The value of local registry data for describing cervical cancer management and outcomes over three decades in Australia.

Abstract

Registry data on invasive cervical cancers (n=1,274) from four major hospitals (1984-2012) were analysed to determine their value for informing local service delivery in Australia. The methodology comprised disease-specific survival analyses using Kaplan-Meier product-limit estimates and Cox proportional hazards models and treatment analyses using logistic regression. Five- and 10-year survivals were 72% and 68%, respectively, equating with relative survival estimates for Australia and the USA. Most common treatments were surgery and radiotherapy. Systemic therapies increased in recent years, generally with radiotherapy, but were less common for residents from less accessible areas. Surgery was more common for younger women and early-stage disease, and radiotherapy for older women and regional and more advanced disease. The proportion of glandular cancers increased in-step with national trends. Little evidence of variation in risk-adjusted survival presented over time or by Local Health District. The study illustrates the value of local registry data for describing local treatment and outcomes. They show the lower use of systemic therapies among residents of less accessible areas which warrants further investigation. Risk-adjusted treatment and outcomes did not vary by socio-economic status, suggesting equity in service delivery. These data are important for local evaluation and were not available from other sources.