508
relapsing nephrotic syndrome treated with cyclophosphamide relapse less than those treated with steroids. If this is so, cyclophosphamide may be working at a more fundamental level in the pathogenic chain, perhaps at a level still accessible to treatment during full steroid-maintained remission. Patients who, after their third relapse, had been treated successfully with steroids were randomly allocated either to steroid withdrawal or to cyclophosphamide (3 mg. per kg. per day) for eight weeks, while their renal function was completely normal, followed by steroid withdrawal. The two groups were then watched for relapse. The results of the trial are highly significant and remarkably clear: such a course of cyclophosphamide greatly reduces the prospect of relapse. The semisequential analysis’s technique 11 used permits stopping at selected times in the trial if success is achieved by then, without the complexity of a full sequential analysis. The ethical difficulties in trial design were avoided, since all the control group who relapsed after steroid withdrawal were treated with cyclophosphamide in an orthodox way during the next relapse. Very surprising was the demonstration that a disease could be treated at a stage when function tests of the target organ were normal. What was treated and how? These drugs, often called immunosuppressive, are perhaps better described as cytotoxic; they affect multiplying cells. The characteristic histological feature is absence of cell proliferation in the glomeruli, even during relapse let alone remission. Since there is evidence that the disease may be immunological,12 perhaps the drug is working at the level of an abnormal proliferation of immunologically reactive cells, which goes on all the time. Though small doses were used, even smaller ones may work and a new trial is to assess a much lower total dose. The international study group have started a trial to confirm this effect over a longer period in patients studied from first attack, and selected according to their more rigorous protocols. The small local trials, such as the one described in this issue, and the large collaborative efforts can both contribute to progress. Use of advanced techniques, not widely available, may well be possible only in single centres; and such techniques are particularly relevant in trials in kidney disease.13 Their use in the azathioprine trial may have revealed effects missed by observation of the qualitative disappearance of proteinuria. Some may argue that it is patients that are being treated, not measurements; but most or all of a group of patients may show some improvement of function while only some reap useful benefit. The function of a trial is to detect an effect if it is there; whether it is useful for any individual is a superimArmitage, P., McPherson, C. K., Rowe, B. C. J. Roy. Statist. Soc. A. 1969, 132, 235. 12. Ngu, J. L., Barratt, T. M., Soothill, J. F. Clin. exp. Immun. 1970, 6, 109. 13. Barratt, T. M., McLaine, P. N., Soothill, J. F. Archs Dis. Child. 1970, 45, 496.
11.
posed judgment. Application of optimum measurement techniques may therefore be an important advantage of the single-centre trial. But it would be wrong to regard a treatment trial as an exceptional event. Modern industry has continuous quality control of its products. Why not medicine? Clearly, when treatment becomes effective, the delivery of maximum benefit to the patient depends on such trials, and that means concentrating patients into a single or a few centres. Such action would be easiest in a small and highly populated country with a Health Service, like Britain.
Treatment of Infertility TREATMENT of
with gonadotrophins is be dangerous unless remarkably effective, elaborate laboratory resources are available for monitoring dosage. The incidence of multiple pregnancies is high, but the most disturbing side-effect is overstimulation associated with ovarian enlargement and cyst formation, and, in severe forms, ascites, pleural effusions, and changes in blood-volume and clotting-time.1 The factors responsible are incompletely defined, but it seems clear that they are related to too high a dose of follicle-stimulating hormone and possibly to the ratio of follicle-stimulating to luteinising hormone in the preparation given. The reactions can be avoided if the ovulating hormone, human chorionic gonadotrophin, is withheld, but this is wasteful in material and time. The assessment of response is based on the quantitative estimation of oestrogen production, measurement of which requires much care. Many simpler methods which give an index of oestrogen secretion have been tried, but none as yet can replace the quantitative method. They include vaginal cytology, endometrial histology, and examination of the cervical mucus; all depend on changes in oestrogen and progesterone secretion and some are affected by the balance of these two hormones. The quantity of cervical mucus increases about tenfold from the time of menstruation to mid-cycle, the maximum at midcycle preceding the rise in basal body temperature by 1-3 days. At this time the mucus is very elastic, showing maximum Spinnbarkeit, or spinability, and sperm will penetrate readily. If the mucus is spread out on a slide it shows a crystalline pattern which is greatest at the time of ovulation and resembles a fern.2 Another end-point is the simple measurement of chloride content by means of strips of silverchromate papery3 Since there is a small margin between what may be judged a safe level of oestrogen and that associated with hyperstimulation, it is
infertility but it
can
Rabau, E., David, A., Serr, D. M., Mashiach, S., Lunenfeld, B. Am. J. Obstet. Gynec. 1967, 98, 92. 2. Macdonald, R. R. J. Obstet. Gynœc. Br. Commonw. 1969, 76, 1090. 3. McSweeney, D. J., Sbarra, A. J. Obstet. Gynec., N.Y. 1965, 26, 201. 1.
