3 MARCH 1979
BRITISH MEDICAL JOURNAL
619
in general practice, so that there are many cimetidine-treated patients in the community with undiagnosed or (if they come up to clinic during or shortly after a course of cimetidine) undiagnosable abdominal pain, some of whom may be harbouring palliated malignancy. It is necessary to emphasise, categorically, that it is in every patient's best interests to have a definitive diagnosis made before treatment with cimetidine is started. There is no doubt that the endoscopic and radiological facilities within the NHS are adequate to cope with the demand. A definitive diagnosis is all the more important since it is clear that prolonged, rather than short-term, cimetidine treatment is necessary for most of these patients with chronic duodenal ulcer disease who are suitable for medical treatment. N R PEDEN K G WORMSLEY
mastalgia rarely have the full premenstrual tension syndrome. We have reviewed 83 patients with typical cyclical mastalgia using a detailed questionnaire and only six had more than four of the typical premenstrual symptoms described by Dalton.' We would also agree that bromocriptine improves breast pain, but only cyclical pain. Although the authors do not comment on the pain pattern, their table appears to show that women with non-cyclical ("constant") pain respond as well as those with cyclical pain. However, in our much larger controlled study using bromocriptine in mastalgia,2 whcre the patients were divided into cyclical and noncyclical groups defined according to previously published patterns,; we have shown that only those women with cyclical mastalgia respond to the drug. A group of 11 women with noncyclical pain failed to show any response to bromocriptine, while the 29 patients in the Department of Pharmacology and Therapeutics, cyclical group responded in the manner Ninewells Hospital, described by Dr Blichert-Toft and his Dundee colleagues. We therefore advocate that patients Ritchie, J A, and Truelove, S C, British Medical with mastalgia should be divided into cyclical JoUrnal, 1979, 1, 376. and non-cyclical groups as only the former are Taylor, R H, et al, Lancet, 1978, 1, 686. Murray, C, et al, Lancet, 1978, 1, 1092. likely to respond to bromocriptine. Minoli, G, Terruzzi, V, and Rossini, A, Lancet, 1978,
works perfectly well. To quote a single example, a physician's 2-year-old child developed a febrile upper respiratory infection followed a week later by recognisable whooping, bacteriologically confirmed as whooping cough. A two weeks' course of erythromycin was given to this child and to the second child in the family, a 6-week-old baby, who would have had little chance of avoiding the disease. The baby snuffled for a day or two but otherwise remained in normal health. In about half the infants admitted to this department during the present whoopingcough epidemic there was a history of a prior case in an older child, so that by applying prophylactic treatment it might be possible to halve the existing low mortality from the disease-surely a very worthwhile aim. ALASTAIR G IRONSIDE Regional Department of Infectious Diseases, University of Manchester
Warin, J F, and Ironside, A G, Lecture Notes on the Infectiouis Diseases, 2nd edn. Oxford, Blackwells, 1975. Arneil, G C, and McAllister, T A, Practitioner, 1977, 219, 855.
1, 1092.
SIR,-As Dr Michael Drury (10 February, p 410) points out, the discussion about cimetidine by Dr J H Baron and others (20 January, p 169) has important implications. No one doubts that cimetidine presents a great advance in the treatment of duodenal ulcer and certain other disorders of the upper gastrointestinal tract, and the majority of patients with duodenal ulcer respond-at least in the short term -to such treatment. The correlation between ulcer healing and remission of symptoms is known to be poor. Misiewiczl has reviewed this problem and concludes, "In the absence of endoscopic evidence, the presence or absence of symptoms cannot be assumed to indicate with certainty the presence or the absence of a peptic ulcer." A large multicentre trial on the treatment of duodenal ulceration with cimetidine, published after Misiewicz's review, shows that 22",, of patients whose ulcers had been shown to have healed continued to have pain. Further, ulcer healing rates in response to placebo may vary from 20",, to 60",,.; Dr Drury suggests that a therapeutic trial, at least in younger patients with upper abdominal pain, should be carried out using cimetidine for one week as a first step in investigation. Many gastroenterologists would find it difficult to accept this proposition. I would suggest that, on the available evidence, it would be impossible to draw any firm diagnostic conclusions from such a therapeutic trial in any individual patient. K F R SCHILLER Department of Gastroenterology, St Peter's Hospital, Chertsev, Surrey
Misiewicz, J J, (,liinics i't (Gastroenterology, 1978, 7, 571. Gult, 1979, 20, 68. I,angman, M J S. C,linics in (Gastroenterology, 7, 583.
