Journal of Medical Screening

Objective To determine whether women who have had a positive serum screening result in one pregnancy have a lower rate of participation in screening in their next pregnancy.

Setting The Women and Infants Hospital triple marker screening programme.

Methods Pregnancy and screening information was collected from laboratory and hospital databases to compare subsequent screening participation in women who were screen negative and screen positive for risk of Down's syndrome (DS) or neural tube defect (NTD) pregnancy.

Results In an age matched comparison, 108 women who had a previous screen positive result were significantly less likely than 108 women who were screen negative to participate in maternal serum screening in their next pregnancy. When examined according to type of screen positive result, the effect was significant for both those who were screen positive for DS and those who were screen positive for NTD. The degree of risk in screen positive women did not significantly affect their uptake of screening in the next pregnancy.

Conclusions Anxiety related to a screen positive result probably causes decreased participation in maternal serum screening in the next pregnancy. Reducing the screen positive rate in prenatal serum screening would alleviate maternal anxiety and would probably lead to more stable participation.

Objective Antenatal screening for syphilis is well established in the United Kingdom. The prevalence of syphilis is now very low, prompting the question as to whether this screening programme is still necessary. This paper aims at identifying possible screening strategy options for the programme and comparing their effectiveness and cost effectiveness.

Methods The cost of the screening programme in the United Kingdom was estimated. This was based on the cost of screening tests, treatment, and follow up of infected women and their infants. This information was obtained from laboratories, antenatal clinics, and genitourinary medicine clinics. Epidemiological data from a survey of women treated for syphilis in pregnancy were analysed to identify groups at increased risk of syphilis. Strategic options for the screening programme were then identified. The effectiveness, number needed to treat, and cost effectiveness of these options were compared.

Results Antenatal screening in the United Kingdom detected at least 40 pregnant women who need treatment for syphilis every year. This means that 18 602 women are screened for every woman detected who needs treatment for syphilis. The marginal annual cost of this screening programme in the United Kingdom is £672 366. This is equivalent to 90p per woman screened, or £16 670 to detect one woman who needs treatment for syphilis. The screening programme could be targeted geographically at pregnant women in the Thames regions. This option has the potential to save £482 185. Other strategic options are to target pregnant women in non-white ethnic groups, or those born outside the United Kingdom. These targeted options would each detect between 70% and 77% of women needing treatment for syphilis. These options could potentially save £592 938 and £562 691 respectively.

Conclusions Targeting or stopping the screening programme would save relatively little money. Although selectively screening groups by country of birth or by ethnic group could detect at least 70% of cases, this would be politically and practically difficult. Targeting by region would also be effective, but would pose similar ethical and medicolegal problems. These facts and the changing international epidemiology of syphilis lead us to recommend that the current universal antenatal screening for syphilis should continue.

Establishment of mammography screening in Sweden has progressed logically from pilot study through clinical trials to service screening. Screening with mammography for early detection of breast cancer has been provided by all Sweden's 26 county councils since 1997. It took 23 years from the initial pilot study through clinical trials to the establishment of mammography service screening through out Sweden.In the screening rounds completed by 1995–96, and provided by all but one county council, 1 040 000 women participated, corresponding to 81% of those invited. The national average recall rate was 2.2%, and consequently 23 000 women were recalled for additional investigations. Eleven county councils invited women aged 40–74, six invited women aged 50–69, the remaining eight invited women between both these age intervals.Mammography outside screening programmes—clinical mammography—is available throughout Sweden. About 100 000 women a year were referred for clinical mammography and about 50% of these were either younger or older than those invited for screening. A negative relation between the use of clinical mammography and participation in the screening programmes was noticed.

Objective To investigate the proportion of interval breast cancers that could have been detected at the previous screening examination, and to gain more insight into the characteristics of these tumours.

Setting Breast cancer screening programme in mid- and southern Limburg, the Netherlands.

