Personal Perspective

Professional misconduct: the Bristol case

Stephen N Solsin

In June 1998, the Professional Conduct Committee of (appointed in 1986); and Dr John Roylance, a former
the General Medical Council of the United Kingdom (the radiologist, and Chief Executive of the Trust from its
body which regulates British doctors) concluded the creation in 1991 until his retirement in 1995.
longest-running case it has considered this century.'
Three medical practitioners were accused of serious The central allegations were that the Chief Executive
professional misconduct relating to 29 deaths (and four and the Medical Director of the Trust allowed to be
survivors with brain damage) in 53 paediatric cardiac carried out, and the two paediatric cardiac surgeons
operations undertaken at the Bristol Royal Infirmary
carried out, operations on children knowing that the
between 1988 and 1995. All three denied the charges
mortality rates for these operations, in the hands of
but, after 65 days of evidence over eight months (cost-
these surgeons, were high.3 Furthermore, the surgeons
ing £2.2 million), all were found guilty.2
were accused of not communicating to the parents the
The doctors concerned are Mr James Wisheart, a correct risk of death for these operations in their henas»
paediatric and adult cardiac surgeon (appointed in
1976, now retired), and the former Medical Director of Stephen Bolsin, a cardiac anaesthetist, "blew the
the United Bristol Healthcare Trust (the hospital group whistle" and then had the courage to follow through until
that includes the Bristol Royal Infirmary); Mr Janardan a full investigation was carried out. The process took
Dhasmana, paediatric and adult cardiac surgeon over six years. Here he tells his story.

join ed the cardiac anaesthetic team at the Bristol Royal realise that the long surgery times (up to three times as long

I Infirmary on 1 September 1988, undertaking a mixture

of adult and paediatric cardiac anaesthesia. The Bristol
Royal Infirmary is a 400-bed tertiary referral hospital serv-
as normal for Mr Wisheart) and the long aortic artery cross-
clamp times (during which the heart is deprived of oxygen)
were leading to heart failure postoperatively as well as a range
ing a population of about 7 million people in southwest Eng- of other problems.
land and South Wales, with a supraregional paediatric I had always kept a record of all cases I had undertaken,
cardiac surgery service. I had come from a Senior Registrar and now began to add details of deaths and survivors. The
post at the Brompton Hospital in London, a national refer- figures were a more objective confirmation that there was a
ral centre for adult and paediatric cardiothoracic surgery (as serious problem with mortality in the unit. Audit meetings
well as cardiology and respiratory medicine) and a post- were held on an irregular basis, and on one occasion in 1990
graduate teaching hospital. I tried to document the concerns of some of the anaesthetists
by taking minutes of the meeting. I was told when I pro-
My first impression was that operations in Bristol were
duced the minutes that they were not needed and that no
taking a lot longer than at the Brompton Hospital, but I ini-
minutes of the meetings would be taken in future. There
tially put this down to increased difficulty of the cases. How-
appeared to be considerable apathy among my surgical, car-
ever, I did notice that the patients, both adults and children,
diological and even some anaesthetic colleagues towards the
seemed to suffer from more cardiac, renal and respiratory continuing high death rates, and my expressions of concern
complications than I had been used to. Slowly I began to were not seen as helpful or constructive.

For editorial comment, see pages 351 and 352 Approach to the District General Manager
Department of Anaesthesia, Perioperative Medicine and
Pain Management, The Geelong Hospital, Box 281, Geelong, In 1990, the Bristol Royal Infirmary became part of the
VIC 3220 United Bristol Healthcare Trust, one of the National Health
Stephen N Bolsin, MB BS, BSc, FRCA, Director of Anaesthesia and Service's first (and largest) Hospital Trusts. I wrote to the
Chief of Critical Care Services. then District General Manager (Dr Roylance, who subse-
Reprints will not be available from the author. Correspondence:
Dr S N Bolsin, The Geelong Hospital, Box 281, Geelong, Victoria, 3220. quently became Chief Executive of the Trust), documenting
E-mail: [email protected] my concerns about the unit and the excessive paediatric car-

MJA Vol 169 5 October 1998 369

Personal Perspective

diac surgical mortality. I described the unit as having "one

1: Crude mortality rates for paediatric cardiac
of the highest mortality rates in the country for open heart operations, 1990-1992"
surgery on children under one year of age". The Director of
Anaesthesia subsequently agreed to speak to the cardiac Bristol UK average
Surgeons
surgeons about the problem ... but surgery continued. (1/90-7/92) (1990) Wisheart Dhasmana
By 1992 I had decided that I no longer wished to work in
a unit which appeared unable to address the high mortality Tetralogy of Fallot