Treatment of mastalgia with bromocriptine
SIR,-We read with interest the short report on the use of bromocriptine in mastalgia by Dr M Blichert-Toft and others (27 January, p 237). We would agree that patients with
R E MANSEL J R WISBEY' Treatment of malaria L E HUGHES SIR,-I was interested in the article (10 University Department of Surgery, Welsh National School of Medicine, February, p 385) on malaria in Birmingham Cardiff and a London Teaching Hospital. I spent some Dalton, K, The Premenstrual Syndrome. London, ten years in Assam and Bangladesh and had an Heinemann, 1964. opportunity of seeing a great deal of malaria Mansel, R E, Preece, P E, and Hughes, L E, British caused by both Plasmodium vivax and P JIournal of Surgery, 1978, 65, 724. Preece, P E, et al, Lancet, 1976, 2, 670.
Whooping-cough vaccination SIR,-The recent controversy on whooping cough has centred entirely on the safety and efficacy of vaccines and has largely ignored the fact that by the time a child is old enough to be fully vaccinated he has little to fear from whooping cough. There is no dispute that the main brunt of the illness and practically all the mortality falls on infants in the first six months of life, so that even the revised "earlier age" schedules of vaccination fail to protect directly this vulnerable age group. It is argued that vaccination reduces the incidence of whooping cough in older siblings, thus indirectly protecting the baby. However, the infant mortality in the last "well vaccinated" whooping cough outbreak was higher than in the present very much larger "unvaccinated" outbreak and this does not suggest that indirect protection is very effective. The only other protective measure possible for the young infant is the use of antibiotic prophylaxis, and this of course depends on the prior occurrence of recognisable whooping cough in an older child in the family. During the present outbreak we have had the opportunity to confirm that erythromycin given for 14 days in a dose of 25 mg/kg/day is virtually fully effective in eradicating Bor-detella per tussis from the upper respiratory tract. Therefore the simple prophylactic procedure is to treat the older child or children and the vulnerable baby with erythromycin for 14 days to eradicate the infection from the family. This procedure has been recommended in the past by several authorities' and, although there are no controlled trials available, my own view from individual experiences is that it
falciparum. There are many interesting points raised by this article, but I want to raise only one. The authors indicate that their treatment for cerebral malaria was either intravenous quinine or chloroquine. In two of the cases thus treated there was a fatal outcome. When I first arrived in Bangladesh there was a strong fear regarding the use of intravenous quinine for cerebral malaria. I had a bitter personal experience of a sudden death resulting from intravenous quinine given for cerebral malaria. However, having referred to MansonBahr's Tropical Diseases, I eventually laid down the following regimen for use in our hospital in cases of cerebral malaria. After confirmation of P falciparum in the blood, the blood pressure and pulse were checked. The patient was then given 1 ml 1/1000 adrenaline over two minutes intramuscularly. A pause of five minutes was allowed for checking the pulse and blood pressure. Intravenous quinine dihydrochloride was given in a dose of 600 mg made up to 20 ml in distilled water and was instilled slowly over ten minutes. The usual result was the return to consciousness of the patient within five to ten minutes. It was then possible to continue oral medication. It is not widely appreciated that the giving of intravenous quinine produces a form of Jarisch-Hexheimer reaction similar to that caused in former days by arsenic given for tertiary cerebral syphilis. The parasites in the cerebral capillaries are rapidly lysed and produce widespread infarcts. If adrenaline is given first, however, the parasites are driven out of the peripheral capillaries by the vasoconstrictive effect of the adrenaline, and the spleen also contracts, driving more of the parasites into the general circulation. If one gives the adrenaline five minutes to work then the peripheral capillaries are largely clear of the parasites and the quinine is allowed to
620
BRITISH MEDICAL JOURNAL
act on the parasites in the general circulation. In addition the adrenaline greatly assists the patient's own peripheral circulation because he is suffering in cerebral malaria from hypovolaemia, and this is usually indicated by a very low blood pressure and a very poor pulse volume. A dose in the region of 600 mg, depending on the weight of the patient, is usually sufficient to lower the general parasitaemia and allow the patient to recover consciousness. One can then continue with oral doses of chloroquine by mouth to complete the treatment. In over 50 cases over some ten years on this regimen we never had a death, and recoveries were as dramatic as those obtained in attacks of left ventricular failure by the use of intravenous frusemide. Some authorities suggest steroids as being useful in cerebral malaria. However, steroids do not have the powerful peripheral vasoconstrictive effect produced by adrenaline, which is now much underused and undervalued in emergencies. I T PATRICK
We would be most grateful if surgeons working in communities with high and with low prevalences of appendicitis would palpate the appendix during abdominal operations, record the frequency with which faecal particles can be detected, and report their findings to us. Such findings could make a significant contribution to the understanding of the pathogenesis of this disease and such assistance will be acknowledged in any publication of accumulated experience.
