Self-selection in a population-based cohort study: impact on health service use and survival for bowel and lung cancer assessed using data linkage

• DOI: 10.1186/s12874-018-0537-3
• Publication Date: 2018-08-08T07:53:53Z
• Authors: Nicola Creighton; Stuart Purdie; Matthew Soeberg; Richard Walton; Deborah Baker; Jane Young
• Keywords: Male; Medicine (General); Lung Neoplasms; Information Storage and Retrieval; Cohort Studies; 03 medical and health sciences; R5-920; 0302 clinical medicine; Humans; Registries; Cancer; Aged; Selection bias; Aged, 80 and over; Health care utilisation; Sociodemographic factors; Health Services; Middle Aged; 16. Peace & justice; Survival Analysis; 3. Good health; Hospitalization; Cohort studies; Female; New South Wales; Colorectal Neoplasms; Emergency Service, Hospital; Research Article

In contrast to aetiological associations, there is little empirical evidence for generalising health service use associations from cohort studies. We compared the of study participants diagnosed with bowel or lung cancer source population people these cancers in New South Wales (NSW), Australia assess representativeness participants. Population-based registry data NSW residents aged ≥45 years at diagnosis were linked 45 and Up Study, a population-based (N~ 267,000). measured hospitalisation, emergency department (ED) attendance all-cause survival, risk factor outcomes using administrative population. assessed bias prevalence ratios relative frequency (RRF) odds (ROR), respectively. People major cities, non-English speaking countries comorbidites under-represented among (n = 1837) 969) by 20–50%. Cohort had similar hospitalisation ED One-year survival after surgical resection was similar, but up 25% higher post-diagnosis (lung 3-year survival: RRF 1.24, 95% confidence interval 1.12,1.37). Except area-based socioeconomic position, factors measures appeared relatively unbiased. Absolute non-representative sample showed bias. Non-comparability non-participants, such as may estimates associations. Primary outpatient care be more vulnerable