Article Text
Abstract
Background Newer approaches to genetic counselling are required for population-based testing. We compare traditional face-to-face genetic counselling with a DVD-assisted approach for population-based BRCA1/2 testing.
Methods A cluster-randomised non-inferiority trial in the London Ashkenazi Jewish population.
Inclusion criteria Ashkenazi Jewish men/women >18 years; exclusion criteria: (a) known BRCA1/2 mutation, (b) previous BRCA1/2 testing and (c) first-degree relative of BRCA1/2 carrier. Ashkenazi Jewish men/women underwent pre-test genetic counselling prior to BRCA1/2 testing in the Genetic Cancer Prediction through Population Screening trial (ISRCTN73338115). Genetic counselling clinics (clusters) were randomised to traditional counselling (TC) and DVD-based counselling (DVD-C) approaches. DVD-C involved a DVD presentation followed by shorter face-to-face genetic counselling. Outcome measures included genetic testing uptake, cancer risk perception, increase in knowledge, counselling time and satisfaction (Genetic Counselling Satisfaction Scale). Random-effects models adjusted for covariates compared outcomes between TC and DVD-C groups. One-sided 97.5% CI was used to determine non-inferiority. Secondary outcomes: relevance, satisfaction, adequacy, emotional impact and improved understanding with the DVD; cost-minimisation analysis for TC and DVD-C approaches.
Results 936 individuals (clusters=256, mean-size=3.6) were randomised to TC (n=527, clusters=134) and DVD-C (n=409, clusters=122) approaches. Groups were similar at baseline, mean age=53.9 (SD=15) years, women=66.8%, men=33.2%. DVD-C was non-inferior to TC for increase in knowledge (d=−0.07; lower 97.5% CI=−0.41), counselling satisfaction (d=−0.38, 97.5% CI=1.2) and risk perception (d=0.08; upper 97.5% CI=3.1). Group differences and CIs did not cross non-inferiority margins. DVD-C was equivalent to TC for uptake of genetic testing (d=−3%; lower/upper 97.5% CI −7.9%/1.7%) and superior for counselling time (20.4 (CI 18.7 to 22.2) min reduction (p<0.005)). 98% people found the DVD length and information satisfactory. 85–89% felt it improved their understanding of risks/benefits/implications/purpose of genetic testing. 95% would recommend it to others. The cost of genetic counselling for DVD-C=£7787 and TC=£17 307. DVD-C resulted in cost savings=£9520 (£14/volunteer).
Conclusions DVD-C is an effective, acceptable, non-inferior, time-saving and cost-efficient alternative to TC.
Trial registration number ISRCTN 73338115.
- BRCA1, BRCA2
- Cluster Randomised trial
- Population based testing
- Genetic screening/counselling
- genetic testing
Statistics from Altmetric.com
- BRCA1, BRCA2
- Cluster Randomised trial
- Population based testing
- Genetic screening/counselling
- genetic testing
Introduction
Genetic testing for high-penetrance BRCA1/2 mutations is usually available to individuals from high-risk families fulfilling stringent family history (FH) criteria following genetic counselling in specialised cancer genetic clinics. Recent studies show that a significant proportion of BRCA1/2 carriers lack a strong FH of cancer but can be identified through population-based approaches, not standard clinical care.1–3 The Genetic Cancer Prediction through Population Screening (GCaPPS) randomised controlled trial (RCT) compared population screening (PS) with FH-based testing for BRCA1/2 mutations in Ashkenazi Jewish (AJ) individuals (ISRCTN73338115). We found that PS for BRCA1/2 mutations in AJ population does not harm quality of life/psychological well-being3 and is extremely cost-effective, leading to 33 days gain in life expectancy and incremental cost-effectiveness ratio (ICER)=‘-£2079/quality-adjusted life-year (QALY)’ well below the £20 000/QALY National Institute for Health and Care Excellence (NICE) threshold.4
