BRITISH MEDICAL JOURNAL
12 FEBRUARY 1977
seemed somewhat unnecessary when one was primarily attempting to determine the incidence of diarrhoea in a large number of patients. Both trials showed the incidence of diarrhoea to be halved in the patients receiving talampicillin and these results were shown to be statistically significant (P < 0-02 in both studies). Dr Grant is perfectly entitled to express his personal opinion of these antibiotics, but other clinicians appear to have come to a different conclusion and amoxycillin is now very widely used in both hospital and general practice in this country and overseas. Finally, I am sure that Dr Grant would be the first to appreciate that he is getting better value for money when he provides drugs that are well absorbed rather than those that, for the most part, go down the drain. E T KNUDSEN Medical Director, Beecham Pharmaceuticals Division
443
between low levels of human placental lactogen and poor outcome of pregnancy but also that, in their practice, use of the assay in high-risk pregnancies can lead to a significant reduction in the antepartum death rate. We have been unable to trace comparable evidence relating to other methods of fetal risk prediction. Urinary oestrogen assay and serial ultrasound cephalometry have been widely used in Britain during recent years and the table printed below documents their increased use in one geographically defined population. Recent research2-4 has raised questions concerning the value of these techniques, but these studies have failed to provide any clear answers. Chards has drawn attention to the fact that "much effort is spent on the identification of new parameters of feto-placental function when the main requirement is for detailed evaluation and refinement of tests which are already available." However, some centres have already replaced urinary oestrogen assays with plasma oestriol measurements and introduced other biochemical and physical methods of antepartum fetal monitoring. Spellacy's example with human placental lactogen should be followed by using randomised controlled trials to assess the value of the increasing variety of methods in use. Who knows, fetal movement counts by the mother6 may prove as effective as other competitors and an even more convenient and economically attractive proposition than your recommended "best buy" of urinary oestrogen assay at 25p a time.
Brentford, Middx Nakazawa, S, Amoxycillin (BRL 2333) International Symposium, London, September 1973, II, 24. 2 Robinson, G N, Nouvelle Presse Medicale, 1975, 4, 2461. 3 Comber, K R, Osborne, Christine D, and Sutherland, R, Antimicrobial Agents and Chemotherapy, 1975, 7, 179. May, J R, and Ingold, A, British Journal of Diseases of the Chest, 1972, 66, 185. Molla, A L, Practitioner, 1974, 212, 123. 6 Pines, A, et al, Chemotherapy, 1977, 23, 58. 7 Ingold, A, British Journal of Diseases of the Chest, 1975, 69, 211. Burgi, H, Chemotherapy, 1973, 18, suppl, p 19. 9 Burns, M W, and Devitt, L, Journal of Infectious Diseases, 1974, 129, S 194. 10 Khan, A A, Journal of International Medical Research, 1975, 3, 230. Scragg, J N, and Rubidge, C J, American Journal of Tropical Medicine and Hygiene, 1975, 24, 860. 12 Scragg, J N, British Medical Journal, 1976, 2, 1031. We are grateful to Professor K T Evans and 3 Afifi, A M, Adnan, M, and Garf, A A, British Medical Dr Huw Gravelle, Department of Radiology, Journal, 1976, 2, 1033. 14 Leigh, D A, et al, British Medical Journal. 1976, 1, Welsh National School of Medicine, for access to 1378. ultrasonogram records. 15 Middleton, R S W, Clinical Trials Journal, 1976, 13, ANNE C DAVIES 23. Final-year medical student 6Jaffe, G, Murphy, J E, and Robinson, 0 P W, Practitioner, 1976, 216, 455. 17Knudson, E T, and Harding, J W, British Journal of IAIN CHALMERS Clinical Practice, 1975, 29, 255. Welsh National School of Medicine,
abnormality. Professor C F George and Dr C M Castleden (1 January, p 47) consider the various other factors which might explain the results of Dr Schneider and his colleagues. As mentioned above, we do not dispute that propranolol is virtually completely absorbed from the gastrointestinal tract and, as indicated in our previous letter, we still consider that altered rates of absorption from the gut are a likely explanation of the results. However, we are sceptical whether accelerated gastric emptying really does occur in coeliac disease as has been previously reported.5 While we accept that a reduction in the first pass metabolism of the drug as a result of liver dysfunction is a possible mechanism in Crohn's disease, in which minor histological abnormality of the liver is a relatively frequent DIANA RIAD FAHMY occurrence,6 this seems unlikely in coeliac
Cardiff
Predicting fetal death SIR,-The antepartum monitoring techniques considered in your recent leading article (15 January, p 123) may well be more or less precise predictors of fetal death, but they were introduced with a view to preventing death, not predicting it. This may seem a statement of the obvious, but the research surrounding these physical and biochemical techniques suggests that there may be some confusion of these two roles. The fact that retrospective studies have demonstrated reasonable correlation between the outcome of pregnancy and measurements made with these techniques does not establish that their use will prevent fetal death. You refer to the only study of which we are aware which was specifically designed to assess the extent to which use of a fetal risk predictor may alter the natural history of a pregnancy.' Spellacy and his colleagues have certainly provided encouraging evidence not only of a reasonable correlation
