BRITISH MEDICAL JOURNAL
26 APRIL 1975
Conclusions It is understandable that more debate has been given to whether or not psychotropic drugs should be prescribed than to which psychotropic drugs should be prescribed. I believe that both debates would be assisted by a smaller range of preparations being available for prescription. Patients would receive drugs more familiar to the doctor, which must reduce side effects and inappropriate medication. Would there be an equal disadvantage in depriving patients of
179
the existing selection of drugs ? The question deserves critical examination.
References 2
4
5
Shepherd, M., et al., Psychiatric Illness in General Practice. London, Oxford University Press, 1966. Wells, F. O., Journal of the Royal College of General Practitioners, 1973, 23, 164. Durno, D., "Activity Analysis in General Practice," M.D. Thesis, Aberdeen, 1972. Durno, D., Personal communication, 1974. Petrie, J. C., Howie, J. G. R., and Durno, D., British Medical Journal, 1974, 2, 262.
Clinical Review Patterns of Incidence in Acute Pancreatitis JOHN E. TRAPNELL, E. H. L. DUNCAN British Medical journal, 1975, 2, 179-183
Summary A review of acute pancreatitis occurring over a 20-year period in the Bristol clinical area is reported. A total of 590 cases were available for analysis. The yearly incidence was 53 8 per million population at risk, with a mortality of 9 0 per million. This compares favourably with 11 4 deaths per million for England and Wales as a whole during the same period but the difference is not statistically significant. When the deaths occurring in the Bristol clinical area were expressed in terms of case mortality rate the figure was 17%. In contrast the mortality for recurrent acute pancreatitis was only 1-5%, and the benign nature of this second condition is confirmed. Aetiological factors and age and sex distribution were also analysed in relation to each other and to mortality. An increase in acute pancreatitis secondary to chronic alcoholism was confirmed and steroid pancreatitis also emerged as a definite entity in this survey. The pattern of recurrence in patients with idiopathic pancreatitis was studied in detail and is analysed on an actuarial basis. Introduction The full pattern of the incidence of acute pancreatitis is hard to determine. Though large series of cases have been reported from many centres in different parts of the world most have been concerned with the disease as it occurs in selected segments of the population. Reports from special clinics in the United States have been on patients referred from a wide area, and the city hospital patients may have related to some large urban conurbation, but these centres serve only low-income groups. Royal Victoria Hospital, Boscombe, Bournemouth JOHN E. TRAPNELL, M.D., F.R.C.S., Consultant Surgeon Department of Community Health, University of Bristol E. H. L. DUNCAN, M.A., B.SC., Senior Lecturer in Medical Statistics
Furthermore, the Veterans Hospitals, the other source of large series of cases in the United States, also represent a highly selected section of the population. In the same way, incidence and mortality statistics from other parts of the world have always been derived from hospital case series and it is never clear how closely they relate to the disease as it occurs in the general population. There has been only one report of a population survey.' This was undertaken in Rochester, Minnesota, and though some of the details, particularly of the follow-up and checking procedures, were scanty it provided the first geographical yearly incidence for all forms of the disease ranging from 150 to 180 cases per million population. For acute pancreatitis alone (but including recurrent acute pancreatitis) the yearly incidence ranged from 100 to 115 cases per million. We present an analysis of all cases of acute pancreatitis occurring in the Bristol clinical area during the 20-year period 1950-69. The patients were drawn from the large urban area of the city of Bristol and its suburbs and from the rural area of south Gloucestershire and north Somerset.
Clinical Material In collecting the cases a most careful search was made in all the hospitals in the Bristol area. Details of the few cases treated in private hospitals were also kindly supplied by the consultants concerned, and these figures were then checked, firstly, by reference to the regional hospital board's hospital activity analysis system and then by examination of the cases certified as dying from the disease during the period. In this connexion detailed reports of the coroners' necropsies were obtained. Interestingly, the regional hospital board's statistics and the coroners' data consistently produced one or two cases of acute pancreatitis each year not identified from the other sources. The diagnostic criteria were either a consistent clinical picture and a serum amylase level over 1000 Somogyi units or acute pancreatitis plainly evident at laparotomy or necropsy. On the basis of these criteria and from the search outlined 590 primary attacks were identified over the 20-year period. There was no significant referral rate of patients with pancreatitis from outside the area and thus the series seemed to be fully representative of the disease pattern in the area at that time. Because the hospital activities analysis data and coroners'
180
BRITISH MEDICAL JOURNAL
records were available only to the end of 1967, for the purpose, of detailed statistical analysis this has been regarded as the las "complete" year. During 1950-67 there were 551 cases, and al1 of these patients who survived their first attack were followec up in the second half of 1968 or in early 1969. Thus follow-up of these patients ranged from one to 18 years. II
I
Overall Incidence and Mortality During the first six of the 18 "complete" years under review (1950-5) there was an increase in the number of cases diagnosed, and another peak occurred during 1961-4 (table I). Over the whole period, however, the mortality rate did not alter significantly, and from a critical review of local factors it appeared that the initial increase was due to a wider application of the serum amylase test, which was part of a national trend at the time. The peak in the early 1960s was due to a period of heightened interest in the condition locally. Thus both increases seem to have been due to improved diagnosis rather than to an absolute change in incidence. With the relative increase in incidence of the disease over the years there was a relative decrease ir mortality (tables Ia, Ib). Critical consideration of the method used to express mortalitythat is, as a percentage of the number of cases diagnosedsuggests that it is apt to give a false impression, for there was, in fact, no significant absolute change in the death rate in the total population over the years in question. TABLE Ia-Incidence of Acute Pancreatitis in Bristol Clinical Area, 1950-67 Year 1950 1951 1952 1953 1954 1955 1956 1957 1958
lNo. 8
Year
No.
2
1959 1960 1961 1962 1963 1964 1965 1966 1967
32 28 35 54 36 47 37 23 30
4 8 6 3
20 20
29 30 36 23
pancreatitis. Eight of these were lost to follow-up. Of the remaining 430 patients 133 had suffered a total of 261 recurrent TABLE Ii-Incidence of Cases and Death Rates per Million Population from Acute Pancreatitis in Bristol Clinical Area during 1961-7 Year
Cases
Population
11
5
31
4 5
32
TABLE ib-Summary of Incidence in Bristol Clinical Area Years
No.
Died
1950-5 1956-61 1962-7
143 181 227
34 42 37
23-8 23-2 16-3
Total
551
113
20-5
Of the 551 patients in the survey 113 died of their first attack, giving a primary case mortality rate of 20-5%. As noted above, the expression of mortality as a percentage of diagnosed cases does not necessarily give a reliable figure. To achieve this deaths must be related to the total population at risk. The Registrar General's Statistical Review of England and Wales2 states that during 1961-7 there were 3790 deaths from acute pancreatitis (I.C.D. 577-0) and that there were 331 669 900 person-years at risk, giving a mean yearly death rate of 11 4 per million population. Analysis of the 1961 Census figures and the Registrar General's estimate for the years between censuses shows that there were 4 867 985 person-years at risk in the Bristol clinical area, and with 44 deaths during that period the death rate was 9 0 per million (table II). Though the death rate for the survey area was lower than the national average the difference is not statistically significant.
