adjuvant chemotherapy could be delivered close to home.
Potential barriers, and I think this is something that we would obviously face - or will face - when we introduce whatever we're going to do, which I hope we will. Changes in schedule, and their report was that the community oncologists at the small centres were expected to fit the schedule of the central site. I've put that there, I suspect that is likely to be what happens. But with time schedules can be managed.
The requirements for a preparation was reported as a barrier but something that was important, so that the quality of presentations were high. This perception of criticism was a barrier and the chair's, the leads, had to work very hard to prevent it. Then some sensitivity regarding the differential - the small town versus the big city.
So in summary, what I would say, optimal patient care clearly requires coordination for the regional patient to get care, optimal, not just treatment but the whole package. This is not only a medical condition. It's never easy to develop and maintain these type of links between the sites because these teams will only work if the people know each other and there is mutual trust and respect.
That's partly when something goes wrong or a diagnosis is delayed. You will present that, they're cases who need to be presented for optimal outcome. But it's going to happen if the presenting person feels that they almost are within friends, that this will not be used against them and no amount of system engineering, which is sort of what we're talking about, is going to do away with the need for trust between members of the treating team. But I do think we have the opportunity and I think there is the goodwill that something along these lines is definitely possible and would be to the benefit of all our patients.
Top of page
(cuts out) that are happening here, along those lines that you can quote?
Prof. Bruce Mann:
No, not really.
So what's it going to take?
Prof. Bruce Mann:
The model that works quite well here amongst surgery is there is the one on one personal relationships where there is trust, where there are a lot of phone calls. "I'll just run a case by you, what should I do?" That's the informal tumour board relying on the knowledge of the individual to have a good understanding of the disease and I think…
(inaudible) us phoning up somebody who knows and you get some personal tutorial down the line.
Prof. Bruce Mann:
Yeah. Exactly. So I'd probably ask David Currow, the CanNET - I'm not aware of the fine details but reading some of the reports, I think it is a promising model.
Although the model varies according to where CanNET was set up and what was trialled.
Question (Mike Poulsen):
Yeah. Mike Poulsen in Brisbane. We've had some experience with running a sort of virtual head and neck clinic using a sort of hub and spoke model where centres in Cairns and Townsville will phone in with a video conference to our head and neck clinic at Princess Alexandria Hospital in Brisbane on a weekly basis. They'll submit CTs electronically and they're reviewed conjointly, so some limited experience. It is challenging and it's not as perfect as a face to face consultation but it saves an awful lot of travel.
Question (unidentified male):
Professor, there is actually quite a bit of work going on with some of those virtual models - sorry, I've got a sore throat - including through the CanNET Victoria project, there's a number of models being implemented. And I think there's one at the Royal Melbourne Hospital on a Thursday morning where we have a virtual MBT for the haematology patients and, in fact, I think your surgical meeting holds us up often starting. So all those people waiting outside the door is the virtual MBT.
Roger Alison, Brisbane. I mean, you partly alluded to this but we were talking about this this morning. One of the benefits of sub-specialisation is not just in the treatment team but the pathology and radiology services which can markedly change and this - obviously Davis recognised this - and they've got very good technology. But we should never underestimate the vital role that these other specialties play in our treatment determination. I think it's very important.
Prof. Bruce Mann:
I couldn't agree more and which discipline is needed does vary a little according to the disease.
Question (Bob Thomas):
Yes, I'm Bob Thomas from Melbourne and I've enjoyed both Cliff and Bruce's talks. I've got a question for both of them. I think Cliff talked about, if you like, hospital systems and Bruce talked about, if you like, referral or professional issues and systems and I really enjoyed Cliff's talk about bricks and mortar. I'm wondering whether it's possible to paint on the walls of the regional cancer centres, as they are being built, the systems that you've been talking about or do we have to wait for a period of time of this gradual aggregation of these sort of systems and programs which do come in place over a period of time? Or can we put them in place right from the start?
It's a good question.
Prof. Bruce Mann:
I think the systems can be set up - each system needs a champion to run it and if the champion is there and the teams are there, we don't have to wait for the building.
I think that's right. I think the other issue is to work out human behaviour here. The reason the QSA has worked is because it's the clinicians who essentially are asking the questions and they want the answer because it's going to impact on their practice. I think if you're building in bricks and mortar, you've got to build in people in exactly the same way - good foundation, keep the world true but you've got to engage people locally. They'll do the work for you. They want to do it. They just don't have the tools.
Does anyone in the room feel - just a straw poll - how many people are here from regional cancer centres? And most people here would feel linked to a regional cancer centre and a network, I would assume. Is that a reasonable assumption? So how many people would feel that either within their network, their tumour group or whatever or their centre, they've got explicit rules understanding systems in place for knowing when somebody should not be operated on or treated locally and what elements of their care should actually happen elsewhere?
