remss

Wei Xu, Susie Su, Yi Zhu

2021-01-13

1. Introduction

remss from github is a package for TNM stage is important in treatment decision-making and outcome predicting. The existing oropharyngeal Cancer (OPC) TNM stages have not made distinction of the two sub sites of HPV+ and HPV- diseases.We developed novel criteria to assess performance of the TNM stage grouping schemes based on parametric modeling adjusting on important clinical factors. These criteria evaluate the TNM stage grouping scheme in five different measures: hazard consistency, hazard discrimination, explained variation, likelihood difference, and balance.

Reference:

“Xu, W., et al. ’Refining evaluation methodology on TNM stage system: assessment on HPV-related oropharyngeal cancer.’Austin Biometrics and Biostatistics 2 (2015): 1014.” (via)

2. Installation

You can install remss from [github] (https://github.com/qiuanzhu/remss):

library("devtools")
install_github("qiuanzhu/remss")

3. Examples

In the following examples, a simulated dataset with 10 variables and 504 observations.

library("remss")
head(Rdata)
#> 
#> Attaching package: 'remss'
#> The following object is masked from 'package:base':
#> 
#>     rank
Id Gender Age OS survmonth Treatment Basic_group Scheme.1 Scheme.2 Scheme.3
1 0 64.0 1 101.55616 Treatment 1 G13 II II III
2 0 60.2 1 37.64384 Treatment 2 G2 I II II
3 1 64.7 0 73.15068 Treatment 1 G17 II III IV
4 1 64.6 0 127.56164 Treatment 1 G12 II II III
5 0 53.2 0 49.24932 Treatment 1 G2 I II II
6 1 51.5 0 83.80274 Treatment 1 G17 II III IV

There are three grouping schemes (Scheme.1,Scheme.2,Scheme.3) base on the classification (Basic_group). Five measures of grouping scheme and weight for each measurement are introduced.

data(Rdata)
Scheme=c('Scheme.1','Scheme.2','Scheme.3')
Covar=c('Age','Treatment')
weight=c(1,1,0.5,0.5,1)
Order=list(c('I','II','III'),c('I','II','III','IV'),c('I','II','III','IV'))
table<-rank(os='OS',ostime='survmonth',groupvar='Basic_group', scheme=Scheme, order=Order, covariate=Covar,weight=weight,data=Rdata)
#> Warning in fitter(X, Y, strats, offset, init, control, weights = weights, :
#> Loglik converged before variable 16,19 ; coefficient may be infinite.

Variable Information:

Variable Name Numeric Mean Standard Deviation Level Set
OS 1 0.2639 0.4412 Null
survmonth 1 72.1517 39.657 Null
Basic_group 0 NA NA c(“G1”, “G10”, “…”)
Scheme.1 0 NA NA c(“I”, “II”, “…”)
Scheme.2 0 NA NA c(“I”, “II”, “…”)
Scheme.3 0 NA NA c(“I”, “II”, “…”)
Age 1 51.6038 12.8468 Null
Treatment 0 NA NA c(“Treatment 1”, “Treatment 2”)

Observation number:

Total obs Complete obs Imcomplete obs
504 504 0

Hazard Consistency Measurement:

Scheme Score Standardized Score Rank
2 Scheme.2 1.659162 0.0000000 1
1 Scheme.1 2.042695 0.8535856 2
3 Scheme.3 2.108482 1.0000000 3

Hazard Discrimination Measurement:

Scheme Score Standardized Score Rank
Scheme.1 0.0969022 0.0000000 1
Scheme.2 0.3479726 0.0928677 2
Scheme.3 2.8004306 1.0000000 3

Likelihood Difference Measurement:

Scheme Score Standardized Score Rank
Scheme.1 20.93218 0.0000000 1
Scheme.2 16.68120 0.6394996 2
Scheme.3 14.28482 1.0000000 3

Explained Variance Measurement:

Scheme Score Std_Score Rank
2 Scheme.2 21.44556 0.0000000 1
1 Scheme.1 19.90845 0.6062326 2
3 Scheme.3 18.91005 1.0000000 3

Balance Measurement:

Scheme Score Standardized Score Rank
1 Scheme.1 0.4126984 0.0000000 1
3 Scheme.3 0.5039683 0.6969697 2
2 Scheme.2 0.5436508 1.0000000 3

Overall Rank:

Scheme Overall Score Rank
Scheme.1 1.156702 1
Scheme.2 1.412618 2
Scheme.3 3.696970 3