Recurrent Acute Pancreatitis Altogether 438 patients survived their initial attack of
acute
No.
Rate per Million
1961 1962 1963 1964 1965 1966 1967
676 770 683-035 689-440 695-810 702-605 707-710 712-615
35 54 36 47 37 23 30
51-7 79 1 52-2 67-5 52-7 32-5 42-1
Total
4 867 985 Person-years
262
53-8
Deaths I. No. Rate per Million 7 10-3 8 11*7 6 8-7 7 10-1 4 5-7 5 7-1 7 9-8 44 90
TABLE III-Results of Follow-up of 430 Patients Surviving First Attack of Acute Pancreatitis Aetiological Group Gall-stone .. Idiopathic .. .. Alcoholic .. Other Total
No. of Patients Available for
No. Having
Recurrence
Total No. of Recurrent Attacks
No. of Deaths
232 148 18 32
88 35 10
182 61 18
2 2
430
133
261
4
Follow-up
acute attacks but only four had died (table III). When these deaths are expressed as a percentage of the total number of recurrent episodes (1-5%) it is clear that the outlook for this group is vastly different from that for patients in their first attack. Thus recurrent acute pancreatitis is by comparison a relatively benign condition with a much improved prognosis. The aetiological factors in the 430 cases are set out in table III. Of the 88 patients with gall stones having a recurrent attack at least 36 were on a waiting list for cholecystectomy. Only a few had had their biliary tract disease long enough for a sequence of recurrences to develop. It was, however, a feature of the pattern of recurrence in this group that while most attacks recurred within a year or two of the previous episode there were on occasions long gaps-10 to 15 years or over in several instances -and during these intervals the patients were entirely symptomfree. These follow-up studies therefore confirm that if biliary tract disease is not treated adequately there is an ever-present danger of recurrent acute pancreatitis. One of the two patients with gall stones who died was a man of 49 with long standing diabetes. After an initial episode in 1950 he succumbed to his first recurrent attack four years later. The other death was also in a man, in this case aged 41. Again it was his first recurrence, this time after an interval of only eight months. One-quarter of the patients in the idiopathic group suffered a recurrent attack, but except for one of the two who died only two patients had more than two recurrent acute episodesnamely, four attacks each. It was again a feature of this group that these further attacks were mild and uncomplicated. Calculated by the actuarial method3 10% of patients with idiopathic pancreatitis are likely to have a first recurrence within one year of the primary attack, 17% within two years, and 25% within six years. No recurrence occurred after a gap of more than six years, though 75 patients were followed up after this period for a total of 444 person-years. Patients with idiopathic pancreatitis who have had one recurrence are likely to have a second. We found, again using the actuarial method, that 12% have a second recurrence within a year, 16% within two years, 31% within three years, and 56% within six years. One of the patients with idiopathic pancreatitis who died of a recurrence was a woman of 62; she succumbed to her third attack in five years. The other patient who died was unique in
I~~~~~~~ Died
26 APRIL 1975
BRITISH MEDICAL JOURNAL
181
26 APRIL 1975
this series. She suffered her first attack of acute pancreatitis in 1952 when she was aged 12. Over the next eight years she had 10 further, separate episodes of varying severity but remained well between attacks, with no evidence of exocrine or endocrine insufficiency and no sign of pancreatic calcification. Partial pancreatectomy with retrograde drainage into a Roux loop was performed after the sixth recurrence but in spite of this the attacks persisted. Finally, she was readmitted in September 1960 with the usual acute presentation. Instead of responding to intravenous fluids and standard measures, however, her condition deteriorated unexpectedly and rapidly. She became shocked and cyanosed and died within 72 hours. Necropsy showed widespread fat necrosis and oedema but no haemorrhage or necrosis of the pancreas. The gland itself was fibrotic, the main duct was not dilated, and the Roux loop had separated from the distal end. The alcoholic group is, perhaps, more difficult to assess, for it is characteristic of these patients that they progress to the chronic relapsing rather than the recurrent acute form of the disease. Indeed, several of the 18 cases followed up pursued such a course, but 10 men all suffered one or more further episode of acute pain with vomiting associated with a marked increase in their serum amylase level at a time when they were otherwise well and entirely symptom-free. They must therefore be included in this survey as examples of recurrent acute episodes. There were no deaths from any of these attacks or, indeed, from the chronic relapsing disease when it supervened in this group of cases.
Aetiology The aetiologically related factors for the whole series of 590 cases are set out in table IV. Analysis of the two major groups, gall-stone and idiopathic pancreatitis, shows that there was no significant change in the patterns of incidence during the 20 years of this survey. In particular, it is disappointing that the incidence in the gall-stone group did not fall. It is now common surgical practice in Great Britain to remove all gall stones more or less "on sight" even when they are symptomless, and it was hoped that one of the benefits might have been a decrease in the prevalence of gall-stone pancreatitis. A partial explanation for a lack of any change in incidence may be that just over a quarter of these patients (81 out of 316) had no past history of any abdominal symptoms before the onset of their attack, and hence there was no pointer to their underlying biliary tract disease. It is also of interest that 60 patients in the idiopathic group (30%) gave a history at their first admission of dyspepsia of varying severity and duration. Some of these patients had been extensively investigated with negative findings. At least 26 who were followed up not only remained recurrence-free but also reported that their previous dyspeptic symptoms had disappeared completely. From this it appears that when a patient presenting with an attack of acute pancreatitis gives a history of abdominal dyspepsia this should not necessarily be taken as a pointer to the presence of gall stones. Furthermore, patients with previous trouble are no more liable to have persisting symptoms or develop recurrence than those without ally past history. In the minority subgroups those patients with acute pant creatitis secondary to mumps and hyperparathyroidism do norequire any special mention, save to say that these are rare predisposing causes. All of the patients with carcinoma, seven of the head of the gland, and one of the ampulla of Vater, eventually died of their disease but, interestingly, apart from the one who died from acute pancreatitis all the survivors lived for a year or more, and one lived for three and a half years before dying of his malignancy. In all these cases there appeared to have been great difficulty in identifying the underlying disease, and in five cases the diagnosis was conclusively established only at necropsy.
TABLE Iv-Aetiological Factors in 590 Cases of Acute Pancreatitis Cases No.
Biliary tract disease Chronic alcoholism .26 Mumps .7 Hyperparathyroidism .1 Carcinoma .8 Steroids .6 Other. Not stated* .14 Idiopathic .203
0
316
53-6 4-4 1-2 0-2
9
1-5 2-4 34-4
1-4
1-0
590 100-0 *In these cases the presence or absence of biliary tract disease niined.
-. .
