cs_combined {clinicalsignificance}

Combined Analysis of Clinical Significance

Description

cs_combined() can be used to determine the clinical significance of intervention studies employing the combination of the distribution-based and statistical approach. For this, it will be assumed that the functional (non-clinical population) and patient (clinical population) scores form two distinct distributions on a continuum. cs_combined() calculates a cutoff point between these two populations as well as a reliable change index (RCI) based on a provided instrument reliability estimate and counts, how many of those patients that showed a reliable change (that is likely to be not due to measurement error) switched from the clinical to the functional population during intervention. Several methods for calculating the cutoff and RCI are available.

Usage

cs_combined(
  data,
  id,
  time,
  outcome,
  group = NULL,
  pre = NULL,
  post = NULL,
  mid_improvement = NULL,
  mid_deterioration = NULL,
  reliability = NULL,
  reliability_post = NULL,
  m_functional = NULL,
  sd_functional = NULL,
  better_is = c("lower", "higher"),
  rci_method = c("JT", "GLN", "HLL", "EN", "NK", "HA", "HLM"),
  cutoff_type = c("a", "b", "c"),
  significance_level = 0.05
)

Arguments

data

A tidy data frame

id

Participant ID

time

Time variable

outcome

Outcome variable

group

Grouping variable (optional)

pre

Pre measurement (only needed if the time variable contains more than two measurements)

post

Post measurement (only needed if the time variable contains more than two measurements)

mid_improvement

Numeric, change that indicates a clinically significant improvement

mid_deterioration

Numeric, change that indicates a clinically significant deterioration (optional). If mid_deterioration is not provided, it will be assumed to be equal to mid_improvement

reliability

The instrument's reliability estimate. If you selected the NK method, the here specified reliability will be the instrument's pre measurement reliability. Not needed for the HLM method.

reliability_post

The instrument's reliability at post measurement (only needed for the NK method)

m_functional

Numeric, mean of functional population.

sd_functional

Numeric, standard deviation of functional population

better_is

Which direction means a better outcome for the used instrument? Available are

• "lower" (lower outcome scores are desirable, the default) and

• "higher" (higher outcome scores are desirable)

rci_method

Clinical significance method. Available are

• "JT" (Jacobson & Truax, 1991, the default)

• "GLN" (Gulliksen, Lord, and Novick; Hsu, 1989, Hsu, 1995)

• "HLL" (Hsu, Linn & Nord; Hsu, 1989)

• "EN" (Edwards & Nunnally; Speer, 1992)

• "NK" (Nunnally & Kotsch, 1983), requires a reliability estimate at post measurement. If this is not supplied, reliability and reliability_post are assumed to be equal

• "HA" (Hageman & Arrindell, 1999)

• "HLM" (Hierarchical Linear Modeling; Raudenbush & Bryk, 2002), requires at least three measurements per patient

cutoff_type

Cutoff type. Available are "a", "b", and "c". Defaults to "a" but "c" is usually recommended. For "b" and "c", summary data from a functional population must be given with arguments m_functional and sd_functional.

significance_level

Significance level alpha, defaults to 0.05. If you choose the "HA" method, this value corresponds to the maximum risk of misclassification

Value

An S3 object of class cs_analysis and cs_combined

Categories

Each individual's change can then be categorized into the following groups:

• Recovered, i.e., the individual showed a reliable change in the beneficial direction and changed from the clinical to the functional population

• Improved, i.e., the individual showed a reliable change in the beneficial direction but did not change populations

• Unchanged, i.e., the individual showed no reliable change

• Deteriorated, i.e., the individual showed a reliable change in the disadvantageous direction but did not change populations

• Harmed, i.e., the individual showed a reliable change in the disadvantageous direction and switched from the functional to the clinincal population

Computational details

There are three available cutoff types, namely a, b, and c which can be used to "draw a line" or separate the functional and clinical population on a continuum. a as a cutoff is defined as the mean of the clinical population minus two times the standard deviation (SD) of the clinical population. b is defined as the mean of the functional population plus also two times the SD of the clinical population. This is true for "negative" outcomes, where a lower instrument score is desirable. For "positive" outcomes, where higher scores are beneficial, a is the mean of the clinical population plus 2 \cdot SD of the clinical population and b is mean of the functional population minus 2 \cdot SD of the clinical population. The summary statistics for the clinical population are estimated from the provided data at pre measurement.

c is defined as the midpoint between both populations based on their respective mean and SD. In order to calculate b and c, descriptive statistics for the functional population must be provided.

From the provided data, a region of change is calculated in which an individual change may likely be due to an inherent measurement of the used instrument. This concept is also known as the minimally detectable change (MDC).

Data preparation

The data set must be tidy, which corresponds to a long data frame in general. It must contain a patient identifier which must be unique per patient. Also, a column containing the different measurements and the outcome must be supplied. Each participant-measurement combination must be unique, so for instance, the data must not contain two "After" measurements for the same patient.

Additionally, if the measurement column contains only two values, the first value based on alphabetical, numerical or factor ordering will be used as the pre measurement. For instance, if the column contains the measurements identifiers "pre" and "post" as strings, then "post" will be sorted before "pre" and thus be used as the "pre" measurement. The function will throw a warning but generally you may want to explicitly define the "pre" and "post" measurement with arguments pre and post. In case of more than two measurement identifiers, you have to define pre and post manually since the function does not know what your pre and post intervention measurements are.

If your data is grouped, you can specify the group by referencing the grouping variable (see examples below). The analysis is then run for every group to compare group differences.

See Also

Main clinical signficance functions cs_anchor(), cs_distribution(), cs_percentage(), cs_statistical()

Examples

cs_results <- claus_2020 |>
  cs_combined(
    id,
    time,
    bdi,
    pre = 1,
    post = 4,
    reliability = 0.80
  )

cs_results
summary(cs_results)
plot(cs_results)


# You can choose a different cutoff but must provide summary statistics for the
# functional population
cs_results_c <- claus_2020 |>
  cs_combined(
    id,
    time,
    bdi,
    pre = 1,
    post = 4,
    reliability = 0.80,
    m_functional = 8,
    sd_functional = 8,
    cutoff_type = "c"
  )

cs_results_c
summary(cs_results_c)
plot(cs_results_c)


# You can group the analysis by providing a grouping variable in the data
cs_results_grouped <- claus_2020 |>
  cs_combined(
    id,
    time,
    bdi,
    pre = 1,
    post = 4,
    group = treatment,
    reliability = 0.80,
    m_functional = 8,
    sd_functional = 8,
    cutoff_type = "c"
  )

cs_results_grouped
summary(cs_results_grouped)
plot(cs_results_grouped)