Skip to main content
. 2015 Jun 5;69(10):985–992. doi: 10.1136/jech-2014-205309

Figure 2.

Figure 2

Funnel plots of predicted random intercepts of mortality within 1 year of diagnosis, controlling for (A) covariates (age, year of diagnosis, sex and deprivation) (model 2), (B) covariates and treatment by curative surgery (model 3), (C) covariates and spending on lung cancer at Primary Care Trust level (model 4) for patients with non-small cell lung cancer (15–99 years) diagnosed during 1998–2006 in England. The target is fixed as zero, which specifies the expected value of the random effect.