Aylesbury, Buckinghamshire
A "personalised" human milk bank
Aetiology of appendicitis
SIR,-The geographical distribution of appendicitis indicates that it is a disease associated with modern Western culture.' The pathological features suggest that obstruction of the appendix lumen precedes inflammatory change.3 In only a minority of cases can a faecalith be demonstrated as the cause of the obstruction, but it has been postulated that in others the cause may be exaggerated muscular contraction occasioned by the presence of solid faecal matter in the lumen of the appendix. If such muscle contraction associated with viscid small-volume faeces can block the wide lumen of the pelvic colon,4 it is not difficult to appreciate that it might much more easily do the same to the much narrower lumen of the appendix. There is some evidence that in communities with low prevalences of appendicitis solid faecal matter is rarely present in the appendix lumen (Rangabashram, personal communication). If the frequency of solid appendicular content could be related to the frequency of appendicitis in various communities a direct relationship could endorse and the absence of relationship would discount the hypothesis that solid appendicular content contributed to the causation of the disease. One of us (ASM), and some of his colleagues, palpated the appendix during 55 abdominal operations in which the appendix was removed as an incidental procedure and subsequently opened to determine the reliability of external palpation as a means of ascertaining the presence or absence of solid faecal material in the appendix. In 32 cases solid faecal particles were neither detected by palpation during operation nor discovered subsequently in opening the organ. In 22 faecal particles were both detected by palpation and their presence subsequently confirmed. In one case a thin faecalith was found on opening the appendix though it was not palpable at operation. In no case was there evidence of solid faecal particles on palpation that was not subsequently confirmed on opening the appendix. Palpation would therefore appear to be a reliable method of determining the presence of solid faeces within the appendix. In this series solid faecal particles were palpated and present in 40", of appendices examined.
D P BURKITT A S MOOLGAOKAR F I TOVEY Unit of Geographical Pathology, St Thomas's Hospital Medical School, London SEI
Burkitt, I) I', British Jouirnal of Surgery,, 1971, 58, 695. Horton, I W L, British Medical journal, 1978, 2, 1672. :'Johnson, R H, British Medical Jouirnal, 1978, 2, 590. Painter, N S, Diverticutlar D)isease of the Colon. London, William Heinemann, London, 1975.
