Thirdly, tamoxifen is known to reduce the incidence of local relapse after surgery, and this may militate against the need for radiotherapy after local excision while simultaneously reducing the incidence of contralateral breast cancer.4 The alternative policy advocated by Drs Hamilton and Buchanan-close observationwould certainly tell us something of the natural course of the disease and establish precisely the incidence of local relapse after local excision alone. But in 10 years or more we would be no further on. The beauty of the two x two factorial trial is that it not only allows us to study the natural course of the disease but also shows whether one of two approaches might reduce the incidence of local relapse while allowing precise estimation of the cost-benefit ratios. MICHAEL BAUM
Cancer Research Campaign, Clinical Trials Centre, King's College School of Medicine and Dentistry, London SE5 9NV 1 Hamilton CR, Buchanan RB. Radiotherapy for ductal carcinoma in situ detected by screening. Br Med J 1990;301:224-5. (28 July.) 2 Schnitt EJ, Silen W, Sadowsky NL, Connolly JL, Harris JR. Ductal carcinoma in situ (intraductal carcinoma) of the breast. N EnglJMed 1988;318:898-903. 3 Haybittle JL, Brinkley D, Houghton J, A'Hern RP, Baum M. Postoperative radiotherapy and late mortality: evidence from the Cancer Research Campaign trial for early breast cancer.
BrMedJ7 1989;298:1611-4. 4 Cusick J, Baum M. Tamoxifen and contralateral breast cancer. Lancet 1985;ii:282.
Fractures of long bones occurring in neonatal intensive therapy units SIR,-Although Drs Rachel R Phillips and Stephen H Lee show that fractures may occur during invasive procedures on very premature infants, they are unnecessarily cautious in recommending that "the utmost delicacy must be used when handling these infants because the slightest trauma may cause fractures."' Most parents share this misconception about the fragility of very premature infants and are initially reluctant to handle their tiny offspring. Premature infants are deceptively resistant to trauma and are much less likely than term infants to suffer fractures during delivery.2 Helping parents to overcome their fear of handling a premature infant is an important aspect of neonatal care which has considerable benefits for the infant. Close physical contact between very premature infants and their parents improves lactation, psychological bonding, and even the infants' respiratory state. These advantages supplement intensive care in developed countries and are actually life saving in developing countries.3 JOHN GIBBS Child Development Centre, Liverpool L12 2AP 1 Phillips RR, Lee SH. Fractures of long bones occurring in neonatal intensive therapy units. Br Med J 1990;301:225-6. (28 July.) 2 Hensinger RN, Jones ET. Orthopaedic problems in the newborn. In: Roberton NRC, ed. Textbook of neonatology. Edinburgh: Churchill Livingstone, 1986:696-720. 3 Whitelaw A. Kangaroo baby care: just a nice experience or an important advance for preterm infants? Pediatrics 1990;85:
604-5.
Bleeding time in patients with hepatic cirrhosis SIR,-Dr J C Blake and colleagues conclude that "bleeding time should be measured when assessing patients for invasive procedures who have a raised
442
bilirubin concentration or poor hepatic function."' Their results cannot justify this conclusion. Defective primary haemostatic mechanisms have long been recognised in liver disease,2 and Dr Blake and colleagues merely confirm this. They made no attempt to show that a prolonged bleeding time is predictive of the risk of haemorrhage from an invasive procedure. Although bleeding time is valuable in investigating haemorrhagic disorders, a recent extensive review questioned the evidence that it is a useful guide to haemorrhagic risk in disease.' Only prospective assessment of the relation between tests of coagulation and haemostasis and occurrence ofhaemorrhage after invasive procedures will resolve the question. We are currently conducting such a study and have been prompted to review our interim results. We have so far measured standardised skin bleeding time (using a Simplate device)2 and done coagulation screening tests in 30 subjects before they underwent percutaneous liver biopsy for diagnostic purposes. Bleeding was assessed by computed tomography of the upper abdomen 24 hours after the procedure as well as by clinical observation and serial measurement of packed cell volume. In five subjects bleeding time was prolonged (>8 minutes, range 8-5-19), and in nine thrombocytopenia was present (platelet count 17 seconds, range 19-33). In two patients with prolonged prothrombin time vitamin K was administered and in one coagulation factor concentrates were given. No excessive haemorrhage or reduction in packed cell volume occurred and all computed tomograms were negative for haematoma formation. These findings are consistent with the low incidence of bleeding after various invasive procedures that has been reported in liver disease.4 Though we share the concern of Dr Blake and colleagues with regard to the haemorrhagic risk in subjects with impaired liver function, we would suggest that recommending the routine performance of a test of unproved predictive value, particularly without comment on its interpretation or any suggestion as to appropriate action, is unhelpful. Further information is required before this can be justified. M GREAVES R NAKIELNY K K HAMPTON F E PRESTON D R TRIGER