Pre-test genetic counselling is a fundamental element of international guidelines5 for informed decision-making prior to genetic testing. A range of decision aids varying from pamphlets, booklets, computer-based programmes, audiotapes, to web-based platforms have been used as adjuncts to counselling to facilitate decision-making in high-risk populations. Decision aids reduce decisional conflict and lead to an increase in knowledge, accuracy of perceived benefits/harms, participation in decision-making process and ability to make informed value-based choices.6 ,7 In addition, group-based and telephone counselling approaches have been found to be beneficial and non-inferior in high-risk women.8–12
For large-scale, population-based genetic testing to become feasible and practical, it is necessary to move away from the ‘traditional face-to-face genetic counselling’ (TC)13 ,14 approach, which is cost–intensive, requiring significant health professional time. At present, there is no established model for providing pre-test genetic counselling for genetic testing on a population basis.15 We hypothesised that using a DVD (audio-visual tool) could significantly reduce the duration and increase cost-efficiency compared with traditional face-to-face counselling, while being non-inferior in terms of knowledge gained, counselling satisfaction, risk perception and equivalent in uptake of genetic testing. We report on outcomes from the only RCT that we are aware of comparing TC and DVD-based genetic counselling (DVD-C) approaches in an unselected population-based setting, undertaken during recruitment to the GCaPPS study.
Methodology
Cluster-randomised non-inferiority trial set within GCaPPS (ISRCTN73338115). Inclusion criteria: (a) individuals >18 years, (b) AJ ethnicity; exclusion criteria: (a) known BRCA1/2 mutation, (b) previous BRCA1/2 testing and (c) first-degree relative of a BRCA1/2 carrier. All volunteers received non-directive pre-test genetic counselling regarding genetic testing for AJ BRCA1/2 founder mutations. Genetic counselling was undertaken by a qualified genetic counsellor with clinical/counselling supervision provided by a regional genetics centre and a clinical fellow experienced in cancer genetics risk assessment and management. It was structured to meet the goals of genetic counselling,16–18 covering interpretation of FH, knowledge about risk, inheritance, management options, advantages, disadvantages and psychosocial implications to promote informed choice and adaptation.
Recruitment clinics (clusters) were randomised to TC and DVD-C approaches. Randomisation of clinics was essential for logistic, organisational and pragmatic reasons. There was an initial DVD development process from November 2008 to January 2009. This study reports on genetic counselling outcomes of clinics randomised from February 2009 until end of recruitment (July 2010) using the final DVD version. Randomisation was undertaken by a computer-generated random number algorithm. Participants were blinded to the type of genetic counselling when making an appointment. Appointments were made and randomisation implemented by the study administrator independent of the counsellors. DVD-C approach involved a DVD presentation (in the recruitment clinic) to small groups of volunteers (2–5) at a time. DVD-C volunteers subsequently saw a genetic counsellor for an individual genetic counselling session (post-DVD) at the same appointment. Participants in the TC group underwent face-to-face genetic counselling only. FH and baseline questionnaires were collected prior to the DVD presentation (DVD-C) or prior to seeing the genetic counsellor (TC group). Time taken for genetic counselling was documented. Postcounselling questionnaires were filled and collected after the genetic counselling session. Individuals deciding to undergo BRCA1/2 genetic testing were consented after genetic counselling.
Outcomes included uptake of genetic testing, change in cancer risk perception, increase in knowledge, counselling time and counselling satisfaction.
Secondary outcomes included relevance, satisfaction, adequacy, emotional impact and improvement of understanding with the DVD, and cost-minimisation analysis.