How many feel that you've got that in place across a reasonable range of tumour sites? That's a resounding affirmation. So clearly a major challenge is how to put that in place. How many of you - can you just put up your hand if a barrier is in fact getting the multidisciplinary team together on a regular basis? Is that the barrier? Maybe. Sorry.
Did you want to follow up your question - and I think that's a part of the agenda for tomorrow. We've got this panel discussion tomorrow. I think - thanks, Cliff - if that's not happening, then in fact we have got unacceptable variation in cancer care across Australia despite this $560 million going into this care, not to mention the recurrent. It has to be happening. How do we actually minimise the variation? This is a huge, huge issue and some of those barriers that Cliff was talking about to getting multidisciplinary teams together, this is big stuff.
I'm sorry, I've…
Question (Michael Penniment):
Michael Penniment, I'm from Adelaide and Darwin and just on that point, we started our telemedicine, telehealth at Royal Adelaide in 1996 and we've evolved it from a sort of friendly relationship to the virtual tumour boards which have been instituting across about four or five regions in South Australia for about two years. We've also included Darwin and Broken Hill and other regional centres.
We use the virtual telehealth for a number of things, from QA meetings to direct patient correspondence in terms of Broken Hill and other centres and we also have the virtual tumour board meetings in a number of sub-specialities. That's been particularly important since we brought the Darwin centre on board. We have institutional agreements between the Darwin centre and the Royal Adelaide to make sure that's evolved beyond just being a friendly relationship, that there is actually institutional linkage.
But the things that you said were true in terms of in those initial years where it was really just a couple of mates getting together across a couple of sites. That did rely very much on those people being in those positions at a particular time, to the point where now everybody that works in the various centres knows that's part of their core role. Thanks.
Thanks for that. It's very useful to know what's happening. Rosemary?
Question (Rosemary Knight):
Rosemary Knight, Principal Adviser in Cancer for the Department of Health and Ageing. Can I just say that perhaps one of the reasons why we didn't get a huge show of hands is having been involved with regional cancer centres in developing the principles and assessing them and setting them up, I'm really consciously aware that many of them aren't even at the bricks and mortar stage yet. So what we've got is, we've got a few regional cancer centres that we've funded nationally which are leveraging or building off some of the existing centres but most of them are still at the architectural stage and getting through, and so there's quite a variability in stages of development.
So let me ask you this, Rosemary, what have you got in place of the Commonwealth that meets that challenge that Bruce has just set?
Question (Rosemary Knight):
We've got principles which everybody had to address and selection criteria in which the regional cancer centres were assessed against and one of them is about integrated network to cancer services with appropriate referrals and triages and it's deliberately being constructed in such a way to set up, so that it works either with the hub and spoke type model or networks of cancer services that ensure safe and effective high quality care. One of the deliverables.
When the rubber hits the road and you've got the odd surgeon who's doing an operation he or she shouldn't, that's when it gets tough but there's no principles to resolve that.
Question (Rosemary Knight):
That's true. We've got two other mechanisms in place, one of which is that this is a shared program with the states and territories and particularly around the recurrent funding and the workforce. And the other thing we've got going at the moment is some work that's been initiated from COAG where we've started to look at some of the key features that might be important in terms of sustainable systems.
Thanks. Mike, do you want to come back for a second bite.
Question (Michael Penniment):
Yeah, if I could. I think we've got to be careful about mandating with regional centres what they can and can't do. There will be some things that they should never do, like paediatric malignancy, taking on (unclear) myoblastoma. But there are other sites where the regional centre will have the expertise, they will have a head and neck trained surgeon, they will have people who are regularly doing head and neck malignancy. I don't see a problem with them doing head and neck radiotherapy.
There'll be other centres where there is no head and neck service, no allied health service and clearly it's inappropriate. So I think what we have to do with the regional centres is if they meet certain standards, then they have to demonstrate that it's safe to deliver that. But I'd be very wary about sort of saying, look, thou shalt not do this, thou shalt not do that and I'd make it across the board.
Yeah, because it can be an accident of fate. For example, with the Royal Perth has a very low volume Hepatobiliary Unit with very good results. So it's very variable.
Bruce, do you want to make any final comments or are you…
Prof. Bruce Mann:
No, I think it's a good discussion and I think it is a discussion that various groups are going to have to get together and one of the big groups who do need to be engaged is the College of Surgeons and, I guess, I'll be asked to help with that.
Speaking of craft groups and affiliations as per our earlier presentation. Bruce, thank you very much.
This publication is available as a downloadable document.