lexpand, estimates various
survival time functions as requested by the user.survtab(data, surv.breaks = NULL, by.vars = NULL, event.values = NULL,
surv.type = "surv.rel", surv.method = "hazard", relsurv.method = "e2",
subset = NULL, agegr.w.breaks = NULL, agegr.w.weights = NULL,
conf.level = 0.95, conf.type = "log-log", format = TRUE,
verbose = FALSE)lexpand[a,b).by.vars = c('sex', 'area') computes
survivals separately for each combination
of 'sex' and 'area'.lex.Xst column;
if NULL, uses all but the first
level of as.factor(lex.Xst) as event.values;
these values are events and all other values are considered t'surv.obs',
'surv.cause', 'surv.rel',
'cif.obs' or 'cif.rel';
defines what kind of survival time function(s) is/are estimated; see Details'lifetable' or 'hazard'; determines
the method of calculating survival time functions'e2' or 'pp';
defines whether to compute relative survival using the
EdererII method or using Pohar-Perme weighting;
ignored if surv.type != "surv.rel"subset = sex == 1;
subsets the data before computationsNULL, given breaks will be
used to define age groups in age group
weighting; given as left inclusive and right exclusive, e.g. [a,b)agegr.w.breaks is not NULL,
the user can define a vector of weights
to give to each age group defined by agegr.w.breaks;
if agegr.w.weights is NULL,
internal weights (see E0.95 for 95 percent confidence intervals"plain",
"log-log" and "log";
defines the transformation used on the survival (and/or relative survival)
function to yield confidence
intervals via the delta methodTRUE, output is formatted into a neat table;
otherwise you get all the raw resultsTRUE, the function is chatty and
returns some messages and timings along the processsurv.obs- observed (raw) survivalCIF_k- cumulative incidence function for causekCIF.rel- cumulative incidence function using excess casesr.e2- relative survival, EdererIIr.pp- relative survival, Pohar-Perme weighted.as implies agegroup-standardisation, and .lo and
.hi imply lower and upper confidence limits, respectively.
The prefix SE. stands for standard error.survtab creates survival tables using data split with e.g.
lexpand. We recommend using lexpand since
it is well tested and one usually needs to merge in population hazards
to ocmpute relative survivals.
By default
survtab makes use of the exact same breaks that were used in
splitting (with e.g. lexpand), so it is not necessary to specify any
surv.breaks. If specified, the
surv.breaks must be a subset of the pertinent
breaks given in lexpand.
The function supplies surv.breaks to cut to
create survival intervals in the data,
e.g.
surv.breaks=0:5 -> [0,1),[1,2), ..., [4,5).
Interval lengths (deltas) are also calculated based
on surv.breaks. The upper limit of the breaks should
be meaningful and never e.g. Inf.
if surv.type = 'surv.obs', only 'raw' observed survival
is calculated over the chosen time intervals. With
surv.type = 'surv.rel', also relative survival estimates
are supplied in addition to observed survival figures.
surv.type = 'cif.obs' requests cumulative incidence functions (CIF)
to be estimated, where all unique event.values
are seen as competing risks indicators
(others are random censoring indicators);
CIFs are estimated for each competing risk are computed based
on a survival-interval-specific proportional hazards
assumption as described by Chiang (1968) using the chosen surv.method.
With surv.type = 'cif.rel', a CIF is estimated with using
excess cases as the ''cause-specific'' cases.
if surv.type = 'surv.cause', cause-specific survivals are estimated
separately for each unique value of event.values.
Relative / net survival
When surv.type = 'surv.rel', the user can choose
relsurv.method = 'pp', whereupon
additional Pohar-Perme weighting is used
to get closer to a true net survival measure.
By default relsurv.method = 'e2'.
Age-standardised survival
The user can also apply age standardisation on top of
everything else. Then the requested survival figures
are calculated for each age group separately, and then a
weighted average of the age-group-specific survivals
is presented.
The user must define the age-standardisation age
groups and their weights with the agegr.w.breaks
and agegr.w.weights arguments. The numbers of
age groups and weights should match; e.g.
with agegr.w.breaks = c(0,45,65,75,Inf) the weights vector
must then have 4 elements.
The agegr.w.weights do not have to sum to one as
they are processed internally to do so.
If one wishes to use one of the three integrated international
standard weighting schemes available,
one must specify the weighting scheme by using e.g.
agegr.w.weights = "ICSS1", and also
by specifying the used agegr.w.breaks. However, as
the weights are available only
for 5-year age groups, the agegr.w.breaks must all
(except the last) be divisible by 5;
e.g.
agegr.w.breaks = c(0, 45, 65, 85, Inf).
You can see the weights integrated
into ICSS into the console.
See also ICSS.
Note that the by.vars should not be confused with age-standardisation.
by.vars simply determine variables, for the unique combinations of which
survivals are computed and outputted separately.
Period analysis / delayed entry
If one wishes to calculate period analysis / delayed entry estimates,
one should limit the data accordingly
when expanding the data; see lexpand.
Data requirements
This function requires the data to contain,
at minimum, the variables
lex.id, lex.dur,lex.Cst,
lex.Xst, and fot; these will be enough to calculate
observed survivals.
Relative survivals require additional information. EdererII
relative survival requires the presence of a pop.haz variable in the data,
and Pohar-Perme weighting requires pp (the inverse cumulative population
survival). Both can be computed with lexpand.
You may take a look at a simulated cohort sire as an example of the
minimum required information for when processing data to be used in calculating
relative survival (in the Finnish context).splitMulti, lexpand,
ICSS, sire## see more examples with explanations in vignette("survtab_examples")
## prepare data for e.g. 5-year "period analysis" for 2008-2012
## note: sire is a simulated cohort integrated into popEpi.
BL <- list(fot=seq(0, 5, by = 1/12),
per = c("2008-01-01", "2013-01-01"))
x <- lexpand(sire, birth = bi_date, entry = dg_date, exit = ex_date,
status = status %in% 1:2,
breaks = BL,
pophaz = popmort)
## calculate relative EdererII period method
## survivals using the fot.breaks given in lexpand()
st <- survtab(x)Run the code above in your browser using DataLab