Computer‐based testing and the 12‐item Screener and Opioid Assessment for Patients with Pain‐Revised: A combined approach to improving efficiency

Abstract

Objectives were to combine computer-based stopping rules with the 12-item form of the Screener and Opioid Assessment for Patients with Pain-Revised (SOAPP-R), and to compare this combined procedure with the fulllength (24-item) SOAPP-R, the computer-based stopping rules alone, and the 12-item short form alone. Three datasets were analyzed, including data from 428 subjects from the initial or cross-validation studies of the SOAPP-R; 84 patients from a pain center; and 110 primary care patients.Subjects completed the full-length SOAPP-R and were assessed for aberrant medication-related behaviors. A realdata simulation was conducted to determine the screening characteristics and mean test length of each version of the SOAPP-R. One procedure combining stochastic curtailment with the 12-item short form reduced the mean length by 37% to 42% compared to the short form alone; by 42% to 45% compared to stochastic curtailment alone;and by 68% to 71% compared to the full-length form. The combined procedure had lower mean test lengths, and

Techniques to Shorten a Screening Tool for Emergency Department Patients

“…Notably, the technique of curtailment does not itself provide an indication of which items are most (and least) predictive of the outcome, nor does it allow us to determine the number of items to be administered in advance. Taking the process a step further, it is then possible to administer the shorter static forms on a computer and apply curtailment techniques, reducing the number of questions even more 27. All of this work supports the concept that lengthy screening tools that have been developed for non-ED settings can potentially be repurposed and made more efficient for the frenetic and time-sensitive environment of the ED without a negative effect on the predictive value of the screening tool.…”

A Comparison of Short Forms of the Screener and Opioid Assessment for Patients With Pain – Revised (SOAPP-R)

“…Previous research has been devoted to developing short forms of the SOAPP-R in order to reduce respondent and administrative burden (Black et al, 2018; Finkelman et al, 2018). However, prior to the current study, no research had compared the 8-item SOAPP-R to other short versions of the SOAPP-R, such as the 12-item short form and the curtailed and stochastically curtailed versions of the full-length SOAPP-R and 12-item form.…”

“…Attention focused on comparing different versions of the 8-item SOAPP-R (the screener in its full-length form, with curtailment, and with stochastic curtailment) with one another and with other versions of the SOAPP-R. Indeed, the other versions of the SOAPP-R examined herein had been investigated previously using the three datasets of this study (Finkelman et al, 2018). However, their performance in comparison to versions of the 8-item SOAPP-R had been an open question.…”

“…The probability of a “high risk” classification can be estimated after each stage of testing (i.e., after each item is presented) based on a logistic regression model (see Finkelman, Smits, Kim, and Riley (2012) for details). Both curtailment and stochastic curtailment have been found to substantially reduce the average test length of the full-length SOAPP-R and the 12-item SOAPP-R without unduly compromising sensitivity or specificity (Finkelman et al, 2018).…”

“…The stopping rules that will be investigated in the current research are curtailment and stochastic curtailment , which will be explained thoroughly in the Materials and Methods section. Curtailment and stochastic curtailment can be applied to both the full-length SOAPP-R and the 12-item short form in order to reduce respondent and administrative burden (Finkelman et al, 2018).…”