3 MARCH 1979
does not correlate with peak plasma drug concentration.2 Furthermore, there is evidence from pharmacokinetic studies2 that diuresis is most closely correlated with the concentration of drug in a peripheral compartment. If diuretic response relates to peak plasma drug concentration it is very difficult to accourt for the efficacy of low-dosage infusion in refractory oedema, where plasma drug concentrations are ostensibly subtherapeutic' and are certainly much lower than the levels achieved by intravenous bolus dosing. Finally, there is published evidence with frusemide and bumetanide that dividing the dosage enhances the total diuretic effect'5; possibly by interrupting homoeostatic recovery. We would therefore suggest that diuresis does not relate to peak plasma drug concentration, and that there is insufficient evidence for the statement that twice daily administration is less effective than the full dosage given once. HENRY L ELLIOTT J R LAWRENCE
SIR,-Soon 50",, of medical students will be women. It is reasonable to assume that the majority of women doctors will have children, and many-will want to return to work while still breast-feeding. We describe a "personalised" human milk bank that enables a working woman doctor to continue breastfeeding. Using this system while working halftime every day, we have both reared our babies until five months old on a diet of only human milk. The equipment is a hand breast-pump, a nursery sterilising tank, plastic bottles, and a deep freeze. About 75 ml/day of milk can be obtained easily either by expressing each breast before the morning feed or by applying the pump to one breast during the baby's first feed of the day. A milk bank can thus be deep frozen during the weeks of maternity leave and at weekends when one is back at work. The same procedure for milk collection continues when the mother is working, but each day's fresh milk is supplemented by some of the frozen milk. The combined milk is given to the baby during the mother's absence at work. We have found this system easy to operate and have encountered no problems. We strongly recommend it to women doctors, who must be sensitive to the writings of the Jelliffes' yet want to return to work before the introduction of cows' milk is desirable.
Department of Matcria Medica, Stobhill General Hospital,
(;lasg(W
Burg, M, et al, AmneriCan journal of Physiology, 1973, 225, 119. Branch, R A, et al, Blritish joturnal of C(linical Pharmlacology. 1977, 4, 121. I.awrence, J R, et a1l, C(linical Pharmacologi!y and TherapetiiCs, 1978, 23, 558. lawson, D H, et,al, British Medical 7ouirtalo, 1978, 2, 476. Wilson, T W, et al, C(linical Pharmiacologv antd Therapeuitics, 1975, 18, 165. Hunter, K R, and Underwood, P N, Postgradiute Medical7ournal. 1975, 51, 91.
Does prazosin induce formation of antinuclear factor? SIR,-Prazosin, a widely used antihypertensive drug, has never, to my knowledge, been implicated in the development of a positive antinuclear factor (ANF). It was therefore surprising to read the observations of Dr A J Marshall and others (20 January, p 165) describing, for the first time, a high incidence of positive ANF in patients treated with prazosin. My experience with this drug dates from early 1974 and covers treatment of a total of 85 patients. In thc first group of 42 patients ANF tests were done before treatment and checked after long-term treatment. Prazosin was given either as the sole
antihypertensive drug or in combination with a beta-blocker. Maintenarnce doses of prazosin ranged from 1-5 mg to 6 mg daily (average 3 2 mg). JOANNA GAZZARD In addition to the initial ANF tests, titres were measured again after treatment with prazosin London SW4 varying from 3 months to 36 months (average 19 CHRISTINE LEE months). In all, 113 tests were carried out in these patients. ANF was measured by a routinc indirect London SW15 immunofluorescence technique using frozen ' Jelliffe, D B, and Jelliffe, E F P, New England Journal sections of monkey kidney and stomach and of Medicine, 1977, 297, 912. FITC-labelled anti-IgG antibodies. The pretreatment test was negative (titre below 8) in all patients cxcept two, in whom titres were 1,16 and /8 respectively. On final follow-up, Management of refractory oedema ANF tests were negative in all patients. The patient with a titre of 16 six months later had SIR,-The leading article on "Management of a titre pretreatment of 118. After 12 months procainamide was refractory oedema" (20 January, p 148) states started owing to ventricular extrasystoles; it was, that diuretic effect relates to peak plasma drug however, withdrawn two months later owing to the concentration, from which it is adduced that development of troublesome joint pains. The ANF any dose given once daily will be more titre was 1 32 when measured one month later. effective than half dosage given twice daily. Prazosin treatment was not interrupted and the This is an area of controversy and in fact there final ANF test, after three years on a maintenance dose of 6 mg daily, was negative. is evidence for an opposing point of view.
