799
watery stools and a positive bedside ’Clinitest’, despite lactose-free feeds, suggest osmotic diarrhoea with sugar malabsorption. In October, 1985, one of four paediatric wards was allocated to accept only patients with dehydrating diarrhoea irrespective of any associated conditions. A policy decision in 1987 not to admit more than one child per cot resulted in a short-term need for patients to be accommodated elsewhere, but subsequently the ward has rarely been full. No additional medical or nursing staff were required. Other aspects of diarrhoea management, such as the use of antibiotics and the "bowel cocktail",2and a graded approach to refeeding, were completed in accordance with established guidelines as before. Before 1986, admissions for diarrhoea accounted for 2-4% of all outpatient attendances. 30% of all inpatients were admitted with diarrhoea and had an inpatient case fatality rate of around 23%.3 The statistical review of the first five years of our new approach is shown below:
uptake were noted in mid and lower zones of the left lung. He was discharged well after 10 days of intravenous antibiotics. Although a gallium scan is not specific for infection, it is a reasonably sensitive detector of inflammation that may not be seen on routine chest radiographs.1-3 We suggest that our patient had a "radiographically silent" pneumococcal alveolitis complicated by bacteraemia rather than a "bacteraemic bronchitis" alone. The reasons for lack of radiographic findings are unclear but may be related to either the organism, such as slower multiplication in alveolar spaces, lower pathogenicity or toxigenicity, or host factors, such as lower leucocyte mobilisation in the alveolar space4 (our patient did not have a leucocytosis during his illness). Unless further sensitive tests such as a gallium scan are used, it is difficult to exclude parenchymal infection with a normal chest radiograph. St John’s Mercy Medical Center, St Louis, Missouri 63141, USA
FARRIN A. MANIAN
1. Coleman DL, Hattner RS, Luce JM, Dodek PM, Golden JA, Murray JF. Correlation between gallium lung scans and fiberoptic bronchoscopy in patients with suspected Pneumocystis carimi pneumonia and the acquired deficiency syndrome. Am Rev
Respir Dis 1984; 130: 1166-69.
Despite improved clinic services and the commissioning of a new hospital in the region, the fall in admissions to 1-2% of outpatient numbers in 1990 is attributable to both oral rehydration and the simplified parenteral fluid schedule. This has resulted in a decreased case-fatality rate. We suggest that the state of the circulation and the need for resuscitation are the important issues in diarrhoeal dehydration. The "percent dehydration" has little management relevance. King Edward VIII Hospital and Department of Paediatrics and Child Health, University of Natal, PO Box 170339, Congella 4013, South Africa
D. F. WITTENBERG S. RAMJI M. BROUGHTON
1. Mackenzie A, Barnes G, Shann F. Clinical signs of dehydration in children. Lancet 1989; ii. 605-07. 2. Bowie MD, Mann MD, Hill ID. The bowel cocktail. Pediatrics 1981; 67: 920-21. 3. Loening WEK Management of acute infective diarrhoea. S Afr J Cont Med Educ
1986; 4: 69-72.
Pneumococcal bacteraemia and the normal chest radiograph SIR,-Iread with interest the reports by Dr Eng and Dr Aronson (Nov 17, p 1266), and Dr Martinez and colleagues (Jan 5, p 57) about Streptococcus pneumoniae bacteraemia, which in the cases of Eng and Aronson were associated with normal chest radiography. However, routine chest radiography may not have been sufficiently sensitive to exclude pneumonitis. A 59-year-old previously healthy non-smoker was admitted to hospital with acute onset of rigors, fever, right-sided chest pain, and productive cough with occasional haemoptysis. His temperature was 40’C, respiratory rate 24/min, pulse 140/min, and blood pressure 138/78 mm Hg. Prominent right-sided crackles were heard in the basilar region of the chest. His white cell count was 62 x 109/1 with 87% segmented neutrophils. Arterial blood gases showed pH 7-53, pC02 27 mm Hg, and Pa02 67 mm Hg on room
air.
