Patient UDS Opioid Weekly Pattern Data

Show the pattern of positive, negative, and missing urine drug screen (UDS) results for opioids by patient over the study protocol. Study "Week 1" starts the day after randomization (for patients who were randomized) or the day after signed consent (for patients who were not randomized).

Usage

data(derived_weeklyOpioidPattern)

Format

A tibble with 3,560 rows and columns:

who

Patient ID

startWeek

The start of the "word" is how many weeks before randomization? This should be -4 for most people, but can be as high as -8. Note that week 0 is included, so a value of -4 represents data in the 5th week before randomization; that is, 29-35 days prior to randomization. Most subjects have timeline follow-back data 30 days before consent, and delays between consent and randomization are common.

randWeek1

Week of first randomization (1, if randomized; NA if not)

randWeek2

Week of second randomization (only for CTN-0030)

endWeek

The end of the "word" is how many weeks after randomization? This depends on the study protocol, but should be close to 16 or 24 weeks for most subjects.

Baseline

A character string of symbols from startWeek to the last week before randWeek1. Symbols are as defined in Phase_1.

Phase_1

A character string of symbols from randWeek1 to endWeek (for subjects from CTN-0027 and CTN-0051) or the last week before randWeek2 (for CTN-0030). These symbols are: "+" = the subject's UDS showed presence of an opioid in that week; "-" = the subject's UDS showed absence of an opioid in that week; "*" = two or more UDS in the same week, where >= 1 UDS was positive and >= 1 UDS was negative; "o" if the subject was supposed to visit a clinic per the study protocol but did not visit or did not complete a UDS screen; and "_" to represent weeks wherein the subject was not scheduled to provide UDS.

Phase_2

A character string of symbols from randWeek2 to endWeek for subjects from CTN-0030 only. Symbols are as defined in Phase_1.

Details

This data set contains a "word" describing weekly non-study opioid use patterns as measured by UDS. Based on the substances screened in this data set, our list of substances classified as an opioid is: Oxymorphone, Opium, Fentanyl, Hydromorphone, Codeine, Suboxone, Tramadol, Morphine, Buprenorphine, Hydrocodone, Opioid, Methadone, Oxycodone, and Heroin. UDS results indicating the presence of one or more of these substances will be marked with "+" for that week. UDS results negative for these substances will be marked with "-". If, by study protocol, the subject was supposed to visit a clinic to complete a UDS but they did not, then the visit for that week will be marked with "o". If, by study protocol, the subject was NOT supposed to visit a clinic to complete a UDS for a given week, then visit for that week will be marked with "_". This data set is a derived data set; the script used to create it is "scripts/create_weeklyOpioidPattern_20211123.R".

NOTE: because our window is a strict weekly window, a subject could have both positive and negative UDS within the same 7-day period. In this case, the week is marked as "*". Depending on the definition of treatment failure or treatment success desired, these dual-status indicators can be re-coded to "+" or "-" as appropriate. Also, some studies include a baseline UDS in the week of consent (before the subject was randomized to a treatment arm). Some subjects were randomized in the same week as the week wherein consent was signed, while other subjects were randomized weeks later. We represent the weeks before consent and the variable number of weeks between consent and randomization with "_" if there were no baseline UDS visits in those weeks (this represents the pre-study period). For subjects who were never randomized, the weeks before consent are also marked as pre-study ("_"), and all subsequent protocol weeks are marked as missing.