It is generally accepted that diuresis depends
Thus in none of the 42 patients did a conversion from negative to positive ANF lumen' and there is evidence that natriuresis test occur which was related to long-term
on the drug concentration within the tubular
BRITISH MEDICAL JOURNAL
619
in general practice, so that there are many cimetidine-treated patients in the community with undiagnosed or (if they come up to clinic during or shortly after a course of cimetidine) undiagnosable abdominal pain, some of whom may be harbouring palliated malignancy. It is necessary to emphasise, categorically, that it is in every patient's best interests to have a definitive diagnosis made before treatment with cimetidine is started. There is no doubt that the endoscopic and radiological facilities within the NHS are adequate to cope with the demand. A definitive diagnosis is all the more important since it is clear that prolonged, rather than short-term, cimetidine treatment is necessary for most of these patients with chronic duodenal ulcer disease who are suitable for medical treatment. N R PEDEN K G WORMSLEY
mastalgia rarely have the full premenstrual tension syndrome. We have reviewed 83 patients with typical cyclical mastalgia using a detailed questionnaire and only six had more than four of the typical premenstrual symptoms described by Dalton.' We would also agree that bromocriptine improves breast pain, but only cyclical pain. Although the authors do not comment on the pain pattern, their table appears to show that women with non-cyclical ("constant") pain respond as well as those with cyclical pain. However, in our much larger controlled study using bromocriptine in mastalgia,2 whcre the patients were divided into cyclical and noncyclical groups defined according to previously published patterns,; we have shown that only those women with cyclical mastalgia respond to the drug. A group of 11 women with noncyclical pain failed to show any response to bromocriptine, while the 29 patients in the Department of Pharmacology and Therapeutics, cyclical group responded in the manner Ninewells Hospital, described by Dr Blichert-Toft and his Dundee colleagues. We therefore advocate that patients Ritchie, J A, and Truelove, S C, British Medical with mastalgia should be divided into cyclical JoUrnal, 1979, 1, 376. and non-cyclical groups as only the former are Taylor, R H, et al, Lancet, 1978, 1, 686. Murray, C, et al, Lancet, 1978, 1, 1092. likely to respond to bromocriptine. Minoli, G, Terruzzi, V, and Rossini, A, Lancet, 1978,
works perfectly well. To quote a single example, a physician's 2-year-old child developed a febrile upper respiratory infection followed a week later by recognisable whooping, bacteriologically confirmed as whooping cough. A two weeks' course of erythromycin was given to this child and to the second child in the family, a 6-week-old baby, who would have had little chance of avoiding the disease. The baby snuffled for a day or two but otherwise remained in normal health. In about half the infants admitted to this department during the present whoopingcough epidemic there was a history of a prior case in an older child, so that by applying prophylactic treatment it might be possible to halve the existing low mortality from the disease-surely a very worthwhile aim. ALASTAIR G IRONSIDE Regional Department of Infectious Diseases, University of Manchester
Warin, J F, and Ironside, A G, Lecture Notes on the Infectiouis Diseases, 2nd edn. Oxford, Blackwells, 1975. Arneil, G C, and McAllister, T A, Practitioner, 1977, 219, 855.
1, 1092.
SIR,-As Dr Michael Drury (10 February, p 410) points out, the discussion about cimetidine by Dr J H Baron and others (20 January, p 169) has important implications. No one doubts that cimetidine presents a great advance in the treatment of duodenal ulcer and certain other disorders of the upper gastrointestinal tract, and the majority of patients with duodenal ulcer respond-at least in the short term -to such treatment. The correlation between ulcer healing and remission of symptoms is known to be poor. Misiewiczl has reviewed this problem and concludes, "In the absence of endoscopic evidence, the presence or absence of symptoms cannot be assumed to indicate with certainty the presence or the absence of a peptic ulcer." A large multicentre trial on the treatment of duodenal ulceration with cimetidine, published after Misiewicz's review, shows that 22",, of patients whose ulcers had been shown to have healed continued to have pain. Further, ulcer healing rates in response to placebo may vary from 20",, to 60",,.; Dr Drury suggests that a therapeutic trial, at least in younger patients with upper abdominal pain, should be carried out using cimetidine for one week as a first step in investigation. Many gastroenterologists would find it difficult to accept this proposition. I would suggest that, on the available evidence, it would be impossible to draw any firm diagnostic conclusions from such a therapeutic trial in any individual patient. K F R SCHILLER Department of Gastroenterology, St Peter's Hospital, Chertsev, Surrey
Misiewicz, J J, (,liinics i't (Gastroenterology, 1978, 7, 571. Gult, 1979, 20, 68. I,angman, M J S. C,linics in (Gastroenterology, 7, 583.