Supraregional Assay Service, Tenovus Institute of Cancer Research, Cardiff. lSpellacy, W N, Buhi, W C, and Birk, S A, American Journal of Obstetrics and Gynecology, 1975, 121, 835. 2Chalmers, I, et al, British Medical Journal, 1976, 1, 735. 3Chalmers, I, Lawson, J G, and Turnbull, A C, British Journal of Obstetrics and Gynaecology, 1976, 83, 921. 4 Chalmers, I, Lawson, J G, and Turnbull, A C, British Journal of Obstetrics and Gynaecology, 1976, 83, 930. 5 Chard, T, Journal of Clinical Pathology, 1976, 29, suppl 10, p 18. 6 Pearson, J F, and Weaver, J B, British Medical Journal, 1976, 1, 1305.
Plasma propranolol levels in Crohn's disease and coeliac disease SIR,-In reply to Dr R E Schneider and others (27 November, p 1324) it is not disputed that propranolol might eventually be fully absorbed in the patients or in the controls in their
Frequency and intensity of antepartum monitoring among parturients resident in Cardiff, 1967-74
Referred for oestrogen assay () Mean number of assays per case referred Referred for diagnostic ultrasound* during pregnancy () Mean number of ultrasound examinations per case referred *All indications.
1967
1968
1969
1970
1971
1972
1973
1974
-
30
3-2
90
19-4
19-7
17-6
16-4
2-6
2-8
2-5
2-9
2-8
3-2
3-6
-
investigation (2 October, p 794). The problem is why the plasma levels of propranolol were higher over 1-6 h after ingestion in patients with Crohn's disease and at 1 h after ingestion in treated coeliac patients than in healthy volunteers. However, an explanation for their results has not been provided, nor has our proposal (6 November, p 1135) of the possible role of the jejunal acid microclimate in regulation of propranolol absorption from the proximal jejunum been disproved. Higher propranolol levels in rheumatoid arthritis and pneumonia' do not conflict with the jejunal acid microclimate hypothesis. The occurrence of folic acid deficiency in rheumatoid arthritis2-4 provides some indirect evidence which suggests that the proximal jejunal acid microclimate could be deficient in these nongastrointestinal disorders. Our own unpublished observations on seven patients with folic acid deficiency and normal jejunal histology have shown that the proximal jejunal luminal surface pH in these patients was significantly less acid than in normal controls. Therefore it is possible that the acid microclimate could be deficient in rheumatoid arthritis in spite of no obvious intestinal
07 10
101
14-3
200
210
2-1
2-3
2-5
2-2
disease, in which liver dysfunction is rare. In our own series of 300 adult patients with coeliac disease only three cases of chronic liver disease have been seen. We were disappointed that Dr R L Parsons and his colleagues (8 January, p 103) have found that two drugs with a low pKa, aspirin and indomethacin, are absorbed normally or even faster than normal in coeliac disease. Unfortunately it is not clear from their letter whether their coeliac patients were on normal or gluten-free diets. In a previous paper,7 referred to in their letter, they found that plasma propranolol levels after ingestion were higher in coeliacs, all of whom were on a gluten-free diet, when compared with normals, and no details of jejunal morphology at the time of study were provided. Obviously in deciding the role of the jejunum and its acid microclimate in the absorption of drugs it is important to have details of the jejunal morphology at the time of study. Nevertheless, it is strange that, whereas two drugs with a low pKa appear to be normally absorbed in coeliac disease, folic acid, which is probably 50°O unionised at pH 3.58 (it has two carboxyl groups and therefore cannot have one pKa value), is malabsorbed in untreated coeliac disease.9-'1 The jejunal acid microclimate must affect absorption of drugs which are weak acids, such as folic acid, aspirin, and indomethacin, or weak bases, such as propranolol, and so its deficiency in untreated coeliac
444
BRITISH MEDICAL JOURNAL
disease12 and in Crohn's disease'3 should have some effect on absorption of these drugs. However, there are so many other factors involved in drug absorption and drug plasma levels that these other factors may obscure or alter effects that changes in the jejunal acid microclimate may produce. B T COOPER W TREVOR COOKE Nutritional and Intestinal Unit, General Hospital, Birmingham
from that in Crohn's disease. Here the difference between the plasma propranolol levels in patients and controls is highly significant at all sampling times-apart from the '-hour oneas are the C.ax values as well as the areas under the curve. We should like to emphasise once more that we do not necessarily equate this with enhanced absorption of the drug in Crohn's and other diseases.2 Other factors are much more likely to be involved.
