dca: Decision Curve Analysis

Description Usage Arguments Value Author(s) Examples

Description

Diagnostic and prognostic models are typically evaluated with measures of accuracy that do not address clinical consequences. Decision-analytic techniques allow assessment of clinical outcomes but often require collection of additional information may be cumbersome to apply to models that yield a continuous result. Decision curve analysis is a method for evaluating and comparing prediction models that incorporates clinical consequences, requires only the data set on which the models are tested, and can be applied to models that have either continuous or dichotomous results. The dca function performs decision curve analysis for binary outcomes. See http://www.decisioncurveanalysis.org for more information.

Usage

1
2
3
4
dca(data, outcome, predictors, xstart = 0.01, xstop = 0.99, xby = 0.01,
  ymin = -0.05, probability = NULL, harm = NULL, graph = TRUE,
  intervention = FALSE, interventionper = 100, smooth = FALSE,
  loess.span = 0.1)

Arguments

data

a data frame containing the outcome of the outcome predictions.

outcome

the outcome, response variable. Must be a variable contained within the data frame specified in data=.

predictors

the predictor variable(s). Must be a variable(s) contained within the data frame specified in data=.

xstart

starting value for x-axis (threshold probability) between 0 and 1. The default is 0.01.

xstop

stopping value for x-axis (threshold probability) between 0 and 1. The default is 0.99.

xby

increment for threshold probability. The default is 0.01.

ymin

minimum bound for graph. The default is -0.05.

probability

specifies whether or not each of the independent variables are probabilities. The default is TRUE.

harm

specifies the harm(s) associated with the independent variable(s). The default is none.

graph

specifies whether or not to display graph of net benefits. The default is TRUE.

intervention

plot net reduction in interventions.

interventionper

number of net reduction in interventions per interger. The default is 100.

smooth

specifies whether or not to smooth net benefit curve. The default is FALSE.

loess.span

specifies the degree of smoothing. The default is 0.10.

Value

Returns a list containing the calculated net benefit, ADD MORE

Author(s)

Daniel D Sjoberg sjobergd@mskcc.org

Examples

1
2
3
4
5
6
7
library(MASS)
data.set <- birthwt
model = glm(low ~ age + lwt, family=binomial(link="logit"), data=data.set)
data.set$predlow = predict(model, type="response")
dca(data=data.set, outcome="low", predictors=c("age", "lwt"), probability=c("FALSE", "FALSE"))
result1 = dca(data=data.set, outcome="low", predictors="age", smooth="TRUE", xstop=0.50, probability="FALSE", intervention="TRUE")
result2 = dca(data=data.set, outcome="low", predictors="predlow", smooth="TRUE", xstop=0.50)

ddsjoberg/dca documentation built on May 17, 2019, 7:03 p.m.