J. Paediatr. Child Health (1990) 26, 150-151
False negative results on newborn screening for cystic fibrosis R. L. HENRY, T. J.C. BOULTON and L.G. RODDICK Department of Paediatrics, University of Newcastle and NewcasNe Mater Misericordiae Hospital, Waratah, New South Wales, Australia
Abstract Over a 5 year period in Newcastle, 18 new cases of cystic fibrosis (CF) were diagnosed in children who had been screened in the newborn period. In six of these children, the screening programme failed. Four of these children had a normal screen and an additional two had elevated immunoreactive trypsin (IRT), but there were problems with the notification procedure. Three of the children missed by the screening process had a significantly delayed diagnosis; in all three cases the diagnosis of CF was suspected clinically, but a sweat test was delayed because of false reassurance from the fact that the child had been screened for CF. In a fourth case, multiple elevated sweat electrolyte levels were obtained, but the diagnosis of CF was considered to be in doubt because of the normal IRT assay. A sweat test should be performed on any child in whom there is clincial suspicion of CF. Key words:
cystic fibrosis; newborn screening
Neonatal screening for cystic fibrosis (CF) by measurement of immunoreactive trypsin (IRT) in blood spot samples was introduced into New South Wales in the early 1980s.' Reports of this procedure have been published from several Much of the debate about these programmes has focused on whether the natural history of CF is modified favourably by early acceptable treatment6 Short-term benefits from screening for CF have been documented,' but there is no conclusive evidence that early diagnosis produces improvement in surviva~.~*~ In spite of the unresolved issue of whether newborn screening for CF increases survival, our impression is that the programme has been well accepted by paediatricians in New South Wales. The clinical presentation of new cases of CF in older children with growth failure, fat malabsorption and recurrent lower respiratory infections has become unusual. However we felt that some paediatricians had started to regard the IRT screening programme as a diagnostic test rather than a screen. The aim of this report is to highlight the need to remain aware of the possibility of CF in any child who has classical signs and symptoms.
METHODS The CF Clinic at the Newcastle Mater Misericordiae Hospital is the only referral centre for CF in a region which has between 7000 and 8000 deliveries per year. In the 5 year period 1984-89, 18 new cases of CF were diagnosed in children who had had blood collected in the neonatal period as part of the newborn
Correspondence: Dr R. L. Henry, University of Newcastle, Newcastle Mater Misericordiae Hospital, Waratah, NSW 2298, Australia. R. L. Henry, MB, BS. Dip. Clin. Epid.. FRACP, Senior Lecturer in Paediatrics; T. C. Boulton, MD, FRACP, Professor of Paediatrics; L. G. Roddick, MB, BS, Dip. Obs., FRACP. Consultant Paediatrician Accepted for publication 11 April 1990.
screening programme. These 18 children excluded children who had been born outside NSW (including a boy who had been missed on newborn screening in another State) and a girl who was born before newborn screening was introduced. In all cases the diagnosis of CF was confirmed by compatible clinical features and by a sweat test, with values of sweat chloride greater than 60 mmol/L.
RESULTS Six of the 18 cases of CF (33%) were missed by the screening procedures. Table 1 summarizes the clinical features at presentation and the results of the newborn screening test. The first child was born with congenital hydrocephalus and had a ventriculoperitoneal shunt inserted soon after birth. He had a long history of decreased growth velocity, with weight
Table 1 Children with cystic fibrosis missed by newborn screening procedure Case Age at diagnosis
Clinical features at diagnosis
Results of newborn screening test
1
2 years
F77,* diarrhoea
Normal
2
20 months
Diarrhoea, cough, salty taste
Normal
3
3 months
F77,' pneumonia, diarrhoea
Abnormal
4
7 weeks
Diarrhoea, pneumonia, F77,' salty taste
Normal
5
6 weeks
F77.' diarrhoea
Abnormal
6
2 days
Meconium ileus
Normal
' F77: Failure to thrive.