Total
Deaths No. % 19-9 63 3 11-5 0 9 0 0 1 12-5 4 66-7 11-1 1 2 14-3 45 22-2 20-2 119 had not been deter-
In the other two groups of cases, alcoholic and steroid pancreatitis, there was a significant change in the pattern of incidence over the 20-year period. It has been recognized for many years that the principal difference in aetiological spectrum between Great Britain and most other countries where acute pancreatitis is relatively common lies in the fact that chronic alcoholism is fairly uncommon as a cause in this country. In published series from the United States,4 Australia,5 6 South Africa,7 and Europe8 the alcoholic contingent constituted as many as 30 to 40% of the total number of cases seen. The present findings, however, suggest that the low-incidence pattern is changing (table V). This statistically significant trend confirms the impression presented in a recent review of chronic relapsing pancreatitis in London 9 and also the preliminary findings of a current study in Glasgow,' 0 where it appears that 20 to 25o% of cases now have an alcoholic background. It is therefore clear from this analysis and from supportive evidence that the figure of 4-4% derived from the present study, stretching as it does over 20 years, does not represent a correct picture of the current pattern of the disease. TABLE v-Five-yearly Incidence of Alcoholic Pancreatitis Period No. of cases
1950-4
1955-9
1960-4
1965-9
2
4
7
13
Total 26
Steroid pancreatitis emerged as a definite entity in the last few years. One of the six cases in this survey occurred in 1955, one in 1960, and the remaining four all after 1965. This will clearly remain an uncommon cause of pancreatitis but the numbers of cases may be expected to increase as more patients are put on long-term steroid therapy. This may therefore pose an additional problem for the transplant patient receiving steroid medication as part of his immunosuppressive regimen."' Finally, pregnancy is not included in the list of aetiological factors in this survey. Several workers in the 1950s suggested that pregnancy may predispose to acute pancreatitis,'2 13 but Howard4 questioned this assumption and suggested that as many of these women were subsequently shown to have biliary tract disease this was the aetiologically important factor rather than the pregnancy. Nevertheless, the association is still being reproduced in lists of aetiological factors,'0 14 and recently Corlett and Mishell'5 reported one episode of pancreatitis in 1066 births. These workers did not give their diagnostic criteria and the clinical details were sparse, so that it is difficult to judge the relevance of this statistic. In the present survey of 590 cases only 17 patients developed acute pancreatitis either during pregnancy (4 cases) or during a six-week post-partum period (13 cases). There was one death. At necropsy the patient was found to have gall stones, and of the survivors 14 also had biliary tract disease, which was subsequently eradicated surgically. Follow-up of these patients showed only one who developed further symptoms, and she, after sterilization, progressed in a typical fashion over 10 years to a chronic relapsing form of the disease. All the remaining patients were fit and well and symptom-free when interviewed.
182
BRITISH MEDICAL JOURNAL
Five women, including the two with idiopathic disease, had had one further child each. Four had had two further pregnancies and one three further children, all without incident. If we accept that the associated biliary tract disease was responsible for the pancreatitis in 15 of these 17 patients it seems likely that the remaining two cases could well have arisen by chance during pregnancy or the puerperium. During the 20year period there were over 220 000 live,births and stillbirths in the Bristol clinical area. If pregnancy does, in fact, predispose to acute pancreatitis one would have expected to see many more cases in a study of this size over this period of time. It must therefore be concluded that there is no evidence for any such association, and until a contrary view can be adequately supported there are no grounds for sterilizing women who have had an attack of acute pancreatitis.
Mortality and Aetiology The primary mortality rates for the two main groups of cases were comparable (table IV), and other, more detailed studies'16 have confirmed this. There were too few cases of alcoholic pancreatitis in this study for any conclusions about mortality patterns to be drawn. We must therefore await a report of a larger series of cases before it will be possible to see whether the death rate for the alcoholics, even in their first attack, coincides with that of the gall-stone and idiopathic groups. Even though the numbers are small it appears that steroid pancreatitis departs, from the overall spectrum for it clearly carries a bad prognosis with a high primary mortality rate.
26 APRIL 1975
and so does mortality. Analysis of the data shows that the four deaths in patients under 30 were all in the idiopathic group, and, in particular, the two very young cases were both postpartum deaths at one and five weeks. There was no evidence of birth trauma or of any other discernible cause in these two patients, and at necropsy the typical changes of acute pancreatitis were found. The remnant of the pancreas was histologically normal and there was no evidence of any underlying disorder to account for this rare occurrence. When age was related to aetiology (table VIII), apart from the characteristic pattern in the alcoholic group referred to above the number of cases of gall-stone pancreatitis was found to rise steadily with age, but in the idiopathic group there was a more even spread of cases, particularly in the under-50s. Finally, though the number of cases of mumps pancreatitis was small it is interesting that only one case developed in a child, aged 8 years. Most of the patients were in their teens or 20s and the oldest was a man of 48. TABLE ViII-Age Distribution and Aetiology Age (Years) 80
23 47 63 83 56 27
Total
316
No. with Alcoholism
2 6 14 4
26
No. with Idiopathic Disease 2 5
No. with Mumps
203
7
10 10 17 49 68 31 11
1 2 2 1 1
Pattern of Sex Distribution
Of the 590 patients in this survey 226 were male and 364 female. This is meaningful only when broken down by aetiology (table VI). Gall-stone pancreatitis was twice as common in women than in men. Chronic alcoholism produced its effect mainly in middle-aged men (see table VIII), while after allowing for there being more women than men at the higher ages in the general population there was no significant sex difference in incidence in the idiopathic group.
TABLE VI-Sex Distribution and Aetiology Aetiological Group .. .. Gall-stone .. .. Alcoholism .. .. Idiopathic .. Total
No. of Male Patients
No. of Female Patients
102 23 89
214 3 114
214
331
TABLE vII-Age Distribution of Patients who Died Age (Years)
No. of Cases
69 10203040506070>80 Total
3 8
30
41 84
120 159 100 45 590
Age The incidence of acute pancreatitis
Deaths No.