II
rates associated with some procedures, so I applied for a post Age <1 year 2 (3) 7 (79)
in Oxford. I asked the Professor of Anaesthetics in Bristol Age >1 year 6 (26) 15 (245)
(also President-elect of the Royal College of Anaesthetists) All ages 8 (29) 22 (324) 4 (12)t 4 (16)f
for a reference - when he found out why I wanted to leave 28%* 7%* 33% 25%
Bristol he recommended that I collect the necessary hard Atrioventricular canal surgery
evidence on the mortality rates so that he could attempt to I Age <1 year 4 (13) 12 (112)
deal with the matter locally to achieve constructive change. Age> 1 year 1 (5) 8 (81)
All ages 5 (18) 20 (193) 6 (10)§ 2 (12)§
28% 10% 60% 17%
Data collection
*Number of deaths given, with number of operations performed in
In the summer of 1992, after failing to be appointed in parentheses.
Oxford, I started to collect and analyse data on paediatric tOne survivor not attributable to a specific surgeon.
tP< 0.001.
cardiac surgical mortality with Dr Andy Black, a Senior Lec- §Data for individual surgeons extends to 12/92, thus totals exceed those
turer in Anaesthetics in Bristol. Dr Black and I had already given in first (Bristol) column.
been collecting and analysing data on adult cardiac surgery
to produce risk-adjusted outcomes for all cardiac surgical
centres in the United Kingdom. Crude outcome data had not collected data on the "arterial switch" procedure (to cor-
been collected since 1977 through the UK Cardiac Surgery rect transposition of the great arteries), although we knew
Register but did not allow comparison of "like with like". that this was carried out particularly badly in Bristol, because
The Department of Health had provided funding of over at the time there were no national comparison data for this
£300000 to further this work, which was carried out under operation in the Cardiac Surgery Register (the register had
the auspices of the Association of Cardiothoracic Anaes- data on surgery for transposition in general, but did not sep-
thetists of Great Britain and Ireland. I had been appointed arate out the different procedures).
their first National Audit Coordinator in 1991.
In 1993 I changed my working pattern to avoid anaes-
Approach to the Chief Executive and
thetising children for cardiac surgery but continued with my
Chairman of the Board
adult practice. I still covered paediatric cardiac surgery at
night, but avoided the heart-breaking task of elective surgery When the data were complete, we presented it to the Pro-
for these pitiful infants. I collaborated with the academic unit fessor of Anaesthesia, who contacted the Chief Executive of
on numerous projects and publications, including the devel- the Hospital Trust (Dr Roylance) and the Director of Anaes-
opment of outcome studies for adult cardiac surgery. thesia. We also presented the data to the newly appointed
Professor of Cardiac Surgery - he was horrified by the
Disturbing evidence results and agreed to take the matter up with the surgeons
directly. He also approached Dr Roylance, as well as the lay
By the spring of 1993 the results of Dr Black's and my audit Chairman of the Trust Board. However, the net result of dis-
were available and they were disturbing (Box 1). The over- seminating these results was minimal - the high-risk oper-
all mortality rates for atrioventricular (AV) canal operations ations by both surgeons continued and the death toll
(for atrial septal defect with ventricular septal defect and mounted.
mitral and tricuspid valve abnormalities) in Bristol were sta-
tistically significantly worse than the national average, and Continuing high-risk operations
the same applied to operations for tetralogy of Fallot (ven-
tricular septal defect, pulmonary stenosis, dextroposition of During this time, Mr Dhasmana began performing arterial
the aorta and right ventricular hypertophy). For Senning's switch procedures (Box 2) on neonates «30 days old), with
procedure (an intra-atrial repair for transposition of the great disastrous consequences. He stopped these in 1994, but not
vessels) and Fontan's procedure (for tricuspid atresia) we before nine (of 13) infants had died" and at least one had
found no difference between the national average and Bris- sustained permanent brain damage. For comparison, a sur-
tol. Furthermore, it appeared that one of the surgeons had geon in a nearby unit had had only one death in 200 cases
a higher mortality than the other. We had used national aver- and had no patients with residual morbidity (Dr W Brawn,
age data from 1990 to compare with Bristol data from Cardiac Surgeon, Birmingham Children's Hospital, personal
1990-92 in order to avoid criticisms of unfair comparisons communication). A subsequently published series of 104
(mortality rates were continually falling over time as neonatal switch operations between 1987 and 1993 reported
expertise and technology improved). We had also deliberately only one death in the first 52 patients, a cluster of deaths

