Study funding and approvals
This study is funded by Macmillan Cancer Support, and is sponsored by the University of Leeds. The study protocol has received ethical and governance approvals from the Leeds East NHS Research Ethics Committee (ref. 10/H1306/65), and from the Leeds Teaching Hospitals NHS Trust and Calderdale and Huddersfield NHS Foundation Trust Research and Development Departments. Small amendments to the protocol, concerning the methods of recruitment and the PROMs used, have also subsequently been approved and are included in this paper.
Patients will be recruited from the Yorkshire Cancer Network (YCN), which covers a population of approximately 2.6 million in the North and West Yorkshire areas of England, UK. North and West Yorkshire include ethnically and socioeconomically diverse urban and rural areas. Patients will be recruited from five hospitals in the Leeds Teaching Hospitals NHS Trust, which is the YCN Cancer Centre, and from two hospitals in the Calderdale and Huddersfield NHS Foundation Trust, which is one of six YCN Cancer Units.
Participants and sample size
Patients will be eligible for study inclusion if they are: (1) over 17 years of age, (2) diagnosed with breast, colorectal or prostate cancer, (3) within the last 6 months, (4) suitable for treatment with curative intent, and (5) English literate. Patients will be excluded if they lack the capacity to give informed consent (e.g. due to psychopathology, cognitive dysfunction, learning difficulties). The primary aim of the study is to test the feasibility of the ePOCS system, and one of the key feasibility outcomes is the number of patients who consent to join-up (as a proportion of all those eligible and invited). The recruited sample size is therefore an unknown outcome, rather than a predetermined target; we aim to approach all eligible consecutive patients and to recruit all those who are willing to consent. However, we have estimated that we may expect to recruit around 600 patients. The number of adults newly diagnosed with breast, colorectal or prostate cancer, within a recent 12 month period at the two participating NHS Trusts, is almost 4,000 (specifically 3,839). However, many patients will not meet the full eligibility criteria (e.g. some patients will be non-English literate, unable to provide informed consent etc.) (3,839 × .50 = 1,919), and we plan to recruit patients for approximately eight months, rather than a whole year (1,919 × .66 = 1,266). Previous similar PROMs-based studies run by our research group have yielded approximately 70% consent rates [25
] (1,266 × .70 = 886), although not all patients will have sufficient interest in and/or access to the internet to consent to join ePOCS-most likely those without home access [29
] (886 × .70 = 620).
Patient Reported Outcome Measures (PROMs)
Patients will be asked to complete a range of generic, cancer-specific and cancer diagnosis-specific PROMS, representative of those likely to be administered in future research studies which use the ePOCS system. Patients will be asked to complete PROMs at three time-points; as soon as possible after consent (≤ 6 months post-diagnosis; T1), 9 months post-diagnosis (T2), and 15 months post-diagnosis (T3). The specific measures are detailed below, and assess a variety of health and quality-of-life domains, including pain, fatigue, psychological well-being, physical, social and sexual functioning and financial concerns.
Illness Perception Questionnaire-Revised (IPQ-R)
The IPQ-R [30
] assesses patients' personal beliefs and expectations about their illness (e.g. about its controllability and consequences), and can be adapted to assess perceptions about any illness (i.e. in this instance, cancer). The IPQ-R comprises nine subscales, eight of which will be used in this study; the 'causes' subscale will be omitted due to concern from patients on our study steering group that enquiring about the (perceived) causes of a patient's cancer would cause distress; this use of subscales is valid (Moss-Morris, personal communication, 1 July 2010). The IPQ-R (minus the causes subscale) comprises 38 statements (e.g. "my cancer is a serious condition", "my cancer will improve in time") rated on a scale of 1 (strongly disagree)
to 5 (strongly agree)
, and 14 symptoms (e.g. 'breathlessness', 'headaches') rated on a yes
(0) scale, with respect to patients' views at the present moment. The IPQ-R, and its predecessor the IPQ [31
], have been shown to be reliable and valid, and to predict various aspects of illness adaptation and recovery, in a range of samples including cancer patients [30
]. Patients will be asked to complete the IPQ-R at T1.
EuroQol-5D (EQ-5D), version 2
The EQ-5D [38
] is a 6-item generic measure of health status which assesses mobility, self-care, usual activities, pain/discomfort and anxiety/depression, using a three-option response format according to the severity of problems experienced that day (no problems, some problems, severe problems). The EQ-5D also includes a visual analogue scale on which health state today
is rated from 0 (worst imaginable health state
) to 100 (best imaginable health state
). The EQ-5D is an internationally used measure, that can be employed in both the clinical and economic evaluation of health care, and which has been shown to be reliable and valid in several previous studies with cancer patients [40
]. Patients will be asked to complete the EQ-5D at all three time-points.