A baseline questionnaire assessed FH and socio-demographic characteristics. Knowledge was assessed by a specially developed 10-item (true=1/false=0) questionnaire (see online supplementary table S1) at baseline and postgenetic counselling. Satisfaction with genetic counselling was assessed postcounselling by the validated six-item Genetic Counselling Satisfaction Scale (GCSS): five-point Likert scale (strongly disagree=1, strongly agree=5) for each item, maximum score=30.19 ,20 Cancer risk perception was measured on a previously used 0–100 scale at baseline and postcounselling.21 A DVD evaluation questionnaire (see online supplementary table S2) assessed DVD impact (secondary outcomes) from May 2009 till July 2010. This was completed by DVD-C volunteers after watching the DVD and before meeting the genetic counsellor. Development of the knowledge questionnaire and DVD is described in online supplementary tables S3 and S4, respectively.
Supplemental material
Participants were recruited from the North London Jewish community. Recruitment was based on self-referral. Study flyers were made available through community charities, a high-street pharmacy (Boots) and website (http://www.gcapps.org). Eligible individuals who registered with the study team were sent a detailed trial information booklet. Genetic counselling was undertaken at high-street/community-based centres outside a hospital setting.
Statistical analysis
Statistical analyses were undertaken in ‘Stata-13.0’ (Stata, Texas, USA).
Baseline characteristics were calculated using descriptive statistics. χ2 tests compared categorical variables and t test (parametric) and Mann–Whitney (non-parametric) tests compared continuous outcome variables between two independent samples.
Random-effects models that included a random intercept term for each cluster (clinic) compared outcomes between TC and DVD-C groups, and were adjusted for potential confounders: FH (high/low risk), age, gender, parity, income, education and marital status. The total knowledge score was calculated as a sum of true=1 and false=0 for all 10 questions. Sensitivity analysis for knowledge scores was undertaken by (a) correcting final score to reflect proportion of valid questions answered and (b) assigning a score=‘0’ for missing answers. As the GCSS scores were highly skewed with a significant peak at 30, the transformation |GCSS score − 30| was considered. The resulting data distribution was approximated by a zero-inflated negative binomial regression model, adjusted with the same confounders. Per-protocol and intention-to-treat analysis were evaluated for outcomes of DVD-C and TC groups. A sensitivity analysis with multivariate imputation using chained equations (MICE)22 for missing data was undertaken for all outcomes. MICE iteratively simulates from suitable univariate imputation models that are fully conditional on all selected predictor variables until convergence is reached. Fifty fully imputed data sets were created to generate valid estimates and SEs, and produce correct statistical inference.
Non-inferiority analysis is needed to determine whether DVD-C is not worse than the current standard (TC) by an acceptable amount. A one-sided 97.5% CI was used to determine non-inferiority for cancer risk perception, increase in knowledge and counselling satisfaction. Non-inferiority was established when the 97.5% CI did not cross the non-inferiority margin. A two-sided 95% CI was used to test equivalency of genetic testing uptake as the aim of genetic counselling is informed decision-making rather than to increase/decrease testing. A superiority analysis was undertaken for counselling time.
The non-inferiority margins were based on clinically meaningful changes where available or set at no more than 0.5 SD worse than that for TC from prior studies19 ,23 or data collected during initial counselling undertaken from November 2008 to January 2009. The non-inferiority margin for knowledge gain=1 unit (minimum possible change on the scale, SD=3); GCSS=2 units (SD=5.6); risk perception=7 (SD=23.7). A ±10% equivalence margin was used for uptake of testing.
The sample size was adjusted by a variance inflation factor calculated for the intraclass correlation (ICC) from clustering. Sample size=K×n/[1+(n−1)×ICC], where K is the number of clusters, n is the cluster size and ICC is the intraclass correlation coefficient. This was further increased by 10% to adjust for relative efficiency between varying and equal cluster sizes.24 Assuming a mean cluster size=5, ICC=0.1, the adjusted sample size=(original sample)×1.54.
