|
Randomized trial of a specialist genetic assessment service for familial breast cancer |
Brain K, Gray J, Norman P, France E, Anglim C, Barton G, Parsons E, Clarke A, Sweetland H, Tischkowitz M, Myring J, Stansfield K, Webster D, Gower-Thomas K, Daoud R, Gateley C, Monypenny I, Singhal H, Branston L, Sampson J, Roberts E, Newcombe R, Cohen D, Rogers C, Mansel R, Harper P |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology The use of a new multidisciplinary genetic and surgical assessment service for screening for breast cancer, in women with a family history of breast cancer. This was performed in addition to cancer surveillance and advice on risk management. The genetics consultation was based on education about breast cancer genetics, a genetic assessment of the individual's breast cancer risk, and possible presymptomatic testing for the BRCA1 and/or BRCA2 gene in women identified as high risk.
Type of intervention Multidisciplinary screening.
Economic study type Cost-effectiveness analysis.
Study population The study population comprised women with a family history of breast cancer. This was defined as having the following:
a first-degree female relative diagnosed with breast cancer before the age of 50 years;
a first-degree female relative with bilateral breast cancer at any age;
at least two first-degree relatives with breast cancer; or
a first- and second-degree relative with breast cancer.
For inclusion, the women also had to have no personal history of breast cancer, no prior genetic counselling, and be resident in Wales. After loss to follow-up, the final sample for the effectiveness analysis comprised 545 women. The mean age was 42.59 years (range: 19 - 73) and, on average, the women had two relatives affected with breast cancer (range: 1 - 9).
Setting The setting was an institution. The study was carried out in two centres, the Breast Test Wales Screening Centre and the Family History Clinic at the University Hospital of Wales. Both were situated in Cardiff (Wales, UK).
Dates to which data relate The dates during which the data on the effectiveness and resources used were gathered were not reported. The prices used referred to 1997 to 1998.
Source of effectiveness data The effectiveness evidence was derived from a single study.
Link between effectiveness and cost data The costing was undertaken prospectively on the same patient sample as that used in the effectiveness analysis.
Study sample Power calculations were performed to determine the appropriate sample size. Data from 1,000 patients were required to obtain an 89% power to detect a statistically significant difference between the groups at the 5% level. The patients included in the study were drawn from all women resident in Wales, who were referred by their general practitioner to a breast surgeon at the district general hospital because of a family history of breast cancer. A total of 1,700 practitioners were contacted about the study, and the recruitment period was 18 months.
Overall, 1,172 women were referred to the trial. Data from the first 1,000 women were considered in the study. Two hundred and sixty women were immediately excluded because 167 of them did not return the questionnaire, and a further 93 did not agree to participate in the study, did not attend the scheduled appointments on several appointments, or were not eligible. Of the remaining 740 women, 369 received the standard intervention and 371 received the trial intervention, although 5 did not attend the surgical component of the multidisciplinary approach. The women in the intervention group were further stratified as low, moderate or high risk. The baseline characteristics were not shown for this sample, but for the group after loss to follow-up.
Study design The study was a randomised, controlled clinical trial carried out in two centres (Breast Test Wales Screening Centre and Family History Clinic at the University Hospital of Wales). A randomisation procedure, which was based on a computer-generated sequence of random numbers, was used to control for differences between randomisation at the two centres. The women were followed for 9 months. The immediate loss to follow-up was 52 women in the control group and 28 women in the intervention group. At 9 months, 35 women in the control group and 75 women in the intervention group were lost to follow-up. As a result, only 282 women in the standard group and 263 women in the trial group completed the study. No blinding method for the outcome assessment was reported.
Analysis of effectiveness The basis of the clinical analysis was not explicitly reported, but appears to have been that of treatment completers only. The primary health outcomes used in the effectiveness analysis were:
general anxiety, measured using the State-Trait Anxiety Inventory (score range: 20 - 80);
breast cancer worry, measure using the Breast Cancer Worries scale (score range: 6 - 24);
perceived risk, measured using two questions (score range: 2 - 10);
knowledge of familial breast cancer (score range: 0 - 4); and
patient satisfaction, measured using the 12-item Satisfaction With Genetic Counselling Questionnaire (score range: 12 - 48).
