Skip to main content

Table 3 Risk ratios of emergency department visits for pediatric asthma for days in the terminal nodes as compared to the referent group, a Atlanta, Georgia, 1999–2009

From: Classification and regression trees for epidemiologic research: an air pollution example

Terminal nodeb

Nc

Risk ratio

95% Confidence interval

Type of days (pollutant quartiles)

    

CO

NO2

O3

PM2.5

Referent

131

1.00

 

1

1

1

1

T1

316

1.10

(1.05, 1.16)

1-4

1,2

1-4

4

T2

279

1.08

(1.03, 1.14)

1-4

3,4

1-4

1

T3

1039

1.05

(1.01, 1.1)

1-4

3,4

1-4

2,3

T4

441

1.07

(1.02, 1.13)

1-4

3,4

4

4

T5

91

1.03

(0.97, 1.1)

1,2

1,2

4

1-3

T6

68

1.07

(0.98, 1.17)

3,4

1,2

4

1-3

T7

76

1.08

(1.01, 1.15)

1-4

3

1-3

4

T8

168

1.07

(1.01, 1.14)

1-4

4

1-3

4

T9

458

1.01

(0.97, 1.05)

1-4

1

1,2

1-3

T10

263

1.04

(0.99, 1.09)

1-4

2

1

1-3

T11

435

1.03

(0.98, 1.07)

1-4

2

2,3

1-3

T12

160

1.02

(0.97, 1.08)

1-4

1

3

1,2

T13

85

1.04

(0.97, 1.11)

1-4

1

3

3

  1. aDays when all pollutants are in the lowest quartile.
  2. bTerminal nodes represent different types of days that can be described in terms of the pollutant quartiles.
  3. cEach day is in one and only one terminal node; the column sums to 4010.
  4. d P-values are associated with the null hypothesis that the risk ratio for the pollutant indicator is 1.0.
Back to article page

Environmental Health

ISSN: 1476-069X