Medical Outcomes Study 36-item Short-Form Health Survey, version 2 (SF-36v2)
The SF-36v2 [44
] is a generic measure of health status and functioning which assesses eight domains including physical functioning, pain and mental health. The measure comprises 36 items (e.g. "have you been happy", "did you feel worn out") rated on a variety of Likert-type response scales (e.g. excellent
, all of the time
- none of the time
), primarily with respect to the past 4 weeks. The SF-36v2, and its predecessor the SF-36, are internationally used measures with extensive normative data, and have been demonstrated reliable and valid in numerous previous studies with cancer patients [45
]. Patients will be asked to complete the SF-36v2 at T2 and T3.
European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ), version 3
The EORTC-QLQ [51
] is a cancer-specific measure assessing health-related functioning and symptoms, which includes a generic core questionnaire and numerous diagnosis and treatment specific modules. This study will use the breast (EORTC-QLQ-BR23), colorectal (EORTC-QLQ-CR29) and prostate (EORTC-QLQ-PR25) modules, each of which contain between 20 and 27 questions (depending on item branching). In addition to the relevant diagnosis-specific module, patients will be asked to complete six subscales from the core questionnaire (nausea and vomiting, dyspnoea, insomnia, appetite loss, constipation and diarrhoea; 7 items), and two from the ovarian module (peripheral neuropathy and chemotherapy side effects; 7 items); this use of subscales is acceptable (EORTC, personal communication, 19 July 2010). All EORTC-QLQ items (e.g. "did you have a dry mouth", "has weight gain been a problem for you") are rated on a scale of 1 (not at all
) to 4 (very much
) with respect to the past week or month. The EORTC-QLQ is an internationally used measure with sound psychometric properties [5
]. Patients will be asked to complete the EORTC-QLQ modules and subscales at T2 and T3.
Social Difficulties Inventory (SDI-21)
The SDI-21 [27
] assesses various everyday difficulties commonly experienced by cancer patients, including relationship difficulties, domestic problems and financial worries; 21 questions (e.g. "have you felt isolated", "have you had any financial difficulties") are answered on a 0 (no difficulty
) to 3 (very much
) scale with respect to the past month. The SDI-21 has been demonstrated reliable and valid, has a clinically meaningful scoring system (including cut-off and minimal change scores indicative of need for action), and is highlighted in the NCSI as a potentially useful screening measure [1
]. Patients will be asked to complete the SDI-21 at T2.
Quality of Life in Adult Cancer Survivors (QLACS) scale
The QLACS [56
] measures health-related quality-of-life in seven generic and five cancer-specific domains, including cognitive problems, social avoidance and appearance. QLACS comprises 47 items (e.g. "you felt tired a lot", "you had difficulty doing activities that require concentrating") rated on a scale of 1 (never
) to 7 (always
) with respect to the past 4 weeks. QLACS is a relatively new measure, although is flagged as potentially useful in the NCSI, and evidence to date indicates acceptable psychometric properties in samples including breast, colorectal and prostate cancer patients [1
]. Patients will be asked to complete the QLACS at T3.
Sociodemographic details (and clinical information)
Patients will be asked to provide information about their ethnicity, relationship status and level of education at T1. Other sociodemographic details (e.g. gender, age, postcode) and clinical information (e.g. date and type of cancer diagnosis, treatment regimens) will be obtained from participants' medical records (following their explicit permission, recorded on the consent form).
Patients will be asked one question about their pre-diagnosis employment status at T1 (e.g. full-time employment, retired), and two questions about their current employment status at all three time-points (e.g. full-time employment, retired, currently working less or more hours than usual).
Past use of mental health services
Patients will be asked two questions about their past utilisation of mental health care services at T1; questions enquire about lifetime use, and use over the last 12 months, and originate from the National Comorbidity Survey [59
Recent use of health and social services
Patients will be asked 20 questions about their use of health and social services at T3. The questions have been devised with health economist colleagues from the University of Leeds, and aim to help estimate some of the financial costs of cancer and its treatment, both to patients and the welfare state. Questions enquire about patients' frequency of use of medications and services (e.g. district nurses, social workers, residential homes, hospices), and the cancer-related costs incurred by patients and carers (e.g. due to increased home-heating costs, and travel and time off work for hospital appointments), over the last 3 months.