151
Cystic fibrosis screening
well below the third percentile and length below the third percentile until the diagnosis of CF was made by sweat test at 2 years of age. Diarrhoea was present and he had fat globules in his stools. He had been seen in two tertiary referral paediatric centres and studies of growth hormone secretion and a small bowel biopsy had been undertaken before a sweat test was performed. Review of his IRT revealed the result was within the laboratory normal range. The second case was diagnosed at 20 months. Her IRT screen was normal. At 3 weeks of age she was admitted to hospital with bloody diarrhoea and was found to have rotavirus in her stools. Subsequently her mother noted a salty taste to her sweat, frequent, loose offensive bulky stools that were difficult to flush and a persistent cough. At the age of 6 months an Early Childhood Nursing Sister suggested the diagnosis of CF but a doctor dismissed this suggestion. Her 3 day faecal fat excretion was measured at diagnosis and was 1 8 9 (normal range < 1.5 glday). The third child had been adopted at 2 months of age. She presented at 3 months with pneumonia, poor growth and a long history of frequent loose fatty stools. Prior to adoption, her foster mother had called her 'Winnie the Pooh' and had sought medical advice. The diagnosis of CF was entertained but the paediatric resident was reassured by the fact that newborn screening had occurred. In fact the IRT was abnormal. Although the hospital where she had been born had been notified of the abnormality, no further action had been taken and the adoptive agency was unaware of the abnormal screening test. The fourth child had loose stools from birth. At 5 weeks of age she was admitted to hospital with pneumonia. Chlamydia trachomatis and rotavirus were isolated from her nasopharynx and stool respectively. She was failing to thrive. Two weeks later she was re-admitted with a second episode of pneumonia. At this time the history of a salty taste was elicited. A sweat test was diagnostic of CF. Her first IRT was abnormal but a repeat was normal and thus she was deemed to be normal. After the second episode of pneumonia she had a series of abnormal sweat tests. Fat globules were present in her stools. Her paediatrician was under the false impression that false negatives on the newborn blood spot screen were virtually non-existent. The fifth case was a girl who presented to an Early Childhood Health Clinic at 6 weeks of age with a weight of 2900 g, 200 g less than her birthweight. She had a history of loose stools. A paediatric registrar suspected the diagnosis of CF, which was confirmed by sweat test. Her newborn IRT had been abnormal and a request for a repeat specimen had been sent 5 weeks previously. Problems with hospital administration procedures meant that the child could not be traced. A second IRT was measured after CF had been diagnosed and was abnormal.
The sixth patient was a girl who presented with meconium ileus and in whom the diagnosis of CF was suspected immediately. The IRT was normal.
DISCUSSION One third of new cases of CF diagnosed over a 5 year period presented problems for the newborn blood spot screen. Two were administrative, in that the laboratory correctly identified an abnormality but this information was not acted upon. In one of these cases the child was adopted. Four of the false negative cases were missed by the laboratory. Two had markedly delayed diagnoses and a third had suffered two episodes of pneumonia before CF was diagnosed. For the child with meconium ileus, the newborn screen was irrelevant, although it was worth noting that if the child had died before a sweat test had been performed, the possibility exists that incorrect genetic advice might have been given on the basis of the normal IRT. We feel that it is most unlikely that problems with newborn screening for CF are limited to the Hunter region. Indeed two of the six children had been seen in Sydney hospitals. The problems that occurred with some of these children serve to reinforce the benefits of early diagnosis in avoiding emotional trauma and physical d i ~ e a s e However .~ we need to be aware that the screening procedure is fallible and that a sweat test should be performed in any child in whom there is a clinical suspicion of cystic fibrosis. This is the only way to minimize the impact of false negatives from the newborn screening programme.
REFERENCES Wilken B , Brown A. R. D.. Urwin R. et al. Cystic fibrosis screening by dried blood spot trypsin assay. Results in 75 000 newborn infants. J. Pediatr 1983: 102: 383-7. Lyon I.C. T , Crossley J. R.. Smith P. A. Screening for cystic fibrosis. N.2 Med J. 1983; 96: 673-5. Wesley A. W , Smith P. A.. Elliottt R B. Experience with neonatal screening for cystic fibrosis in New Zealand using measurement of irnmunoreactive trypsinogen. Aust. Paediatr. J. 1989; 25: 151-5 Cassio A., Bernardi F.. Piazzi S. et ai. Neonatal screening for cystic fibrosis by dried blood spot trypsin assay. Acta Paediatr. Scand, 1984; 73: 554-8 Healey A. F., Heeley M. E.,King D. N. et al. Screening for cystic fibrosis by dried blood spot trypsin assay. Arch. Dis. Child. 1982, 57' 18-21 Wilson J. M. G.. Jungner G. Principles and Practice of Screening for Disease. World Health Organization, Geneva 1966. Wilcken B., Towns S. J., Mellis C. M. Diagnostic delay in cystic fibrosis. Lessons from new born Screening. Arch. Dis. Chiid. 1983; 58: 863-6. Hudson I., Phelan P. D. Are sex, age at diagnosis or mode of presentation prognostic factors for cystic fibrosis? Paediatr. Pulmonol. 1987: 3: 288-97.