%
2 1 1 7 7 18 42 18
66-7 12-5 3-3 17-1 8-3 15-0 26-4 23 0 40-0
119
20-2
23
rises
with
age
(table VII),
Conclusion The overall incidence of acute pancreatitis in Great Britain is not known. The only reliable statistic available2 gives the mortality from the condition as 11-4 per million population. The present survey, with a slightly though not significantly lower absolute mortality, has shown an overall incidence of acute pancreatitis over a seven-year period of 53-8 per million in the Bristol clinical area. Critical evaluation of these data, however, suggests that this figure does not give a correct picture of the present position. It is probably too low, and for two reasons. Firstly, it appears that alcoholic pancreatitis is on the increase, especially in some urban areas. This subgroup will not only form a larger proportion of any collected series but will also swell the overall total. Then, secondly, it is becoming apparent that the composition of any group of cases will depend not only on the absolute factors which may be operating in a given areaaetiology and age of the population at risk are both importantbut also on the enthusiasm with which the serum amylase test is applied to patients being admitted with an acute abdomen. Clearly, the more widely the test is used the greater the number of cases that will be uncovered. This will lead to a relative increase in incidence and also to a shift in the clinical spectrum because of the inclusion of more mild forms of the disease. As a result there may be a misleading lowering of the apparent mortality, particularly if this continues to be expressed as case mortality rates rather than in terms related to the total population at risk. This needs to be stressed at present, for there has been great interest in comparative mortality figures in connexion with claims which have been made for various forms of treatment. Mortality rates of 12%,17 115%,'18 5-5% in a subgroup,'0 and 4 7%1 have all been reported, but in each instance the numbers of patients from which these statistics were derived were small, varying from 30 to 78 cases. Also the composition of this clinical material and the diagnostic criteria were not always mentioned. In particular, no details were given in these reports of whether cases of recurrent acute or chronic relapsing pancreatitis, with
BRITISH MEDICAL JOURNAL
183
26 APRIL 1975
their much lower mortality, had been included, and all this data must be provided if the claims made are to be substantiated. Except for one report of a controlled trial of aprotinin (Trasylol)20 no drug or regimen of treatment has been shown to be of unequivocal value in the treatment of acute pancreatitis, and during the period of this survey no special treatments were given to the patients under review. Twenty-two patients did receive aprotinin in effective dosage but this was not until 1968 and 1969, after the "complete" years used for the detailed statistical analysis. It is therefore interesting that when the absolute mortality for a first attack of acute pancreatitis, which was found to be 9 per million population, was taken and expressed as a percentage of the total number of cases collected, the case mortality rate was still of the order of 17% during the years 1961-7 in the Bristol clinical area.
References O'Sullivan, J. N., et al., Gastroenterology, 1972, 62, 373. 2
1
Registrar General's Statistical Review of England and Wales, 1969, part I, Medical Tables. London, H.M.S.O., 1971.
Hill, A. B., Principles of Medical Statistics, 8th edn., p. 228. London, Lancet, 1966. J. M., and Jordan, G. L., Surgical Diseases of the Pancreas. Philadelphia, Lippincott, 1960. 6 White, T. T., Pancreatitis. London, Arnold, 1966. ' Mayday, G. B., and Pheils, M. T., Medical Journal of Australia, 1970, 1, 1142. 7Marks, I. N., and Bank, S., South African Medical journal, 1963, 37, 1039. 8 Sarles, H., et al., Gut, 1965, 6, 545. 9 James, O., Agnew, J. E., and Bouchier, I. A. D., British Medical Journal, 1974, 2, 34. 10 Imrie, C. W., British Journal of Surgery, 1974, 61, 539. 11 Johnson, W. C., and Nabseth, D. C., Annals of Surgery, 1970, 171, 309. 12 Langmade, C. F., and Edmondson, H. A., Surgery, Gynecology and Obstetrics, 1951, 92, 43. 13 Joske, R. A., British Medical journal, 1955, 1, 124. 14 Wyatt, A. P., Annals of the Royal College of Surgeons of England, 1974, 54, 229. 16 Corlett, R. C., and Mishell, D. R., American Journal of Obstetrics and Gynecology, 1972, 113, 28. 16 Trapnell, J. E., M.D. thesis, 1966, University of Cambridge. 17 Condon, J. R., Knight, M., and Day, J. L., British Journal of Surgery, 1973, 60, 509. 18 Imrie, C. W., and Blumgart, L. H., British Medical Journal, 1974, 1, 38. 19 Slade, A. J., British Medical J7ournal, 1974, 1, 201. 20 Trapnell, J. E., et al., British journal of Surgery, 1974, 61, 177. 21 Haemmerli, U. P., Hefti, M. L., and Schmid, M., Bibliotheca Gastroenterologica, 1965, 7, 58. 3
4 Howard,
Merrison Committee Report of G.M.C. Inquiry British
Medical_Journal,
1975, 2, 183-188
The Report of the Committee
of Inquiry into the Regulation of the Medical Profession
(Cmnd 6018, price £1.75)
published on 16 April. The Committee,* chaired by Dr. A. W. Merrison, was . | 2 set up in 1972 by Sir Keith X; Joseph "to consider what changes need to be made in the existing provisions for the regulation of the medical profession; what functions should be assigned to the body charged with the responsibility for its Dr. A. W. Merrison, F.R.S. regulation; and how that body should be constituted to enable it to discharge its functions most effectively; and to make recommendations." Printed below are extracts from the report, together with the conclusions and recommendations of each of its sections. was
*Members were: Dr. A. W. Merrison, F.R.S. (vice-chancellor of Bristol University); Dr. J. R. Bennett (consultant physician, Hull Royal Infirmary); Mr. C. M. Clothier (recorder, Master of the Bench of the Inner Temple, Judge of Appeal, Isle of Man); Miss Margaret Drabble (writer); Miss Catherine M. Hall (general secretary, Royal College of Nursing); Mr. N. G. C. Hendry (consultant orthopaedic surgeon, Aberdeen Royal Infirmary); Dr. D. H. Irvine (general practitioner, Northumberland, honorary secretary of the Royal College of General Practitioners); Mr. Ian MacDonald (president of the Council of Industrial Tribunals, Scotland); Professor D. C. Marsh (professor of applied social science, University of Nottingham); Miss Audrey M. Prime (staff side secretary, General Whitley Council for the Health Services of Great Britain); Professor K. Rawnsley (professor of psychological medicine, Welsh National School of Medicine, dean of the Royal College of Psychiatrists); Professor G. A. Smart (director of the British Postgraduate Medical Federation); Mrs. Jean G. C. Turner (surgical registrar); Mrs. Mary Warnock (research fellow, Lady Margaret Hall, University of Oxford); Dr. W. B. Whowell (general practitioner, Leicestershire).
General Views of Committee In its opening chapter the report sets out the general views of the committee, stating: "In developing our views on the regulation of the medical profession, we come to the conclusion that these powers could be exercised only by a regulatory body (and we retain for it the name 'General Medical Council') constituted in a way substantially different from the present G.M.C. Our proposals for education and the judging of a doctor's fitness to practise must be read with this always in mind. When, therefore, we refer to the G.M.C. in the rest of this report we mean (unless the context is historical) the G.M.C. which we recommend to take the place of the present one. "We do not attempt, in the report which follows, to solve all the problems of regulating the medical profession. Our task has been primarily to recommend machinery for the solution of problems and in some areas to point the direction of possible solutions which the profession itself must work out. What we have suggested is a framework within which difficulties can be resolved and which, we hope, will satisfy the profession and the community it serves; and be sufficiently efficient and flexible to take account of rapid continuing progress in science and technology, the changing use of medical resources, and the movement in attitude and outlook of the profession and public alike." CONCLUSIONS
"Medical registration provides a means of recognizing the competent practitioner. "It is advantageous to the public to be able to recognize, and to a member of the medical profession to be regarded as, a competent medical practitioner. "The medical register is used by the public at second hand. "A medical register necessarily involves a registeringbodywith considerable powers, particularly over the providers of medical education.
