- Email: [email protected]
PREDNISOLONE DOES NOT PREVENT POST-HERPETIC NEURALGIA
576
regular pulse rate of 120/min with blood pressure of 110/70 mm Hg, raised venous pressure, loud gallop rhythm, late inspiratory crackles at both lung bases, and pedal oedema. There was a 38-week uterus and a dilated cervical os of 2-3 cm. Chest X-ray showed cardiomegaly and moderate pulmonary oedema. ECG confirmed the previous findings, with echocardiography demonstrating a dilated, hypokinetic left ventricle. As labour appeared imminent, she was immediately transferred to the intensive therapy unit. Plasma protein fraction was administered to maintain a right atrial pressure of 6-10 cm H20 during the establishment of epidural blockade with 0375% bupivacaine. 12 h later a healthy girl was delivered by forceps without maternal effort. After delivery the patient was gently mobilised. However, 96 h later her condition suddenly began to deteriorate with evidence of pump failure that required high-dose inotropic support. Emergency heart transplantation was required 17 days postpartum with an excellent result. One of the characteristic features of peripartum cardiomyopathy is its presentation or progression in the early puerperium,!,2 this patient being a most extreme example. It has been suggested that this timing may represent a rebound from the metabolic changes of pregnancy.’ However, the observations of Dr Robson and colleagues (July 11, p 111) of an increase in stroke volume in the early puerperium in normal women, accompanied by increased cardiac chamber dimensions, may explain the phenomenon. The suggested mechanism of augmented venous return due to auto-transfusion from the uteroplacental bed could also represent the critical haemodynamic stress that leads to the often dramatic and progressive clinical deterioration seen in peripartum cardiomyopathy. As a permissive feature it may be amenable to the early use of vasodilator therapy. If this approach proved unsuccessful immediate referral with a view to cardiac
University College Hospital, London WC1E 6AU
1.
A. A. GRACE R. E. JENKINS M. J. KING
Julian DG, Szekely P. Peripartum cardiomyopathy. Prog Cardiovasc
223-40. 2. Homans DC.
Peripartum cardiomyopathy. N Engl J Med 1985; 312:
Dis
1985;
27:
1432-37.
PREDNISOLONE DOES NOT PREVENT POST-HERPETIC NEURALGIA
SIR,-Professor Esmann and colleagues (July 18, p 126) show prednisolone does not prevent post-herpetic neuralgia. We have reported in an open trial that acyclovir, if given within 24 h of the appearance of the rash, will arrest the development of the lesions such that they remain at the papular staged The average duration of the rash in Esmann et al’s cases was 23 days and most patients already had large numbers of vesicles, which suggests that they were treated after 48 h. Most cases in this country are managed at home by general practitioners and it is important to emphasise that these patients should be treated as early as possible. In our experience, about 50 % of cases present to their general practitioners within 24 h of the development of the rash and although acyclovir provides some protection at later stages, our data suggest that this is too late. The first 24 h after the development of the rash are critical and any synergistic protective action of acyclovir and steroids would be of major benefit in patients treated at this stage. Esmann and colleagues show prednisolone has an effect on early pain and it is possible that acyclovir and prednisolone will have a protective action against post-herpetic neuralgia, if given early enough. If given for 3 weeks as reported, steroids may produce side-effects, but the use of steroids for 5 days should be safe and may have a protective action in patients treated within 48 h of the development of the rash. This possibility merits further study, because there is no effective treatment for post-herpetic neuralgia, which can be a prolonged and disabling condition. that
Royal Liverpool Hospital, Liverpool L7 8XP 1. Finn
R,
Smith
SiR,—Your editorial on the eventual association between diabetic microangiopathy and joint disorders (Aug 8, p 313) once again reminds us of the clinical epidemiological perspective-namely, that coexisting common diseases seen in the hospital ward may be explained by chance alone. I was prompted by your editorial to check the database of the Gothenburg studyl of a general population cohort aged 61-70 years (N 47 124) and followed up for over 10 years for any admission to the only general hospital of the area with diabetic retinopathy and/or nephropathy (N = 232) as well as skin and joint disorders. Significant (p<0001) associations, as expected, were found with cutaneous chronic ulcers and osteomyelitis, while no relation was found between diabetic microangiopathy and rheumatoid arthritis or osteoarthritis. A previous analysis on the same community-based study suggested a significant association between diabetes in general (ie, with or without microangiopathy) and psoriasis.2 =
Department of Social Medicine and Clinical Epidemiology, University of Gothenburg, PO Box 5008, 42105 Vastra Frolunda, Sweden 1. 2.
BENGT LINDEGARD
Lindegard B. Malignant diabetic retinopathy. Lancet 1985; i: 271. Lindegård B. Diseases associated with psoriasis in a general population of 159,200 middle-aged, urban, native Swedes. Dermatologica 1986; 172: 298—304.
