csd {SEERaBomb}

Cancer risk vs years Since Diagnosis of other cancer

Description

Computes relative risks (RR) of 2nd cancers over specified intervals of times since diagnoses of a 1st cancer. 2D spline fits are used to produce expected cases E controlling for background risk dedepence on age and calendar year. RR is then O/E where O is the number of observed cases.

Usage

csd(seerSet,brkst=c(0),brksy=c(1975),brksa=c(0),trts=NULL,
                    PYLong=FALSE,firstS="all",exclUnkSurv=FALSE) 

Arguments

seerSet

A seerSet object produced by mk2D().

brkst

Vector of breaks in years used to form Time intervals/bins since diagnosis. An upper limit of 100, well beyond 40 years of SEER follow up currently available, is assumed/added to brkst internally, and should thus not be in brkst.

brksy

Vector of breaks used to form groups of calendar Year at 1st cancer diagnosis intervals/bins. An upper limit of yearEnd (last year in SEER; a seerSet field) is assumed/added to brksy internally.

brksa

Vector of breaks used to form groups of Age at 1st cancer diagnosis intervals/bins. An upper limit of 126 is assumed.

trts

Character vector of treatments of interest. Default of NULL => all levels in seerSet's canc$trt.

PYLong

Set true if in addition to O and E for each tsd interval you also want PY strips for each individual; having these big dataframes slows saving seerSets, so only fetch if needed.

firstS

Character vector of first cancers of interest. Default of "all" sets it to the vector of all cancers in the seerSet field cancerS, which is created when the object is first created by seerSet().

exclUnkSurv

Set true if you wish to exclude all cases with unknown survival times as a marker of bad data.

Value

The input with an L component added to it or extended it if it already existed. Each component of L is a nested list of lists that can yield second cancer relative risks as a function of time since 1st cancer diagnosis. The most recent component of L is also provided as a data.frame seerSet$DF produced internally using getDF.

Note

This function was developed with support from the Cleveland Clinic Foundation.

Author(s)

Tom Radivoyevitch ([e-mail])

See Also

SEERaBomb-package, mk2D,seerSet

Examples

## Not run: 
library(SEERaBomb)
pm=simSeerSet()
pm=mk2D(pm)
pm$canc
pm=csd(pm,brkst=c(0,5),brksy=c(1975,2000),brksa=c(0,50),trts=c("noRad","rad"))
pm
library(ggplot2)
theme_set(theme_gray(base_size = 16)) 
theme_update(legend.position = "top")
g=qplot(x=t,y=RR,data=subset(pm$DF,cancer1=="A"&cancer2=="B"),col=trt,geom=c("line","point"),
        xlab="Years Since First Cancer Diagnosis",ylab="Relative Risk")
g=g+facet_grid(yearG~ageG,scales="free")+geom_abline(intercept=1, slope=0) 
g+geom_errorbar(aes(ymin=rrL,ymax=rrU,width=.15))

## End(Not run)