variaTIon in Referral Thresholds in the Chronic Lymphocytic LeukaEmia Pathway

Overview

This study will use surveys within primary and secondary care to explore referral patterns and healthcare professional (HCP) decision-making and advice for haematological malignancies, identify inequalities, and establish best practice for CLL referrals and specialist support. The project will allow us to understand and 'map' variations in practice across the West Midlands region with regard to referrals for CLL from primary care physicians to haematology services in secondary care. Specific objectives are to: 1. Identify service variations by using surveys to assess differences in CLL service provision across primary and secondary care in the West Midlands 2. Investigate service variations by describing the key characteristics of responding primary care practices (practice size, number of GPs, Integrated Care Board (ICB) locality, deprivation quartile) 3. Propose strategies to improve equity that will enhanced patient outcomes and reduce inequalities in care for CLL patients, aiming to ensure consistent standards for referral and management across different care settings.

Chronic lymphocytic leukaemia (CLL) is a haematological malignancy characterised by a clonal proliferation of CD5+ mature B lymphocytes which accumulate in the bone marrow, ymph nodes and blood. CLL is the most common adult leukaemia, with an annual incidence of around 4,500 cases in the UK.1 The median age at diagnosis is 72, and around 40% of cases affect people aged 75 and above, with prevalence substantially higher in males than females. CLL develops slowly and is often symptomatic, with a high proportion of cases found incidentally during routine full blood count (FBC) tests and identified by a persistent lymphocytosis. An advanced stage of disease is associated with bulky lymphadenopathy, fatigue, unexplained weight loss, fever, night sweats, organomegaly and cytopenias (anaemia, thrombocytopenia and neutropenia) due to bone marrow infiltration. Around two thirds of patients diagnosed with CLL do not need treatment during their lifetime, however the immune perturbation associated with the condition is associated with an increased risk of secondary malignancy and infection morbidity, even in those with early stage disease.2 If a CLL diagnosis is recognised, intervention can be taken to mitigate against these associated conditions. These include appropriate counselling e.g. taking care with sun exposure to lower skin cancer risk and reduce treatment-induced UV sensitivity. The Department of Health (DH) and British Society for Haematology (BSH) guideline also recommends annual influenza and covid-19 vaccination, as well as immunisation against Pneumococcus and Varicella Zoster at diagnosis to reduce infection morbidity and mortality risk.3 Early implementation of vaccination is known to maximise its effectiveness, thus a timely diagnosis of CLL can reduce the risk of other potentially life-threatening conditions.4 Much of the diagnostic work up and advice for patients regarding ongoing management for early stage CLL could be delivered within Advice and Guidance (A\&G) or as a one-off secondary care consultation. Once diagnosed, early stage CLL that requires active observation can be managed effectively within the primary care setting.5 However, referral rates for CLL can be influenced by both primary care referral practices and secondary care referral acceptance criteria. Variation in accessing services increases health inequalities, and in CLL, patients with lower socioeconomic status are more likely to experience delayed diagnosis, have a greater likelihood of being diagnosed after emergency presentation, and have significantly higher mortality risk than the least deprived patients.6,7 Locally, evidence from University Hospitals Birmingham (UHB) Trust dashboard data previously showed substantial differences in referral rates between primary and secondary care for lymphocytosis in the West Midlands and community follow up practice for early stage CLL in primary care, varying across the region. A\&G services and timely patient review are more likely to be accessed by GP practices located within affluent areas. It is not clear why this variation exists within the West Midlands, thus exploring primary care referral patterns and decision-making for haematological malignancies represents an important area to be addressed, in order to reduce unwarranted variation in patient referrals and CLL management practice. The investigators will develop and disseminate two surveys to collect data to describe a) referral behaviours and understanding of CLL (GPs) \[survey 1\], and b) how different NHS Trusts across the West Midlands manage CLL Clinical Nurse Specialist (CNS) services \[survey 2\]. In addition to benchmarking current practice for CLL referral and management across the region, the study will help to identify practices where direction to local referral guidance for lymphocytosis and lymphadenopathy (recently developed with BSol ICB) may be highlighted. It may also demonstrate the differential knowledge among specialist nurses and how this may be improved in local hospitals. PRIMARY CARE PROVIDER SURVEY (target n=100) Setting/participants The investigators aim to survey all GPs within the Birmingham and Solihull (BSol) ICB area. BSol ICB is organised around a locality model, with around 175 GP practices located across Central, East, North, South, West and Solihull localities. This will equate to a maximum of 800 GPs eligible to receive the primary care survey. A response of 100 surveys would be sufficient to estimate the proportion of GPs in the West Midlands who may be receptive to primary care monitoring for early stage CLL. General practices will be categorised according to key characteristics (e.g. practice size, number of GPs, ICB locality, deprivation quartile) so that the representativeness of survey responses can be assessed. A 'thank you' of £25 will be offered for each survey returned, to reimburse participants for their time. Survey content The GP survey will assess multiple dimensions of CLL referral and management practice in the primary care setting, including: * Referrals: knowledge of local (BSol) CLL referral guidelines and pathways; use of referral to haematology services; mode of referral (e.g. direct referral vs. advice and guidance) * Awareness: understanding of CLL amongst primary care physicians; awareness of requirements for referrals and management of secondary malignancy/vaccination guidelines * Monitoring: receptiveness towards primary care monitoring and 'watch and wait' for early stage CLL Surveys will be designed to minimise burden on