The total sample sizes needed for 80% power to detect ‘equivalence’ of uptake of testing=830 and ‘non-inferiority’ for knowledge=437, counselling satisfaction=382 and risk perception=554. Sample size for 15 min reduction in counselling time (SD=9.9)=37 and for non-inferiority margin of 0.5 SD of counselling time=265. Based on the final sample size of 936, cluster size=3.6, uptake of testing=89%, the study has >90% power for determining equivalence of uptake (ICC=0.21) and >95% power for establishing non-inferiority of knowledge gain (ICC=0.007), counselling satisfaction (ICC=0.0005), risk perception (ICC=0.053), and superiority for counselling time (ICC=0.15).
Cost-minimisation analysis was undertaken for TC and DVD-C approaches. The costs of filming the DVD=£300/- and burning a blank DVD=£0.60. The per-person cost=[DVD cost (unit cost=£((300/409)+0.60) per-volunteer)+genetic counselling cost]. The unit cost assumed for genetic counselling=£44/h of client contact, and the cost assumed for a psychologist appointment (if needed)=£73/h face-to-face contact (from Personal Social Services Research Unit's unit costs of health and social care 201025).
Patient/community involvement
The study was preceded by an extensive broad-based consultation/engagement with all sections of the Jewish community, which lasted almost a year (see online supplementary table S5).
Results
Between February 2009 and July 2010, 936 people underwent genetic counselling in GCaPPS and were cluster randomised by recruitment clinics (256 clusters) to TC (134 clusters, n=527) and DVD-C (122 clusters, n=409) groups. The mean cluster size=3.6 (TC=3.8, DVD-C=3.4). Baseline characteristics of participants were not significantly different between these groups (table 1). The mean age of participants was 53.9 (SD 15) years; 66.8% were women and 33.2% men. Our findings suggest a significant proportion of the AJ population are interested in BRCA1/2 testing and find it acceptable. Most (89%) of the participants opted for genetic testing following counselling. The uptake of testing rates and means (SD) for knowledge, GCSS, counselling time and risk perception is given in table 1. The consort flow chart is given in figure 1.
We found DVD-C was non-inferior to TC for increase in knowledge (d=−0.07; lower 97.5% CI=−0.41), counselling satisfaction (d=−0.38, 97.5% CI=1.2) and change in risk perception (d=0.08, upper 97.5% CI=3.1) (figure 2, table 2). Group differences and 97.5% CIs did not cross non-inferiority margins. Sensitivity analysis for knowledge scores and use of zero-inflated negative binomial regression for GCSS scores gave the same results of DVD-C being non-inferior to TC. DVD-C was equivalent to TC for uptake of genetic testing (d=−3%, lower/upper 97.5% CI −7.9%/1.7%) (figure 3, table 2). DVD-C was superior to TC in terms of counselling time leading to 20.5 (95% CI 18.7 to 22.2) min reduction in counselling time (p<0.005) (figure 3, table 2). Sensitivity analysis following multiple imputation of missing data also showed similar results (table 2).
Baseline knowledge level was significantly associated with decreasing age, and increasing levels of income and education, but independent of FH, gender, marital status and having children (table 3). Overall genetic counselling led to a significant increase in knowledge scores (p<0.0005).
Responses (n=316) to the DVD evaluation questionnaire are given in table 4. Ninety-eight per cent people were satisfied with the overall information, amount of information and DVD length. Thirteen per cent felt certain parts required more detailed explanation. Only 2% felt some parts could be left out (see online supplementary table S5). Ninety-five per cent would recommend the DVD to others, and 85–89% indicated it improved their understanding of risks/benefits/implications and purpose of genetic testing. Emotionally, 77% felt reassured; 87–95% felt no significant degree of worry/concern/upset; 11% felt somewhat worried/concerned, 3% somewhat upset and 1.3% ‘quite a lot’ worried/concerned after watching the DVD. Table 5 summarises responses on parts making people feel worried/concerned/upset/reassured.
The total genetic counselling cost estimate=£7786.65 (£19/volunteer) for DVD-C and £17 306.68 (£33/volunteer) for TC groups. The reduction in face-to-face health professional consultation time with the DVD translated into a total cost difference=£9520.03. DVD-based counselling led to a cost saving=£14/volunteer counselled. Although the cost minimisation of £14/volunteer may seem to be small in individual terms, when extrapolated across a whole population it actually amounts to quite a substantial saving for the healthcare system.
