Skip to main content
. Author manuscript; available in PMC: 2019 Oct 1.
Published in final edited form as: Ann Epidemiol. 2018 Jul 31;28(10):730–735. doi: 10.1016/j.annepidem.2018.07.015

Table 4.

Results of multivariate analysis between cell phone legislation and fatal collision by sub-groupa

Characteristic Sub-group Model 1 IRRb
(95% CI)
Model 2 IRRb
(95% CI)
Overall
Hand-held ban 0.88 (0.81, 0.96) 0.90 (0.84, 0.97)
Texting ban 1.03 (0.96, 1.10) 1.03 (0.91, 1.16)
Age
16–24 year olds
Hand-held ban 0.84 (0.72, 0.97) 0.86 (0.76, 0.97)
Texting ban 0.99 (0.90, 1.09) 1.00 (0.71, 1.40)
25–39 year olds
Hand-held ban 0.84 (0.79, 0.90) 0.86 (0.80, 0.93)
Texting ban 1.05 (0.93, 1.18) 1.05 (0.77, 1.44)
40–59 year olds
Hand-held ban 0.86 (0.80, 0.93) 0.89 (0.84, 0.95)
Texting ban 1.01 (0.93, 1.10) 1.01 (0.74, 1.38)
>60 year olds
Hand-held ban 0.91 (0.84, 1.00) 0.92 (0.86, 0.99)
Texting ban 1.02 (0.97, 1.08) 1.02 (0.97, 1.07)
Sex
Males
Hand-held ban 0.89 (0.83, 0.96) 0.91 (0.85, 0.97)
Texting ban 1.03 (0.95, 1.12) 1.03 (0.86, 1.24)
Females
Hand-held ban 0.86 (0.78, 0.96) 0.88 (0.81, 0.97)
Texting ban 1.02 (0.95, 1.08) 1.02 (0.83, 1.25)
Race/ethnicity
White non-Hispanic
Hand held ban 0.90 (0.76, 1.05) 0.93 (0.79, 1.09)
Texting ban 1.12 (0.97, 1.30) 1.11 (0.96, 1.28)
White Hispanic
Hand held ban 0.87 (0.72, 1.06) 0.83 (0.67, 1.04)
Texting ban 0.88 (0.63, 1.24) 0.90 (0.52, 1.55)
Black non-Hispanic
Hand held ban 0.86 (0.72, 1.02) 0.89 (0.75, 1.05)
Texting ban 0.97 (0.80, 1.16) 0.94 (0.74, 1.18)
Other
Hand held ban 0.88 (0.47, 1.68) 0.93 (0.75, 1.15)
Texting ban 0.85 (0.02, 36.85) 0.85 (0.06, 12.35)
Location
Urban
Hand held ban 0.85 (0.67, 1.07) 0.88 (0.77, 1.00)
Texting ban 1.01 (0.88, 1.15) 0.99 (0.92, 1.07)
Rural
Hand held ban 0.89 (0.82, 0.95) 0.90 (0.45, 1.83)
Texting ban 1.03 (0.96, 1.11) 1.04 (0.96, 1.13)
a:

Abbreviations: CI=confidence interval; IRR=incident rate ratio

b:

Model 1 contained variables for presence of universal texting ban (binary), presence of universal hand-held ban (binary), quarter (continuous), spline knots and year (continuous). Model 2 included variables for presence of universal texting ban (binary), presence of universal hand-held ban (binary), quarter (continuous), spline knots, year (continuous), log of adjusted per capita income, log of unemployment rate, log of adjusted gas price, maximum state speed limits ≥70 miles per hour (binary), presence of administrative license suspension (binary), presence of per se minimum blood alcohol concentration ≥ 0.08 g/dl (binary), presence of primary seat belt laws for front seat passengers (binary), log of cell phone coverage, presence of graduated drivers licensing laws (binary), and log of police population. Model 1 and Model 2 were stratified analyses. Because the focus of this analysis was the relationship between legislation and fatal collision rate, only the IRR pertaining to the bans were shown for ease of presentation.