Sputum gram stain showed no white blood cells and occasional gram-positive diplococci. Chest radiography showed no evidence of infiltrates and sinus radiographs were normal. Lung ventilation and perfusion scan showed no evidence of pulmonary emboli. Blood cultures grew S pneumoniae. He
was treated with intravenous and later penicillin G. His fever resolved within 48 h, but cough and chest pain persisted for several days. Repeat chest radiography on days 2 and 4 of hospital stay was normal, as was echocardiography. A gallium scan showed uptake (ie, infiltrates) in mid and lower zones of the right lung, consistent with both symptoms and clinical fmdings. Areas of less intense
ampidllin/sulbactam initially,
2. Roberts WC. Morphologic aspects of pulmonary sarcoidosis. Ann Intern Med 1981; 94: 73-76. 3. Line BR, Fulmer JD, Reynolds HY, et al. Gallium-67 citrate scanning in the staging of idiopathic pulmonary fibrosis: correlation with physiologic and morphologic features and bronchoalveolar lavage. Am Rev Respir Dis 1978; 118: 355-65. 4. Mufson MA. Streptococcus pneumomae. In: Mandell GL, Douglas RG, Bennett JE, eds. Principles and practice of infectious diseases, 3rd ed. New York: Churchill Livingstone, 1990: 1539-50.
Is vasculitis in subacute bacterial endocarditis associated with ANCA? SIR,- The demonstration of antineutrophil cytoplasmic (ANCA) is of considerable help in the diagnosis of necrotising crescentic glomerulonephritis.l Conditions other than the various forms of primary vasculitis with crescentic necrotising antibodies
glomerulonephritis are associated with ANCA positivity-eg, Kawasaki syndromeAIDS,3 Pneumocystis carinii pneumonia,4 Crohn’s disease,s or ulcerative colitiS.6 All these disorders are characterised by intense stimulation of the immune system. In the differential diagnosis of crescentic necrotising glomerulonephritis, several other diseases should be considered, which are generally not associated with ANCA—eg, cryoglobulinaemia, polychondritis, and subacute bacterial endocarditis (SBE). We report a patient with bacteriologically confirmed SBE that was associated not only with extrarenal signs of vasculitis, but also with the presence of ANCA. A 26-year-old woman had closure of an atrial septum primum defect in 1971 and mycoplasma pneumonia in 1988. In March, 1990, she had lumbar pain with sacroiliitis (confirmed by leucocyte scintigraphy). Serological findings pointed to Lyme disease (borrelia titres: IgG1 240, IgM1 128). IgG persisted until January, 1991, but IgM became negative. She was given oral cephalosporin treatment for borreliosis. Subsequent de novo appearance of microhaematuria, temporally related to dental treatment, prompted a consultation with a nephrologist who found a nephritic sediment with erythrocyte casts and many dysmorphic erythrocytes. On this occasion a grade 3/6 systolic murmur was noted over the aortic valve. Sonography revealed hepatomegaly and splenomegaly. The patient was subfebrile; erythrocyte sedimentation rate (ESR) was raised (55/33 mm/h), as was C-reactive protein (CRP) (3-8 mg/dl, normal 0-5 mg/dl). Polyclonal elevation of IgG (2040 mg/dl) was noted but no leucocytosis (5’7/nl). Blood culture showed a haemolytic viridans streptococcus. Renal biopsy showed focal segmental glomerulonephritis with crescents of 2 or 4 glomeruli. Massive deposits of IgM and C3 were noted on immunohistology. The patient had classic Osler signs and painful nodules in the subcutaneous fat of the forearm. She was ANCA-negative on admission but the titre rose within 4 weeks (160, fluorescence pattern c-ANCA), and the 29 kD-ELISA became positive (29%, normal 10%). Treatment with high doses of penicillin G (30 MU) and tobramycin (2 x 80 mg) was started. This led to reduction of fever and a return to normal of urinary sediment. ANCA titres also decreased progressively without immunosuppression (Sept 5,1990,
800
titre 40; Sept 20, 20; Jan 24, 1991, negative). At the last analysis, urinary findings, routine laboratory tests (ESR, CRP, blood cells), and cardiological examination were unremarkable. This case raises important questions. Was the demonstration of ANCA in this patient with subacute bacterial endocarditis coincidental or was it related to the appearance of extrarenal vasculitis (Osler nodules)? Do ANCA have a pathogenetic role in extrarenal vasculitis? These questions can be answered only by investigation of larger numbers of patients. Department of Internal Medicine, Nephrology, University of Heidelberg, 6900 Heidelberg 1, Germany
J. WAGNER
Division
K. ANDRASSY E. RITZ
Andrassy K, Koderisch J, Rufer M, et al. Detection and clinical implication of antineutrophil cytoplasmic antibodies in Wegener’s granulomatosis and rapidly progressive glomerulonephritis. Clin Nephrol 1989; 32: 159-67. 2. Savage COS, Winearls CG, Jones S, Marshall PD, Lockwood CM. Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vaculitis. Lancet 1987; i: 1389-93. 3. Koderisch J, Andrassy K, Rasmussen N, Hartman M, Tilgen W. "False-positive" anti-neutrophil cytoplasmic antibodies in HIV infection. Lancet 1990; 335: 1.