Royal Hallamshire Hospital, Sheffield S1O 2JF
All three studies have screened outpatients at high risk-because they were mainly from genitourinary clinics,' because of their relationship to an index patient positive for antibodies to HTLV-I with tropical spastic paraparesis,4 or because they were pregnant.2 An overview of these shows that none of the 631 Afro-Caribbeans born in the United Kingdom were seropositive compared with 16 out of 461 (3-6%, 95% confidence interval 3-2 to 3-8%) of those in the United Kingdom who had been born in the West Indies (table). This is roughly the prevalence expected in people of this age from Jamaican data. Prevalence of antibodies to HTLV-I in Afro-Caribbean people living in United Kingdom according to place of birth Born in United Kingdom
Born in Caribbean
No positive
No tested
No positive
No tested
Mowbray et alt Cruickshank et al4 Toswill etal'
0 0 0
403 14 214
7 4 5
192 9 260
Total
0
631
16
461
Source
Certainly, British born Afro-Caribbeans can be positive for antibodies to HTLV-I, as evidenced by two such relatives of patients with leukaemial and two of patients with paraparesis' in incompletely ascertained families, but it is much less common. Though there are still insufficient data, this important disparity is providing clues to transmission. But the puzzle has not yet been solved. Breast feeding, seemingly equally prevalent among Afro-Caribbean mothers wherever they were born,6 does not seem to be the only (or main) perinatal route of infection. Clearly, a centrally coordinated and planned approach to collecting data prospectively would be most helpful. Even in an area of low prevalence work on this virus has already thrown considerable light not only on the aetiology of one tumour (T cell leukaemia) but also on the neuropathology of the spastic paraparesis, which may be regarded as the tropical equivalent of multiple sclerosis-a disease so far still in search of its retrovirus. Despite this progress treatment for all three conditions remains ineffective. J K CRUICKSHANK Northwick Park Hospital/Clinical Research Centre, Harrow HAl 3UJ I Weber J. HTLV-I infection in Britain. BrMedJ 1990;301:71-2.
I Blake JC, Sprengers D, Grech P, McCormick PA, McIntyre N,
Burroughs AK. Bleeding time in patients with hepatic cirrhosis. BrMedJ 1990;301:12-5. (7 July.) 2 Greaves M, Preston FE. The laboratory investigation of acquired and congenital platelet disorders. In: Thomson JM, ed. Blood coagulation and haemostasis: a practical guide. 3rd ed. Edinburgh: Churchill Livingstone, 1985. 3 Rodgers RPC. A critical reappraisal of the bleeding time. Semin ThromnbHaemost 1990;16:1-138. 4 Friedman EW, Sussman II. Safety of invasive procedures in patients with the coagulopathy of liver disease. Clin Lab Haematol 1989;11: 199-204.
(14 July.) 2 Tosswill JHC, Ades AE, Peckham C, Mortimer P, Weber JN. Infection with human T cell leukaemia/lymphoma virus type I in patients attending an antenatal clinic in London. Br MedJ
1990;301:95-6. (14 July.) 3 Mowbray J, Morson S, Chawira A, et al. Epidemiology of HTLV-I infections in a sub-population of Afro-Caribbean origin in England.J Med Virol 1989;29:289-95. 4 Cruickshank JK, Richardson JH, Morgan OStC, et al. Screening for prolonged incubation of HTLV-I infection in British and Jamaican relatives of British patients with tropical spastic
paraparesis. BrMedJ7 1990;300:300-4. (3 February.)
5 Matutes E, Dalgleish AG, Weiss RA, Joseph AP, Catovsky D. Studies in HTLV-I carriers from the Caribbean. Int J Cancer
1986;38:41-5.
HTLV-I infection in Britain
6 Douglas J. Food type preferences and trends among AfroCaribbeans in Britain. In: Cruickshank JK, Beevers DG, eds. Ethnic factors in health and disease. Sevenoaks: Butterworths, 1989.
SIR,-Dr Jonathan Weber's editorial discusses the current position and limitations of data on infection with human T cell leukaemia/lymphoma virus type I (HTLV-I) in Britain.' He does not refer to the paper in the same issue (of which he is an author) looking at antenatal screening of mothers at potential risk.2 This is the third study in the past year documenting an important unexplained paradox that people of Afro-Caribbean origin born in the United Kingdom have considerably less risk of carrying the virus than AfroCaribbeans in Britain born in the West Indies.