370 MJA Vol 169 5 October 1998

 Personal Perspective

(in June 1998) quoted a report from the Department of

2: Transposition of the great arteries, and the arterial
switch procedure
Health and Social Security, dated 1988, claiming to highlight
the problems of poor results in the paediatric cardiac surgery
In this congenital condition, the aorta arises from the unit in Bristol. It would seem that the Department of Health
right ventricle and the pulmonary arteries from the left and the Bristol Royal Infirmary had known about the prob-
ventricle, resulting in two separate and parallel lem before I even started to work there.
circulations. Without any communication between the In June 1994, I and five of my fellow anaesthetists
two circulations, affected babies would die at birth; thus,
requested a full and open review of the results of the arter-
almost all have at least some form of interatrial
communication.
Without treatment: 30% die in the first week of life,
50% die in the first month, and 90% in the first year.
II ial switch operations. This review did not take place for
another seven months and two more children died in the
meantime (in July 1994 and January 1995). The Trust had
by now appointed a new paediatric cardiac surgeon (from
Treatment: The arterial switch procedure, which I the Royal Children's Hospital in Melbourne) who was sched-
reconnects the transposed aorta and pulmonary arteries Ii uled to start operating in May 1995, and the Professor of
into their correct anatomical positions, and relocates the m Cardiac Surgery believed he had made an agreement with
coronary arteries to the "new" aorta, became the I the incumbent surgeons that no further switch operations
standard of care in the 1980s. It replaced the palliative
Senning and Mustard procedures, which correct the would be done in the hospital.
physiology by redirecting blood flow in the atria, but
work only for as long as the right ventricle can withstand The last fatal operation
the systemic load.
Prognosis: In 1996-97 in the UK, the operative However, in late December 1994, the Professor and I dis-
mortality rate was 6.5%.5 Prognosis varies with the exact covered, from the routinely circulated operating list, that Mr
nature of the defect, but survival at 10 years after Dhasmana intended to undertake one more switch opera-
successful correction of the defect approaches 100%. tion, in January 1995. The child was 18 months old and had
previously undergone palliative surgery. We immediately
notified all the clinical and non-clinical managers within the
(followed by the surgeon retraining), and nine deaths over- hospital who we thought would be able to influence Mr
all." Dhasmana's results for switch operations in older chil- Dhasmana to cancel the operation, but no one was prepared
dren were also poor, with a mortality rate of around 30% (in to help. Consequently we both, independently, applied to the
the best hands, mortality for this was as low as 1%). Over- Department of Health to try to ensure that common sense
all, his mortality rate for the switch procedure was thus about prevailed. We now know that the Chief Executive of the
50%. Trust (Dr Roylance) was contacted by the Department and
Mr Wisheart, the senior paediatric cardiac surgeon, had was requested to have the operation postponed or the child
particularly worrying results for AV canal operations - by transferred to another unit. Dr Roylance was not prepared
the time he retired from paediatric cardiac surgery, seven of to take that action.
the last eight children undergoing this operation had died." The night before the operation was due to take place, a
The national average mortality rate was 10%. meeting of paediatric cardiologists, cardiac surgeons and
anaesthetists was hastily convened to review all arterial
switch operations in the hospital (over the last seven
Conflict of interest
years).The results showed that the arterial switch operation
Despite all this very clear evidence, and the surgeons' own had a mortality rate of 69% (9 of 13) in neonates, and 36%
records of the complications and outcomes of all their oper- (10 of 28) in children over 30 days (slightly different from
ations, the unit's performance was not reviewed. One of the the estimates that I had produced and previously circulated
reasons for this was Mr Wisheart's increasing positions of of 67% and 40%, respectively). The data were presented in
responsibility within the Trust, eventually becoming Med- a selective way that used subgroups with small numbers,
ical Director (responsible for medical standards of all prac- information from different (non-comparable) years, and
titioners within the Trust). This meant he was the medical "comparative data" that included non-switch operations, and
adviser to the Chief Executive of the Trust when the prob- thus the excess mortality in the Bristol group was dismissed
lems with paediatric cardiac surgery (which involved him) as not statistically significant. I argued that there was more
were raised. than enough information in the data presented to indicate
that there was an institutional problem with the switch oper-
ation at the Bristol Royal Infirmary and that we should thus
Approach to the Department of Health
postpone the operation until the new surgeon arrived. Any
In December 1993, I contacted the Department of Health urgent switches could be transferred to Birmingham Chil-
with my concerns, to find that another Professor of dren's Hospital, or Great Ormond Street Children's Hos-
Surgery from the Bristol Royal Infirmary had already pital, in London. However, the meeting agreed that the
approached the Department with the same concerns. operation should proceed and I was in a minority of one (the
Intriguingly, a television program screened several years later Professor of Cardiac Surgery had not been invited).