Treatment of mastalgia with bromocriptine
SIR,-We read with interest the short report on the use of bromocriptine in mastalgia by Dr M Blichert-Toft and others (27 January, p 237). We would agree that patients with
R E MANSEL J R WISBEY' Treatment of malaria L E HUGHES SIR,-I was interested in the article (10 University Department of Surgery, Welsh National School of Medicine, February, p 385) on malaria in Birmingham Cardiff and a London Teaching Hospital. I spent some Dalton, K, The Premenstrual Syndrome. London, ten years in Assam and Bangladesh and had an Heinemann, 1964. opportunity of seeing a great deal of malaria Mansel, R E, Preece, P E, and Hughes, L E, British caused by both Plasmodium vivax and P JIournal of Surgery, 1978, 65, 724. Preece, P E, et al, Lancet, 1976, 2, 670.
Whooping-cough vaccination SIR,-The recent controversy on whooping cough has centred entirely on the safety and efficacy of vaccines and has largely ignored the fact that by the time a child is old enough to be fully vaccinated he has little to fear from whooping cough. There is no dispute that the main brunt of the illness and practically all the mortality falls on infants in the first six months of life, so that even the revised "earlier age" schedules of vaccination fail to protect directly this vulnerable age group. It is argued that vaccination reduces the incidence of whooping cough in older siblings, thus indirectly protecting the baby. However, the infant mortality in the last "well vaccinated" whooping cough outbreak was higher than in the present very much larger "unvaccinated" outbreak and this does not suggest that indirect protection is very effective. The only other protective measure possible for the young infant is the use of antibiotic prophylaxis, and this of course depends on the prior occurrence of recognisable whooping cough in an older child in the family. During the present outbreak we have had the opportunity to confirm that erythromycin given for 14 days in a dose of 25 mg/kg/day is virtually fully effective in eradicating Bor-detella per tussis from the upper respiratory tract. Therefore the simple prophylactic procedure is to treat the older child or children and the vulnerable baby with erythromycin for 14 days to eradicate the infection from the family. This procedure has been recommended in the past by several authorities' and, although there are no controlled trials available, my own view from individual experiences is that it
falciparum. There are many interesting points raised by this article, but I want to raise only one. The authors indicate that their treatment for cerebral malaria was either intravenous quinine or chloroquine. In two of the cases thus treated there was a fatal outcome. When I first arrived in Bangladesh there was a strong fear regarding the use of intravenous quinine for cerebral malaria. I had a bitter personal experience of a sudden death resulting from intravenous quinine given for cerebral malaria. However, having referred to MansonBahr's Tropical Diseases, I eventually laid down the following regimen for use in our hospital in cases of cerebral malaria. After confirmation of P falciparum in the blood, the blood pressure and pulse were checked. The patient was then given 1 ml 1/1000 adrenaline over two minutes intramuscularly. A pause of five minutes was allowed for checking the pulse and blood pressure. Intravenous quinine dihydrochloride was given in a dose of 600 mg made up to 20 ml in distilled water and was instilled slowly over ten minutes. The usual result was the return to consciousness of the patient within five to ten minutes. It was then possible to continue oral medication. It is not widely appreciated that the giving of intravenous quinine produces a form of Jarisch-Hexheimer reaction similar to that caused in former days by arsenic given for tertiary cerebral syphilis. The parasites in the cerebral capillaries are rapidly lysed and produce widespread infarcts. If adrenaline is given first, however, the parasites are driven out of the peripheral capillaries by the vasoconstrictive effect of the adrenaline, and the spleen also contracts, driving more of the parasites into the general circulation. If one gives the adrenaline five minutes to work then the peripheral capillaries are largely clear of the parasites and the quinine is allowed to
620
BRITISH MEDICAL JOURNAL
act on the parasites in the general circulation. In addition the adrenaline greatly assists the patient's own peripheral circulation because he is suffering in cerebral malaria from hypovolaemia, and this is usually indicated by a very low blood pressure and a very poor pulse volume. A dose in the region of 600 mg, depending on the weight of the patient, is usually sufficient to lower the general parasitaemia and allow the patient to recover consciousness. One can then continue with oral doses of chloroquine by mouth to complete the treatment. In over 50 cases over some ten years on this regimen we never had a death, and recoveries were as dramatic as those obtained in attacks of left ventricular failure by the use of intravenous frusemide. Some authorities suggest steroids as being useful in cerebral malaria. However, steroids do not have the powerful peripheral vasoconstrictive effect produced by adrenaline, which is now much underused and undervalued in emergencies. I T PATRICK
We would be most grateful if surgeons working in communities with high and with low prevalences of appendicitis would palpate the appendix during abdominal operations, record the frequency with which faecal particles can be detected, and report their findings to us. Such findings could make a significant contribution to the understanding of the pathogenesis of this disease and such assistance will be acknowledged in any publication of accumulated experience.
