Calculate MCID using Distribution-Based Methods — calculate_mcid_distribution • pharmhand

Calculates MCID estimates using distribution-based approaches.

Usage

calculate_mcid_distribution(
  data,
  score_var,
  reliability = NULL,
  methods = c("half_sd", "one_sem", "third_sd", "fifth_sd")
)

Arguments

data: Data frame with baseline PRO scores
score_var: Character. Name of the baseline score variable
reliability: Numeric. Test-retest reliability for SEM method (default: NULL, SEM not calculated)
methods: Character vector. Methods to use: "half_sd", "one_sem", "third_sd", "fifth_sd" (default: all)

Value

A list with MCID estimates:

half_sd: 0.5 * SD of baseline scores
third_sd: 0.33 * SD of baseline scores
fifth_sd: 0.2 * SD of baseline scores
one_sem: 1 * SEM (requires reliability)
sd: Standard deviation of baseline scores
n: Number of observations
reliability: Test-retest reliability if provided

Examples

if (FALSE) { # \dontrun{
# Distribution-based MCID for a PRO measure
result <- calculate_mcid_distribution(
  data = pro_data,
  score_var = "AVAL",
  reliability = 0.85
)

# Access 0.5 SD MCID
result$half_sd
} # }