All outcomes were measured at baseline, immediately after the clinic, and 9 months after the clinic. Since the drop-out rate was quite high, the authors assessed the participation bias at each stage (baseline, immediately after clinic, and 9 months after clinic). For example, at baseline, the effectiveness outcome results of the study participants, irrespective of the group (735 women), were compared with those of women who completed the questionnaire but then declined to participate in the study (75 women). The authors reported that the characteristics of the group at baseline were not statistically significantly different at the 5% level.
Effectiveness results The average state anxiety score (+/- standard deviation, SD) was as follows.
Trial group: 35.9 (+/- 11.11) at baseline, 34.33 (+/- 10.79) after the clinic, and 36.38 (+/- 12.34) 9 months after the clinic.
Control group: 35.54 (+/- 10.87) at baseline, 33.14 (+/- 10.11) after the clinic, and 35.18 (+/- 11.75) 9 months after the clinic.
The average breast cancer worry (+/- SD) was as follows.
Trial group: 11.79 (+/- 3.37) at baseline, 10.55 (+/- 2.91) after the clinic, and 10.55 (+/- 3.21) 9 months after the clinic.
Control group: 11.49 (+/- 2.97) at baseline, 10.50 (+/- 2.70) after the clinic, and 10.63 (+/- 2.90) 9 months after the clinic.
The average perceived risk (+/- SD) was as follows:
Trial group: 7.29 (+/- 1.24) at baseline, 6.44 (+/- 1.30) after the clinic, and 6.74 (+/- 1.30) 9 months after the clinic.
Control group: 7.33 (+/- 1.17) at baseline, 6.62 (+/- 1.14) after the clinic, and 6.90 (+/- 1.25) 9 months after the clinic.
The average knowledge (+/- SD) was as follows.
Trial group: 1.54 (+/- 1.09) at baseline, and 2.17 (+/- 1.08) after the clinic.
Control group: 1.45 (+/- 1.06) at baseline, and 1.89 (+/- 1.08) after the clinic.
Patient satisfaction after the clinic was 42.82 (+/- 5.23) in the trial group and 42.29 (+/- 4.57) in the control group.
In general, there were no statistically significant differences between the groups in terms of the reported outcomes. The only statistically significant difference observed was for the knowledge score. The trial group had a greater knowledge score after the clinic than did the control group. Finally, in terms of participation bias, women who dropped out at baseline reported a higher trait anxiety and state of anxiety, and a lower personal risk, than those participating in the study did. At the first assessment after the clinic, the women who dropped out were younger and reported a higher baseline cancer worry. At the 9-month assessment, the drop-outs were younger and reported higher baseline state anxiety, cancer worry, and perceived risk. The rate of drop-outs in the control group was higher after the clinic than in the intervention group.
Clinical conclusions Both screening methods were generally effective in alleviating worry, anxiety, and perceived risk. However, the new multidisciplinary approach was slightly, but statistically significantly, more effective in improving the knowledge level when compared with the traditional approach.
Measure of benefits used in the economic analysis The health outcomes were left disaggregated because it was not possible to specify any single outcome as being the key variable. It was stated that a cost-consequence analysis was therefore carried out.
Direct costs Discounting was irrelevant due to the short time horizon of the study. The unit costs and the quantities of resources used were not reported separately. The cost/resource boundary adopted was that of a society. The costing was undertaken following the TRACE project (Trial of Genetic Assessment in Breast Cancer; see Other Publications of Related Interest). Only the marginal costs of sending a woman to a special genetics service rather than the normal practice of referral, were included in the analysis. Those items that were the same for both groups were excluded. The time of the health professionals was assessed at the gross employment costs, according to grade. The travel time of the patients and/or relatives was determined by survey. The travel time of the professionals included the direct costs and the opportunity costs of time spent travelling.
The costs were presumably estimated from the two centres involved in the study. The quantities of resources were estimated on the basis of three tests on the samples of extracted DNA and, if a mutation was confirmed, then the presenting woman and her relatives were offered a further test with additional counselling before and after the test procedure. The dates during which the resources were gathered were not reported. All the costs were in 1997 to 1998 prices.
Statistical analysis of costs Statistical analyses of the costs were not conducted, although the costs were treated stochastically.
Indirect Costs The indirect costs were not included.
Currency UK pounds sterling (). US dollars ($) were also used. The exchange rate was 1 = $1.58.
Sensitivity analysis No sensitivity analyses were carried out.
Estimated benefits used in the economic analysis See the 'Effectiveness Results' section.