12 FEBRUARY 1977
injectable form of diazepam was positive on two separate occasions. I think there was little doubt of the diagnosis in that case. The case we described was an example of an epidermal (eczematous) reaction to a systemically administered drug. J S COMAISH Department of Dermatology, Royal Victoria Infirmary, Newcastle upon Tyne I
Felix, R H, and Comaish, J S, Lancet, 1974, 1, 1017.
M L LUCAS We should like to thank Dr J A H Waterhouse JOHN A BLAIR from Cancer Registry, Queen Elizabeth Medical Atrial fibrillation in the elderly Department of Chemistry, University of Aston in Birmingham, Birmingham
Centre, for checking our statistical calculations.
Babb, J, et al, Lancet, 1976, 1, 1413. 2Gough, K R, et al, British Medical Journal, 1964, 1, 212. 3 Deller, D J, et al, British Medical3Journal, 1966, 1, 765. ' Ratanashien, K, et al, Journal of Clinical Pathology. In press. 5Mobert, S, and Carlberger, F, Scandinavian Journal of Gastroenterology, 1974, 9, 17. Eade, M, et al, Scandinavian Journal of Gastroenterology, 1971, 6, 199. 7Parsons, R L, et al, Gut, 1976, 17, 139. Blair, J A, and Matty, A J, Clinics in Gastroenterology, 1974, 3, 183. 9 Hepner, G W, et al, Lancet, 1968, 2, 302. 0 Gerson, C D, et al, American Journal of Digestive Diseases, 1974, 19, 911. Mackenzie, J F, and Russell, R I, Clinical Science and Molecular Medicine, 1976, 51, 363. Lucas, M L, et al. Submitted for publication. s Cooper, B T, et al. Submitted for publication.
Department of Medicine, University of Birmingham
I
SIR,-We were interested in the comments by Dr R L Parsons and other (8 January, p 103) on our work on plasma propranolol levels in patients with coeliac disease (2 Ocotber, p 794). We cannot accept their claim to have shown a statistically significant rise in these levels at 1, 6, and 8 h after the oral administration of 40 mg of the drug. Using Student's t test, as they did, recalculation of the t values for the difference between the plasma propranolol levels in controls and coeliac patients from their own data' gave values of 1-8, 1-73, and 1-3 respectively. The corresponding P values were therefore on no occasion below or even near the 0-05 level of significance. It also seems strange that the difference between the areas under the curve for plasma propranolol levels in controls and coeliac patients should be significant when calculated according to one programme (Wagner and Nelson), but not significant when calculated according to another (Saunders and Natumen). In our patients the area under the curve in coeliacs was not significantly raised when using Simpson's rule. The claim that the discrepancy between their and our results is due to the difference in the numbers of patients used in the two groups as well as to a difference in the duration of treatment on a gluten-free diet also is not acceptable. Although Dr Parsons and his colleagues' allegedly studied 14 patients compared with our eight, they gave data on only 13 and for their calculations indeed only use 11 of these. This reduces the difference between their and our series to only three patients instead of five as claimed in their letter. Although one can only guess which patients were included it does seem likely that the average duration of treatment in their series was shorter than in ours. This would suggest that duration of treatment does not significantly affect plasma levels of propranolol in coeliac disease as the data of Dr Parsons and his colleagues do not substantiate the claim of a significant difference at any sampling time. This situation is very different
R E SCHNEIDER
C F HAWKINS Queen Elizabeth Hospital,
Birmingham
'Parsons, R L, et al, Gut, 1976, 17, 139. 2Babb, J, et al, Lancet, 1976, 1, 1413.
