Description
The burn data set from KMsurv is from Ichida et al. (1993) which is linked on
the web site. It represents a study of an attempt to improve infection control by
replacing routine bathing with total body washing with an antimicrobial agent. The
main outcome of interest is infection with straphylocous aureaus, the time and status
of which are represented by T3 and D3.
There are 11 ordinary covariates Z1–Z11
and two time dependent predictors that can be constructed for surgical excision of
burn tissue (T1, D1) and prophylactic antibiotic treatment (T2, D2). The input
R file on the web site recodes and renames the ordinary covariates, as shown on
the next page, to make interpretation easier. The cases with routine bathing are
historical controls so this does not represent a single cohort study and in particular
the assignment of treatment to patient is not random.
Your assignment is to analyze the data and write a report as if to the head burn
surgeon in the hospital. You will be graded on the analysis and also on the quality
of the communication in the presentation.
Below are some possible things to look
at. There may be others as well, and maybe you don’t need all of these. If you
make a plot or do an analysis, then explain why you did it and what the meaning
of the results might be.
1. Plot the Kaplan-Meier curves for the treated and untreated patients and use
survdiff to test for whether the curves are different.
2. Plot the cumulative hazards vs. time and the complimentary log-log survival
vs. log time.
3. Construct Cox model using only the time-independent predictors, Maybe start
with one using only Treatment. Decide if the burn site variables will be separately included after analysis or included or excluded as a group. Note that
this is not a factor, because a patient may have burns at many sites. The
respiratory tract burn site variable is different from the others since it does
not focus on skin.
4. Run the usual suite of model checking methods and report any interesting
findings. Possibly alter the model as a result.
1
5. Construct the data set with the the time-dependent covariates for surgical
excision and prophylactic antibiotic treatment and find a good model which
includes useful time-dependent covariates as well as useful time-independent
ones.
6. Run the usual suite of model checking methods and report any interesting
findings. Possibly alter the model as a result.
7. Interpret the results and comment on the implications for clinical management.
Note that, from other studies, for the endpoint of survival (an outcome not
included in this data set), burn percentage and burn degree (first, second,
third) are quite important, but the first one may or may not be important for
infection control, and information on the second factor is not included in the
data set.
2
Burn Data Set
Var. Definition
Obs Observation number
Z1 Treatment: 0-routine bathing 1-Body cleansing
Z2 Gender (0=male 1=female)
Z3 Race: 0=nonwhite 1=white
Z4 Percentage of total surface area burned
Z5 Burn site indicator: head 1=yes, 0=no
Z6 Burn site indicator: buttock 1=yes, 0=no
Z7 Burn site indicator: trunk 1=yes, 0=no
Z8 Burn site indicator: upper leg 1=yes, 0=no
Z9 Burn site indicator: lower leg 1=yes, 0=no
Z10 Burn site indicator: respiratory tract 1=yes, 0=no
Z11 Type of burn: 1=chemical, 2=scald, 3=electric, 4=flame
T1 Time to excision or on study time
D1 Excision indicator: 1=yes 0=no
T2 Time to prophylactic antibiotic treatment or on study time
D2 Prophylactic antibiotic treatment: 1=yes 0=no
T3 Time to straphylocous aureaus infection or on study time
D3 Straphylocous aureaus infection: 1=yes 0=no
Recoded Burn Data Set
Var. New Var. Definition and Factor Levels
Obs Observation number
Z1 Treatment Routine/Cleansing
Z2 Gender Male/Female
Z3 Race Nonwhite/White
Z4 PercentBurned Percentage of total surface area burned
Z5 SiteHead NotBurned/Burned
Z6 SiteButtock NotBurned/Burned
Z7 SiteTrunk NotBurned/Burned
Z8 SiteUpperLeg NotBurned/Burned
Z9 SiteLowerLeg NotBurned/Burned
Z10 SiteRespTract NotBurned/Burned
Z11 BurnType Chemical/Scald/Electric/Flame
T1 Time to excision or on study time
D1 Excision indicator: 1=yes 0=no
T2 Time to prophylactic antibiotic treatment or on study time
D2 Prophylactic antibiotic treatment: 1=yes 0=no
T3 Time to straphylocous aureaus infection or on study time
D3 Straphylocous aureaus infection: 1=yes 0=no
3