SOCIAL-CLASS GRADIENTS, SPECIFIC CAUSE MORTALITY, AND ACCURACY OF DIAGNOSIS OF CAUSE OF DEATH p 197) examine social-class differences in ischaemic heart disease and there has been correspondence in your columns about Professor Stehbens’ review on the accuracy of death certification in respect of coronary heart disease (March 14, p 606). Stehbens is rightly sceptical of epidemiological studies which analyse international differences and time changes in specific cause mortality without reference to differences in diagnostic accuracy and certifying practices. However, it would be unfortunate if Stehbens’ strictures were to be applied unthinkingly to studies of social-class differences in specific cause mortality such as the British Regional Heart Study or the OPCS decennial supplement.1 Differences in diagnostic accuracy can only influence social-class gradients in specific cause mortality if misdiagnosis occurs differentially across social classes: class gradients may then be inflated or deflated artefactually by the systematic misallocation of certain social groups to particular
SIR,-Dr Pocock and colleagues (July 25,
transplantation seems appropriate. Medical Unit and Intensive Therapy Unit,
DIABETIC MICROANGIOPATHY AND JOINT DISORDERS
RONALD FINN
MA. Oral acydovir for herpes zoster. Lancet 1984;
ii. 575.
diagnoses. We have directly investigated social class variations in the of cause of death. Our analysis suggests that systematic misallocation does exist, although its impact on specific cause mortality gradients is likely to be limited. Class gradients in cardiovascular diseases would appear to be unaffected. We were able to match the occupation of the deceased (obtained at death registration) with data comparing clinicians’ and pathologists’ diagnoses of cause of death in 1152 consecutive hospital deaths sent for necropsy in the South Lothian District between May, 1975, and June, 1977 (we thank Dr H. M. Cameron and Dr E. McGoogan, department of pathology, University of Edinburgh, and the General Register Office for Scotland). The methodology of the necropsy study has been reported.2 Due to inadequate information on the occupations of the females, analysis was confined to the 637 male deaths. To obtain sufficient numbers for analysis the Registrar General’s classes had to be aggregated into three groups: nonmanual (classes I, II, and IlIa), skilled manual (IIIb), and semi-skilled and unskilled manual (IV and V). The table shows the distribution of diagnostic disagreement between clinician and pathologist by social class and International Classification of Diseases (ICD) diagnostic chapter. A highly restrictive measure of disagreement was chosen: disagreement was only judged to have occurred when the major underlying cause of death was allocated to two different ICD chapters, although ICD chapter VII has been divided into cardiovascular and cerebrovascular diseases.
diagnosis
regular pulse rate of 120/min with blood pressure of 110/70 mm Hg, raised venous pressure, loud gallop rhythm, late inspiratory crackles at both lung bases, and pedal oedema. There was a 38-week uterus and a dilated cervical os of 2-3 cm. Chest X-ray showed cardiomegaly and moderate pulmonary oedema. ECG confirmed the previous findings, with echocardiography demonstrating a dilated, hypokinetic left ventricle. As labour appeared imminent, she was immediately transferred to the intensive therapy unit. Plasma protein fraction was administered to maintain a right atrial pressure of 6-10 cm H20 during the establishment of epidural blockade with 0375% bupivacaine. 12 h later a healthy girl was delivered by forceps without maternal effort. After delivery the patient was gently mobilised. However, 96 h later her condition suddenly began to deteriorate with evidence of pump failure that required high-dose inotropic support. Emergency heart transplantation was required 17 days postpartum with an excellent result. One of the characteristic features of peripartum cardiomyopathy is its presentation or progression in the early puerperium,!,2 this patient being a most extreme example. It has been suggested that this timing may represent a rebound from the metabolic changes of pregnancy.’ However, the observations of Dr Robson and colleagues (July 11, p 111) of an increase in stroke volume in the early puerperium in normal women, accompanied by increased cardiac chamber dimensions, may explain the phenomenon. The suggested mechanism of augmented venous return due to auto-transfusion from the uteroplacental bed could also represent the critical haemodynamic stress that leads to the often dramatic and progressive clinical deterioration seen in peripartum cardiomyopathy. As a permissive feature it may be amenable to the early use of vasodilator therapy. If this approach proved unsuccessful immediate referral with a view to cardiac
University College Hospital, London WC1E 6AU
1.
A. A. GRACE R. E. JENKINS M. J. KING
Julian DG, Szekely P. Peripartum cardiomyopathy. Prog Cardiovasc
223-40. 2. Homans DC.
Peripartum cardiomyopathy. N Engl J Med 1985; 312:
Dis
1985;
27:
1432-37.