Allergy to diazepam SIR,-I read with interest the report by Dr Louis Milner (15 January, p 144). The patient under review did give a history of being allergic to drugs including chlordiazepoxide and chlorazepate, both of which are members of the benzodiazepine family to which diazepam also belongs, implying that she had taken them in some form or fashion earlier in her life, when her allergy to those drugs was detected. May I point out the basic mechanism of this very rare anaphylactic reaction? The active and common metabolite of all benzodiazepines (which include chlordiazepoxide, diazepam, nitrazepam, chlorazepate, oxazepam, chonazepam, and medazepam) is desmethyldiazepam,l which seems to be the real antigenic moiety and it is a case of cross-allergenicity between different members of the same chemical family. As the patient had taken chlordiazepoxide and chlorazepate earlier in her life tissue-fixed antibody to the common antigenic moiety was already present in the body; otherwise the very first administration of diazepam would not have led to the reported anaphylactic reactions (type I immune reaction). It would have been better not to give diazepam to a patient known to be allergic to other chemically allied members of the same family. This reaction has thrown some light on the basic cause of this cross-allergenicity between different members of the benzodiazepine group of drugs which are prescribed so commonly today. SISIR K MAJUMDAR Elmdene Alcoholic Treatment Unit, Bexley Hospital, Bexley, Kent
SIR,-In his answer (1 January, p 42) to the question "How should one treat atrial fibrillation in the elderly ?" your expert has not included any mention of tachycardia. I am no cardiologist but have learned a little about this common problem in general practice. Gross tachycardia (110-150/min) with frank failure, or the threat of failure, is common at the onset of atrial fibrillation and may occur without obvious precipitating cause at any subsequent time. Adequate digoxin controls the pulse rate, usually cures the failure if present, and certainly lessens the risk of failure if it is not already present. Digoxin is less effective if thyrotoxicosis is present and there may be other, rarer, reasons for it to fail or be best avoided. I know of no adequate modem substitute. Incidentally, a return to normal rhythm within two or three days of digitalisation is fairly common too. Having digitalised one of these patients, it has been my custom to maintain digitalisation indefinitely unless normal rhythm returns. I am aware that some, at least, relapse into tachycardia within a week or two of stopping their digoxin. I should much like to know-if anyone really knows-how many of them need life-time maintenance. Have we-the whole profession-or have we not a duty to go to some trouble to try to ensure long continuous treatment in every such case ? E B GROGONO Woodford Green, Essex
***Our expert writes: "If a patient with atrial fibrillation has persistent tachycardia or goes into failure when digitalis is withdrawn, then clearly he or she should be digitalised permanently, thyrotoxicosis having first been excluded. This is not, however, the case with the majority of old people with atrial fibrillation and normal ventricular rates. Such patients often only go into heart failure when they have an intercurrent chest infection or an incident of cardiac infarction. In these cases permanent digitalisation is not necessary except during the incident of failure. Digitalis toxicity is now well documented and may provoke all kinds of dysrhythmia in the elderly, sometimes with fatal results. As with all powerful drugs it is better administered when it is needed rather than for long periods."-ED, BM7.
Lader, M, in Advanced Medicine-Topics in Therapeutics-2, ed P Turner, p 212. London, Pitman Medical, 1976.
Progestasert SIR,-Dr Louis Milner (15 January, p 144) comments that "hypersensitivity reactions to the benzodiazepine derivatives have not been reported," but Dr R H Felix and I reported an allergic reaction to Valium (diazepam) in the Lancet in 1974 in a paper entitled "The value of patch and other skin tests in drug eruptions." Patch-testing with the
SIR,-The recent widespread publicity in the national press and the active promotion of Progestasert (a new intrauterine contraceptive device which slowly releases progesterone over a period of one year) to general practitioners may have created the false impression that this item is prescribable on an FPIO prescription form.