26 APRIL 1975
Conclusions It is understandable that more debate has been given to whether or not psychotropic drugs should be prescribed than to which psychotropic drugs should be prescribed. I believe that both debates would be assisted by a smaller range of preparations being available for prescription. Patients would receive drugs more familiar to the doctor, which must reduce side effects and inappropriate medication. Would there be an equal disadvantage in depriving patients of
179
the existing selection of drugs ? The question deserves critical examination.
References 2
4
5
Shepherd, M., et al., Psychiatric Illness in General Practice. London, Oxford University Press, 1966. Wells, F. O., Journal of the Royal College of General Practitioners, 1973, 23, 164. Durno, D., "Activity Analysis in General Practice," M.D. Thesis, Aberdeen, 1972. Durno, D., Personal communication, 1974. Petrie, J. C., Howie, J. G. R., and Durno, D., British Medical Journal, 1974, 2, 262.
Clinical Review Patterns of Incidence in Acute Pancreatitis JOHN E. TRAPNELL, E. H. L. DUNCAN British Medical journal, 1975, 2, 179-183
Summary A review of acute pancreatitis occurring over a 20-year period in the Bristol clinical area is reported. A total of 590 cases were available for analysis. The yearly incidence was 53 8 per million population at risk, with a mortality of 9 0 per million. This compares favourably with 11 4 deaths per million for England and Wales as a whole during the same period but the difference is not statistically significant. When the deaths occurring in the Bristol clinical area were expressed in terms of case mortality rate the figure was 17%. In contrast the mortality for recurrent acute pancreatitis was only 1-5%, and the benign nature of this second condition is confirmed. Aetiological factors and age and sex distribution were also analysed in relation to each other and to mortality. An increase in acute pancreatitis secondary to chronic alcoholism was confirmed and steroid pancreatitis also emerged as a definite entity in this survey. The pattern of recurrence in patients with idiopathic pancreatitis was studied in detail and is analysed on an actuarial basis. Introduction The full pattern of the incidence of acute pancreatitis is hard to determine. Though large series of cases have been reported from many centres in different parts of the world most have been concerned with the disease as it occurs in selected segments of the population. Reports from special clinics in the United States have been on patients referred from a wide area, and the city hospital patients may have related to some large urban conurbation, but these centres serve only low-income groups. Royal Victoria Hospital, Boscombe, Bournemouth JOHN E. TRAPNELL, M.D., F.R.C.S., Consultant Surgeon Department of Community Health, University of Bristol E. H. L. DUNCAN, M.A., B.SC., Senior Lecturer in Medical Statistics
Furthermore, the Veterans Hospitals, the other source of large series of cases in the United States, also represent a highly selected section of the population. In the same way, incidence and mortality statistics from other parts of the world have always been derived from hospital case series and it is never clear how closely they relate to the disease as it occurs in the general population. There has been only one report of a population survey.' This was undertaken in Rochester, Minnesota, and though some of the details, particularly of the follow-up and checking procedures, were scanty it provided the first geographical yearly incidence for all forms of the disease ranging from 150 to 180 cases per million population. For acute pancreatitis alone (but including recurrent acute pancreatitis) the yearly incidence ranged from 100 to 115 cases per million. We present an analysis of all cases of acute pancreatitis occurring in the Bristol clinical area during the 20-year period 1950-69. The patients were drawn from the large urban area of the city of Bristol and its suburbs and from the rural area of south Gloucestershire and north Somerset.
Clinical Material In collecting the cases a most careful search was made in all the hospitals in the Bristol area. Details of the few cases treated in private hospitals were also kindly supplied by the consultants concerned, and these figures were then checked, firstly, by reference to the regional hospital board's hospital activity analysis system and then by examination of the cases certified as dying from the disease during the period. In this connexion detailed reports of the coroners' necropsies were obtained. Interestingly, the regional hospital board's statistics and the coroners' data consistently produced one or two cases of acute pancreatitis each year not identified from the other sources. The diagnostic criteria were either a consistent clinical picture and a serum amylase level over 1000 Somogyi units or acute pancreatitis plainly evident at laparotomy or necropsy. On the basis of these criteria and from the search outlined 590 primary attacks were identified over the 20-year period. There was no significant referral rate of patients with pancreatitis from outside the area and thus the series seemed to be fully representative of the disease pattern in the area at that time. Because the hospital activities analysis data and coroners'
180
BRITISH MEDICAL JOURNAL
records were available only to the end of 1967, for the purpose, of detailed statistical analysis this has been regarded as the las "complete" year. During 1950-67 there were 551 cases, and al1 of these patients who survived their first attack were followec up in the second half of 1968 or in early 1969. Thus follow-up of these patients ranged from one to 18 years. II
I
Overall Incidence and Mortality During the first six of the 18 "complete" years under review (1950-5) there was an increase in the number of cases diagnosed, and another peak occurred during 1961-4 (table I). Over the whole period, however, the mortality rate did not alter significantly, and from a critical review of local factors it appeared that the initial increase was due to a wider application of the serum amylase test, which was part of a national trend at the time. The peak in the early 1960s was due to a period of heightened interest in the condition locally. Thus both increases seem to have been due to improved diagnosis rather than to an absolute change in incidence. With the relative increase in incidence of the disease over the years there was a relative decrease ir mortality (tables Ia, Ib). Critical consideration of the method used to express mortalitythat is, as a percentage of the number of cases diagnosedsuggests that it is apt to give a false impression, for there was, in fact, no significant absolute change in the death rate in the total population over the years in question. TABLE Ia-Incidence of Acute Pancreatitis in Bristol Clinical Area, 1950-67 Year 1950 1951 1952 1953 1954 1955 1956 1957 1958
lNo. 8
Year
No.
2
1959 1960 1961 1962 1963 1964 1965 1966 1967
32 28 35 54 36 47 37 23 30
4 8 6 3
20 20
29 30 36 23
pancreatitis. Eight of these were lost to follow-up. Of the remaining 430 patients 133 had suffered a total of 261 recurrent TABLE Ii-Incidence of Cases and Death Rates per Million Population from Acute Pancreatitis in Bristol Clinical Area during 1961-7 Year
Cases
Population
11
5
31
4 5
32
TABLE ib-Summary of Incidence in Bristol Clinical Area Years
No.
Died
1950-5 1956-61 1962-7
143 181 227
34 42 37
23-8 23-2 16-3
Total
551
113
20-5
Of the 551 patients in the survey 113 died of their first attack, giving a primary case mortality rate of 20-5%. As noted above, the expression of mortality as a percentage of diagnosed cases does not necessarily give a reliable figure. To achieve this deaths must be related to the total population at risk. The Registrar General's Statistical Review of England and Wales2 states that during 1961-7 there were 3790 deaths from acute pancreatitis (I.C.D. 577-0) and that there were 331 669 900 person-years at risk, giving a mean yearly death rate of 11 4 per million population. Analysis of the 1961 Census figures and the Registrar General's estimate for the years between censuses shows that there were 4 867 985 person-years at risk in the Bristol clinical area, and with 44 deaths during that period the death rate was 9 0 per million (table II). Though the death rate for the survey area was lower than the national average the difference is not statistically significant.
