Check Participant Number Consistency

Verifies that reported participant numbers are internally consistent: randomised = analysed + lost to follow-up.

Usage

check_n_consistency(
  n_randomised_int,
  n_randomised_ctrl,
  n_analysed_int = NULL,
  n_analysed_ctrl = NULL,
  n_lost_int = NULL,
  n_lost_ctrl = NULL,
  n_randomised_total = NULL
)

Arguments

n_randomised_int, n_randomised_ctrl

Integer. Per-arm randomised counts.

n_analysed_int, n_analysed_ctrl

Integer or NULL. Per-arm analysed.

n_lost_int, n_lost_ctrl

Integer or NULL. Per-arm lost to follow-up.

n_randomised_total

Integer or NULL. Total randomised.

Value

A list with components:

consistent

Logical. TRUE if all checks pass.

checks

Data frame of individual checks.

n_checks

Number of checks performed.

n_failed

Number of failed checks.

Details

Implements INSPECT-SR check 4.6. Missing values (NULL) are skipped.

Examples

check_n_consistency(
  n_randomised_int = 100, n_randomised_ctrl = 100,
  n_analysed_int = 95, n_analysed_ctrl = 92,
  n_lost_int = 5, n_lost_ctrl = 8,
  n_randomised_total = 200
)
#> $consistent
#> [1] TRUE
#> 
#> $checks
#>                                        check expected observed pass
#> 1  Total randomised = Intervention + Control      200      200 TRUE
#> 2 Intervention: Randomised = Analysed + Lost      100      100 TRUE
#> 3      Control: Randomised = Analysed + Lost      100      100 TRUE
#> 4           Intervention: Lost <= Randomised      100        5 TRUE
#> 5                Control: Lost <= Randomised      100        8 TRUE
#> 
#> $n_checks
#> [1] 5
#> 
#> $n_failed
#> [1] 0
#>