Full ethical approval for this study has been obtained from the Cancer Council Victoria Human Research Ethics Committee and the University of Newcastle Human Research Ethics Committee.
People with a diagnosis of bowel cancer within the last 10 months will be recruited from population-based cancer Victorian Cancer Registry covering the entire Australian state of Victoria (population 5.5 million).
Patient eligibility and recruitment
People with bowel cancer (index cases) who meet the following criteria will be identified from the cancer registry: 1) aged 18 or older; 2) registered with cancer registry within 10 months of diagnosis. The registry will write to the notifying doctor of each potentially eligible person and ask the doctor to contact the registry within 4 weeks if there is any reason why the patient should not be contacted about the study. Index cases whose doctors perceive them to be unsuitable for the study will be excluded. The registry will contact remaining index cases and ask their permission to pass their contact details onto the research team. Two reminders will be sent to non responders. Those who consent will be invited to participate in the trial with two reminders sent to non responders.
Assessment of family history
Patients will complete a baseline computer assisted telephone interview (CATI) to assess family history. On the basis of their report of family history each index case and their family will be classified as 1) average risk; 2) increased risk or 3) potentially high risk. These categories, defined according to the definition of the National Health and Medical Research Council (NHMRC), are each associated with different types of screening recommendations. Participants will also complete questions on age, gender, marital status, education and employment, disease characteristics, treatment, screening history, surveillance intentions and quality of life. Quality of life will be assessed using the European Quality of Life Scale (EQ-5D version) which covers 5 dimensions of health: mobility, self-care, usual activities, pain/discomfort and anxiety/depression [12
First degree relatives of people with bowel cancer will be eligible to participate in the trial if they are a) aged 18 or older; b) English speaking; and c) able to provide informed consent. Those with a prior diagnosis of bowel cancer, Crohn's disease, or inflammatory bowel disease will be ineligible.
Recruitment of FDRs
Patients will be asked to provide the names and contact details of any living first degree relatives aged 18 or older. They will be asked to indicate which (if any) of their FDRs they are willing for the research team to contact about the study. A choice will be given regarding method of contact of FDRs. Option one will involve the patient being given a letter about the study which they can pass onto their relatives; while option 2 will involve the research team contacting the relative(s) directly by mail. Patients will be asked to seek their relatives' permission prior to selecting the latter option. Non responders for option one receive one reminder letter. The information pack will contain a cover letter, information statement about the study, and ask the person to contact the research team if they were interested in participating. If the index case has provided the relative's phone number, the letter will ask the relative to return a "do not contact" form within the next two weeks if they did not want a researcher to contact them about the study. FDRs that consent to being contacted will be telephoned and asked to participate in a brief screening interview to assess trial eligibility. Consenting eligible FDRs will complete a baseline computer-assisted telephone interview (CATI) which will include questions about age, gender, marital status, education, employment, knowledge about screening and prior screening history and future intentions of screening, use of complementary therapies, and quality of life. Quality of life will be assessed using the European Quality of Life Scale [12
Consenting patients and their first degree relatives will be randomized as a family unit to either "minimal ethical care" or a tailored print based intervention. Randomization will be conducted centrally using a computer generated procedure. Interviewers will be blind to participants' allocation.
Patients assigned to the intervention group will receive a letter detailing recommendations for follow up care. A fact sheet with information about recommended follow up care will be mailed to patients. The patient's GP and surgeon will also be sent a fact sheet with information about best evidence surveillance care. First degree relatives will receive a tailored letter which provides advice on recommended bowel cancer screening tests and intervals based on their level of family risk. A brochure detailing the three risk levels and their corresponding screening recommendations will be enclosed. Family risk will be determined based on patient self report information and will correspond to the three levels of risk identified in the National Health and Medical Research Guidelines. The GP of the FDR will also receive a tailored letter indicating the likely risk category of the first degree relative. A brochure detailing screening recommendations for each risk category will be enclosed.
Minimal ethical care
Patients and first degree relatives assigned to the minimal ethical care group will receive a generic booklet on bowel cancer and generic pamphlet on bowel cancer screening respectively. First degree relatives categorized at very high risk will sent the tailored information as per the intervention group as part of our duty of care.