Recurrent Acute Pancreatitis Altogether 438 patients survived their initial attack of
acute
No.
Rate per Million
1961 1962 1963 1964 1965 1966 1967
676 770 683-035 689-440 695-810 702-605 707-710 712-615
35 54 36 47 37 23 30
51-7 79 1 52-2 67-5 52-7 32-5 42-1
Total
4 867 985 Person-years
262
53-8
Deaths I. No. Rate per Million 7 10-3 8 11*7 6 8-7 7 10-1 4 5-7 5 7-1 7 9-8 44 90
TABLE III-Results of Follow-up of 430 Patients Surviving First Attack of Acute Pancreatitis Aetiological Group Gall-stone .. Idiopathic .. .. Alcoholic .. Other Total
No. of Patients Available for
No. Having
Recurrence
Total No. of Recurrent Attacks
No. of Deaths
232 148 18 32
88 35 10
182 61 18
2 2
430
133
261
4
Follow-up
acute attacks but only four had died (table III). When these deaths are expressed as a percentage of the total number of recurrent episodes (1-5%) it is clear that the outlook for this group is vastly different from that for patients in their first attack. Thus recurrent acute pancreatitis is by comparison a relatively benign condition with a much improved prognosis. The aetiological factors in the 430 cases are set out in table III. Of the 88 patients with gall stones having a recurrent attack at least 36 were on a waiting list for cholecystectomy. Only a few had had their biliary tract disease long enough for a sequence of recurrences to develop. It was, however, a feature of the pattern of recurrence in this group that while most attacks recurred within a year or two of the previous episode there were on occasions long gaps-10 to 15 years or over in several instances -and during these intervals the patients were entirely symptomfree. These follow-up studies therefore confirm that if biliary tract disease is not treated adequately there is an ever-present danger of recurrent acute pancreatitis. One of the two patients with gall stones who died was a man of 49 with long standing diabetes. After an initial episode in 1950 he succumbed to his first recurrent attack four years later. The other death was also in a man, in this case aged 41. Again it was his first recurrence, this time after an interval of only eight months. One-quarter of the patients in the idiopathic group suffered a recurrent attack, but except for one of the two who died only two patients had more than two recurrent acute episodesnamely, four attacks each. It was again a feature of this group that these further attacks were mild and uncomplicated. Calculated by the actuarial method3 10% of patients with idiopathic pancreatitis are likely to have a first recurrence within one year of the primary attack, 17% within two years, and 25% within six years. No recurrence occurred after a gap of more than six years, though 75 patients were followed up after this period for a total of 444 person-years. Patients with idiopathic pancreatitis who have had one recurrence are likely to have a second. We found, again using the actuarial method, that 12% have a second recurrence within a year, 16% within two years, 31% within three years, and 56% within six years. One of the patients with idiopathic pancreatitis who died of a recurrence was a woman of 62; she succumbed to her third attack in five years. The other patient who died was unique in
I~~~~~~~ Died
26 APRIL 1975
BRITISH MEDICAL JOURNAL
181
26 APRIL 1975
this series. She suffered her first attack of acute pancreatitis in 1952 when she was aged 12. Over the next eight years she had 10 further, separate episodes of varying severity but remained well between attacks, with no evidence of exocrine or endocrine insufficiency and no sign of pancreatic calcification. Partial pancreatectomy with retrograde drainage into a Roux loop was performed after the sixth recurrence but in spite of this the attacks persisted. Finally, she was readmitted in September 1960 with the usual acute presentation. Instead of responding to intravenous fluids and standard measures, however, her condition deteriorated unexpectedly and rapidly. She became shocked and cyanosed and died within 72 hours. Necropsy showed widespread fat necrosis and oedema but no haemorrhage or necrosis of the pancreas. The gland itself was fibrotic, the main duct was not dilated, and the Roux loop had separated from the distal end. The alcoholic group is, perhaps, more difficult to assess, for it is characteristic of these patients that they progress to the chronic relapsing rather than the recurrent acute form of the disease. Indeed, several of the 18 cases followed up pursued such a course, but 10 men all suffered one or more further episode of acute pain with vomiting associated with a marked increase in their serum amylase level at a time when they were otherwise well and entirely symptom-free. They must therefore be included in this survey as examples of recurrent acute episodes. There were no deaths from any of these attacks or, indeed, from the chronic relapsing disease when it supervened in this group of cases.
Aetiology The aetiologically related factors for the whole series of 590 cases are set out in table IV. Analysis of the two major groups, gall-stone and idiopathic pancreatitis, shows that there was no significant change in the patterns of incidence during the 20 years of this survey. In particular, it is disappointing that the incidence in the gall-stone group did not fall. It is now common surgical practice in Great Britain to remove all gall stones more or less "on sight" even when they are symptomless, and it was hoped that one of the benefits might have been a decrease in the prevalence of gall-stone pancreatitis. A partial explanation for a lack of any change in incidence may be that just over a quarter of these patients (81 out of 316) had no past history of any abdominal symptoms before the onset of their attack, and hence there was no pointer to their underlying biliary tract disease. It is also of interest that 60 patients in the idiopathic group (30%) gave a history at their first admission of dyspepsia of varying severity and duration. Some of these patients had been extensively investigated with negative findings. At least 26 who were followed up not only remained recurrence-free but also reported that their previous dyspeptic symptoms had disappeared completely. From this it appears that when a patient presenting with an attack of acute pancreatitis gives a history of abdominal dyspepsia this should not necessarily be taken as a pointer to the presence of gall stones. Furthermore, patients with previous trouble are no more liable to have persisting symptoms or develop recurrence than those without ally past history. In the minority subgroups those patients with acute pant creatitis secondary to mumps and hyperparathyroidism do norequire any special mention, save to say that these are rare predisposing causes. All of the patients with carcinoma, seven of the head of the gland, and one of the ampulla of Vater, eventually died of their disease but, interestingly, apart from the one who died from acute pancreatitis all the survivors lived for a year or more, and one lived for three and a half years before dying of his malignancy. In all these cases there appeared to have been great difficulty in identifying the underlying disease, and in five cases the diagnosis was conclusively established only at necropsy.
TABLE Iv-Aetiological Factors in 590 Cases of Acute Pancreatitis Cases No.
Biliary tract disease Chronic alcoholism .26 Mumps .7 Hyperparathyroidism .1 Carcinoma .8 Steroids .6 Other. Not stated* .14 Idiopathic .203
0
316
53-6 4-4 1-2 0-2
9
1-5 2-4 34-4
1-4
1-0
590 100-0 *In these cases the presence or absence of biliary tract disease niined.
-. .