Aylesbury, Buckinghamshire
A "personalised" human milk bank
Aetiology of appendicitis
SIR,-The geographical distribution of appendicitis indicates that it is a disease associated with modern Western culture.' The pathological features suggest that obstruction of the appendix lumen precedes inflammatory change.3 In only a minority of cases can a faecalith be demonstrated as the cause of the obstruction, but it has been postulated that in others the cause may be exaggerated muscular contraction occasioned by the presence of solid faecal matter in the lumen of the appendix. If such muscle contraction associated with viscid small-volume faeces can block the wide lumen of the pelvic colon,4 it is not difficult to appreciate that it might much more easily do the same to the much narrower lumen of the appendix. There is some evidence that in communities with low prevalences of appendicitis solid faecal matter is rarely present in the appendix lumen (Rangabashram, personal communication). If the frequency of solid appendicular content could be related to the frequency of appendicitis in various communities a direct relationship could endorse and the absence of relationship would discount the hypothesis that solid appendicular content contributed to the causation of the disease. One of us (ASM), and some of his colleagues, palpated the appendix during 55 abdominal operations in which the appendix was removed as an incidental procedure and subsequently opened to determine the reliability of external palpation as a means of ascertaining the presence or absence of solid faecal material in the appendix. In 32 cases solid faecal particles were neither detected by palpation during operation nor discovered subsequently in opening the organ. In 22 faecal particles were both detected by palpation and their presence subsequently confirmed. In one case a thin faecalith was found on opening the appendix though it was not palpable at operation. In no case was there evidence of solid faecal particles on palpation that was not subsequently confirmed on opening the appendix. Palpation would therefore appear to be a reliable method of determining the presence of solid faeces within the appendix. In this series solid faecal particles were palpated and present in 40", of appendices examined.
D P BURKITT A S MOOLGAOKAR F I TOVEY Unit of Geographical Pathology, St Thomas's Hospital Medical School, London SEI
Burkitt, I) I', British Jouirnal of Surgery,, 1971, 58, 695. Horton, I W L, British Medical journal, 1978, 2, 1672. :'Johnson, R H, British Medical Jouirnal, 1978, 2, 590. Painter, N S, Diverticutlar D)isease of the Colon. London, William Heinemann, London, 1975.
