Conditional change score for serial PFT measurements

pft_change() computes the conditional change score (CCS) defined in Box 2 of the Stanojevic et al. ERS/ATS 2022 interpretation standard. The CCS evaluates whether the change between two FEV1 z-scores is larger than would be expected from within-subject variability and regression to the mean alone.

Formula (paper Box 2 p. 12): $$CCS = (z_2 - r \cdot z_1) / \sqrt{1 - r^2}$$

Where the autocorrelation r is itself a function of the time interval between measurements and the patient's age at the first time point: $$r = 0.642 - 0.04 \cdot time(years) + 0.020 \cdot age(years)$$

Changes within +/- 1.96 change scores are considered within the normal limits per the paper.

This formula was derived from a children/young-people cohort (Stanojevic 2022 references the underlying study and notes the approach has "yet to be validated, extended to adults" but permits its use as "a reasonable tool to facilitate interpretation"). For adults the 2022 standard alternatively recommends FEV1Q (Box 3); see pft_fev1q().

Usage

pft_change(z1, z2, age_t1 = NULL, time_years = NULL, r = NULL)

Arguments

z1, z2

Numeric vectors of FEV1 z-scores at time 1 and time 2.

age_t1

Numeric. Patient age (in years) at the first measurement.

time_years

Numeric. Elapsed time between measurements in years (e.g. 0.25 for 3 months, 4 for 4 years).

r

Optional. Numeric in (-1, 1). If supplied, used directly in place of the paper's age/time formula – useful for callers who have a population-specific autocorrelation estimate. If NULL (the default), r is computed from age_t1 and time_years via the Box 2 formula.

Value

A data frame with columns:

• ccs: the conditional change score.

• r_used: the autocorrelation actually used in the calculation (returned so callers can audit the value chosen).

• is_significant: logical, TRUE when |ccs| > 1.96 (i.e. outside the paper's normal-limits range).

References

Stanojevic S, Kaminsky DA, Miller MR, et al. ERS/ATS technical standard on interpretive strategies for routine lung function tests. Eur Respir J. 2022;60(1):2101499. doi:10.1183/13993003.01499-2021 . Box 2 (p. 12).

See also

pft_spirometry() to produce the FEV1 z-scores at each time point.

Examples

# Stanojevic 2022 Box 2 worked example: a 14-year-old male whose
# FEV1 z-score dropped from -0.78 to -1.60 over 3 months.
pft_change(z1 = -0.78, z2 = -1.60, age_t1 = 14, time_years = 0.25)
#> # A tibble: 1 × 3
#>     ccs r_used is_significant
#>   <dbl>  <dbl> <lgl>         
#> 1 -2.17  0.912 TRUE          
# -> r_used = 0.912, ccs ~= -2.17, is_significant = TRUE

# Same drop spread over 4 years
pft_change(z1 = -0.78, z2 = -1.60, age_t1 = 14, time_years = 4)
#> # A tibble: 1 × 3
#>     ccs r_used is_significant
#>   <dbl>  <dbl> <lgl>         
#> 1 -1.55  0.762 FALSE         
# -> r_used = 0.762, ccs ~= -1.55, is_significant = FALSE