A flexible parametric approach to examining spatial variation in relative survival

Susanna M. Cramb, Kerrie L. Mengersen, Paul C. Lambert, Louise M. Ryan, Peter D. Baade

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Most of the few published models used to obtain small-area estimates of relative survival are based on a generalized linear model with piecewise constant hazards under a Bayesian formulation. Limitations of these models include the need to artificially split the time scale, restricted ability to include continuous covariates, and limited predictive capacity. Here, an alternative Bayesian approach is proposed: a spatial flexible parametric relative survival model. This overcomes previous limitations by combining the benefits of flexible parametric models: the smooth, well-fitting baseline hazard functions and predictive ability, with the Bayesian benefits of robust and reliable small-area estimates. Both spatially structured and unstructured frailty components are included. Spatial smoothing is conducted using the intrinsic conditional autoregressive prior. The model was applied to breast, colorectal, and lung cancer data from the Queensland Cancer Registry across 478 geographical areas. Advantages of this approach include the ease of including more realistic complexity, the feasibility of using individual-level input data, and the capacity to conduct overall, cause-specific, and relative survival analysis within the same framework. Spatial flexible parametric survival models have great potential for exploring small-area survival inequalities, and we hope to stimulate further use of these models within wider contexts.

Original languageEnglish
Pages (from-to)5448-5463
Number of pages16
JournalStatistics in Medicine
Volume35
Issue number29
DOIs
Publication statusPublished - 20 Dec 2016
Externally publishedYes

Fingerprint

Queensland
Bayes Theorem
Survival Analysis
Registries
Colorectal Neoplasms
Linear Models
Lung Neoplasms
Breast Neoplasms
Survival Model
Parametric Model
Neoplasms
Colorectal Cancer
Frailty
Hazard Function
Lung Cancer
Generalized Linear Model
Breast Cancer
Bayesian Approach
Hazard
Model

Cite this

Cramb, S. M., Mengersen, K. L., Lambert, P. C., Ryan, L. M., & Baade, P. D. (2016). A flexible parametric approach to examining spatial variation in relative survival. Statistics in Medicine, 35(29), 5448-5463. https://doi.org/10.1002/sim.7071
Cramb, Susanna M. ; Mengersen, Kerrie L. ; Lambert, Paul C. ; Ryan, Louise M. ; Baade, Peter D. / A flexible parametric approach to examining spatial variation in relative survival. In: Statistics in Medicine. 2016 ; Vol. 35, No. 29. pp. 5448-5463.
@article{b62c63913dc949d59f9598f836c81851,
title = "A flexible parametric approach to examining spatial variation in relative survival",
abstract = "Most of the few published models used to obtain small-area estimates of relative survival are based on a generalized linear model with piecewise constant hazards under a Bayesian formulation. Limitations of these models include the need to artificially split the time scale, restricted ability to include continuous covariates, and limited predictive capacity. Here, an alternative Bayesian approach is proposed: a spatial flexible parametric relative survival model. This overcomes previous limitations by combining the benefits of flexible parametric models: the smooth, well-fitting baseline hazard functions and predictive ability, with the Bayesian benefits of robust and reliable small-area estimates. Both spatially structured and unstructured frailty components are included. Spatial smoothing is conducted using the intrinsic conditional autoregressive prior. The model was applied to breast, colorectal, and lung cancer data from the Queensland Cancer Registry across 478 geographical areas. Advantages of this approach include the ease of including more realistic complexity, the feasibility of using individual-level input data, and the capacity to conduct overall, cause-specific, and relative survival analysis within the same framework. Spatial flexible parametric survival models have great potential for exploring small-area survival inequalities, and we hope to stimulate further use of these models within wider contexts.",
keywords = "Australia, Bayesian, cancer, flexible parametric, relative survival, small area",
author = "Cramb, {Susanna M.} and Mengersen, {Kerrie L.} and Lambert, {Paul C.} and Ryan, {Louise M.} and Baade, {Peter D.}",
year = "2016",
month = "12",
day = "20",
doi = "10.1002/sim.7071",
language = "English",
volume = "35",
pages = "5448--5463",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley & Sons",
number = "29",

}

Cramb, SM, Mengersen, KL, Lambert, PC, Ryan, LM & Baade, PD 2016, 'A flexible parametric approach to examining spatial variation in relative survival', Statistics in Medicine, vol. 35, no. 29, pp. 5448-5463. https://doi.org/10.1002/sim.7071

A flexible parametric approach to examining spatial variation in relative survival. / Cramb, Susanna M.; Mengersen, Kerrie L.; Lambert, Paul C.; Ryan, Louise M.; Baade, Peter D.

In: Statistics in Medicine, Vol. 35, No. 29, 20.12.2016, p. 5448-5463.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - A flexible parametric approach to examining spatial variation in relative survival

AU - Cramb, Susanna M.

AU - Mengersen, Kerrie L.

AU - Lambert, Paul C.

AU - Ryan, Louise M.

AU - Baade, Peter D.

PY - 2016/12/20

Y1 - 2016/12/20

N2 - Most of the few published models used to obtain small-area estimates of relative survival are based on a generalized linear model with piecewise constant hazards under a Bayesian formulation. Limitations of these models include the need to artificially split the time scale, restricted ability to include continuous covariates, and limited predictive capacity. Here, an alternative Bayesian approach is proposed: a spatial flexible parametric relative survival model. This overcomes previous limitations by combining the benefits of flexible parametric models: the smooth, well-fitting baseline hazard functions and predictive ability, with the Bayesian benefits of robust and reliable small-area estimates. Both spatially structured and unstructured frailty components are included. Spatial smoothing is conducted using the intrinsic conditional autoregressive prior. The model was applied to breast, colorectal, and lung cancer data from the Queensland Cancer Registry across 478 geographical areas. Advantages of this approach include the ease of including more realistic complexity, the feasibility of using individual-level input data, and the capacity to conduct overall, cause-specific, and relative survival analysis within the same framework. Spatial flexible parametric survival models have great potential for exploring small-area survival inequalities, and we hope to stimulate further use of these models within wider contexts.

AB - Most of the few published models used to obtain small-area estimates of relative survival are based on a generalized linear model with piecewise constant hazards under a Bayesian formulation. Limitations of these models include the need to artificially split the time scale, restricted ability to include continuous covariates, and limited predictive capacity. Here, an alternative Bayesian approach is proposed: a spatial flexible parametric relative survival model. This overcomes previous limitations by combining the benefits of flexible parametric models: the smooth, well-fitting baseline hazard functions and predictive ability, with the Bayesian benefits of robust and reliable small-area estimates. Both spatially structured and unstructured frailty components are included. Spatial smoothing is conducted using the intrinsic conditional autoregressive prior. The model was applied to breast, colorectal, and lung cancer data from the Queensland Cancer Registry across 478 geographical areas. Advantages of this approach include the ease of including more realistic complexity, the feasibility of using individual-level input data, and the capacity to conduct overall, cause-specific, and relative survival analysis within the same framework. Spatial flexible parametric survival models have great potential for exploring small-area survival inequalities, and we hope to stimulate further use of these models within wider contexts.

KW - Australia

KW - Bayesian

KW - cancer

KW - flexible parametric

KW - relative survival

KW - small area

UR - http://www.scopus.com/inward/record.url?scp=84982921033&partnerID=8YFLogxK

U2 - 10.1002/sim.7071

DO - 10.1002/sim.7071

M3 - Article

VL - 35

SP - 5448

EP - 5463

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 29

ER -