Patient feedback on the ePOCS system
At the end of the study, participants will be offered the opportunity to provide comment on the ePOCS system through a 'feedback questionnaire' containing a mix of 28 closed and open questions (e.g. "what did you like about the electronic questionnaire system", "would you have preferred to complete the questionnaires on paper"). For patients who join the feasibility study and use the ePOCS system to complete PROMs, we want to know about their experiences, and whether they have any suggestions for the system's improvement. For patients who join the study but do not go online and complete (all the) PROMs, we would like to know if there were any system-related reasons for this, and if there are things we can change to make the system easier to use. It will be important to keep feedback on the system distinct from the system itself, and to include those patients who may join the study but do not in fact engage with the system and complete PROMs. Therefore, all consented patients (who have not actively withdrawn from the study) will be posted a paper copy of the feedback questionnaire, after completion of the T3 PROMs (or after the time-window for completion has expired). To minimise burden, a pre-paid addressed envelope will be included for return, and non-returns will not be followed-up with reminders.
Recruitment will be undertaken by oncology clinical care teams, at the two participating NHS Trusts, supported by dedicated research nurses funded by the West Yorkshire Comprehensive Local Research Network. Eligible patients will be identified during discussions in routine MDT meetings and/or through consultation of medical notes. All eligible patients will be given a comprehensive information sheet containing the contact details of the ePOCS study team, and which emphasises the voluntary nature of participation, and patients' right to withdraw consent, at any time, without the need for explanation, and without their personal care being affected. Patients who wish to join the study will be asked to read, complete and sign a consent form; this will be countersigned by the recruiting research nurse/clinician, and a copy filed in patients' medical notes. Participants' General Practitioners will be sent a letter informing them of their patients' participation in the study. Clinical care teams will provide anonymous information about those patients who decline to participate (gender, age, postcode, date and type of cancer diagnosis); if forthcoming, the recruiting teams will also note any reasons offered by patients for non-consent (e.g. lack of time, no interest in computers).
Prior to this study, we conducted a small opinion-gathering interview study (NHS Research Ethics Committee ref. 09/H1313/73) to obtain advice from oncology clinicians and patients about the 'best' time and person in secondary care to approach patients about joining ongoing PROMs-based research. Following the strong unanimous views expressed in that study, patients will first be approached about participation in the current feasibility study by an oncology clinician with whom they are familiar (e.g. consultant, registrar, specialist nurse) and, if possible, with whom they have established a 'good relationship'. Wherever possible, patients will be approached about the study in-person, typically during a routine outpatient appointment; if this is not possible, patients will be sent a letter about the study, signed by their consultant. Perhaps unsurprisingly, given that every patient is an individual, and every care pathway is unique, opinion in the interview study did not converge on a single best time for approach. Clinicians and patients did, however, agree on what were non-optimal times; notably, very close to the time of diagnosis (i.e. within a few weeks), and when patients are facing uncertainty (e.g. when the treatment course is yet to be decided, when the results of a short-term intervention like surgery are as yet unknown). The general consensus was that patients several weeks out from diagnosis, who are well-embarked on a decided course of treatment, will be open to receiving information about the study. However, the exact time-point when this will be (e.g. 4 weeks or 3 months post-diagnosis) will differ for each patient. The clinical teams recruiting patients will use the findings from our preparatory interview study to inform and guide their recruitment approach.
Consenting patients will be given a unique ID and password on a postcard, which contains instructions for accessing the ePOCS website and our study team's contact details. Patients will be asked to log-on to the ePOCS system and complete the T1 PROMs as soon as possible. Patients will then be sent an email or letter (depending on their preferred mode of correspondence) when they are approaching (i.e. 3 weeks before) 9 months (T2) and 15 months (T3) post-diagnosis, inviting them to log-on to the system and complete PROMs again. At each time-point, patients will have up to 6 weeks to complete the PROMs (i.e. 3 weeks either side of the exact time-point; aside from the first time-point, which is variable, as patients will be recruited at any time within 6 months post-diagnosis). Up to two email/letter reminders will be sent to patients who do not complete the PROMs. For all PROMs questions, patients will have the option to respond 'I would prefer not to answer this question'. Completion of the PROMs at each time-point is likely to take approximately 20-30 minutes in total; patients can spread-out completion if they wish (e.g. over a number of days). All correspondence will remind patients that their PROMs responses are not fed back to their clinical team, and to contact their doctor or nurse if they have any problems or concerns about their health. The study website also contains links to relevant supportive organisations (e.g. Macmillan Cancer Support, Prostate Cancer Charity). At the end of the study, patients will be sent a paper feedback questionnaire inviting their thoughts and opinions on the ePOCS system.
