Cancer does not affect everyone equally. Even within universal healthcare systems like the National Health Service (NHS), we have observed persistent inequalities along the whole cancer journey for decades. Some patients may wait longer than others to get a confirmed diagnosis (delays in diagnosis) or to get treatment after diagnosis (delays in treatment), while some might not receive the care that would benefit them most (suboptimal treatment). Unlike what most researchers think, patient and tumour factors (such as pre-existing conditions and cancer stage) explain only part of these observed inequalities, and healthcare system factors (such as the number of patients registered with the same GP practice) might also have a part to play in these disparities. Using data from primary care (such as GPs), secondary care (hospitals) and cancer registries, we will study how some characteristics of the healthcare system and individual factors play a role together to generate sociodemographic inequalities along the cancer patients’ journey, including delays in diagnosis and access to the best treatment. With more cancer patients now living longer, side effects of cancer treatment are also becoming a more important issue in medium and long terms. We shall explore potential inequalities in side effects related to cancer treatment and build a prediction tool for doctors and patients to make informed decisions. Our study will provide evidence for policymakers and NHS to tackle inequalities in cancer care and outcomes.
Background: Wide inequalities in cancer outcomes have been reported for decades in England. Previous research suggested that individual (patient and tumour) factors explain only part of these observed inequalities and healthcare system-wide issues might also contribute to these disparities.
Objectives: This study aims to investigate the interplay between individual and healthcare system factors on sociodemographic inequalities along the cancer patient pathway, from diagnosis to treatment and post-treatment.
Design: Retrospective cohort design.
Methods: The National Cancer Registration Analysis Service (NCRAS) database will be used to include a cohort of adult patients with the 22 most common cancers from 2014 to the latest date of available data. Sociodemographic exposures include age, sex, ethnicity and socioeconomic deprivation, and outcomes include (i) delays in cancer diagnosis (e.g., advanced stage), (ii) access to timely and optimal systemic treatment and radiotherapy, and (iii) adverse events following these treatments. Exposures and other individual characteristics and healthcare system factors will be retrieved from primary care and secondary care health records..
Statistical analyses: We will apply a three-step analytic plan for each pathway interval. First, we will map the processes between patient characteristics and the system component of interest, which potentially lead to inequalities in cancer outcomes. Second, we shall examine associations between the system components and sociodemographic characteristics of the patients, after adjusting for individual characteristics such as comorbidities. For categorical outcomes, clustering will be accounted for using Generalised Estimating Equations and/or Generalised Linear Mixed Models allowing to quantify the inter-cluster variability. Time-to-event outcomes (e.g., time to treatment or occurrence of adverse events) will be analysed using flexible multivariable hazards models or general hazard structure models, while accounting for the hierarchical data structure. Third, we shall quantify the causal effect of system-level components and individual characteristics (overall and mediated by other factors) using causal inference methods.
- Delays in cancer diagnosis (e.g., referral pathway, route to diagnosis, severity of cancer such as stage)
- Timely access to optimal systemic treatment (chemotherapy, hormonotherapy, immunotherapy) and radiotherapy (e.g., type and duration of treatment, time to decide and start the treatment)
- Identification of adverse events and complications following these treatments. For example, most common ones include fatigue, diarrhoea, skin rashes, itching, pain, loss of appetite, insomnia and depression which are recorded in primary care. More serious ones include autoimmune reactions, infection, anaemia, heart failure, acute kidney diseases, etc., that required hospitalisations and were recorded in secondary care). To deal with potential inconsistent recording of conditions related to side effects, we will consult clinicians who are practicing in the UK to determine the list of clinical codes for these conditions, and we will also explore statistical methods to deal with potential misclassification bias in electronic health records (Beesley and Mukherjee 2022).
Bernard Rachet - Chief Investigator - London School of Hygiene & Tropical Medicine ( LSHTM )
Suping Ling - Corresponding Applicant - London School of Hygiene & Tropical Medicine ( LSHTM )
Aimilia Exarchakou - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Ananya Malhotra - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Clemence Leyrat - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Manuela Quaresma - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
HES Admitted Patient Care;HES Diagnostic Imaging Dataset;NCRAS Cancer Registration Data;NCRAS National Radiotherapy Dataset (RTDS) data;NCRAS Systemic Anti-Cancer Treatment (SACT) data;ONS Death Registration Data;Patient Level Index of Multiple Deprivation Domains;Practice Level Index of Multiple Deprivation