Stacey A McCaffrey1, Ryan A Black1,2, Albert J Villapiano1, Robert N Jamison3, Stephen F Butler1. 1. Inflexxion Inc., Waltham, Massachusetts. 2. Nova Southeastern University, College of Psychology, Fort Lauderdale, Florida. 3. Anesthesia and Psychiatry, Brigham and Women's Hospital Harvard Medical Center, Pain Management Center, Chestnut Hill, Massachusetts, USA.
Abstract
Objective: The Current Opioid Misuse Measure (COMM) is a commonly used self-report instrument to identify and monitor aberrant opioid-related behavior in chronic pain patients on opioid therapy. However, the length of the COMM may limit its clinical utility. Additionally, this paper-and-pencil screener requires hand scoring, which increases paperwork and staff burden. Therefore, the current study presents development of the "COMM-9," a brief electronically administered form of the COMM. Methods: Patients (N = 517) with chronic noncancer pain on opioid therapy completed the COMM. Patients were classified as either being positive or negative for aberrant drug-related behavior based on self-report data from a structured interview, physician-report data, and urine toxicology screen (the Aberrant Drug Behavior Index [ADBI]). COMM items with the strongest classification accuracy were identified using the LASSO method as the selection criterion in conjunction with the leave-one-out cross-validation method as the stop criterion. A sub-set of patients (n=55) completed a second administration of the COMM one week later to evaluate test-retest reliability. Results: Nine items were identified before the selection criterion stopped, and logistic regression was utilized to predict probabilities of positive ADBI from the 9 COMM items using all data and the cross-validation procedure. Receiver operating characteristic curves revealed high levels of classification accuracy that were essentially equivalent to the full COMM. Cut-points were identified to classify patients as having no/low risk, moderate risk, and high risk for aberrant opioid-related behavior. Test-retest reliability of the COMM-9 was comparable to the full 17-item COMM. Conclusions: This study presents the successful development of a brief electronic screener to identify current aberrant opioid-related behavior in chronic pain patients on long-term opioid therapy.
Objective: The Current Opioid Misuse Measure (COMM) is a commonly used self-report instrument to identify and monitor aberrant opioid-related behavior in chronic painpatients on opioid therapy. However, the length of the COMM may limit its clinical utility. Additionally, this paper-and-pencil screener requires hand scoring, which increases paperwork and staff burden. Therefore, the current study presents development of the "COMM-9," a brief electronically administered form of the COMM. Methods:Patients (N = 517) with chronic noncancer pain on opioid therapy completed the COMM. Patients were classified as either being positive or negative for aberrant drug-related behavior based on self-report data from a structured interview, physician-report data, and urine toxicology screen (the Aberrant Drug Behavior Index [ADBI]). COMM items with the strongest classification accuracy were identified using the LASSO method as the selection criterion in conjunction with the leave-one-out cross-validation method as the stop criterion. A sub-set of patients (n=55) completed a second administration of the COMM one week later to evaluate test-retest reliability. Results: Nine items were identified before the selection criterion stopped, and logistic regression was utilized to predict probabilities of positive ADBI from the 9 COMM items using all data and the cross-validation procedure. Receiver operating characteristic curves revealed high levels of classification accuracy that were essentially equivalent to the full COMM. Cut-points were identified to classify patients as having no/low risk, moderate risk, and high risk for aberrant opioid-related behavior. Test-retest reliability of the COMM-9 was comparable to the full 17-item COMM. Conclusions: This study presents the successful development of a brief electronic screener to identify current aberrant opioid-related behavior in chronic painpatients on long-term opioid therapy.
Authors: Amanda M Brandow; C Patrick Carroll; Susan Creary; Ronisha Edwards-Elliott; Jeffrey Glassberg; Robert W Hurley; Abdullah Kutlar; Mohamed Seisa; Jennifer Stinson; John J Strouse; Fouza Yusuf; William Zempsky; Eddy Lang Journal: Blood Adv Date: 2020-06-23