Calculate ratio of two SIRs by providing observed and expected counts to sir_ratio The related functions sir_ratio_lci and sir_ratio_uci can also calculate lower and upper estimates of the confidence interval Calculations are based on formulas suggested by Breslow & Day 1987

sir_ratio(o1, o2, e1, e2)

sir_ratio_lci(o1, o2, e1, e2, alpha = 0.05)

sir_ratio_uci(o1, o2, e1, e2, alpha = 0.05)

Arguments

o1

observed count for SIR 1

o2

observed count for SIR 2

e1

expected count for SIR 1

e2

observed count for SIR 2

alpha

alpha significance level for confidence interval calculations. Default is alpha = 0.05 which will give 95 percent confidence intervals.

Value

num numeric value of SIR / SMR estimate

References

Breslow NE, Day NE. Statistical Methods in Cancer Research Volume II: The Design and Analysis of Cohort Studies. Lyon, France: IARC; 1987. (IARC Scientific Publications IARC Scientific Publications No. 82). Available from: http://publications.iarc.fr/Book-And-Report-Series/Iarc-Scientific-Publications/Statistical-Methods-In-Cancer-Research-Volume-II-The-Design-And-Analysis-Of-Cohort-Studies-1986

Examples

#provide the two expected and observed count to get the ratio of SIRs/SMRs
msSPChelpR::sir_ratio(o1 = 2140, o2 = 3158, e1 = 1993, e2 = 2123)
#> [1] 1.385338

#calculate lower confidence limit
msSPChelpR::sir_ratio_lci(o1 = 2140, o2 = 3158, e1 = 1993, e2 = 2123, alpha = 0.05)
#> [1] 1.310944

#calculate upper confidence limit
msSPChelpR::sir_ratio_uci(o1 = 2140, o2 = 3158, e1 = 1993, e2 = 2123, alpha = 0.05)
#> [1] 1.464168

#functions can be easily used inside dplyr::mutate function
library(dplyr)
#> 
#> Attaching package: ‘dplyr’
#> The following objects are masked from ‘package:stats’:
#> 
#>     filter, lag
#> The following objects are masked from ‘package:base’:
#> 
#>     intersect, setdiff, setequal, union
test_df <- data.frame(sir_oth = c(1.07, 1.36, 0.96), 
                  sir_smo = c(1.49, 1.81, 1.41),
                  observed_oth = c(2140, 748, 1392),
                  expected_oth = c(1993, 550, 1443),
                  observed_smo = c(3158, 744, 2414),
                  expected_smo = c(2123, 412, 1711))

test_df %>%
  mutate(smo_ratio = sir_ratio(observed_oth, observed_smo, expected_oth, expected_smo),
         smo_ratio_lci = sir_ratio_lci(observed_oth, observed_smo, expected_oth, expected_smo),
         smo_ratio_uci = sir_ratio_uci(observed_oth, observed_smo, expected_oth, expected_smo))
#>   sir_oth sir_smo observed_oth expected_oth observed_smo expected_smo smo_ratio
#> 1    1.07    1.49         2140         1993         3158         2123  1.385338
#> 2    1.36    1.81          748          550          744          412  1.327813
#> 3    0.96    1.41         1392         1443         2414         1711  1.462562
#>   smo_ratio_lci smo_ratio_uci
#> 1      1.310944      1.464168
#> 2      1.198055      1.471613
#> 3      1.368629      1.563409