participants (estimated time for completion 10-15 minutes), and will use short, tick-box style questions and Likert scales, with optional free text boxes to allow participants to provide additional information to substantiate their responses if they wish. The draft survey will receive input from clinical stakeholders before being finalised, to ensure that question coverage, wording and comprehension is relevant and appropriate. Survey administration All GPs currently practicing in the six localities that constitute BSol ICB will be eligible to participate in the survey. Surveys will be created by, and administered using the Joint Information Systems Committee (JISC) Online Surveys tool (formerly Bristol Online Surveys), for which the University of Birmingham holds an institutional licence. Online surveys have been chosen as the survey platform because it is designed to protect respondent anonymity - the system does not use cookies for survey completion, and information about respondents' IP addresses cannot be accessed. Survey respondents will be able to take part in the survey from any location, at a time of their choice, minimising any potential inconvenience associated with research participation. Online Surveys also allows participants to save their answers at any point and return to completion later. A list of GP practices within BSol ICB, including contact details of the lead GP is publicly available and will be used to populate the distribution list, with the expectation that the lead GP will cascade the survey to others within their practice. The survey will be accessible via a weblink embedded in an invitation email. Consent Written informed consent will not be obtained, and submission of the survey will be taken as consent to participate. The front page of the survey (before any survey questions) will include a Participant Information Sheet (PIS). There will also be a mandatory tick box embedded within the survey for respondents to acknowledge that they have read and understood a series of statements assuring them of the confidential handling of any information they provide in their response; their right to refuse to answer specific survey questions; that they can opt out of submitting a survey response at any time, but that once their survey has been submitted, their response cannot be removed. Participants will not be able to progress into the main body of the survey until the content box has been checked. They will also be required to type their name into a mandatory text box underneath the consent statements, which acts as a simple form of signed consent. CLINICAL NURSE SPECIALIST SURVEY (target n=30) Setting/participants Clinical Nurse Specialists are employed by NHS Trusts. Across the six ICBs that currently constitute the West Midlands region, there are 19 NHS Trusts and a limited number of people employed within the CNS role. As with the GP survey, a £25 honorarium will be offered to CNS survey participants, to reimburse them for the time spent on survey completion. Survey content The aim of the CNS survey is to understand how different NHS Trusts across the West Midlands manage CLL CNS services according to catchment population size. Areas of questioning will include: * CNS levels of confidence and skill in advising CLL patients about their condition and its management * Understanding how specialist CLL clinic and physician input varies across different Trusts * CNS awareness of useful websites and sources of information for patients with CLL e.g. vaccination passport, CLL Support Association and extent to which participants signpost patients to these information sources. As with the GP survey, surveys will be designed to be completed within 10-15 minutes to maximise participation and minimise participant burden. A combination of tick-box questions, Likert scales and free text options will be included. The CNS survey will also be assessed by clinical stakeholders before the final version is developed. Survey administration The CNS survey will be electronic, using the JISC Online Survey platform used for the GP survey but with a dedicated weblink specific to the secondary care survey version. The investigators will use their clinical contacts to send surveys to all known CNSs, with the expectation of receiving around 30 surveys back. The embedded PIS and consent process will be the same for the CNS survey as for the primary care survey. GP and CNS survey analysis Surveys returned from primary care and CNS participants will be analysed descriptively, as the sample size will not be large enough to allow for inferential statistics or sub-group analyses. The investigators expect to be able to characterise the sample in terms of the characteristics of GP practices who have responded (e.g. size, deprivation) and in terms of the NHS Trusts from which secondary care surveys are returned, to describe referral practices, behaviours, decision-making etc. Some stratified analyses may be possible (e.g. less affluent vs. more affluent GP practices, or comparisons across ICB localities), in order to identify any differences between primary care organisations on these factors. Data handling (GP and CNS surveys) Online Surveys assigns each survey respondent a unique identifying number, and does not record any personal information, cookies, IP addresses or email addresses. Participants in the GP and CNS surveys will not be asked to provide any personal information apart from basic demographic data (sex, age group, ethnic group) and some brief information about their job (NHS Trust/GP practice, role, time working in healthcare services). If any survey respondents disclose personal information via their free text responses, these will be removed by the research team at the point of exporting survey responses from the Online Survey database to the database that will be kept securely at University of Birmingham for the purposes of analysis. If respondents have provided any contact details to claim the £25 reimbursement being offered, these details will be removed from the database before analysis. The survey database will be password-protected and stored electronically within a secure folder on the IT network for the Health Services Management Centre at University of Birmingham, using the University of Birmingham's research drive which provides secure, backed up storage of research data. The secure study folder will only be accessible to members of the core research team involved in the management and analysis of the survey data. Data will not be shared outside of University of Birmingham, and anonymised data will be archived under the University of Birmingham's regulations on the management of research data and retained for 10 years.