Total
Deaths No. % 19-9 63 3 11-5 0 9 0 0 1 12-5 4 66-7 11-1 1 2 14-3 45 22-2 20-2 119 had not been deter-
In the other two groups of cases, alcoholic and steroid pancreatitis, there was a significant change in the pattern of incidence over the 20-year period. It has been recognized for many years that the principal difference in aetiological spectrum between Great Britain and most other countries where acute pancreatitis is relatively common lies in the fact that chronic alcoholism is fairly uncommon as a cause in this country. In published series from the United States,4 Australia,5 6 South Africa,7 and Europe8 the alcoholic contingent constituted as many as 30 to 40% of the total number of cases seen. The present findings, however, suggest that the low-incidence pattern is changing (table V). This statistically significant trend confirms the impression presented in a recent review of chronic relapsing pancreatitis in London 9 and also the preliminary findings of a current study in Glasgow,' 0 where it appears that 20 to 25o% of cases now have an alcoholic background. It is therefore clear from this analysis and from supportive evidence that the figure of 4-4% derived from the present study, stretching as it does over 20 years, does not represent a correct picture of the current pattern of the disease. TABLE v-Five-yearly Incidence of Alcoholic Pancreatitis Period No. of cases
1950-4
1955-9
1960-4
1965-9
2
4
7
13
Total 26
Steroid pancreatitis emerged as a definite entity in the last few years. One of the six cases in this survey occurred in 1955, one in 1960, and the remaining four all after 1965. This will clearly remain an uncommon cause of pancreatitis but the numbers of cases may be expected to increase as more patients are put on long-term steroid therapy. This may therefore pose an additional problem for the transplant patient receiving steroid medication as part of his immunosuppressive regimen."' Finally, pregnancy is not included in the list of aetiological factors in this survey. Several workers in the 1950s suggested that pregnancy may predispose to acute pancreatitis,'2 13 but Howard4 questioned this assumption and suggested that as many of these women were subsequently shown to have biliary tract disease this was the aetiologically important factor rather than the pregnancy. Nevertheless, the association is still being reproduced in lists of aetiological factors,'0 14 and recently Corlett and Mishell'5 reported one episode of pancreatitis in 1066 births. These workers did not give their diagnostic criteria and the clinical details were sparse, so that it is difficult to judge the relevance of this statistic. In the present survey of 590 cases only 17 patients developed acute pancreatitis either during pregnancy (4 cases) or during a six-week post-partum period (13 cases). There was one death. At necropsy the patient was found to have gall stones, and of the survivors 14 also had biliary tract disease, which was subsequently eradicated surgically. Follow-up of these patients showed only one who developed further symptoms, and she, after sterilization, progressed in a typical fashion over 10 years to a chronic relapsing form of the disease. All the remaining patients were fit and well and symptom-free when interviewed.
182
BRITISH MEDICAL JOURNAL
Five women, including the two with idiopathic disease, had had one further child each. Four had had two further pregnancies and one three further children, all without incident. If we accept that the associated biliary tract disease was responsible for the pancreatitis in 15 of these 17 patients it seems likely that the remaining two cases could well have arisen by chance during pregnancy or the puerperium. During the 20year period there were over 220 000 live,births and stillbirths in the Bristol clinical area. If pregnancy does, in fact, predispose to acute pancreatitis one would have expected to see many more cases in a study of this size over this period of time. It must therefore be concluded that there is no evidence for any such association, and until a contrary view can be adequately supported there are no grounds for sterilizing women who have had an attack of acute pancreatitis.
Mortality and Aetiology The primary mortality rates for the two main groups of cases were comparable (table IV), and other, more detailed studies'16 have confirmed this. There were too few cases of alcoholic pancreatitis in this study for any conclusions about mortality patterns to be drawn. We must therefore await a report of a larger series of cases before it will be possible to see whether the death rate for the alcoholics, even in their first attack, coincides with that of the gall-stone and idiopathic groups. Even though the numbers are small it appears that steroid pancreatitis departs, from the overall spectrum for it clearly carries a bad prognosis with a high primary mortality rate.
26 APRIL 1975
and so does mortality. Analysis of the data shows that the four deaths in patients under 30 were all in the idiopathic group, and, in particular, the two very young cases were both postpartum deaths at one and five weeks. There was no evidence of birth trauma or of any other discernible cause in these two patients, and at necropsy the typical changes of acute pancreatitis were found. The remnant of the pancreas was histologically normal and there was no evidence of any underlying disorder to account for this rare occurrence. When age was related to aetiology (table VIII), apart from the characteristic pattern in the alcoholic group referred to above the number of cases of gall-stone pancreatitis was found to rise steadily with age, but in the idiopathic group there was a more even spread of cases, particularly in the under-50s. Finally, though the number of cases of mumps pancreatitis was small it is interesting that only one case developed in a child, aged 8 years. Most of the patients were in their teens or 20s and the oldest was a man of 48. TABLE ViII-Age Distribution and Aetiology Age (Years) 80
23 47 63 83 56 27
Total
316
No. with Alcoholism
2 6 14 4
26
No. with Idiopathic Disease 2 5
No. with Mumps
203
7
10 10 17 49 68 31 11
1 2 2 1 1
Pattern of Sex Distribution
Of the 590 patients in this survey 226 were male and 364 female. This is meaningful only when broken down by aetiology (table VI). Gall-stone pancreatitis was twice as common in women than in men. Chronic alcoholism produced its effect mainly in middle-aged men (see table VIII), while after allowing for there being more women than men at the higher ages in the general population there was no significant sex difference in incidence in the idiopathic group.
TABLE VI-Sex Distribution and Aetiology Aetiological Group .. .. Gall-stone .. .. Alcoholism .. .. Idiopathic .. Total
No. of Male Patients
No. of Female Patients
102 23 89
214 3 114
214
331
TABLE vII-Age Distribution of Patients who Died Age (Years)
No. of Cases
69 10203040506070>80 Total
3 8
30
41 84
120 159 100 45 590
Age The incidence of acute pancreatitis
Deaths No.