False negative results on newborn screening for cystic fibrosis R. L. HENRY, T. J.C. BOULTON and L.G. RODDICK Department of Paediatrics, University of Newcastle and NewcasNe Mater Misericordiae Hospital, Waratah, New South Wales, Australia
Abstract Over a 5 year period in Newcastle, 18 new cases of cystic fibrosis (CF) were diagnosed in children who had been screened in the newborn period. In six of these children, the screening programme failed. Four of these children had a normal screen and an additional two had elevated immunoreactive trypsin (IRT), but there were problems with the notification procedure. Three of the children missed by the screening process had a significantly delayed diagnosis; in all three cases the diagnosis of CF was suspected clinically, but a sweat test was delayed because of false reassurance from the fact that the child had been screened for CF. In a fourth case, multiple elevated sweat electrolyte levels were obtained, but the diagnosis of CF was considered to be in doubt because of the normal IRT assay. A sweat test should be performed on any child in whom there is clincial suspicion of CF. Key words:
cystic fibrosis; newborn screening
Neonatal screening for cystic fibrosis (CF) by measurement of immunoreactive trypsin (IRT) in blood spot samples was introduced into New South Wales in the early 1980s.' Reports of this procedure have been published from several Much of the debate about these programmes has focused on whether the natural history of CF is modified favourably by early acceptable treatment6 Short-term benefits from screening for CF have been documented,' but there is no conclusive evidence that early diagnosis produces improvement in surviva~.~*~ In spite of the unresolved issue of whether newborn screening for CF increases survival, our impression is that the programme has been well accepted by paediatricians in New South Wales. The clinical presentation of new cases of CF in older children with growth failure, fat malabsorption and recurrent lower respiratory infections has become unusual. However we felt that some paediatricians had started to regard the IRT screening programme as a diagnostic test rather than a screen. The aim of this report is to highlight the need to remain aware of the possibility of CF in any child who has classical signs and symptoms.
METHODS The CF Clinic at the Newcastle Mater Misericordiae Hospital is the only referral centre for CF in a region which has between 7000 and 8000 deliveries per year. In the 5 year period 1984-89, 18 new cases of CF were diagnosed in children who had had blood collected in the neonatal period as part of the newborn
Correspondence: Dr R. L. Henry, University of Newcastle, Newcastle Mater Misericordiae Hospital, Waratah, NSW 2298, Australia. R. L. Henry, MB, BS. Dip. Clin. Epid.. FRACP, Senior Lecturer in Paediatrics; T. C. Boulton, MD, FRACP, Professor of Paediatrics; L. G. Roddick, MB, BS, Dip. Obs., FRACP. Consultant Paediatrician Accepted for publication 11 April 1990.
screening programme. These 18 children excluded children who had been born outside NSW (including a boy who had been missed on newborn screening in another State) and a girl who was born before newborn screening was introduced. In all cases the diagnosis of CF was confirmed by compatible clinical features and by a sweat test, with values of sweat chloride greater than 60 mmol/L.
RESULTS Six of the 18 cases of CF (33%) were missed by the screening procedures. Table 1 summarizes the clinical features at presentation and the results of the newborn screening test. The first child was born with congenital hydrocephalus and had a ventriculoperitoneal shunt inserted soon after birth. He had a long history of decreased growth velocity, with weight
Table 1 Children with cystic fibrosis missed by newborn screening procedure Case Age at diagnosis
Clinical features at diagnosis
Results of newborn screening test
1
2 years
F77,* diarrhoea
Normal
2
20 months
Diarrhoea, cough, salty taste
Normal
3
3 months
F77,' pneumonia, diarrhoea
Abnormal
4
7 weeks
Diarrhoea, pneumonia, F77,' salty taste
Normal
5
6 weeks
F77.' diarrhoea
Abnormal
6
2 days
Meconium ileus
Normal
' F77: Failure to thrive.
