CTNote (version 0.1.0)

weight_positive_visits: Weight Visits in a Subject Use Pattern

Description

Add numeric weights to a use pattern string

Usage

weight_positive_visits(
  use_pattern,
  weights_num = c(`+` = 1, `*` = 0.5, o = 0.22, `-` = 0),
  posPenalty_num = NULL,
  missPenalty_num = NULL,
  scaleMax = 120L,
  scale = TRUE
)

Value

A numeric value: the results of the sum of all the visit weight values for the use pattern

Arguments

use_pattern

A character string showing the daily, by visit, or weekly substance use pattern for a single subject

weights_num

A named numeric vector mapping the symbols in the use pattern to non-negative numeric values. The default values, 1 for positive UDS ("+"), 0.22 for missing UDS ("o"), and 0 for negative UDS ("-"), are the values set in Ling et al. (1976); see the "Details" section for more information. The default value for a mixed weekly result ("*": at least one positive and at least one negative UDS in a single week) is the average of a positive and negative weight (0.5 by default).

posPenalty_num

A numeric vector showing the penalty for having a positive or mixed UDS ("+" or "*") at each week in the use pattern. Defaults to NULL, implying that a positive UDS should have the same weight at any week in the pattern. One other useful option would be to set the weights to increase along the length of the study protocol, so that the penalty for having a positive UDS is grows larger the longer a participant is in treatment. See the Examples for more information.

missPenalty_num

A numeric vector showing the penalty for having a missing UDS ("o") at each week in the use pattern. Defaults to NULL, implying that a missing UDS should have the same weight at any week in the pattern. One other useful option would be to set the weights to decrease along the length of the study protocol, so that the penalty for missing a clinic visit grows smaller the longer a participant has continued in treatment.

scaleMax

Standardize the score to which maximum? Defaults to 120 to match the scoring scale in Ling et al. (1976). See "Details" for more.

scale

Scale the resulting score to a standard maximum (given by the scaleMax argument). This defaults to TRUE; this should not be changed unless you are a developer and you are trying to debug your code.

Details

This function exists to code the treatment outcome defined in Ling et al. (1976): tools:::Rd_expr_doi("10.1001/archpsyc.1976.01770060043007"). This definition requires other CTNote:: functions as well, but this function was written specifically for that definition.

The weights_num argument is the static "penalty" for positive and missing UDS values; this will not change over the protocol weeks. These values are then multiplied by the penalty vectors (posPenalty_num and missPenalty_num).

Examples

  pattern_char <- "++o+*-------+--+-o-o-o+o+"
  
  
  ###  Defaults  ###
  # See how the weights map to the symbols (DO NOT use in practice)
  weight_positive_visits(pattern_char, scale = FALSE)
  
  # Score this use pattern via default settings
  weight_positive_visits(pattern_char)
  
  
  ###  Increase Static Weight of Missing UDS  ###
  # Because the score for a missing UDS from the Ling et al. (1976) paper was
  #   an estimated value from their data, other weights may better represent
  #   modern addiction behavior patterns. For instance, we believe that 
  #   missing UDS values may be worse than a positive UDS in some instances,
  #   because they indicate that the subject is no longer participating in 
  #   treatment at all. We then should change the static weights to reflect
  #   this.
  weight_positive_visits(
    pattern_char,
    weights_num = c(`+` = 0.8, `*` = 0.4, `o` = 1, `-` = 0)
  )
  
  
  ###  Increasing Positive UDS Penalty  ###
  # Score this use pattern using an increasing positive UDS penalty (similar
  #   to that shown in Lint et al. (1976))
  newPosPenal_num <- seq(
    from = 1, to  = 5, 
    length = stringr::str_length(pattern_char)
  )
  weight_positive_visits(pattern_char, posPenalty_num = newPosPenal_num)
  
  
  ###  Variable Missing UDS Penalty  ###
  # Score this use pattern using a step-down missing UDS penalty (based on
  #   the idea that missing values during the treatment induction period are
  #   much worse than missing any other time in the study)
  newMissPenal_num <- rep(1, stringr::str_length(pattern_char))
  newMissPenal_num[1:4] <- 3
  weight_positive_visits(pattern_char, missPenalty_num = newMissPenal_num)
  
  
  ###  Composite Penalties  ###
  # Score this use pattern with both increasing positive UDS and step-down
  #   missing UDS penalties, while adjusting the weights.
  weight_positive_visits(
    pattern_char,
    weights_num = c(`+` = 0.8, `*` = 0.4, `o` = 1, `-` = 0),
    posPenalty_num = newPosPenal_num,
    missPenalty_num = newMissPenal_num
  )