PREDNISOLONE DOES NOT PREVENT POST-HERPETIC NEURALGIA
SIR,-Professor Esmann and colleagues (July 18, p 126) show prednisolone does not prevent post-herpetic neuralgia. We have reported in an open trial that acyclovir, if given within 24 h of the appearance of the rash, will arrest the development of the lesions such that they remain at the papular staged The average duration of the rash in Esmann et al’s cases was 23 days and most patients already had large numbers of vesicles, which suggests that they were treated after 48 h. Most cases in this country are managed at home by general practitioners and it is important to emphasise that these patients should be treated as early as possible. In our experience, about 50 % of cases present to their general practitioners within 24 h of the development of the rash and although acyclovir provides some protection at later stages, our data suggest that this is too late. The first 24 h after the development of the rash are critical and any synergistic protective action of acyclovir and steroids would be of major benefit in patients treated at this stage. Esmann and colleagues show prednisolone has an effect on early pain and it is possible that acyclovir and prednisolone will have a protective action against post-herpetic neuralgia, if given early enough. If given for 3 weeks as reported, steroids may produce side-effects, but the use of steroids for 5 days should be safe and may have a protective action in patients treated within 48 h of the development of the rash. This possibility merits further study, because there is no effective treatment for post-herpetic neuralgia, which can be a prolonged and disabling condition. that
Royal Liverpool Hospital, Liverpool L7 8XP 1. Finn
R,
Smith
SiR,—Your editorial on the eventual association between diabetic microangiopathy and joint disorders (Aug 8, p 313) once again reminds us of the clinical epidemiological perspective-namely, that coexisting common diseases seen in the hospital ward may be explained by chance alone. I was prompted by your editorial to check the database of the Gothenburg studyl of a general population cohort aged 61-70 years (N 47 124) and followed up for over 10 years for any admission to the only general hospital of the area with diabetic retinopathy and/or nephropathy (N = 232) as well as skin and joint disorders. Significant (p<0001) associations, as expected, were found with cutaneous chronic ulcers and osteomyelitis, while no relation was found between diabetic microangiopathy and rheumatoid arthritis or osteoarthritis. A previous analysis on the same community-based study suggested a significant association between diabetes in general (ie, with or without microangiopathy) and psoriasis.2 =
Department of Social Medicine and Clinical Epidemiology, University of Gothenburg, PO Box 5008, 42105 Vastra Frolunda, Sweden 1. 2.
BENGT LINDEGARD
Lindegard B. Malignant diabetic retinopathy. Lancet 1985; i: 271. Lindegård B. Diseases associated with psoriasis in a general population of 159,200 middle-aged, urban, native Swedes. Dermatologica 1986; 172: 298—304.
SOCIAL-CLASS GRADIENTS, SPECIFIC CAUSE MORTALITY, AND ACCURACY OF DIAGNOSIS OF CAUSE OF DEATH p 197) examine social-class differences in ischaemic heart disease and there has been correspondence in your columns about Professor Stehbens’ review on the accuracy of death certification in respect of coronary heart disease (March 14, p 606). Stehbens is rightly sceptical of epidemiological studies which analyse international differences and time changes in specific cause mortality without reference to differences in diagnostic accuracy and certifying practices. However, it would be unfortunate if Stehbens’ strictures were to be applied unthinkingly to studies of social-class differences in specific cause mortality such as the British Regional Heart Study or the OPCS decennial supplement.1 Differences in diagnostic accuracy can only influence social-class gradients in specific cause mortality if misdiagnosis occurs differentially across social classes: class gradients may then be inflated or deflated artefactually by the systematic misallocation of certain social groups to particular
SIR,-Dr Pocock and colleagues (July 25,
transplantation seems appropriate. Medical Unit and Intensive Therapy Unit,
DIABETIC MICROANGIOPATHY AND JOINT DISORDERS
RONALD FINN
MA. Oral acydovir for herpes zoster. Lancet 1984;
ii. 575.
diagnoses. We have directly investigated social class variations in the of cause of death. Our analysis suggests that systematic misallocation does exist, although its impact on specific cause mortality gradients is likely to be limited. Class gradients in cardiovascular diseases would appear to be unaffected. We were able to match the occupation of the deceased (obtained at death registration) with data comparing clinicians’ and pathologists’ diagnoses of cause of death in 1152 consecutive hospital deaths sent for necropsy in the South Lothian District between May, 1975, and June, 1977 (we thank Dr H. M. Cameron and Dr E. McGoogan, department of pathology, University of Edinburgh, and the General Register Office for Scotland). The methodology of the necropsy study has been reported.2 Due to inadequate information on the occupations of the females, analysis was confined to the 637 male deaths. To obtain sufficient numbers for analysis the Registrar General’s classes had to be aggregated into three groups: nonmanual (classes I, II, and IlIa), skilled manual (IIIb), and semi-skilled and unskilled manual (IV and V). The table shows the distribution of diagnostic disagreement between clinician and pathologist by social class and International Classification of Diseases (ICD) diagnostic chapter. A highly restrictive measure of disagreement was chosen: disagreement was only judged to have occurred when the major underlying cause of death was allocated to two different ICD chapters, although ICD chapter VII has been divided into cardiovascular and cerebrovascular diseases.
diagnosis