151
Cystic fibrosis screening
well below the third percentile and length below the third percentile until the diagnosis of CF was made by sweat test at 2 years of age. Diarrhoea was present and he had fat globules in his stools. He had been seen in two tertiary referral paediatric centres and studies of growth hormone secretion and a small bowel biopsy had been undertaken before a sweat test was performed. Review of his IRT revealed the result was within the laboratory normal range. The second case was diagnosed at 20 months. Her IRT screen was normal. At 3 weeks of age she was admitted to hospital with bloody diarrhoea and was found to have rotavirus in her stools. Subsequently her mother noted a salty taste to her sweat, frequent, loose offensive bulky stools that were difficult to flush and a persistent cough. At the age of 6 months an Early Childhood Nursing Sister suggested the diagnosis of CF but a doctor dismissed this suggestion. Her 3 day faecal fat excretion was measured at diagnosis and was 1 8 9 (normal range < 1.5 glday). The third child had been adopted at 2 months of age. She presented at 3 months with pneumonia, poor growth and a long history of frequent loose fatty stools. Prior to adoption, her foster mother had called her 'Winnie the Pooh' and had sought medical advice. The diagnosis of CF was entertained but the paediatric resident was reassured by the fact that newborn screening had occurred. In fact the IRT was abnormal. Although the hospital where she had been born had been notified of the abnormality, no further action had been taken and the adoptive agency was unaware of the abnormal screening test. The fourth child had loose stools from birth. At 5 weeks of age she was admitted to hospital with pneumonia. Chlamydia trachomatis and rotavirus were isolated from her nasopharynx and stool respectively. She was failing to thrive. Two weeks later she was re-admitted with a second episode of pneumonia. At this time the history of a salty taste was elicited. A sweat test was diagnostic of CF. Her first IRT was abnormal but a repeat was normal and thus she was deemed to be normal. After the second episode of pneumonia she had a series of abnormal sweat tests. Fat globules were present in her stools. Her paediatrician was under the false impression that false negatives on the newborn blood spot screen were virtually non-existent. The fifth case was a girl who presented to an Early Childhood Health Clinic at 6 weeks of age with a weight of 2900 g, 200 g less than her birthweight. She had a history of loose stools. A paediatric registrar suspected the diagnosis of CF, which was confirmed by sweat test. Her newborn IRT had been abnormal and a request for a repeat specimen had been sent 5 weeks previously. Problems with hospital administration procedures meant that the child could not be traced. A second IRT was measured after CF had been diagnosed and was abnormal.
The sixth patient was a girl who presented with meconium ileus and in whom the diagnosis of CF was suspected immediately. The IRT was normal.
DISCUSSION One third of new cases of CF diagnosed over a 5 year period presented problems for the newborn blood spot screen. Two were administrative, in that the laboratory correctly identified an abnormality but this information was not acted upon. In one of these cases the child was adopted. Four of the false negative cases were missed by the laboratory. Two had markedly delayed diagnoses and a third had suffered two episodes of pneumonia before CF was diagnosed. For the child with meconium ileus, the newborn screen was irrelevant, although it was worth noting that if the child had died before a sweat test had been performed, the possibility exists that incorrect genetic advice might have been given on the basis of the normal IRT. We feel that it is most unlikely that problems with newborn screening for CF are limited to the Hunter region. Indeed two of the six children had been seen in Sydney hospitals. The problems that occurred with some of these children serve to reinforce the benefits of early diagnosis in avoiding emotional trauma and physical d i ~ e a s e However .~ we need to be aware that the screening procedure is fallible and that a sweat test should be performed in any child in whom there is a clinical suspicion of cystic fibrosis. This is the only way to minimize the impact of false negatives from the newborn screening programme.
REFERENCES Wilken B , Brown A. R. D.. Urwin R. et al. Cystic fibrosis screening by dried blood spot trypsin assay. Results in 75 000 newborn infants. J. Pediatr 1983: 102: 383-7. Lyon I.C. T , Crossley J. R.. Smith P. A. Screening for cystic fibrosis. N.2 Med J. 1983; 96: 673-5. Wesley A. W , Smith P. A.. Elliottt R B. Experience with neonatal screening for cystic fibrosis in New Zealand using measurement of irnmunoreactive trypsinogen. Aust. Paediatr. J. 1989; 25: 151-5 Cassio A., Bernardi F.. Piazzi S. et ai. Neonatal screening for cystic fibrosis by dried blood spot trypsin assay. Acta Paediatr. Scand, 1984; 73: 554-8 Healey A. F., Heeley M. E.,King D. N. et al. Screening for cystic fibrosis by dried blood spot trypsin assay. Arch. Dis. Child. 1982, 57' 18-21 Wilson J. M. G.. Jungner G. Principles and Practice of Screening for Disease. World Health Organization, Geneva 1966. Wilcken B., Towns S. J., Mellis C. M. Diagnostic delay in cystic fibrosis. Lessons from new born Screening. Arch. Dis. Chiid. 1983; 58: 863-6. Hudson I., Phelan P. D. Are sex, age at diagnosis or mode of presentation prognostic factors for cystic fibrosis? Paediatr. Pulmonol. 1987: 3: 288-97.
