SUMMARY The problems associated with recurrent Crohn's disease were examined in a series of 168 patients who had undergone primary resection for this condition at the General Infirmary at Leeds from 1939 to 1968 inclusive.The overall recurrence rate was 34.2%. The risk of recurrence was less in patients with involvement mainly of large bowel rather than small. It was also affected by the age of the patient, being greatest in children or adolescents, less in adults, and least in those over 60 years of age.Recurrent disease was most commonly found in the small bowel proximal to an anastomosis, and usually manifested itself either in the first year or two after surgery or some five to 15 years later. Patients with 'early' recurrence had a shorter history of symptoms at operation, and a graver outlook than those with 'late' recurrence.The status of recurrent cases was much better than might have been anticipated in so far as nearly 70% of them were considered at review to be in very good or good general health. Moreover the risk of further recurrence after a second or third operation was found to be no greater than after a first operation.

No abstract

SUMMARYAn analysis is presented of the early results of 415 operations performed for Crohn's disease on 295 patients at the Leeds General Infirmary from 1939 to 1968 inclusive. Two hundred and forty-four of the operations were primary and the other 171 were reoperations.The operative mortality was 4.1 per cent after primary operation, 8.2 per cent after reoperation, and 5.8 per cent overall. The most important factor influencing the immediate mortality after primary surgery was the age of the patient, the mortality in patients under 50 years being 2-6 per cent and that in patients over that age 8.9 per cent.The main postoperative complications were sepsis, fluid and electrolyte imbalance, and delay in the healing of the perineal wound, which persisted unhealed for over 6 months in 42 per cent of the patients who had had rectal excision.In the interests of simplicity assessineizt of the ultimate outcome of surgery is confined to an examination of the late results in the 234 immediate survivors of primary operations by means of a followup ranging from 6 months to 30 years. Four patients could not be traced and 2 had had their operations less than 6 months before the completion of the review in 1968, which left 228 patients for analysis. The overall late mortality in this group was 13 per cent. Much of it was due to intercurrent conditions such as would be expected in a matched sample of the general population, but 7 per cent was related to Crohn's disease, being due mostly to septic complications following reoperations.Thirty-four per cent of the patients needed at least one further operation after their initial one for Crohn's disease. However, after reoperation the chances of a third or fourth operation being necessary were not increased. The reoperation rate varied considerably after different primary operations, being lowest after 56 proctocolectomy and highest after by-pass operations and colectomy and ileorectal anastomosis.Symptomatically most of the patients have done well. At the time of review 87 per cent of them were enjoying good or very good general health (76 per cent of patients after small-bowel operations, but 96 per cent of those after large-bowel operations), and over the period of follow-up 72 per cent of the 'patient-years' were spent free of symptoms.

). Our methods seem to prevent symptomatic hypoglycaemia without the need for repeated laboratory monitoring of the blood glucose level.The reason for maintaining very large energy intakes was to prevent an interruption in brain growth with consequent possible permanent reduction in brain capacity (Dobbing, 1968). During drip feeding the reduction in handling of the small infants lowers the risk of cross-infection, diminishes the loss of energy due to cooling, and saves valuable nursing time. Feeding the smallest infants is the most time-consuming task of the staff of a special care unit, producing considerable tension at times, especially when staff is depleted. There were no insuperable difficulties in managing this trial in spite of several changes in nursing staff. Our unit is staffed mainly by pupil midwives, as in many other district hospitals, and these methods have been dictated by necessity as well as providing better medical care. COMPARISON OF MILKSAlthough this trial was designed to determine the physical and biochemical safety of large volume intakes, the differences in weight gains between the different groups are of interest. The weight gains on human and SMA-S26 milk were similar throughout the trial period. In the first week for all patients the weight gain on half-cream Regal milk was significantly greater than on SMA-26 (Tables IV and V). The infants in the half-cream Regal group received the highest mineral load, and this may have been responsible for the larger weight gains in the first two weeks, although Davidson et al. (1967) showed that variations in the mineral content of feeds over a wide range had no effect on weight gains.It has been traditional for many years to follow the dictum of Levine (Gordon et al., 1947) that breast milk is inadequate for babies of low birth weight. Hence a valuable link between mother and infant in the early days is denied to many of these infants all over the world, with a grave risk of rejection or even child abuse. In our unit mothers are encouraged to provide breast milk for their infants and have been grateful for the satisfaction of feeling that they are playing a vital part in the care of their infants. In fact, there is a high incidence of established breast-feeding of babies of low birth weight in our unit.The results of this trial show that human milk is an adequate food for infants of low birth weight provided that a sufficient daily intake is maintained. SMA-S26 has been shown to be a safe and satisfactory milk feed if breast milk is not available. However, our experience suggests that half-cream cows' milk preparations in these large volumes cannot be recommended unreservedly as a safe alternative.We thank Miss M. A. Woodward, superintendent midwife, and Sisters J. Ryans and J. Burley, who were in charge of the special care unit and without whose valuable co-operation this study could not have been completed. We are indebted to Michael

SUmMARY An analysis is presented of the course and outcome of the initial referred attack in a group of 332 patients with Crohn's disease treated at the General Infirmary at Leeds from 1939 to 1968 inclusive. A further 50 cases were excluded on the basis of insufficient diagnostic evidence: usually such patients had acute terminal ileitis.Only 5 % of patients achieved full remission on conservative management, and most eventually came to surgery. The overall fatality rate was some 3.3 % and this was affected chiefly by the severity of attack and the age of the patient. The severity of attack was classified as mild, moderate, or severe according to a system which took into account factors such as bowelhabit, rectal bleeding, abdominal pain, pulse rate, temperature, haemoglobin, and weight. In attacks graded as 'mild', no deaths occurred; in those graded 'severe' the fatality rate was 8-8%. Similarly, whilst no patient under 20 years of age died, the fatality rate in the over 60s was some 14.8 %.Further comparison between the present group of 332 patients and an earlier series of 204 patients suffering from ulcerative colitis (Watts, de Dombal, Watkinson, and Goligher, 1966a) shows considerable differences in the course and prognosis of the two diseases. In particular 70% of patients with ulcerative colitis achieved remission on conservative management alone whereas only 5% of Crohn's disease patients did so.These differences, and in particular the poor response to conservative therapy, are discussed, together with their implications for management.