Method Firstly, previous screening mammograms of 92 interval cancer cases were blindly reread by the radiologists from two different units as part of their daily screening workload. Secondly, a separate informed review was conducted, in which all of the cases (except for two) were subclassified into four categories: screening error; minimal signs; radiologically occult both at previous screening and at diagnosis; or radiologically occult at previous screening. Trends in recall rates and false and true positive rates were calculated to study potential changes during and after the blind review.

Results In the blind review, 15% of the interval cancers were detected in both rereadings and 21% in one rereading. In the informed review, 25% were classified as a screening error and 24% as minimal signs present; 74% of the cases classified as a screening error had been recalled at least once in the blind review. Recall rates and false positive rates in daily screening practice increased significantly during and after the study period.

Conclusions About one third of the interval cancers could have been detected in the previous screening round. In order to diminish the number of avoidable interval cancers, review and classification of interval cancers is an important tool for continuing the education of screening radiologists. Therefore, further development of review training procedures is necessary.

Objective Examination of figures from screening at several different centres, and from the incidence of interval cancers in those programmes, suggests that a single radiologist reading by him/herself would miss 16 to 31% of cancers detectable in the mammograms. This study investigates how the accuracy of breast cancer screening could be improved.

Method Draw a comparison between screening for breast cancer and other inspection tasks that have been studied in the experimental psychological laboratory to suggest two low cost fixes that might improve the accuracy of breast cancer screening.

Recommendations (1) Radiologists should not work for more than half an hour at a time; (2) two radiologists double reading should inspect the same series of mammograms in inverse order; (3) there should be a clerical check to see whether the efficiency of detection decreases with time on task; and (4) a more sophisticated engineering of the screening procedure might be accomplished by assembling a library of mammograms for which the correct diagnosis is known. If these library mammograms were mixed randomly with those to be inspected in some suitable ratio, it would then be possible to provide the radiologist with immediate feedback on the accuracy of (some of) his/her diagnoses. From experimental studies of inspection tasks this should improve diagnostic accuracy. These recommendations, of course, require evaluation in field trials.

Objective To estimate the costs of a quality management (QM) system as proposed by the Quality Management for Screening report for a future national colorectal cancer screening programme.

Methods Estimates of the costs of the QM system, including the associated costs of education and training and information provision, were based on expert opinion, the existing literature, and the experience of the current National Health Service (NHS) breast cancer screening programme (BSP) and the NHS cervical cancer screening programme (CSP).

Results The cost of a QM system to support a national colorectal cancer programme in the UK was estimated as approximately £3.8 million a year. Further annual costs related to QM will include £500 000 for education and training and £200 000 for information provision. Adding these additional costs to a previously published UK economic evaluation of colorectal cancer screening increases the cost-utility ratio to approximately £6500 per quality adjusted life year gained (over an eight year follow up period).

Conclusions Any new screening programme, or an existing one, must have QM to ensure that the quality of screening is high and to maintain the right balance between benefit and harm. The significant costs of such a QM system should be included in any economic evaluation of a screening programme.

Objective To compare two immunochemical faecal occult blood tests based on reversed passive haemagglutination (RPHA) or latex agglutination (Hdia) in a population based screening setting.

Method Hdia was interpreted according to three positivity thresholds: 100, 150, or 200 ng of haemoglobin/mg of specimen solution. A total of 5844 subjects were recruited into the study, from 17 432 invited subjects aged 50–70.

Results Positivity rates were 3.3% for RPHA, Hdia100 3.5%, Hdia150 2.5%, Hdia200 2.0%. Among subjects complying with the diagnostic work up, colorectal cancer (CRC) was detected in 19 subjects (17 RPHA positive, 16 Hdia100 positive, 15 Hdia150 positive, 14 Hdia200 positive) and high risk adenoma/s in 41 subjects (28 RPHA positive, 32 Hdia100 positive, 29 Hdia150 positive, 25 Hdia200 positive). The prevalence of screen positive CRC in the population was for RPHA 2.9‰, Hdia100 2.7‰, Hdia150 2.6‰, Hdia200 2.4‰. The prevalence of screen positive high risk adenomas in the population was for RPHA 4.8‰, Hdia100 5.5‰, Hdia150 5.0‰, Hdia200 4.3‰.