Members of our study team will administrate the ePOCS system during feasibility testing. We will use the Tracker on a daily basis to monitor patients' PROMs completion progress, and to prepare and send the necessary related communications (e.g. reminder notices, thank you acknowledgments). Our study team will also be accessible to patients (via email, telephone and/or letter) to provide assistance with using the ePOCS system.
The current feasibility study is funded for 2 years. If the ePOCS system proves feasible, we would like to expand the research by, amongst other things, following patients for longer than 15 months post-diagnosis. If we obtain further funding to do this, we would like to invite the patients who participate in this feasibility study to join our future studies (and complete more PROMs at longer-term follow-ups). We will explicitly seek patients' permission to contact them in the future (for up to 10 years after the study) if we obtain further funding-to inform them about, and to seek their consent to participate in, new ePOCS-related research studies. Patients will be informed of this via the study information sheet, and if they give permission for future contact, will be asked to sign a separate statement on the consent form to indicate this.
Study outcomes and data analysis
The feasibility of the ePOCS system will be assessed by examining a number of different outcomes. Notably: (1) the proportion of patients recruited into the system, and their representativeness, relative to all those invited to join, (2) the proportion and representativeness of patients who remain in the system and complete PROMs at the different time-points, relative to all those who join, (3) the completeness, timeliness and reliability of the PROMs data obtained at each time-point (e.g. extent of missing responses, internal reliability of responses), (4) patients' views and opinions on the system provided via the feedback questionnaire, (5) the success and reliability of the underpinning informatics (e.g. frequency and nature of technical problems, frequency and nature of patient enquiries requesting assistance in using the system), and (6) the system running costs. The efficiency and practicality of the patient recruitment procedures (i.e. hospital-based clinician-led recruitment) will also be examined.
The different feasibility outcomes will be primarily analysed quantitatively using descriptive statistics such as proportions, frequencies and means (e.g. proportion of invited patients who consent, frequency of patient requests for help with using the system, mean running costs per month), chi-square and t-tests (e.g. to examine group differences in gender, age etc. between consenters and non-consenters) and measures of internal reliability (e.g. Cronbach's alpha, to assess quality of PROMs responses). Open-ended patient feedback questions will be examined qualitatively, as will the type of reasons offered for non-consent and withdrawal, and the nature of technical problems and patient difficulties with using the system. The efficiency and practicality of the patient recruitment procedures will also be examined qualitatively (including using process mapping tools and software).
If the system proves feasible, and a sufficient number of patients provide data of satisfactory quality, we will analyse these data for two main purposes: (1) to explore and check the psychometric properties (e.g. reliability, validity) of the newer, less well-established PROMs requiring of further psychometric analyses, and (2) to assess and describe patients' self-reported health and quality-of-life outcomes, and explore if and how they relate to disease, treatment and/or individual differences characteristics.
Psychometric analyses will be undertaken using descriptive statistics such as proportions and means (e.g. to determine the presence of floor and ceiling effects), measures of correlation (e.g. between responses to different PROMs to determine convergent and discriminant validity) and measures of internal reliability (e.g. Cronbach's alpha). The health and quality-of-life outcomes reported by patients will be assessed and described using statistics such as proportions, means and standard deviations (e.g. sample means and standard deviations of PROMs scores) and t-tests (e.g. to compare the sample scores with population norms). If and how the outcomes relate to disease, treatment and/or individual differences characteristics (e.g. sociodemographic and psychological variables) will be explored using multiple and logistic regression analyses (e.g. to explore if illness perceptions within 6 months of diagnosis are predictive of health outcomes 15 months post-diagnosis) and/or analysis of variance (ANOVA) (e.g. to examine if breast, colorectal and prostate cancer patients experience significantly different levels of social difficulties 9 months post-diagnosis).
Study organisation and management
The study research team and wider steering group collectively includes expertise in psychosocial oncology research, (electronic) PROMs, clinical oncology, nursing, data management, statistics, epidemiology, health informatics and cancer registration. The conduct and progress of the study will be discussed and reviewed in fortnightly to monthly meetings of the core day-to-day research team and chief investigator, and in quarterly meetings of all study co-investigators/the wider steering group. The steering group includes four patient representatives, and we will provide verbal reports on the study at biannual meetings of our wider research group's patient and carer Research Advisory Group. The study is included in the UK National Institute for Health Research (NIHR) Clinical Research Network (CRN) Portfolio, and we will provide monthly anonymised reports on study accrual to the NIHR CRN office. We will also provide regular reports on study recruitment and progress to the National Cancer Research Institute clinical studies groups (the breast, colorectal and prostate cancer site-specific groups, and the psychosocial oncology group).