The Screener and Opioid Assessment for Patients with Pain - Revised (SOAPP-R) is a 24-item questionnaire that has been validated as a measure of risk of opioid abuse. While previous studies have compared the SOAPP-R favorably to other assessments in terms of sensitivity and specificity, the screener is currently only available in paper-and-pencil format. Moreover, it may be too long for certain respondents, particularly compromised individuals such as people in poor health and people who have difficulty with reading comprehension. Shorter versions of the SOAPP-R that maintain adequate sensitivity and specificity are therefore needed. The proposed research represents a fundamental first step toward filling this gap by providing a "proof of principle" for shorter computer-based versions of the instrument. These computer-based versions utilize interim analyses (calculations that are performed while assessment is in progress) to customize the questionnaire at the level of the individual respondent. The customizable procedures to be studied include curtailment, stochastic curtailment, and decision trees, all of which have been successfully applied to assessments in other domains. Additionally, static (non-customizable) short forms of the SOAPP-R will be developed that can be used by practitioners who do not have the capability to implement computer-based questionnaires. The first goal of the proposed research is to develop a computer program that performs and evaluates all procedures (curtailment, stochastic curtailment, decision trees, short forms, and the full-length SOAPP-R) on a given dataset. Evaluation of procedures will include statistics related to classification accuracy (e.g., sensitivity and specificity) as well as statistics related to respondent burden (e.., mean and standard deviation of the number of items administered). The second goal of the proposed research is to compare the aforementioned procedures using a dataset of 428 individuals who took the full-length SOAPP-R and were followed up to determine whether they had engaged in aberrant medication-related behavior. The follow-up determination was made via the Aberrant Drug Behavior Index (ADBI), a measure that incorporates self-report via structured interview, physician report, and urine toxicology results. Based on previous research in other domains, it is hypothesized that the customizable computer-based versions of the SOAPP-R will exhibit greater measurement efficiency than both the short forms and the full-length instrument. If this hypothesis is borne out in the data, the proposed research will constitute an important advance in helping clinicians make a decision to prescribe or not prescribe opioids.
The proposed research will provide a proof of principle for computer-based versions of the SOAPP-R, a screener that predicts risk of opioid abuse. The computer-based versions use information from a respondent's answers to customize the screener to the individual taking it. Computer-based instruments have the potential to reduce respondent burden, thereby making the screener more accessible to compromised populations and aiding clinicians in their decision of whether to prescribe opioids.