209
EDITORIALS
allow women choices about
make and a to the birth of establishing programme prevent affected babies and thereby save money for the taxpayer. The first type of programme, evaluated by Watson et al (p 217), is based in primary care, and individuals are screened before they have to make reproductive decisions. Screening via this route can adopt several forms, and can be undertaken by invitation, through schools medical services, or programme
informed
to
couples to reproduction,
or
opportunistically at general practice surgeries or family-planning clinics. Watson et a15 had previously found that screening by invitation achieves a much lower uptake (10%) than screening through family planning clinics or general surgeries. Advantages of
Screening for cystic fibrosis Since the discovery of the genetic locus responsible for cystic fibrosis (CF) in 1989,1-3 and the subsequent characterisation of most of the mutations responsible for producing the disease, there have been many calls for general population screening and numerous suggestions for ways of implementing such a service.4,5 In this issue there are two reports about different approaches to screening. Watson and colleagues (p 217) studied a primary care population whereas Mennie and co-workers (p 214) chose the more familiar setting of the antenatal clinic. In most discussions about screening the general assumption is that, because a test is possible, testing should be implemented as a service; the moral, ethical, social, and economic issues that underpin the introduction of such a service tend to be overshadowed by questions of logistics. In the USA, the Office of Technology Assessment report on CF screening due out shortly will take no position on whether screening ought to be instituted on a massive scale. The screening test for CF is based on mouthwash samples of cells whose DNA is amplified, digested, and subjected to separation techniques that identify three or four common mutations. In the UK about 85% of cases of CF would be associated with one or other of these abnormalities. A much larger proportion of cases can be detected by use of more complex techniques and in certain subpopulations, such as that lately described in Brittany.6 However, any screening programme must be considered in the light of the additional costs involved in detecting higher proportions of cases and the gene frequency in the local population. Detection of 85% of mutations means that only 72% of couples at risk of an affected child could be identified. In some countries detection of 85 % of mutations is unrealistic, and the proportion detectable and thus the efficacy of the test may vary within a country. Before a screening programme is established the organisers must make explicit the objectives of the programme. For example, there is a considerable difference between establishing a screening
this type of programme are that the information about CF, counselling, and the test itself are separated from immediate decisions regarding reproduction; this approach allows time for reflection and contemplation free from the worry of having to make immediate decisions about a pregnancy. This strategy has its disadvantages. Because the test is undertaken outside the context of reproduction, its results and meaning may be forgotten when decisions have to be made. Even after counselling, a significant minority were unable to state correctly their risk of having an affected child, even to the extent that 13% believed that two carriers could not have a normal child. In the primary care context, only 57% suggested a test to their partner, and only 87 % of partners underwent a test, so less than half (49%) the couples at higher risk had carrier status determined in both partners, in each case with only 85 % efficacy. Thus only 35 % of couples at risk of having a child with CF were detectable by this screening approach. What about stigmatisation of carriers? Although most people who were screened shared the information with relatives, partners, and friends, a sizeable minority did not. The second model of a screening service-to offer screening to women who are attending antenatal clinics-includes several variations. General practice antenatal clinics7 can be used for earlier diagnosis; and screening may be offered only to couples rather than in two stages (woman first and partner only if woman is positive). This model has the advantage of higher uptake rates, and information is imparted in the direct context in which it is to be used, thereby minimising the risk of misinterpretation, forgetfulness, and stigma. Again, there are disadvantages. Testing that takes place during pregnancy has to be done quickly and decisions on action must be taken speedily to allow termination in the first or early second trimester. Pregnancy itself is a time of stress, especially with a first baby, and this could make the couple act hastily and ultimately regret their decision, particularly if counselling is perceived as directive. Individuals need to be retested if their partner changes. Moreover, screening of couples as a unit prevents individuals’
knowing their status family members be
or recommending that other screened. There is also the
210
difficulty of advising a couple when one member is positive. They cannot be reassured, since their risk of having an affected baby is still 1 in 500 (because of the mutations not included in the screening test); nor can they be offered prenatal diagnosis, since the falsepositive rate would mean that twenty normal pregnancies were terminated for each affected one.4 If the aim were to reduce the incidence of cystic fibrosis, both types of screening programme would require very high response rates to justify their application. In the antenatal clinic strategy, 73% of potentially eligible women were actually screened. In this group, only 85 % of mutations could be detected, and likewise in their partners, so only 52% of couples at potential risk could be detected by this method. Furthermore, Watson et al showed there is considerable uncertainty about termination of an affected pregnancy in the primary care setting. Although the single affected pregnancy in Edinburgh was terminated, one cannot assume that this would always be so; even after counselling, only just over one third of couples would definitely consider termination, half were undecided, and the option of having no children is not available to those already
an
pregnant.
