Development and clinical feasibility study of a brief version of an addiction‐focused phenotyping battery in females receiving buprenorphine for opioid use disorder

Abstract Introduction We aimed to streamline the NIDA Phenotyping Assessment Battery (PhAB), a package of self‐report scales and neurobehavioral tasks used in substance use disorder (SUD) clinical trials, for clinical administration ease. Tailoring the PhAB to shorten administration time for a treatment setting is critical to expanding its acceptability in SUD clinical trials. This study's primary objectives were to develop a brief version of PhAB (PhAB‐B) and assess its operational feasibility and acceptability in a female clinical treatment sample. Methods Assessments of the original PhAB were evaluated along several criteria to identify a subset for the PhAB‐B. Non‐pregnant females (N=55) between ages 18–65, stabilized on buprenorphine for opioid use disorder (OUD) at an outpatient addiction clinic, completed this abbreviated battery remotely or after a provider visit in clinic. Participant satisfaction questions were administered. REDCap recorded the time to complete PhAB‐B measures. Results The PhAB‐B included 11 measures that probed reward, cognition, negative emotionality, interoception, metacognition, and sleep. Participants who completed the PhAB‐B (N =55) were 36.1 ± 8.9 years of age, White (54.5%), Black (34.5%), and non‐Latinx (96.0%). Most participants completed the PhAB‐B remotely (n = 42, 76.4%). Some participants completed it in‐person (n = 13, 23.6%). PhAB‐B mean completion time was 23.0 ± 12.0 min. Participant experiences were positive, and 96% of whom reported that they would participate in the study again. Conclusion Our findings support the clinical feasibility and acceptability of the PhAB‐B among a female opioid use disorder outpatient addiction treatment sample. Future studies should assess the PhAB‐B psychometric properties among broader treatment samples.


INTRODUCTION
Addiction represents an important public health concern as life expectancy in the United States continues to decrease, largely due to increased substance use-related deaths like overdose (Woolf & Schoomaker, 2019). Evidence-based addiction treatments, both pharmacological and behavioral, are effective (Witkiewitz et al., 2019).
However, most people with substance use disorder (SUD) do not receive treatment . Among those who do, treatment attrition and substance use recurrence are common (McLellan et al., 2000). Strategies to expand and personalize SUD treatment options using precision medicine principles are needed.
Precision medicine entails tailoring treatment regimens based on individual characteristics to improve outcomes (Volkow, 2020). SUD treatment provides an opportunity for advances in precision medicine because people with SUD comprise an extremely heterogeneous group in terms of their biological, psychosocial, and neurobehavioral characteristics and related drivers of substance use (Leggio et al., 2009;Litten et al., 2015). For example, according to the Diagnostic and Statistical Manual of Mental Disorders-5, over 2000 different symptom combinations would all meet the criteria for alcohol use disorder (AUD) (APA, 2013;Witkiewitz et al., 2019). Current diagnostic and treatment practices do not usually account for these individual characteristics which limits treatment efficacy and contributes to heterogeneity in SUD treatment response (Huhn et al., 2019;Panlilio et al., 2019;Saraiya et al., 2020).
The utility of deep phenotyping of individuals with SUD to characterize the function of specific neurofunction domains has received attention as a strategy to better address heterogeneity in addiction (Keyser-Marcus et al., 2021;Kwako et al., 2016). The Addictions Neuroclinical Assessment (ANA) battery was developed by the National Institute on Alcohol Abuse and Alcoholism for use in AUD clinical trials to probe negative emotionality, incentive salience, and executive function as domains considered critical to AUD (Kwako et al., 2019). However, a comprehensive ANA battery takes approximately 10 hours to complete across 2 days of administration (Kwako et al., 2016), limiting its feasibility in a clinical trial and addiction treatment setting.
To account for a broader range of neurofunctional domains involved in the addiction process, the National Institute for Drug Abuse (NIDA) subsequently developed the Phenotyping Assessment Battery (PhAB), which expanded upon the three-domain ANA model to also include interoception, metacognition, and sleep domains. The PhAB is a comprehensive package of self-report psychometric scales and neurocognitive behavioral tasks, with an optional resting state fMRI, designed for deployment in any type of SUD clinical trial (Keyser-Marcus et al., 2021). A prior PhAB feasibility study conducted among a sample of individuals with and without opioid, stimulant, and/or cannabis use disorders found the PhAB to be feasible in a research laboratory setting with substantially lower participant burden than the ANA battery (<2hour completion time) (Keyser-Marcus et al., 2021). The PhAB was also rated as highly acceptable to participants. However, for prospective studies to be conducted within outpatient SUD treatment clinics, investigators may find the practicality of even the 2-hour in-person PhAB assessment to be unfeasible. Recruiting SUD treatment patients to complete study assessments can be challenging, especially if the assessments are lengthy, necessitate in-person completion, or require a separate visit to a research testing facility.
Therefore, while the NIDA PhAB is a promising deep phenotyping battery, its acceptance into clinical research practice is slow, driven in part by the impracticality of administering an extensive battery in a fast-paced clinic (Keyser-Marcus et al., 2021). A shortened version of the PhAB to optimize administration in a realistic SUD treatment clinical setting is needed. Additionally, sex is an important variable that modifies SUD trajectories (Huhn et al., 2019). A first step to developing a sex-informed phenotyping tool is to stratify samples by sex when assessing abbreviated PhAB options. Females have been largely underrepresented in phenotyping studies among SUD clinical samples, warranting their priority in these initial investigations (Kwako et al., 2019;Nieto et al., 2021). The current study's objectives were to (1) describe the development of a brief version of the PhAB (PhAB-B), (2) assess the feasibility of the PhAB-B in a treatment setting, and (3) evaluate the acceptability of the PhAB-B among a female outpatient treatment sample receiving buprenorphine for opioid use disorder (OUD).