Discussion
To the best of our knowledge, this is the first RCT to report on systematic pre-test genetic counselling in a low-risk population (unselected for FH) of men and women undergoing BRCA1/2 mutation testing. The finding that DVD-C is not inferior to TC with respect to increase in knowledge, risk perception or counselling satisfaction, equivalent in uptake of testing and more cost-efficient (cost saving=£14/volunteer) is of great importance and suggests that DVD-C can be used as an effective and efficient alternative to traditional pre-test genetic counselling.
Group genetic counselling is reported to reduce the duration of counselling in high-risk populations,8 but this is the first report of using a DVD in this situation. DVD is an audio-visual tool with several advantages. It can be distributed/accessed by post, the web, general practitioner surgeries, community centres or other high-street sources and watched by people prior to their genetics appointment. Unlike group/telephone counselling, it does not require a health professional to deliver the educational material. Printed educational material is also effective in increasing knowledge and facilitating decision-making.26 ,27 We did not directly compare a printed decision aid with a DVD in this study. Pre-test genetic counselling reduces distress, improves patients’ risk perception28 and currently remains part of international guidelines for genetic testing.5 Although no pre-test genetic counselling was undertaken in two single-arm contemporaneous Canadian2 and Israeli29 population studies, post-test counselling was provided, and good satisfaction reported by participants with the testing process. Such an approach of ‘no pre-test counselling’ or only ‘post-test counselling’ has not yet been directly compared with TC in a randomised trial.
For population-based testing to be feasible, newer models for providing information for informed decision-making prior to genetic testing are necessary, which need to be properly evaluated in well-designed trials and ideally compared with the gold standard of TC. While we have demonstrated a viable DVD-based model, other models are also being explored/developed. Telephone genetic counselling has been successfully used for triaging women from high-risk families for TC10 and disclosure of test result.9 ,30 ,31 Three RCTs compared telephone counselling to TC in high-risk women attending genetics clinics. No difference in satisfaction32 was reported in one. Two were non-inferiority trials and found telephone counselling was non-inferior to TC,11 ,12 though lower testing uptake was reported in one.11 Telegenetics has been compared with TC in an RCT and reported to costless with no difference in satisfaction, though it was associated with 10% lower attendance.33 Telephone counselling/telegenetics have not yet been evaluated in a low-risk population unselected for FH. Newer models like mainstreaming counselling by the non-cancer genetics professional community34 or trained nurse specialists35 are currently being explored in clinical practice, but have not yet been directly compared with TC or other approaches in an RCT. It is likely that different models/pathways may be needed for different populations and different countries or healthcare systems. Further well-designed high-quality research is needed in this area.
The strengths of this report include the cluster-randomised design, non-inferiority analysis, community-based model for undergoing genetic testing and a high questionnaire response rate (73–100%). The differences in number of volunteers between the two study arms are explained by the randomisation of clinics (not volunteers), varying clinic times and differences in clinic sizes. But as expected, the baseline characteristics of the groups were in balance (table 1). Lack of qualitative data may be considered a weakness, and restriction to AJ participants may limit generalisability to other populations. We were also unable to analyse long-term outcomes postdisclosure of the test result, and this may be a limitation of the analysis. We did not include the 15 min patient time taken to watch the DVD in the cost-minimisation analysis because our analysis covers a healthcare perspective in line with NICE methods guidance, and therefore as per NICE guidance, patient costs are excluded. Besides, in practice, we would expect patients to have watched the DVD before attending a genetic counselling session. We guaranteed compliance and maximised questionnaire response by making people watch the DVD prior to counselling. Hence, in the future, when the DVD is delivered at home, it is important to ensure that people do watch the DVD at home prior to attending the genetic counselling session to ensure generalisability of results.
