The Importance of Full Participation in Childhood Cancer Studies
Author Information
Author(s): GR Law, AG Smith, E Roman
Primary Institution: Leukaemia Research Fund, Centre for Clinical Epidemiology, University of Leeds
Hypothesis
Differential participation between cases and controls can lead to biased estimates of risk.
Conclusion
The study found that parents of participating controls lived in more affluent areas than initially selected controls and their matched cases, indicating potential bias in risk estimates.
Supporting Evidence
- Participation rates were 87% for cases and 72% for controls, indicating a disparity in participation.
- Parents of controls who agreed to participate lived in more affluent areas than those initially selected.
- The study highlights the importance of understanding non-participation to avoid bias in risk estimates.
Takeaway
When studying childhood cancer, it's important to have everyone participate, because if some people don't, it can make the results unfair and confusing.
Methodology
A case-control approach was employed, involving face-to-face interviews with parents of children diagnosed with cancer, and census data was used to analyze participation bias.
Potential Biases
Differential participation may introduce significant bias in estimating risks associated with childhood cancer.
Limitations
The study may not fully account for all factors influencing participation, and the motivations of non-participants are unclear.
Participant Demographics
Participants included children aged 0-14 diagnosed with cancer and their matched controls, with a focus on socio-economic status based on residential area.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website