1227-28.
4. Andrassy K, Koderisch J, Adler D. Diagnostische Bedeutung von neutrophilen zytoplasmatischen Antikörpern in der Nephrologie. Immun Infekt 1990; 18: 53-55. 5. Peen E, Tejle K, Skogh Th. Anti-granulocyte antibodies in Crohn’s disease. APMIS 1990; 19 (suppl): 98. 6. Jorgensen BB, Wiik A, Hoier-Madsen M, et al. Serum antibodies to neutrophils in patients with chronic inflammatory bowel disease. ANCA Workshop, Washington, 1990: abstr 66.
Alarm bells for enuresis SIR,-In your March 2 editorial you omit two important points. Parents should be asked if the child sleeps very deeply and whether bedwetting is more likely when the child is very tired. I believe that bedwetting occurs merely because the child is too deeply asleep to be woken by a full bladder. In the Lancet series Disabilities and how to live with them many years ago you published the story of a hospital sister who had this complaint. She remarked that it was more likely to happen after a tiring day. In the case of a child who had double incontinence a small dose of amphetamine prevented anal incontinence, and a larger dose also prevented urinary incontinence. It should always be made clear to parents that nocturnal enuresis does not result from laziness and is not the child’s fault. "Shame therapy" is cruel and causes the child much misery-furthermore it does not work. It is difficult to know whether alarms work well. Does the child get better spontaneously? Are parents reluctant to upset the caring doctor by telling him that his treatment does not work? Or have they gone to see another doctor?
Fig Left
panel,
1-Intact PTH with calcitriol treatment.
PTH 50-200
pmol/I (n=5); right panel, PTH
watery stools and a positive bedside ’Clinitest’, despite lactose-free feeds, suggest osmotic diarrhoea with sugar malabsorption. In October, 1985, one of four paediatric wards was allocated to accept only patients with dehydrating diarrhoea irrespective of any associated conditions. A policy decision in 1987 not to admit more than one child per cot resulted in a short-term need for patients to be accommodated elsewhere, but subsequently the ward has rarely been full. No additional medical or nursing staff were required. Other aspects of diarrhoea management, such as the use of antibiotics and the "bowel cocktail",2and a graded approach to refeeding, were completed in accordance with established guidelines as before. Before 1986, admissions for diarrhoea accounted for 2-4% of all outpatient attendances. 30% of all inpatients were admitted with diarrhoea and had an inpatient case fatality rate of around 23%.3 The statistical review of the first five years of our new approach is shown below:
uptake were noted in mid and lower zones of the left lung. He was discharged well after 10 days of intravenous antibiotics. Although a gallium scan is not specific for infection, it is a reasonably sensitive detector of inflammation that may not be seen on routine chest radiographs.1-3 We suggest that our patient had a "radiographically silent" pneumococcal alveolitis complicated by bacteraemia rather than a "bacteraemic bronchitis" alone. The reasons for lack of radiographic findings are unclear but may be related to either the organism, such as slower multiplication in alveolar spaces, lower pathogenicity or toxigenicity, or host factors, such as lower leucocyte mobilisation in the alveolar space4 (our patient did not have a leucocytosis during his illness). Unless further sensitive tests such as a gallium scan are used, it is difficult to exclude parenchymal infection with a normal chest radiograph. St John’s Mercy Medical Center, St Louis, Missouri 63141, USA
FARRIN A. MANIAN
1. Coleman DL, Hattner RS, Luce JM, Dodek PM, Golden JA, Murray JF. Correlation between gallium lung scans and fiberoptic bronchoscopy in patients with suspected Pneumocystis carimi pneumonia and the acquired deficiency syndrome. Am Rev
Respir Dis 1984; 130: 1166-69.