and the costs of medical care for unaffected individuals. Re-evaluation will be needed each time the efficacy of the test is improved and refined. The technical ability and practical skills to implement screening programmes for detection of CF carriers in the general population are now available. We still need to think very carefully about whether nationwide programmes are what we really want, and what the aims and ethical base of such an approach might be.
Testing inevitably results in some anxiety--during the process, during the wait for results, after notification of a positive result, and before the first counselling session. Longer term follow-up of screening in primary care revealed that, for most individuals, anxiety levels return to normal after 6 months. This is a short period of follow-up for an event that may have lifelong implications. Screening in antenatal clinics likewise induces anxiety, which seems to revert to normal when a negative result is obtained for the parmer. There is no knowing whether similar anxiety would be experienced in subsequent pregnancies, or what happens to those couples where both are found to be carriers. Uncertainty about decisions to terminate pregnancy, and about residual anxiety, makes detailed economic evaluation difficult. Nevertheless, crude economic statements-cost comparisons rather than cost-benefit or cost-utility analyses-have been made in both settings. The results suggest that financial savings may accrue to a health service from use of screening programmes, but several important factors have not been evaluated. For example, the cost of caring for all children born with CF whose parents were offered screening but either declined or who had undetectable mutations must be included as a cost of the programme. It is questionable whether patients with CF can be considered wholly as a cost in a cost-benefit analysis. Similarly, can normal individuals be regarded as entirely cost-free? Use of crude cost comparisons may greatly overestimate the value of the programme and underestimates the cost prevented. approach to population screening, it is implicit that, at least for some families, termination of
per
case
For either
affected pregnancy or a decision to remain childless would be preferable to having a child with CF. This decision may reduce the strain placed on families by a child with chronic illness, but what do we know about prognosis and quality of life for people with CF? Although median expectations of life of 20-25 years are usually cited, Britton and Knox8 pointed out that these figures are for a birth cohort of 20-25 years ago, when treatment was less effective, and they suggest that most patients bom with CF today could expect to live well into middle age, working and living independently. Termination of pregnancy is not itself without cost, just as having a child with CF in the family is not all negative. One cannot assume that an affected pregnancy will be followed by an unaffected one; that some families will choose instead to have no further children may itself cause considerable
morbidity. If justification of a screening programme for CF is economic, then decisions should be informed by full cost-benefit analyses which include costs of screening failures, the potential contribution of people with CF to society, loss of normal fetuses through antenatal diagnosis, effects on the family of children with CF,
1.
Rommens JM,Iannuzi MC, Kerem B-S, et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989; 245:
1059-65. 2. Riordan JR, Rommens JM, Kerem B-S, et al. Identification of the cystic fibrosis gene: cloning and characterisation of complementary DNA. Science 1989; 245: 1066-72. 3. Kerem B-S, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245: 1073-80. 4. Wald NJ. Couple screening for cystic fibrosis. Lancet 1991; 338: 1318-19. 5. Watson EK, Mayall E, Chapple J, et al. Screening for carriers of cystic fibrosis through the primary health care services. BMJ 1991; 303: 504-07. 6. Ferec C, Audrezet MP, Mercier B, et al. Detection of over 98% of cystic fibrosis mutations in a Celtic population. Nature Genet 1992; I: 188-91. 7. Harris HJ, Scotcher D, Craufurd D, Wallace A, Harris R. Cystic fibrosis carrier screening at first diagnosis of pregnancy in general practice. Lancet 1992; 339: 1539. 8. Britton J, Knox AJ. Screening for cystic fibrosis. Lancet 1991; 338:1524.