MJA Vol 169 5 October 1998 371

Personal Perspective

The next day the operation went ahead and the child died The General Medical Council inquiry
on the table.
In April 1996, I wrote to the GMC in the UK, requesting
My wife, Maggie, and I were mortified. We had considered
an inquiry into the events in Bristol and suggesting that the
sending Maggie to the children's ward to tell the parents of
actions of some of the doctors amounted to serious profes-
the problems and advise them to get a second opinion in
sional misconduct. I believe that I am the only doctor to have
Birmingham. We now felt guilty that we had not done this,
taken such action. The Professional Conduct Committee
even though we realised such an act would have exposed us
(consisting of Dr Donald Irvine [President of the GMC],
to the possible allegation of serious professional misconduct
two anaesthetists, a surgeon, a haematologist, two members
ourselves.
of the public and a legal assessor) started that inquiry in
1997, concluding it in June this year by finding all three doc-
My employment threatened tors guilty of serious professional misconduct. Dr Roylance
The disclosure, in April 1995, of the problems in Bristol on and Mr Wisheart have been struck off the Medical Register,
the front page of the Daily Telegraph, and also in the British and Mr Dhasmana has been disqualified from practising pae-
Medical Journal, 6 led to my being asked if I would reduce my diatric cardiac surgery for three years, although he can con-
cardiac anaesthetic commitment to one day per week in tinue operating on adults. An action group of over 90
order to "reduce contact with the surgeons". I agreed, on a families pressed for a full independent public inquiry into the
temporary basis, but when I requested to resume my normal events in Bristol - the government has now announced that
cardiac allocation I received threats to my continuing this will take place, that it will be chaired by Ian Kennedy,
employment from the Director of Anaesthesia and the Chief Professor of Medical Law and Ethics at University College,
Executive of the Trust." London, will cover operations performed between 1984 and
1995, and will look more widely at the involvement of the
National Health Service and the Department of Health. 10
First independent inquiry The local police are also now examining the evidence with
In January 1995, the Trust was advised by the Department a view to criminal charges. Compensation claims are likely
of Health that there should be an inquiry into paediatric car- to run into tens of millions of pounds.
diac surgery in Bristol and that this should be undertaken by
credible outside experts. The Trust approached a Paediatric Would I do it again?
Cardiac Surgeon from London and a Paediatric Cardiolo-
gist from Newcastle to conduct the inquiry: their original My role was to stand up for the best interests of the patient.
report identified Mr Wisheart as a "higher risk surgeon" and For that, I suffered at the hands of a profession that was not
recommended that he carry out no further paediatric cardiac prepared to act against senior colleagues in positions of
operations. However, the Chief Executive referred to this power. My career has been altered irreconcilably and I have
report as "a draft working document" and immediately asked had to move to another country. My family have been sim-
for alterations to be made. A second report failed to men- ilarly affected. Despite this, I would, if necessary, do the
tion a higher-risk surgeon and claimed that concern by same thing again. The challenge for the medical profession
anaesthetists had raised anxiety in the surgeons and may globally is to make sure that such stands on patient safety
have contributed to the high mortality. This report was never have to be taken again.
released to the local press.
Mr Wisheart continued to operate on children for closed
cardiac conditions. On 1 May 1995 he undertook his last References
paediatric cardiac procedure - a reoperation to repair coarc-
1. Horton R. Doctors, the General Medical Council, and Bristol. Lancet 1998; 351:
tation of the aorta. This child died several weeks later. 1525-1526.
Also, on 1 May 1995, Mr Ashe Pawade commenced work 2. Dyer C. Bristol doctors found guilty of serious professional misconduct. BMJ 1998;
316: 1924.
as Paediatric Cardiac Surgeon for the United Bristol Health-
3. Dyer C. Two surgeons are accused of ignoring high death rates. BMJ 1997; 315: 967.
care Trust. Since then, mortality rates have fallen dramati- 4. Morris K. UK misconduct case raises informed consent issues. Lancet 1998; 351:
cally and the unit is now one of the best in the UK. 9 885.
5. Treasure T. Lessons from the Bristol case. BMJ 1998; 316: 1685-1686.
6. Dyer O. Hospital banned from doing open heart operation. BMJ 1995; 310; 960.
Move to Australia 7. de Leval MR, Francois K, Bull C, et al. Analysis of a cluster of surgical failures. Appli-
cation to a series of neonatal arterial switch operations. J Thorac Cardiovasc Surg
In February 1996, I moved to The Geelong Hospital, Vic- 1994; 107: 914-923.
toria, to take up the post of Director of Anaesthesia. I have 8. General Medical Council. Proceedings of the disciplinary hearing. London: GMC,
1997.
since been involved in setting up an extremely successful car-
9. Willie P, Hughes D. Cardiac services in Bristol are now of high quality. BMJ 1998;
diac surgery unit, with excellent results and very low com- 316: 1986.
plication rates. 10. Warden J. Cardiac surgery inquiry given wide remit. BMJ 1998; 317: 489. 0

372 MJA Vol 169 5 October 1998