Despite improved clinic services and the commissioning of a new hospital in the region, the fall in admissions to 1-2% of outpatient numbers in 1990 is attributable to both oral rehydration and the simplified parenteral fluid schedule. This has resulted in a decreased case-fatality rate. We suggest that the state of the circulation and the need for resuscitation are the important issues in diarrhoeal dehydration. The "percent dehydration" has little management relevance. King Edward VIII Hospital and Department of Paediatrics and Child Health, University of Natal, PO Box 170339, Congella 4013, South Africa
D. F. WITTENBERG S. RAMJI M. BROUGHTON
1. Mackenzie A, Barnes G, Shann F. Clinical signs of dehydration in children. Lancet 1989; ii. 605-07. 2. Bowie MD, Mann MD, Hill ID. The bowel cocktail. Pediatrics 1981; 67: 920-21. 3. Loening WEK Management of acute infective diarrhoea. S Afr J Cont Med Educ
1986; 4: 69-72.
Pneumococcal bacteraemia and the normal chest radiograph SIR,-Iread with interest the reports by Dr Eng and Dr Aronson (Nov 17, p 1266), and Dr Martinez and colleagues (Jan 5, p 57) about Streptococcus pneumoniae bacteraemia, which in the cases of Eng and Aronson were associated with normal chest radiography. However, routine chest radiography may not have been sufficiently sensitive to exclude pneumonitis. A 59-year-old previously healthy non-smoker was admitted to hospital with acute onset of rigors, fever, right-sided chest pain, and productive cough with occasional haemoptysis. His temperature was 40’C, respiratory rate 24/min, pulse 140/min, and blood pressure 138/78 mm Hg. Prominent right-sided crackles were heard in the basilar region of the chest. His white cell count was 62 x 109/1 with 87% segmented neutrophils. Arterial blood gases showed pH 7-53, pC02 27 mm Hg, and Pa02 67 mm Hg on room
air.
Sputum gram stain showed no white blood cells and occasional gram-positive diplococci. Chest radiography showed no evidence of infiltrates and sinus radiographs were normal. Lung ventilation and perfusion scan showed no evidence of pulmonary emboli. Blood cultures grew S pneumoniae. He
was treated with intravenous and later penicillin G. His fever resolved within 48 h, but cough and chest pain persisted for several days. Repeat chest radiography on days 2 and 4 of hospital stay was normal, as was echocardiography. A gallium scan showed uptake (ie, infiltrates) in mid and lower zones of the right lung, consistent with both symptoms and clinical fmdings. Areas of less intense
ampidllin/sulbactam initially,
2. Roberts WC. Morphologic aspects of pulmonary sarcoidosis. Ann Intern Med 1981; 94: 73-76. 3. Line BR, Fulmer JD, Reynolds HY, et al. Gallium-67 citrate scanning in the staging of idiopathic pulmonary fibrosis: correlation with physiologic and morphologic features and bronchoalveolar lavage. Am Rev Respir Dis 1978; 118: 355-65. 4. Mufson MA. Streptococcus pneumomae. In: Mandell GL, Douglas RG, Bennett JE, eds. Principles and practice of infectious diseases, 3rd ed. New York: Churchill Livingstone, 1990: 1539-50.
Is vasculitis in subacute bacterial endocarditis associated with ANCA? SIR,- The demonstration of antineutrophil cytoplasmic (ANCA) is of considerable help in the diagnosis of necrotising crescentic glomerulonephritis.l Conditions other than the various forms of primary vasculitis with crescentic necrotising antibodies
glomerulonephritis are associated with ANCA positivity-eg, Kawasaki syndromeAIDS,3 Pneumocystis carinii pneumonia,4 Crohn’s disease,s or ulcerative colitiS.6 All these disorders are characterised by intense stimulation of the immune system. In the differential diagnosis of crescentic necrotising glomerulonephritis, several other diseases should be considered, which are generally not associated with ANCA—eg, cryoglobulinaemia, polychondritis, and subacute bacterial endocarditis (SBE). We report a patient with bacteriologically confirmed SBE that was associated not only with extrarenal signs of vasculitis, but also with the presence of ANCA. A 26-year-old woman had closure of an atrial septum primum defect in 1971 and mycoplasma pneumonia in 1988. In March, 1990, she had lumbar pain with sacroiliitis (confirmed by leucocyte scintigraphy). Serological findings pointed to Lyme disease (borrelia titres: IgG1 240, IgM1 128). IgG persisted until