Conclusion Hdia100 was as sensitive as RPHA for cancer and high risk adenomas. As Hdia is less technically complex than RPHA, it is a valid alternative to the latter, provided that full automation of the development procedure is available. Increasing the positivity threshold of Hdia up to 150 or 200 ng of haemoglobin/mg of specimen solution is not advisable as the increase in specificity is too small to justify the corresponding decrease in the detection of screen positive cancers in the population.

Objectives To assess the acceptability of bowel cancer screening using flexible sigmoidoscopy (FS).

Setting Adults aged 55 to 64 recruited from general practices in Welwyn Garden City and Leicester, which were the pilot and start up centres of a multicentred randomised controlled trial of FS screening (the ICRF/MRC Flexible Sigmoidoscopy Screening Trial).

Methods Screenees (n=4422) were sent a three month follow up questionnaire that included measures of satisfaction with information given before the test, facilities at the test unit, attitudes of the staff, and explanation of the results. Measures of pain, embarrassment, feelings of being “in control” during the test, willingness to encourage others to have the test, and gladness to have participated were also included. In addition, semistructured telephone interviews were conducted with 60 screenees, stratified by screening outcome and gender.

Results The follow up questionnaire was completed by 94% of screenees. Responses indicated a high level of satisfaction with the procedure: 99% were satisfied with the information given before the test, the facilities, the attitudes of the staff, and the explanation of their results; 91% reported only mild or no pain; 97% reported little or no embarrassment; and 99% were glad they had the test. Satisfaction ratings varied little by gender or outcome group. The quantitative results were reinforced by the qualitative data, which also revealed high acceptability.

Conclusion In the context of a clinical trial with dedicated trial staff, FS is a well tolerated procedure. There are high levels of satisfaction with service provision and positive attitudes towards the programme.

Objectives To provide preliminary evidence that the currently employed semiquantitative method of screening for glucose-6-phosphate dehydrogenase (G6PD) deficiency can only detect infants who are totally deficient for G6PD and misses all cases of partial G6PD deficiency.

Setting General population: 2150 randomly selected blood samples from the Blood Donation Department, Speliopouleion General Hospital, Athens, Greece. Neonate population: 2000 samples from neonates (50% male; 50% female) in maternity hospitals in the greater Athens area. High risk population: a total of 545 individuals from 133 families in the Athens area, the minimum acceptance criteria being the parents and any brother or sister.

Method Blood specimens from neonates or adults were collected and either spotted and dried on special filter paper (Schleicher and Schull 2992, Darmstadt, Germany) or used in tubes after being heparinised. For the quantitative evaluation of G6PD enzyme activity, the Quantase G6PD screening kit (Quantase Limited, Perth, UK) was used. Quantase G6PD controls (Quantase Limited) were used at three levels of G6PD. These controls are rated at 24, 30, and 37°C. Alternatively, we used the Sigma G6PDH controls (Sigma Chemical Company, St Louis, USA) which are rated at 30 and 37°C. The assay was performed according to the instructions included in the kit with the modification for haemoglobin normalisation.

Results General population: 36 females who were classified as having normal enzymatic activity with the semiquantitative test, were classified as partially deficient with the quantitative test. Neonate population: using the quantitative test, the percentage of G6PD deficient neonates in this population was 5.5%, compared with 3.17% reported in routine screening using the semiquantitative method. High risk population: the quantitative method detected 28 cases of total or partial G6PD deficiency in sisters of males with known total deficiency. The semiquantitative method only detected 32% (nine out of 28) of these cases.

Conclusions A considerable amount of partially G6PD deficient female neonates (heterozygotes) are undetected and classified as having normal enzymatic activity using the semiquantitative method, which uses a cut off of 2.1 U/g haemoglobin (Hb). The use of a fully quantitative G6PD screening kit is proposed, employing the automated haemoglobin normalisation and a cut off of 6.4 U/g Hb. Any neonate with an activity below this mark should be regarded as G6PD deficient, and all preventive measures should be taken.