12 FEBRUARY 1977
seemed somewhat unnecessary when one was primarily attempting to determine the incidence of diarrhoea in a large number of patients. Both trials showed the incidence of diarrhoea to be halved in the patients receiving talampicillin and these results were shown to be statistically significant (P < 0-02 in both studies). Dr Grant is perfectly entitled to express his personal opinion of these antibiotics, but other clinicians appear to have come to a different conclusion and amoxycillin is now very widely used in both hospital and general practice in this country and overseas. Finally, I am sure that Dr Grant would be the first to appreciate that he is getting better value for money when he provides drugs that are well absorbed rather than those that, for the most part, go down the drain. E T KNUDSEN Medical Director, Beecham Pharmaceuticals Division
443
between low levels of human placental lactogen and poor outcome of pregnancy but also that, in their practice, use of the assay in high-risk pregnancies can lead to a significant reduction in the antepartum death rate. We have been unable to trace comparable evidence relating to other methods of fetal risk prediction. Urinary oestrogen assay and serial ultrasound cephalometry have been widely used in Britain during recent years and the table printed below documents their increased use in one geographically defined population. Recent research2-4 has raised questions concerning the value of these techniques, but these studies have failed to provide any clear answers. Chards has drawn attention to the fact that "much effort is spent on the identification of new parameters of feto-placental function when the main requirement is for detailed evaluation and refinement of tests which are already available." However, some centres have already replaced urinary oestrogen assays with plasma oestriol measurements and introduced other biochemical and physical methods of antepartum fetal monitoring. Spellacy's example with human placental lactogen should be followed by using randomised controlled trials to assess the value of the increasing variety of methods in use. Who knows, fetal movement counts by the mother6 may prove as effective as other competitors and an even more convenient and economically attractive proposition than your recommended "best buy" of urinary oestrogen assay at 25p a time.
Brentford, Middx Nakazawa, S, Amoxycillin (BRL 2333) International Symposium, London, September 1973, II, 24. 2 Robinson, G N, Nouvelle Presse Medicale, 1975, 4, 2461. 3 Comber, K R, Osborne, Christine D, and Sutherland, R, Antimicrobial Agents and Chemotherapy, 1975, 7, 179. May, J R, and Ingold, A, British Journal of Diseases of the Chest, 1972, 66, 185. Molla, A L, Practitioner, 1974, 212, 123. 6 Pines, A, et al, Chemotherapy, 1977, 23, 58. 7 Ingold, A, British Journal of Diseases of the Chest, 1975, 69, 211. Burgi, H, Chemotherapy, 1973, 18, suppl, p 19. 9 Burns, M W, and Devitt, L, Journal of Infectious Diseases, 1974, 129, S 194. 10 Khan, A A, Journal of International Medical Research, 1975, 3, 230. Scragg, J N, and Rubidge, C J, American Journal of Tropical Medicine and Hygiene, 1975, 24, 860. 12 Scragg, J N, British Medical Journal, 1976, 2, 1031. We are grateful to Professor K T Evans and 3 Afifi, A M, Adnan, M, and Garf, A A, British Medical Dr Huw Gravelle, Department of Radiology, Journal, 1976, 2, 1033. 14 Leigh, D A, et al, British Medical Journal. 1976, 1, Welsh National School of Medicine, for access to 1378. ultrasonogram records. 15 Middleton, R S W, Clinical Trials Journal, 1976, 13, ANNE C DAVIES 23. Final-year medical student 6Jaffe, G, Murphy, J E, and Robinson, 0 P W, Practitioner, 1976, 216, 455. 17Knudson, E T, and Harding, J W, British Journal of IAIN CHALMERS Clinical Practice, 1975, 29, 255. Welsh National School of Medicine,
abnormality. Professor C F George and Dr C M Castleden (1 January, p 47) consider the various other factors which might explain the results of Dr Schneider and his colleagues. As mentioned above, we do not dispute that propranolol is virtually completely absorbed from the gastrointestinal tract and, as indicated in our previous letter, we still consider that altered rates of absorption from the gut are a likely explanation of the results. However, we are sceptical whether accelerated gastric emptying really does occur in coeliac disease as has been previously reported.5 While we accept that a reduction in the first pass metabolism of the drug as a result of liver dysfunction is a possible mechanism in Crohn's disease, in which minor histological abnormality of the liver is a relatively frequent DIANA RIAD FAHMY occurrence,6 this seems unlikely in coeliac
Cardiff
Predicting fetal death SIR,-The antepartum monitoring techniques considered in your recent leading article (15 January, p 123) may well be more or less precise predictors of fetal death, but they were introduced with a view to preventing death, not predicting it. This may seem a statement of the obvious, but the research surrounding these physical and biochemical techniques suggests that there may be some confusion of these two roles. The fact that retrospective studies have demonstrated reasonable correlation between the outcome of pregnancy and measurements made with these techniques does not establish that their use will prevent fetal death. You refer to the only study of which we are aware which was specifically designed to assess the extent to which use of a fetal risk predictor may alter the natural history of a pregnancy.' Spellacy and his colleagues have certainly provided encouraging evidence not only of a reasonable correlation