Development of the PhAB-B
A multidisciplinary workgroup (N = 7) of clinician (CEM and FGM) and  (Kroenke et al., 2001;Measures, 2020;Spitzer et al., 1960). A post-traumatic stress disorder (PTSD) symptom measure (PCL-5 PTSD Checklist) was added given the high prevalence of PTSD and comorbidity with other mental health conditions common in the SUD treatment population (Blevins et al., 2015).

Participants and study design
The pilot study of the PhAB-B battery was a cross-sectional survey nested within a larger study assessing the prevalence and correlates of insomnia among individuals in outpatient OUD treatment with buprenorphine. The parent study intentionally focused on recruiting a convenience sample of female participants, given the known higher burden of insomnia among women than men (Suh et al., 2018). Inclusion criteria were age between 18 and 65 years, female sex, not currently pregnant or within six weeks of pregnancy, having an OUD diagnosis, receiving buprenorphine for OUD for at least 6 weeks, and speaking English.
Potential participants were approached in the outpatient SUD treatment clinic between February and November 2022. All participants provided informed consent. After consent, participants completed a brief research assistant interview (e.g., medical history, current medications, and substance use) and then were offered two options for completion of the parent study's electronic survey: (1) in-person on a tablet provided in the outpatient clinic or (2)

SUD treatment clinical setting
The outpatient SUD treatment clinic is affiliated with a safety net academic medical center in a Medicaid-expanded southern state.
The clinic provides outpatient SUD treatment for over 500 adults yearly, predominately low-income, racially, and ethnically minoritized people, with approximately 90% receiving buprenorphine for OUD.
Most patients are referred from within the academic medical center (e.g., inpatient consults and primary care physicians). A comprehensive, recovery-oriented care model is utilized throughout treatment in which patients have access to medical, psychiatric, mental health, case management, and social work services. The clinic prioritizes a lowthreshold, harm reduction approach, meaning that established patients with substance use recurrence are not excluded from treatment, but instead provided with increased wrap-around support. Additionally, there is no time limitation on how long patients can continue SUD treatment with the clinic. Therefore, the patient population in the clinical study setting represents many stages across the OUD treatment and recovery cascade.

PhAB-B survey measures
Within the reward domain, the 20-item short version of the UPPS-P (SUPPS-P) Impulsive Behavior Scale assessed facets of impulsivity across five subscales (urgency, premeditation, perseverance, sensation seeking, and positive urgency) (Cyders et al., 2014).
Within the cognition domain, the five-trial delay discounting task assessed an individual's discount rate by presenting a series of questions in which the participant chooses between some amount of a delayed commodity or an amount available immediately (Koffarnus & Bickel, 2014). The brief task is based on the premise that individuals tend to value rewards less as the amount of time increases until those rewards would be received.
Within the negative emotionality domain, the nine-item PHQ-9 assessed the frequency of depressive symptoms over the past 2 weeks (Kroenke et al., 2001). One item on the PHQ-9 assesses the frequency of suicidal ideation. If a participant endorsed suicidality (n = 15), the study research assistant informed the principal investigator and contacted the participant within 24 hours to conduct a risk assessment. If the research assistant was unable to reach the participant after three attempts, she contacted the emergency contact listed in the medical record. The seven-item Generalized Anxiety Disorder-7 (GAD-7) assessed the frequency of core symptoms of GAD over the past 2 weeks (Spitzer et al., 1960). The 15-item Distress Tolerance Scale measured emotional distress tolerance across four factors (tolerance, absorption, appraisal, and regulation) (Simons & Gaher, 2005).