3 MARCH 1979
does not correlate with peak plasma drug concentration.2 Furthermore, there is evidence from pharmacokinetic studies2 that diuresis is most closely correlated with the concentration of drug in a peripheral compartment. If diuretic response relates to peak plasma drug concentration it is very difficult to accourt for the efficacy of low-dosage infusion in refractory oedema, where plasma drug concentrations are ostensibly subtherapeutic' and are certainly much lower than the levels achieved by intravenous bolus dosing. Finally, there is published evidence with frusemide and bumetanide that dividing the dosage enhances the total diuretic effect'5; possibly by interrupting homoeostatic recovery. We would therefore suggest that diuresis does not relate to peak plasma drug concentration, and that there is insufficient evidence for the statement that twice daily administration is less effective than the full dosage given once. HENRY L ELLIOTT J R LAWRENCE
SIR,-Soon 50",, of medical students will be women. It is reasonable to assume that the majority of women doctors will have children, and many-will want to return to work while still breast-feeding. We describe a "personalised" human milk bank that enables a working woman doctor to continue breastfeeding. Using this system while working halftime every day, we have both reared our babies until five months old on a diet of only human milk. The equipment is a hand breast-pump, a nursery sterilising tank, plastic bottles, and a deep freeze. About 75 ml/day of milk can be obtained easily either by expressing each breast before the morning feed or by applying the pump to one breast during the baby's first feed of the day. A milk bank can thus be deep frozen during the weeks of maternity leave and at weekends when one is back at work. The same procedure for milk collection continues when the mother is working, but each day's fresh milk is supplemented by some of the frozen milk. The combined milk is given to the baby during the mother's absence at work. We have found this system easy to operate and have encountered no problems. We strongly recommend it to women doctors, who must be sensitive to the writings of the Jelliffes' yet want to return to work before the introduction of cows' milk is desirable.
Department of Matcria Medica, Stobhill General Hospital,
(;lasg(W
Burg, M, et al, AmneriCan journal of Physiology, 1973, 225, 119. Branch, R A, et al, Blritish joturnal of C(linical Pharmlacology. 1977, 4, 121. I.awrence, J R, et a1l, C(linical Pharmacologi!y and TherapetiiCs, 1978, 23, 558. lawson, D H, et,al, British Medical 7ouirtalo, 1978, 2, 476. Wilson, T W, et al, C(linical Pharmiacologv antd Therapeuitics, 1975, 18, 165. Hunter, K R, and Underwood, P N, Postgradiute Medical7ournal. 1975, 51, 91.
Does prazosin induce formation of antinuclear factor? SIR,-Prazosin, a widely used antihypertensive drug, has never, to my knowledge, been implicated in the development of a positive antinuclear factor (ANF). It was therefore surprising to read the observations of Dr A J Marshall and others (20 January, p 165) describing, for the first time, a high incidence of positive ANF in patients treated with prazosin. My experience with this drug dates from early 1974 and covers treatment of a total of 85 patients. In thc first group of 42 patients ANF tests were done before treatment and checked after long-term treatment. Prazosin was given either as the sole
antihypertensive drug or in combination with a beta-blocker. Maintenarnce doses of prazosin ranged from 1-5 mg to 6 mg daily (average 3 2 mg). JOANNA GAZZARD In addition to the initial ANF tests, titres were measured again after treatment with prazosin London SW4 varying from 3 months to 36 months (average 19 CHRISTINE LEE months). In all, 113 tests were carried out in these patients. ANF was measured by a routinc indirect London SW15 immunofluorescence technique using frozen ' Jelliffe, D B, and Jelliffe, E F P, New England Journal sections of monkey kidney and stomach and of Medicine, 1977, 297, 912. FITC-labelled anti-IgG antibodies. The pretreatment test was negative (titre below 8) in all patients cxcept two, in whom titres were 1,16 and /8 respectively. On final follow-up, Management of refractory oedema ANF tests were negative in all patients. The patient with a titre of 16 six months later had SIR,-The leading article on "Management of a titre pretreatment of 118. After 12 months procainamide was refractory oedema" (20 January, p 148) states started owing to ventricular extrasystoles; it was, that diuretic effect relates to peak plasma drug however, withdrawn two months later owing to the concentration, from which it is adduced that development of troublesome joint pains. The ANF any dose given once daily will be more titre was 1 32 when measured one month later. effective than half dosage given twice daily. Prazosin treatment was not interrupted and the This is an area of controversy and in fact there final ANF test, after three years on a maintenance dose of 6 mg daily, was negative. is evidence for an opposing point of view.
It is generally accepted that diuresis depends
Thus in none of the 42 patients did a conversion from negative to positive ANF lumen' and there is evidence that natriuresis test occur which was related to long-term
on the drug concentration within the tubular