Cost results The mean cost of an initial consultation at the specialist genetic assessment clinic was 14.27 (95% confidence interval, CI: 12.30 - 16.24), or $22.55 (95% CI: 19.43 - 25.66), higher than the cost obtained with routine care.
The total extra cost of initial consultation for the whole cohort of patients (412 women) was 5,879 ($9,289).
Forty-eight women at high risk attended the preliminary genetics clinic at a cost of 2,576 ($4,070). Subsequently, there was a further additional cost in visits, counselling, and a mutation search for 39 affected relatives. This resulted in an additional cost of 11,462 ($18,110).
Six presenting women and 13 relatives received mutation testing at a total cost of 5,204 ($8,222). Overall, the total extra cost per presenting woman was 60.98 ($96.35).
Synthesis of costs and benefits Authors' conclusions The health benefits associated with the new multidisciplinary genetic screening for breast cancer were quite modest: consulting with a genetic or surgical specialist was equally reassuring for at-risk women. The small improvement in knowledge in the group who experienced the new approach should be weighted against the extra cost required, compared with the traditional approach.
CRD COMMENTARY - Selection of comparators The rationale for the choice of the comparators was clear. Contact with a specialist breast surgeon was selected as the comparator because it represented the standard practice in the UK. You should assess whether it represents a commonly used intervention in your own setting.
Validity of estimate of measure of effectiveness The effectiveness analysis used randomisation procedures, and power calculations were conducted to determine the sample size. In addition, the groups were reported to be similar at baseline. However, the rates of drop-out were quite high and participation bias, between those who completed the study and those who dropped out from the initial sample, was found at all stages of the assessment. As a consequence, there was a lack of statistically significant differences among some of the health outcomes assessed in the study groups. This could have been due to the reduced statistical power resulting from the substantial drop-out rate. The time period during which the effectiveness evidence was gathered was unclear. All these issues could have limited the internal validity of the analysis, therefore, the results should be interpreted with caution.
Validity of estimate of measure of benefit No summary health benefit measure was adopted because the authors reported that it was not easy to select a single outcome as the key variable among those reported in the study. As a result, a cost-consequence analysis was conducted.
Validity of estimate of costs The analysis of the costs was conducted according to the TRACE project (see Other Publications of Related Interest), but the authors noted that the different costing method would lead to different cost estimations. The costs were treated stochastically but were not statistically analysed. In addition, they appear to have been fairly specific to the study setting. Only those costs that differed between the study groups were included in the analysis. The dates during which the resources were collected were not reported. Although it was stated that a societal perspective was adopted, indirect costs (such as productivity losses) do not appear to have been included in the analysis.
Other issues The results of the study were not compared with the existing literature. The issue of the generalisability of the study findings to other settings was not addressed, and sensitivity analyses were not conducted. As a consequence, the external validity of the study was limited. However, the results were presented in full and the authors' conclusions were in keeping with the study population.
Implications of the study The authors pointed out that "additional psychological support and possible referral to specialist services may be required to manage distress in the subset of at-risk women who are psychologically vulnerable". Further studies focusing on factors that mediate the impact of genetic assessments, and sensitivity analyses of the current data, are reported to be in progress.
Bibliographic details Brain K, Gray J, Norman P, France E, Anglim C, Barton G, Parsons E, Clarke A, Sweetland H, Tischkowitz M, Myring J, Stansfield K, Webster D, Gower-Thomas K, Daoud R, Gateley C, Monypenny I, Singhal H, Branston L, Sampson J, Roberts E, Newcombe R, Cohen D, Rogers C, Mansel R, Harper P. Randomized trial of a specialist genetic assessment service for familial breast cancer. Journal of the National Cancer Institute 2000; 92(16): 1345-1351 Other publications of related interest Gray J, Brain K, Norman P, Anglim C, Barton G, Branston L, et al. A model protocol evaluating the introduction of genetic assessment for women with a family history of breast cancer. Journal of Medical Genetics 2000;37:192-6.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Anxiety /etiology; Breast Neoplasms /economics /genetics /psychology; Cost-Benefit Analysis; Female; Genetic Testing /economics; Humans; Middle Aged; Patient Care Team /economics; Patient Satisfaction; Prospective Studies; Risk; Wales AccessionNumber 22000001369 Date bibliographic record published 31/08/2002 Date abstract record published 31/08/2002 |
|
|
|