SIR,-Dr Jonathan Weber stressed the need for active surveillance for sporadic cases of infection with human T cell leukaemia/lymphoma virus type I (HTLV-I) in Britain.' We suggest that recurrent infection with Strongyloides stercoralis is an indication for screening for HTLV-I. A 39 year old West Indian woman presented in January 1988 with abdominal pain and vomiting. Small bowel obstruction was diagnosed radioBMJ
VOLUME 301
1 SEPTEMBER 1990
Cancer Research Campaign, Clinical Trials Centre, King's College School of Medicine and Dentistry, London SE5 9NV 1 Hamilton CR, Buchanan RB. Radiotherapy for ductal carcinoma in situ detected by screening. Br Med J 1990;301:224-5. (28 July.) 2 Schnitt EJ, Silen W, Sadowsky NL, Connolly JL, Harris JR. Ductal carcinoma in situ (intraductal carcinoma) of the breast. N EnglJMed 1988;318:898-903. 3 Haybittle JL, Brinkley D, Houghton J, A'Hern RP, Baum M. Postoperative radiotherapy and late mortality: evidence from the Cancer Research Campaign trial for early breast cancer.
BrMedJ7 1989;298:1611-4. 4 Cusick J, Baum M. Tamoxifen and contralateral breast cancer. Lancet 1985;ii:282.
Fractures of long bones occurring in neonatal intensive therapy units SIR,-Although Drs Rachel R Phillips and Stephen H Lee show that fractures may occur during invasive procedures on very premature infants, they are unnecessarily cautious in recommending that "the utmost delicacy must be used when handling these infants because the slightest trauma may cause fractures."' Most parents share this misconception about the fragility of very premature infants and are initially reluctant to handle their tiny offspring. Premature infants are deceptively resistant to trauma and are much less likely than term infants to suffer fractures during delivery.2 Helping parents to overcome their fear of handling a premature infant is an important aspect of neonatal care which has considerable benefits for the infant. Close physical contact between very premature infants and their parents improves lactation, psychological bonding, and even the infants' respiratory state. These advantages supplement intensive care in developed countries and are actually life saving in developing countries.3 JOHN GIBBS Child Development Centre, Liverpool L12 2AP 1 Phillips RR, Lee SH. Fractures of long bones occurring in neonatal intensive therapy units. Br Med J 1990;301:225-6. (28 July.) 2 Hensinger RN, Jones ET. Orthopaedic problems in the newborn. In: Roberton NRC, ed. Textbook of neonatology. Edinburgh: Churchill Livingstone, 1986:696-720. 3 Whitelaw A. Kangaroo baby care: just a nice experience or an important advance for preterm infants? Pediatrics 1990;85:
604-5.
Bleeding time in patients with hepatic cirrhosis SIR,-Dr J C Blake and colleagues conclude that "bleeding time should be measured when assessing patients for invasive procedures who have a raised
442
bilirubin concentration or poor hepatic function."' Their results cannot justify this conclusion. Defective primary haemostatic mechanisms have long been recognised in liver disease,2 and Dr Blake and colleagues merely confirm this. They made no attempt to show that a prolonged bleeding time is predictive of the risk of haemorrhage from an invasive procedure. Although bleeding time is valuable in investigating haemorrhagic disorders, a recent extensive review questioned the evidence that it is a useful guide to haemorrhagic risk in disease.' Only prospective assessment of the relation between tests of coagulation and haemostasis and occurrence ofhaemorrhage after invasive procedures will resolve the question. We are currently conducting such a study and have been prompted to review our interim results. We have so far measured standardised skin bleeding time (using a Simplate device)2 and done coagulation screening tests in 30 subjects before they underwent percutaneous liver biopsy for diagnostic purposes. Bleeding was assessed by computed tomography of the upper abdomen 24 hours after the procedure as well as by clinical observation and serial measurement of packed cell volume. In five subjects bleeding time was prolonged (>8 minutes, range 8-5-19), and in nine thrombocytopenia was present (platelet count 17 seconds, range 19-33). In two patients with prolonged prothrombin time vitamin K was administered and in one coagulation factor concentrates were given. No excessive haemorrhage or reduction in packed cell volume occurred and all computed tomograms were negative for haematoma formation. These findings are consistent with the low incidence of bleeding after various invasive procedures that has been reported in liver disease.4 Though we share the concern of Dr Blake and colleagues with regard to the haemorrhagic risk in subjects with impaired liver function, we would suggest that recommending the routine performance of a test of unproved predictive value, particularly without comment on its interpretation or any suggestion as to appropriate action, is unhelpful. Further information is required before this can be justified. M GREAVES R NAKIELNY K K HAMPTON F E PRESTON D R TRIGER