The high genetic testing uptake rate found in our study has also been reported by others.2 ,36 ,37 This may also be a function of a self-selected population and/or non-directive informative pre-test counselling received by participants. Our knowledge questionnaire was able to detect changes in knowledge (sensitivity to change). The increase in knowledge following pre-test counselling found in a low-risk population is similar to previous reports from high-risk populations.26 ,38 ,39 Older studies reported lower levels of knowledge about genetic testing and understanding of cancer risk.26 ,39 However, our relatively higher mean baseline score (>7) suggests that the average person coming forward for BRCA1/2 testing today may have greater levels of awareness/knowledge, which is reassuring. The lack of difference in knowledge scores between those with and without a strong FH of cancer re-emphasises this point and is contrary to previous findings of an association between knowledge and FH of cancer.38 The high baseline levels of knowledge may be a reflection of number of factors such as (a) self-selected trial participants, (b) the higher education and income levels known to be prevalent in the UK Jewish community compared with the rest of the non-Jewish general population and (c) ever-increasing public information and awareness on this issue. Our finding that level of knowledge is associated with education and income is consistent with earlier reports,38 ,40 and with the positive correlation (Spearman's r=0.3, p<0.005) between income and education levels, expected in a general population. Younger people had greater knowledge about genetic testing than older people. To the best of our knowledge, this has not been reported before. Factors that could have contributed to this include greater awareness of genetics, its recent incorporation into school curriculums, proactive behaviour and better access to sources of information in younger age groups.
Decision-making where each option has benefits/risks that people may value differently can be a difficult process. Overall, our DVD was well received with high satisfaction levels and enabled people to make specific, deliberated choices appropriate for them. The increase in knowledge is consistent with the effectiveness of the DVD in providing relevant information and improving the understanding of purpose/benefits/risks/implications related to genetic testing. Getting the right balance between DVD length and amount of information provided is challenging. The 98% satisfaction with length/information, 88% feeling no need for further explanation and 95% willingness to recommend it suggest our 15 min DVD struck the right balance for most people. A longer/more detailed DVD would yield small improvements, while greatly increasing the proportion of disaffected people.7 That the same information/content on a topic generated different reactions (reassurance/worry) suggests the DVD helped facilitate variable responses consistent with individual personal values. Need for more information on insurance/risks/inheritance highlighted by a small proportion represent areas for further development. The DVD quality can also be improved by incorporating qualitative data and using better production, film making and editing facilities.
The ability to identify 50% additional carriers, lack of psychological harm and cost-effectiveness of population testing for BRCA1/2 mutations in AJ individuals3 ,4 ,29 calls for changing the clinical paradigm to population testing for BRCA1/2 founder mutations in this population. DVD-based counselling approach is an effective, acceptable, non-inferior and cost-efficient alternative to TC and could be implemented for population testing in AJ. This can generate cost savings that is relevant for health authorities and commissioners of genetic counselling services and could enable more resources being directed to individuals who have difficulty coping with the genetic test result and/or needing greater support from genetics services following genetic testing.
Advances in high-throughput genetic testing technology, computational analytics and falling costs have made non-AJ general population testing technically feasible.41 ,42 The identification of newer moderate penetrance genes (RAD51C/RAD51D/BRIP1)43–45 and availability of panel testing will lead to an ever-increasing demand for genetic services with newer challenges for pre-test education and genetic counselling. Future research needs to compare telegenetics, telephone counselling, use of dial-in/web-based helplines, web apps along with DVD/other decision tools to identify/develop cost-efficient mass-based strategies to optimise education and facilitate informed decision-making without negatively affecting satisfaction, knowledge or psychological well-being in the general non-AJ population. A move away from TC is necessary to achieve the full benefit of genomic advances to deliver predictive, preventive, personalised and participatory (P4) medicine for cancer prevention.