%
2 1 1 7 7 18 42 18
66-7 12-5 3-3 17-1 8-3 15-0 26-4 23 0 40-0
119
20-2
23
rises
with
age
(table VII),
Conclusion The overall incidence of acute pancreatitis in Great Britain is not known. The only reliable statistic available2 gives the mortality from the condition as 11-4 per million population. The present survey, with a slightly though not significantly lower absolute mortality, has shown an overall incidence of acute pancreatitis over a seven-year period of 53-8 per million in the Bristol clinical area. Critical evaluation of these data, however, suggests that this figure does not give a correct picture of the present position. It is probably too low, and for two reasons. Firstly, it appears that alcoholic pancreatitis is on the increase, especially in some urban areas. This subgroup will not only form a larger proportion of any collected series but will also swell the overall total. Then, secondly, it is becoming apparent that the composition of any group of cases will depend not only on the absolute factors which may be operating in a given areaaetiology and age of the population at risk are both importantbut also on the enthusiasm with which the serum amylase test is applied to patients being admitted with an acute abdomen. Clearly, the more widely the test is used the greater the number of cases that will be uncovered. This will lead to a relative increase in incidence and also to a shift in the clinical spectrum because of the inclusion of more mild forms of the disease. As a result there may be a misleading lowering of the apparent mortality, particularly if this continues to be expressed as case mortality rates rather than in terms related to the total population at risk. This needs to be stressed at present, for there has been great interest in comparative mortality figures in connexion with claims which have been made for various forms of treatment. Mortality rates of 12%,17 115%,'18 5-5% in a subgroup,'0 and 4 7%1 have all been reported, but in each instance the numbers of patients from which these statistics were derived were small, varying from 30 to 78 cases. Also the composition of this clinical material and the diagnostic criteria were not always mentioned. In particular, no details were given in these reports of whether cases of recurrent acute or chronic relapsing pancreatitis, with
BRITISH MEDICAL JOURNAL
183
26 APRIL 1975
their much lower mortality, had been included, and all this data must be provided if the claims made are to be substantiated. Except for one report of a controlled trial of aprotinin (Trasylol)20 no drug or regimen of treatment has been shown to be of unequivocal value in the treatment of acute pancreatitis, and during the period of this survey no special treatments were given to the patients under review. Twenty-two patients did receive aprotinin in effective dosage but this was not until 1968 and 1969, after the "complete" years used for the detailed statistical analysis. It is therefore interesting that when the absolute mortality for a first attack of acute pancreatitis, which was found to be 9 per million population, was taken and expressed as a percentage of the total number of cases collected, the case mortality rate was still of the order of 17% during the years 1961-7 in the Bristol clinical area.
References O'Sullivan, J. N., et al., Gastroenterology, 1972, 62, 373. 2
1
Registrar General's Statistical Review of England and Wales, 1969, part I, Medical Tables. London, H.M.S.O., 1971.
Hill, A. B., Principles of Medical Statistics, 8th edn., p. 228. London, Lancet, 1966. J. M., and Jordan, G. L., Surgical Diseases of the Pancreas. Philadelphia, Lippincott, 1960. 6 White, T. T., Pancreatitis. London, Arnold, 1966. ' Mayday, G. B., and Pheils, M. T., Medical Journal of Australia, 1970, 1, 1142. 7Marks, I. N., and Bank, S., South African Medical journal, 1963, 37, 1039. 8 Sarles, H., et al., Gut, 1965, 6, 545. 9 James, O., Agnew, J. E., and Bouchier, I. A. D., British Medical Journal, 1974, 2, 34. 10 Imrie, C. W., British Journal of Surgery, 1974, 61, 539. 11 Johnson, W. C., and Nabseth, D. C., Annals of Surgery, 1970, 171, 309. 12 Langmade, C. F., and Edmondson, H. A., Surgery, Gynecology and Obstetrics, 1951, 92, 43. 13 Joske, R. A., British Medical journal, 1955, 1, 124. 14 Wyatt, A. P., Annals of the Royal College of Surgeons of England, 1974, 54, 229. 16 Corlett, R. C., and Mishell, D. R., American Journal of Obstetrics and Gynecology, 1972, 113, 28. 16 Trapnell, J. E., M.D. thesis, 1966, University of Cambridge. 17 Condon, J. R., Knight, M., and Day, J. L., British Journal of Surgery, 1973, 60, 509. 18 Imrie, C. W., and Blumgart, L. H., British Medical Journal, 1974, 1, 38. 19 Slade, A. J., British Medical J7ournal, 1974, 1, 201. 20 Trapnell, J. E., et al., British journal of Surgery, 1974, 61, 177. 21 Haemmerli, U. P., Hefti, M. L., and Schmid, M., Bibliotheca Gastroenterologica, 1965, 7, 58. 3
4 Howard,
Merrison Committee Report of G.M.C. Inquiry British
Medical_Journal,
1975, 2, 183-188
The Report of the Committee
of Inquiry into the Regulation of the Medical Profession
(Cmnd 6018, price £1.75)
published on 16 April. The Committee,* chaired by Dr. A. W. Merrison, was . | 2 set up in 1972 by Sir Keith X; Joseph "to consider what changes need to be made in the existing provisions for the regulation of the medical profession; what functions should be assigned to the body charged with the responsibility for its Dr. A. W. Merrison, F.R.S. regulation; and how that body should be constituted to enable it to discharge its functions most effectively; and to make recommendations." Printed below are extracts from the report, together with the conclusions and recommendations of each of its sections. was
*Members were: Dr. A. W. Merrison, F.R.S. (vice-chancellor of Bristol University); Dr. J. R. Bennett (consultant physician, Hull Royal Infirmary); Mr. C. M. Clothier (recorder, Master of the Bench of the Inner Temple, Judge of Appeal, Isle of Man); Miss Margaret Drabble (writer); Miss Catherine M. Hall (general secretary, Royal College of Nursing); Mr. N. G. C. Hendry (consultant orthopaedic surgeon, Aberdeen Royal Infirmary); Dr. D. H. Irvine (general practitioner, Northumberland, honorary secretary of the Royal College of General Practitioners); Mr. Ian MacDonald (president of the Council of Industrial Tribunals, Scotland); Professor D. C. Marsh (professor of applied social science, University of Nottingham); Miss Audrey M. Prime (staff side secretary, General Whitley Council for the Health Services of Great Britain); Professor K. Rawnsley (professor of psychological medicine, Welsh National School of Medicine, dean of the Royal College of Psychiatrists); Professor G. A. Smart (director of the British Postgraduate Medical Federation); Mrs. Jean G. C. Turner (surgical registrar); Mrs. Mary Warnock (research fellow, Lady Margaret Hall, University of Oxford); Dr. W. B. Whowell (general practitioner, Leicestershire).
General Views of Committee In its opening chapter the report sets out the general views of the committee, stating: "In developing our views on the regulation of the medical profession, we come to the conclusion that these powers could be exercised only by a regulatory body (and we retain for it the name 'General Medical Council') constituted in a way substantially different from the present G.M.C. Our proposals for education and the judging of a doctor's fitness to practise must be read with this always in mind. When, therefore, we refer to the G.M.C. in the rest of this report we mean (unless the context is historical) the G.M.C. which we recommend to take the place of the present one. "We do not attempt, in the report which follows, to solve all the problems of regulating the medical profession. Our task has been primarily to recommend machinery for the solution of problems and in some areas to point the direction of possible solutions which the profession itself must work out. What we have suggested is a framework within which difficulties can be resolved and which, we hope, will satisfy the profession and the community it serves; and be sufficiently efficient and flexible to take account of rapid continuing progress in science and technology, the changing use of medical resources, and the movement in attitude and outlook of the profession and public alike." CONCLUSIONS
"Medical registration provides a means of recognizing the competent practitioner. "It is advantageous to the public to be able to recognize, and to a member of the medical profession to be regarded as, a competent medical practitioner. "The medical register is used by the public at second hand. "A medical register necessarily involves a registeringbodywith considerable powers, particularly over the providers of medical education.