Supraregional Assay Service, Tenovus Institute of Cancer Research, Cardiff. lSpellacy, W N, Buhi, W C, and Birk, S A, American Journal of Obstetrics and Gynecology, 1975, 121, 835. 2Chalmers, I, et al, British Medical Journal, 1976, 1, 735. 3Chalmers, I, Lawson, J G, and Turnbull, A C, British Journal of Obstetrics and Gynaecology, 1976, 83, 921. 4 Chalmers, I, Lawson, J G, and Turnbull, A C, British Journal of Obstetrics and Gynaecology, 1976, 83, 930. 5 Chard, T, Journal of Clinical Pathology, 1976, 29, suppl 10, p 18. 6 Pearson, J F, and Weaver, J B, British Medical Journal, 1976, 1, 1305.
Plasma propranolol levels in Crohn's disease and coeliac disease SIR,-In reply to Dr R E Schneider and others (27 November, p 1324) it is not disputed that propranolol might eventually be fully absorbed in the patients or in the controls in their
Frequency and intensity of antepartum monitoring among parturients resident in Cardiff, 1967-74
Referred for oestrogen assay () Mean number of assays per case referred Referred for diagnostic ultrasound* during pregnancy () Mean number of ultrasound examinations per case referred *All indications.
1967
1968
1969
1970
1971
1972
1973
1974
-
30
3-2
90
19-4
19-7
17-6
16-4
2-6
2-8
2-5
2-9
2-8
3-2
3-6
-
investigation (2 October, p 794). The problem is why the plasma levels of propranolol were higher over 1-6 h after ingestion in patients with Crohn's disease and at 1 h after ingestion in treated coeliac patients than in healthy volunteers. However, an explanation for their results has not been provided, nor has our proposal (6 November, p 1135) of the possible role of the jejunal acid microclimate in regulation of propranolol absorption from the proximal jejunum been disproved. Higher propranolol levels in rheumatoid arthritis and pneumonia' do not conflict with the jejunal acid microclimate hypothesis. The occurrence of folic acid deficiency in rheumatoid arthritis2-4 provides some indirect evidence which suggests that the proximal jejunal acid microclimate could be deficient in these nongastrointestinal disorders. Our own unpublished observations on seven patients with folic acid deficiency and normal jejunal histology have shown that the proximal jejunal luminal surface pH in these patients was significantly less acid than in normal controls. Therefore it is possible that the acid microclimate could be deficient in rheumatoid arthritis in spite of no obvious intestinal
07 10
101
14-3
200
210
2-1
2-3
2-5
2-2
disease, in which liver dysfunction is rare. In our own series of 300 adult patients with coeliac disease only three cases of chronic liver disease have been seen. We were disappointed that Dr R L Parsons and his colleagues (8 January, p 103) have found that two drugs with a low pKa, aspirin and indomethacin, are absorbed normally or even faster than normal in coeliac disease. Unfortunately it is not clear from their letter whether their coeliac patients were on normal or gluten-free diets. In a previous paper,7 referred to in their letter, they found that plasma propranolol levels after ingestion were higher in coeliacs, all of whom were on a gluten-free diet, when compared with normals, and no details of jejunal morphology at the time of study were provided. Obviously in deciding the role of the jejunum and its acid microclimate in the absorption of drugs it is important to have details of the jejunal morphology at the time of study. Nevertheless, it is strange that, whereas two drugs with a low pKa appear to be normally absorbed in coeliac disease, folic acid, which is probably 50°O unionised at pH 3.58 (it has two carboxyl groups and therefore cannot have one pKa value), is malabsorbed in untreated coeliac disease.9-'1 The jejunal acid microclimate must affect absorption of drugs which are weak acids, such as folic acid, aspirin, and indomethacin, or weak bases, such as propranolol, and so its deficiency in untreated coeliac
444
BRITISH MEDICAL JOURNAL
disease12 and in Crohn's disease'3 should have some effect on absorption of these drugs. However, there are so many other factors involved in drug absorption and drug plasma levels that these other factors may obscure or alter effects that changes in the jejunal acid microclimate may produce. B T COOPER W TREVOR COOKE Nutritional and Intestinal Unit, General Hospital, Birmingham
from that in Crohn's disease. Here the difference between the plasma propranolol levels in patients and controls is highly significant at all sampling times-apart from the '-hour oneas are the C.ax values as well as the areas under the curve. We should like to emphasise once more that we do not necessarily equate this with enhanced absorption of the drug in Crohn's and other diseases.2 Other factors are much more likely to be involved.