Acknowledgments
The authors are particularly grateful to the women and men who participated in the trial. They are grateful to the entire medical, nursing and administrative staff who work on the GCaPPS trial and to the independent members of the trial steering committee (chaired by Prof Michael Baum) and data monitoring committee (chaired by Prof Jack Cuzick). They are also grateful to the numerous supporting Jewish charities, community and religious organisations, Rabbis as well as numerous members of the Jewish community for their time, advice and support. They are particularly grateful to the teams at Boots Pharmacy, Norwood, Jewish Care, Ovacome, Agudas Israel Housing Association, Academic Study group on Israel and the Middle East, Liberal Judaism, Movement for Reform Judaism, Indian Jewish Association, Stamford Hill Group Practice, Lane End Medical Centre, Jewish Medical Association and Chai Cancer Care, for their input and support. They are grateful to Robert Liston, Vijay Devineni and Andy Ryan for their help with designing the trial management system and for IT support. They are also grateful to the various regional genetic units in London (Great Ormond Street Hospital, Kennedy Galton Centre Northwick Park Hospital, Guys Hospital and Royal Marsden Hospital) and the West Midlands Regional Genetics Service for their support of the study. Thanks also to Katriina Whitaker, Prof Mahesh Parmar, Dr Anthony Silverstone, Rabbi Margaret Jacobi, Marlena Schmool, Rabbi Danny Rich, Rabbi Miriam Berger, Rabbi Eli Kernkraut, Rabbi Tony Bayfield, Rabbi Helen Freeman, Angela Brady, Elizabeth Bancroft, Imelda Udeh, Judith Soloway, Jennifer Wiggins, Adina Roth, Hannah Lyons, Jane Lyons, Sarah Chamberlain, Michelle Johnson, Helen Mitchell, Katherine Duerden, Gemma Byrne, Fiona MacDonald, Louise Bayne, Ruth Payne for their support of the study.
References
Footnotes
Contributors RM, IJ and UM were responsible for literature search and design of the study. RM, IJ, UM, JW, KL, SG, SCS and AK were involved in developing interventional questionnaires. RM, KL, MB, AMG and MR were involved in data collection and analysis. RM and MB did the statistical analysis and prepared the tables and figures. RM, MR, AMG and RL did the cost-minimisation analysis. RT and CJ were collaborators and helped with study development and data collection from genetic laboratories. RM and IJ prepared the first draft of the manuscript. RM, IJ, UM, KL, JW, SG, LS, NB, RD, AK, HD, YW, CC, IT, AMG and UB were involved in running the study. YW did the genetic testing. All authors critically contributed to and revised the manuscript and approved the final version.
Funding The study was funded by ‘The Eve Appeal’ charity (GTCV). The corresponding author had full access to all data in the study. The GCaPPS investigators had final responsibility for the decision to submit the report for publication. The study was registered with the International Standard Randomised Controlled Trial Number Register—ISRCTN 73338115 (http://www.controlled-trials.com/ISRCTN73338115).
Competing interests All authors have completed the ICMJE uniform disclosure form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author). All authors declare support from the Eve Appeal charity for research funding of this work. IJ declares consultancy arrangements with Becton Dickinson, in the field of molecular markers for ovarian cancer. IJ and UM have a financial interest in Abcodia, Ltd., a company formed to develop academic and commercial development of biomarkers for screening and risk prediction. IJ is a member of the board of Abcodia Ltd and a Director of Women's Health Specialists Ltd.
Patient consent Participants gave informed consent for use of data for publication. Separate consent for data sharing of non-identifiable data was sought and provided by a number of participants.
Ethics approval The GCaPPS study received full ethics approval from the Institute of Child Health/Great Ormond Street Hospital Research Ethics Committee on 8th June 2008 (REC Reference number 08/H0713/44).
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Relevant anonymised patient level data can be obtained upon reasonable request from the authors.
Transparency declaration The corresponding author (the manuscript's guarantor) affirms that the manuscript is an honest, accurate and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.