Colorectal cancer: new evidence for the adenoma/carcinoma sequence The debate about the ability of some colorectal adenomas to undergo malignant change has centred largely on epidemiological and histopathological studies.1-3 More refined work, involving analysis of cell lines and genetic changes, is emerging in support
EDITORIALS
allow women choices about
make and a to the birth of establishing programme prevent affected babies and thereby save money for the taxpayer. The first type of programme, evaluated by Watson et al (p 217), is based in primary care, and individuals are screened before they have to make reproductive decisions. Screening via this route can adopt several forms, and can be undertaken by invitation, through schools medical services, or programme
informed
to
couples to reproduction,
or
opportunistically at general practice surgeries or family-planning clinics. Watson et a15 had previously found that screening by invitation achieves a much lower uptake (10%) than screening through family planning clinics or general surgeries. Advantages of
Screening for cystic fibrosis Since the discovery of the genetic locus responsible for cystic fibrosis (CF) in 1989,1-3 and the subsequent characterisation of most of the mutations responsible for producing the disease, there have been many calls for general population screening and numerous suggestions for ways of implementing such a service.4,5 In this issue there are two reports about different approaches to screening. Watson and colleagues (p 217) studied a primary care population whereas Mennie and co-workers (p 214) chose the more familiar setting of the antenatal clinic. In most discussions about screening the general assumption is that, because a test is possible, testing should be implemented as a service; the moral, ethical, social, and economic issues that underpin the introduction of such a service tend to be overshadowed by questions of logistics. In the USA, the Office of Technology Assessment report on CF screening due out shortly will take no position on whether screening ought to be instituted on a massive scale. The screening test for CF is based on mouthwash samples of cells whose DNA is amplified, digested, and subjected to separation techniques that identify three or four common mutations. In the UK about 85% of cases of CF would be associated with one or other of these abnormalities. A much larger proportion of cases can be detected by use of more complex techniques and in certain subpopulations, such as that lately described in Brittany.6 However, any screening programme must be considered in the light of the additional costs involved in detecting higher proportions of cases and the gene frequency in the local population. Detection of 85% of mutations means that only 72% of couples at risk of an affected child could be identified. In some countries detection of 85 % of mutations is unrealistic, and the proportion detectable and thus the efficacy of the test may vary within a country. Before a screening programme is established the organisers must make explicit the objectives of the programme. For example, there is a considerable difference between establishing a screening
this type of programme are that the information about CF, counselling, and the test itself are separated from immediate decisions regarding reproduction; this approach allows time for reflection and contemplation free from the worry of having to make immediate decisions about a pregnancy. This strategy has its disadvantages. Because the test is undertaken outside the context of reproduction, its results and meaning may be forgotten when decisions have to be made. Even after counselling, a significant minority were unable to state correctly their risk of having an affected child, even to the extent that 13% believed that two carriers could not have a normal child. In the primary care context, only 57% suggested a test to their partner, and only 87 % of partners underwent a test, so less than half (49%) the couples at higher risk had carrier status determined in both partners, in each case with only 85 % efficacy. Thus only 35 % of couples at risk of having a child with CF were detectable by this screening approach. What about stigmatisation of carriers? Although most people who were screened shared the information with relatives, partners, and friends, a sizeable minority did not. The second model of a screening service-to offer screening to women who are attending antenatal clinics-includes several variations. General practice antenatal clinics7 can be used for earlier diagnosis; and screening may be offered only to couples rather than in two stages (woman first and partner only if woman is positive). This model has the advantage of higher uptake rates, and information is imparted in the direct context in which it is to be used, thereby minimising the risk of misinterpretation, forgetfulness, and stigma. Again, there are disadvantages. Testing that takes place during pregnancy has to be done quickly and decisions on action must be taken speedily to allow termination in the first or early second trimester. Pregnancy itself is a time of stress, especially with a first baby, and this could make the couple act hastily and ultimately regret their decision, particularly if counselling is perceived as directive. Individuals need to be retested if their partner changes. Moreover, screening of couples as a unit prevents individuals’
knowing their status family members be
or recommending that other screened. There is also the
210
difficulty of advising a couple when one member is positive. They cannot be reassured, since their risk of having an affected baby is still 1 in 500 (because of the mutations not included in the screening test); nor can they be offered prenatal diagnosis, since the falsepositive rate would mean that twenty normal pregnancies were terminated for each affected one.4 If the aim were to reduce the incidence of cystic fibrosis, both types of screening programme would require very high response rates to justify their application. In the antenatal clinic strategy, 73% of potentially eligible women were actually screened. In this group, only 85 % of mutations could be detected, and likewise in their partners, so only 52% of couples at potential risk could be detected by this method. Furthermore, Watson et al showed there is considerable uncertainty about termination of an affected pregnancy in the primary care setting. Although the single affected pregnancy in Edinburgh was terminated, one cannot assume that this would always be so; even after counselling, only just over one third of couples would definitely consider termination, half were undecided, and the option of having no children is not available to those already
an
pregnant.