12 FEBRUARY 1977
injectable form of diazepam was positive on two separate occasions. I think there was little doubt of the diagnosis in that case. The case we described was an example of an epidermal (eczematous) reaction to a systemically administered drug. J S COMAISH Department of Dermatology, Royal Victoria Infirmary, Newcastle upon Tyne I
Felix, R H, and Comaish, J S, Lancet, 1974, 1, 1017.
M L LUCAS We should like to thank Dr J A H Waterhouse JOHN A BLAIR from Cancer Registry, Queen Elizabeth Medical Atrial fibrillation in the elderly Department of Chemistry, University of Aston in Birmingham, Birmingham
Centre, for checking our statistical calculations.
Babb, J, et al, Lancet, 1976, 1, 1413. 2Gough, K R, et al, British Medical Journal, 1964, 1, 212. 3 Deller, D J, et al, British Medical3Journal, 1966, 1, 765. ' Ratanashien, K, et al, Journal of Clinical Pathology. In press. 5Mobert, S, and Carlberger, F, Scandinavian Journal of Gastroenterology, 1974, 9, 17. Eade, M, et al, Scandinavian Journal of Gastroenterology, 1971, 6, 199. 7Parsons, R L, et al, Gut, 1976, 17, 139. Blair, J A, and Matty, A J, Clinics in Gastroenterology, 1974, 3, 183. 9 Hepner, G W, et al, Lancet, 1968, 2, 302. 0 Gerson, C D, et al, American Journal of Digestive Diseases, 1974, 19, 911. Mackenzie, J F, and Russell, R I, Clinical Science and Molecular Medicine, 1976, 51, 363. Lucas, M L, et al. Submitted for publication. s Cooper, B T, et al. Submitted for publication.
Department of Medicine, University of Birmingham
I
SIR,-We were interested in the comments by Dr R L Parsons and other (8 January, p 103) on our work on plasma propranolol levels in patients with coeliac disease (2 Ocotber, p 794). We cannot accept their claim to have shown a statistically significant rise in these levels at 1, 6, and 8 h after the oral administration of 40 mg of the drug. Using Student's t test, as they did, recalculation of the t values for the difference between the plasma propranolol levels in controls and coeliac patients from their own data' gave values of 1-8, 1-73, and 1-3 respectively. The corresponding P values were therefore on no occasion below or even near the 0-05 level of significance. It also seems strange that the difference between the areas under the curve for plasma propranolol levels in controls and coeliac patients should be significant when calculated according to one programme (Wagner and Nelson), but not significant when calculated according to another (Saunders and Natumen). In our patients the area under the curve in coeliacs was not significantly raised when using Simpson's rule. The claim that the discrepancy between their and our results is due to the difference in the numbers of patients used in the two groups as well as to a difference in the duration of treatment on a gluten-free diet also is not acceptable. Although Dr Parsons and his colleagues' allegedly studied 14 patients compared with our eight, they gave data on only 13 and for their calculations indeed only use 11 of these. This reduces the difference between their and our series to only three patients instead of five as claimed in their letter. Although one can only guess which patients were included it does seem likely that the average duration of treatment in their series was shorter than in ours. This would suggest that duration of treatment does not significantly affect plasma levels of propranolol in coeliac disease as the data of Dr Parsons and his colleagues do not substantiate the claim of a significant difference at any sampling time. This situation is very different
R E SCHNEIDER
C F HAWKINS Queen Elizabeth Hospital,
Birmingham
'Parsons, R L, et al, Gut, 1976, 17, 139. 2Babb, J, et al, Lancet, 1976, 1, 1413.