509
doubtful whether any of these
is suffibe useful in ciently precise, although they may of treatment. A other forms novel monitoring observation concerning the relationship of oestrogens to neutrophil counts has been made by CRUICKSHANK et al.A positive correlation between oestrogen levels and neutrophil counts and a negative correlation with haemoglobin was noted in a group of 20 infertile women undergoing treatment with gonadotrophin.5 A regression equation was used to predict preovulatory oestrogen levels during treatment, and in the preliminary series accuracy was achieved in 70% of samples. Although this is encouraging, there was a fairly high standard deviation, so that in its present form the method cannot safely replace the determination of oestrogens. A number of shortened methods for oestrogen determinations have now been described. BROWN et al.6measure total oestrogens and are able to make results available within 31/2 hours of receiving the sample. Using this method they reported in a series of 222 courses of treatment a pregnancy from every 5-2 courses and clinical symptoms of overstimulation in only 7 cycles (3-2%). The oestrogens averaged about 200 ;jt.g. per 24 hours at the time of giving human chorionic gonadotrophin and 50-100 g. in the normal cycle; but it is hard to know what to accept as a safe level. Another important factor may be the rate of increase in oestrogens at the time of giving human chorionic gonadotrophin. BUTLER7 compared the oestrone excretion in the normal cycle with that in induced cycles and showed that many treatments produced rates of increase which were too high or too low. He found that the closest fit resulted from the administration of gonadotrophin in three injections on alternate days. BROWN et al. gave the less convenient, but more commonly used, daily injections. This produced a latent period of 3-4 days during which there was no change in oestrogen excretion, followed by progressive increases. A convenient criterion to accept for excessive stimulation was an oestrogen value exceeding 100 {jLg. per 24 hours within 5 days of starting treatment. The latest work is described on p. 482 by Professor ScoTT and his colleagues, who have attempted to devise a more exact method of monitoring treatment. They observed a linear relationship between the log of the 24-hour urinary oestrone excretion and time. They suggest that it may be possible to predict oestrone levels on subsequent days, so that the best time to give chorionic gonadotrophin may be pinpointed. Treatment could also be withheld if the steepness of the slope of the increasing oestrone was too great and threatened hyperstimulation. 4.
end-points
Cruickshank, J. M., Morris, R., Butt, W. R., Crooke, A. C. J. Obstet Gynœc. Br. Commonw. 1970, 77, 634. 5. Cruickshank, J. M. ibid. p. 644. 6. Brown, J. B., Evans, J. H., Adey, F. D., Taft, H. P., Townsend, L ibid. 1969, 76, 289. 7. Butler, J. K. Proc. R. Soc. Med. 1969, 62, 34.
CHILDREN IN ISOLATION UNITS LARGE numbers of children are admitted to isolation units every year. Some are admitted because they need hospital care and must be isolated because of an infective risk to other patients. Others are admitted to isolation units, not because they are seriously ill, but because they develop an infectious disease in circumstances which do not allow home care. Examples are children in residential accommodation and children from deprived and overcrowded households. The stress of isolation in hospital may be added to an already disturbed background of, for example, parental separation. Until lately, the only solutions to this problem on the hospital side were a policy of rapid discharge to make the child’s stay as short as possible, and genuinely unrestricted visiting. But although infectious-disease units have a better record than paediatric wards in encouraging unrestricted visiting, the same factors which first lead to admission often also mean poor visiting of the children in hospital. For similar reasons, rooming-in of the mother is rarely possible, even where facilities are offered.
For
some
years the Save the Children Fund has
supported the formation of playgroups in hospitals, and has helped to train playleaders to run them. The usual type of playgroup cannot be organised in an isolation ward because it is rarely possible to put the children into groups, but children in isolation may benefit from this kind of help no less, and possibly more, than those in ordinary wards. An experiment in applying the methods of play therapy to children in isolation was started three years ago at St. George’s Hospital, in Tooting, London, when, with the help of the Save the Children Fund, a playleader joined the staff of the infectious-disease unit. The result was a notable increase in happiness and activity in the wards; fears that the nursing staff would be left with all the painful tasks while the playleader took over all the pleasant relationships with the children proved groundless. On the contrary, the special skills and methods of the playleader have been disseminated through the wards, and have had a generally beneficial effect on the standard of child care. But the playleader’s task is much more difficult in an isolation unit than it is in an open ward. She can usually be with only one child at a time, and barrier-nursing precautions must often be used, so that the time available for each child is much less than in the conventional playgroup. An addition to the facilities in the infectious-disease unit at St. George’s has helped to reduce the isolation barrier further. This is the installation of an internal telephone system for the children’s use. Each cubicle is equipped with a socket; and the handsets, eight for the two wards, can be distributed to the cubicles occupied by children of suitable age and clinical condition. The handsets are of a conventional type, and with the aid of a very simple internal directory the children can phone other children in the ward, or the playleader, or nurse, and can be telephoned by them. The system was installed by an industrial company at cost price, and was paid for by the Friends of St. George’s Hospital. It has rapidly become very popular with the patients, and almost as popular with the medical and nursing staff.