Institute of Health & Society

Socio-demographic inequalities in the diagnosis of

Exploring socio-demographic inequalities in the diagnosis of sarcoma, with a particular focus on deprivation

Funded by Sarcoma UK.

In a study of national cancer data, we recently revealed a marked association between increasing levels of deprivation and worse early mortality for younger patients with osteosarcoma. The reasons for this worse early mortality are unknown. It is possible that more deprived patients (or those resident in more deprived areas) have different pathways to, and experiences, of diagnosis than more affluent patients, and that this may affect their subsequent experiences and clinical outcomes.

The project will investigate

  1. socio-demographic (including deprivation, age and gender) variations in routes to, and experiences of, diagnosis in bone and soft tissue sarcoma patients
  2. associations between routes to diagnosis and survival
  3. associations between “poor” diagnostic experiences and poorer treatment, rehabilitation and follow-up experiences

The study will use two existing datasets. The first dataset is the national “Routes To Diagnosis” dataset for 2009-14, which includes all bone and soft tissue sarcoma patients in England categorised into diagnostic groups (eg, “Two-week wait”, “Emergency presentation”, “GP referral”, “Other outpatient” and “Inpatient Elective”). We are comparing routes to diagnosis across deprivation categories, age-groups, and gender. Comparisons will be made between bone tumours and soft tissue sarcomas. Associations between diagnostic route and survival will be investigated, to determine whether these could explain relationships between deprivation and survival.

In the second dataset (National Sarcoma Patient Experience Survey), we will analyse variations in patient-reported experiences of diagnosis, and assocations between diagnosis experiences and experiences of treatment, rehabiltation and follow-up. We will use descriptive analyses, multivariable logistic regression and/or multinomial regression, and relative survival models.