periodontal {DOS2}

Smoking and Periodontal Disease

Description

Data from NHANES 2011-2012 containing 441 matched pairs of a daily cigarette smoker and a never smoker, recording the extent of periodontal disease. See Rosenbaum (2017) and Chapter 20 of "Design of Observational Studies", second edition.

Usage

data("periodontal")

Format

A data frame with 882 observations on the following 12 variables.

SEQN

NHANES 2011-2012 sequence number

female

=1 for female, 0 for male

age

Age in years

black

=1 for black, 0 for other

educf

Education, in five categories. An ordered factor with levels <9 for less than 9th grade, 9 to 11 for 9th to 11th grade, HS/GED for high school or GED degree, SomeCol for some college, College for college degree.

income

Ratio of family income to the poverty level, capped at 5 times the poverty level.

cigsperday

Cigarettes smoked per day for daily smokers, 0 for never smokers

either

Number of periodonal measurements indicative of periodontal disease.

neither

Number of periodonal measurements

pcteither

Percent indicative of periodontal disease, =100*either/neither.

z

Treatment indicator, 1=daily smoker, 0=never smoker

mset

Matched set indicator, 1 to 441.

Details

Excluding wisdom teeth, 6 measurements are taken for each tooth that is present, up to 28 teeth. Following Tomar and Asma (2000), a measurement indicates periodontal disease if either there is a loss of attachment of at least 4mm or a pocket depth of at least 4mm. The first individual has 11 measurements indicative of periodontal disease, out of 106 measurements, so pcteither is 100*11/106 = 10.38 percent. A related data set in DOS2 with bivariate outcome is teeth.

Source

Data are from the National Health and Nutrition Examination Survey 2011-2012 and were used as an example in Rosenbaum (2017). In the second edition of Design of Observational Studies, these data are discussed in Chapter 20, Evidence Factors.

References

Rosenbaum, P. R. (2015) <https://obsstudies.org/two-r-packages-for-sensitivity-analysis-in-observational-studies/> "Two R packages for sensitivity analysis in observational studies". Observational Studies, 1(1), 1-17.

Rosenbaum, P. R. (2017) <doi:10.1214/17-STS621> "The general structure of evidence factors in observational studies". Statistical Science 32, 514-530.

Tomar, S. L. and Asma, S. (2000) <doi:10.1902/jop.2000.71.5.743> "Smoking attributable periodontitis in the US: Findings from NHANES III". J Periodont 71, 743-751.

"US National Health and Nutrition Examination Survey 2011-2012". www.cdc.gov/nchs/nhanes/index.htm

Examples

# Figure 1 in Rosenbaum (2017)
data(periodontal)
attach(periodontal)
oldpar<-par()
m<-matrix(1:2,1,2)
layout(m,widths=c(1,2))
boxplot(pcteither[z==1]-pcteither[z==0],ylab="Smoker-Control Difference",
        main="(i)",xlab="Matched Pairs",ylim=c(-100,100))
abline(h=0,lty=2)
crosscutplot(cigsperday[z==1],pcteither[z==1]-pcteither[z==0],ylab="Smoker-Control Difference",
    xlab="Cigarettes per Day",main="(ii)",ylim=c(-100,100))
abline(h=0,lty=2)

# Sensitivity analysis in Section 2.3 of Rosenbaum (2017)
y<-pcteither[z==1]-pcteither[z==0]
x<-cigsperday[z==1]
senWilcox(y,gamma=2.76)
# The following is the same as sensitivitymw::senmw(y,gamma=2.77,method="p")
sensitivitymult::senm(pcteither,z,mset,gamma=2.77,inner=.5,trim=2)
# The following is the same as sensitivitymw::senmw(y,gamma=3.5,method="p")
sensitivitymult::senm(pcteither,z,mset,gamma=3.5,inner=.5,trim=2)
# Second evidence factor
crosscut(x,y)
crosscut(x,y,gamma=1.6)

# Note, however, that other statistics report greater insensitivity to
# bias by virtue of having larger design sensitivity:
sensitivitymult::senm(pcteither,z,mset,gamma=3.5,inner=1,trim=4)
sensitivitymult::senm(pcteither,z,mset,gamma=4.2,inner=1,trim=4)
senU(y,m1=4,m2=5,m=5,gamma=2.77)
senU(y,m1=6,m2=8,m=8,gamma=2.77)
senU(y,m1=6,m2=8,m=8,gamma=3.5)
detach(periodontal)
par(oldpar)