DOCUMENTATION FOR oxford.dat  (free format)


 Variable                        
------------------------------------------------
1  year of birth*                 1944 to 1964

2  age at death                   0 to 9

3  in-utero exposure to x-ray     1 = yes ; 0 = no
		
4  N_Cases                        Number of cases

5  N_Ctls                         Number of controls
									   
----------------------------------------------------

(Taken from pp 239 and 284-289 of Breslow and Day, Volume I)

Summary data from the Oxford Childhood Cancer Survey and 
associated studies reported by Kneale (Kneale GW (1971) 
Problems arising in estimating from retrospective survey data 
the latent period of juvenile cancers initiated by obstetric 
radiography. Biometrics, 27, 563-590) are presented in 
Appendix II.

Cases were ascertained as all children under ten years of age 
in England and Wales who died of cancer (leukaemia or solid 
tumours) during the period 1954-65.

For each of these a neighbourhood control of the same age was 
selected who was alive and well at the time the case died. 
Only "traced" pairs, for whom both case and control mothers 
could be found and interviewed, were analysed. The published 
data ignore the exact pairing but do preserve the 
stratification by age and year of birth.

Exposure in this example is simply a question of whether or 
not the study subjects received in utero irradiation, as 
reported by the mother.

The stratification variables were age at death, from 0 to 9 
years, and year of birth, from 1944 to 1964. Because of the 
limited period of case ascertainment, not all 210 possible 
combinations of these factors appear. For example, among 
childhood cancer patients born in 1944, only those who died 
at age 9 are represented. A total of 120 such strata were 
available.

In order to estimate the overall relative risk of 
obstetric radiation, and to determine whether, and if so how, 
it varied with age and year, we ...


* Appendix II puts the column header "Year of DEATH" for the
  years 1944 to 1964. From the text, it is clear that this
  wording is in error and that the data do in fact refer to 
  year of BIRTH.