January, 1991, but IgM became negative. She was given oral cephalosporin treatment for borreliosis. Subsequent de novo appearance of microhaematuria, temporally related to dental treatment, prompted a consultation with a nephrologist who found a nephritic sediment with erythrocyte casts and many dysmorphic erythrocytes. On this occasion a grade 3/6 systolic murmur was noted over the aortic valve. Sonography revealed hepatomegaly and splenomegaly. The patient was subfebrile; erythrocyte sedimentation rate (ESR) was raised (55/33 mm/h), as was C-reactive protein (CRP) (3-8 mg/dl, normal 0-5 mg/dl). Polyclonal elevation of IgG (2040 mg/dl) was noted but no leucocytosis (5’7/nl). Blood culture showed a haemolytic viridans streptococcus. Renal biopsy showed focal segmental glomerulonephritis with crescents of 2 or 4 glomeruli. Massive deposits of IgM and C3 were noted on immunohistology. The patient had classic Osler signs and painful nodules in the subcutaneous fat of the forearm. She was ANCA-negative on admission but the titre rose within 4 weeks (160, fluorescence pattern c-ANCA), and the 29 kD-ELISA became positive (29%, normal 10%). Treatment with high doses of penicillin G (30 MU) and tobramycin (2 x 80 mg) was started. This led to reduction of fever and a return to normal of urinary sediment. ANCA titres also decreased progressively without immunosuppression (Sept 5,1990,
800
titre 40; Sept 20, 20; Jan 24, 1991, negative). At the last analysis, urinary findings, routine laboratory tests (ESR, CRP, blood cells), and cardiological examination were unremarkable. This case raises important questions. Was the demonstration of ANCA in this patient with subacute bacterial endocarditis coincidental or was it related to the appearance of extrarenal vasculitis (Osler nodules)? Do ANCA have a pathogenetic role in extrarenal vasculitis? These questions can be answered only by investigation of larger numbers of patients. Department of Internal Medicine, Nephrology, University of Heidelberg, 6900 Heidelberg 1, Germany
J. WAGNER
Division
K. ANDRASSY E. RITZ
Andrassy K, Koderisch J, Rufer M, et al. Detection and clinical implication of antineutrophil cytoplasmic antibodies in Wegener’s granulomatosis and rapidly progressive glomerulonephritis. Clin Nephrol 1989; 32: 159-67. 2. Savage COS, Winearls CG, Jones S, Marshall PD, Lockwood CM. Prospective study of radioimmunoassay for antibodies against neutrophil cytoplasm in diagnosis of systemic vaculitis. Lancet 1987; i: 1389-93. 3. Koderisch J, Andrassy K, Rasmussen N, Hartman M, Tilgen W. "False-positive" anti-neutrophil cytoplasmic antibodies in HIV infection. Lancet 1990; 335: 1.
1227-28.
4. Andrassy K, Koderisch J, Adler D. Diagnostische Bedeutung von neutrophilen zytoplasmatischen Antikörpern in der Nephrologie. Immun Infekt 1990; 18: 53-55. 5. Peen E, Tejle K, Skogh Th. Anti-granulocyte antibodies in Crohn’s disease. APMIS 1990; 19 (suppl): 98. 6. Jorgensen BB, Wiik A, Hoier-Madsen M, et al. Serum antibodies to neutrophils in patients with chronic inflammatory bowel disease. ANCA Workshop, Washington, 1990: abstr 66.
Alarm bells for enuresis SIR,-In your March 2 editorial you omit two important points. Parents should be asked if the child sleeps very deeply and whether bedwetting is more likely when the child is very tired. I believe that bedwetting occurs merely because the child is too deeply asleep to be woken by a full bladder. In the Lancet series Disabilities and how to live with them many years ago you published the story of a hospital sister who had this complaint. She remarked that it was more likely to happen after a tiring day. In the case of a child who had double incontinence a small dose of amphetamine prevented anal incontinence, and a larger dose also prevented urinary incontinence. It should always be made clear to parents that nocturnal enuresis does not result from laziness and is not the child’s fault. "Shame therapy" is cruel and causes the child much misery-furthermore it does not work. It is difficult to know whether alarms work well. Does the child get better spontaneously? Are parents reluctant to upset the caring doctor by telling him that his treatment does not work? Or have they gone to see another doctor?
Fig Left
panel,
1-Intact PTH with calcitriol treatment.
PTH 50-200
pmol/I (n=5); right panel, PTH