Royal Hallamshire Hospital, Sheffield S1O 2JF
All three studies have screened outpatients at high risk-because they were mainly from genitourinary clinics,' because of their relationship to an index patient positive for antibodies to HTLV-I with tropical spastic paraparesis,4 or because they were pregnant.2 An overview of these shows that none of the 631 Afro-Caribbeans born in the United Kingdom were seropositive compared with 16 out of 461 (3-6%, 95% confidence interval 3-2 to 3-8%) of those in the United Kingdom who had been born in the West Indies (table). This is roughly the prevalence expected in people of this age from Jamaican data. Prevalence of antibodies to HTLV-I in Afro-Caribbean people living in United Kingdom according to place of birth Born in United Kingdom
Born in Caribbean
No positive
No tested
No positive
No tested
Mowbray et alt Cruickshank et al4 Toswill etal'
0 0 0
403 14 214
7 4 5
192 9 260
Total
0
631
16
461
Source
Certainly, British born Afro-Caribbeans can be positive for antibodies to HTLV-I, as evidenced by two such relatives of patients with leukaemial and two of patients with paraparesis' in incompletely ascertained families, but it is much less common. Though there are still insufficient data, this important disparity is providing clues to transmission. But the puzzle has not yet been solved. Breast feeding, seemingly equally prevalent among Afro-Caribbean mothers wherever they were born,6 does not seem to be the only (or main) perinatal route of infection. Clearly, a centrally coordinated and planned approach to collecting data prospectively would be most helpful. Even in an area of low prevalence work on this virus has already thrown considerable light not only on the aetiology of one tumour (T cell leukaemia) but also on the neuropathology of the spastic paraparesis, which may be regarded as the tropical equivalent of multiple sclerosis-a disease so far still in search of its retrovirus. Despite this progress treatment for all three conditions remains ineffective. J K CRUICKSHANK Northwick Park Hospital/Clinical Research Centre, Harrow HAl 3UJ I Weber J. HTLV-I infection in Britain. BrMedJ 1990;301:71-2.
I Blake JC, Sprengers D, Grech P, McCormick PA, McIntyre N,
Burroughs AK. Bleeding time in patients with hepatic cirrhosis. BrMedJ 1990;301:12-5. (7 July.) 2 Greaves M, Preston FE. The laboratory investigation of acquired and congenital platelet disorders. In: Thomson JM, ed. Blood coagulation and haemostasis: a practical guide. 3rd ed. Edinburgh: Churchill Livingstone, 1985. 3 Rodgers RPC. A critical reappraisal of the bleeding time. Semin ThromnbHaemost 1990;16:1-138. 4 Friedman EW, Sussman II. Safety of invasive procedures in patients with the coagulopathy of liver disease. Clin Lab Haematol 1989;11: 199-204.
(14 July.) 2 Tosswill JHC, Ades AE, Peckham C, Mortimer P, Weber JN. Infection with human T cell leukaemia/lymphoma virus type I in patients attending an antenatal clinic in London. Br MedJ
1990;301:95-6. (14 July.) 3 Mowbray J, Morson S, Chawira A, et al. Epidemiology of HTLV-I infections in a sub-population of Afro-Caribbean origin in England.J Med Virol 1989;29:289-95. 4 Cruickshank JK, Richardson JH, Morgan OStC, et al. Screening for prolonged incubation of HTLV-I infection in British and Jamaican relatives of British patients with tropical spastic
paraparesis. BrMedJ7 1990;300:300-4. (3 February.)
5 Matutes E, Dalgleish AG, Weiss RA, Joseph AP, Catovsky D. Studies in HTLV-I carriers from the Caribbean. Int J Cancer
1986;38:41-5.
HTLV-I infection in Britain
6 Douglas J. Food type preferences and trends among AfroCaribbeans in Britain. In: Cruickshank JK, Beevers DG, eds. Ethnic factors in health and disease. Sevenoaks: Butterworths, 1989.
SIR,-Dr Jonathan Weber's editorial discusses the current position and limitations of data on infection with human T cell leukaemia/lymphoma virus type I (HTLV-I) in Britain.' He does not refer to the paper in the same issue (of which he is an author) looking at antenatal screening of mothers at potential risk.2 This is the third study in the past year documenting an important unexplained paradox that people of Afro-Caribbean origin born in the United Kingdom have considerably less risk of carrying the virus than AfroCaribbeans in Britain born in the West Indies.
SIR,-Dr Jonathan Weber stressed the need for active surveillance for sporadic cases of infection with human T cell leukaemia/lymphoma virus type I (HTLV-I) in Britain.' We suggest that recurrent infection with Strongyloides stercoralis is an indication for screening for HTLV-I. A 39 year old West Indian woman presented in January 1988 with abdominal pain and vomiting. Small bowel obstruction was diagnosed radioBMJ
VOLUME 301
1 SEPTEMBER 1990