Allergy to diazepam SIR,-I read with interest the report by Dr Louis Milner (15 January, p 144). The patient under review did give a history of being allergic to drugs including chlordiazepoxide and chlorazepate, both of which are members of the benzodiazepine family to which diazepam also belongs, implying that she had taken them in some form or fashion earlier in her life, when her allergy to those drugs was detected. May I point out the basic mechanism of this very rare anaphylactic reaction? The active and common metabolite of all benzodiazepines (which include chlordiazepoxide, diazepam, nitrazepam, chlorazepate, oxazepam, chonazepam, and medazepam) is desmethyldiazepam,l which seems to be the real antigenic moiety and it is a case of cross-allergenicity between different members of the same chemical family. As the patient had taken chlordiazepoxide and chlorazepate earlier in her life tissue-fixed antibody to the common antigenic moiety was already present in the body; otherwise the very first administration of diazepam would not have led to the reported anaphylactic reactions (type I immune reaction). It would have been better not to give diazepam to a patient known to be allergic to other chemically allied members of the same family. This reaction has thrown some light on the basic cause of this cross-allergenicity between different members of the benzodiazepine group of drugs which are prescribed so commonly today. SISIR K MAJUMDAR Elmdene Alcoholic Treatment Unit, Bexley Hospital, Bexley, Kent
SIR,-In his answer (1 January, p 42) to the question "How should one treat atrial fibrillation in the elderly ?" your expert has not included any mention of tachycardia. I am no cardiologist but have learned a little about this common problem in general practice. Gross tachycardia (110-150/min) with frank failure, or the threat of failure, is common at the onset of atrial fibrillation and may occur without obvious precipitating cause at any subsequent time. Adequate digoxin controls the pulse rate, usually cures the failure if present, and certainly lessens the risk of failure if it is not already present. Digoxin is less effective if thyrotoxicosis is present and there may be other, rarer, reasons for it to fail or be best avoided. I know of no adequate modem substitute. Incidentally, a return to normal rhythm within two or three days of digitalisation is fairly common too. Having digitalised one of these patients, it has been my custom to maintain digitalisation indefinitely unless normal rhythm returns. I am aware that some, at least, relapse into tachycardia within a week or two of stopping their digoxin. I should much like to know-if anyone really knows-how many of them need life-time maintenance. Have we-the whole profession-or have we not a duty to go to some trouble to try to ensure long continuous treatment in every such case ? E B GROGONO Woodford Green, Essex
***Our expert writes: "If a patient with atrial fibrillation has persistent tachycardia or goes into failure when digitalis is withdrawn, then clearly he or she should be digitalised permanently, thyrotoxicosis having first been excluded. This is not, however, the case with the majority of old people with atrial fibrillation and normal ventricular rates. Such patients often only go into heart failure when they have an intercurrent chest infection or an incident of cardiac infarction. In these cases permanent digitalisation is not necessary except during the incident of failure. Digitalis toxicity is now well documented and may provoke all kinds of dysrhythmia in the elderly, sometimes with fatal results. As with all powerful drugs it is better administered when it is needed rather than for long periods."-ED, BM7.
Lader, M, in Advanced Medicine-Topics in Therapeutics-2, ed P Turner, p 212. London, Pitman Medical, 1976.
Progestasert SIR,-Dr Louis Milner (15 January, p 144) comments that "hypersensitivity reactions to the benzodiazepine derivatives have not been reported," but Dr R H Felix and I reported an allergic reaction to Valium (diazepam) in the Lancet in 1974 in a paper entitled "The value of patch and other skin tests in drug eruptions." Patch-testing with the
SIR,-The recent widespread publicity in the national press and the active promotion of Progestasert (a new intrauterine contraceptive device which slowly releases progesterone over a period of one year) to general practitioners may have created the false impression that this item is prescribable on an FPIO prescription form.