and the costs of medical care for unaffected individuals. Re-evaluation will be needed each time the efficacy of the test is improved and refined. The technical ability and practical skills to implement screening programmes for detection of CF carriers in the general population are now available. We still need to think very carefully about whether nationwide programmes are what we really want, and what the aims and ethical base of such an approach might be.
Testing inevitably results in some anxiety--during the process, during the wait for results, after notification of a positive result, and before the first counselling session. Longer term follow-up of screening in primary care revealed that, for most individuals, anxiety levels return to normal after 6 months. This is a short period of follow-up for an event that may have lifelong implications. Screening in antenatal clinics likewise induces anxiety, which seems to revert to normal when a negative result is obtained for the parmer. There is no knowing whether similar anxiety would be experienced in subsequent pregnancies, or what happens to those couples where both are found to be carriers. Uncertainty about decisions to terminate pregnancy, and about residual anxiety, makes detailed economic evaluation difficult. Nevertheless, crude economic statements-cost comparisons rather than cost-benefit or cost-utility analyses-have been made in both settings. The results suggest that financial savings may accrue to a health service from use of screening programmes, but several important factors have not been evaluated. For example, the cost of caring for all children born with CF whose parents were offered screening but either declined or who had undetectable mutations must be included as a cost of the programme. It is questionable whether patients with CF can be considered wholly as a cost in a cost-benefit analysis. Similarly, can normal individuals be regarded as entirely cost-free? Use of crude cost comparisons may greatly overestimate the value of the programme and underestimates the cost prevented. approach to population screening, it is implicit that, at least for some families, termination of
per
case
For either
affected pregnancy or a decision to remain childless would be preferable to having a child with CF. This decision may reduce the strain placed on families by a child with chronic illness, but what do we know about prognosis and quality of life for people with CF? Although median expectations of life of 20-25 years are usually cited, Britton and Knox8 pointed out that these figures are for a birth cohort of 20-25 years ago, when treatment was less effective, and they suggest that most patients bom with CF today could expect to live well into middle age, working and living independently. Termination of pregnancy is not itself without cost, just as having a child with CF in the family is not all negative. One cannot assume that an affected pregnancy will be followed by an unaffected one; that some families will choose instead to have no further children may itself cause considerable
morbidity. If justification of a screening programme for CF is economic, then decisions should be informed by full cost-benefit analyses which include costs of screening failures, the potential contribution of people with CF to society, loss of normal fetuses through antenatal diagnosis, effects on the family of children with CF,
1.
Rommens JM,Iannuzi MC, Kerem B-S, et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989; 245:
1059-65. 2. Riordan JR, Rommens JM, Kerem B-S, et al. Identification of the cystic fibrosis gene: cloning and characterisation of complementary DNA. Science 1989; 245: 1066-72. 3. Kerem B-S, Rommens JM, Buchanan JA, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245: 1073-80. 4. Wald NJ. Couple screening for cystic fibrosis. Lancet 1991; 338: 1318-19. 5. Watson EK, Mayall E, Chapple J, et al. Screening for carriers of cystic fibrosis through the primary health care services. BMJ 1991; 303: 504-07. 6. Ferec C, Audrezet MP, Mercier B, et al. Detection of over 98% of cystic fibrosis mutations in a Celtic population. Nature Genet 1992; I: 188-91. 7. Harris HJ, Scotcher D, Craufurd D, Wallace A, Harris R. Cystic fibrosis carrier screening at first diagnosis of pregnancy in general practice. Lancet 1992; 339: 1539. 8. Britton J, Knox AJ. Screening for cystic fibrosis. Lancet 1991; 338:1524.
Colorectal cancer: new evidence for the adenoma/carcinoma sequence The debate about the ability of some colorectal adenomas to undergo malignant change has centred largely on epidemiological and histopathological studies.1-3 More refined work, involving analysis of cell lines and genetic changes, is emerging in support