Feasibility and tolerability metrics
The primary feasibility outcomes assessed the ease of PhAB-B administration and participation burden in a clinical SUD treatment setting.
Feasibility outcomes were operationalized as individual measure completion times and PhAB-B total completion time, both measured by REDCap, as well as the type of study visit (in-person or remote). Additionally, we assessed representativeness of those who completed the

PhAB-B battery (PhAB-B completers).
Secondarily, the acceptability of the PhAB-B was assessed by participant satisfaction with the overall study (i.e., "I would participate in this study again" and "I would recommend this study to family and friends") upon completion of the PhAB-B measures in the survey.

Additional descriptive measures
Recent non-prescribed drug and alcohol use were assessed in the interview using the Timeline Follow Back, a calendar method used to collect daily alcohol and non-prescribed drug use data over the past 28 days (Sobell & Sobell, 1992). For those who did not complete the interview (n = 9), recent non-alcohol substance use was assessed by a review of recent urine drug test results (within the past 28 days) documented in the medical record. Age, race, and ethnicity were selfreported in the parent study's survey. For participants who consented to the parent study but did not complete any other parent study procedures (n = 9) or did not initiate the parent study survey (n = 27), these variables were abstracted from the medical record. For all participants, insurance status and length of time receiving buprenorphine were abstracted from the medical record.

Development of the PhAB-B
The PhAB-B includes a total of 11 self-report measures across the six addiction-relevant neurofunctional domains in the original PhAB.

Pilot testing the PhAB-B
As shown in the participant flow diagram (Figure 1), N = 95 female patients consented to parent study participation (76.0% recruitment rate). Of those, n = 40 (42.2%) participants were lost to follow-up between study enrollment and initiating the PhAB-B battery (PhAB-B non-completers). All participants who initiated the PhAB-B battery (n = 55, PhAB-B completers) completed the full battery (100% PhAB-B completion rate) indicating the battery was tolerable to all of those who started it during pilot testing. PhAB-B completers and non-completers did not differ on demographic variables or recent substance use (Table 2). PhAB-B completers had been receiving buprenorphine for significantly shorter time than PhAB-B non-completers (p = .009).

DISCUSSION
To advance the use of precision medicine principles in addiction treatment, assessments that identify psychobiologically meaningful individual differences with sufficient brevity to be acceptable to patients and research participants are essential. (Volkow, 2018) The current study sought to describe the development of the PhAB-B as well as assess the feasibility and acceptability of the PhAB-B in a clinical research setting.
Among an outpatient sample of female patients receiving buprenorphine, we found that the PhAB-B was feasible to administer either remotely or in-person in this treatment setting, and it was not considered burdensome to participants. Notably, the sample largely consisted of patients identifying with marginalized groups in a safety-net care system at various stages of the OUD treatment and recovery cascade. Even when research studies are designed to be low-burden, it can still be challenging to recruit and retain participants from an outpatient SUD treatment setting (Fallin-Bennett et al., 2022). We observed a relatively high recruitment rate (76%) and a high PhAB-B completion rate (100%) among those who initiated the battery (higher than the 83% completion rate in the original PhAB feasibility study). How-ever, there was a substantial group of participants who were lost to follow-up between study enrollment and PhAB-B initiation (42%). In  We hypothesize that the primary factors that positively contributed to the feasibility and acceptability of the PhAB-B in this outpatient SUD treatment setting were the ability to complete the entire battery remotely and its brevity. During the COVID-19 pandemic, telehealthdelivered addiction care was adopted by many SUD treatment centers and has persisted as an important mode of healthcare delivery in outpatient SUD treatment (Mark et al., 2022  completion rate, these findings are preliminary and intended to inform future research that can build upon this study by piloting the PhAB-B among broader treatment samples, in terms of SUD type, sex, and gender, and assessing its psychometric properties.

CONCLUSIONS
This study is the first to describe the development and pilot test-

DATA AVAILABILITY STATEMENT
Research data are not shared.