Follow up assessments
Both first degree relatives and patients will be mailed follow up surveys at 12, 24 and 36 months. First degree relatives will be asked to provide self report information on the type of bowel cancer screening undertaken in the past 12 months (if any), and any days off work due screening. Patients will be asked to provide self report information about participation in surveillance colonoscopy within the past 12 months, quality of life. Patients or first degree relatives who report having had a colonoscopy or bowel cancer screening test in the past 12 months will be asked for permission to contact their doctor to verify the timing and results of the test. A random sample of 10% of self reported screening will be verified in this way to determine accuracy of self reported screening behavior.
Baseline characteristics of index patients and first degree relatives will be reported for intervention and control groups. Aim 1. The proportion of index patients and the proportion of first degree relatives who undertake screening/surveillance tests in line with NHMRC recommendations for their risk category in years 1, 2 and 3 following diagnosis in the index case will be compared between the two experimental groups using the chi-square test. Aim 2. Sociodemographic characteristics, screening history, risk status, concurrent engagement in alternative programs of surveillance and symptom status, of those who are appropriately and inappropriately screened at each time point will be compared using the chi-square test for categorical variables and the t-test or a non-parametric equivalent for continuous variables. Logistic regression will then be used to examine factors associated with appropriate screening/surveillance while adjusting for potential confounders. Variables will be included in the initial logistic model if they have a p value of 0.2 or less on univariate analyses, with backward stepwise methods used to exclude variables with a p value of 0.10 or more on the likelihood ratio test. Analysis will be undertaken separately for the index cases and for first degree relatives. Analyses will be adjusted for clustering of relatives within index patients using Generalised Estimating Equations.
Approximately 3000 eligible new cases of colorectal cancer are expected to be identified through the Victorian Cancer Registry each year. Based on previous experience, 3% of clinicians will refuse permission for their patient to be approached, therefore we expect permission to contact 2900. Based on prior research, we expect 60% of patients contacted by the VCR agree to be contacted by the research team and 70% of these consent to participate in the study. Recruitment will be undertaken for 12 months, thus 1200 index cases will be recruited in total, or 600 per group. Aim 1. Assuming that 20% of participants will be lost at each follow-up time (leaving 480, 380, 300 index cases per group available at 12, 24, 36 month), the study will have 80% power, with a 5% significance level, to detect a difference of 10% between experimental groups in the proportion of index cases appropriately surveillance at 12 month follow-up, 12% difference at 24 and 36 month follow-up. It is estimated that there will be an average of 2 eligible first degree relatives for each index case, and that 70% of these will consent and be available for 12 month follow-up, with a 20% loss to follow-up at each of the remaining assessment times. This will provide 1680, 1350 and 1050 relatives at 12, 24 and 36 months. Allowing for a design effect of 2 will provide an effective sample size of 400, 330 and 260 per group at each time point, which is adequate to detect a 10% difference in the proportion of first degree relatives appropriately screened at 12 months, 12% at 24 months and 13% at 36 months with 5% significance level and 80-85% power. A 10% difference at the first follow-up is considered to be clinically meaningful, given the burden of disease from colorectal cancer and the intervention effect is anticipated to increase over time, as the intervention will continue for the three year follow-up period. Aim 2. Assuming at least one third of individuals will be appropriately screened at each follow up time, the study will also have at least 80% power to detect differences in characteristics of those who are and are not followed up/screened appropriately of 10-13% for binary variables and 0.2-0.25 standard deviations for continuous variables for index cases, and 11-14% for binary variables and 0.22 -0.27 standard deviations for continuous variables for FDRs.
Cost effectiveness analysis
The cost-effectiveness analysis will examine the null hypothesis that the mean cost-effectiveness of the proposed health care intervention is no different to the mean cost-effectiveness of usual care. The perspective adopted for this analysis is societal and will encapsulate the viewpoint of the health sector in which resources associated with the provision of care (i.e., colorectal screening) are combined with cost-savings arising from that care, and the viewpoint of the patient in terms of out of pocket expenditures associated with treatment. Costs will then be compared with changes in patients' quality adjusted life years.